Training courses

Kernel and Embedded Linux

Bootlin training courses

Embedded Linux, kernel,
Yocto Project, Buildroot, real-time,
graphics, boot time, debugging...

Bootlin logo

Elixir Cross Referencer

   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
3530
3531
3532
3533
3534
3535
3536
3537
3538
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
3552
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
3566
3567
3568
3569
3570
3571
3572
3573
3574
3575
3576
3577
3578
3579
3580
3581
3582
3583
3584
3585
3586
3587
3588
3589
3590
3591
3592
3593
3594
3595
3596
3597
3598
3599
3600
3601
3602
3603
3604
3605
3606
3607
3608
3609
3610
3611
3612
3613
3614
3615
3616
3617
3618
3619
3620
3621
3622
3623
3624
3625
3626
3627
3628
3629
3630
3631
3632
3633
3634
3635
3636
3637
3638
3639
3640
3641
3642
3643
3644
3645
3646
3647
3648
3649
3650
3651
3652
3653
3654
3655
3656
3657
3658
3659
3660
3661
3662
3663
3664
3665
3666
3667
3668
3669
3670
3671
3672
3673
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
3686
3687
3688
3689
3690
3691
3692
3693
3694
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
3706
3707
3708
3709
3710
3711
3712
3713
3714
3715
3716
3717
3718
3719
3720
3721
3722
3723
3724
3725
3726
3727
3728
3729
3730
3731
3732
3733
3734
3735
3736
3737
3738
3739
3740
3741
3742
3743
3744
3745
3746
3747
3748
3749
3750
3751
3752
3753
3754
3755
3756
3757
3758
3759
3760
3761
3762
3763
3764
3765
3766
3767
3768
3769
3770
3771
3772
3773
3774
3775
3776
3777
3778
3779
3780
3781
3782
3783
3784
3785
3786
3787
3788
3789
3790
3791
3792
3793
3794
3795
3796
3797
3798
3799
3800
3801
3802
3803
3804
3805
3806
3807
3808
3809
3810
3811
3812
3813
3814
3815
3816
3817
3818
3819
3820
3821
3822
3823
3824
3825
3826
3827
3828
3829
3830
3831
3832
3833
3834
3835
3836
3837
3838
3839
3840
3841
3842
3843
3844
3845
3846
3847
3848
3849
3850
3851
3852
3853
3854
3855
3856
3857
3858
3859
3860
3861
3862
3863
3864
3865
3866
3867
3868
3869
3870
3871
3872
3873
3874
3875
3876
3877
3878
3879
3880
3881
3882
3883
3884
3885
3886
3887
3888
3889
3890
3891
3892
3893
3894
3895
3896
3897
3898
3899
3900
3901
3902
3903
3904
3905
3906
3907
3908
3909
3910
3911
3912
3913
3914
3915
3916
3917
3918
3919
3920
3921
3922
3923
3924
3925
3926
3927
3928
3929
3930
3931
3932
3933
3934
3935
3936
3937
3938
3939
3940
3941
3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
3953
3954
3955
3956
3957
3958
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
3969
3970
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
3981
3982
3983
3984
3985
3986
3987
3988
3989
3990
3991
3992
3993
3994
3995
3996
3997
3998
3999
4000
4001
4002
4003
4004
4005
4006
4007
4008
4009
4010
4011
4012
4013
4014
4015
4016
4017
4018
4019
4020
4021
4022
4023
4024
4025
4026
4027
4028
4029
4030
4031
4032
4033
4034
4035
4036
4037
4038
4039
4040
4041
4042
4043
4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
4068
4069
4070
4071
4072
4073
4074
4075
4076
4077
4078
4079
4080
4081
4082
4083
4084
4085
4086
4087
4088
4089
4090
4091
4092
4093
4094
4095
4096
4097
4098
4099
4100
4101
4102
4103
4104
4105
4106
4107
4108
4109
4110
4111
4112
4113
4114
4115
4116
4117
4118
4119
4120
4121
4122
4123
4124
4125
4126
4127
4128
4129
4130
4131
4132
4133
4134
4135
4136
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
4162
4163
4164
4165
4166
4167
4168
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
4179
4180
4181
4182
4183
4184
4185
4186
4187
4188
4189
4190
4191
4192
4193
4194
4195
4196
4197
4198
4199
4200
4201
4202
4203
4204
4205
4206
4207
4208
4209
4210
4211
4212
4213
4214
4215
4216
4217
4218
4219
4220
4221
4222
4223
4224
4225
4226
4227
4228
4229
4230
4231
4232
4233
4234
4235
4236
4237
4238
4239
4240
4241
4242
4243
4244
4245
4246
4247
4248
4249
4250
4251
4252
4253
4254
4255
4256
4257
4258
4259
4260
4261
4262
4263
4264
4265
4266
4267
4268
4269
4270
4271
4272
4273
4274
4275
4276
4277
4278
4279
4280
4281
4282
4283
4284
4285
4286
4287
4288
4289
4290
4291
4292
4293
4294
4295
4296
4297
4298
4299
4300
4301
4302
4303
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
4315
4316
4317
4318
4319
4320
4321
4322
4323
4324
4325
4326
4327
4328
4329
4330
4331
4332
4333
4334
4335
4336
4337
4338
4339
4340
4341
4342
4343
4344
4345
4346
4347
4348
4349
4350
4351
4352
4353
4354
4355
4356
4357
4358
4359
4360
4361
4362
4363
4364
4365
4366
4367
4368
4369
4370
4371
4372
4373
4374
4375
4376
4377
4378
4379
4380
4381
4382
4383
4384
4385
4386
4387
4388
4389
4390
4391
4392
4393
4394
4395
4396
4397
4398
4399
4400
4401
4402
4403
4404
4405
4406
4407
4408
4409
4410
4411
4412
4413
4414
4415
4416
4417
4418
4419
4420
4421
4422
4423
4424
4425
4426
4427
4428
4429
4430
4431
4432
4433
4434
4435
4436
4437
4438
4439
4440
4441
4442
4443
4444
4445
4446
4447
4448
4449
4450
4451
4452
4453
4454
4455
4456
4457
4458
4459
4460
4461
4462
4463
4464
4465
4466
4467
4468
4469
4470
4471
4472
4473
4474
4475
4476
4477
4478
4479
4480
4481
4482
4483
4484
4485
4486
4487
4488
4489
4490
4491
4492
4493
4494
4495
4496
4497
4498
4499
4500
4501
4502
4503
4504
4505
4506
4507
4508
4509
4510
4511
4512
4513
4514
4515
4516
4517
4518
4519
4520
4521
4522
4523
4524
4525
4526
4527
4528
4529
4530
4531
4532
4533
4534
4535
4536
4537
4538
4539
4540
4541
4542
4543
4544
4545
4546
4547
4548
4549
4550
4551
4552
4553
4554
4555
4556
4557
4558
4559
4560
4561
4562
4563
4564
4565
4566
4567
4568
4569
4570
4571
4572
4573
4574
4575
4576
4577
4578
4579
4580
4581
4582
4583
4584
4585
4586
4587
4588
4589
4590
4591
4592
4593
4594
4595
4596
4597
4598
4599
4600
4601
4602
4603
4604
4605
4606
4607
4608
4609
4610
4611
4612
4613
4614
4615
4616
4617
4618
4619
4620
4621
4622
4623
4624
4625
4626
4627
4628
4629
4630
4631
4632
4633
4634
4635
4636
4637
4638
4639
4640
4641
4642
4643
4644
4645
4646
4647
4648
4649
4650
4651
4652
4653
4654
4655
4656
4657
4658
4659
4660
4661
4662
4663
4664
4665
4666
4667
4668
4669
4670
4671
4672
4673
4674
4675
4676
4677
4678
4679
4680
4681
4682
4683
4684
4685
4686
4687
4688
4689
4690
4691
4692
4693
4694
4695
4696
4697
4698
4699
4700
4701
4702
4703
4704
4705
4706
4707
4708
4709
4710
4711
4712
4713
4714
4715
4716
4717
4718
4719
4720
4721
4722
4723
4724
4725
4726
4727
4728
4729
4730
4731
4732
4733
4734
4735
4736
4737
4738
4739
4740
4741
4742
4743
4744
4745
4746
4747
4748
4749
4750
4751
4752
4753
4754
4755
4756
4757
4758
4759
4760
4761
4762
4763
4764
4765
4766
4767
4768
4769
4770
4771
4772
4773
4774
4775
4776
4777
4778
4779
4780
4781
4782
4783
4784
4785
4786
4787
4788
4789
4790
4791
4792
4793
4794
4795
4796
4797
4798
4799
4800
4801
4802
4803
4804
4805
4806
4807
4808
4809
4810
4811
4812
4813
4814
4815
4816
4817
4818
4819
4820
4821
4822
4823
4824
4825
4826
4827
4828
4829
4830
4831
4832
4833
4834
4835
4836
4837
4838
4839
4840
4841
4842
4843
4844
4845
4846
4847
4848
4849
4850
4851
4852
4853
4854
4855
4856
4857
4858
4859
4860
4861
4862
4863
4864
4865
4866
4867
4868
4869
4870
4871
4872
4873
4874
4875
4876
4877
4878
4879
4880
4881
4882
4883
4884
4885
4886
4887
4888
4889
4890
4891
4892
4893
4894
4895
4896
4897
4898
4899
4900
4901
4902
4903
4904
4905
4906
4907
4908
4909
4910
4911
4912
4913
4914
4915
4916
4917
4918
4919
4920
4921
4922
4923
4924
4925
4926
4927
4928
4929
4930
4931
4932
4933
4934
4935
4936
4937
4938
4939
4940
4941
4942
4943
4944
4945
4946
4947
4948
4949
4950
4951
4952
4953
4954
4955
4956
4957
4958
4959
4960
4961
4962
4963
4964
4965
4966
4967
4968
4969
4970
4971
4972
4973
4974
4975
4976
4977
4978
4979
4980
4981
4982
4983
4984
4985
4986
4987
4988
4989
4990
4991
4992
4993
4994
4995
4996
4997
4998
4999
5000
5001
5002
5003
5004
5005
5006
5007
5008
5009
5010
5011
5012
5013
5014
5015
5016
5017
5018
5019
5020
5021
5022
5023
5024
5025
5026
5027
5028
5029
5030
5031
5032
5033
5034
5035
5036
5037
5038
5039
5040
5041
5042
5043
5044
5045
5046
5047
5048
5049
5050
5051
5052
5053
5054
5055
5056
5057
5058
5059
5060
5061
5062
5063
5064
5065
5066
5067
5068
5069
5070
5071
5072
5073
5074
5075
5076
5077
5078
5079
5080
5081
5082
5083
5084
5085
5086
5087
5088
5089
5090
5091
5092
5093
5094
5095
5096
5097
5098
5099
5100
5101
5102
5103
5104
5105
5106
5107
5108
5109
5110
5111
5112
5113
5114
5115
5116
5117
5118
5119
5120
5121
5122
5123
5124
5125
5126
5127
5128
5129
5130
5131
5132
5133
5134
5135
5136
5137
5138
5139
5140
5141
5142
5143
5144
5145
5146
5147
5148
5149
5150
5151
5152
5153
5154
5155
5156
5157
5158
5159
5160
5161
5162
5163
5164
5165
5166
5167
5168
5169
5170
5171
5172
5173
5174
5175
5176
5177
5178
5179
5180
5181
5182
5183
5184
5185
5186
5187
5188
5189
5190
5191
5192
5193
5194
5195
5196
5197
5198
5199
5200
5201
5202
5203
5204
5205
5206
5207
5208
5209
5210
5211
5212
5213
5214
5215
5216
5217
5218
5219
5220
5221
5222
5223
5224
5225
5226
5227
5228
5229
5230
5231
5232
5233
5234
5235
5236
5237
5238
5239
5240
5241
5242
5243
5244
5245
5246
5247
5248
5249
5250
5251
5252
5253
5254
5255
5256
5257
5258
5259
5260
5261
5262
5263
5264
5265
5266
5267
5268
5269
5270
5271
5272
5273
5274
5275
5276
5277
5278
5279
5280
5281
5282
5283
5284
5285
5286
5287
5288
5289
5290
5291
5292
5293
5294
5295
5296
5297
5298
5299
5300
5301
5302
5303
5304
5305
5306
5307
5308
5309
5310
5311
5312
5313
5314
5315
5316
5317
5318
5319
5320
5321
5322
5323
5324
5325
5326
5327
5328
5329
5330
5331
5332
5333
5334
5335
5336
5337
5338
5339
5340
5341
5342
5343
5344
5345
5346
5347
5348
5349
5350
5351
5352
5353
5354
5355
5356
5357
5358
5359
5360
5361
5362
5363
5364
5365
5366
5367
5368
5369
5370
5371
5372
5373
5374
5375
5376
5377
5378
5379
5380
5381
5382
5383
5384
5385
5386
5387
5388
5389
5390
5391
5392
5393
5394
5395
5396
5397
5398
5399
5400
5401
5402
5403
5404
5405
5406
5407
5408
5409
5410
5411
5412
5413
5414
5415
5416
5417
5418
5419
5420
5421
5422
5423
5424
5425
5426
5427
5428
5429
5430
5431
5432
5433
5434
5435
5436
5437
5438
5439
5440
5441
5442
5443
5444
5445
5446
5447
5448
5449
5450
5451
5452
5453
5454
5455
5456
5457
5458
5459
5460
5461
5462
5463
5464
5465
5466
5467
5468
5469
5470
5471
5472
5473
5474
5475
5476
5477
5478
5479
5480
5481
5482
5483
5484
5485
5486
5487
5488
5489
5490
5491
5492
5493
5494
5495
5496
5497
5498
5499
5500
5501
5502
5503
5504
5505
5506
5507
5508
5509
5510
5511
5512
5513
5514
5515
5516
5517
5518
5519
5520
5521
5522
5523
5524
5525
5526
5527
5528
5529
5530
5531
5532
5533
5534
5535
5536
5537
5538
5539
5540
5541
5542
5543
5544
5545
5546
5547
5548
5549
5550
5551
5552
5553
5554
5555
5556
5557
5558
5559
5560
5561
5562
5563
5564
5565
5566
5567
5568
5569
5570
5571
5572
5573
5574
5575
5576
5577
5578
5579
5580
5581
5582
5583
5584
5585
5586
5587
5588
5589
5590
5591
5592
5593
5594
5595
5596
5597
5598
5599
5600
5601
5602
5603
5604
5605
5606
5607
5608
5609
5610
5611
5612
5613
5614
5615
5616
5617
5618
5619
5620
5621
5622
5623
5624
5625
5626
5627
5628
5629
5630
5631
5632
5633
5634
5635
5636
5637
5638
5639
5640
5641
5642
5643
5644
5645
5646
5647
5648
5649
5650
5651
5652
5653
5654
5655
5656
5657
5658
5659
5660
5661
5662
5663
5664
5665
5666
5667
5668
5669
5670
5671
5672
5673
5674
5675
5676
5677
5678
5679
5680
5681
5682
5683
5684
5685
5686
5687
5688
5689
5690
5691
5692
5693
5694
5695
5696
5697
5698
5699
5700
5701
5702
5703
5704
5705
5706
5707
5708
5709
5710
5711
5712
5713
5714
5715
5716
5717
5718
5719
5720
5721
5722
5723
5724
5725
5726
5727
5728
5729
5730
5731
5732
5733
5734
5735
5736
5737
5738
5739
5740
5741
5742
5743
5744
5745
5746
5747
5748
5749
5750
5751
5752
5753
5754
5755
5756
5757
5758
5759
5760
5761
5762
5763
5764
5765
5766
5767
5768
5769
5770
5771
5772
5773
5774
5775
5776
5777
5778
5779
5780
5781
5782
5783
5784
5785
5786
5787
5788
5789
5790
5791
5792
5793
5794
5795
5796
5797
5798
5799
5800
5801
5802
5803
5804
5805
5806
5807
5808
5809
5810
5811
5812
5813
5814
5815
5816
5817
5818
5819
5820
5821
5822
5823
5824
5825
5826
5827
5828
5829
5830
5831
5832
5833
5834
5835
5836
5837
5838
5839
5840
5841
5842
5843
5844
5845
5846
5847
5848
5849
5850
5851
5852
5853
5854
5855
5856
5857
5858
5859
5860
5861
5862
5863
5864
5865
5866
5867
5868
5869
5870
5871
5872
5873
5874
5875
5876
5877
5878
5879
5880
5881
5882
5883
5884
5885
5886
5887
5888
5889
5890
5891
5892
5893
5894
5895
5896
5897
5898
5899
5900
5901
5902
5903
5904
5905
5906
5907
5908
5909
5910
5911
5912
5913
5914
5915
5916
5917
5918
5919
5920
5921
5922
5923
5924
5925
5926
5927
5928
5929
5930
5931
5932
5933
5934
5935
5936
5937
5938
5939
5940
5941
5942
5943
5944
5945
5946
5947
5948
5949
5950
5951
5952
5953
5954
5955
5956
5957
5958
5959
5960
5961
5962
5963
5964
5965
5966
5967
5968
5969
5970
5971
5972
5973
5974
5975
5976
5977
5978
5979
5980
5981
5982
5983
5984
5985
5986
5987
5988
5989
5990
5991
5992
5993
5994
5995
5996
5997
5998
5999
6000
6001
6002
6003
6004
6005
6006
6007
6008
6009
6010
6011
6012
6013
6014
6015
6016
6017
6018
6019
6020
6021
6022
6023
6024
6025
6026
6027
6028
6029
6030
6031
6032
6033
6034
6035
6036
6037
6038
6039
6040
6041
6042
6043
6044
6045
6046
6047
6048
6049
6050
6051
6052
6053
6054
6055
6056
6057
6058
6059
6060
6061
6062
6063
6064
6065
6066
6067
6068
6069
6070
6071
6072
6073
6074
6075
6076
6077
6078
6079
6080
6081
6082
6083
6084
6085
6086
6087
6088
6089
6090
6091
6092
6093
6094
6095
6096
6097
6098
6099
6100
6101
6102
6103
6104
6105
6106
6107
6108
6109
6110
6111
6112
6113
6114
6115
6116
6117
6118
6119
6120
6121
6122
6123
6124
6125
6126
6127
6128
6129
6130
6131
6132
6133
6134
6135
6136
6137
6138
6139
6140
6141
6142
6143
6144
6145
6146
6147
6148
6149
6150
6151
6152
6153
6154
6155
6156
6157
6158
6159
6160
6161
6162
6163
6164
6165
6166
6167
6168
6169
6170
6171
6172
6173
6174
6175
6176
6177
6178
6179
6180
6181
6182
6183
6184
6185
6186
6187
6188
6189
6190
6191
6192
6193
6194
6195
6196
6197
6198
6199
6200
6201
6202
6203
6204
6205
6206
6207
6208
6209
6210
6211
6212
6213
6214
6215
6216
6217
6218
6219
6220
6221
6222
6223
6224
6225
6226
6227
6228
6229
6230
6231
6232
6233
6234
6235
6236
6237
6238
6239
6240
6241
6242
6243
6244
6245
6246
6247
6248
6249
6250
6251
6252
6253
6254
6255
6256
6257
6258
6259
6260
6261
6262
6263
6264
6265
6266
6267
6268
6269
6270
6271
6272
6273
6274
6275
6276
6277
6278
6279
6280
6281
6282
6283
6284
6285
6286
6287
6288
6289
6290
6291
6292
6293
6294
6295
6296
6297
6298
6299
6300
6301
6302
6303
6304
6305
6306
6307
6308
6309
6310
6311
6312
6313
6314
6315
6316
6317
6318
6319
6320
6321
6322
6323
6324
6325
6326
6327
6328
6329
6330
6331
6332
6333
6334
6335
6336
6337
6338
6339
6340
6341
6342
6343
6344
6345
6346
6347
6348
6349
6350
6351
6352
6353
6354
6355
6356
6357
6358
6359
6360
6361
6362
6363
6364
6365
6366
6367
6368
6369
6370
6371
6372
6373
6374
6375
6376
6377
6378
6379
6380
6381
6382
6383
6384
6385
6386
6387
6388
6389
6390
6391
6392
6393
6394
6395
6396
6397
6398
6399
6400
6401
6402
6403
6404
6405
6406
6407
6408
6409
6410
6411
6412
6413
6414
6415
6416
6417
6418
6419
6420
6421
6422
6423
6424
6425
6426
6427
6428
6429
6430
6431
6432
6433
6434
6435
6436
6437
6438
6439
6440
6441
6442
6443
6444
6445
6446
6447
6448
6449
6450
6451
6452
6453
6454
6455
6456
6457
6458
6459
6460
6461
6462
6463
6464
6465
6466
6467
6468
6469
6470
6471
6472
6473
6474
6475
6476
6477
6478
6479
6480
6481
6482
6483
6484
6485
6486
6487
6488
6489
6490
6491
6492
6493
6494
6495
6496
6497
6498
6499
6500
6501
6502
6503
6504
6505
6506
6507
6508
6509
6510
6511
6512
6513
6514
6515
6516
6517
6518
6519
6520
6521
6522
6523
6524
6525
6526
6527
6528
6529
6530
6531
6532
6533
6534
6535
6536
6537
6538
6539
6540
6541
6542
6543
6544
6545
6546
6547
6548
6549
6550
6551
6552
6553
6554
6555
6556
6557
6558
6559
6560
6561
6562
6563
6564
6565
6566
6567
6568
6569
6570
6571
6572
6573
6574
6575
6576
6577
6578
6579
6580
6581
6582
6583
6584
6585
6586
6587
/*	$NetBSD: failover.c,v 1.2 2018/04/07 22:37:30 christos Exp $	*/

/* failover.c

   Failover protocol support code... */

/*
 * Copyright (c) 2004-2017 by Internet Systems Consortium, Inc. ("ISC")
 * Copyright (c) 1999-2003 by Internet Software Consortium
 *
 * This Source Code Form is subject to the terms of the Mozilla Public
 * License, v. 2.0. If a copy of the MPL was not distributed with this
 * file, You can obtain one at http://mozilla.org/MPL/2.0/.
 *
 * 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.
 *
 *   Internet Systems Consortium, Inc.
 *   950 Charter Street
 *   Redwood City, CA 94063
 *   <info@isc.org>
 *   https://www.isc.org/
 *
 */

#include <sys/cdefs.h>
__RCSID("$NetBSD: failover.c,v 1.2 2018/04/07 22:37:30 christos Exp $");

