summaryrefslogtreecommitdiff
path: root/lib/dns/dispatch.c
blob: 1d65ce43f4c157f900c1fedb3c14c1881152e973 (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
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
3485
3486
3487
3488
3489
3490
3491
3492
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
3514
3515
3516
3517
3518
3519
3520
3521
3522
3523
3524
3525
3526
3527
3528
3529
/*
 * Copyright (C) 2004-2009  Internet Systems Consortium, Inc. ("ISC")
 * Copyright (C) 1999-2003  Internet Software Consortium.
 *
 * Permission to use, copy, modify, and/or distribute this software for any
 * purpose with or without fee is hereby granted, provided that the above
 * copyright notice and this permission notice appear in all copies.
 *
 * THE SOFTWARE IS PROVIDED "AS IS" AND ISC DISCLAIMS ALL WARRANTIES WITH
 * REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY
 * AND FITNESS.  IN NO EVENT SHALL ISC BE LIABLE FOR ANY SPECIAL, DIRECT,
 * INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM
 * LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE
 * OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
 * PERFORMANCE OF THIS SOFTWARE.
 */

/* $Id: dispatch.c,v 1.168 2009/12/02 23:15:14 marka Exp $ */

/*! \file */

#include <config.h>

#include <stdlib.h>
#include <sys/types.h>
#include <unistd.h>
#include <stdlib.h>

#include <isc/entropy.h>
#include <isc/mem.h>
#include <isc/mutex.h>
#include <isc/portset.h>
#include <isc/print.h>
#include <isc/random.h>
#include <isc/stats.h>
#include <isc/string.h>
#include <isc/task.h>
#include <isc/time.h>
#include <isc/util.h>

#include <dns/acl.h>
#include <dns/dispatch.h>
#include <dns/events.h>
#include <dns/log.h>
#include <dns/message.h>
#include <dns/portlist.h>
#include <dns/stats.h>
#include <dns/tcpmsg.h>
#include <dns/types.h>

typedef ISC_LIST(dns_dispentry_t)	dns_displist_t;

typedef struct dispsocket		dispsocket_t;
typedef ISC_LIST(dispsocket_t)		dispsocketlist_t;

typedef struct dispportentry		dispportentry_t;
typedef ISC_LIST(dispportentry_t)	dispportlist_t;

/* ARC4 Random generator state */
typedef struct arc4ctx {
	isc_uint8_t	i;
	isc_uint8_t	j;
	isc_uint8_t	s[256];
	int		count;
	isc_entropy_t	*entropy;	/*%< entropy source for ARC4 */
	isc_mutex_t	*lock;
} arc4ctx_t;

typedef struct dns_qid {
	unsigned int	magic;
	unsigned int	qid_nbuckets;	/*%< hash table size */
	unsigned int	qid_increment;	/*%< id increment on collision */
	isc_mutex_t	lock;
	dns_displist_t	*qid_table;	/*%< the table itself */
	dispsocketlist_t *sock_table;	/*%< socket table */
} dns_qid_t;

struct dns_dispatchmgr {
	/* Unlocked. */
	unsigned int			magic;
	isc_mem_t		       *mctx;
	dns_acl_t		       *blackhole;
	dns_portlist_t		       *portlist;
	isc_stats_t		       *stats;
	isc_entropy_t		       *entropy; /*%< entropy source */

	/* Locked by "lock". */
	isc_mutex_t			lock;
	unsigned int			state;
	ISC_LIST(dns_dispatch_t)	list;

	/* Locked by arc4_lock. */
	isc_mutex_t			arc4_lock;
	arc4ctx_t			arc4ctx;    /*%< ARC4 context for QID */

	/* locked by buffer lock */
	dns_qid_t			*qid;
	isc_mutex_t			buffer_lock;
	unsigned int			buffers;    /*%< allocated buffers */
	unsigned int			buffersize; /*%< size of each buffer */
	unsigned int			maxbuffers; /*%< max buffers */

	/* Locked internally. */
	isc_mutex_t			pool_lock;
	isc_mempool_t		       *epool;	/*%< memory pool for events */
	isc_mempool_t		       *rpool;	/*%< memory pool for replies */
	isc_mempool_t		       *dpool;  /*%< dispatch allocations */
	isc_mempool_t		       *bpool;	/*%< memory pool for buffers */
	isc_mempool_t		       *spool;	/*%< memory pool for dispsocs */

	/*%
	 * Locked by qid->lock if qid exists; otherwise, can be used without
	 * being locked.
	 * Memory footprint considerations: this is a simple implementation of
	 * available ports, i.e., an ordered array of the actual port numbers.
	 * This will require about 256KB of memory in the worst case (128KB for
	 * each of IPv4 and IPv6).  We could reduce it by representing it as a
	 * more sophisticated way such as a list (or array) of ranges that are
	 * searched to identify a specific port.  Our decision here is the saved
	 * memory isn't worth the implementation complexity, considering the
	 * fact that the whole BIND9 process (which is mainly named) already
	 * requires a pretty large memory footprint.  We may, however, have to
	 * revisit the decision when we want to use it as a separate module for
	 * an environment where memory requirement is severer.
	 */
	in_port_t	*v4ports;	/*%< available ports for IPv4 */
	unsigned int	nv4ports;	/*%< # of available ports for IPv4 */
	in_port_t	*v6ports;	/*%< available ports for IPv4 */
	unsigned int	nv6ports;	/*%< # of available ports for IPv4 */
};

#define MGR_SHUTTINGDOWN		0x00000001U
#define MGR_IS_SHUTTINGDOWN(l)	(((l)->state & MGR_SHUTTINGDOWN) != 0)

#define IS_PRIVATE(d)	(((d)->attributes & DNS_DISPATCHATTR_PRIVATE) != 0)

struct dns_dispentry {
	unsigned int			magic;
	dns_dispatch_t		       *disp;
	dns_messageid_t			id;
	in_port_t			port;
	unsigned int			bucket;
	isc_sockaddr_t			host;
	isc_task_t		       *task;
	isc_taskaction_t		action;
	void			       *arg;
	isc_boolean_t			item_out;
	dispsocket_t			*dispsocket;
	ISC_LIST(dns_dispatchevent_t)	items;
	ISC_LINK(dns_dispentry_t)	link;
};

/*%
 * Maximum number of dispatch sockets that can be pooled for reuse.  The
 * appropriate value may vary, but experiments have shown a busy caching server
 * may need more than 1000 sockets concurrently opened.  The maximum allowable
 * number of dispatch sockets (per manager) will be set to the double of this
 * value.
 */
#ifndef DNS_DISPATCH_POOLSOCKS
#define DNS_DISPATCH_POOLSOCKS			2048
#endif

/*%
 * Quota to control the number of dispatch sockets.  If a dispatch has more
 * than the quota of sockets, new queries will purge oldest ones, so that
 * a massive number of outstanding queries won't prevent subsequent queries
 * (especially if the older ones take longer time and result in timeout).
 */
#ifndef DNS_DISPATCH_SOCKSQUOTA
#define DNS_DISPATCH_SOCKSQUOTA			3072
#endif

struct dispsocket {
	unsigned int			magic;
	isc_socket_t			*socket;
	dns_dispatch_t			*disp;
	isc_sockaddr_t			host;
	in_port_t			localport; /* XXX: should be removed later */
	dispportentry_t			*portentry;
	dns_dispentry_t			*resp;
	isc_task_t			*task;
	ISC_LINK(dispsocket_t)		link;
	unsigned int			bucket;
	ISC_LINK(dispsocket_t)		blink;
};

/*%
 * A port table entry.  We remember every port we first open in a table with a
 * reference counter so that we can 'reuse' the same port (with different
 * destination addresses) using the SO_REUSEADDR socket option.
 */
struct dispportentry {
	in_port_t			port;
	unsigned int			refs;
	ISC_LINK(struct dispportentry)	link;
};

#ifndef DNS_DISPATCH_PORTTABLESIZE
#define DNS_DISPATCH_PORTTABLESIZE	1024
#endif

#define INVALID_BUCKET		(0xffffdead)

/*%
 * Number of tasks for each dispatch that use separate sockets for different
 * transactions.  This must be a power of 2 as it will divide 32 bit numbers
 * to get an uniformly random tasks selection.  See get_dispsocket().
 */
#define MAX_INTERNAL_TASKS	64

struct dns_dispatch {
	/* Unlocked. */
	unsigned int		magic;		/*%< magic */
	dns_dispatchmgr_t      *mgr;		/*%< dispatch manager */
	int			ntasks;
	/*%
	 * internal task buckets.  We use multiple tasks to distribute various
	 * socket events well when using separate dispatch sockets.  We use the
	 * 1st task (task[0]) for internal control events.
	 */
	isc_task_t	       *task[MAX_INTERNAL_TASKS];
	isc_socket_t	       *socket;		/*%< isc socket attached to */
	isc_sockaddr_t		local;		/*%< local address */
	in_port_t		localport;	/*%< local UDP port */
	unsigned int		maxrequests;	/*%< max requests */
	isc_event_t	       *ctlevent;

	/*% Locked by mgr->lock. */
	ISC_LINK(dns_dispatch_t) link;

	/* Locked by "lock". */
	isc_mutex_t		lock;		/*%< locks all below */
	isc_sockettype_t	socktype;
	unsigned int		attributes;
	unsigned int		refcount;	/*%< number of users */
	dns_dispatchevent_t    *failsafe_ev;	/*%< failsafe cancel event */
	unsigned int		shutting_down : 1,
				shutdown_out : 1,
				connected : 1,
				tcpmsg_valid : 1,
				recv_pending : 1; /*%< is a recv() pending? */
	isc_result_t		shutdown_why;
	ISC_LIST(dispsocket_t)	activesockets;
	ISC_LIST(dispsocket_t)	inactivesockets;
	unsigned int		nsockets;
	unsigned int		requests;	/*%< how many requests we have */
	unsigned int		tcpbuffers;	/*%< allocated buffers */
	dns_tcpmsg_t		tcpmsg;		/*%< for tcp streams */
	dns_qid_t		*qid;
	arc4ctx_t		arc4ctx;	/*%< for QID/UDP port num */
	dispportlist_t		*port_table;	/*%< hold ports 'owned' by us */
	isc_mempool_t		*portpool;	/*%< port table entries  */
};

#define QID_MAGIC		ISC_MAGIC('Q', 'i', 'd', ' ')
#define VALID_QID(e)		ISC_MAGIC_VALID((e), QID_MAGIC)

#define RESPONSE_MAGIC		ISC_MAGIC('D', 'r', 's', 'p')
#define VALID_RESPONSE(e)	ISC_MAGIC_VALID((e), RESPONSE_MAGIC)

#define DISPSOCK_MAGIC		ISC_MAGIC('D', 's', 'o', 'c')
#define VALID_DISPSOCK(e)	ISC_MAGIC_VALID((e), DISPSOCK_MAGIC)

#define DISPATCH_MAGIC		ISC_MAGIC('D', 'i', 's', 'p')
#define VALID_DISPATCH(e)	ISC_MAGIC_VALID((e), DISPATCH_MAGIC)

#define DNS_DISPATCHMGR_MAGIC	ISC_MAGIC('D', 'M', 'g', 'r')
#define VALID_DISPATCHMGR(e)	ISC_MAGIC_VALID((e), DNS_DISPATCHMGR_MAGIC)

#define DNS_QID(disp) ((disp)->socktype == isc_sockettype_tcp) ? \
		       (disp)->qid : (disp)->mgr->qid
#define DISP_ARC4CTX(disp) ((disp)->socktype == isc_sockettype_udp) ? \
			(&(disp)->arc4ctx) : (&(disp)->mgr->arc4ctx)

/*%
 * Locking a query port buffer is a bit tricky.  We access the buffer without
 * locking until qid is created.  Technically, there is a possibility of race
 * between the creation of qid and access to the port buffer; in practice,
 * however, this should be safe because qid isn't created until the first
 * dispatch is created and there should be no contending situation until then.
 */
#define PORTBUFLOCK(mgr) if ((mgr)->qid != NULL) LOCK(&((mgr)->qid->lock))
#define PORTBUFUNLOCK(mgr) if ((mgr)->qid != NULL) UNLOCK((&(mgr)->qid->lock))

/*
 * Statics.
 */
static dns_dispentry_t *entry_search(dns_qid_t *, isc_sockaddr_t *,
				     dns_messageid_t, in_port_t, unsigned int);
static isc_boolean_t destroy_disp_ok(dns_dispatch_t *);
static void destroy_disp(isc_task_t *task, isc_event_t *event);
static void destroy_dispsocket(dns_dispatch_t *, dispsocket_t **);
static void deactivate_dispsocket(dns_dispatch_t *, dispsocket_t *);
static void udp_exrecv(isc_task_t *, isc_event_t *);
static void udp_shrecv(isc_task_t *, isc_event_t *);
static void udp_recv(isc_event_t *, dns_dispatch_t *, dispsocket_t *);
static void tcp_recv(isc_task_t *, isc_event_t *);
static isc_result_t startrecv(dns_dispatch_t *, dispsocket_t *);
static isc_uint32_t dns_hash(dns_qid_t *, isc_sockaddr_t *, dns_messageid_t,
			     in_port_t);
static void free_buffer(dns_dispatch_t *disp, void *buf, unsigned int len);
static void *allocate_udp_buffer(dns_dispatch_t *disp);
static inline void free_event(dns_dispatch_t *disp, dns_dispatchevent_t *ev);
static inline dns_dispatchevent_t *allocate_event(dns_dispatch_t *disp);
static void do_cancel(dns_dispatch_t *disp);
static dns_dispentry_t *linear_first(dns_qid_t *disp);
static dns_dispentry_t *linear_next(dns_qid_t *disp,
				    dns_dispentry_t *resp);
static void dispatch_free(dns_dispatch_t **dispp);
static isc_result_t get_udpsocket(dns_dispatchmgr_t *mgr,
				  dns_dispatch_t *disp,
				  isc_socketmgr_t *sockmgr,
				  isc_sockaddr_t *localaddr,
				  isc_socket_t **sockp);
static isc_result_t dispatch_createudp(dns_dispatchmgr_t *mgr,
				       isc_socketmgr_t *sockmgr,
				       isc_taskmgr_t *taskmgr,
				       isc_sockaddr_t *localaddr,
				       unsigned int maxrequests,
				       unsigned int attributes,
				       dns_dispatch_t **dispp);
static isc_boolean_t destroy_mgr_ok(dns_dispatchmgr_t *mgr);
static void destroy_mgr(dns_dispatchmgr_t **mgrp);
static isc_result_t qid_allocate(dns_dispatchmgr_t *mgr, unsigned int buckets,
				 unsigned int increment, dns_qid_t **qidp,
				 isc_boolean_t needaddrtable);
static void qid_destroy(isc_mem_t *mctx, dns_qid_t **qidp);
static isc_result_t open_socket(isc_socketmgr_t *mgr, isc_sockaddr_t *local,
				unsigned int options, isc_socket_t **sockp);
static isc_boolean_t portavailable(dns_dispatchmgr_t *mgr, isc_socket_t *sock,
				   isc_sockaddr_t *sockaddrp);

#define LVL(x) ISC_LOG_DEBUG(x)

static void
mgr_log(dns_dispatchmgr_t *mgr, int level, const char *fmt, ...)
     ISC_FORMAT_PRINTF(3, 4);

static void
mgr_log(dns_dispatchmgr_t *mgr, int level, const char *fmt, ...) {
	char msgbuf[2048];
	va_list ap;

	if (! isc_log_wouldlog(dns_lctx, level))
		return;

	va_start(ap, fmt);
	vsnprintf(msgbuf, sizeof(msgbuf), fmt, ap);
	va_end(ap);

	isc_log_write(dns_lctx,
		      DNS_LOGCATEGORY_DISPATCH, DNS_LOGMODULE_DISPATCH,
		      level, "dispatchmgr %p: %s", mgr, msgbuf);
}

static inline void
inc_stats(dns_dispatchmgr_t *mgr, isc_statscounter_t counter) {
	if (mgr->stats != NULL)
		isc_stats_increment(mgr->stats, counter);
}

static void
dispatch_log(dns_dispatch_t *disp, int level, const char *fmt, ...)
     ISC_FORMAT_PRINTF(3, 4);

static void
dispatch_log(dns_dispatch_t *disp, int level, const char *fmt, ...) {
	char msgbuf[2048];
	va_list ap;

	if (! isc_log_wouldlog(dns_lctx, level))
		return;

	va_start(ap, fmt);
	vsnprintf(msgbuf, sizeof(msgbuf), fmt, ap);
	va_end(ap);

	isc_log_write(dns_lctx,
		      DNS_LOGCATEGORY_DISPATCH, DNS_LOGMODULE_DISPATCH,
		      level, "dispatch %p: %s", disp, msgbuf);
}

static void
request_log(dns_dispatch_t *disp, dns_dispentry_t *resp,
	    int level, const char *fmt, ...)
     ISC_FORMAT_PRINTF(4, 5);

static void
request_log(dns_dispatch_t *disp, dns_dispentry_t *resp,
	    int level, const char *fmt, ...)
{
	char msgbuf[2048];
	char peerbuf[256];
	va_list ap;

	if (! isc_log_wouldlog(dns_lctx, level))
		return;

	va_start(ap, fmt);
	vsnprintf(msgbuf, sizeof(msgbuf), fmt, ap);
	va_end(ap);

	if (VALID_RESPONSE(resp)) {
		isc_sockaddr_format(&resp->host, peerbuf, sizeof(peerbuf));
		isc_log_write(dns_lctx, DNS_LOGCATEGORY_DISPATCH,
			      DNS_LOGMODULE_DISPATCH, level,
			      "dispatch %p response %p %s: %s", disp, resp,
			      peerbuf, msgbuf);
	} else {
		isc_log_write(dns_lctx, DNS_LOGCATEGORY_DISPATCH,
			      DNS_LOGMODULE_DISPATCH, level,
			      "dispatch %p req/resp %p: %s", disp, resp,
			      msgbuf);
	}
}

/*%
 * ARC4 random number generator derived from OpenBSD.
 * Only dispatch_random() and dispatch_uniformrandom() are expected
 * to be called from general dispatch routines; the rest of them are subroutines
 * for these two.
 *
 * The original copyright follows:
 * Copyright (c) 1996, David Mazieres <dm@uun.org>
 * Copyright (c) 2008, Damien Miller <djm@openbsd.org>
 *
 * Permission to use, copy, modify, and distribute this software for any
 * purpose with or without fee is hereby granted, provided that the above
 * copyright notice and this permission notice appear in all copies.
 *
 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
 */
#ifdef BIND9
static void
dispatch_initrandom(arc4ctx_t *actx, isc_entropy_t *entropy,
		    isc_mutex_t *lock)
{
	int n;
	for (n = 0; n < 256; n++)
		actx->s[n] = n;
	actx->i = 0;
	actx->j = 0;
	actx->count = 0;
	actx->entropy = entropy; /* don't have to attach */
	actx->lock = lock;
}

static void
dispatch_arc4addrandom(arc4ctx_t *actx, unsigned char *dat, int datlen) {
	int n;
	isc_uint8_t si;

	actx->i--;
	for (n = 0; n < 256; n++) {
		actx->i = (actx->i + 1);
		si = actx->s[actx->i];
		actx->j = (actx->j + si + dat[n % datlen]);
		actx->s[actx->i] = actx->s[actx->j];
		actx->s[actx->j] = si;
	}
	actx->j = actx->i;
}

static inline isc_uint8_t
dispatch_arc4get8(arc4ctx_t *actx) {
	isc_uint8_t si, sj;

	actx->i = (actx->i + 1);
	si = actx->s[actx->i];
	actx->j = (actx->j + si);
	sj = actx->s[actx->j];
	actx->s[actx->i] = sj;
	actx->s[actx->j] = si;

	return (actx->s[(si + sj) & 0xff]);
}

static inline isc_uint16_t
dispatch_arc4get16(arc4ctx_t *actx) {
	isc_uint16_t val;

	val = dispatch_arc4get8(actx) << 8;
	val |= dispatch_arc4get8(actx);

	return (val);
}

static void
dispatch_arc4stir(arc4ctx_t *actx) {
	int i;
	union {
		unsigned char rnd[128];
		isc_uint32_t rnd32[32];
	} rnd;
	isc_result_t result;

	if (actx->entropy != NULL) {
		/*
		 * We accept any quality of random data to avoid blocking.
		 */
		result = isc_entropy_getdata(actx->entropy, rnd.rnd,
					     sizeof(rnd), NULL, 0);
		RUNTIME_CHECK(result == ISC_R_SUCCESS);
	} else {
		for (i = 0; i < 32; i++)
			isc_random_get(&rnd.rnd32[i]);
	}
	dispatch_arc4addrandom(actx, rnd.rnd, sizeof(rnd.rnd));

	/*
	 * Discard early keystream, as per recommendations in:
	 * http://www.wisdom.weizmann.ac.il/~itsik/RC4/Papers/Rc4_ksa.ps
	 */
	for (i = 0; i < 256; i++)
		(void)dispatch_arc4get8(actx);

	/*
	 * Derived from OpenBSD's implementation.  The rationale is not clear,
	 * but should be conservative enough in safety, and reasonably large
	 * for efficiency.
	 */
	actx->count = 1600000;
}

static isc_uint16_t
dispatch_random(arc4ctx_t *actx) {
	isc_uint16_t result;

	if (actx->lock != NULL)
		LOCK(actx->lock);

	actx->count -= sizeof(isc_uint16_t);
	if (actx->count <= 0)
		dispatch_arc4stir(actx);
	result = dispatch_arc4get16(actx);

	if (actx->lock != NULL)
		UNLOCK(actx->lock);

	return (result);
}
#else
/*
 * For general purpose library, we don't have to be too strict about the
 * quality of random values.  Performance doesn't matter much, either.
 * So we simply use the isc_random module to keep the library as small as
 * possible.
 */

static void
dispatch_initrandom(arc4ctx_t *actx, isc_entropy_t *entropy,
		    isc_mutex_t *lock)
{
	UNUSED(actx);
	UNUSED(entropy);
	UNUSED(lock);

	return;
}

static isc_uint16_t
dispatch_random(arc4ctx_t *actx) {
	isc_uint32_t r;

	UNUSED(actx);

	isc_random_get(&r);
	return (r & 0xffff);
}
#endif	/* BIND9 */

static isc_uint16_t
dispatch_uniformrandom(arc4ctx_t *actx, isc_uint16_t upper_bound) {
	isc_uint16_t min, r;

	if (upper_bound < 2)
		return (0);

	/*
	 * Ensure the range of random numbers [min, 0xffff] be a multiple of
	 * upper_bound and contain at least a half of the 16 bit range.
	 */

	if (upper_bound > 0x8000)
		min = 1 + ~upper_bound; /* 0x8000 - upper_bound */
	else
		min = (isc_uint16_t)(0x10000 % (isc_uint32_t)upper_bound);

	/*
	 * This could theoretically loop forever but each retry has
	 * p > 0.5 (worst case, usually far better) of selecting a
	 * number inside the range we need, so it should rarely need
	 * to re-roll.
	 */
	for (;;) {
		r = dispatch_random(actx);
		if (r >= min)
			break;
	}

	return (r % upper_bound);
}

/*
 * Return a hash of the destination and message id.
 */
static isc_uint32_t
dns_hash(dns_qid_t *qid, isc_sockaddr_t *dest, dns_messageid_t id,
	 in_port_t port)
{
	unsigned int ret;

	ret = isc_sockaddr_hash(dest, ISC_TRUE);
	ret ^= (id << 16) | port;
	ret %= qid->qid_nbuckets;

	INSIST(ret < qid->qid_nbuckets);

	return (ret);
}

/*
 * Find the first entry in 'qid'.  Returns NULL if there are no entries.
 */
static dns_dispentry_t *
linear_first(dns_qid_t *qid) {
	dns_dispentry_t *ret;
	unsigned int bucket;

	bucket = 0;

	while (bucket < qid->qid_nbuckets) {
		ret = ISC_LIST_HEAD(qid->qid_table[bucket]);
		if (ret != NULL)
			return (ret);
		bucket++;
	}

	return (NULL);
}

/*
 * Find the next entry after 'resp' in 'qid'.  Return NULL if there are
 * no more entries.
 */
static dns_dispentry_t *
linear_next(dns_qid_t *qid, dns_dispentry_t *resp) {
	dns_dispentry_t *ret;
	unsigned int bucket;

	ret = ISC_LIST_NEXT(resp, link);
	if (ret != NULL)
		return (ret);

	bucket = resp->bucket;
	bucket++;
	while (bucket < qid->qid_nbuckets) {
		ret = ISC_LIST_HEAD(qid->qid_table[bucket]);
		if (ret != NULL)
			return (ret);
		bucket++;
	}

	return (NULL);
}

/*
 * The dispatch must be locked.
 */
static isc_boolean_t
destroy_disp_ok(dns_dispatch_t *disp)
{
	if (disp->refcount != 0)
		return (ISC_FALSE);

	if (disp->recv_pending != 0)
		return (ISC_FALSE);

	if (!ISC_LIST_EMPTY(disp->activesockets))
		return (ISC_FALSE);

	if (disp->shutting_down == 0)
		return (ISC_FALSE);

	return (ISC_TRUE);
}

/*
 * Called when refcount reaches 0 (and safe to destroy).
 *
 * The dispatcher must not be locked.
 * The manager must be locked.
 */
static void
destroy_disp(isc_task_t *task, isc_event_t *event) {
	dns_dispatch_t *disp;
	dns_dispatchmgr_t *mgr;
	isc_boolean_t killmgr;
	dispsocket_t *dispsocket;
	int i;

	INSIST(event->ev_type == DNS_EVENT_DISPATCHCONTROL);

	UNUSED(task);

	disp = event->ev_arg;
	mgr = disp->mgr;

	LOCK(&mgr->lock);
	ISC_LIST_UNLINK(mgr->list, disp, link);

	dispatch_log(disp, LVL(90),
		     "shutting down; detaching from sock %p, task %p",
		     disp->socket, disp->task[0]); /* XXXX */

	if (disp->socket != NULL)
		isc_socket_detach(&disp->socket);
	while ((dispsocket = ISC_LIST_HEAD(disp->inactivesockets)) != NULL) {
		ISC_LIST_UNLINK(disp->inactivesockets, dispsocket, link);
		destroy_dispsocket(disp, &dispsocket);
	}
	for (i = 0; i < disp->ntasks; i++)
		isc_task_detach(&disp->task[i]);
	isc_event_free(&event);

	dispatch_free(&disp);

	killmgr = destroy_mgr_ok(mgr);
	UNLOCK(&mgr->lock);
	if (killmgr)
		destroy_mgr(&mgr);
}

/*%
 * Manipulate port table per dispatch: find an entry for a given port number,
 * create a new entry, and decrement a given entry with possible clean-up.
 */
static dispportentry_t *
port_search(dns_dispatch_t *disp, in_port_t port) {
	dispportentry_t *portentry;

	REQUIRE(disp->port_table != NULL);

	portentry = ISC_LIST_HEAD(disp->port_table[port %
						   DNS_DISPATCH_PORTTABLESIZE]);
	while (portentry != NULL) {
		if (portentry->port == port)
			return (portentry);
		portentry = ISC_LIST_NEXT(portentry, link);
	}

	return (NULL);
}

static dispportentry_t *
new_portentry(dns_dispatch_t *disp, in_port_t port) {
	dispportentry_t *portentry;

	REQUIRE(disp->port_table != NULL);

	portentry = isc_mempool_get(disp->portpool);
	if (portentry == NULL)
		return (portentry);

	portentry->port = port;
	portentry->refs = 0;
	ISC_LINK_INIT(portentry, link);
	ISC_LIST_APPEND(disp->port_table[port % DNS_DISPATCH_PORTTABLESIZE],
			portentry, link);

	return (portentry);
}

/*%
 * The caller must not hold the qid->lock.
 */
static void
deref_portentry(dns_dispatch_t *disp, dispportentry_t **portentryp) {
	dispportentry_t *portentry = *portentryp;
	dns_qid_t *qid;

	REQUIRE(disp->port_table != NULL);
	REQUIRE(portentry != NULL && portentry->refs > 0);

	qid = DNS_QID(disp);
	LOCK(&qid->lock);
	portentry->refs--;
	if (portentry->refs == 0) {
		ISC_LIST_UNLINK(disp->port_table[portentry->port %
						 DNS_DISPATCH_PORTTABLESIZE],
				portentry, link);
		isc_mempool_put(disp->portpool, portentry);
	}

	*portentryp = NULL;
	UNLOCK(&qid->lock);
}

/*%
 * Find a dispsocket for socket address 'dest', and port number 'port'.
 * Return NULL if no such entry exists.
 */
static dispsocket_t *
socket_search(dns_qid_t *qid, isc_sockaddr_t *dest, in_port_t port,
	      unsigned int bucket)
{
	dispsocket_t *dispsock;