#include "cdefs.h"
#include "dhcpd.h"
#include <omapip/omapip_p.h>

#if defined (FAILOVER_PROTOCOL)
dhcp_failover_state_t *failover_states;
static isc_result_t do_a_failover_option (omapi_object_t *,
					  dhcp_failover_link_t *);
dhcp_failover_listener_t *failover_listeners;

static isc_result_t failover_message_reference (failover_message_t **,
						failover_message_t *,
						const char *file, int line);
static isc_result_t failover_message_dereference (failover_message_t **,
						  const char *file, int line);

static void dhcp_failover_pool_balance(dhcp_failover_state_t *state);
static void dhcp_failover_pool_reqbalance(dhcp_failover_state_t *state);
static int dhcp_failover_pool_dobalance(dhcp_failover_state_t *state,
					isc_boolean_t *sendreq);
static inline int secondary_not_hoarding(dhcp_failover_state_t *state,
					 struct pool *p);
static void scrub_lease(struct lease* lease, const char *file, int line);

int check_secs_byte_order = 0; /* enables byte order check of secs field if 1 */

/*!
 * \brief Performs a "pre-flight" sanity check of failover configuration
 *
 * Provides an opportunity to do post-parse pre-startup sanity checking
 * of failover configuration.  This allows checks to be done under test
 * mode (-T), without requiring full startup for validation.
 *
 * Currently, it enforces all failover peers be used in at lease one
 * pool. This logic was formerly located in dhcp_failover_startup.
 *
 * On failure, a fatal error is logged.
 *
 */
void dhcp_failover_sanity_check() {
	dhcp_failover_state_t *state;
	int fail_count = 0;

	for (state = failover_states; state; state = state->next) {
		if (state->pool_count == 0) {
			log_error ("ERROR: Failover peer, %s, has no referring"
				   " pools. You must refer to each peer in at"
                                   " least one pool declaration.",
				   state->name);
			fail_count++;
		}

		if (state->load_balance_max_secs == 0) {
			log_info ("WARNING: load balancing will be disabled "
                                  "for failover peer, %s, "
				  "because its load balance max secs is 0",
				  state->name);
		}
	}

	if (fail_count) {
		log_fatal ("Failover configuration sanity check failed");
	}

}

void dhcp_failover_startup ()
{
	dhcp_failover_state_t *state;
	isc_result_t status;
	struct timeval tv;

	for (state = failover_states; state; state = state -> next) {
		dhcp_failover_state_transition (state, "startup");
		/* In case the peer is already running, immediately try
		   to establish a connection with it. */
		status = dhcp_failover_link_initiate ((omapi_object_t *)state);
		if (status != ISC_R_SUCCESS && status != DHCP_R_INCOMPLETE) {
#if defined (DEBUG_FAILOVER_TIMING)
			log_info ("add_timeout +90 dhcp_failover_reconnect");
#endif
			tv . tv_sec = cur_time + 90;
			tv . tv_usec = 0;
			add_timeout (&tv,
				     dhcp_failover_reconnect, state,
				     (tvref_t)
				     dhcp_failover_state_reference,
				     (tvunref_t)
				     dhcp_failover_state_dereference);
			log_error ("failover peer %s: %s", state -> name,
				   isc_result_totext (status));
		}

		status = (dhcp_failover_listen
			  ((omapi_object_t *)state));
		if (status != ISC_R_SUCCESS) {
#if defined (DEBUG_FAILOVER_TIMING)
			log_info ("add_timeout +90 %s",
				  "dhcp_failover_listener_restart");
#endif
			tv . tv_sec = cur_time + 90;
			tv . tv_usec = 0;
			add_timeout (&tv,
				     dhcp_failover_listener_restart,
				     state,
				     (tvref_t)omapi_object_reference,
				     (tvunref_t)omapi_object_dereference);
		}
	}
}

int dhcp_failover_write_all_states ()
{
	dhcp_failover_state_t *state;

	for (state = failover_states; state; state = state -> next) {
		if (!write_failover_state (state))
			return 0;
	}
	return 1;
}

isc_result_t enter_failover_peer (peer)
	dhcp_failover_state_t *peer;
{
	dhcp_failover_state_t *dup = (dhcp_failover_state_t *)0;
	isc_result_t status;

	status = find_failover_peer (&dup, peer -> name, MDL);
	if (status == ISC_R_NOTFOUND) {
		if (failover_states) {
			dhcp_failover_state_reference (&peer -> next,
						      failover_states, MDL);
			dhcp_failover_state_dereference (&failover_states,
							 MDL);
		}
		dhcp_failover_state_reference (&failover_states, peer, MDL);
		return ISC_R_SUCCESS;
	}
	dhcp_failover_state_dereference (&dup, MDL);
	if (status == ISC_R_SUCCESS)
		return ISC_R_EXISTS;
	return status;
}

isc_result_t find_failover_peer (peer, name, file, line)
	dhcp_failover_state_t **peer;
	const char *name;
	const char *file;
	int line;
{
	dhcp_failover_state_t *p;

	for (p = failover_states; p; p = p -> next)
		if (!strcmp (name, p -> name))
			break;
	if (p)
		return dhcp_failover_state_reference (peer, p, file, line);
	return ISC_R_NOTFOUND;
}

/* The failover protocol has three objects associated with it.  For
   each failover partner declaration in the dhcpd.conf file, primary
   or secondary, there is a failover_state object.  For any primary or
   secondary state object that has a connection to its peer, there is
   also a failover_link object, which has its own input state separate
   from the failover protocol state for managing the actual bytes
   coming in off the wire.  Finally, there will be one listener object
   for every distinct port number associated with a secondary
   failover_state object.  Normally all secondary failover_state
   objects are expected to listen on the same port number, so there
   need be only one listener object, but if different port numbers are
   specified for each failover object, there could be as many as one
   listener object for each secondary failover_state object. */

/* This, then, is the implementation of the failover link object. */

isc_result_t dhcp_failover_link_initiate (omapi_object_t *h)
{
	isc_result_t status;
	dhcp_failover_link_t *obj;
	dhcp_failover_state_t *state;
	omapi_object_t *o;
	int i;
	struct data_string ds;
	omapi_addr_list_t *addrs = (omapi_addr_list_t *)0;
	omapi_addr_t local_addr;

	/* Find the failover state in the object chain. */
	for (o = h; o -> outer; o = o -> outer)
		;
	for (; o; o = o -> inner) {
		if (o -> type == dhcp_type_failover_state)
			break;
	}
	if (!o)
		return DHCP_R_INVALIDARG;
	state = (dhcp_failover_state_t *)o;

	obj = (dhcp_failover_link_t *)0;
	status = dhcp_failover_link_allocate (&obj, MDL);
	if (status != ISC_R_SUCCESS)
		return status;
	option_cache_reference (&obj -> peer_address,
				state -> partner.address, MDL);
	obj -> peer_port = state -> partner.port;
	dhcp_failover_state_reference (&obj -> state_object, state, MDL);

	memset (&ds, 0, sizeof ds);
	if (!evaluate_option_cache (&ds, (struct packet *)0, (struct lease *)0,
				    (struct client_state *)0,
				    (struct option_state *)0,
				    (struct option_state *)0,
				    &global_scope, obj -> peer_address, MDL)) {
		dhcp_failover_link_dereference (&obj, MDL);
		return ISC_R_UNEXPECTED;
	}

	/* Make an omapi address list out of a buffer containing zero or more
	   IPv4 addresses. */
	status = omapi_addr_list_new (&addrs, ds.len / 4, MDL);
	if (status != ISC_R_SUCCESS) {
		dhcp_failover_link_dereference (&obj, MDL);
		return status;
	}

	for (i = 0; i  < addrs -> count; i++) {
		addrs -> addresses [i].addrtype = AF_INET;
		addrs -> addresses [i].addrlen = sizeof (struct in_addr);
		memcpy (addrs -> addresses [i].address,
			&ds.data [i * 4], sizeof (struct in_addr));
		addrs -> addresses [i].port = obj -> peer_port;
	}
	data_string_forget (&ds, MDL);

	/* Now figure out the local address that we're supposed to use. */
	if (!state -> me.address ||
	    !evaluate_option_cache (&ds, (struct packet *)0,
				    (struct lease *)0,
				    (struct client_state *)0,
				    (struct option_state *)0,
				    (struct option_state *)0,
				    &global_scope, state -> me.address,
				    MDL)) {
		memset (&local_addr, 0, sizeof local_addr);
		local_addr.addrtype = AF_INET;
		local_addr.addrlen = sizeof (struct in_addr);
		if (!state -> server_identifier.len) {
			log_fatal ("failover peer %s: no local address.",
				   state -> name);
		}
	} else {
		if (ds.len != sizeof (struct in_addr)) {
			log_error("failover peer %s: 'address' parameter "
				  "fails to resolve to an IPv4 address",
				  state->name);
			data_string_forget (&ds, MDL);
			dhcp_failover_link_dereference (&obj, MDL);
			omapi_addr_list_dereference (&addrs, MDL);
			return DHCP_R_INVALIDARG;
		}
		local_addr.addrtype = AF_INET;
		local_addr.addrlen = ds.len;
		memcpy (local_addr.address, ds.data, ds.len);
		if (!state -> server_identifier.len)
			data_string_copy (&state -> server_identifier,
					  &ds, MDL);
		data_string_forget (&ds, MDL);
		local_addr.port = 0;  /* Let the O.S. choose. */
	}

	status = omapi_connect_list ((omapi_object_t *)obj,
				     addrs, &local_addr);
	omapi_addr_list_dereference (&addrs, MDL);

	dhcp_failover_link_dereference (&obj, MDL);
	return status;
}

isc_result_t dhcp_failover_link_signal (omapi_object_t *h,
					const char *name, va_list ap)
{
	isc_result_t status;
	dhcp_failover_link_t *link;
	omapi_object_t *c;
	dhcp_failover_state_t *s, *state = (dhcp_failover_state_t *)0;
	char *sname;
	int slen;
	struct timeval tv;

	if (h -> type != dhcp_type_failover_link) {
		/* XXX shouldn't happen.   Put an assert here? */
		return ISC_R_UNEXPECTED;
	}
	link = (dhcp_failover_link_t *)h;

	if (!strcmp (name, "connect")) {
	    if (link -> state_object -> i_am == primary) {
		status = dhcp_failover_send_connect (h);
		if (status != ISC_R_SUCCESS) {
		    log_info ("dhcp_failover_send_connect: %s",
			      isc_result_totext (status));
		    omapi_disconnect (h -> outer, 1);
		}
	    } else
		status = ISC_R_SUCCESS;
	    /* Allow the peer fifteen seconds to send us a
	       startup message. */
#if defined (DEBUG_FAILOVER_TIMING)
	    log_info ("add_timeout +15 %s",
		      "dhcp_failover_link_startup_timeout");
#endif
	    tv . tv_sec = cur_time + 15;
	    tv . tv_usec = 0;
	    add_timeout (&tv,
			 dhcp_failover_link_startup_timeout,
			 link,
			 (tvref_t)dhcp_failover_link_reference,
			 (tvunref_t)dhcp_failover_link_dereference);
	    return status;
	}

	if (!strcmp (name, "disconnect")) {
	    if (link -> state_object) {
		dhcp_failover_state_reference (&state,
					       link -> state_object, MDL);
		link -> state = dhcp_flink_disconnected;

		/* Make the transition. */
		if (state->link_to_peer == link)
		    dhcp_failover_state_transition(link->state_object, name);

		/* Schedule an attempt to reconnect. */
#if defined (DEBUG_FAILOVER_TIMING)
		log_info("add_timeout +5 dhcp_failover_reconnect");
#endif
		tv.tv_sec = cur_time + 5;
		tv.tv_usec = cur_tv.tv_usec;
		add_timeout(&tv, dhcp_failover_reconnect, state,
			    (tvref_t)dhcp_failover_state_reference,
			    (tvunref_t)dhcp_failover_state_dereference);

		dhcp_failover_state_dereference (&state, MDL);
	    }
	    return ISC_R_SUCCESS;
	}

	if (!strcmp (name, "status")) {
	  if (link -> state_object) {
	    isc_result_t	status;

	    status = va_arg(ap, isc_result_t);

	    if ((status == ISC_R_HOSTUNREACH) || (status == ISC_R_TIMEDOUT)) {
	      dhcp_failover_state_reference (&state,
					     link -> state_object, MDL);
	      link -> state = dhcp_flink_disconnected;

	      /* Make the transition. */
	      dhcp_failover_state_transition (link -> state_object,
					      "disconnect");

	      /* Start trying to reconnect. */
#if defined (DEBUG_FAILOVER_TIMING)
	      log_info ("add_timeout +5 %s",
			"dhcp_failover_reconnect");
#endif
	      tv . tv_sec = cur_time + 5;
	      tv . tv_usec = 0;
	      add_timeout (&tv, dhcp_failover_reconnect,
			   state,
			   (tvref_t)dhcp_failover_state_reference,
			   (tvunref_t)dhcp_failover_state_dereference);
	    }
	    dhcp_failover_state_dereference (&state, MDL);
	  }
	  return ISC_R_SUCCESS;
	}

	/* Not a signal we recognize? */
	if (strcmp (name, "ready")) {
		if (h -> inner && h -> inner -> type -> signal_handler)
			return (*(h -> inner -> type -> signal_handler))
				(h -> inner, name, ap);
		return ISC_R_NOTFOUND;
	}

	if (!h -> outer || h -> outer -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;
	c = h -> outer;

	/* We get here because we requested that we be woken up after
           some number of bytes were read, and that number of bytes
           has in fact been read. */
	switch (link -> state) {
	      case dhcp_flink_start:
		link -> state = dhcp_flink_message_length_wait;
		if ((omapi_connection_require (c, 2)) != ISC_R_SUCCESS)
			break;
	      case dhcp_flink_message_length_wait:
	      next_message:
		link -> state = dhcp_flink_message_wait;
		link -> imsg = dmalloc (sizeof (failover_message_t), MDL);
		if (!link -> imsg) {
			status = ISC_R_NOMEMORY;
		      dhcp_flink_fail:
			if (link -> imsg) {
				failover_message_dereference (&link->imsg,
							      MDL);
			}
			link -> state = dhcp_flink_disconnected;
			log_info ("message length wait: %s",
				  isc_result_totext (status));
			omapi_disconnect (c, 1);
			/* XXX just blow away the protocol state now?
			   XXX or will disconnect blow it away? */
			return ISC_R_UNEXPECTED;
		}
		memset (link -> imsg, 0, sizeof (failover_message_t));
		link -> imsg -> refcnt = 1;
		/* Get the length: */
		omapi_connection_get_uint16 (c, &link -> imsg_len);
		link -> imsg_count = 0;	/* Bytes read. */

		/* Ensure the message is of valid length. */
		if (link->imsg_len < DHCP_FAILOVER_MIN_MESSAGE_SIZE ||
		    link->imsg_len > DHCP_FAILOVER_MAX_MESSAGE_SIZE) {
			status = ISC_R_UNEXPECTED;
			goto dhcp_flink_fail;
		}

		if ((omapi_connection_require (c, link -> imsg_len - 2U)) !=
		    ISC_R_SUCCESS)
			break;
	      case dhcp_flink_message_wait:
		/* Read in the message.  At this point we have the
		   entire message in the input buffer.  For each
		   incoming value ID, set a bit in the bitmask
		   indicating that we've gotten it.  Maybe flag an
		   error message if the bit is already set.  Once
		   we're done reading, we can check the bitmask to
		   make sure that the required fields for each message
		   have been included. */

		link -> imsg_count += 2;	/* Count the length as read. */

		/* Get message type. */
		omapi_connection_copyout (&link -> imsg -> type, c, 1);
		link -> imsg_count++;

		/* Get message payload offset. */
		omapi_connection_copyout (&link -> imsg_payoff, c, 1);
		link -> imsg_count++;

		/* Get message time. */
		omapi_connection_get_uint32 (c, &link -> imsg -> time);
		link -> imsg_count += 4;

		/* Get transaction ID. */
		omapi_connection_get_uint32 (c, &link -> imsg -> xid);
		link -> imsg_count += 4;

#if defined (DEBUG_FAILOVER_MESSAGES)
# if !defined(DEBUG_FAILOVER_CONTACT_MESSAGES)
		if (link->imsg->type == FTM_CONTACT)
			goto skip_contact;
# endif
		log_info ("link: message %s  payoff %d  time %ld  xid %ld",
			  dhcp_failover_message_name (link -> imsg -> type),
			  link -> imsg_payoff,
			  (unsigned long)link -> imsg -> time,
			  (unsigned long)link -> imsg -> xid);
# if !defined(DEBUG_FAILOVER_CONTACT_MESSAGES)
	      skip_contact:
# endif
#endif
		/* Skip over any portions of the message header that we
		   don't understand. */
		if (link -> imsg_payoff - link -> imsg_count) {
			omapi_connection_copyout ((unsigned char *)0, c,
						  (link -> imsg_payoff -
						   link -> imsg_count));
			link -> imsg_count = link -> imsg_payoff;
		}

		/* Now start sucking options off the wire. */
		while (link -> imsg_count < link -> imsg_len) {
			status = do_a_failover_option (c, link);
			if (status != ISC_R_SUCCESS)
				goto dhcp_flink_fail;
		}

		/* If it's a connect message, try to associate it with
		   a state object. */
		/* XXX this should be authenticated! */
		if (link -> imsg -> type == FTM_CONNECT) {
		    const char *errmsg;
		    int reason;

		    if (!(link->imsg->options_present &
				FTB_RELATIONSHIP_NAME)) {
			errmsg = "missing relationship-name";
			reason = FTR_INVALID_PARTNER;
			goto badconnect;
		    }

		    /* See if we can find a failover_state object that
		       matches this connection.  This message should only
		       be received by a secondary from a primary. */
		    for (s = failover_states; s; s = s -> next) {
			if (dhcp_failover_state_match_by_name(s,
			    &link->imsg->relationship_name))
				state = s;
		    }

		    /* If we can't find a failover protocol state
		       for this remote host, drop the connection */
		    if (!state) {
			    errmsg = "unknown failover relationship name";
			    reason = FTR_INVALID_PARTNER;

			  badconnect:
				/* XXX Send a refusal message first?
				   XXX Look in protocol spec for guidance. */

			    if (state != NULL) {
				sname = state->name;
				slen = strlen(sname);
			    } else if (link->imsg->options_present &
				       FTB_RELATIONSHIP_NAME) {
				sname = (char *)link->imsg->
						relationship_name.data;
				slen = link->imsg->relationship_name.count;
			    } else {
				sname = "unknown";
				slen = strlen(sname);
			    }

			    log_error("Failover CONNECT from %.*s: %s",
				      slen, sname, errmsg);
			    dhcp_failover_send_connectack
				    ((omapi_object_t *)link, state,
				     reason, errmsg);
			    log_info ("failover: disconnect: %s", errmsg);
			    omapi_disconnect (c, 0);
			    link -> state = dhcp_flink_disconnected;
			    return ISC_R_SUCCESS;
		    }

		    if ((cur_time > link -> imsg -> time &&
			 cur_time - link -> imsg -> time > 60) ||
			(cur_time < link -> imsg -> time &&
			 link -> imsg -> time - cur_time > 60)) {
			    errmsg = "time offset too large";
			    reason = FTR_TIMEMISMATCH;
			    goto badconnect;
		    }

		    if (!(link -> imsg -> options_present & FTB_HBA) ||
			link -> imsg -> hba.count != 32) {
			    errmsg = "invalid HBA";
			    reason = FTR_HBA_CONFLICT; /* XXX */
			    goto badconnect;
		    }
		    if (state -> hba)
			    dfree (state -> hba, MDL);
		    state -> hba = dmalloc (32, MDL);
		    if (!state -> hba) {
			    errmsg = "no memory";
			    reason = FTR_MISC_REJECT;
			    goto badconnect;
		    }
		    memcpy (state -> hba, link -> imsg -> hba.data, 32);

		    if (!link -> state_object)
			    dhcp_failover_state_reference
				    (&link -> state_object, state, MDL);
		    if (!link -> peer_address)
			    option_cache_reference
				    (&link -> peer_address,
				     state -> partner.address, MDL);
		}

		/* If we don't have a state object at this point, it's
		   some kind of bogus situation, so just drop the
		   connection. */
		if (!link -> state_object) {
			log_info ("failover: connect: no matching state.");
			omapi_disconnect (c, 1);
			link -> state = dhcp_flink_disconnected;
			return DHCP_R_INVALIDARG;
		}

		/* Once we have the entire message, and we've validated
		   it as best we can here, pass it to the parent. */
		omapi_signal ((omapi_object_t *)link -> state_object,
			      "message", link);
		link -> state = dhcp_flink_message_length_wait;
		if (link -> imsg)
			failover_message_dereference (&link -> imsg, MDL);
		/* XXX This is dangerous because we could get into a tight
		   XXX loop reading input without servicing any other stuff.
		   XXX There needs to be a way to relinquish control but
		   XXX get it back immediately if there's no other work to
		   XXX do. */
		if ((omapi_connection_require (c, 2)) == ISC_R_SUCCESS)
			goto next_message;
		break;

	      default:
		log_fatal("Impossible case at %s:%d.", MDL);
		break;
	}
	return ISC_R_SUCCESS;
}

static isc_result_t do_a_failover_option (c, link)
	omapi_object_t *c;
	dhcp_failover_link_t *link;
{
	u_int16_t option_code;
	u_int16_t option_len;
	unsigned char *op;
	unsigned op_size;
	unsigned op_count;
	int i;

	if (link -> imsg_count + 2 > link -> imsg_len) {
		log_error ("FAILOVER: message overflow at option code.");
		return DHCP_R_PROTOCOLERROR;
	}

	if (link->imsg->type > FTM_MAX) {
		log_error ("FAILOVER: invalid message type: %d",
			   link->imsg->type);
		return DHCP_R_PROTOCOLERROR;
	}

	/* Get option code. */
	omapi_connection_get_uint16 (c, &option_code);
	link -> imsg_count += 2;

	if (link -> imsg_count + 2 > link -> imsg_len) {
		log_error ("FAILOVER: message overflow at length.");
		return DHCP_R_PROTOCOLERROR;
	}

	/* Get option length. */
	omapi_connection_get_uint16 (c, &option_len);
	link -> imsg_count += 2;

	if (link -> imsg_count + option_len > link -> imsg_len) {
		log_error ("FAILOVER: message overflow at data.");
		return DHCP_R_PROTOCOLERROR;
	}

	/* If it's an unknown code, skip over it. */
	if ((option_code > FTO_MAX) ||
	    (ft_options[option_code].type == FT_UNDEF)) {
#if defined (DEBUG_FAILOVER_MESSAGES)
		log_debug ("  option code %d (%s) len %d (not recognized)",
			   option_code,
			   dhcp_failover_option_name (option_code),
			   option_len);
#endif
		omapi_connection_copyout ((unsigned char *)0, c, option_len);
		link -> imsg_count += option_len;
		return ISC_R_SUCCESS;
	}

	/* If it's the digest, do it now. */
	if (ft_options [option_code].type == FT_DIGEST) {
		link -> imsg_count += option_len;
		if (link -> imsg_count != link -> imsg_len) {
			log_error ("FAILOVER: digest not at end of message");
			return DHCP_R_PROTOCOLERROR;
		}
#if defined (DEBUG_FAILOVER_MESSAGES)
		log_debug ("  option %s len %d",
			   ft_options [option_code].name, option_len);
#endif
		/* For now, just dump it. */
		omapi_connection_copyout ((unsigned char *)0, c, option_len);
		return ISC_R_SUCCESS;
	}

	/* Only accept an option once. */
	if (link -> imsg -> options_present & ft_options [option_code].bit) {
		log_error ("FAILOVER: duplicate option %s",
			   ft_options [option_code].name);
		return DHCP_R_PROTOCOLERROR;
	}