	REQUIRE(bucket < qid->qid_nbuckets);

	dispsock = ISC_LIST_HEAD(qid->sock_table[bucket]);

	while (dispsock != NULL) {
		if (dispsock->portentry != NULL &&
		    dispsock->portentry->port == port &&
		    isc_sockaddr_equal(dest, &dispsock->host))
			return (dispsock);
		dispsock = ISC_LIST_NEXT(dispsock, blink);
	}

	return (NULL);
}

/*%
 * Make a new socket for a single dispatch with a random port number.
 * The caller must hold the disp->lock and qid->lock.
 */
static isc_result_t
get_dispsocket(dns_dispatch_t *disp, isc_sockaddr_t *dest,
	       isc_socketmgr_t *sockmgr, dns_qid_t *qid,
	       dispsocket_t **dispsockp, in_port_t *portp)
{
	int i;
	isc_uint32_t r;
	dns_dispatchmgr_t *mgr = disp->mgr;
	isc_socket_t *sock = NULL;
	isc_result_t result = ISC_R_FAILURE;
	in_port_t port;
	isc_sockaddr_t localaddr;
	unsigned int bucket = 0;
	dispsocket_t *dispsock;
	unsigned int nports;
	in_port_t *ports;
	unsigned int bindoptions;
	dispportentry_t *portentry = NULL;

	if (isc_sockaddr_pf(&disp->local) == AF_INET) {
		nports = disp->mgr->nv4ports;
		ports = disp->mgr->v4ports;
	} else {
		nports = disp->mgr->nv6ports;
		ports = disp->mgr->v6ports;
	}
	if (nports == 0)
		return (ISC_R_ADDRNOTAVAIL);

	dispsock = ISC_LIST_HEAD(disp->inactivesockets);
	if (dispsock != NULL) {
		ISC_LIST_UNLINK(disp->inactivesockets, dispsock, link);
		sock = dispsock->socket;
		dispsock->socket = NULL;
	} else {
		dispsock = isc_mempool_get(mgr->spool);
		if (dispsock == NULL)
			return (ISC_R_NOMEMORY);

		disp->nsockets++;
		dispsock->socket = NULL;
		dispsock->disp = disp;
		dispsock->resp = NULL;
		dispsock->portentry = NULL;
		isc_random_get(&r);
		dispsock->task = NULL;
		isc_task_attach(disp->task[r % disp->ntasks], &dispsock->task);
		ISC_LINK_INIT(dispsock, link);
		ISC_LINK_INIT(dispsock, blink);
		dispsock->magic = DISPSOCK_MAGIC;
	}

	/*
	 * Pick up a random UDP port and open a new socket with it.  Avoid
	 * choosing ports that share the same destination because it will be
	 * very likely to fail in bind(2) or connect(2).
	 */
	localaddr = disp->local;
	for (i = 0; i < 64; i++) {
		port = ports[dispatch_uniformrandom(DISP_ARC4CTX(disp),
							nports)];
		isc_sockaddr_setport(&localaddr, port);

		bucket = dns_hash(qid, dest, 0, port);
		if (socket_search(qid, dest, port, bucket) != NULL)
			continue;
		bindoptions = 0;
		portentry = port_search(disp, port);
		if (portentry != NULL)
			bindoptions |= ISC_SOCKET_REUSEADDRESS;
		result = open_socket(sockmgr, &localaddr, bindoptions, &sock);
		if (result == ISC_R_SUCCESS) {
			if (portentry == NULL) {
				portentry = new_portentry(disp, port);
				if (portentry == NULL) {
					result = ISC_R_NOMEMORY;
					break;
				}
			}
			portentry->refs++;
			break;
		} else if (result != ISC_R_ADDRINUSE)
			break;
	}

	if (result == ISC_R_SUCCESS) {
		dispsock->socket = sock;
		dispsock->host = *dest;
		dispsock->portentry = portentry;
		dispsock->bucket = bucket;
		ISC_LIST_APPEND(qid->sock_table[bucket], dispsock, blink);
		*dispsockp = dispsock;
		*portp = port;
	} else {
		/*
		 * We could keep it in the inactive list, but since this should
		 * be an exceptional case and might be resource shortage, we'd
		 * rather destroy it.
		 */
		if (sock != NULL)
			isc_socket_detach(&sock);
		destroy_dispsocket(disp, &dispsock);
	}

	return (result);
}

/*%
 * Destroy a dedicated dispatch socket.
 */
static void
destroy_dispsocket(dns_dispatch_t *disp, dispsocket_t **dispsockp) {
	dispsocket_t *dispsock;
	dns_qid_t *qid;

	/*
	 * The dispatch must be locked.
	 */

	REQUIRE(dispsockp != NULL && *dispsockp != NULL);
	dispsock = *dispsockp;
	REQUIRE(!ISC_LINK_LINKED(dispsock, link));

	disp->nsockets--;
	dispsock->magic = 0;
	if (dispsock->portentry != NULL)
		deref_portentry(disp, &dispsock->portentry);
	if (dispsock->socket != NULL)
		isc_socket_detach(&dispsock->socket);
	if (ISC_LINK_LINKED(dispsock, blink)) {
		qid = DNS_QID(disp);
		LOCK(&qid->lock);
		ISC_LIST_UNLINK(qid->sock_table[dispsock->bucket], dispsock,
				blink);
		UNLOCK(&qid->lock);
	}
	if (dispsock->task != NULL)
		isc_task_detach(&dispsock->task);
	isc_mempool_put(disp->mgr->spool, dispsock);

	*dispsockp = NULL;
}

/*%
 * Deactivate a dedicated dispatch socket.  Move it to the inactive list for
 * future reuse unless the total number of sockets are exceeding the maximum.
 */
static void
deactivate_dispsocket(dns_dispatch_t *disp, dispsocket_t *dispsock) {
	isc_result_t result;
	dns_qid_t *qid;

	/*
	 * The dispatch must be locked.
	 */
	ISC_LIST_UNLINK(disp->activesockets, dispsock, link);
	if (dispsock->resp != NULL) {
		INSIST(dispsock->resp->dispsocket == dispsock);
		dispsock->resp->dispsocket = NULL;
	}

	INSIST(dispsock->portentry != NULL);
	deref_portentry(disp, &dispsock->portentry);

#ifdef BIND9
	if (disp->nsockets > DNS_DISPATCH_POOLSOCKS)
		destroy_dispsocket(disp, &dispsock);
	else {
		result = isc_socket_close(dispsock->socket);

		qid = DNS_QID(disp);
		LOCK(&qid->lock);
		ISC_LIST_UNLINK(qid->sock_table[dispsock->bucket], dispsock,
				blink);
		UNLOCK(&qid->lock);

		if (result == ISC_R_SUCCESS)
			ISC_LIST_APPEND(disp->inactivesockets, dispsock, link);
		else {
			/*
			 * If the underlying system does not allow this
			 * optimization, destroy this temporary structure (and
			 * create a new one for a new transaction).
			 */
			INSIST(result == ISC_R_NOTIMPLEMENTED);
			destroy_dispsocket(disp, &dispsock);
		}
	}
#else
	/* This kind of optimization isn't necessary for normal use */
	UNUSED(qid);
	UNUSED(result);

	destroy_dispsocket(disp, &dispsock);
#endif
}

/*
 * Find an entry for query ID 'id', socket address 'dest', and port number
 * 'port'.
 * Return NULL if no such entry exists.
 */
static dns_dispentry_t *
entry_search(dns_qid_t *qid, isc_sockaddr_t *dest, dns_messageid_t id,
	     in_port_t port, unsigned int bucket)
{
	dns_dispentry_t *res;

	REQUIRE(bucket < qid->qid_nbuckets);

	res = ISC_LIST_HEAD(qid->qid_table[bucket]);

	while (res != NULL) {
		if (res->id == id && isc_sockaddr_equal(dest, &res->host) &&
		    res->port == port) {
			return (res);
		}
		res = ISC_LIST_NEXT(res, link);
	}

	return (NULL);
}

static void
free_buffer(dns_dispatch_t *disp, void *buf, unsigned int len) {
	INSIST(buf != NULL && len != 0);


	switch (disp->socktype) {
	case isc_sockettype_tcp:
		INSIST(disp->tcpbuffers > 0);
		disp->tcpbuffers--;
		isc_mem_put(disp->mgr->mctx, buf, len);
		break;
	case isc_sockettype_udp:
		LOCK(&disp->mgr->buffer_lock);
		INSIST(disp->mgr->buffers > 0);
		INSIST(len == disp->mgr->buffersize);
		disp->mgr->buffers--;
		isc_mempool_put(disp->mgr->bpool, buf);
		UNLOCK(&disp->mgr->buffer_lock);
		break;
	default:
		INSIST(0);
		break;
	}
}

static void *
allocate_udp_buffer(dns_dispatch_t *disp) {
	void *temp;

	LOCK(&disp->mgr->buffer_lock);
	temp = isc_mempool_get(disp->mgr->bpool);

	if (temp != NULL)
		disp->mgr->buffers++;
	UNLOCK(&disp->mgr->buffer_lock);

	return (temp);
}

static inline void
free_event(dns_dispatch_t *disp, dns_dispatchevent_t *ev) {
	if (disp->failsafe_ev == ev) {
		INSIST(disp->shutdown_out == 1);
		disp->shutdown_out = 0;

		return;
	}

	isc_mempool_put(disp->mgr->epool, ev);
}

static inline dns_dispatchevent_t *
allocate_event(dns_dispatch_t *disp) {
	dns_dispatchevent_t *ev;

	ev = isc_mempool_get(disp->mgr->epool);
	if (ev == NULL)
		return (NULL);
	ISC_EVENT_INIT(ev, sizeof(*ev), 0, NULL, 0,
		       NULL, NULL, NULL, NULL, NULL);

	return (ev);
}

static void
udp_exrecv(isc_task_t *task, isc_event_t *ev) {
	dispsocket_t *dispsock = ev->ev_arg;

	UNUSED(task);

	REQUIRE(VALID_DISPSOCK(dispsock));
	udp_recv(ev, dispsock->disp, dispsock);
}

static void
udp_shrecv(isc_task_t *task, isc_event_t *ev) {
	dns_dispatch_t *disp = ev->ev_arg;

	UNUSED(task);

	REQUIRE(VALID_DISPATCH(disp));
	udp_recv(ev, disp, NULL);
}

/*
 * General flow:
 *
 * If I/O result == CANCELED or error, free the buffer.
 *
 * If query, free the buffer, restart.
 *
 * If response:
 *	Allocate event, fill in details.
 *		If cannot allocate, free buffer, restart.
 *	find target.  If not found, free buffer, restart.
 *	if event queue is not empty, queue.  else, send.
 *	restart.
 */
static void
udp_recv(isc_event_t *ev_in, dns_dispatch_t *disp, dispsocket_t *dispsock) {
	isc_socketevent_t *ev = (isc_socketevent_t *)ev_in;
	dns_messageid_t id;
	isc_result_t dres;
	isc_buffer_t source;
	unsigned int flags;
	dns_dispentry_t *resp = NULL;
	dns_dispatchevent_t *rev;
	unsigned int bucket;
	isc_boolean_t killit;
	isc_boolean_t queue_response;
	dns_dispatchmgr_t *mgr;
	dns_qid_t *qid;
	isc_netaddr_t netaddr;
	int match;
	int result;
	isc_boolean_t qidlocked = ISC_FALSE;

	LOCK(&disp->lock);

	mgr = disp->mgr;
	qid = mgr->qid;

	dispatch_log(disp, LVL(90),
		     "got packet: requests %d, buffers %d, recvs %d",
		     disp->requests, disp->mgr->buffers, disp->recv_pending);

	if (dispsock == NULL && ev->ev_type == ISC_SOCKEVENT_RECVDONE) {
		/*
		 * Unless the receive event was imported from a listening
		 * interface, in which case the event type is
		 * DNS_EVENT_IMPORTRECVDONE, receive operation must be pending.
		 */
		INSIST(disp->recv_pending != 0);
		disp->recv_pending = 0;
	}

	if (dispsock != NULL &&
	    (ev->result == ISC_R_CANCELED || dispsock->resp == NULL)) {
		/*
		 * dispsock->resp can be NULL if this transaction was canceled
		 * just after receiving a response.  Since this socket is
		 * exclusively used and there should be at most one receive
		 * event the canceled event should have been no effect.  So
		 * we can (and should) deactivate the socket right now.
		 */
		deactivate_dispsocket(disp, dispsock);
		dispsock = NULL;
	}

	if (disp->shutting_down) {
		/*
		 * This dispatcher is shutting down.
		 */
		free_buffer(disp, ev->region.base, ev->region.length);

		isc_event_free(&ev_in);
		ev = NULL;

		killit = destroy_disp_ok(disp);
		UNLOCK(&disp->lock);
		if (killit)
			isc_task_send(disp->task[0], &disp->ctlevent);

		return;
	}

	if ((disp->attributes & DNS_DISPATCHATTR_EXCLUSIVE) != 0) {
		if (dispsock != NULL) {
			resp = dispsock->resp;
			id = resp->id;
			if (ev->result != ISC_R_SUCCESS) {
				/*
				 * This is most likely a network error on a
				 * connected socket.  It makes no sense to
				 * check the address or parse the packet, but it
				 * will help to return the error to the caller.
				 */
				goto sendresponse;
			}
		} else {
			free_buffer(disp, ev->region.base, ev->region.length);

			UNLOCK(&disp->lock);
			isc_event_free(&ev_in);
			return;
		}
	} else if (ev->result != ISC_R_SUCCESS) {
		free_buffer(disp, ev->region.base, ev->region.length);

		if (ev->result != ISC_R_CANCELED)
			dispatch_log(disp, ISC_LOG_ERROR,
				     "odd socket result in udp_recv(): %s",
				     isc_result_totext(ev->result));

		UNLOCK(&disp->lock);
		isc_event_free(&ev_in);
		return;
	}

	/*
	 * If this is from a blackholed address, drop it.
	 */
	isc_netaddr_fromsockaddr(&netaddr, &ev->address);
	if (disp->mgr->blackhole != NULL &&
	    dns_acl_match(&netaddr, NULL, disp->mgr->blackhole,
			  NULL, &match, NULL) == ISC_R_SUCCESS &&
	    match > 0)
	{
		if (isc_log_wouldlog(dns_lctx, LVL(10))) {
			char netaddrstr[ISC_NETADDR_FORMATSIZE];
			isc_netaddr_format(&netaddr, netaddrstr,
					   sizeof(netaddrstr));
			dispatch_log(disp, LVL(10),
				     "blackholed packet from %s",
				     netaddrstr);
		}
		free_buffer(disp, ev->region.base, ev->region.length);
		goto restart;
	}

	/*
	 * Peek into the buffer to see what we can see.
	 */
	isc_buffer_init(&source, ev->region.base, ev->region.length);
	isc_buffer_add(&source, ev->n);
	dres = dns_message_peekheader(&source, &id, &flags);
	if (dres != ISC_R_SUCCESS) {
		free_buffer(disp, ev->region.base, ev->region.length);
		dispatch_log(disp, LVL(10), "got garbage packet");
		goto restart;
	}

	dispatch_log(disp, LVL(92),
		     "got valid DNS message header, /QR %c, id %u",
		     ((flags & DNS_MESSAGEFLAG_QR) ? '1' : '0'), id);

	/*
	 * Look at flags.  If query, drop it. If response,
	 * look to see where it goes.
	 */
	queue_response = ISC_FALSE;
	if ((flags & DNS_MESSAGEFLAG_QR) == 0) {
		/* query */
		free_buffer(disp, ev->region.base, ev->region.length);
		goto restart;
	}