	/* Make sure the option is appropriate for this type of message.
	   Really, any option is generally allowed for any message, and the
	   cases where this is not true are too complicated to represent in
	   this way - what this code is doing is to just avoid saving the
	   value of an option we don't have any way to use, which allows
	   us to make the failover_message structure smaller. */
	if (ft_options [option_code].bit &&
	    !(fto_allowed [link -> imsg -> type] &
	      ft_options [option_code].bit)) {
		omapi_connection_copyout ((unsigned char *)0, c, option_len);
		link -> imsg_count += option_len;
		return ISC_R_SUCCESS;
	}

	/* Figure out how many elements, how big they are, and where
	   to store them. */
	if (ft_options [option_code].num_present) {
		/* If this option takes a fixed number of elements,
		   we expect the space for them to be preallocated,
		   and we can just read the data in. */

		op = ((unsigned char *)link -> imsg) +
			ft_options [option_code].offset;
		op_size = ft_sizes [ft_options [option_code].type];
		op_count = ft_options [option_code].num_present;

		if (option_len != op_size * op_count) {
			log_error ("FAILOVER: option size (%d:%d), option %s",
				   option_len,
				   (ft_sizes [ft_options [option_code].type] *
				    ft_options [option_code].num_present),
				   ft_options [option_code].name);
			return DHCP_R_PROTOCOLERROR;
		}
	} else {
		failover_option_t *fo;

		/* FT_DDNS* are special - one or two bytes of status
		   followed by the client FQDN. */

		/* Note: FT_DDNS* option support appears to be incomplete.
		   ISC-Bugs #36996 has been opened to address this. */
		if (ft_options [option_code].type == FT_DDNS ||
		    ft_options [option_code].type == FT_DDNS1) {
			ddns_fqdn_t *ddns =
				((ddns_fqdn_t *)
				 (((char *)link -> imsg) +
				  ft_options [option_code].offset));

			op_count = (ft_options [option_code].type == FT_DDNS1
				    ? 1 : 2);

			omapi_connection_copyout (&ddns -> codes [0],
						  c, op_count);
			link -> imsg_count += op_count;
			if (op_count == 1)
				ddns -> codes [1] = 0;
			op_size = 1;
			op_count = option_len - op_count;

			ddns -> length = op_count;
			ddns -> data = dmalloc (op_count, MDL);
			if (!ddns -> data) {
				log_error ("FAILOVER: no memory getting%s(%d)",
					   " DNS data ", op_count);

				/* Actually, NO_MEMORY, but if we lose here
				   we have to drop the connection. */
				return DHCP_R_PROTOCOLERROR;
			}
			omapi_connection_copyout (ddns -> data, c, op_count);
			goto out;
		}

		/* A zero for num_present means that any number of
		   elements can appear, so we have to figure out how
		   many we got from the length of the option, and then
		   fill out a failover_option structure describing the
		   data. */
		op_size = ft_sizes [ft_options [option_code].type];

		/* Make sure that option data length is a multiple of the
		   size of the data type being sent. */
		if (op_size > 1 && option_len % op_size) {
			log_error ("FAILOVER: option_len %d not %s%d",
				   option_len, "multiple of ", op_size);
			return DHCP_R_PROTOCOLERROR;
		}

		op_count = option_len / op_size;

		fo = ((failover_option_t *)
		      (((char *)link -> imsg) +
		       ft_options [option_code].offset));

		fo -> count = op_count;
		fo -> data = dmalloc (option_len, MDL);
		if (!fo -> data) {
			log_error ("FAILOVER: no memory getting %s (%d)",
				   "option data", op_count);

			return DHCP_R_PROTOCOLERROR;
		}
		op = fo -> data;
	}

	/* For single-byte message values and multi-byte values that
           don't need swapping, just read them in all at once. */
	if (op_size == 1 || ft_options [option_code].type == FT_IPADDR) {
		omapi_connection_copyout ((unsigned char *)op, c, option_len);
		link -> imsg_count += option_len;

		/*
		 * As of 3.1.0, many option codes were changed to conform to
		 * draft revision 12 (which alphabetized, then renumbered all
		 * the option codes without preserving the version option code
		 * nor bumping its value).  As it turns out, the message codes
		 * for CONNECT and CONNECTACK turn out the same, so it tries
		 * its darndest to connect, and falls short (when TLS_REQUEST
		 * comes up size 2 rather than size 1 as draft revision 12 also
		 * mandates).
		 *
		 * The VENDOR_CLASS code in 3.0.x was 11, which is now the HBA
		 * code.  Both work out to be arbitrarily long text-or-byte
		 * strings, so they pass parsing.
		 *
		 * Note that it is possible (or intentional), if highly
		 * improbable, for the HBA bit array to exactly match
		 * isc-V3.0.x.  Warning here is not an issue; if it really is
		 * 3.0.x, there will be a protocol error later on.  If it isn't
		 * actually 3.0.x, then I guess the lucky user will have to
		 * live with a weird warning.
		 */
		if ((option_code == 11) && (option_len > 9) &&
		    (strncmp((const char *)op, "isc-V3.0.", 9) == 0)) {
		        log_error("WARNING: failover as of versions 3.1.0 and "
				  "on are not reverse compatible with "
				  "versions 3.0.x.");
		}

		goto out;
	}

	/* For values that require swapping, read them in one at a time
	   using routines that swap bytes. */
	for (i = 0; i < op_count; i++) {
		switch (ft_options [option_code].type) {
		      case FT_UINT32:
			omapi_connection_get_uint32 (c, (u_int32_t *)op);
			op += 4;
			link -> imsg_count += 4;
			break;

		      case FT_UINT16:
			omapi_connection_get_uint16 (c, (u_int16_t *)op);
			op += 2;
			link -> imsg_count += 2;
			break;

		      default:
			/* Everything else should have been handled
			   already. */
			log_error ("FAILOVER: option %s: bad type %d",
				   ft_options [option_code].name,
				   ft_options [option_code].type);
			return DHCP_R_PROTOCOLERROR;
		}
	}
      out:
	/* Remember that we got this option. */
	link -> imsg -> options_present |= ft_options [option_code].bit;
	return ISC_R_SUCCESS;
}

isc_result_t dhcp_failover_link_set_value (omapi_object_t *h,
					   omapi_object_t *id,
					   omapi_data_string_t *name,
					   omapi_typed_data_t *value)
{
	if (h -> type != omapi_type_protocol)
		return DHCP_R_INVALIDARG;

	/* Never valid to set these. */
	if (!omapi_ds_strcmp (name, "link-port") ||
	    !omapi_ds_strcmp (name, "link-name") ||
	    !omapi_ds_strcmp (name, "link-state"))
		return ISC_R_NOPERM;

	if (h -> inner && h -> inner -> type -> set_value)
		return (*(h -> inner -> type -> set_value))
			(h -> inner, id, name, value);
	return ISC_R_NOTFOUND;
}

isc_result_t dhcp_failover_link_get_value (omapi_object_t *h,
					   omapi_object_t *id,
					   omapi_data_string_t *name,
					   omapi_value_t **value)
{
	dhcp_failover_link_t *link;

	if (h -> type != omapi_type_protocol)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)h;

	if (!omapi_ds_strcmp (name, "link-port")) {
		return omapi_make_int_value (value, name,
					     (int)link -> peer_port, MDL);
	} else if (!omapi_ds_strcmp (name, "link-state")) {
		if (link -> state >= dhcp_flink_state_max)
			return omapi_make_string_value (value, name,
							"invalid link state",
							MDL);
		return omapi_make_string_value
			(value, name,
			 dhcp_flink_state_names [link -> state], MDL);
	}

	if (h -> inner && h -> inner -> type -> get_value)
		return (*(h -> inner -> type -> get_value))
			(h -> inner, id, name, value);
	return ISC_R_NOTFOUND;
}

isc_result_t dhcp_failover_link_destroy (omapi_object_t *h,
					 const char *file, int line)
{
	dhcp_failover_link_t *link;
	if (h -> type != dhcp_type_failover_link)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)h;

	if (link -> peer_address)
		option_cache_dereference (&link -> peer_address, file, line);
	if (link -> imsg)
		failover_message_dereference (&link -> imsg, file, line);
	if (link -> state_object)
		dhcp_failover_state_dereference (&link -> state_object,
						 file, line);
	return ISC_R_SUCCESS;
}

/* Write all the published values associated with the object through the
   specified connection. */

isc_result_t dhcp_failover_link_stuff_values (omapi_object_t *c,
					      omapi_object_t *id,
					      omapi_object_t *l)
{
	dhcp_failover_link_t *link;
	isc_result_t status;

	if (l -> type != dhcp_type_failover_link)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)l;

	status = omapi_connection_put_name (c, "link-port");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (int));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, link -> peer_port);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "link-state");
	if (status != ISC_R_SUCCESS)
		return status;
	if (link -> state >= dhcp_flink_state_max)
		status = omapi_connection_put_string (c, "invalid link state");
	else
		status = (omapi_connection_put_string
			  (c, dhcp_flink_state_names [link -> state]));
	if (status != ISC_R_SUCCESS)
		return status;

	if (link -> inner && link -> inner -> type -> stuff_values)
		return (*(link -> inner -> type -> stuff_values)) (c, id,
								link -> inner);
	return ISC_R_SUCCESS;
}

/* Set up a listener for the omapi protocol.    The handle stored points to
   a listener object, not a protocol object. */

isc_result_t dhcp_failover_listen (omapi_object_t *h)
{
	isc_result_t status;
	dhcp_failover_listener_t *obj, *l;
	omapi_value_t *value = (omapi_value_t *)0;
	omapi_addr_t local_addr;
	unsigned long port;

	status = omapi_get_value_str (h, (omapi_object_t *)0,
				      "local-port", &value);
	if (status != ISC_R_SUCCESS)
		return status;
	if (!value -> value) {
		omapi_value_dereference (&value, MDL);
		return DHCP_R_INVALIDARG;
	}

	status = omapi_get_int_value (&port, value -> value);
	omapi_value_dereference (&value, MDL);
	if (status != ISC_R_SUCCESS)
		return status;
	local_addr.port = port;

	status = omapi_get_value_str (h, (omapi_object_t *)0,
				      "local-address", &value);
	if (status != ISC_R_SUCCESS)
		return status;
	if (!value -> value) {
	      nogood:
		omapi_value_dereference (&value, MDL);
		return DHCP_R_INVALIDARG;
	}

	if (value -> value -> type != omapi_datatype_data ||
	    value -> value -> u.buffer.len != sizeof (struct in_addr))
		goto nogood;

	memcpy (local_addr.address, value -> value -> u.buffer.value,
		value -> value -> u.buffer.len);
	local_addr.addrlen = value -> value -> u.buffer.len;
	local_addr.addrtype = AF_INET;

	omapi_value_dereference (&value, MDL);

	/* Are we already listening on this port and address? */
	for (l = failover_listeners; l; l = l -> next) {
		if (l -> address.port == local_addr.port &&
		    l -> address.addrtype == local_addr.addrtype &&
		    l -> address.addrlen == local_addr.addrlen &&
		    !memcmp (l -> address.address, local_addr.address,
			     local_addr.addrlen))
			break;
	}
	/* Already listening. */
	if (l)
		return ISC_R_SUCCESS;

	obj = (dhcp_failover_listener_t *)0;
	status = dhcp_failover_listener_allocate (&obj, MDL);
	if (status != ISC_R_SUCCESS)
		return status;
	obj -> address = local_addr;

	status = omapi_listen_addr ((omapi_object_t *)obj, &obj -> address, 1);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_object_reference (&h -> outer,
					 (omapi_object_t *)obj, MDL);
	if (status != ISC_R_SUCCESS) {
		dhcp_failover_listener_dereference (&obj, MDL);
		return status;
	}
	status = omapi_object_reference (&obj -> inner, h, MDL);
	if (status != ISC_R_SUCCESS) {
		dhcp_failover_listener_dereference (&obj, MDL);
		return status;
	}

	/* Put this listener on the list. */
	if (failover_listeners) {
		dhcp_failover_listener_reference (&obj -> next,
						  failover_listeners, MDL);
		dhcp_failover_listener_dereference (&failover_listeners, MDL);
	}
	dhcp_failover_listener_reference (&failover_listeners, obj, MDL);

	return dhcp_failover_listener_dereference (&obj, MDL);
}

/* Signal handler for protocol listener - if we get a connect signal,
   create a new protocol connection, otherwise pass the signal down. */

isc_result_t dhcp_failover_listener_signal (omapi_object_t *o,
					    const char *name, va_list ap)
{
	isc_result_t status;
	omapi_connection_object_t *c;
	dhcp_failover_link_t *obj;
	dhcp_failover_listener_t *p;
	dhcp_failover_state_t *s, *state = (dhcp_failover_state_t *)0;

	if (!o || o -> type != dhcp_type_failover_listener)
		return DHCP_R_INVALIDARG;
	p = (dhcp_failover_listener_t *)o;

	/* Not a signal we recognize? */
	if (strcmp (name, "connect")) {
		if (p -> inner && p -> inner -> type -> signal_handler)
			return (*(p -> inner -> type -> signal_handler))
				(p -> inner, name, ap);
		return ISC_R_NOTFOUND;
	}

	c = va_arg (ap, omapi_connection_object_t *);
	if (!c || c -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	/* See if we can find a failover_state object that
	   matches this connection. */
	for (s = failover_states; s; s = s -> next) {
		if (dhcp_failover_state_match
		    (s, (u_int8_t *)&c -> remote_addr.sin_addr,
		    sizeof c -> remote_addr.sin_addr)) {
			state = s;
			break;
		}
	}
	if (!state) {
		log_info ("failover: listener: no matching state");
		omapi_disconnect ((omapi_object_t *)c, 1);
		return(ISC_R_NOTFOUND);
	}

	obj = (dhcp_failover_link_t *)0;
	status = dhcp_failover_link_allocate (&obj, MDL);
	if (status != ISC_R_SUCCESS)
		return status;
	obj -> peer_port = ntohs (c -> remote_addr.sin_port);

	status = omapi_object_reference (&obj -> outer,
					 (omapi_object_t *)c, MDL);
	if (status != ISC_R_SUCCESS) {
	      lose:
		dhcp_failover_link_dereference (&obj, MDL);
		log_info ("failover: listener: picayune failure.");
		omapi_disconnect ((omapi_object_t *)c, 1);
		return status;
	}

	status = omapi_object_reference (&c -> inner,
					 (omapi_object_t *)obj, MDL);
	if (status != ISC_R_SUCCESS)
		goto lose;

	status = dhcp_failover_state_reference (&obj -> state_object,
						state, MDL);
	if (status != ISC_R_SUCCESS)
		goto lose;

	omapi_signal_in ((omapi_object_t *)obj, "connect");

	return dhcp_failover_link_dereference (&obj, MDL);
}

isc_result_t dhcp_failover_listener_set_value (omapi_object_t *h,
						omapi_object_t *id,
						omapi_data_string_t *name,
						omapi_typed_data_t *value)
{
	if (h -> type != dhcp_type_failover_listener)
		return DHCP_R_INVALIDARG;

	if (h -> inner && h -> inner -> type -> set_value)
		return (*(h -> inner -> type -> set_value))
			(h -> inner, id, name, value);
	return ISC_R_NOTFOUND;
}

isc_result_t dhcp_failover_listener_get_value (omapi_object_t *h,
						omapi_object_t *id,
						omapi_data_string_t *name,
						omapi_value_t **value)
{
	if (h -> type != dhcp_type_failover_listener)
		return DHCP_R_INVALIDARG;

	if (h -> inner && h -> inner -> type -> get_value)
		return (*(h -> inner -> type -> get_value))
			(h -> inner, id, name, value);
	return ISC_R_NOTFOUND;
}

isc_result_t dhcp_failover_listener_destroy (omapi_object_t *h,
					      const char *file, int line)
{
	dhcp_failover_listener_t *l;

	if (h -> type != dhcp_type_failover_listener)
		return DHCP_R_INVALIDARG;
	l = (dhcp_failover_listener_t *)h;
	if (l -> next)
		dhcp_failover_listener_dereference (&l -> next, file, line);

	return ISC_R_SUCCESS;
}

/* Write all the published values associated with the object through the
   specified connection. */

isc_result_t dhcp_failover_listener_stuff (omapi_object_t *c,
					   omapi_object_t *id,
					   omapi_object_t *p)
{
	if (p -> type != dhcp_type_failover_listener)
		return DHCP_R_INVALIDARG;

	if (p -> inner && p -> inner -> type -> stuff_values)
		return (*(p -> inner -> type -> stuff_values)) (c, id,
								p -> inner);
	return ISC_R_SUCCESS;
}

/* Set up master state machine for the failover protocol. */

isc_result_t dhcp_failover_register (omapi_object_t *h)
{
	isc_result_t status;
	dhcp_failover_state_t *obj;
	unsigned long port;
	omapi_value_t *value = (omapi_value_t *)0;

	status = omapi_get_value_str (h, (omapi_object_t *)0,
				      "local-port", &value);
	if (status != ISC_R_SUCCESS)
		return status;
	if (!value -> value) {
		omapi_value_dereference (&value, MDL);
		return DHCP_R_INVALIDARG;
	}

	status = omapi_get_int_value (&port, value -> value);
	omapi_value_dereference (&value, MDL);
	if (status != ISC_R_SUCCESS)
		return status;

	obj = (dhcp_failover_state_t *)0;
	dhcp_failover_state_allocate (&obj, MDL);
	obj -> me.port = port;

	status = omapi_listen ((omapi_object_t *)obj, port, 1);
	if (status != ISC_R_SUCCESS) {
		dhcp_failover_state_dereference (&obj, MDL);
		return status;
	}

	status = omapi_object_reference (&h -> outer, (omapi_object_t *)obj,
					 MDL);
	if (status != ISC_R_SUCCESS) {
		dhcp_failover_state_dereference (&obj, MDL);
		return status;
	}
	status = omapi_object_reference (&obj -> inner, h, MDL);
	dhcp_failover_state_dereference (&obj, MDL);
	return status;
}

/* Signal handler for protocol state machine. */

isc_result_t dhcp_failover_state_signal (omapi_object_t *o,
					 const char *name, va_list ap)
{
	isc_result_t status;
	dhcp_failover_state_t *state;
	dhcp_failover_link_t *link;
	struct timeval tv;

	if (!o || o -> type != dhcp_type_failover_state)
		return DHCP_R_INVALIDARG;
	state = (dhcp_failover_state_t *)o;

	/* Not a signal we recognize? */
	if (strcmp (name, "disconnect") &&
	    strcmp (name, "message")) {
		if (state -> inner && state -> inner -> type -> signal_handler)
			return (*(state -> inner -> type -> signal_handler))
				(state -> inner, name, ap);
		return ISC_R_NOTFOUND;
	}

	/* Handle connect signals by seeing what state we're in
	   and potentially doing a state transition. */
	if (!strcmp (name, "disconnect")) {
		link = va_arg (ap, dhcp_failover_link_t *);

		dhcp_failover_link_dereference (&state -> link_to_peer, MDL);
		dhcp_failover_state_transition (state, "disconnect");
		if (state -> i_am == primary) {
#if defined (DEBUG_FAILOVER_TIMING)
			log_info ("add_timeout +90 %s",
				  "dhcp_failover_reconnect");
#endif
			tv . tv_sec = cur_time + 90;
			tv . tv_usec = 0;
			add_timeout (&tv, dhcp_failover_reconnect,
				     state,
				     (tvref_t)dhcp_failover_state_reference,
				     (tvunref_t)
				     dhcp_failover_state_dereference);
		}
	} else if (!strcmp (name, "message")) {
		link = va_arg (ap, dhcp_failover_link_t *);

		if (link -> imsg -> type == FTM_CONNECT) {
			/* If we already have a link to the peer, it must be
			   dead, so drop it.
			   XXX Is this the right thing to do?
			   XXX Probably not - what if both peers start at
			   XXX the same time? */
			if (state -> link_to_peer) {
				dhcp_failover_send_connectack
					((omapi_object_t *)link, state,
					 FTR_DUP_CONNECTION,
					 "already connected");
				omapi_disconnect (link -> outer, 1);
				return ISC_R_SUCCESS;
			}
			if (!(link -> imsg -> options_present & FTB_MCLT)) {
				dhcp_failover_send_connectack
					((omapi_object_t *)link, state,
					 FTR_INVALID_MCLT,
					 "no MCLT provided");
				omapi_disconnect (link -> outer, 1);
				return ISC_R_SUCCESS;
			}

			dhcp_failover_link_reference (&state -> link_to_peer,
						      link, MDL);
			status = (dhcp_failover_send_connectack
				  ((omapi_object_t *)link, state, 0, 0));
			if (status != ISC_R_SUCCESS) {
				dhcp_failover_link_dereference
					(&state -> link_to_peer, MDL);
				log_info ("dhcp_failover_send_connectack: %s",
					  isc_result_totext (status));
				omapi_disconnect (link -> outer, 1);
				return ISC_R_SUCCESS;
			}
			if (link -> imsg -> options_present & FTB_MAX_UNACKED)
				state -> partner.max_flying_updates =
					link -> imsg -> max_unacked;
			if (link -> imsg -> options_present & FTB_RECEIVE_TIMER)
				state -> partner.max_response_delay =
					link -> imsg -> receive_timer;
			state -> mclt = link -> imsg -> mclt;
			dhcp_failover_send_state (state);
			cancel_timeout (dhcp_failover_link_startup_timeout,
					link);
		} else if (link -> imsg -> type == FTM_CONNECTACK) {
		    const char *errmsg;
		    char errbuf[1024];
		    int reason;

		    cancel_timeout (dhcp_failover_link_startup_timeout,
				    link);

		    if (!(link->imsg->options_present &
			  FTB_RELATIONSHIP_NAME)) {
			errmsg = "missing relationship-name";
			reason = FTR_INVALID_PARTNER;
			goto badconnectack;
		    }

		    if (link->imsg->options_present & FTB_REJECT_REASON) {
			/* XXX: add message option to text output. */
			log_error ("Failover CONNECT to %s rejected: %s",
				   state ? state->name : "unknown",
				   (dhcp_failover_reject_reason_print
				    (link -> imsg -> reject_reason)));
			/* XXX print message from peer if peer sent message. */
			omapi_disconnect (link -> outer, 1);
			return ISC_R_SUCCESS;
		    }

		    if (!dhcp_failover_state_match_by_name(state,
			&link->imsg->relationship_name)) {
			/* XXX: Overflow results in log truncation, safe. */
			snprintf(errbuf, sizeof(errbuf), "remote failover "
				 "relationship name %.*s does not match",
				 (int)link->imsg->relationship_name.count,
				 link->imsg->relationship_name.data);
			errmsg = errbuf;
			reason = FTR_INVALID_PARTNER;
		      badconnectack:
			log_error("Failover CONNECTACK from %s: %s",
				  state->name, errmsg);
			dhcp_failover_send_disconnect ((omapi_object_t *)link,
						       reason, errmsg);
			omapi_disconnect (link -> outer, 0);
			return ISC_R_SUCCESS;
		    }

		    if (state -> link_to_peer) {
			errmsg = "already connected";
			reason = FTR_DUP_CONNECTION;
			goto badconnectack;
		    }

		    if ((cur_time > link -> imsg -> time &&
			 cur_time - link -> imsg -> time > 60) ||
			(cur_time < link -> imsg -> time &&
			 link -> imsg -> time - cur_time > 60)) {
			    errmsg = "time offset too large";
			    reason = FTR_TIMEMISMATCH;
			    goto badconnectack;
		    }

		    dhcp_failover_link_reference (&state -> link_to_peer,
						  link, MDL);
#if 0
		    /* XXX This is probably the right thing to do, but
		       XXX for release three, to make the smallest possible
		       XXX change, we are doing this when the peer state
		       XXX changes instead. */
		    if (state -> me.state == startup)
			    dhcp_failover_set_state (state,
						     state -> saved_state);
		    else
#endif
			    dhcp_failover_send_state (state);

		    if (link -> imsg -> options_present & FTB_MAX_UNACKED)
			    state -> partner.max_flying_updates =
				    link -> imsg -> max_unacked;
		    if (link -> imsg -> options_present & FTB_RECEIVE_TIMER)
			    state -> partner.max_response_delay =
				    link -> imsg -> receive_timer;
#if defined (DEBUG_FAILOVER_CONTACT_TIMING)
		    log_info ("add_timeout +%d %s",
			      (int)state -> partner.max_response_delay / 3,
			      "dhcp_failover_send_contact");
#endif
		    tv . tv_sec = cur_time +
			    (int)state -> partner.max_response_delay / 3;
		    tv . tv_usec = 0;
		    add_timeout (&tv,
				 dhcp_failover_send_contact, state,
				 (tvref_t)dhcp_failover_state_reference,
				 (tvunref_t)dhcp_failover_state_dereference);
#if defined (DEBUG_FAILOVER_CONTACT_TIMING)
		    log_info ("add_timeout +%d %s",
			      (int)state -> me.max_response_delay,
			      "dhcp_failover_timeout");
#endif
		    tv . tv_sec = cur_time +
			    (int)state -> me.max_response_delay;
		    tv . tv_usec = 0;
		    add_timeout (&tv,
				 dhcp_failover_timeout, state,
				 (tvref_t)dhcp_failover_state_reference,
				 (tvunref_t)dhcp_failover_state_dereference);
		} else if (link -> imsg -> type == FTM_DISCONNECT) {
		    if (link -> imsg -> reject_reason) {
			log_error ("Failover DISCONNECT from %s: %s",
				   state ? state->name : "unknown",
				   (dhcp_failover_reject_reason_print
				    (link -> imsg -> reject_reason)));
		    }
		    omapi_disconnect (link -> outer, 1);
		} else if (link -> imsg -> type == FTM_BNDUPD) {
			dhcp_failover_process_bind_update (state,
							   link -> imsg);
		} else if (link -> imsg -> type == FTM_BNDACK) {
			dhcp_failover_process_bind_ack (state, link -> imsg);
		} else if (link -> imsg -> type == FTM_UPDREQ) {
			dhcp_failover_process_update_request (state,
							      link -> imsg);
		} else if (link -> imsg -> type == FTM_UPDREQALL) {
			dhcp_failover_process_update_request_all
				(state, link -> imsg);
		} else if (link -> imsg -> type == FTM_UPDDONE) {
			dhcp_failover_process_update_done (state,
							   link -> imsg);
		} else if (link -> imsg -> type == FTM_POOLREQ) {
			dhcp_failover_pool_reqbalance(state);
		} else if (link -> imsg -> type == FTM_POOLRESP) {
			log_info ("pool response: %ld leases",
				  (unsigned long)
				  link -> imsg -> addresses_transferred);
		} else if (link -> imsg -> type == FTM_STATE) {
			dhcp_failover_peer_state_changed (state,
							  link -> imsg);
		}

		/* Add a timeout so that if the partner doesn't send
		   another message for the maximum transmit idle time
		   plus a grace of one second, we close the
		   connection. */
		if (state -> link_to_peer &&
		    state -> link_to_peer == link &&
		    state -> link_to_peer -> state != dhcp_flink_disconnected)
		{
#if defined (DEBUG_FAILOVER_CONTACT_TIMING)
		    log_info ("add_timeout +%d %s",
			      (int)state -> me.max_response_delay,
			      "dhcp_failover_timeout");
#endif
		    tv . tv_sec = cur_time +
			    (int)state -> me.max_response_delay;
		    tv . tv_usec = 0;
		    add_timeout (&tv,
				 dhcp_failover_timeout, state,
				 (tvref_t)dhcp_failover_state_reference,
				 (tvunref_t)dhcp_failover_state_dereference);

		}
	}

	/* Handle all the events we care about... */
	return ISC_R_SUCCESS;
}

isc_result_t dhcp_failover_state_transition (dhcp_failover_state_t *state,
					     const char *name)
{
	isc_result_t status;

	/* XXX Check these state transitions against the spec! */
	if (!strcmp (name, "disconnect")) {
		if (state -> link_to_peer) {
		    log_info ("peer %s: disconnected", state -> name);
		    if (state -> link_to_peer -> state_object)
			dhcp_failover_state_dereference
				(&state -> link_to_peer -> state_object, MDL);
		    dhcp_failover_link_dereference (&state -> link_to_peer,
						    MDL);
		}
		cancel_timeout (dhcp_failover_send_contact, state);
		cancel_timeout (dhcp_failover_timeout, state);
		cancel_timeout (dhcp_failover_startup_timeout, state);

		switch (state -> me.state == startup ?
			state -> saved_state : state -> me.state) {
		      /* In these situations, we remain in the current
		       * state, or if in startup enter those states.
		       */
		      case conflict_done:
			/* As the peer may not have received or may have
			 * lost track of updates we sent previously we
			 * rescind them, causing us to retransmit them
			 * on an update request.
			 */
			dhcp_failover_rescind_updates(state);
			/* fall through */

		      case communications_interrupted:
		      case partner_down:
		      case paused:
		      case recover:
		      case recover_done:
		      case recover_wait:
		      case resolution_interrupted:
		      case shut_down:
			/* Already in the right state? */
			if (state -> me.state == startup)
				return (dhcp_failover_set_state
					(state, state -> saved_state));
			return ISC_R_SUCCESS;

		      case potential_conflict:
			return dhcp_failover_set_state
				(state, resolution_interrupted);

		      case normal:
			return dhcp_failover_set_state
				(state, communications_interrupted);

		      case unknown_state:
			return dhcp_failover_set_state
				(state, resolution_interrupted);

		      default:
			log_fatal("Impossible case at %s:%d.", MDL);
			break;	/* can't happen. */
		}
	} else if (!strcmp (name, "connect")) {
		switch (state -> me.state) {
		      case communications_interrupted:
			status = dhcp_failover_set_state (state, normal);
			dhcp_failover_send_updates (state);
			return status;

		      case resolution_interrupted:
			return dhcp_failover_set_state (state,
							potential_conflict);

		      case conflict_done:
		      case partner_down:
		      case potential_conflict:
		      case normal:
		      case recover:
		      case shut_down:
		      case paused:
		      case unknown_state:
		      case recover_done:
		      case startup:
		      case recover_wait:
			return dhcp_failover_send_state (state);

		      default:
			log_fatal("Impossible case at %s:%d.", MDL);
			break;
		}
	} else if (!strcmp (name, "startup")) {
		dhcp_failover_set_state (state, startup);
		return ISC_R_SUCCESS;
	} else if (!strcmp (name, "connect-timeout")) {
		switch (state -> me.state) {
		      case communications_interrupted:
		      case partner_down:
		      case resolution_interrupted:
		      case paused:
		      case startup:
		      case shut_down:
		      case conflict_done:
			return ISC_R_SUCCESS;

		      case normal:
		      case recover:
		      case recover_wait:
		      case recover_done:
		      case unknown_state:
			return dhcp_failover_set_state
				(state, communications_interrupted);

		      case potential_conflict:
			return dhcp_failover_set_state
				(state, resolution_interrupted);

		      default:
			log_fatal("Impossible case at %s:%d.", MDL);
			break;
		}
	}
	return DHCP_R_INVALIDARG;
}

isc_result_t dhcp_failover_set_service_state (dhcp_failover_state_t *state)
{
	switch (state -> me.state) {
	      case unknown_state:
		state -> service_state = not_responding;
		state -> nrr = " (my state unknown)";
		break;

	      case partner_down:
		state -> service_state = service_partner_down;
		state -> nrr = "";
		break;

	      case normal:
		state -> service_state = cooperating;
		state -> nrr = "";
		break;

	      case communications_interrupted:
		state -> service_state = not_cooperating;
		state -> nrr = "";
		break;

	      case resolution_interrupted:
	      case potential_conflict:
	      case conflict_done:
		state -> service_state = not_responding;
		state -> nrr = " (resolving conflicts)";
		break;

	      case recover:
		state -> service_state = not_responding;
		state -> nrr = " (recovering)";
		break;

	      case shut_down:
		state -> service_state = not_responding;
		state -> nrr = " (shut down)";
		break;

	      case paused:
		state -> service_state = not_responding;
		state -> nrr = " (paused)";
		break;

	      case recover_wait:
		state -> service_state = not_responding;
		state -> nrr = " (recover wait)";
		break;

	      case recover_done:
		state -> service_state = not_responding;
		state -> nrr = " (recover done)";
		break;

	      case startup:
		state -> service_state = service_startup;
		state -> nrr = " (startup)";
		break;

	      default:
		log_fatal("Impossible case at %s:%d.\n", MDL);
		break;
	}

	/* Some peer states can require us not to respond, even if our
	   state doesn't. */
	/* XXX hm.   I suspect this isn't true anymore. */
	if (state -> service_state != not_responding) {
		switch (state -> partner.state) {
		      case partner_down:
			state -> service_state = not_responding;
			state -> nrr = " (peer demands: recovering)";
			break;

		      case potential_conflict:
		      case conflict_done:
		      case resolution_interrupted:
			state -> service_state = not_responding;
			state -> nrr = " (peer demands: resolving conflicts)";
			break;

			/* Other peer states don't affect our behaviour. */
		      default:
			break;
		}
	}

	return ISC_R_SUCCESS;
}

/*!
 * \brief Return any leases on the ack queue back to the update queue
 *
 * Re-schedule any pending updates by moving them from the ack queue
 * (update sent awaiting response) back to the update queue (need to
 * send an update for this lease).  This will result in a retransmission
 * of the update.
 *
 * \param state is the state block for the failover connection we are
 * updating.
 */

void dhcp_failover_rescind_updates (dhcp_failover_state_t *state)
{
    struct lease *lp;

    if (state->ack_queue_tail == NULL)
	    return;

    /* Zap the flags. */
    for (lp = state->ack_queue_head; lp; lp = lp->next_pending)
	    lp->flags = ((lp->flags & ~ON_ACK_QUEUE) | ON_UPDATE_QUEUE);

    /* Now hook the ack queue to the beginning of the update queue. */
    if (state->update_queue_head) {
	    lease_reference(&state->ack_queue_tail->next_pending,
			    state->update_queue_head, MDL);
	    lease_dereference(&state->update_queue_head, MDL);
    }
    lease_reference(&state->update_queue_head, state->ack_queue_head, MDL);

    if (!state->update_queue_tail) {
#if defined (POINTER_DEBUG)
	    if (state->ack_queue_tail->next_pending) {
		    log_error("next pending on ack queue tail.");
		    abort();
	    }
#endif
	    lease_reference(&state->update_queue_tail,
			    state->ack_queue_tail, MDL);
    }
    lease_dereference(&state->ack_queue_tail, MDL);
    lease_dereference(&state->ack_queue_head, MDL);
    state->cur_unacked_updates = 0;
}

isc_result_t dhcp_failover_set_state (dhcp_failover_state_t *state,
				      enum failover_state new_state)
{
    enum failover_state saved_state;
    TIME saved_stos;
    struct pool *p;
    struct shared_network *s;
    struct lease *l;
    struct timeval tv;

    /* If we're in certain states where we're sending updates, and the peer
     * state changes, we need to re-schedule any pending updates just to
     * be on the safe side.  This results in retransmission.
     */
    switch (state -> me.state) {
      case normal:
      case potential_conflict:
      case partner_down:
	/* Move the ack queue to the update queue */
	dhcp_failover_rescind_updates(state);

	/* We will re-queue a timeout later, if applicable. */
	cancel_timeout (dhcp_failover_keepalive, state);
	break;

      default:
	break;
    }

    /* Tentatively make the transition. */
    saved_state = state -> me.state;
    saved_stos = state -> me.stos;

    /* Keep the old stos if we're going into recover_wait or if we're
       coming into or out of startup. */
    if (new_state != recover_wait && new_state != startup &&
	saved_state != startup)
	    state -> me.stos = cur_time;

    /* If we're in shutdown, peer is in partner_down, and we're moving
       to recover, we can skip waiting for MCLT to expire.    This happens
       when a server is moved administratively into shutdown prior to
       actually shutting down.   Of course, if there are any updates
       pending we can't actually do this. */
    if (new_state == recover && saved_state == shut_down &&
	state -> partner.state == partner_down &&
	!state -> update_queue_head && !state -> ack_queue_head)
	    state -> me.stos = cur_time - state -> mclt;

    state -> me.state = new_state;
    if (new_state == startup && saved_state != startup)
	state -> saved_state = saved_state;

    /* If we can't record the new state, we can't make a state transition. */
    if (!write_failover_state (state) || !commit_leases ()) {
	    log_error ("Unable to record current failover state for %s",
		       state -> name);
	    state -> me.state = saved_state;
	    state -> me.stos = saved_stos;
	    return ISC_R_IOERROR;
    }

    log_info ("failover peer %s: I move from %s to %s",
	      state -> name, dhcp_failover_state_name_print (saved_state),
	      dhcp_failover_state_name_print (state -> me.state));

    /* If both servers are now normal log it */
    if ((state->me.state == normal) && (state->partner.state == normal))
	    log_info("failover peer %s: Both servers normal", state->name);

    /* If we were in startup and we just left it, cancel the timeout. */
    if (new_state != startup && saved_state == startup)
	cancel_timeout (dhcp_failover_startup_timeout, state);

    /*
     * If the state changes for any reason, cancel 'delayed auto state
     * changes' (currently there is just the one).
     */
    cancel_timeout(dhcp_failover_auto_partner_down, state);

    /* Set our service state. */
    dhcp_failover_set_service_state (state);

    /* Tell the peer about it. */
    if (state -> link_to_peer)
	    dhcp_failover_send_state (state);

    switch (new_state) {
	  case communications_interrupted:
	    /*
	     * There is an optional feature to automatically enter partner
	     * down after a timer expires, upon entering comms-interrupted.
	     * This feature is generally not safe except in specific
	     * circumstances.
	     *
	     * A zero value (also the default) disables it.
	     */
	    if (state->auto_partner_down == 0)
		break;

#if defined (DEBUG_FAILOVER_TIMING)
	    log_info("add_timeout +%lu dhcp_failover_auto_partner_down",
		      (unsigned long)state->auto_partner_down);
#endif
	    tv.tv_sec = cur_time + state->auto_partner_down;
	    tv.tv_usec = 0;
	    add_timeout(&tv, dhcp_failover_auto_partner_down, state,
			(tvref_t)omapi_object_reference,
			(tvunref_t)omapi_object_dereference);
	    break;

	  case normal:
	    /* Upon entering normal state, the server is expected to retransmit
	     * all pending binding updates.  This is a good opportunity to
	     * rebalance the pool (potentially making new pending updates),
	     * which also schedules the next pool rebalance.
	     */
	    dhcp_failover_pool_balance(state);
	    dhcp_failover_generate_update_queue(state, 0);

	    if (state->update_queue_tail != NULL) {
		dhcp_failover_send_updates(state);
		log_info("Sending updates to %s.", state->name);
	    }

	    break;

	  case potential_conflict:
	    if ((state->i_am == primary) ||
		((state->i_am == secondary) &&
		 (state->partner.state == conflict_done)))
		    dhcp_failover_send_update_request (state);
	    break;

	  case startup:
#if defined (DEBUG_FAILOVER_TIMING)
	    log_info ("add_timeout +15 %s",
		      "dhcp_failover_startup_timeout");
#endif
	    tv . tv_sec = cur_time + 15;
	    tv . tv_usec = 0;
	    add_timeout (&tv,
			 dhcp_failover_startup_timeout,
			 state,
			 (tvref_t)omapi_object_reference,
			 (tvunref_t)
			 omapi_object_dereference);
	    break;

	    /* If we come back in recover_wait and there's still waiting
	       to do, set a timeout. */
	  case recover_wait:
	    if (state -> me.stos + state -> mclt > cur_time) {
#if defined (DEBUG_FAILOVER_TIMING)
		    log_info ("add_timeout +%d %s",
			      (int)(cur_time -
				    state -> me.stos + state -> mclt),
			      "dhcp_failover_startup_timeout");
#endif
		    tv . tv_sec = (int)(state -> me.stos + state -> mclt);
		    tv . tv_usec = 0;
		    add_timeout (&tv,
				 dhcp_failover_recover_done,
				 state,
				 (tvref_t)omapi_object_reference,
				 (tvunref_t)
				 omapi_object_dereference);
	    } else
		    dhcp_failover_recover_done (state);
	    break;

	  case recover:
	    /* XXX: We're supposed to calculate if updreq or updreqall is
	     * needed.  In theory, we should only have to updreqall if we
	     * are positive we lost our stable storage.
	     */
	    if (state -> link_to_peer)
		    dhcp_failover_send_update_request_all (state);
	    break;

	  case partner_down:
	    /* For every expired lease, set a timeout for it to become free. */
	    for (s = shared_networks; s; s = s->next) {
		for (p = s->pools; p; p = p->next) {
#if defined (BINARY_LEASES)
		    long int tiebreaker = 0;
#endif
		    if (p->failover_peer == state) {
			for (l = LEASE_GET_FIRST(p->expired);
			     l != NULL;
			     l = LEASE_GET_NEXT(p->expired, l)) {
			    l->tsfp = state->me.stos + state->mclt;
			    l->sort_time = (l->tsfp > l->ends) ?
					   l->tsfp : l->ends;
#if defined (BINARY_LEASES)
			    /* If necessary fix up the tiebreaker so the leases
			     * maintain proper sort order.
			     */
			    l->sort_tiebreaker = tiebreaker;
			    if (tiebreaker != LONG_MAX)
			        tiebreaker++;
#endif

			}

			l = LEASE_GET_FIRST(p->expired);
			if (l && (l->sort_time < p->next_event_time)) {

			    p->next_event_time = l->sort_time;
#if defined (DEBUG_FAILOVER_TIMING)
			    log_info ("add_timeout +%d %s",
				      (int)(cur_time - p->next_event_time),
				      "pool_timer");
#endif
			    tv.tv_sec = p->next_event_time;
			    tv.tv_usec = 0;
			    add_timeout(&tv, pool_timer, p,
					(tvref_t)pool_reference,
					(tvunref_t)pool_dereference);
			}
		    }
		}
	    }
	    break;

	  default:
	    break;
    }

    return ISC_R_SUCCESS;
}

isc_result_t dhcp_failover_peer_state_changed (dhcp_failover_state_t *state,
					       failover_message_t *msg)
{
	enum failover_state previous_state = state -> partner.state;
	enum failover_state new_state;
	int startupp;

	new_state = msg -> server_state;
	startupp = (msg -> server_flags & FTF_SERVER_STARTUP) ? 1 : 0;

	if (state -> partner.state == new_state && state -> me.state) {
		switch (state -> me.state) {
		      case startup:
			/*
			 * If we have a peer state we must be connected.
			 * If so we should move to potential_conflict
			 * instead of resolution_interrupted, otherwise
			 * back to whereever we were before we stopped.
			 */
			if (state->saved_state == resolution_interrupted)
				dhcp_failover_set_state(state,
							potential_conflict);
			else
				dhcp_failover_set_state(state,
							state->saved_state);
			return ISC_R_SUCCESS;

		      case unknown_state:
		      case normal:
		      case potential_conflict:
		      case recover_done:
		      case shut_down:
		      case paused:
		      case recover_wait:
			return ISC_R_SUCCESS;

			/* If we get a peer state change when we're
			   disconnected, we always process it. */
		      case partner_down:
		      case communications_interrupted:
		      case resolution_interrupted:
		      case recover:
		      case conflict_done:
			break;

		      default:
			log_fatal("Impossible case at %s:%d.", MDL);
			break;
		}
	}

	state -> partner.state = new_state;
	state -> partner.stos = cur_time;

	log_info ("failover peer %s: peer moves from %s to %s",
		  state -> name,
		  dhcp_failover_state_name_print (previous_state),
		  dhcp_failover_state_name_print (state -> partner.state));

	/* If both servers are now normal log it */
	if ((state->me.state == normal) && (state->partner.state == normal))
		log_info("failover peer %s: Both servers normal", state->name);

	if (!write_failover_state (state) || !commit_leases ()) {
		/* This is bad, but it's not fatal.  Of course, if we
		   can't write to the lease database, we're not going to
		   get much done anyway. */
		log_error ("Unable to record current failover state for %s",
			   state -> name);
	}

	/* Quickly validate the new state as being one of the 13 known
	 * states.
	 */
	switch (new_state) {
	      case unknown_state:
	      case startup:
	      case normal:
	      case communications_interrupted:
	      case partner_down:
	      case potential_conflict:
	      case recover:
	      case paused:
	      case shut_down:
	      case recover_done:
	      case resolution_interrupted:
	      case conflict_done:
	      case recover_wait:
		break;

	      default:
		log_error("failover peer %s: Invalid state: %d", state->name,
			  new_state);
		dhcp_failover_set_state(state, shut_down);
		return ISC_R_SUCCESS;
	}