	/*
	 * Search for the corresponding response.  If we are using an exclusive
	 * socket, we've already identified it and we can skip the search; but
	 * the ID and the address must match the expected ones.
	 */
	if (resp == NULL) {
		bucket = dns_hash(qid, &ev->address, id, disp->localport);
		LOCK(&qid->lock);
		qidlocked = ISC_TRUE;
		resp = entry_search(qid, &ev->address, id, disp->localport,
				    bucket);
		dispatch_log(disp, LVL(90),
			     "search for response in bucket %d: %s",
			     bucket, (resp == NULL ? "not found" : "found"));

		if (resp == NULL) {
			inc_stats(mgr, dns_resstatscounter_mismatch);
			free_buffer(disp, ev->region.base, ev->region.length);
			goto unlock;
		}
	} else if (resp->id != id || !isc_sockaddr_equal(&ev->address,
							 &resp->host)) {
		dispatch_log(disp, LVL(90),
			     "response to an exclusive socket doesn't match");
		inc_stats(mgr, dns_resstatscounter_mismatch);
		free_buffer(disp, ev->region.base, ev->region.length);
		goto unlock;
	}

	/*
	 * Now that we have the original dispatch the query was sent
	 * from check that the address and port the response was
	 * sent to make sense.
	 */
	if (disp != resp->disp) {
		isc_sockaddr_t a1;
		isc_sockaddr_t a2;

		/*
		 * Check that the socket types and ports match.
		 */
		if (disp->socktype != resp->disp->socktype ||
		    isc_sockaddr_getport(&disp->local) !=
		    isc_sockaddr_getport(&resp->disp->local)) {
			free_buffer(disp, ev->region.base, ev->region.length);
			goto unlock;
		}

		/*
		 * If both dispatches are bound to an address then fail as
		 * the addresses can't be equal (enforced by the IP stack).
		 *
		 * Note under Linux a packet can be sent out via IPv4 socket
		 * and the response be received via a IPv6 socket.
		 *
		 * Requests sent out via IPv6 should always come back in
		 * via IPv6.
		 */
		if (isc_sockaddr_pf(&resp->disp->local) == PF_INET6 &&
		    isc_sockaddr_pf(&disp->local) != PF_INET6) {
			free_buffer(disp, ev->region.base, ev->region.length);
			goto unlock;
		}
		isc_sockaddr_anyofpf(&a1, isc_sockaddr_pf(&resp->disp->local));
		isc_sockaddr_anyofpf(&a2, isc_sockaddr_pf(&disp->local));
		if (!isc_sockaddr_eqaddr(&a1, &resp->disp->local) &&
		    !isc_sockaddr_eqaddr(&a2, &disp->local)) {
			free_buffer(disp, ev->region.base, ev->region.length);
			goto unlock;
		}
	}

  sendresponse:
	queue_response = resp->item_out;
	rev = allocate_event(resp->disp);
	if (rev == NULL) {
		free_buffer(disp, ev->region.base, ev->region.length);
		goto unlock;
	}

	/*
	 * At this point, rev contains the event we want to fill in, and
	 * resp contains the information on the place to send it to.
	 * Send the event off.
	 */
	isc_buffer_init(&rev->buffer, ev->region.base, ev->region.length);
	isc_buffer_add(&rev->buffer, ev->n);
	rev->result = ev->result;
	rev->id = id;
	rev->addr = ev->address;
	rev->pktinfo = ev->pktinfo;
	rev->attributes = ev->attributes;
	if (queue_response) {
		ISC_LIST_APPEND(resp->items, rev, ev_link);
	} else {
		ISC_EVENT_INIT(rev, sizeof(*rev), 0, NULL,
			       DNS_EVENT_DISPATCH,
			       resp->action, resp->arg, resp, NULL, NULL);
		request_log(disp, resp, LVL(90),
			    "[a] Sent event %p buffer %p len %d to task %p",
			    rev, rev->buffer.base, rev->buffer.length,
			    resp->task);
		resp->item_out = ISC_TRUE;
		isc_task_send(resp->task, ISC_EVENT_PTR(&rev));
	}
 unlock:
	if (qidlocked)
		UNLOCK(&qid->lock);

	/*
	 * Restart recv() to get the next packet.
	 */
 restart:
	result = startrecv(disp, dispsock);
	if (result != ISC_R_SUCCESS && dispsock != NULL) {
		/*
		 * XXX: wired. There seems to be no recovery process other than
		 * deactivate this socket anyway (since we cannot start
		 * receiving, we won't be able to receive a cancel event
		 * from the user).
		 */
		deactivate_dispsocket(disp, dispsock);
	}
	UNLOCK(&disp->lock);

	isc_event_free(&ev_in);
}

/*
 * General flow:
 *
 * If I/O result == CANCELED, EOF, or error, notify everyone as the
 * various queues drain.
 *
 * If query, restart.
 *
 * If response:
 *	Allocate event, fill in details.
 *		If cannot allocate, restart.
 *	find target.  If not found, restart.
 *	if event queue is not empty, queue.  else, send.
 *	restart.
 */
static void
tcp_recv(isc_task_t *task, isc_event_t *ev_in) {
	dns_dispatch_t *disp = ev_in->ev_arg;
	dns_tcpmsg_t *tcpmsg = &disp->tcpmsg;
	dns_messageid_t id;
	isc_result_t dres;
	unsigned int flags;
	dns_dispentry_t *resp;
	dns_dispatchevent_t *rev;
	unsigned int bucket;
	isc_boolean_t killit;
	isc_boolean_t queue_response;
	dns_qid_t *qid;
	int level;
	char buf[ISC_SOCKADDR_FORMATSIZE];

	UNUSED(task);

	REQUIRE(VALID_DISPATCH(disp));

	qid = disp->qid;

	dispatch_log(disp, LVL(90),
		     "got TCP packet: requests %d, buffers %d, recvs %d",
		     disp->requests, disp->tcpbuffers, disp->recv_pending);

	LOCK(&disp->lock);

	INSIST(disp->recv_pending != 0);
	disp->recv_pending = 0;

	if (disp->refcount == 0) {
		/*
		 * This dispatcher is shutting down.  Force cancelation.
		 */
		tcpmsg->result = ISC_R_CANCELED;
	}

	if (tcpmsg->result != ISC_R_SUCCESS) {
		switch (tcpmsg->result) {
		case ISC_R_CANCELED:
			break;

		case ISC_R_EOF:
			dispatch_log(disp, LVL(90), "shutting down on EOF");
			do_cancel(disp);
			break;

		case ISC_R_CONNECTIONRESET:
			level = ISC_LOG_INFO;
			goto logit;

		default:
			level = ISC_LOG_ERROR;
		logit:
			isc_sockaddr_format(&tcpmsg->address, buf, sizeof(buf));
			dispatch_log(disp, level, "shutting down due to TCP "
				     "receive error: %s: %s", buf,
				     isc_result_totext(tcpmsg->result));
			do_cancel(disp);
			break;
		}

		/*
		 * The event is statically allocated in the tcpmsg
		 * structure, and destroy_disp() frees the tcpmsg, so we must
		 * free the event *before* calling destroy_disp().
		 */
		isc_event_free(&ev_in);

		disp->shutting_down = 1;
		disp->shutdown_why = tcpmsg->result;

		/*
		 * If the recv() was canceled pass the word on.
		 */
		killit = destroy_disp_ok(disp);
		UNLOCK(&disp->lock);
		if (killit)
			isc_task_send(disp->task[0], &disp->ctlevent);
		return;
	}

	dispatch_log(disp, LVL(90), "result %d, length == %d, addr = %p",
		     tcpmsg->result,
		     tcpmsg->buffer.length, tcpmsg->buffer.base);

	/*
	 * Peek into the buffer to see what we can see.
	 */
	dres = dns_message_peekheader(&tcpmsg->buffer, &id, &flags);
	if (dres != ISC_R_SUCCESS) {
		dispatch_log(disp, LVL(10), "got garbage packet");
		goto restart;
	}

	dispatch_log(disp, LVL(92),
		     "got valid DNS message header, /QR %c, id %u",
		     ((flags & DNS_MESSAGEFLAG_QR) ? '1' : '0'), id);

	/*
	 * Allocate an event to send to the query or response client, and
	 * allocate a new buffer for our use.
	 */

	/*
	 * Look at flags.  If query, drop it. If response,
	 * look to see where it goes.
	 */
	queue_response = ISC_FALSE;
	if ((flags & DNS_MESSAGEFLAG_QR) == 0) {
		/*
		 * Query.
		 */
		goto restart;
	}

	/*
	 * Response.
	 */
	bucket = dns_hash(qid, &tcpmsg->address, id, disp->localport);
	LOCK(&qid->lock);
	resp = entry_search(qid, &tcpmsg->address, id, disp->localport, bucket);
	dispatch_log(disp, LVL(90),
		     "search for response in bucket %d: %s",
		     bucket, (resp == NULL ? "not found" : "found"));

	if (resp == NULL)
		goto unlock;
	queue_response = resp->item_out;
	rev = allocate_event(disp);
	if (rev == NULL)
		goto unlock;

	/*
	 * At this point, rev contains the event we want to fill in, and
	 * resp contains the information on the place to send it to.
	 * Send the event off.
	 */
	dns_tcpmsg_keepbuffer(tcpmsg, &rev->buffer);
	disp->tcpbuffers++;
	rev->result = ISC_R_SUCCESS;
	rev->id = id;
	rev->addr = tcpmsg->address;
	if (queue_response) {
		ISC_LIST_APPEND(resp->items, rev, ev_link);
	} else {
		ISC_EVENT_INIT(rev, sizeof(*rev), 0, NULL, DNS_EVENT_DISPATCH,
			       resp->action, resp->arg, resp, NULL, NULL);
		request_log(disp, resp, LVL(90),
			    "[b] Sent event %p buffer %p len %d to task %p",
			    rev, rev->buffer.base, rev->buffer.length,
			    resp->task);
		resp->item_out = ISC_TRUE;
		isc_task_send(resp->task, ISC_EVENT_PTR(&rev));
	}
 unlock:
	UNLOCK(&qid->lock);

	/*
	 * Restart recv() to get the next packet.
	 */
 restart:
	(void)startrecv(disp, NULL);

	UNLOCK(&disp->lock);

	isc_event_free(&ev_in);
}

/*
 * disp must be locked.
 */
static isc_result_t
startrecv(dns_dispatch_t *disp, dispsocket_t *dispsock) {
	isc_result_t res;
	isc_region_t region;
	isc_socket_t *socket;

	if (disp->shutting_down == 1)
		return (ISC_R_SUCCESS);

	if ((disp->attributes & DNS_DISPATCHATTR_NOLISTEN) != 0)
		return (ISC_R_SUCCESS);

	if (disp->recv_pending != 0 && dispsock == NULL)
		return (ISC_R_SUCCESS);

	if (disp->mgr->buffers >= disp->mgr->maxbuffers)
		return (ISC_R_NOMEMORY);

	if ((disp->attributes & DNS_DISPATCHATTR_EXCLUSIVE) != 0 &&
	    dispsock == NULL)
		return (ISC_R_SUCCESS);

	if (dispsock != NULL)
		socket = dispsock->socket;
	else
		socket = disp->socket;
	INSIST(socket != NULL);

	switch (disp->socktype) {
		/*
		 * UDP reads are always maximal.
		 */
	case isc_sockettype_udp:
		region.length = disp->mgr->buffersize;
		region.base = allocate_udp_buffer(disp);
		if (region.base == NULL)
			return (ISC_R_NOMEMORY);
		if (dispsock != NULL) {
			res = isc_socket_recv(socket, &region, 1,
					      dispsock->task, udp_exrecv,
					      dispsock);
			if (res != ISC_R_SUCCESS) {
				free_buffer(disp, region.base, region.length);
				return (res);
			}
		} else {
			res = isc_socket_recv(socket, &region, 1,
					      disp->task[0], udp_shrecv, disp);
			if (res != ISC_R_SUCCESS) {
				free_buffer(disp, region.base, region.length);
				disp->shutdown_why = res;
				disp->shutting_down = 1;
				do_cancel(disp);
				return (ISC_R_SUCCESS); /* recover by cancel */
			}
			INSIST(disp->recv_pending == 0);
			disp->recv_pending = 1;
		}
		break;

	case isc_sockettype_tcp:
		res = dns_tcpmsg_readmessage(&disp->tcpmsg, disp->task[0],
					     tcp_recv, disp);
		if (res != ISC_R_SUCCESS) {
			disp->shutdown_why = res;
			disp->shutting_down = 1;
			do_cancel(disp);
			return (ISC_R_SUCCESS); /* recover by cancel */
		}
		INSIST(disp->recv_pending == 0);
		disp->recv_pending = 1;
		break;
	default:
		INSIST(0);
		break;
	}

	return (ISC_R_SUCCESS);
}

/*
 * Mgr must be locked when calling this function.
 */
static isc_boolean_t
destroy_mgr_ok(dns_dispatchmgr_t *mgr) {
	mgr_log(mgr, LVL(90),
		"destroy_mgr_ok: shuttingdown=%d, listnonempty=%d, "
		"epool=%d, rpool=%d, dpool=%d",
		MGR_IS_SHUTTINGDOWN(mgr), !ISC_LIST_EMPTY(mgr->list),
		isc_mempool_getallocated(mgr->epool),
		isc_mempool_getallocated(mgr->rpool),
		isc_mempool_getallocated(mgr->dpool));
	if (!MGR_IS_SHUTTINGDOWN(mgr))
		return (ISC_FALSE);
	if (!ISC_LIST_EMPTY(mgr->list))
		return (ISC_FALSE);
	if (isc_mempool_getallocated(mgr->epool) != 0)
		return (ISC_FALSE);
	if (isc_mempool_getallocated(mgr->rpool) != 0)
		return (ISC_FALSE);
	if (isc_mempool_getallocated(mgr->dpool) != 0)
		return (ISC_FALSE);

	return (ISC_TRUE);
}

/*
 * Mgr must be unlocked when calling this function.
 */
static void
destroy_mgr(dns_dispatchmgr_t **mgrp) {
	isc_mem_t *mctx;
	dns_dispatchmgr_t *mgr;

	mgr = *mgrp;
	*mgrp = NULL;

	mctx = mgr->mctx;

	mgr->magic = 0;
	mgr->mctx = NULL;
	DESTROYLOCK(&mgr->lock);
	mgr->state = 0;

	DESTROYLOCK(&mgr->arc4_lock);

	isc_mempool_destroy(&mgr->epool);
	isc_mempool_destroy(&mgr->rpool);
	isc_mempool_destroy(&mgr->dpool);
	isc_mempool_destroy(&mgr->bpool);
	isc_mempool_destroy(&mgr->spool);

	DESTROYLOCK(&mgr->pool_lock);

#ifdef BIND9
	if (mgr->entropy != NULL)
		isc_entropy_detach(&mgr->entropy);
#endif /* BIND9 */
	if (mgr->qid != NULL)
		qid_destroy(mctx, &mgr->qid);

	DESTROYLOCK(&mgr->buffer_lock);

	if (mgr->blackhole != NULL)
		dns_acl_detach(&mgr->blackhole);

	if (mgr->stats != NULL)
		isc_stats_detach(&mgr->stats);

	if (mgr->v4ports != NULL) {
		isc_mem_put(mctx, mgr->v4ports,
			    mgr->nv4ports * sizeof(in_port_t));
	}
	if (mgr->v6ports != NULL) {
		isc_mem_put(mctx, mgr->v6ports,
			    mgr->nv6ports * sizeof(in_port_t));
	}
	isc_mem_put(mctx, mgr, sizeof(dns_dispatchmgr_t));
	isc_mem_detach(&mctx);
}

static isc_result_t
open_socket(isc_socketmgr_t *mgr, isc_sockaddr_t *local,
	    unsigned int options, isc_socket_t **sockp)
{
	isc_socket_t *sock;
	isc_result_t result;

	sock = *sockp;
	if (sock == NULL) {
		result = isc_socket_create(mgr, isc_sockaddr_pf(local),
					   isc_sockettype_udp, &sock);
		if (result != ISC_R_SUCCESS)
			return (result);
		isc_socket_setname(sock, "dispatcher", NULL);
	} else {
#ifdef BIND9
		result = isc_socket_open(sock);
		if (result != ISC_R_SUCCESS)
			return (result);
#else
		INSIST(0);
#endif
	}

#ifndef ISC_ALLOW_MAPPED
	isc_socket_ipv6only(sock, ISC_TRUE);
#endif
	result = isc_socket_bind(sock, local, options);
	if (result != ISC_R_SUCCESS) {
		if (*sockp == NULL)
			isc_socket_detach(&sock);
		else {
#ifdef BIND9
			isc_socket_close(sock);
#else
			INSIST(0);
#endif
		}
		return (result);
	}

	*sockp = sock;
	return (ISC_R_SUCCESS);
}

/*%
 * Create a temporary port list to set the initial default set of dispatch
 * ports: [1024, 65535].  This is almost meaningless as the application will
 * normally set the ports explicitly, but is provided to fill some minor corner
 * cases.
 */
static isc_result_t
create_default_portset(isc_mem_t *mctx, isc_portset_t **portsetp) {
	isc_result_t result;

	result = isc_portset_create(mctx, portsetp);
	if (result != ISC_R_SUCCESS)
		return (result);
	isc_portset_addrange(*portsetp, 1024, 65535);

	return (ISC_R_SUCCESS);
}

/*
 * Publics.
 */

isc_result_t
dns_dispatchmgr_create(isc_mem_t *mctx, isc_entropy_t *entropy,
		       dns_dispatchmgr_t **mgrp)
{
	dns_dispatchmgr_t *mgr;
	isc_result_t result;
	isc_portset_t *v4portset = NULL;
	isc_portset_t *v6portset = NULL;

	REQUIRE(mctx != NULL);
	REQUIRE(mgrp != NULL && *mgrp == NULL);

	mgr = isc_mem_get(mctx, sizeof(dns_dispatchmgr_t));
	if (mgr == NULL)
		return (ISC_R_NOMEMORY);

	mgr->mctx = NULL;
	isc_mem_attach(mctx, &mgr->mctx);

	mgr->blackhole = NULL;
	mgr->stats = NULL;

	result = isc_mutex_init(&mgr->lock);
	if (result != ISC_R_SUCCESS)
		goto deallocate;

	result = isc_mutex_init(&mgr->arc4_lock);
	if (result != ISC_R_SUCCESS)
		goto kill_lock;

	result = isc_mutex_init(&mgr->buffer_lock);
	if (result != ISC_R_SUCCESS)
		goto kill_arc4_lock;

	result = isc_mutex_init(&mgr->pool_lock);
	if (result != ISC_R_SUCCESS)
		goto kill_buffer_lock;

	mgr->epool = NULL;
	if (isc_mempool_create(mgr->mctx, sizeof(dns_dispatchevent_t),
			       &mgr->epool) != ISC_R_SUCCESS) {
		result = ISC_R_NOMEMORY;
		goto kill_pool_lock;
	}

	mgr->rpool = NULL;
	if (isc_mempool_create(mgr->mctx, sizeof(dns_dispentry_t),
			       &mgr->rpool) != ISC_R_SUCCESS) {
		result = ISC_R_NOMEMORY;
		goto kill_epool;
	}

	mgr->dpool = NULL;
	if (isc_mempool_create(mgr->mctx, sizeof(dns_dispatch_t),
			       &mgr->dpool) != ISC_R_SUCCESS) {
		result = ISC_R_NOMEMORY;
		goto kill_rpool;
	}

	isc_mempool_setname(mgr->epool, "dispmgr_epool");
	isc_mempool_setfreemax(mgr->epool, 1024);
	isc_mempool_associatelock(mgr->epool, &mgr->pool_lock);

	isc_mempool_setname(mgr->rpool, "dispmgr_rpool");
	isc_mempool_setfreemax(mgr->rpool, 1024);
	isc_mempool_associatelock(mgr->rpool, &mgr->pool_lock);

	isc_mempool_setname(mgr->dpool, "dispmgr_dpool");
	isc_mempool_setfreemax(mgr->dpool, 1024);
	isc_mempool_associatelock(mgr->dpool, &mgr->pool_lock);

	mgr->buffers = 0;
	mgr->buffersize = 0;
	mgr->maxbuffers = 0;
	mgr->bpool = NULL;
	mgr->spool = NULL;
	mgr->entropy = NULL;
	mgr->qid = NULL;
	mgr->state = 0;
	ISC_LIST_INIT(mgr->list);
	mgr->v4ports = NULL;
	mgr->v6ports = NULL;
	mgr->nv4ports = 0;
	mgr->nv6ports = 0;
	mgr->magic = DNS_DISPATCHMGR_MAGIC;

	result = create_default_portset(mctx, &v4portset);
	if (result == ISC_R_SUCCESS) {
		result = create_default_portset(mctx, &v6portset);
		if (result == ISC_R_SUCCESS) {
			result = dns_dispatchmgr_setavailports(mgr,
							       v4portset,
							       v6portset);
		}
	}
	if (v4portset != NULL)
		isc_portset_destroy(mctx, &v4portset);
	if (v6portset != NULL)
		isc_portset_destroy(mctx, &v6portset);
	if (result != ISC_R_SUCCESS)
		goto kill_dpool;

#ifdef BIND9
	if (entropy != NULL)
		isc_entropy_attach(entropy, &mgr->entropy);
#else
	UNUSED(entropy);
#endif

	dispatch_initrandom(&mgr->arc4ctx, mgr->entropy, &mgr->arc4_lock);

	*mgrp = mgr;
	return (ISC_R_SUCCESS);

 kill_dpool:
	isc_mempool_destroy(&mgr->dpool);
 kill_rpool:
	isc_mempool_destroy(&mgr->rpool);
 kill_epool:
	isc_mempool_destroy(&mgr->epool);
 kill_pool_lock:
	DESTROYLOCK(&mgr->pool_lock);
 kill_buffer_lock:
	DESTROYLOCK(&mgr->buffer_lock);
 kill_arc4_lock:
	DESTROYLOCK(&mgr->arc4_lock);
 kill_lock:
	DESTROYLOCK(&mgr->lock);
 deallocate:
	isc_mem_put(mctx, mgr, sizeof(dns_dispatchmgr_t));
	isc_mem_detach(&mctx);

	return (result);
}

void
dns_dispatchmgr_setblackhole(dns_dispatchmgr_t *mgr, dns_acl_t *blackhole) {
	REQUIRE(VALID_DISPATCHMGR(mgr));
	if (mgr->blackhole != NULL)
		dns_acl_detach(&mgr->blackhole);
	dns_acl_attach(blackhole, &mgr->blackhole);
}

dns_acl_t *
dns_dispatchmgr_getblackhole(dns_dispatchmgr_t *mgr) {
	REQUIRE(VALID_DISPATCHMGR(mgr));
	return (mgr->blackhole);
}

void
dns_dispatchmgr_setblackportlist(dns_dispatchmgr_t *mgr,
				 dns_portlist_t *portlist)
{
	REQUIRE(VALID_DISPATCHMGR(mgr));
	UNUSED(portlist);

	/* This function is deprecated: use dns_dispatchmgr_setavailports(). */
	return;
}

dns_portlist_t *
dns_dispatchmgr_getblackportlist(dns_dispatchmgr_t *mgr) {
	REQUIRE(VALID_DISPATCHMGR(mgr));
	return (NULL);		/* this function is deprecated */
}

isc_result_t
dns_dispatchmgr_setavailports(dns_dispatchmgr_t *mgr, isc_portset_t *v4portset,
			      isc_portset_t *v6portset)
{
	in_port_t *v4ports, *v6ports, p;
	unsigned int nv4ports, nv6ports, i4, i6;

	REQUIRE(VALID_DISPATCHMGR(mgr));

	nv4ports = isc_portset_nports(v4portset);
	nv6ports = isc_portset_nports(v6portset);

	v4ports = NULL;
	if (nv4ports != 0) {
		v4ports = isc_mem_get(mgr->mctx, sizeof(in_port_t) * nv4ports);
		if (v4ports == NULL)
			return (ISC_R_NOMEMORY);
	}
	v6ports = NULL;
	if (nv6ports != 0) {
		v6ports = isc_mem_get(mgr->mctx, sizeof(in_port_t) * nv6ports);
		if (v6ports == NULL) {
			if (v4ports != NULL) {
				isc_mem_put(mgr->mctx, v4ports,
					    sizeof(in_port_t) *
					    isc_portset_nports(v4portset));
			}
			return (ISC_R_NOMEMORY);
		}
	}

	p = 0;
	i4 = 0;
	i6 = 0;
	do {
		if (isc_portset_isset(v4portset, p)) {
			INSIST(i4 < nv4ports);
			v4ports[i4++] = p;
		}
		if (isc_portset_isset(v6portset, p)) {
			INSIST(i6 < nv6ports);
			v6ports[i6++] = p;
		}
	} while (p++ < 65535);
	INSIST(i4 == nv4ports && i6 == nv6ports);

	PORTBUFLOCK(mgr);
	if (mgr->v4ports != NULL) {
		isc_mem_put(mgr->mctx, mgr->v4ports,
			    mgr->nv4ports * sizeof(in_port_t));
	}
	mgr->v4ports = v4ports;
	mgr->nv4ports = nv4ports;

	if (mgr->v6ports != NULL) {
		isc_mem_put(mgr->mctx, mgr->v6ports,
			    mgr->nv6ports * sizeof(in_port_t));
	}
	mgr->v6ports = v6ports;
	mgr->nv6ports = nv6ports;
	PORTBUFUNLOCK(mgr);

	return (ISC_R_SUCCESS);
}

static isc_result_t
dns_dispatchmgr_setudp(dns_dispatchmgr_t *mgr,
		       unsigned int buffersize, unsigned int maxbuffers,
		       unsigned int maxrequests, unsigned int buckets,
		       unsigned int increment)
{
	isc_result_t result;

	REQUIRE(VALID_DISPATCHMGR(mgr));
	REQUIRE(buffersize >= 512 && buffersize < (64 * 1024));
	REQUIRE(maxbuffers > 0);
	REQUIRE(buckets < 2097169);  /* next prime > 65536 * 32 */
	REQUIRE(increment > buckets);

	/*
	 * Keep some number of items around.  This should be a config
	 * option.  For now, keep 8, but later keep at least two even
	 * if the caller wants less.  This allows us to ensure certain
	 * things, like an event can be "freed" and the next allocation
	 * will always succeed.
	 *
	 * Note that if limits are placed on anything here, we use one
	 * event internally, so the actual limit should be "wanted + 1."
	 *
	 * XXXMLG
	 */

	if (maxbuffers < 8)
		maxbuffers = 8;

	LOCK(&mgr->buffer_lock);

	/* Create or adjust buffer pool */
	if (mgr->bpool != NULL) {
		/*
		 * We only increase the maxbuffers to avoid accidental buffer
		 * shortage.  Ideally we'd separate the manager-wide maximum
		 * from per-dispatch limits and respect the latter within the
		 * global limit.  But at this moment that's deemed to be
		 * overkilling and isn't worth additional implementation
		 * complexity.
		 */
		if (maxbuffers > mgr->maxbuffers) {
			isc_mempool_setmaxalloc(mgr->bpool, maxbuffers);
			mgr->maxbuffers = maxbuffers;
		}
	} else {
		result = isc_mempool_create(mgr->mctx, buffersize, &mgr->bpool);
		if (result != ISC_R_SUCCESS) {
			UNLOCK(&mgr->buffer_lock);
			return (result);
		}
		isc_mempool_setname(mgr->bpool, "dispmgr_bpool");
		isc_mempool_setmaxalloc(mgr->bpool, maxbuffers);
		isc_mempool_associatelock(mgr->bpool, &mgr->pool_lock);
	}

	/* Create or adjust socket pool */
	if (mgr->spool != NULL) {
		isc_mempool_setmaxalloc(mgr->spool, DNS_DISPATCH_POOLSOCKS * 2);
		UNLOCK(&mgr->buffer_lock);
		return (ISC_R_SUCCESS);
	}
	result = isc_mempool_create(mgr->mctx, sizeof(dispsocket_t),
				    &mgr->spool);
	if (result != ISC_R_SUCCESS) {
		UNLOCK(&mgr->buffer_lock);
		goto cleanup;
	}
	isc_mempool_setname(mgr->spool, "dispmgr_spool");
	isc_mempool_setmaxalloc(mgr->spool, maxrequests);
	isc_mempool_associatelock(mgr->spool, &mgr->pool_lock);

	result = qid_allocate(mgr, buckets, increment, &mgr->qid, ISC_TRUE);
	if (result != ISC_R_SUCCESS)
		goto cleanup;

	mgr->buffersize = buffersize;
	mgr->maxbuffers = maxbuffers;
	UNLOCK(&mgr->buffer_lock);
	return (ISC_R_SUCCESS);

 cleanup:
	isc_mempool_destroy(&mgr->bpool);
	if (mgr->spool != NULL)
		isc_mempool_destroy(&mgr->spool);
	UNLOCK(&mgr->buffer_lock);
	return (result);
}

void
dns_dispatchmgr_destroy(dns_dispatchmgr_t **mgrp) {
	dns_dispatchmgr_t *mgr;
	isc_boolean_t killit;

	REQUIRE(mgrp != NULL);
	REQUIRE(VALID_DISPATCHMGR(*mgrp));

	mgr = *mgrp;
	*mgrp = NULL;