	/* Do any state transitions that are required as a result of the
	   peer's state transition. */

	switch (state -> me.state == startup ?
		state -> saved_state : state -> me.state) {
	      case normal:
		switch (new_state) {
		      case normal:
			dhcp_failover_state_pool_check (state);
			break;

		      case partner_down:
			if (state -> me.state == startup)
				dhcp_failover_set_state (state, recover);
			else
				dhcp_failover_set_state (state,
							 potential_conflict);
			break;

		      case potential_conflict:
		      case resolution_interrupted:
		      case conflict_done:
			/* None of these transitions should ever occur. */
			log_error("Peer %s: Invalid state transition %s "
				"to %s.", state->name,
				dhcp_failover_state_name_print(previous_state),
				dhcp_failover_state_name_print(new_state));
			dhcp_failover_set_state (state, shut_down);
			break;

		      case recover:
		      case shut_down:
			dhcp_failover_set_state (state, partner_down);
			break;

		      case paused:
			dhcp_failover_set_state (state,
						 communications_interrupted);
			break;

		      default:
			/* recover_wait, recover_done, unknown_state, startup,
			 * communications_interrupted
			 */
			break;
		}
		break;

	      case recover:
		switch (new_state) {
		      case recover:
			log_info ("failover peer %s: requesting %s",
				  state -> name, "full update from peer");
			/* Don't send updreqall if we're really in the
			   startup state, because that will result in two
			   being sent. */
			if (state -> me.state == recover)
				dhcp_failover_send_update_request_all (state);
			break;

		      case potential_conflict:
		      case resolution_interrupted:
		      case conflict_done:
		      case normal:
			dhcp_failover_set_state (state, potential_conflict);
			break;

		      case partner_down:
		      case communications_interrupted:
			/* We're supposed to send an update request at this
			   point. */
			/* XXX we don't currently have code here to do any
			   XXX clever detection of when we should send an
			   XXX UPDREQALL message rather than an UPDREQ
			   XXX message.   What to do, what to do? */
			/* Currently when we enter recover state, no matter
			 * the reason, we send an UPDREQALL.  So, it makes
			 * the most sense to stick to that until something
			 * better is done.
			 * Furthermore, we only want to send the update
			 * request if we are not in startup state.
			 */
			if (state -> me.state == recover)
				dhcp_failover_send_update_request_all (state);
			break;

		      case shut_down:
			/* XXX We're not explicitly told what to do in this
			   XXX case, but this transition is consistent with
			   XXX what is elsewhere in the draft. */
			dhcp_failover_set_state (state, partner_down);
			break;

			/* We can't really do anything in this case. */
		      default:
			/* paused, recover_done, recover_wait, unknown_state,
			 * startup.
			 */
			break;
		}
		break;

	      case potential_conflict:
		switch (new_state) {
		      case normal:
			/* This is an illegal transition. */
			log_error("Peer %s moves to normal during conflict "
				  "resolution - panic, shutting down.",
				  state->name);
			dhcp_failover_set_state(state, shut_down);
			break;

		      case conflict_done:
			if (previous_state == potential_conflict)
				dhcp_failover_send_update_request (state);
			else
				log_error("Peer %s: Unexpected move to "
					  "conflict-done.", state->name);
			break;

		      case recover_done:
		      case recover_wait:
		      case potential_conflict:
		      case partner_down:
		      case communications_interrupted:
		      case resolution_interrupted:
		      case paused:
			break;

		      case recover:
			dhcp_failover_set_state (state, recover);
			break;

		      case shut_down:
			dhcp_failover_set_state (state, partner_down);
			break;

		      default:
			/* unknown_state, startup */
			break;
		}
		break;

	      case conflict_done:
		switch (new_state) {
		      case normal:
		      case shut_down:
			dhcp_failover_set_state(state, new_state);
			break;

		      case potential_conflict:
		      case resolution_interrupted:
			/*
			 * This can happen when the connection is lost and
			 * recovered after the primary has moved to
			 * conflict-done but the secondary is still in
			 * potential-conflict.  In that case, we have to
			 * remain in conflict-done.
			 */
			break;

		      default:
			log_fatal("Peer %s: Invalid attempt to move from %s "
				"to %s while local state is conflict-done.",
				state->name,
				dhcp_failover_state_name_print(previous_state),
				dhcp_failover_state_name_print(new_state));
		}
		break;

	      case partner_down:
		/* Take no action if other server is starting up. */
		if (startupp)
			break;

		switch (new_state) {
			/* This is where we should be. */
		      case recover:
		      case recover_wait:
			break;

		      case recover_done:
			dhcp_failover_set_state (state, normal);
			break;

		      case normal:
		      case potential_conflict:
		      case partner_down:
		      case communications_interrupted:
		      case resolution_interrupted:
		      case conflict_done:
			dhcp_failover_set_state (state, potential_conflict);
			break;

		      default:
			/* shut_down, paused, unknown_state, startup */
			break;
		}
		break;

	      case communications_interrupted:
		switch (new_state) {
		      case paused:
			/* Stick with the status quo. */
			break;

			/* If we're in communications-interrupted and an
			   amnesic peer connects, go to the partner_down
			   state immediately. */
		      case recover:
			dhcp_failover_set_state (state, partner_down);
			break;

		      case normal:
		      case communications_interrupted:
		      case recover_done:
		      case recover_wait:
			/* XXX so we don't need to do this specially in
			   XXX the CONNECT and CONNECTACK handlers. */
			dhcp_failover_send_updates (state);
			dhcp_failover_set_state (state, normal);
			break;

		      case potential_conflict:
		      case partner_down:
		      case resolution_interrupted:
		      case conflict_done:
			dhcp_failover_set_state (state, potential_conflict);
			break;

		      case shut_down:
			dhcp_failover_set_state (state, partner_down);
			break;

		      default:
			/* unknown_state, startup */
			break;
		}
		break;

	      case resolution_interrupted:
		switch (new_state) {
		      case normal:
		      case recover:
		      case potential_conflict:
		      case partner_down:
		      case communications_interrupted:
		      case resolution_interrupted:
		      case conflict_done:
		      case recover_done:
		      case recover_wait:
			dhcp_failover_set_state (state, potential_conflict);
			break;

		      case shut_down:
			dhcp_failover_set_state (state, partner_down);
			break;

		      default:
			/* paused, unknown_state, startup */
			break;
		}
		break;

	      /* Make no transitions while in recover_wait...just wait. */
	      case recover_wait:
		break;

	      case recover_done:
		switch (new_state) {
		      case recover_done:
			log_error("Both servers have entered recover-done!");
			/* Fall through and tranistion to normal anyway */

		      case normal:
			dhcp_failover_set_state (state, normal);
			break;

		      case shut_down:
			dhcp_failover_set_state (state, partner_down);
			break;

		      default:
			/* potential_conflict, partner_down,
			 * communications_interrupted, resolution_interrupted,
			 * paused, recover, recover_wait, unknown_state,
			 * startup.
			 */
			break;
		}
		break;

		/* We are essentially dead in the water when we're in
		   either shut_down or paused states, and do not do any
		   automatic state transitions. */
	      case shut_down:
	      case paused:
		break;

	      /* XXX: Shouldn't this be a fatal condition? */
	      case unknown_state:
		break;

	      default:
		log_fatal("Impossible condition at %s:%d.", MDL);
		break;

	}

	/* If we didn't make a transition out of startup as a result of
	   the peer's state change, do it now as a result of the fact that
	   we got a state change from the peer. */
	if (state -> me.state == startup && state -> saved_state != startup)
		dhcp_failover_set_state (state, state -> saved_state);

	/* For now, just set the service state based on the peer's state
	   if necessary. */
	dhcp_failover_set_service_state (state);

	return ISC_R_SUCCESS;
}

/*
 * Balance operation manual entry; startup, entrance to normal state.  No
 * sense sending a POOLREQ at this stage; the peer is likely about to schedule
 * their own rebalance event upon entering normal themselves.
 */
static void
dhcp_failover_pool_balance(dhcp_failover_state_t *state)
{
	/* Cancel pending event. */
	cancel_timeout(dhcp_failover_pool_rebalance, state);
	state->sched_balance = 0;

	dhcp_failover_pool_dobalance(state, NULL);
}

/*
 * Balance operation entry from timer event.  Once per timer interval is
 * the only time we want to emit POOLREQs (asserting an interrupt in our
 * peer).
 */
void
dhcp_failover_pool_rebalance(void *failover_state)
{
	dhcp_failover_state_t *state;
	isc_boolean_t sendreq = ISC_FALSE;

	state = (dhcp_failover_state_t *)failover_state;

	/* Clear scheduled event indicator. */
	state->sched_balance = 0;

	if (dhcp_failover_pool_dobalance(state, &sendreq))
		dhcp_failover_send_updates(state);

	if (sendreq)
		dhcp_failover_send_poolreq(state);
}

/*
 * Balance operation entry from POOLREQ protocol message.  Do not permit a
 * POOLREQ to send back a POOLREQ.  Ping pong.
 */
static void
dhcp_failover_pool_reqbalance(dhcp_failover_state_t *state)
{
	int queued;

	/* Cancel pending event. */
	cancel_timeout(dhcp_failover_pool_rebalance, state);
	state->sched_balance = 0;

	queued = dhcp_failover_pool_dobalance(state, NULL);

	dhcp_failover_send_poolresp(state, queued);

	if (queued)
		dhcp_failover_send_updates(state);
	else
		log_info("peer %s: Got POOLREQ, answering negatively!  "
			 "Peer may be out of leases or database inconsistent.",
			 state->name);
}

/*
 * Do the meat of the work common to all forms of pool rebalance.  If the
 * caller deems it appropriate to transmit POOLREQ messages, it can use the
 * sendreq pointer to pass in the address of a FALSE value which this function
 * will conditionally turn TRUE if a POOLREQ is determined to be necessary.
 * A NULL value may be passed, in which case no action is taken.
 */
static int
dhcp_failover_pool_dobalance(dhcp_failover_state_t *state,
			    isc_boolean_t *sendreq)
{
	int lts, total, thresh, hold, panic, pass;
	int leases_queued = 0;
	struct lease *lp = NULL;
	struct lease *next = NULL;
	struct lease *ltemp = NULL;
	struct shared_network *s;
	struct pool *p;
	binding_state_t peer_lease_state;
	/* binding_state_t my_lease_state; */
        /* XXX Why is this my_lease_state never used? */
	LEASE_STRUCT_PTR lq;
	int (*log_func)(const char *, ...);
	const char *result, *reqlog;

	if (state -> me.state != normal)
		return 0;

	state->last_balance = cur_time;

	for (s = shared_networks ; s ; s = s->next) {
	    for (p = s->pools ; p ; p = p->next) {
		if (p->failover_peer != state)
		    continue;

		/* Right now we're giving the peer half of the free leases.
		   If we have more leases than the peer (i.e., more than
		   half), then the number of leases we have, less the number
		   of leases the peer has, will be how many more leases we
		   have than the peer has.   So if we send half that number
		   to the peer, we should be even. */
		if (p->failover_peer->i_am == primary) {
			lts = (p->free_leases - p->backup_leases) / 2;
			peer_lease_state = FTS_BACKUP;
			/* my_lease_state = FTS_FREE; */
			lq = &p->free;
		} else {
			lts = (p->backup_leases - p->free_leases) / 2;
			peer_lease_state = FTS_FREE;
			/* my_lease_state = FTS_BACKUP; */
			lq = &p->backup;
		}

		total = p->backup_leases + p->free_leases;

		thresh = ((total * state->max_lease_misbalance) + 50) / 100;
		hold = ((total * state->max_lease_ownership) + 50) / 100;

		/*
		 * If we need leases (so lts is negative) more than negative
		 * double the thresh%, panic and send poolreq to hopefully wake
		 * up the peer (but more likely the db is inconsistent).  But,
		 * if this comes out zero, switch to -1 so that the POOLREQ is
		 * sent on lts == -2 rather than right away at -1.
		 *
		 * Note that we do not subtract -1 from panic all the time
		 * because thresh% and hold% may come out to the same number,
		 * and that is correct operation...where thresh% and hold% are
		 * both -1, we want to send poolreq when lts reaches -3.  So,
		 * "-3 < -2", lts < panic.
		 */
		panic = thresh * -2;

		if (panic == 0)
			panic = -1;

		if ((sendreq != NULL) && (lts < panic)) {
			reqlog = "  (requesting peer rebalance!)";
			*sendreq = ISC_TRUE;
		} else
			reqlog = "";

		log_info("balancing pool %lx %s  total %d  free %d  "
			 "backup %d  lts %d  max-own (+/-)%d%s",
			 (unsigned long)p,
			 (p->shared_network ?
			  p->shared_network->name : ""), p->lease_count,
			 p->free_leases, p->backup_leases, lts, hold,
			 reqlog);

		/* In the first pass, try to allocate leases to the
		 * peer which it would normally be responsible for (if
		 * the lease has a hardware address or client-identifier,
		 * and the load-balance-algorithm chooses the peer to
		 * answer that address), up to a hold% excess in the peer's
		 * favor.  In the second pass, just send the oldest (first
		 * on the list) leases up to a hold% excess in our favor.
		 *
		 * This could make for additional pool rebalance
		 * events, but preserving MAC possession should be
		 * worth it.
		 */
		pass = 0;
		lease_reference(&lp, LEASE_GET_FIRSTP(lq), MDL);

		while (lp) {
			if (next)
			    lease_dereference(&next, MDL);
			ltemp = LEASE_GET_NEXTP(lq, lp);
			if (ltemp != NULL)
			    lease_reference(&next, ltemp, MDL);

			/*
			 * Stop if the pool is 'balanced enough.'
			 *
			 * The pool is balanced enough if:
			 *
			 * 1) We're on the first run through and the peer has
			 *    its fair share of leases already (lts reaches
			 *    -hold).
			 * 2) We're on the second run through, we are shifting
			 *    never-used leases, and there is a perfectly even
			 *    balance (lts reaches zero).
			 * 3) Second run through, we are shifting previously
			 *    used leases, and the local system has its fair
			 *    share but no more (lts reaches hold).
			 *
			 * Note that this is implemented below in 3,2,1 order.
			 */
			if (pass) {
				if (lp->ends) {
					if (lts <= hold)
						break;
				} else {
					if (lts <= 0)
						break;
				}
			} else if (lts <= -hold)
				break;

			if (pass || peer_wants_lease(lp)) {
			    --lts;
			    ++leases_queued;
			    lp->next_binding_state = peer_lease_state;
			    lp->tstp = cur_time;
			    lp->starts = cur_time;

			    scrub_lease(lp, MDL);
			    if (!supersede_lease(lp, NULL, 0, 1, 0, 0) ||
			        !write_lease(lp))
			    	    log_error("can't commit lease %s on "
					      "giveaway", piaddr(lp->ip_addr));
			}

			lease_dereference(&lp, MDL);
			if (next)
				lease_reference(&lp, next, MDL);
			else if (!pass) {
				pass = 1;
				lease_reference(&lp, LEASE_GET_FIRSTP(lq), MDL);
			}
		}

		if (next)
			lease_dereference(&next, MDL);
		if (lp)
			lease_dereference(&lp, MDL);

		if (lts > thresh) {
			result = "IMBALANCED";
			log_func = log_error;
		} else {
			result = "balanced";
			log_func = log_info;
		}

		log_func("%s pool %lx %s  total %d  free %d  backup %d  "
			 "lts %d  max-misbal %d", result, (unsigned long)p,
			 (p->shared_network ?
			  p->shared_network->name : ""), p->lease_count,
			 p->free_leases, p->backup_leases, lts, thresh);

		/* Recalculate next rebalance event timer. */
		dhcp_failover_pool_check(p);
	    }
	}

	if (leases_queued)
		commit_leases();

	return leases_queued;
}

/* dhcp_failover_pool_check: Called whenever FREE or BACKUP leases change
 * states, on both servers.  Check the scheduled time to rebalance the pool
 * and lower it if applicable.
 */
void
dhcp_failover_pool_check(struct pool *pool)
{
	dhcp_failover_state_t *peer;
	TIME est1, est2;
	struct timeval tv;
	struct lease *ltemp;

	peer = pool->failover_peer;

	if(!peer || peer->me.state != normal)
		return;

	/* Estimate the time left until lease exhaustion.
	 * The first lease on the backup or free lists is also the oldest
	 * lease.  It is reasonable to guess that it will take at least
	 * as much time for a pool to run out of leases, as the present
	 * age of the oldest lease (seconds since it expired).
	 *
	 * Note that this isn't so sane of an assumption if the oldest
	 * lease is a virgin (ends = 0), we wind up sending this against
	 * the max_balance bounds check.
	 */
	ltemp = LEASE_GET_FIRST(pool->free);
	if(ltemp && ltemp->ends < cur_time)
		est1 = cur_time - ltemp->ends;
	else
		est1 = 0;

	ltemp = LEASE_GET_FIRST(pool->backup);
	if(ltemp && ltemp->ends < cur_time)
		est2 = cur_time - ltemp->ends;
	else
		est2 = 0;

	/* We don't want to schedule rebalance for when we think we'll run
	 * out of leases, we want to schedule the rebalance for when we think
	 * the disparity will be 'large enough' to warrant action.
	 */
	est1 = ((est1 * peer->max_lease_misbalance) + 50) / 100;
	est2 = ((est2 * peer->max_lease_misbalance) + 50) / 100;

	/* Guess when the local system will begin issuing POOLREQ panic
	 * attacks because "max_lease_misbalance*2" has been exceeded.
	 */
	if(peer->i_am == primary)
		est1 *= 2;
	else
		est2 *= 2;

	/* Select the smallest time. */
	if(est1 > est2)
		est1 = est2;

	/* Bounded by the maximum configured value. */
	if(est1 > peer->max_balance)
		est1 = peer->max_balance;

	/* Project this time into the future. */
	est1 += cur_time;

	/* Do not move the time down under the minimum. */
	est2 = peer->last_balance + peer->min_balance;
	if(peer->last_balance && (est1 < est2))
		est1 = est2;

	/* Introduce a random delay. */
	est1 += random() % 5;

	/* Do not move the time forward, or reset to the same time. */
	if(peer->sched_balance) {
		if (est1 >= peer->sched_balance)
			return;

		/* We are about to schedule the time down, cancel the
		 * current timeout.
		 */
		cancel_timeout(dhcp_failover_pool_rebalance, peer);
	}

	/* The time is different, and lower, use it. */
	peer->sched_balance = est1;

#if defined(DEBUG_FAILOVER_TIMING)
	log_info("add_timeout +%d dhcp_failover_pool_rebalance",
		 (int)(est1 - cur_time));
#endif
	tv.tv_sec = est1;
	tv.tv_usec = 0;
	add_timeout(&tv, dhcp_failover_pool_rebalance, peer,
			(tvref_t)dhcp_failover_state_reference,
			(tvunref_t)dhcp_failover_state_dereference);
}

int dhcp_failover_state_pool_check (dhcp_failover_state_t *state)
{
	struct shared_network *s;
	struct pool *p;

	for (s = shared_networks; s; s = s -> next) {
		for (p = s -> pools; p; p = p -> next) {
			if (p -> failover_peer != state)
				continue;
			dhcp_failover_pool_check (p);
		}
	}
	return 0;
}

isc_result_t dhcp_failover_send_updates (dhcp_failover_state_t *state)
{
	struct lease *lp = (struct lease *)0;
	isc_result_t status;

	/* Can't update peer if we're not talking to it! */
	if (!state -> link_to_peer)
		return ISC_R_SUCCESS;

	/* If there are acks pending, transmit them prior to potentially
	 * sending new updates for the same lease.
	 */
	if (state->toack_queue_head != NULL)
		dhcp_failover_send_acks(state);

	while ((state -> partner.max_flying_updates >
		state -> cur_unacked_updates) && state -> update_queue_head) {
		/* Grab the head of the update queue. */
		lease_reference (&lp, state -> update_queue_head, MDL);

		/* Send the update to the peer. */
		status = dhcp_failover_send_bind_update (state, lp);
		if (status != ISC_R_SUCCESS) {
			lease_dereference (&lp, MDL);
			return status;
		}
		lp -> flags &= ~ON_UPDATE_QUEUE;

		/* Take it off the head of the update queue and put the next
		   item in the update queue at the head. */
		lease_dereference (&state -> update_queue_head, MDL);
		if (lp -> next_pending) {
			lease_reference (&state -> update_queue_head,
					 lp -> next_pending, MDL);
			lease_dereference (&lp -> next_pending, MDL);
		} else {
			lease_dereference (&state -> update_queue_tail, MDL);
		}

		if (state -> ack_queue_head) {
			lease_reference
				(&state -> ack_queue_tail -> next_pending,
				 lp, MDL);
			lease_dereference (&state -> ack_queue_tail, MDL);
		} else {
			lease_reference (&state -> ack_queue_head, lp, MDL);
		}
#if defined (POINTER_DEBUG)
		if (lp -> next_pending) {
			log_error ("ack_queue_tail: lp -> next_pending");
			abort ();
		}
#endif
		lease_reference (&state -> ack_queue_tail, lp, MDL);
		lp -> flags |= ON_ACK_QUEUE;
		lease_dereference (&lp, MDL);

		/* Count the object as an unacked update. */
		state -> cur_unacked_updates++;
	}
	return ISC_R_SUCCESS;
}

/* Queue an update for a lease.   Always returns 1 at this point - it's
   not an error for this to be called on a lease for which there's no
   failover peer. */

int dhcp_failover_queue_update (struct lease *lease, int immediate)
{
	dhcp_failover_state_t *state;

	if (!lease -> pool ||
	    !lease -> pool -> failover_peer)
		return 1;

	/* If it's already on the update queue, leave it there. */
	if (lease -> flags & ON_UPDATE_QUEUE)
		return 1;

	/* Get the failover state structure for this lease. */
	state = lease -> pool -> failover_peer;

	/* If it's on the ack queue, take it off. */
	if (lease -> flags & ON_ACK_QUEUE)
		dhcp_failover_ack_queue_remove (state, lease);

	if (state -> update_queue_head) {
		lease_reference (&state -> update_queue_tail -> next_pending,
				 lease, MDL);
		lease_dereference (&state -> update_queue_tail, MDL);
	} else {
		lease_reference (&state -> update_queue_head, lease, MDL);
	}
#if defined (POINTER_DEBUG)
	if (lease -> next_pending) {
		log_error ("next pending on update queue lease.");
#if defined (DEBUG_RC_HISTORY)
		dump_rc_history (lease);
#endif
		abort ();
	}
#endif
	lease_reference (&state -> update_queue_tail, lease, MDL);
	lease -> flags |= ON_UPDATE_QUEUE;
	if (immediate)
		dhcp_failover_send_updates (state);
	return 1;
}

int dhcp_failover_send_acks (dhcp_failover_state_t *state)
{
	failover_message_t *msg = (failover_message_t *)0;

	/* Must commit all leases prior to acking them. */
	if (!commit_leases ())
		return 0;

	while (state -> toack_queue_head) {
		failover_message_reference
			(&msg, state -> toack_queue_head, MDL);
		failover_message_dereference
			(&state -> toack_queue_head, MDL);
		if (msg -> next) {
			failover_message_reference
				(&state -> toack_queue_head, msg -> next, MDL);
		}

		dhcp_failover_send_bind_ack (state, msg, 0, (const char *)0);

		failover_message_dereference (&msg, MDL);
	}

	if (state -> toack_queue_tail)
		failover_message_dereference (&state -> toack_queue_tail, MDL);
	state -> pending_acks = 0;

	return 1;
}

void dhcp_failover_toack_queue_timeout (void *vs)
{
	dhcp_failover_state_t *state = vs;

#if defined (DEBUG_FAILOVER_TIMING)
	log_info ("dhcp_failover_toack_queue_timeout");
#endif

	dhcp_failover_send_acks (state);
}

/* Queue an ack for a message.  There is currently no way to queue a
   negative ack -- these need to be sent directly. */

int dhcp_failover_queue_ack (dhcp_failover_state_t *state,
			     failover_message_t *msg)
{
	struct timeval tv;

	if (state -> toack_queue_head) {
		failover_message_reference
			(&state -> toack_queue_tail -> next, msg, MDL);
		failover_message_dereference (&state -> toack_queue_tail, MDL);
	} else {
		failover_message_reference (&state -> toack_queue_head,
					    msg, MDL);
	}
	failover_message_reference (&state -> toack_queue_tail, msg, MDL);

	state -> pending_acks++;

	/* Flush the toack queue whenever we exceed half the number of
	   allowed unacked updates. */
	if (state -> pending_acks >= state -> partner.max_flying_updates / 2) {
		dhcp_failover_send_acks (state);
	}

	/* Schedule a timeout to flush the ack queue. */
	if (state -> pending_acks > 0) {
#if defined (DEBUG_FAILOVER_TIMING)
		log_info ("add_timeout +2 %s",
			  "dhcp_failover_toack_queue_timeout");
#endif
		tv . tv_sec = cur_time + 2;
		tv . tv_usec = 0;
		add_timeout (&tv,
			     dhcp_failover_toack_queue_timeout, state,
			     (tvref_t)dhcp_failover_state_reference,
			     (tvunref_t)dhcp_failover_state_dereference);
	}

	return 1;
}

void dhcp_failover_ack_queue_remove (dhcp_failover_state_t *state,
				     struct lease *lease)
{
	struct lease *lp;

	if (!(lease -> flags & ON_ACK_QUEUE))
		return;

	if (state -> ack_queue_head == lease) {
		lease_dereference (&state -> ack_queue_head, MDL);
		if (lease -> next_pending) {
			lease_reference (&state -> ack_queue_head,
					 lease -> next_pending, MDL);
			lease_dereference (&lease -> next_pending, MDL);
		} else {
			lease_dereference (&state -> ack_queue_tail, MDL);
		}
	} else {
		for (lp = state -> ack_queue_head;
		     lp && lp -> next_pending != lease;
		     lp = lp -> next_pending)
			;

		if (!lp)
			return;

		lease_dereference (&lp -> next_pending, MDL);
		if (lease -> next_pending) {
			lease_reference (&lp -> next_pending,
					 lease -> next_pending, MDL);
			lease_dereference (&lease -> next_pending, MDL);
		} else {
			lease_dereference (&state -> ack_queue_tail, MDL);
			if (lp -> next_pending) {
				log_error ("state -> ack_queue_tail");
				abort ();
			}
			lease_reference (&state -> ack_queue_tail, lp, MDL);
		}
	}

	lease -> flags &= ~ON_ACK_QUEUE;
	/* Multiple acks on one XID is an error and may cause badness. */
	lease->last_xid = 0;
	/* XXX: this violates draft-failover.  We can't send another
	 * update just because we forgot about an old one that hasn't
	 * been acked yet.
	 */
	state -> cur_unacked_updates--;

	/*
	 * When updating leases as a result of an ack, we defer the commit
	 * for performance reasons.  When there are no more acks pending,
	 * do a commit.
	 */
	if (state -> cur_unacked_updates == 0) {
		commit_leases();
	}
}

isc_result_t dhcp_failover_state_set_value (omapi_object_t *h,
					    omapi_object_t *id,
					    omapi_data_string_t *name,
					    omapi_typed_data_t *value)
{
	isc_result_t status;

	if (h -> type != dhcp_type_failover_state)
		return DHCP_R_INVALIDARG;

	/* This list of successful returns is completely wrong, but the
	   fastest way to make dhcpctl do something vaguely sane when
	   you try to change the local state. */

	if (!omapi_ds_strcmp (name, "name")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "partner-address")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "local-address")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "partner-port")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "local-port")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "max-outstanding-updates")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "mclt")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "load-balance-max-secs")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "load-balance-hba")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "partner-state")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "local-state")) {
		unsigned long l;
		status = omapi_get_int_value (&l, value);
		if (status != ISC_R_SUCCESS)
			return status;
		return dhcp_failover_set_state ((dhcp_failover_state_t *)h, l);
	} else if (!omapi_ds_strcmp (name, "partner-stos")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "local-stos")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "hierarchy")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "last-packet-sent")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "last-timestamp-received")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "skew")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "max-response-delay")) {
		return ISC_R_SUCCESS;
	} else if (!omapi_ds_strcmp (name, "cur-unacked-updates")) {
		return ISC_R_SUCCESS;
	}

	if (h -> inner && h -> inner -> type -> set_value)
		return (*(h -> inner -> type -> set_value))
			(h -> inner, id, name, value);
	return ISC_R_NOTFOUND;
}

void dhcp_failover_keepalive (void *vs)
{
}

void dhcp_failover_reconnect (void *vs)
{
	dhcp_failover_state_t *state = vs;
	isc_result_t status;
	struct timeval tv;

#if defined (DEBUG_FAILOVER_TIMING)
	log_info ("dhcp_failover_reconnect");
#endif
	/* If we already connected the other way, let the connection
           recovery code initiate any retry that may be required. */
	if (state -> link_to_peer)
		return;

	status = dhcp_failover_link_initiate ((omapi_object_t *)state);
	if (status != ISC_R_SUCCESS && status != DHCP_R_INCOMPLETE) {
		log_info ("failover peer %s: %s", state -> name,
			  isc_result_totext (status));
#if defined (DEBUG_FAILOVER_TIMING)
		log_info("add_timeout +90 dhcp_failover_reconnect");
#endif
		tv . tv_sec = cur_time + 90;
		tv . tv_usec = 0;
		add_timeout(&tv, dhcp_failover_reconnect, state,
			    (tvref_t)dhcp_failover_state_reference,
			    (tvunref_t)dhcp_failover_state_dereference);
	}
}

void dhcp_failover_startup_timeout (void *vs)
{
	dhcp_failover_state_t *state = vs;

#if defined (DEBUG_FAILOVER_TIMING)
	log_info ("dhcp_failover_startup_timeout");
#endif

	dhcp_failover_state_transition (state, "disconnect");
}

void dhcp_failover_link_startup_timeout (void *vl)
{
	dhcp_failover_link_t *link = vl;
	omapi_object_t *p;

	for (p = (omapi_object_t *)link; p -> inner; p = p -> inner)
		;
	for (; p; p = p -> outer)
		if (p -> type == omapi_type_connection)
			break;
	if (p) {
		log_info ("failover: link startup timeout");
		omapi_disconnect (p, 1);
	}
}

void dhcp_failover_listener_restart (void *vs)
{
	dhcp_failover_state_t *state = vs;
	isc_result_t status;
	struct timeval tv;

#if defined (DEBUG_FAILOVER_TIMING)
	log_info ("dhcp_failover_listener_restart");
#endif

	status = dhcp_failover_listen ((omapi_object_t *)state);
	if (status != ISC_R_SUCCESS) {
		log_info ("failover peer %s: %s", state -> name,
			  isc_result_totext (status));
#if defined (DEBUG_FAILOVER_TIMING)
		log_info ("add_timeout +90 %s",
			  "dhcp_failover_listener_restart");
#endif
		tv . tv_sec = cur_time + 90;
		tv . tv_usec = 0;
		add_timeout (&tv,
			     dhcp_failover_listener_restart, state,
			     (tvref_t)dhcp_failover_state_reference,
			     (tvunref_t)dhcp_failover_state_dereference);
	}
}

void
dhcp_failover_auto_partner_down(void *vs)
{
	dhcp_failover_state_t *state = vs;

#if defined (DEBUG_FAILOVER_TIMING)
	log_info("dhcp_failover_auto_partner_down");
#endif

	dhcp_failover_set_state(state, partner_down);
}

isc_result_t dhcp_failover_state_get_value (omapi_object_t *h,
					    omapi_object_t *id,
					    omapi_data_string_t *name,
					    omapi_value_t **value)
{
	dhcp_failover_state_t *s;
	struct option_cache *oc;
	struct data_string ds;
	isc_result_t status;

	if (h -> type != dhcp_type_failover_state)
		return DHCP_R_INVALIDARG;
	s = (dhcp_failover_state_t *)h;

	if (!omapi_ds_strcmp (name, "name")) {
		if (s -> name)
			return omapi_make_string_value (value,
							name, s -> name, MDL);
		return ISC_R_NOTFOUND;
	} else if (!omapi_ds_strcmp (name, "partner-address")) {
		oc = s -> partner.address;
	      getaddr:
		memset (&ds, 0, sizeof ds);
		if (!evaluate_option_cache (&ds, (struct packet *)0,
					    (struct lease *)0,
					    (struct client_state *)0,
					    (struct option_state *)0,
					    (struct option_state *)0,
					    &global_scope, oc, MDL)) {
			return ISC_R_NOTFOUND;
		}
		status = omapi_make_const_value (value,
						 name, ds.data, ds.len, MDL);
		/* Disgusting kludge: */
		if (oc == s -> me.address && !s -> server_identifier.len)
			data_string_copy (&s -> server_identifier, &ds, MDL);
		data_string_forget (&ds, MDL);
		return status;
	} else if (!omapi_ds_strcmp (name, "local-address")) {
		oc = s -> me.address;
		goto getaddr;
	} else if (!omapi_ds_strcmp (name, "partner-port")) {
		return omapi_make_int_value (value, name,
					     s -> partner.port, MDL);
	} else if (!omapi_ds_strcmp (name, "local-port")) {
		return omapi_make_int_value (value,
					     name, s -> me.port, MDL);
	} else if (!omapi_ds_strcmp (name, "max-outstanding-updates")) {
		return omapi_make_uint_value (value, name,
					      s -> me.max_flying_updates,
					      MDL);
	} else if (!omapi_ds_strcmp (name, "mclt")) {
		return omapi_make_uint_value (value, name, s -> mclt, MDL);
	} else if (!omapi_ds_strcmp (name, "load-balance-max-secs")) {
		return omapi_make_int_value (value, name,
					     s -> load_balance_max_secs, MDL);
	} else if (!omapi_ds_strcmp (name, "load-balance-hba")) {
		if (s -> hba)
			return omapi_make_const_value (value, name,
						       s -> hba, 32, MDL);
		return ISC_R_NOTFOUND;
	} else if (!omapi_ds_strcmp (name, "partner-state")) {
		return omapi_make_uint_value (value, name,
					     s -> partner.state, MDL);
	} else if (!omapi_ds_strcmp (name, "local-state")) {
		return omapi_make_uint_value (value, name,
					      s -> me.state, MDL);
	} else if (!omapi_ds_strcmp (name, "partner-stos")) {
		return omapi_make_int_value (value, name,
					     s -> partner.stos, MDL);
	} else if (!omapi_ds_strcmp (name, "local-stos")) {
		return omapi_make_int_value (value, name,
					     s -> me.stos, MDL);
	} else if (!omapi_ds_strcmp (name, "hierarchy")) {
		return omapi_make_uint_value (value, name, s -> i_am, MDL);
	} else if (!omapi_ds_strcmp (name, "last-packet-sent")) {
		return omapi_make_int_value (value, name,
					     s -> last_packet_sent, MDL);
	} else if (!omapi_ds_strcmp (name, "last-timestamp-received")) {
		return omapi_make_int_value (value, name,
					     s -> last_timestamp_received,
					     MDL);
	} else if (!omapi_ds_strcmp (name, "skew")) {
		return omapi_make_int_value (value, name, s -> skew, MDL);
	} else if (!omapi_ds_strcmp (name, "max-response-delay")) {
		return omapi_make_uint_value (value, name,
					     s -> me.max_response_delay,
					      MDL);
	} else if (!omapi_ds_strcmp (name, "cur-unacked-updates")) {
		return omapi_make_int_value (value, name,
					     s -> cur_unacked_updates, MDL);
	}

	if (h -> inner && h -> inner -> type -> get_value)
		return (*(h -> inner -> type -> get_value))
			(h -> inner, id, name, value);
	return ISC_R_NOTFOUND;
}

isc_result_t dhcp_failover_state_destroy (omapi_object_t *h,
					      const char *file, int line)
{
	dhcp_failover_state_t *s;

	if (h -> type != dhcp_type_failover_state)
		return DHCP_R_INVALIDARG;
	s = (dhcp_failover_state_t *)h;

	if (s -> link_to_peer)
	    dhcp_failover_link_dereference (&s -> link_to_peer, file, line);
	if (s -> name) {
		dfree (s -> name, MDL);
		s -> name = (char *)0;
	}
	if (s -> partner.address)
		option_cache_dereference (&s -> partner.address, file, line);
	if (s -> me.address)
		option_cache_dereference (&s -> me.address, file, line);
	if (s -> hba) {
		dfree (s -> hba, file, line);
		s -> hba = (u_int8_t *)0;
	}
	if (s -> update_queue_head)
		lease_dereference (&s -> update_queue_head, file, line);
	if (s -> update_queue_tail)
		lease_dereference (&s -> update_queue_tail, file, line);
	if (s -> ack_queue_head)
		lease_dereference (&s -> ack_queue_head, file, line);
	if (s -> ack_queue_tail)
		lease_dereference (&s -> ack_queue_tail, file, line);
	if (s -> send_update_done)
		lease_dereference (&s -> send_update_done, file, line);
	if (s -> toack_queue_head)
		failover_message_dereference (&s -> toack_queue_head,
					      file, line);
	if (s -> toack_queue_tail)
		failover_message_dereference (&s -> toack_queue_tail,
					      file, line);
	return ISC_R_SUCCESS;
}

/* Write all the published values associated with the object through the
   specified connection. */

isc_result_t dhcp_failover_state_stuff (omapi_object_t *c,
					omapi_object_t *id,
					omapi_object_t *h)
{
	/* In this function c should be a (omapi_connection_object_t *) */

	dhcp_failover_state_t *s;
	isc_result_t status;

	if (c -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	if (h -> type != dhcp_type_failover_state)
		return DHCP_R_INVALIDARG;
	s = (dhcp_failover_state_t *)h;

	status = omapi_connection_put_name (c, "name");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_string (c, s -> name);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "partner-address");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof s -> partner.address);
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_copyin (c, (u_int8_t *)&s -> partner.address,
					  sizeof s -> partner.address);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "partner-port");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, (u_int32_t)s -> partner.port);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "local-address");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof s -> me.address);
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_copyin (c, (u_int8_t *)&s -> me.address,
					  sizeof s -> me.address);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "local-port");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, (u_int32_t)s -> me.port);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "max-outstanding-updates");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c,
					      s -> me.max_flying_updates);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "mclt");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, s -> mclt);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "load-balance-max-secs");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = (omapi_connection_put_uint32
		  (c, (u_int32_t)s -> load_balance_max_secs));
	if (status != ISC_R_SUCCESS)
		return status;


	if (s -> hba) {
		status = omapi_connection_put_name (c, "load-balance-hba");
		if (status != ISC_R_SUCCESS)
			return status;
		status = omapi_connection_put_uint32 (c, 32);
		if (status != ISC_R_SUCCESS)
			return status;
		status = omapi_connection_copyin (c, s -> hba, 32);
		if (status != ISC_R_SUCCESS)
			return status;
	}

	status = omapi_connection_put_name (c, "partner-state");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, s -> partner.state);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "local-state");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, s -> me.state);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "partner-stos");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c,
					      (u_int32_t)s -> partner.stos);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "local-stos");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, (u_int32_t)s -> me.stos);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "hierarchy");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, s -> i_am);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "last-packet-sent");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = (omapi_connection_put_uint32
		  (c, (u_int32_t)s -> last_packet_sent));
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "last-timestamp-received");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = (omapi_connection_put_uint32
		  (c, (u_int32_t)s -> last_timestamp_received));
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "skew");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, (u_int32_t)s -> skew);
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "max-response-delay");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = (omapi_connection_put_uint32
		  (c, (u_int32_t)s -> me.max_response_delay));
	if (status != ISC_R_SUCCESS)
		return status;

	status = omapi_connection_put_name (c, "cur-unacked-updates");
	if (status != ISC_R_SUCCESS)
		return status;
	status = omapi_connection_put_uint32 (c, sizeof (u_int32_t));
	if (status != ISC_R_SUCCESS)
		return status;
	status = (omapi_connection_put_uint32
		  (c, (u_int32_t)s -> cur_unacked_updates));
	if (status != ISC_R_SUCCESS)
		return status;

	if (h -> inner && h -> inner -> type -> stuff_values)
		return (*(h -> inner -> type -> stuff_values)) (c, id,
								h -> inner);
	return ISC_R_SUCCESS;
}

isc_result_t dhcp_failover_state_lookup (omapi_object_t **sp,
					 omapi_object_t *id,
					 omapi_object_t *ref)
{
	omapi_value_t *tv = (omapi_value_t *)0;
	isc_result_t status;
	dhcp_failover_state_t *s;

	if (!ref)
		return DHCP_R_NOKEYS;

	/* First see if we were sent a handle. */
	status = omapi_get_value_str (ref, id, "handle", &tv);
	if (status == ISC_R_SUCCESS) {
		status = omapi_handle_td_lookup (sp, tv -> value);

		omapi_value_dereference (&tv, MDL);
		if (status != ISC_R_SUCCESS)
			return status;

		/* Don't return the object if the type is wrong. */
		if ((*sp) -> type != dhcp_type_failover_state) {
			omapi_object_dereference (sp, MDL);
			return DHCP_R_INVALIDARG;
		}
	}

	/* Look the failover state up by peer name. */
	status = omapi_get_value_str (ref, id, "name", &tv);
	if (status == ISC_R_SUCCESS) {
		for (s = failover_states; s; s = s -> next) {
			unsigned l = strlen (s -> name);
			if (l == tv -> value -> u.buffer.len &&
			    !memcmp (s -> name,
				     tv -> value -> u.buffer.value, l))
				break;
		}
		omapi_value_dereference (&tv, MDL);

		/* If we already have a lease, and it's not the same one,
		   then the query was invalid. */
		if (*sp && *sp != (omapi_object_t *)s) {
			omapi_object_dereference (sp, MDL);
			return DHCP_R_KEYCONFLICT;
		} else if (!s) {
			if (*sp)
				omapi_object_dereference (sp, MDL);
			return ISC_R_NOTFOUND;
		} else if (!*sp)
			/* XXX fix so that hash lookup itself creates
			   XXX the reference. */
			omapi_object_reference (sp, (omapi_object_t *)s, MDL);
	}

	/* If we get to here without finding a lease, no valid key was
	   specified. */
	if (!*sp)
		return DHCP_R_NOKEYS;
	return ISC_R_SUCCESS;
}

isc_result_t dhcp_failover_state_create (omapi_object_t **sp,
					 omapi_object_t *id)
{
	return ISC_R_NOTIMPLEMENTED;
}

isc_result_t dhcp_failover_state_remove (omapi_object_t *sp,
					 omapi_object_t *id)
{
	return ISC_R_NOTIMPLEMENTED;
}

int dhcp_failover_state_match (dhcp_failover_state_t *state,
			       u_int8_t *addr, unsigned addrlen)
{
	struct data_string ds;
	int i;

	memset (&ds, 0, sizeof ds);
	if (evaluate_option_cache (&ds, (struct packet *)0,
				   (struct lease *)0,
				   (struct client_state *)0,
				   (struct option_state *)0,
				   (struct option_state *)0,
				   &global_scope,
				   state -> partner.address, MDL)) {
		for (i = 0; i + addrlen - 1 < ds.len; i += addrlen) {
			if (!memcmp (&ds.data [i],
				     addr, addrlen)) {
				data_string_forget (&ds, MDL);
				return 1;
			}
		}
		data_string_forget (&ds, MDL);
	}
	return 0;
}

int
dhcp_failover_state_match_by_name(state, name)
	dhcp_failover_state_t *state;
	failover_option_t *name;
{
	if ((strlen(state->name) == name->count) &&
	    (memcmp(state->name, name->data, name->count) == 0))
		return 1;

	return 0;
}

const char *dhcp_failover_reject_reason_print (int reason)
{
    static char resbuf[sizeof("Undefined-255: This reason code is not defined "
			      "in the protocol standard.")];

    if ((reason > 0xff) || (reason < 0))
	return "Reason code out of range.";

    switch (reason) {
      case FTR_ILLEGAL_IP_ADDR:
	return "Illegal IP address (not part of any address pool).";

      case FTR_FATAL_CONFLICT:
	return "Fatal conflict exists: address in use by other client.";

      case FTR_MISSING_BINDINFO:
	return "Missing binding information.";

      case FTR_TIMEMISMATCH:
	return "Connection rejected, time mismatch too great.";

      case FTR_INVALID_MCLT:
	return "Connection rejected, invalid MCLT.";

      case FTR_MISC_REJECT:
	return "Connection rejected, unknown reason.";

      case FTR_DUP_CONNECTION:
	return "Connection rejected, duplicate connection.";

      case FTR_INVALID_PARTNER:
	return "Connection rejected, invalid failover partner.";

      case FTR_TLS_UNSUPPORTED:
	return "TLS not supported.";

      case FTR_TLS_UNCONFIGURED:
	return "TLS supported but not configured.";

      case FTR_TLS_REQUIRED:
	return "TLS required but not supported by partner.";

      case FTR_DIGEST_UNSUPPORTED:
	return "Message digest not supported.";

      case FTR_DIGEST_UNCONFIGURED:
	return "Message digest not configured.";

      case FTR_VERSION_MISMATCH:
	return "Protocol version mismatch.";

      case FTR_OUTDATED_BIND_INFO:
	return "Outdated binding information.";

      case FTR_LESS_CRIT_BIND_INFO:
	return "Less critical binding information.";

      case FTR_NO_TRAFFIC:
	return "No traffic within sufficient time.";

      case FTR_HBA_CONFLICT:
	return "Hash bucket assignment conflict.";

      case FTR_IP_NOT_RESERVED:
	return "IP not reserved on this server.";

      case FTR_IP_DIGEST_FAILURE:
	return "Message digest failed to compare.";

      case FTR_IP_MISSING_DIGEST:
	return "Missing message digest.";

      case FTR_UNKNOWN:
	return "Unknown Error.";

      default:
	sprintf(resbuf, "Undefined-%d: This reason code is not defined in the "
			"protocol standard.", reason);
	return resbuf;
    }
}

const char *dhcp_failover_state_name_print (enum failover_state state)
{
	switch (state) {
	      default:
	      case unknown_state:
		return "unknown-state";

	      case partner_down:
		return "partner-down";

	      case normal:
		return "normal";

	      case conflict_done:
		return "conflict-done";

	      case communications_interrupted:
		return "communications-interrupted";

	      case resolution_interrupted:
		return "resolution-interrupted";

	      case potential_conflict:
		return "potential-conflict";

	      case recover:
		return "recover";

	      case recover_done:
		return "recover-done";

	      case recover_wait:
		return "recover-wait";

	      case shut_down:
		return "shutdown";

	      case paused:
		return "paused";

	      case startup:
		return "startup";
	}
}

const char *dhcp_failover_message_name (unsigned type)
{
	static char messbuf[sizeof("unknown-message-255")];

	if (type > 0xff)
		return "invalid-message";

	switch (type) {
	      case FTM_POOLREQ:
		return "pool-request";

	      case FTM_POOLRESP:
		return "pool-response";

	      case FTM_BNDUPD:
		return "bind-update";

	      case FTM_BNDACK:
		return "bind-ack";

	      case FTM_CONNECT:
		return "connect";

	      case FTM_CONNECTACK:
		return "connect-ack";

	      case FTM_UPDREQ:
		return "update-request";

	      case FTM_UPDDONE:
		return "update-done";

	      case FTM_UPDREQALL:
		return "update-request-all";

	      case FTM_STATE:
		return "state";

	      case FTM_CONTACT:
		return "contact";

	      case FTM_DISCONNECT:
		return "disconnect";

	      default:
		sprintf(messbuf, "unknown-message-%u", type);
		return messbuf;
	}
}

const char *dhcp_failover_option_name (unsigned type)
{
	static char optbuf[sizeof("unknown-option-65535")];

	if (type > 0xffff)
		return "invalid-option";

	switch (type) {
	    case FTO_ADDRESSES_TRANSFERRED:
		return "addresses-transferred";

	    case FTO_ASSIGNED_IP_ADDRESS:
		return "assigned-ip-address";

	    case FTO_BINDING_STATUS:
		return "binding-status";

	    case FTO_CLIENT_IDENTIFIER:
		return "client-identifier";

	    case FTO_CHADDR:
		return "chaddr";

	    case FTO_CLTT:
		return "cltt";

	    case FTO_DDNS:
		return "ddns";

	    case FTO_DELAYED_SERVICE:
		return "delayed-service";

	    case FTO_HBA:
		return "hba";

	    case FTO_IP_FLAGS:
		return "ip-flags";

	    case FTO_LEASE_EXPIRY:
		return "lease-expiry";

	    case FTO_MAX_UNACKED:
		return "max-unacked";

	    case FTO_MCLT:
		return "mclt";

	    case FTO_MESSAGE:
		return "message";

	    case FTO_MESSAGE_DIGEST:
		return "message-digest";

	    case FTO_POTENTIAL_EXPIRY:
		return "potential-expiry";

	    case FTO_PROTOCOL_VERSION:
		return "protocol-version";

	    case FTO_RECEIVE_TIMER:
		return "receive-timer";

	    case FTO_REJECT_REASON:
		return "reject-reason";

	    case FTO_RELATIONSHIP_NAME:
		return "relationship-name";

	    case FTO_REPLY_OPTIONS:
		return "reply-options";

	    case FTO_REQUEST_OPTIONS:
		return "request-options";

	    case FTO_SERVER_FLAGS:
		return "server-flags";

	    case FTO_SERVER_STATE:
		return "server-state";

	    case FTO_STOS:
		return "stos";

	    case FTO_TLS_REPLY:
		return "tls-reply";

	    case FTO_TLS_REQUEST:
		return "tls-request";

	    case FTO_VENDOR_CLASS:
		return "vendor-class";

	    case FTO_VENDOR_OPTIONS:
		return "vendor-options";

	    default:
		sprintf(optbuf, "unknown-option-%u", type);
		return optbuf;
	}
}

failover_option_t *dhcp_failover_option_printf (unsigned code,
						char *obuf,
						unsigned *obufix,
						unsigned obufmax,
						const char *fmt, ...)
{
	va_list va;
	char tbuf [256];