	LOCK(&mgr->lock);
	mgr->state |= MGR_SHUTTINGDOWN;

	killit = destroy_mgr_ok(mgr);
	UNLOCK(&mgr->lock);

	mgr_log(mgr, LVL(90), "destroy: killit=%d", killit);

	if (killit)
		destroy_mgr(&mgr);
}

void
dns_dispatchmgr_setstats(dns_dispatchmgr_t *mgr, isc_stats_t *stats) {
	REQUIRE(VALID_DISPATCHMGR(mgr));
	REQUIRE(ISC_LIST_EMPTY(mgr->list));
	REQUIRE(mgr->stats == NULL);

	isc_stats_attach(stats, &mgr->stats);
}

static int
port_cmp(const void *key, const void *ent) {
	in_port_t p1 = *(const in_port_t *)key;
	in_port_t p2 = *(const in_port_t *)ent;

	if (p1 < p2)
		return (-1);
	else if (p1 == p2)
		return (0);
	else
		return (1);
}

static isc_boolean_t
portavailable(dns_dispatchmgr_t *mgr, isc_socket_t *sock,
	      isc_sockaddr_t *sockaddrp)
{
	isc_sockaddr_t sockaddr;
	isc_result_t result;
	in_port_t *ports, port;
	unsigned int nports;
	isc_boolean_t available = ISC_FALSE;

	REQUIRE(sock != NULL || sockaddrp != NULL);

	PORTBUFLOCK(mgr);
	if (sock != NULL) {
		sockaddrp = &sockaddr;
		result = isc_socket_getsockname(sock, sockaddrp);
		if (result != ISC_R_SUCCESS)
			goto unlock;
	}

	if (isc_sockaddr_pf(sockaddrp) == AF_INET) {
		ports = mgr->v4ports;
		nports = mgr->nv4ports;
	} else {
		ports = mgr->v6ports;
		nports = mgr->nv6ports;
	}
	if (ports == NULL)
		goto unlock;

	port = isc_sockaddr_getport(sockaddrp);
	if (bsearch(&port, ports, nports, sizeof(in_port_t), port_cmp) != NULL)
		available = ISC_TRUE;

unlock:
	PORTBUFUNLOCK(mgr);
	return (available);
}

#define ATTRMATCH(_a1, _a2, _mask) (((_a1) & (_mask)) == ((_a2) & (_mask)))

static isc_boolean_t
local_addr_match(dns_dispatch_t *disp, isc_sockaddr_t *addr) {
	isc_sockaddr_t sockaddr;
	isc_result_t result;

	REQUIRE(disp->socket != NULL);

	if (addr == NULL)
		return (ISC_TRUE);

	/*
	 * Don't match wildcard ports unless the port is available in the
	 * current configuration.
	 */
	if (isc_sockaddr_getport(addr) == 0 &&
	    isc_sockaddr_getport(&disp->local) == 0 &&
	    !portavailable(disp->mgr, disp->socket, NULL)) {
		return (ISC_FALSE);
	}

	/*
	 * Check if we match the binding <address,port>.
	 * Wildcard ports match/fail here.
	 */
	if (isc_sockaddr_equal(&disp->local, addr))
		return (ISC_TRUE);
	if (isc_sockaddr_getport(addr) == 0)
		return (ISC_FALSE);

	/*
	 * Check if we match a bound wildcard port <address,port>.
	 */
	if (!isc_sockaddr_eqaddr(&disp->local, addr))
		return (ISC_FALSE);
	result = isc_socket_getsockname(disp->socket, &sockaddr);
	if (result != ISC_R_SUCCESS)
		return (ISC_FALSE);

	return (isc_sockaddr_equal(&sockaddr, addr));
}

/*
 * Requires mgr be locked.
 *
 * No dispatcher can be locked by this thread when calling this function.
 *
 *
 * NOTE:
 *	If a matching dispatcher is found, it is locked after this function
 *	returns, and must be unlocked by the caller.
 */
static isc_result_t
dispatch_find(dns_dispatchmgr_t *mgr, isc_sockaddr_t *local,
	      unsigned int attributes, unsigned int mask,
	      dns_dispatch_t **dispp)
{
	dns_dispatch_t *disp;
	isc_result_t result;

	/*
	 * Make certain that we will not match a private or exclusive dispatch.
	 */
	attributes &= ~(DNS_DISPATCHATTR_PRIVATE|DNS_DISPATCHATTR_EXCLUSIVE);
	mask |= (DNS_DISPATCHATTR_PRIVATE|DNS_DISPATCHATTR_EXCLUSIVE);

	disp = ISC_LIST_HEAD(mgr->list);
	while (disp != NULL) {
		LOCK(&disp->lock);
		if ((disp->shutting_down == 0)
		    && ATTRMATCH(disp->attributes, attributes, mask)
		    && local_addr_match(disp, local))
			break;
		UNLOCK(&disp->lock);
		disp = ISC_LIST_NEXT(disp, link);
	}

	if (disp == NULL) {
		result = ISC_R_NOTFOUND;
		goto out;
	}

	*dispp = disp;
	result = ISC_R_SUCCESS;
 out:

	return (result);
}

static isc_result_t
qid_allocate(dns_dispatchmgr_t *mgr, unsigned int buckets,
	     unsigned int increment, dns_qid_t **qidp,
	     isc_boolean_t needsocktable)
{
	dns_qid_t *qid;
	unsigned int i;
	isc_result_t result;

	REQUIRE(VALID_DISPATCHMGR(mgr));
	REQUIRE(buckets < 2097169);  /* next prime > 65536 * 32 */
	REQUIRE(increment > buckets);
	REQUIRE(qidp != NULL && *qidp == NULL);

	qid = isc_mem_get(mgr->mctx, sizeof(*qid));
	if (qid == NULL)
		return (ISC_R_NOMEMORY);

	qid->qid_table = isc_mem_get(mgr->mctx,
				     buckets * sizeof(dns_displist_t));
	if (qid->qid_table == NULL) {
		isc_mem_put(mgr->mctx, qid, sizeof(*qid));
		return (ISC_R_NOMEMORY);
	}

	qid->sock_table = NULL;
	if (needsocktable) {
		qid->sock_table = isc_mem_get(mgr->mctx, buckets *
					      sizeof(dispsocketlist_t));
		if (qid->sock_table == NULL) {
			isc_mem_put(mgr->mctx, qid, sizeof(*qid));
			isc_mem_put(mgr->mctx, qid->qid_table,
				    buckets * sizeof(dns_displist_t));
			return (ISC_R_NOMEMORY);
		}
	}

	result = isc_mutex_init(&qid->lock);
	if (result != ISC_R_SUCCESS) {
		if (qid->sock_table != NULL) {
			isc_mem_put(mgr->mctx, qid->sock_table,
				    buckets * sizeof(dispsocketlist_t));
		}
		isc_mem_put(mgr->mctx, qid->qid_table,
			    buckets * sizeof(dns_displist_t));
		isc_mem_put(mgr->mctx, qid, sizeof(*qid));
		return (result);
	}

	for (i = 0; i < buckets; i++) {
		ISC_LIST_INIT(qid->qid_table[i]);
		if (qid->sock_table != NULL)
			ISC_LIST_INIT(qid->sock_table[i]);
	}

	qid->qid_nbuckets = buckets;
	qid->qid_increment = increment;
	qid->magic = QID_MAGIC;
	*qidp = qid;
	return (ISC_R_SUCCESS);
}

static void
qid_destroy(isc_mem_t *mctx, dns_qid_t **qidp) {
	dns_qid_t *qid;

	REQUIRE(qidp != NULL);
	qid = *qidp;

	REQUIRE(VALID_QID(qid));

	*qidp = NULL;
	qid->magic = 0;
	isc_mem_put(mctx, qid->qid_table,
		    qid->qid_nbuckets * sizeof(dns_displist_t));
	if (qid->sock_table != NULL) {
		isc_mem_put(mctx, qid->sock_table,
			    qid->qid_nbuckets * sizeof(dispsocketlist_t));
	}
	DESTROYLOCK(&qid->lock);
	isc_mem_put(mctx, qid, sizeof(*qid));
}

/*
 * Allocate and set important limits.
 */
static isc_result_t
dispatch_allocate(dns_dispatchmgr_t *mgr, unsigned int maxrequests,
		  dns_dispatch_t **dispp)
{
	dns_dispatch_t *disp;
	isc_result_t result;

	REQUIRE(VALID_DISPATCHMGR(mgr));
	REQUIRE(dispp != NULL && *dispp == NULL);

	/*
	 * Set up the dispatcher, mostly.  Don't bother setting some of
	 * the options that are controlled by tcp vs. udp, etc.
	 */

	disp = isc_mempool_get(mgr->dpool);
	if (disp == NULL)
		return (ISC_R_NOMEMORY);

	disp->magic = 0;
	disp->mgr = mgr;
	disp->maxrequests = maxrequests;
	disp->attributes = 0;
	ISC_LINK_INIT(disp, link);
	disp->refcount = 1;
	disp->recv_pending = 0;
	memset(&disp->local, 0, sizeof(disp->local));
	disp->localport = 0;
	disp->shutting_down = 0;
	disp->shutdown_out = 0;
	disp->connected = 0;
	disp->tcpmsg_valid = 0;
	disp->shutdown_why = ISC_R_UNEXPECTED;
	disp->requests = 0;
	disp->tcpbuffers = 0;
	disp->qid = NULL;
	ISC_LIST_INIT(disp->activesockets);
	ISC_LIST_INIT(disp->inactivesockets);
	disp->nsockets = 0;
	dispatch_initrandom(&disp->arc4ctx, mgr->entropy, NULL);
	disp->port_table = NULL;
	disp->portpool = NULL;

	result = isc_mutex_init(&disp->lock);
	if (result != ISC_R_SUCCESS)
		goto deallocate;

	disp->failsafe_ev = allocate_event(disp);
	if (disp->failsafe_ev == NULL) {
		result = ISC_R_NOMEMORY;
		goto kill_lock;
	}

	disp->magic = DISPATCH_MAGIC;

	*dispp = disp;
	return (ISC_R_SUCCESS);

	/*
	 * error returns
	 */
 kill_lock:
	DESTROYLOCK(&disp->lock);
 deallocate:
	isc_mempool_put(mgr->dpool, disp);

	return (result);
}


/*
 * MUST be unlocked, and not used by anything.
 */
static void
dispatch_free(dns_dispatch_t **dispp)
{
	dns_dispatch_t *disp;
	dns_dispatchmgr_t *mgr;
	int i;

	REQUIRE(VALID_DISPATCH(*dispp));
	disp = *dispp;
	*dispp = NULL;

	mgr = disp->mgr;
	REQUIRE(VALID_DISPATCHMGR(mgr));

	if (disp->tcpmsg_valid) {
		dns_tcpmsg_invalidate(&disp->tcpmsg);
		disp->tcpmsg_valid = 0;
	}

	INSIST(disp->tcpbuffers == 0);
	INSIST(disp->requests == 0);
	INSIST(disp->recv_pending == 0);
	INSIST(ISC_LIST_EMPTY(disp->activesockets));
	INSIST(ISC_LIST_EMPTY(disp->inactivesockets));

	isc_mempool_put(mgr->epool, disp->failsafe_ev);
	disp->failsafe_ev = NULL;

	if (disp->qid != NULL)
		qid_destroy(mgr->mctx, &disp->qid);

	if (disp->port_table != NULL) {
		for (i = 0; i < DNS_DISPATCH_PORTTABLESIZE; i++)
			INSIST(ISC_LIST_EMPTY(disp->port_table[i]));
		isc_mem_put(mgr->mctx, disp->port_table,
			    sizeof(disp->port_table[0]) *
			    DNS_DISPATCH_PORTTABLESIZE);
	}

	if (disp->portpool != NULL)
		isc_mempool_destroy(&disp->portpool);

	disp->mgr = NULL;
	DESTROYLOCK(&disp->lock);
	disp->magic = 0;
	isc_mempool_put(mgr->dpool, disp);
}

isc_result_t
dns_dispatch_createtcp(dns_dispatchmgr_t *mgr, isc_socket_t *sock,
		       isc_taskmgr_t *taskmgr, unsigned int buffersize,
		       unsigned int maxbuffers, unsigned int maxrequests,
		       unsigned int buckets, unsigned int increment,
		       unsigned int attributes, dns_dispatch_t **dispp)
{
	isc_result_t result;
	dns_dispatch_t *disp;

	UNUSED(maxbuffers);
	UNUSED(buffersize);

	REQUIRE(VALID_DISPATCHMGR(mgr));
	REQUIRE(isc_socket_gettype(sock) == isc_sockettype_tcp);
	REQUIRE((attributes & DNS_DISPATCHATTR_TCP) != 0);
	REQUIRE((attributes & DNS_DISPATCHATTR_UDP) == 0);

	attributes |= DNS_DISPATCHATTR_PRIVATE;  /* XXXMLG */

	LOCK(&mgr->lock);

	/*
	 * dispatch_allocate() checks mgr for us.
	 * qid_allocate() checks buckets and increment for us.
	 */
	disp = NULL;
	result = dispatch_allocate(mgr, maxrequests, &disp);
	if (result != ISC_R_SUCCESS) {
		UNLOCK(&mgr->lock);
		return (result);
	}

	result = qid_allocate(mgr, buckets, increment, &disp->qid, ISC_FALSE);
	if (result != ISC_R_SUCCESS)
		goto deallocate_dispatch;

	disp->socktype = isc_sockettype_tcp;
	disp->socket = NULL;
	isc_socket_attach(sock, &disp->socket);

	disp->ntasks = 1;
	disp->task[0] = NULL;
	result = isc_task_create(taskmgr, 0, &disp->task[0]);
	if (result != ISC_R_SUCCESS)
		goto kill_socket;

	disp->ctlevent = isc_event_allocate(mgr->mctx, disp,
					    DNS_EVENT_DISPATCHCONTROL,
					    destroy_disp, disp,
					    sizeof(isc_event_t));
	if (disp->ctlevent == NULL) {
		result = ISC_R_NOMEMORY;
		goto kill_task;
	}

	isc_task_setname(disp->task[0], "tcpdispatch", disp);

	dns_tcpmsg_init(mgr->mctx, disp->socket, &disp->tcpmsg);
	disp->tcpmsg_valid = 1;

	disp->attributes = attributes;

	/*
	 * Append it to the dispatcher list.
	 */
	ISC_LIST_APPEND(mgr->list, disp, link);
	UNLOCK(&mgr->lock);

	mgr_log(mgr, LVL(90), "created TCP dispatcher %p", disp);
	dispatch_log(disp, LVL(90), "created task %p", disp->task[0]);

	*dispp = disp;

	return (ISC_R_SUCCESS);

	/*
	 * Error returns.
	 */
 kill_task:
	isc_task_detach(&disp->task[0]);
 kill_socket:
	isc_socket_detach(&disp->socket);
 deallocate_dispatch:
	dispatch_free(&disp);