	/* %Audit% Truncation causes panic. %2004.06.17,Revisit%
	 * It is unclear what the effects of truncation here are, or
	 * how that condition should be handled.  It seems that this
	 * function is used for formatting messages in the failover
	 * command channel.  For now the safest thing is for
	 * overflow-truncation to cause a fatal log.
	 */
	va_start (va, fmt);
	if (vsnprintf (tbuf, sizeof tbuf, fmt, va) >= sizeof tbuf)
		log_fatal ("%s: vsnprintf would truncate",
				"dhcp_failover_make_option");
	va_end (va);

	return dhcp_failover_make_option (code, obuf, obufix, obufmax,
					  strlen (tbuf), tbuf);
}

failover_option_t *dhcp_failover_make_option (unsigned code,
					      char *obuf, unsigned *obufix,
					      unsigned obufmax, ...)
{
	va_list va;
	struct failover_option_info *info;
	int i;
	unsigned size, count;
	unsigned val;
	u_int8_t *iaddr;
	unsigned ilen = 0;
	u_int8_t *bval;
	char *txt = NULL;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char tbuf [256];
#endif

	/* Note that the failover_option structure is used differently on
	   input than on output - on input, count is an element count, and
	   on output it's the number of bytes total in the option, including
	   the option code and option length. */
	failover_option_t option, *op;


	/* Bogus option code? */
	if (code < 1 || code > FTO_MAX || ft_options [code].type == FT_UNDEF) {
		return &null_failover_option;
	}
	info = &ft_options [code];

	va_start (va, obufmax);

	/* Get the number of elements and the size of the buffer we need
	   to allocate. */
	if (info -> type == FT_DDNS || info -> type == FT_DDNS1) {
		count = info -> type == FT_DDNS ? 1 : 2;
		size = va_arg (va, int) + count;
	} else {
		/* Find out how many items in this list. */
		if (info -> num_present)
			count = info -> num_present;
		else
			count = va_arg (va, int);

		/* Figure out size. */
		switch (info -> type) {
		      case FT_UINT8:
		      case FT_BYTES:
		      case FT_DIGEST:
			size = count;
			break;

		      case FT_TEXT_OR_BYTES:
		      case FT_TEXT:
			txt = va_arg (va, char *);
			size = count;
			break;

		      case FT_IPADDR:
			ilen = va_arg (va, unsigned);
			size = count * ilen;
			break;

		      case FT_UINT32:
			size = count * 4;
			break;

		      case FT_UINT16:
			size = count * 2;
			break;

		      default:
			/* shouldn't get here. */
			log_fatal ("bogus type in failover_make_option: %d",
				   info -> type);
			return &null_failover_option;
		}
	}

	size += 4;

	/* Allocate a buffer for the option. */
	option.count = size;
	option.data = dmalloc (option.count, MDL);
	if (!option.data) {
		va_end (va);
		return &null_failover_option;
	}

	/* Put in the option code and option length. */
	putUShort (option.data, code);
	putUShort (&option.data [2], size - 4);

#if defined (DEBUG_FAILOVER_MESSAGES)
	/* %Audit% Truncation causes panic. %2004.06.17,Revisit%
	 * It is unclear what the effects of truncation here are, or
	 * how that condition should be handled.  It seems that this
	 * message may be sent over the failover command channel.
	 * For now the safest thing is for overflow-truncation to cause
	 * a fatal log.
	 */
	if (snprintf (tbuf, sizeof tbuf, " (%s<%d>", info -> name,
			option.count) >= sizeof tbuf)
		log_fatal ("dhcp_failover_make_option: tbuf overflow");
	failover_print (obuf, obufix, obufmax, tbuf);
#endif

	/* Now put in the data. */
	switch (info -> type) {
	      case FT_UINT8:
		for (i = 0; i < count; i++) {
			val = va_arg (va, unsigned);
#if defined (DEBUG_FAILOVER_MESSAGES)
			/* %Audit% Cannot exceed 24 bytes. %2004.06.17,Safe% */
			sprintf (tbuf, " %d", val);
			failover_print (obuf, obufix, obufmax, tbuf);
#endif
			option.data [i + 4] = val;
		}
		break;

	      case FT_IPADDR:
		for (i = 0; i < count; i++) {
			iaddr = va_arg (va, u_int8_t *);
			if (ilen != 4) {
				dfree (option.data, MDL);
				log_error ("IP addrlen=%d, should be 4.",
					   ilen);
				va_end (va);
				return &null_failover_option;
			}

#if defined (DEBUG_FAILOVER_MESSAGES)
			/*%Audit% Cannot exceed 17 bytes.  %2004.06.17,Safe%*/
			sprintf (tbuf, " %u.%u.%u.%u",
				  iaddr [0], iaddr [1], iaddr [2], iaddr [3]);
			failover_print (obuf, obufix, obufmax, tbuf);
#endif
			memcpy (&option.data [4 + i * ilen], iaddr, ilen);
		}
		break;

	      case FT_UINT32:
		for (i = 0; i < count; i++) {
			val = va_arg (va, unsigned);
#if defined (DEBUG_FAILOVER_MESSAGES)
			/*%Audit% Cannot exceed 24 bytes.  %2004.06.17,Safe%*/
			sprintf (tbuf, " %d", val);
			failover_print (obuf, obufix, obufmax, tbuf);
#endif
			putULong (&option.data [4 + i * 4], val);
		}
		break;

	      case FT_BYTES:
	      case FT_DIGEST:
		bval = va_arg (va, u_int8_t *);
#if defined (DEBUG_FAILOVER_MESSAGES)
		for (i = 0; i < count; i++) {
			/* 23 bytes plus nul, safe. */
			sprintf (tbuf, " %d", bval [i]);
			failover_print (obuf, obufix, obufmax, tbuf);
		}
#endif
		memcpy (&option.data [4], bval, count);
		break;

		/* On output, TEXT_OR_BYTES is _always_ text, and always NUL
		   terminated.  Note that the caller should be careful not
		   to provide a format and data that amount to more than 256
		   bytes of data, since it will cause a fatal error. */
	      case FT_TEXT_OR_BYTES:
	      case FT_TEXT:
#if defined (DEBUG_FAILOVER_MESSAGES)
		/* %Audit% Truncation causes panic. %2004.06.17,Revisit%
		 * It is unclear what the effects of truncation here are, or
		 * how that condition should be handled.  It seems that this
		 * function is used for formatting messages in the failover
		 * command channel.  For now the safest thing is for
		 * overflow-truncation to cause a fatal log.
		 */
		if (snprintf (tbuf, sizeof tbuf, "\"%s\"", txt) >= sizeof tbuf)
			log_fatal ("dhcp_failover_make_option: tbuf overflow");
		failover_print (obuf, obufix, obufmax, tbuf);
#endif
		memcpy (&option.data [4], txt, count);
		break;

	      case FT_DDNS:
	      case FT_DDNS1:
		option.data [4] = va_arg (va, unsigned);
		if (count == 2)
			option.data [5] = va_arg (va, unsigned);
		bval = va_arg (va, u_int8_t *);
		memcpy (&option.data [4 + count], bval, size - count - 4);
#if defined (DEBUG_FAILOVER_MESSAGES)
		for (i = 4; i < size; i++) {
			/*%Audit% Cannot exceed 24 bytes. %2004.06.17,Safe%*/
			sprintf (tbuf, " %d", option.data [i]);
			failover_print (obuf, obufix, obufmax, tbuf);
		}
#endif
		break;

	      case FT_UINT16:
		for (i = 0; i < count; i++) {
			val = va_arg (va, u_int32_t);
#if defined (DEBUG_FAILOVER_MESSAGES)
			/*%Audit% Cannot exceed 24 bytes. %2004.06.17,Safe%*/
			sprintf (tbuf, " %d", val);
			failover_print (obuf, obufix, obufmax, tbuf);
#endif
			putUShort (&option.data [4 + i * 2], val);
		}
		break;

	      case FT_UNDEF:
	      default:
		break;
	}

#if defined DEBUG_FAILOVER_MESSAGES
	failover_print (obuf, obufix, obufmax, ")");
#endif
	va_end (va);

	/* Now allocate a place to store what we just set up. */
	op = dmalloc (sizeof (failover_option_t), MDL);
	if (!op) {
		dfree (option.data, MDL);
		return &null_failover_option;
	}

	*op = option;
	return op;
}

/* Send a failover message header. */

isc_result_t dhcp_failover_put_message (dhcp_failover_link_t *link,
					omapi_object_t *connection,
					int msg_type, u_int32_t xid, ...)
{
	unsigned size = 0;
	int bad_option = 0;
	int opix = 0;
	va_list list;
	failover_option_t *option;
	unsigned char *opbuf;
	isc_result_t status = ISC_R_SUCCESS;
	unsigned char cbuf;
	struct timeval tv;

	/* Run through the argument list once to compute the length of
	   the option portion of the message. */
	va_start (list, xid);
	while ((option = va_arg (list, failover_option_t *))) {
		if (option != &skip_failover_option)
			size += option -> count;
		if (option == &null_failover_option)
			bad_option = 1;
	}
	va_end (list);

	/* Allocate an option buffer, unless we got an error. */
	if (!bad_option && size) {
		opbuf = dmalloc (size, MDL);
		if (!opbuf)
			status = ISC_R_NOMEMORY;
	} else
		opbuf = (unsigned char *)0;

	va_start (list, xid);
	while ((option = va_arg (list, failover_option_t *))) {
		if (option == &skip_failover_option)
		    continue;
		if (!bad_option && opbuf)
			memcpy (&opbuf [opix],
				option -> data, option -> count);
		if (option != &null_failover_option &&
		    option != &skip_failover_option) {
			opix += option -> count;
			dfree (option -> data, MDL);
			dfree (option, MDL);
		}
	}
	va_end(list);

	if (bad_option)
		return DHCP_R_INVALIDARG;

	/* Now send the message header. */

	/* Message length. */
	status = omapi_connection_put_uint16 (connection, size + 12);
	if (status != ISC_R_SUCCESS)
		goto err;

	/* Message type. */
	cbuf = msg_type;
	status = omapi_connection_copyin (connection, &cbuf, 1);
	if (status != ISC_R_SUCCESS)
		goto err;

	/* Payload offset. */
	cbuf = 12;
	status = omapi_connection_copyin (connection, &cbuf, 1);
	if (status != ISC_R_SUCCESS)
		goto err;

	/* Current time. */
	status = omapi_connection_put_uint32 (connection, (u_int32_t)cur_time);
	if (status != ISC_R_SUCCESS)
		goto err;

	/* Transaction ID. */
	status = omapi_connection_put_uint32(connection, xid);
	if (status != ISC_R_SUCCESS)
		goto err;

	/* Payload. */
	if (opbuf) {
		status = omapi_connection_copyin (connection, opbuf, size);
		if (status != ISC_R_SUCCESS)
			goto err;
		dfree (opbuf, MDL);
	}
	if (link -> state_object &&
	    link -> state_object -> link_to_peer == link) {
#if defined (DEBUG_FAILOVER_CONTACT_TIMING)
		log_info ("add_timeout +%d %s",
			  (int)(link -> state_object ->
				partner.max_response_delay) / 3,
			  "dhcp_failover_send_contact");
#endif
		tv . tv_sec = cur_time +
			(int)(link -> state_object ->
			      partner.max_response_delay) / 3;
		tv . tv_usec = 0;
		add_timeout (&tv,
			     dhcp_failover_send_contact, link -> state_object,
			     (tvref_t)dhcp_failover_state_reference,
			     (tvunref_t)dhcp_failover_state_dereference);
	}
	return status;

      err:
	if (opbuf)
		dfree (opbuf, MDL);
	log_info ("dhcp_failover_put_message: something went wrong.");
	omapi_disconnect (connection, 1);
	return status;
}

void dhcp_failover_timeout (void *vstate)
{
	dhcp_failover_state_t *state = vstate;
	dhcp_failover_link_t *link;

#if defined (DEBUG_FAILOVER_TIMING)
	log_info ("dhcp_failover_timeout");
#endif

	if (!state || state -> type != dhcp_type_failover_state)
		return;
	link = state -> link_to_peer;
	if (!link ||
	    !link -> outer ||
	    link -> outer -> type != omapi_type_connection)
		return;

	log_error ("timeout waiting for failover peer %s", state -> name);

	/* If we haven't gotten a timely response, blow away the connection.
	   This will cause the state to change automatically. */
	omapi_disconnect (link -> outer, 1);
}

void dhcp_failover_send_contact (void *vstate)
{
	dhcp_failover_state_t *state = vstate;
	dhcp_failover_link_t *link;
	isc_result_t status;

#if defined(DEBUG_FAILOVER_MESSAGES) && \
    defined(DEBUG_FAILOVER_CONTACT_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

	failover_print(obuf, &obufix, sizeof(obuf), "(contact");
#endif

#if defined (DEBUG_FAILOVER_CONTACT_TIMING)
	log_info ("dhcp_failover_send_contact");
#endif

	if (!state || state -> type != dhcp_type_failover_state)
		return;
	link = state -> link_to_peer;
	if (!link ||
	    !link -> outer ||
	    link -> outer -> type != omapi_type_connection)
		return;

	status = (dhcp_failover_put_message
		  (link, link -> outer,
		   FTM_CONTACT, link->xid++,
		   (failover_option_t *)0));

#if defined(DEBUG_FAILOVER_MESSAGES) && \
    defined(DEBUG_FAILOVER_CONTACT_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print(obuf, &obufix, sizeof(obuf), " (failed)");
	failover_print(obuf, &obufix, sizeof(obuf), ")");
	if (obufix) {
		log_debug ("%s", obuf);
	}
#else
        IGNORE_UNUSED(status);
#endif
	return;
}

isc_result_t dhcp_failover_send_state (dhcp_failover_state_t *state)
{
	dhcp_failover_link_t *link;
	isc_result_t status;

#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(state");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!state || state -> type != dhcp_type_failover_state)
		return DHCP_R_INVALIDARG;
	link = state -> link_to_peer;
	if (!link ||
	    !link -> outer ||
	    link -> outer -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	status = (dhcp_failover_put_message
		  (link, link -> outer,
		   FTM_STATE, link->xid++,
		   dhcp_failover_make_option (FTO_SERVER_STATE, FMA,
					      (state -> me.state == startup
					       ? state -> saved_state
					       : state -> me.state)),
		   dhcp_failover_make_option
		   (FTO_SERVER_FLAGS, FMA,
		    (state -> service_state == service_startup
		     ? FTF_SERVER_STARTUP : 0)),
		   dhcp_failover_make_option (FTO_STOS, FMA, state -> me.stos),
		   (failover_option_t *)0));

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print (FMA, " (failed)");
	failover_print (FMA, ")");
	if (obufix) {
		log_debug ("%s", obuf);
	}
#else
        IGNORE_UNUSED(status);
#endif
	return ISC_R_SUCCESS;
}

/* Send a connect message. */

isc_result_t dhcp_failover_send_connect (omapi_object_t *l)
{
	dhcp_failover_link_t *link;
	dhcp_failover_state_t *state;
	isc_result_t status;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(connect");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!l || l -> type != dhcp_type_failover_link)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)l;
	state = link -> state_object;
	if (!l -> outer || l -> outer -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	status =
	    (dhcp_failover_put_message
	     (link, l -> outer,
	      FTM_CONNECT, link->xid++,
	      dhcp_failover_make_option(FTO_RELATIONSHIP_NAME, FMA,
					strlen(state->name), state->name),
	      dhcp_failover_make_option (FTO_MAX_UNACKED, FMA,
					 state -> me.max_flying_updates),
	      dhcp_failover_make_option (FTO_RECEIVE_TIMER, FMA,
					 state -> me.max_response_delay),
	      dhcp_failover_option_printf(FTO_VENDOR_CLASS, FMA,
					  "isc-%s", PACKAGE_VERSION),
	      dhcp_failover_make_option (FTO_PROTOCOL_VERSION, FMA,
					 DHCP_FAILOVER_VERSION),
	      dhcp_failover_make_option (FTO_TLS_REQUEST, FMA,
					 0, 0),
	      dhcp_failover_make_option (FTO_MCLT, FMA,
					 state -> mclt),
	      (state -> hba
	       ? dhcp_failover_make_option (FTO_HBA, FMA, 32, state -> hba)
	       : &skip_failover_option),
	      (failover_option_t *)0));

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print (FMA, " (failed)");
	failover_print (FMA, ")");
	if (obufix) {
		log_debug ("%s", obuf);
	}
#endif
	return status;
}

isc_result_t dhcp_failover_send_connectack (omapi_object_t *l,
					    dhcp_failover_state_t *state,
					    int reason, const char *errmsg)
{
	dhcp_failover_link_t *link;
	isc_result_t status;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(connectack");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!l || l -> type != dhcp_type_failover_link)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)l;
	if (!l -> outer || l -> outer -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	status =
	    (dhcp_failover_put_message
	     (link, l -> outer,
	      FTM_CONNECTACK, link->imsg->xid,
	      state
	       ? dhcp_failover_make_option(FTO_RELATIONSHIP_NAME, FMA,
					   strlen(state->name), state->name)
	       : (link->imsg->options_present & FTB_RELATIONSHIP_NAME)
		  ? dhcp_failover_make_option(FTO_RELATIONSHIP_NAME, FMA,
					      link->imsg->relationship_name.count,
					      link->imsg->relationship_name.data)
		  : &skip_failover_option,
	      state
	       ? dhcp_failover_make_option (FTO_MAX_UNACKED, FMA,
					    state -> me.max_flying_updates)
	       : &skip_failover_option,
	      state
	       ? dhcp_failover_make_option (FTO_RECEIVE_TIMER, FMA,
					    state -> me.max_response_delay)
	       : &skip_failover_option,
	      dhcp_failover_option_printf(FTO_VENDOR_CLASS, FMA,
					  "isc-%s", PACKAGE_VERSION),
	      dhcp_failover_make_option (FTO_PROTOCOL_VERSION, FMA,
					 DHCP_FAILOVER_VERSION),
	      (link->imsg->options_present & FTB_TLS_REQUEST)
	       ? dhcp_failover_make_option(FTO_TLS_REPLY, FMA,
					   0, 0)
	       : &skip_failover_option,
	      reason
	       ? dhcp_failover_make_option (FTO_REJECT_REASON,
					    FMA, reason)
	       : &skip_failover_option,
	      (reason && errmsg)
	       ? dhcp_failover_make_option (FTO_MESSAGE, FMA,
					    strlen (errmsg), errmsg)
	       : &skip_failover_option,
	      (failover_option_t *)0));

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print (FMA, " (failed)");
	failover_print (FMA, ")");
	if (obufix) {
		log_debug ("%s", obuf);
	}
#endif
	return status;
}

isc_result_t dhcp_failover_send_disconnect (omapi_object_t *l,
					    int reason,
					    const char *message)
{
	dhcp_failover_link_t *link;
	isc_result_t status;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(disconnect");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!l || l -> type != dhcp_type_failover_link)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)l;
	if (!l -> outer || l -> outer -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	if (!message && reason)
		message = dhcp_failover_reject_reason_print (reason);

	status = (dhcp_failover_put_message
		  (link, l -> outer,
		   FTM_DISCONNECT, link->xid++,
		   dhcp_failover_make_option (FTO_REJECT_REASON,
					      FMA, reason),
		   (message
		    ? dhcp_failover_make_option (FTO_MESSAGE, FMA,
						 strlen (message), message)
		    : &skip_failover_option),
		   (failover_option_t *)0));

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print (FMA, " (failed)");
	failover_print (FMA, ")");
	if (obufix) {
		log_debug ("%s", obuf);
	}
#endif
	return status;
}

/* Send a Bind Update message. */

isc_result_t dhcp_failover_send_bind_update (dhcp_failover_state_t *state,
					     struct lease *lease)
{
	dhcp_failover_link_t *link;
	isc_result_t status;
	int flags = 0;
	binding_state_t transmit_state;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(bndupd");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!state -> link_to_peer ||
	    state -> link_to_peer -> type != dhcp_type_failover_link)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)state -> link_to_peer;

	if (!link -> outer || link -> outer -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	transmit_state = lease->desired_binding_state;
	if (lease->flags & RESERVED_LEASE) {
		/* If we are listing an allocable (not yet ACTIVE etc) lease
		 * as reserved, toggle to the peer's 'free state', per the
		 * draft.  This gives the peer permission to alloc it to the
		 * chaddr/uid-named client.
		 */
		if ((state->i_am == primary) && (transmit_state == FTS_FREE))
			transmit_state = FTS_BACKUP;
		else if ((state->i_am == secondary) &&
			 (transmit_state == FTS_BACKUP))
			transmit_state = FTS_FREE;

		flags |= FTF_IP_FLAG_RESERVE;
	}
	if (lease->flags & BOOTP_LEASE)
		flags |= FTF_IP_FLAG_BOOTP;