	UNLOCK(&mgr->lock);

	return (result);
}

isc_result_t
dns_dispatch_getudp(dns_dispatchmgr_t *mgr, isc_socketmgr_t *sockmgr,
		    isc_taskmgr_t *taskmgr, isc_sockaddr_t *localaddr,
		    unsigned int buffersize,
		    unsigned int maxbuffers, unsigned int maxrequests,
		    unsigned int buckets, unsigned int increment,
		    unsigned int attributes, unsigned int mask,
		    dns_dispatch_t **dispp)
{
	isc_result_t result;
	dns_dispatch_t *disp = NULL;

	REQUIRE(VALID_DISPATCHMGR(mgr));
	REQUIRE(sockmgr != NULL);
	REQUIRE(localaddr != NULL);
	REQUIRE(taskmgr != NULL);
	REQUIRE(buffersize >= 512 && buffersize < (64 * 1024));
	REQUIRE(maxbuffers > 0);
	REQUIRE(buckets < 2097169);  /* next prime > 65536 * 32 */
	REQUIRE(increment > buckets);
	REQUIRE(dispp != NULL && *dispp == NULL);
	REQUIRE((attributes & DNS_DISPATCHATTR_TCP) == 0);

	result = dns_dispatchmgr_setudp(mgr, buffersize, maxbuffers,
					maxrequests, buckets, increment);
	if (result != ISC_R_SUCCESS)
		return (result);

	LOCK(&mgr->lock);

	if ((attributes & DNS_DISPATCHATTR_EXCLUSIVE) != 0) {
		REQUIRE(isc_sockaddr_getport(localaddr) == 0);
		goto createudp;
	}

	/*
	 * See if we have a dispatcher that matches.
	 */
	result = dispatch_find(mgr, localaddr, attributes, mask, &disp);
	if (result == ISC_R_SUCCESS) {
		disp->refcount++;

		if (disp->maxrequests < maxrequests)
			disp->maxrequests = maxrequests;

		if ((disp->attributes & DNS_DISPATCHATTR_NOLISTEN) == 0 &&
		    (attributes & DNS_DISPATCHATTR_NOLISTEN) != 0)
		{
			disp->attributes |= DNS_DISPATCHATTR_NOLISTEN;
			if (disp->recv_pending != 0)
				isc_socket_cancel(disp->socket, disp->task[0],
						  ISC_SOCKCANCEL_RECV);
		}

		UNLOCK(&disp->lock);
		UNLOCK(&mgr->lock);

		*dispp = disp;

		return (ISC_R_SUCCESS);
	}

 createudp:
	/*
	 * Nope, create one.
	 */
	result = dispatch_createudp(mgr, sockmgr, taskmgr, localaddr,
				    maxrequests, attributes, &disp);
	if (result != ISC_R_SUCCESS) {
		UNLOCK(&mgr->lock);
		return (result);
	}

	UNLOCK(&mgr->lock);
	*dispp = disp;
	return (ISC_R_SUCCESS);
}

/*
 * mgr should be locked.
 */

#ifndef DNS_DISPATCH_HELD
#define DNS_DISPATCH_HELD 20U
#endif

static isc_result_t
get_udpsocket(dns_dispatchmgr_t *mgr, dns_dispatch_t *disp,
	      isc_socketmgr_t *sockmgr, isc_sockaddr_t *localaddr,
	      isc_socket_t **sockp)
{
	unsigned int i, j;
	isc_socket_t *held[DNS_DISPATCH_HELD];
	isc_sockaddr_t localaddr_bound;
	isc_socket_t *sock = NULL;
	isc_result_t result = ISC_R_SUCCESS;
	isc_boolean_t anyport;

	INSIST(sockp != NULL && *sockp == NULL);

	localaddr_bound = *localaddr;
	anyport = ISC_TF(isc_sockaddr_getport(localaddr) == 0);

	if (anyport) {
		unsigned int nports;
		in_port_t *ports;

		/*
		 * If no port is specified, we first try to pick up a random
		 * port by ourselves.
		 */
		if (isc_sockaddr_pf(&disp->local) == AF_INET) {
			nports = disp->mgr->nv4ports;
			ports = disp->mgr->v4ports;
		} else {
			nports = disp->mgr->nv6ports;
			ports = disp->mgr->v6ports;
		}
		if (nports == 0)
			return (ISC_R_ADDRNOTAVAIL);

		for (i = 0; i < 1024; i++) {
			in_port_t prt;

			prt = ports[dispatch_uniformrandom(
					DISP_ARC4CTX(disp),
					nports)];
			isc_sockaddr_setport(&localaddr_bound, prt);
			result = open_socket(sockmgr, &localaddr_bound,
					     0, &sock);
			if (result == ISC_R_SUCCESS ||
			    result != ISC_R_ADDRINUSE) {
				disp->localport = prt;
				*sockp = sock;
				return (result);
			}
		}

		/*
		 * If this fails 1024 times, we then ask the kernel for
		 * choosing one.
		 */
	} else {
		/* Allow to reuse address for non-random ports. */
		result = open_socket(sockmgr, localaddr,
				     ISC_SOCKET_REUSEADDRESS, &sock);

		if (result == ISC_R_SUCCESS)
			*sockp = sock;

		return (result);
	}

	memset(held, 0, sizeof(held));
	i = 0;

	for (j = 0; j < 0xffffU; j++) {
		result = open_socket(sockmgr, localaddr, 0, &sock);
		if (result != ISC_R_SUCCESS)
			goto end;
		else if (!anyport)
			break;
		else if (portavailable(mgr, sock, NULL))
			break;
		if (held[i] != NULL)
			isc_socket_detach(&held[i]);
		held[i++] = sock;
		sock = NULL;
		if (i == DNS_DISPATCH_HELD)
			i = 0;
	}
	if (j == 0xffffU) {
		mgr_log(mgr, ISC_LOG_ERROR,
			"avoid-v%s-udp-ports: unable to allocate "
			"an available port",
			isc_sockaddr_pf(localaddr) == AF_INET ? "4" : "6");
		result = ISC_R_FAILURE;
		goto end;
	}
	*sockp = sock;

end:
	for (i = 0; i < DNS_DISPATCH_HELD; i++) {
		if (held[i] != NULL)
			isc_socket_detach(&held[i]);
	}

	return (result);
}

static isc_result_t
dispatch_createudp(dns_dispatchmgr_t *mgr, isc_socketmgr_t *sockmgr,
		   isc_taskmgr_t *taskmgr,
		   isc_sockaddr_t *localaddr,
		   unsigned int maxrequests,
		   unsigned int attributes,
		   dns_dispatch_t **dispp)
{
	isc_result_t result;
	dns_dispatch_t *disp;
	isc_socket_t *sock = NULL;
	int i = 0;

	/*
	 * dispatch_allocate() checks mgr for us.
	 */
	disp = NULL;
	result = dispatch_allocate(mgr, maxrequests, &disp);
	if (result != ISC_R_SUCCESS)
		return (result);

	if ((attributes & DNS_DISPATCHATTR_EXCLUSIVE) == 0) {
		result = get_udpsocket(mgr, disp, sockmgr, localaddr, &sock);
		if (result != ISC_R_SUCCESS)
			goto deallocate_dispatch;
	} else {
		isc_sockaddr_t sa_any;

		/*
		 * For dispatches using exclusive sockets with a specific
		 * source address, we only check if the specified address is
		 * available on the system.  Query sockets will be created later
		 * on demand.
		 */
		isc_sockaddr_anyofpf(&sa_any, isc_sockaddr_pf(localaddr));
		if (!isc_sockaddr_eqaddr(&sa_any, localaddr)) {
			result = open_socket(sockmgr, localaddr, 0, &sock);
			if (sock != NULL)
				isc_socket_detach(&sock);
			if (result != ISC_R_SUCCESS)
				goto deallocate_dispatch;
		}

		disp->port_table = isc_mem_get(mgr->mctx,
					       sizeof(disp->port_table[0]) *
					       DNS_DISPATCH_PORTTABLESIZE);
		if (disp->port_table == NULL)
			goto deallocate_dispatch;
		for (i = 0; i < DNS_DISPATCH_PORTTABLESIZE; i++)
			ISC_LIST_INIT(disp->port_table[i]);

		result = isc_mempool_create(mgr->mctx, sizeof(dispportentry_t),
					    &disp->portpool);
		if (result != ISC_R_SUCCESS)
			goto deallocate_dispatch;
		isc_mempool_setname(disp->portpool, "disp_portpool");
		isc_mempool_setfreemax(disp->portpool, 128);
	}
	disp->socktype = isc_sockettype_udp;
	disp->socket = sock;
	disp->local = *localaddr;

	if ((attributes & DNS_DISPATCHATTR_EXCLUSIVE) != 0)
		disp->ntasks = MAX_INTERNAL_TASKS;
	else
		disp->ntasks = 1;
	for (i = 0; i < disp->ntasks; i++) {
		disp->task[i] = NULL;
		result = isc_task_create(taskmgr, 0, &disp->task[i]);
		if (result != ISC_R_SUCCESS) {
			while (--i >= 0) {
				isc_task_shutdown(disp->task[i]);
				isc_task_detach(&disp->task[i]);
			}
			goto kill_socket;
		}
		isc_task_setname(disp->task[i], "udpdispatch", disp);
	}

	disp->ctlevent = isc_event_allocate(mgr->mctx, disp,
					    DNS_EVENT_DISPATCHCONTROL,
					    destroy_disp, disp,
					    sizeof(isc_event_t));
	if (disp->ctlevent == NULL) {
		result = ISC_R_NOMEMORY;
		goto kill_task;
	}

	attributes &= ~DNS_DISPATCHATTR_TCP;
	attributes |= DNS_DISPATCHATTR_UDP;
	disp->attributes = attributes;

	/*
	 * Append it to the dispatcher list.
	 */
	ISC_LIST_APPEND(mgr->list, disp, link);

	mgr_log(mgr, LVL(90), "created UDP dispatcher %p", disp);
	dispatch_log(disp, LVL(90), "created task %p", disp->task[0]); /* XXX */
	if (disp->socket != NULL)
		dispatch_log(disp, LVL(90), "created socket %p", disp->socket);

	*dispp = disp;
	return (result);

	/*
	 * Error returns.
	 */
 kill_task:
	for (i = 0; i < disp->ntasks; i++)
		isc_task_detach(&disp->task[i]);
 kill_socket:
	if (disp->socket != NULL)
		isc_socket_detach(&disp->socket);
 deallocate_dispatch:
	dispatch_free(&disp);

	return (result);
}

void
dns_dispatch_attach(dns_dispatch_t *disp, dns_dispatch_t **dispp) {
	REQUIRE(VALID_DISPATCH(disp));
	REQUIRE(dispp != NULL && *dispp == NULL);

	LOCK(&disp->lock);
	disp->refcount++;
	UNLOCK(&disp->lock);

	*dispp = disp;
}

/*
 * It is important to lock the manager while we are deleting the dispatch,
 * since dns_dispatch_getudp will call dispatch_find, which returns to
 * the caller a dispatch but does not attach to it until later.  _getudp
 * locks the manager, however, so locking it here will keep us from attaching
 * to a dispatcher that is in the process of going away.
 */
void
dns_dispatch_detach(dns_dispatch_t **dispp) {
	dns_dispatch_t *disp;
	dispsocket_t *dispsock;
	isc_boolean_t killit;

	REQUIRE(dispp != NULL && VALID_DISPATCH(*dispp));

	disp = *dispp;
	*dispp = NULL;

	LOCK(&disp->lock);

	INSIST(disp->refcount > 0);
	disp->refcount--;
	killit = ISC_FALSE;
	if (disp->refcount == 0) {
		if (disp->recv_pending > 0)
			isc_socket_cancel(disp->socket, disp->task[0],
					  ISC_SOCKCANCEL_RECV);
		for (dispsock = ISC_LIST_HEAD(disp->activesockets);
		     dispsock != NULL;
		     dispsock = ISC_LIST_NEXT(dispsock, link)) {
			isc_socket_cancel(dispsock->socket, dispsock->task,
					  ISC_SOCKCANCEL_RECV);
		}
		disp->shutting_down = 1;
	}

	dispatch_log(disp, LVL(90), "detach: refcount %d", disp->refcount);

	killit = destroy_disp_ok(disp);
	UNLOCK(&disp->lock);
	if (killit)
		isc_task_send(disp->task[0], &disp->ctlevent);
}

isc_result_t
dns_dispatch_addresponse2(dns_dispatch_t *disp, isc_sockaddr_t *dest,
			  isc_task_t *task, isc_taskaction_t action, void *arg,
			  dns_messageid_t *idp, dns_dispentry_t **resp,
			  isc_socketmgr_t *sockmgr)
{
	dns_dispentry_t *res;
	unsigned int bucket;
	in_port_t localport = 0;
	dns_messageid_t id;
	int i;
	isc_boolean_t ok;
	dns_qid_t *qid;
	dispsocket_t *dispsocket = NULL;
	isc_result_t result;

	REQUIRE(VALID_DISPATCH(disp));
	REQUIRE(task != NULL);
	REQUIRE(dest != NULL);
	REQUIRE(resp != NULL && *resp == NULL);
	REQUIRE(idp != NULL);
	if ((disp->attributes & DNS_DISPATCHATTR_EXCLUSIVE) != 0)
		REQUIRE(sockmgr != NULL);

	LOCK(&disp->lock);

	if (disp->shutting_down == 1) {
		UNLOCK(&disp->lock);
		return (ISC_R_SHUTTINGDOWN);
	}

	if (disp->requests >= disp->maxrequests) {
		UNLOCK(&disp->lock);
		return (ISC_R_QUOTA);
	}

	if ((disp->attributes & DNS_DISPATCHATTR_EXCLUSIVE) != 0 &&
	    disp->nsockets > DNS_DISPATCH_SOCKSQUOTA) {
		dispsocket_t *oldestsocket;
		dns_dispentry_t *oldestresp;
		dns_dispatchevent_t *rev;

		/*
		 * Kill oldest outstanding query if the number of sockets
		 * exceeds the quota to keep the room for new queries.
		 */
		oldestsocket = ISC_LIST_HEAD(disp->activesockets);
		oldestresp = oldestsocket->resp;
		if (oldestresp != NULL && !oldestresp->item_out) {
			rev = allocate_event(oldestresp->disp);
			if (rev != NULL) {
				rev->buffer.base = NULL;
				rev->result = ISC_R_CANCELED;
				rev->id = oldestresp->id;
				ISC_EVENT_INIT(rev, sizeof(*rev), 0,
					       NULL, DNS_EVENT_DISPATCH,
					       oldestresp->action,
					       oldestresp->arg, oldestresp,
					       NULL, NULL);
				oldestresp->item_out = ISC_TRUE;
				isc_task_send(oldestresp->task,
					      ISC_EVENT_PTR(&rev));
				inc_stats(disp->mgr,
					  dns_resstatscounter_dispabort);
			}
		}

		/*
		 * Move this entry to the tail so that it won't (easily) be
		 * examined before actually being canceled.
		 */
		ISC_LIST_UNLINK(disp->activesockets, oldestsocket, link);
		ISC_LIST_APPEND(disp->activesockets, oldestsocket, link);
	}

	qid = DNS_QID(disp);
	LOCK(&qid->lock);

	if ((disp->attributes & DNS_DISPATCHATTR_EXCLUSIVE) != 0) {
		/*
		 * Get a separate UDP socket with a random port number.
		 */
		result = get_dispsocket(disp, dest, sockmgr, qid, &dispsocket,
					&localport);
		if (result != ISC_R_SUCCESS) {
			UNLOCK(&qid->lock);
			UNLOCK(&disp->lock);
			inc_stats(disp->mgr, dns_resstatscounter_dispsockfail);
			return (result);
		}
	} else {
		localport = disp->localport;
	}

	/*
	 * Try somewhat hard to find an unique ID.
	 */
	id = (dns_messageid_t)dispatch_random(DISP_ARC4CTX(disp));
	bucket = dns_hash(qid, dest, id, localport);
	ok = ISC_FALSE;
	for (i = 0; i < 64; i++) {
		if (entry_search(qid, dest, id, localport, bucket) == NULL) {
			ok = ISC_TRUE;
			break;
		}
		id += qid->qid_increment;
		id &= 0x0000ffff;
		bucket = dns_hash(qid, dest, id, localport);
	}

	if (!ok) {
		UNLOCK(&qid->lock);
		UNLOCK(&disp->lock);
		return (ISC_R_NOMORE);
	}

	res = isc_mempool_get(disp->mgr->rpool);
	if (res == NULL) {
		UNLOCK(&qid->lock);
		UNLOCK(&disp->lock);
		if (dispsocket != NULL)
			destroy_dispsocket(disp, &dispsocket);
		return (ISC_R_NOMEMORY);
	}

	disp->refcount++;
	disp->requests++;
	res->task = NULL;
	isc_task_attach(task, &res->task);
	res->disp = disp;
	res->id = id;
	res->port = localport;
	res->bucket = bucket;
	res->host = *dest;
	res->action = action;
	res->arg = arg;
	res->dispsocket = dispsocket;
	if (dispsocket != NULL)
		dispsocket->resp = res;
	res->item_out = ISC_FALSE;
	ISC_LIST_INIT(res->items);
	ISC_LINK_INIT(res, link);
	res->magic = RESPONSE_MAGIC;
	ISC_LIST_APPEND(qid->qid_table[bucket], res, link);
	UNLOCK(&qid->lock);

	request_log(disp, res, LVL(90),
		    "attached to task %p", res->task);

	if (((disp->attributes & DNS_DISPATCHATTR_UDP) != 0) ||
	    ((disp->attributes & DNS_DISPATCHATTR_CONNECTED) != 0)) {
		result = startrecv(disp, dispsocket);
		if (result != ISC_R_SUCCESS) {
			LOCK(&qid->lock);
			ISC_LIST_UNLINK(qid->qid_table[bucket], res, link);
			UNLOCK(&qid->lock);

			if (dispsocket != NULL)
				destroy_dispsocket(disp, &dispsocket);

			disp->refcount--;
			disp->requests--;

			UNLOCK(&disp->lock);
			isc_task_detach(&res->task);
			isc_mempool_put(disp->mgr->rpool, res);
			return (result);
		}
	}

	if (dispsocket != NULL)
		ISC_LIST_APPEND(disp->activesockets, dispsocket, link);

	UNLOCK(&disp->lock);

	*idp = id;
	*resp = res;

	if ((disp->attributes & DNS_DISPATCHATTR_EXCLUSIVE) != 0)
		INSIST(res->dispsocket != NULL);

	return (ISC_R_SUCCESS);
}

isc_result_t
dns_dispatch_addresponse(dns_dispatch_t *disp, isc_sockaddr_t *dest,
			 isc_task_t *task, isc_taskaction_t action, void *arg,
			 dns_messageid_t *idp, dns_dispentry_t **resp)
{
	REQUIRE(VALID_DISPATCH(disp));
	REQUIRE((disp->attributes & DNS_DISPATCHATTR_EXCLUSIVE) == 0);

	return (dns_dispatch_addresponse2(disp, dest, task, action, arg,
					  idp, resp, NULL));
}

void
dns_dispatch_starttcp(dns_dispatch_t *disp) {

	REQUIRE(VALID_DISPATCH(disp));

	dispatch_log(disp, LVL(90), "starttcp %p", disp->task[0]);

	LOCK(&disp->lock);
	disp->attributes |= DNS_DISPATCHATTR_CONNECTED;
	(void)startrecv(disp, NULL);
	UNLOCK(&disp->lock);
}

void
dns_dispatch_removeresponse(dns_dispentry_t **resp,
			    dns_dispatchevent_t **sockevent)
{
	dns_dispatchmgr_t *mgr;
	dns_dispatch_t *disp;
	dns_dispentry_t *res;
	dispsocket_t *dispsock;
	dns_dispatchevent_t *ev;
	unsigned int bucket;
	isc_boolean_t killit;
	unsigned int n;
	isc_eventlist_t events;
	dns_qid_t *qid;

	REQUIRE(resp != NULL);
	REQUIRE(VALID_RESPONSE(*resp));

	res = *resp;
	*resp = NULL;

	disp = res->disp;
	REQUIRE(VALID_DISPATCH(disp));
	mgr = disp->mgr;
	REQUIRE(VALID_DISPATCHMGR(mgr));

	qid = DNS_QID(disp);

	if (sockevent != NULL) {
		REQUIRE(*sockevent != NULL);
		ev = *sockevent;
		*sockevent = NULL;
	} else {
		ev = NULL;
	}

	LOCK(&disp->lock);

	INSIST(disp->requests > 0);
	disp->requests--;
	INSIST(disp->refcount > 0);
	disp->refcount--;
	killit = ISC_FALSE;
	if (disp->refcount == 0) {
		if (disp->recv_pending > 0)
			isc_socket_cancel(disp->socket, disp->task[0],
					  ISC_SOCKCANCEL_RECV);
		for (dispsock = ISC_LIST_HEAD(disp->activesockets);
		     dispsock != NULL;
		     dispsock = ISC_LIST_NEXT(dispsock, link)) {
			isc_socket_cancel(dispsock->socket, dispsock->task,
					  ISC_SOCKCANCEL_RECV);
		}
		disp->shutting_down = 1;
	}

	bucket = res->bucket;

	LOCK(&qid->lock);
	ISC_LIST_UNLINK(qid->qid_table[bucket], res, link);
	UNLOCK(&qid->lock);

	if (ev == NULL && res->item_out) {
		/*
		 * We've posted our event, but the caller hasn't gotten it
		 * yet.  Take it back.
		 */
		ISC_LIST_INIT(events);
		n = isc_task_unsend(res->task, res, DNS_EVENT_DISPATCH,
				    NULL, &events);
		/*
		 * We had better have gotten it back.
		 */
		INSIST(n == 1);
		ev = (dns_dispatchevent_t *)ISC_LIST_HEAD(events);
	}

	if (ev != NULL) {
		REQUIRE(res->item_out == ISC_TRUE);
		res->item_out = ISC_FALSE;
		if (ev->buffer.base != NULL)
			free_buffer(disp, ev->buffer.base, ev->buffer.length);
		free_event(disp, ev);
	}

	request_log(disp, res, LVL(90), "detaching from task %p", res->task);
	isc_task_detach(&res->task);

	if (res->dispsocket != NULL) {
		isc_socket_cancel(res->dispsocket->socket,
				  res->dispsocket->task, ISC_SOCKCANCEL_RECV);
		res->dispsocket->resp = NULL;
	}

	/*
	 * Free any buffered requests as well
	 */
	ev = ISC_LIST_HEAD(res->items);
	while (ev != NULL) {
		ISC_LIST_UNLINK(res->items, ev, ev_link);
		if (ev->buffer.base != NULL)
			free_buffer(disp, ev->buffer.base, ev->buffer.length);
		free_event(disp, ev);
		ev = ISC_LIST_HEAD(res->items);
	}
	res->magic = 0;
	isc_mempool_put(disp->mgr->rpool, res);
	if (disp->shutting_down == 1)
		do_cancel(disp);
	else
		(void)startrecv(disp, NULL);

	killit = destroy_disp_ok(disp);
	UNLOCK(&disp->lock);
	if (killit)
		isc_task_send(disp->task[0], &disp->ctlevent);
}

static void
do_cancel(dns_dispatch_t *disp) {
	dns_dispatchevent_t *ev;
	dns_dispentry_t *resp;
	dns_qid_t *qid;

	if (disp->shutdown_out == 1)
		return;

	qid = DNS_QID(disp);

	/*
	 * Search for the first response handler without packets outstanding
	 * unless a specific hander is given.
	 */
	LOCK(&qid->lock);
	for (resp = linear_first(qid);
	     resp != NULL && resp->item_out;
	     /* Empty. */)
		resp = linear_next(qid, resp);

	/*
	 * No one to send the cancel event to, so nothing to do.
	 */
	if (resp == NULL)
		goto unlock;

	/*
	 * Send the shutdown failsafe event to this resp.
	 */
	ev = disp->failsafe_ev;
	ISC_EVENT_INIT(ev, sizeof(*ev), 0, NULL, DNS_EVENT_DISPATCH,
		       resp->action, resp->arg, resp, NULL, NULL);
	ev->result = disp->shutdown_why;
	ev->buffer.base = NULL;
	ev->buffer.length = 0;
	disp->shutdown_out = 1;
	request_log(disp, resp, LVL(10),
		    "cancel: failsafe event %p -> task %p",
		    ev, resp->task);
	resp->item_out = ISC_TRUE;
	isc_task_send(resp->task, ISC_EVENT_PTR(&ev));
 unlock:
	UNLOCK(&qid->lock);
}

isc_socket_t *
dns_dispatch_getsocket(dns_dispatch_t *disp) {
	REQUIRE(VALID_DISPATCH(disp));

	return (disp->socket);
}

isc_socket_t *
dns_dispatch_getentrysocket(dns_dispentry_t *resp) {
	REQUIRE(VALID_RESPONSE(resp));

	if (resp->dispsocket != NULL)
		return (resp->dispsocket->socket);
	else
		return (NULL);
}

isc_result_t
dns_dispatch_getlocaladdress(dns_dispatch_t *disp, isc_sockaddr_t *addrp) {

	REQUIRE(VALID_DISPATCH(disp));
	REQUIRE(addrp != NULL);

	if (disp->socktype == isc_sockettype_udp) {
		*addrp = disp->local;
		return (ISC_R_SUCCESS);
	}
	return (ISC_R_NOTIMPLEMENTED);
}

void
dns_dispatch_cancel(dns_dispatch_t *disp) {
	REQUIRE(VALID_DISPATCH(disp));

	LOCK(&disp->lock);

	if (disp->shutting_down == 1) {
		UNLOCK(&disp->lock);
		return;
	}

	disp->shutdown_why = ISC_R_CANCELED;
	disp->shutting_down = 1;
	do_cancel(disp);

	UNLOCK(&disp->lock);

	return;
}

unsigned int
dns_dispatch_getattributes(dns_dispatch_t *disp) {
	REQUIRE(VALID_DISPATCH(disp));

	/*
	 * We don't bother locking disp here; it's the caller's responsibility
	 * to use only non volatile flags.
	 */
	return (disp->attributes);
}

void
dns_dispatch_changeattributes(dns_dispatch_t *disp,
			      unsigned int attributes, unsigned int mask)
{
	REQUIRE(VALID_DISPATCH(disp));
	/* Exclusive attribute can only be set on creation */
	REQUIRE((attributes & DNS_DISPATCHATTR_EXCLUSIVE) == 0);
	/* Also, a dispatch with randomport specified cannot start listening */
	REQUIRE((disp->attributes & DNS_DISPATCHATTR_EXCLUSIVE) == 0 ||
		(attributes & DNS_DISPATCHATTR_NOLISTEN) == 0);

	/* XXXMLG
	 * Should check for valid attributes here!
	 */

	LOCK(&disp->lock);

	if ((mask & DNS_DISPATCHATTR_NOLISTEN) != 0) {
		if ((disp->attributes & DNS_DISPATCHATTR_NOLISTEN) != 0 &&
		    (attributes & DNS_DISPATCHATTR_NOLISTEN) == 0) {
			disp->attributes &= ~DNS_DISPATCHATTR_NOLISTEN;
			(void)startrecv(disp, NULL);
		} else if ((disp->attributes & DNS_DISPATCHATTR_NOLISTEN)
			   == 0 &&
			   (attributes & DNS_DISPATCHATTR_NOLISTEN) != 0) {
			disp->attributes |= DNS_DISPATCHATTR_NOLISTEN;
			if (disp->recv_pending != 0)
				isc_socket_cancel(disp->socket, disp->task[0],
						  ISC_SOCKCANCEL_RECV);
		}
	}

	disp->attributes &= ~mask;
	disp->attributes |= (attributes & mask);
	UNLOCK(&disp->lock);
}

void
dns_dispatch_importrecv(dns_dispatch_t *disp, isc_event_t *event) {
	void *buf;
	isc_socketevent_t *sevent, *newsevent;

	REQUIRE(VALID_DISPATCH(disp));
	REQUIRE((disp->attributes & DNS_DISPATCHATTR_NOLISTEN) != 0);
	REQUIRE(event != NULL);

	sevent = (isc_socketevent_t *)event;

	INSIST(sevent->n <= disp->mgr->buffersize);
	newsevent = (isc_socketevent_t *)
		    isc_event_allocate(disp->mgr->mctx, NULL,
				      DNS_EVENT_IMPORTRECVDONE, udp_shrecv,
				      disp, sizeof(isc_socketevent_t));
	if (newsevent == NULL)
		return;

	buf = allocate_udp_buffer(disp);
	if (buf == NULL) {
		isc_event_free(ISC_EVENT_PTR(&newsevent));
		return;
	}
	memcpy(buf, sevent->region.base, sevent->n);
	newsevent->region.base = buf;
	newsevent->region.length = disp->mgr->buffersize;
	newsevent->n = sevent->n;
	newsevent->result = sevent->result;
	newsevent->address = sevent->address;
	newsevent->timestamp = sevent->timestamp;
	newsevent->pktinfo = sevent->pktinfo;
	newsevent->attributes = sevent->attributes;

	isc_task_send(disp->task[0], ISC_EVENT_PTR(&newsevent));
}

#if 0
void
dns_dispatchmgr_dump(dns_dispatchmgr_t *mgr) {
	dns_dispatch_t *disp;
	char foo[1024];

	disp = ISC_LIST_HEAD(mgr->list);
	while (disp != NULL) {
		isc_sockaddr_format(&disp->local, foo, sizeof(foo));
		printf("\tdispatch %p, addr %s\n", disp, foo);
		disp = ISC_LIST_NEXT(disp, link);
	}
}
#endif