	/* last_xid == 0 is illegal, seek past zero if we hit it. */
	if (link->xid == 0)
		link->xid = 1;

	lease->last_xid = link->xid++;

	/*
	 * Our very next action is to transmit a binding update relating to
	 * this lease over the wire, and although there is a BNDACK, there is
	 * no BNDACKACK or BNDACKACKACK...the basic issue as we send a BNDUPD,
	 * we may not receive a BNDACK.  This non-reception does not imply the
	 * peer did not receive and process the BNDUPD.  So at this point, we
	 * must divest any state that would be dangerous to retain under the
	 * impression the peer has been updated.  Normally state changes like
	 * this are processed in supersede_lease(), but in this case we need a
	 * very late binding.
	 *
	 * In failover rules, a server is permitted to work forward in certain
	 * directions from a given lease's state; active leases may be
	 * extended, so forth.  There is an 'optimization' in the failover
	 * draft that permits a server to 'rewind' any work they have not
	 * informed the peer.  Since we can't know if the peer received our
	 * update but was unable to acknowledge it, we make this change on
	 * transmit rather than upon receiving the acknowledgement.
	 *
	 * XXX: Frequent lease commits are undesirable.  This should hopefully
	 * only trigger when a server is sending a lease /state change/, and
	 * not merely an update such as with a renewal.
	 */
	if (lease->rewind_binding_state != lease->binding_state) {
		lease->rewind_binding_state = lease->binding_state;

		write_lease(lease);
		commit_leases();
	}

	/* Send the update. */
	status = (dhcp_failover_put_message
		  (link, link -> outer,
		   FTM_BNDUPD, lease->last_xid,
		   dhcp_failover_make_option (FTO_ASSIGNED_IP_ADDRESS, FMA,
					      lease -> ip_addr.len,
					      lease -> ip_addr.iabuf),
		   dhcp_failover_make_option (FTO_BINDING_STATUS, FMA,
					      lease -> desired_binding_state),
		   lease -> uid_len
		   ? dhcp_failover_make_option (FTO_CLIENT_IDENTIFIER, FMA,
						lease -> uid_len,
						lease -> uid)
		   : &skip_failover_option,
		   lease -> hardware_addr.hlen
		   ? dhcp_failover_make_option (FTO_CHADDR, FMA,
						lease -> hardware_addr.hlen,
						lease -> hardware_addr.hbuf)
		   : &skip_failover_option,
		   dhcp_failover_make_option (FTO_LEASE_EXPIRY, FMA,
					      lease -> ends),
		   dhcp_failover_make_option (FTO_POTENTIAL_EXPIRY, FMA,
					      lease -> tstp),
		   dhcp_failover_make_option (FTO_STOS, FMA,
					      lease -> starts),
		   (lease->cltt != 0) ?
			dhcp_failover_make_option(FTO_CLTT, FMA, lease->cltt) :
			&skip_failover_option, /* No CLTT */
		   flags ? dhcp_failover_make_option(FTO_IP_FLAGS, FMA,
						     flags) :
			   &skip_failover_option, /* No IP_FLAGS */
		   &skip_failover_option,	/* XXX DDNS */
		   &skip_failover_option,	/* XXX request options */
		   &skip_failover_option,	/* XXX reply options */
		   (failover_option_t *)0));

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print (FMA, " (failed)");
	failover_print (FMA, ")");
	if (obufix) {
		log_debug ("%s", obuf);
	}
#endif
	return status;
}

/* Send a Bind ACK message. */

isc_result_t dhcp_failover_send_bind_ack (dhcp_failover_state_t *state,
					  failover_message_t *msg,
					  int reason, const char *message)
{
	dhcp_failover_link_t *link;
	isc_result_t status;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(bndack");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!state -> link_to_peer ||
	    state -> link_to_peer -> type != dhcp_type_failover_link)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)state -> link_to_peer;

	if (!link -> outer || link -> outer -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	if (!message && reason)
		message = dhcp_failover_reject_reason_print (reason);

	/* Send the update. */
	status = (dhcp_failover_put_message
		  (link, link -> outer,
		   FTM_BNDACK, msg->xid,
		   dhcp_failover_make_option (FTO_ASSIGNED_IP_ADDRESS, FMA,
					      sizeof msg -> assigned_addr,
					      &msg -> assigned_addr),
#ifdef DO_BNDACK_SHOULD_NOT
		   dhcp_failover_make_option (FTO_BINDING_STATUS, FMA,
					      msg -> binding_status),
		   (msg -> options_present & FTB_CLIENT_IDENTIFIER)
		   ? dhcp_failover_make_option (FTO_CLIENT_IDENTIFIER, FMA,
						msg -> client_identifier.count,
						msg -> client_identifier.data)
		   : &skip_failover_option,
		   (msg -> options_present & FTB_CHADDR)
		   ? dhcp_failover_make_option (FTO_CHADDR, FMA,
						msg -> chaddr.count,
						msg -> chaddr.data)
		   : &skip_failover_option,
		   dhcp_failover_make_option (FTO_LEASE_EXPIRY, FMA,
					      msg -> expiry),
		   dhcp_failover_make_option (FTO_POTENTIAL_EXPIRY, FMA,
					      msg -> potential_expiry),
		   dhcp_failover_make_option (FTO_STOS, FMA,
					      msg -> stos),
		   (msg->options_present & FTB_CLTT) ?
			dhcp_failover_make_option(FTO_CLTT, FMA, msg->cltt) :
			&skip_failover_option, /* No CLTT in the msg to ack. */
		   ((msg->options_present & FTB_IP_FLAGS) && msg->ip_flags) ?
			dhcp_failover_make_option(FTO_IP_FLAGS, FMA,
						  msg->ip_flags)
			: &skip_failover_option,
#endif /* DO_BNDACK_SHOULD_NOT */
		   reason
		    ? dhcp_failover_make_option(FTO_REJECT_REASON, FMA, reason)
		    : &skip_failover_option,
		   (reason && message)
		    ? dhcp_failover_make_option (FTO_MESSAGE, FMA,
						 strlen (message), message)
		    : &skip_failover_option,
#ifdef DO_BNDACK_SHOULD_NOT
		   &skip_failover_option,	/* XXX DDNS */
		   &skip_failover_option,	/* XXX request options */
		   &skip_failover_option,	/* XXX reply options */
#endif /* DO_BNDACK_SHOULD_NOT */
		   (failover_option_t *)0));

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print (FMA, " (failed)");
	failover_print (FMA, ")");
	if (obufix) {
		log_debug ("%s", obuf);
	}
#endif
	return status;
}

isc_result_t dhcp_failover_send_poolreq (dhcp_failover_state_t *state)
{
	dhcp_failover_link_t *link;
	isc_result_t status;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(poolreq");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!state -> link_to_peer ||
	    state -> link_to_peer -> type != dhcp_type_failover_link)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)state -> link_to_peer;

	if (!link -> outer || link -> outer -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	status = (dhcp_failover_put_message
		  (link, link -> outer,
		   FTM_POOLREQ, link->xid++,
		   (failover_option_t *)0));

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print (FMA, " (failed)");
	failover_print (FMA, ")");
	if (obufix) {
		log_debug ("%s", obuf);
	}
#endif
	return status;
}

isc_result_t dhcp_failover_send_poolresp (dhcp_failover_state_t *state,
					  int leases)
{
	dhcp_failover_link_t *link;
	isc_result_t status;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(poolresp");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!state -> link_to_peer ||
	    state -> link_to_peer -> type != dhcp_type_failover_link)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)state -> link_to_peer;

	if (!link -> outer || link -> outer -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	status = (dhcp_failover_put_message
		  (link, link -> outer,
		   FTM_POOLRESP, link->imsg->xid,
		   dhcp_failover_make_option (FTO_ADDRESSES_TRANSFERRED, FMA,
					      leases),
		   (failover_option_t *)0));

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print (FMA, " (failed)");
	failover_print (FMA, ")");
	if (obufix) {
		log_debug ("%s", obuf);
	}
#endif
	return status;
}

isc_result_t dhcp_failover_send_update_request (dhcp_failover_state_t *state)
{
	dhcp_failover_link_t *link;
	isc_result_t status;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(updreq");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!state->link_to_peer ||
	    state->link_to_peer->type != dhcp_type_failover_link)
		return (DHCP_R_INVALIDARG);
	link = (dhcp_failover_link_t *)state->link_to_peer;

	if (!link->outer || link->outer->type != omapi_type_connection)
		return (DHCP_R_INVALIDARG);

	/* We allow an update to be restarted in case we requested an update
	 * and were interrupted by something. If we had an ALL going we need
	 * to restart that.  Otherwise we simply continue with the request */
	if (state->curUPD == FTM_UPDREQALL) {
		return (dhcp_failover_send_update_request_all(state));
	}

	status = (dhcp_failover_put_message(link, link->outer, FTM_UPDREQ,
					    link->xid++, NULL));

	state->curUPD = FTM_UPDREQ;

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print(FMA, " (failed)");
	failover_print(FMA, ")");
	if (obufix) {
		log_debug("%s", obuf);
	}
#endif

	if (status == ISC_R_SUCCESS) {
		log_info("Sent update request message to %s", state->name);
	} else {
		log_error("Failed to send update request all message to %s: %s",
			 state->name, isc_result_totext(status));
	}
	return (status);
}

isc_result_t dhcp_failover_send_update_request_all (dhcp_failover_state_t
						    *state)
{
	dhcp_failover_link_t *link;
	isc_result_t status;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(updreqall");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!state->link_to_peer ||
	    state->link_to_peer->type != dhcp_type_failover_link)
		return (DHCP_R_INVALIDARG);
	link = (dhcp_failover_link_t *)state->link_to_peer;

	if (!link->outer || link->outer->type != omapi_type_connection)
		return (DHCP_R_INVALIDARG);

	/* We allow an update to be restarted in case we requested an update
	 * and were interrupted by something.
	 */

	status = (dhcp_failover_put_message(link, link->outer, FTM_UPDREQALL,
					    link->xid++, NULL));

	state->curUPD = FTM_UPDREQALL;

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print(FMA, " (failed)");
	failover_print(FMA, ")");
	if (obufix) {
		log_debug("%s", obuf);
	}
#endif

	if (status == ISC_R_SUCCESS) {
		log_info("Sent update request all message to %s", state->name);
	} else {
		log_error("Failed to send update request all message to %s: %s",
			 state->name, isc_result_totext(status));
	}
	return (status);
}

isc_result_t dhcp_failover_send_update_done (dhcp_failover_state_t *state)
{
	dhcp_failover_link_t *link;
	isc_result_t status;
#if defined (DEBUG_FAILOVER_MESSAGES)
	char obuf [64];
	unsigned obufix = 0;

# define FMA obuf, &obufix, sizeof obuf
	failover_print (FMA, "(upddone");
#else
# define FMA (char *)0, (unsigned *)0, 0
#endif

	if (!state -> link_to_peer ||
	    state -> link_to_peer -> type != dhcp_type_failover_link)
		return DHCP_R_INVALIDARG;
	link = (dhcp_failover_link_t *)state -> link_to_peer;

	if (!link -> outer || link -> outer -> type != omapi_type_connection)
		return DHCP_R_INVALIDARG;

	status = (dhcp_failover_put_message
		  (link, link -> outer,
		   FTM_UPDDONE, state->updxid,
		   (failover_option_t *)0));

#if defined (DEBUG_FAILOVER_MESSAGES)
	if (status != ISC_R_SUCCESS)
		failover_print (FMA, " (failed)");
	failover_print (FMA, ")");
	if (obufix) {
		log_debug ("%s", obuf);
	}
#endif

	log_info ("Sent update done message to %s", state -> name);

	state->updxid--; /* Paranoia, just so it mismatches. */

	/* There may be uncommitted leases at this point (since
	   dhcp_failover_process_bind_ack() doesn't commit leases);
	   commit the lease file. */
	commit_leases();

	return status;
}

/*
 * failover_lease_is_better() compares the binding update in 'msg' with
 * the current lease in 'lease'.  If the determination is that the binding
 * update shouldn't be allowed to update/crush more critical binding info
 * on the lease, the lease is preferred.  A value of true is returned if the
 * local lease is preferred, or false if the remote binding update is
 * preferred.
 *
 * For now this function is hopefully simplistic and trivial.  It may be that
 * a more detailed system of preferences is required, so this is something we
 * should monitor as we gain experience with these dueling events.
 */
static isc_boolean_t
failover_lease_is_better(dhcp_failover_state_t *state, struct lease *lease,
			 failover_message_t *msg)
{
	binding_state_t local_state;
	TIME msg_cltt;

	if (lease->binding_state != lease->desired_binding_state)
		local_state = lease->desired_binding_state;
	else
		local_state = lease->binding_state;

	if ((msg->options_present & FTB_CLTT) != 0)
		msg_cltt = msg->cltt;
	else
		msg_cltt = 0;

	switch(local_state) {
	      case FTS_ACTIVE:
		if (msg->binding_status == FTS_ACTIVE) {
			if (msg_cltt < lease->cltt)
				return ISC_TRUE;
			else if (msg_cltt > lease->cltt)
				return ISC_FALSE;
			else if (state->i_am == primary)
				return ISC_TRUE;
			else
				return ISC_FALSE;
		} else if (msg->binding_status == FTS_EXPIRED) {
			return ISC_FALSE;
		}
		/* FALL THROUGH */

	      case FTS_FREE:
	      case FTS_BACKUP:
	      case FTS_EXPIRED:
	      case FTS_RELEASED:
	      case FTS_ABANDONED:
	      case FTS_RESET:
		if (msg->binding_status == FTS_ACTIVE)
			return ISC_FALSE;
		else if (state->i_am == primary)
			return ISC_TRUE;
		else
			return ISC_FALSE;
		/* FALL THROUGH to impossible condition */

	      default:
		log_fatal("Impossible condition at %s:%d.", MDL);
	}

	log_fatal("Impossible condition at %s:%d.", MDL);
	/* Silence compiler warning. */
	return ISC_FALSE;
}

isc_result_t dhcp_failover_process_bind_update (dhcp_failover_state_t *state,
					       failover_message_t *msg)
{
	struct lease *lt = NULL, *lease = NULL;
	struct iaddr ia;
	int reason = FTR_MISC_REJECT;
	const char *message;
	int new_binding_state;
	int send_to_backup = 0;
	int required_options;
	isc_boolean_t chaddr_changed = ISC_FALSE;
	isc_boolean_t ident_changed = ISC_FALSE;

	/* Validate the binding update. */
	required_options = FTB_ASSIGNED_IP_ADDRESS | FTB_BINDING_STATUS;
	if ((msg->options_present & required_options) != required_options) {
		message = "binding update lacks required options";
		reason = FTR_MISSING_BINDINFO;
		goto bad;
	}

	ia.len = sizeof msg -> assigned_addr;
	memcpy (ia.iabuf, &msg -> assigned_addr, ia.len);

	if (!find_lease_by_ip_addr (&lease, ia, MDL)) {
		message = "unknown IP address";
		reason = FTR_ILLEGAL_IP_ADDR;
		goto bad;
	}

	/*
	 * If this lease is covered by a different failover peering
	 * relationship, assert an error.
	 */
	if ((lease->pool == NULL) || (lease->pool->failover_peer == NULL) ||
	    (lease->pool->failover_peer != state)) {
		message = "IP address is covered by a different failover "
			  "relationship state";
		reason = FTR_ILLEGAL_IP_ADDR;
		goto bad;
	}

	/*
	 * Dueling updates:  This happens when both servers send a BNDUPD
	 * at the same time.  We want the best update to win, which means
	 * we reject if we think ours is better, or cancel if we think the
	 * peer's is better.  We only assert a problem if the lease is on
	 * the ACK queue, not on the UPDATE queue.  This means that after
	 * accepting this server's BNDUPD, we will send our own BNDUPD
	 * /after/ sending the BNDACK (this order was recently enforced in
	 * queue processing).
	 */
	if ((lease->flags & ON_ACK_QUEUE) != 0) {
		if (failover_lease_is_better(state, lease, msg)) {
			message = "incoming update is less critical than "
				  "outgoing update";
			reason = FTR_LESS_CRIT_BIND_INFO;
			goto bad;
		} else {
			/* This makes it so we ignore any spurious ACKs. */
			dhcp_failover_ack_queue_remove(state, lease);
		}
	}

	/* Install the new info.  Start by taking a copy to markup. */
	if (!lease_copy (&lt, lease, MDL)) {
		message = "no memory";
		goto bad;
	}

	if (msg -> options_present & FTB_CHADDR) {
		if (msg->binding_status == FTS_ABANDONED) {
			message = "BNDUPD to ABANDONED with a CHADDR";
			goto bad;
		}
		if (msg -> chaddr.count > sizeof lt -> hardware_addr.hbuf) {
			message = "chaddr too long";
			goto bad;
		}

		if ((lt->hardware_addr.hlen != msg->chaddr.count) ||
		    (memcmp(lt->hardware_addr.hbuf, msg->chaddr.data,
			    msg->chaddr.count) != 0))
			chaddr_changed = ISC_TRUE;

		lt -> hardware_addr.hlen = msg -> chaddr.count;
		memcpy (lt -> hardware_addr.hbuf, msg -> chaddr.data,
			msg -> chaddr.count);
	} else if (msg->binding_status == FTS_ACTIVE ||
		   msg->binding_status == FTS_EXPIRED ||
		   msg->binding_status == FTS_RELEASED) {
		message = "BNDUPD without CHADDR";
		reason = FTR_MISSING_BINDINFO;
		goto bad;
	} else if (msg->binding_status == FTS_ABANDONED) {
		chaddr_changed = ISC_TRUE;
		lt->hardware_addr.hlen = 0;
		if (lt->scope)
			binding_scope_dereference(&lt->scope, MDL);
	}

	/* There is no explicit message content to indicate that the client
	 * supplied no client-identifier.  So if we don't hear of a value,
	 * we discard the last one.
	 */
	if (msg->options_present & FTB_CLIENT_IDENTIFIER) {
		if (msg->binding_status == FTS_ABANDONED) {
			message = "BNDUPD to ABANDONED with client-id";
			goto bad;
		}

		if ((lt->uid_len != msg->client_identifier.count) ||
		    (lt->uid == NULL) || /* Sanity; should never happen. */
		    (memcmp(lt->uid, msg->client_identifier.data,
			    lt->uid_len) != 0))
			ident_changed = ISC_TRUE;

		lt->uid_len = msg->client_identifier.count;

		/* Allocate the lt->uid buffer if we haven't already, or
		 * re-allocate the lt-uid buffer if we have one that is not
		 * large enough.  Otherwise, just use the extant buffer.
		 */
		if (!lt->uid || lt->uid == lt->uid_buf ||
		    lt->uid_len > lt->uid_max) {
			if (lt->uid && lt->uid != lt->uid_buf)
				dfree(lt->uid, MDL);

			if (lt->uid_len > sizeof(lt->uid_buf)) {
				lt->uid_max = lt->uid_len;
				lt->uid = dmalloc(lt->uid_len, MDL);
				if (!lt->uid) {
					message = "no memory";
					goto bad;
				}
			} else {
				lt->uid_max = sizeof(lt->uid_buf);
				lt->uid = lt->uid_buf;
			}
		}
		memcpy (lt -> uid,
			msg -> client_identifier.data, lt -> uid_len);
	} else if (lt->uid && msg->binding_status != FTS_RESET &&
		   msg->binding_status != FTS_FREE &&
		   msg->binding_status != FTS_BACKUP) {
		ident_changed = ISC_TRUE;
		if (lt->uid != lt->uid_buf)
			dfree (lt->uid, MDL);
		lt->uid = NULL;
		lt->uid_max = lt->uid_len = 0;
	}

	/*
	 * A server's configuration can assign a 'binding scope';
	 *
	 *	set var = "value";
	 *
	 * The problem with these binding scopes is that they are refreshed
	 * when the server processes a client's DHCP packet.  A local binding
	 * scope is trash, then, when the lease has been assigned by the
	 * partner server.  There is no real way to detect this, a peer may
	 * be updating us (as through potential conflict) with a binding we
	 * sent them, but we can trivially detect the /problematic/ case;
	 *
	 *	lease is free.
	 *	primary allocates lease to client A, assigns ddns name A.
	 *	primary fails.
	 *	secondary enters partner down.
	 *	lease expires, and is set free.
	 *	lease is allocated to client B and given ddns name B.
	 *	primary recovers.
	 *
	 * The binding update in this case will be active->active, but the
	 * client identification on the lease will have changed.  The ddns
	 * update on client A will have leaked if we just remove the binding
	 * scope blindly.
	 */
	if (msg->binding_status == FTS_ACTIVE &&
	    (chaddr_changed || ident_changed)) {
#if defined (NSUPDATE)
		(void) ddns_removals(lease, NULL, NULL, ISC_FALSE);
#endif /* NSUPDATE */

		if (lease->scope != NULL)
			binding_scope_dereference(&lease->scope, MDL);
	}

	/* XXX Times may need to be adjusted based on clock skew! */
	if (msg -> options_present & FTB_STOS) {
		lt -> starts = msg -> stos;
	}
	if (msg -> options_present & FTB_LEASE_EXPIRY) {
		lt -> ends = msg -> expiry;
	}
	if (msg->options_present & FTB_POTENTIAL_EXPIRY) {
		lt->atsfp = lt->tsfp = msg->potential_expiry;
	}
	if (msg->options_present & FTB_IP_FLAGS) {
		if (msg->ip_flags & FTF_IP_FLAG_RESERVE) {
			if ((((state->i_am == primary) &&
			      (lease->binding_state == FTS_FREE)) ||
			     ((state->i_am == secondary) &&
			      (lease->binding_state == FTS_BACKUP))) &&
			    !(lease->flags & RESERVED_LEASE)) {
				message = "Address is not reserved.";
				reason = FTR_IP_NOT_RESERVED;
				goto bad;
			}

			lt->flags |= RESERVED_LEASE;
		} else
			lt->flags &= ~RESERVED_LEASE;

		if (msg->ip_flags & FTF_IP_FLAG_BOOTP) {
			if ((((state->i_am == primary) &&
			      (lease->binding_state == FTS_FREE)) ||
			     ((state->i_am == secondary) &&
			      (lease->binding_state == FTS_BACKUP))) &&
			    !(lease->flags & BOOTP_LEASE)) {
				message = "Address is not allocated to BOOTP.";
				goto bad;
			}
			lt->flags |= BOOTP_LEASE;
		} else
			lt->flags &= ~BOOTP_LEASE;

		if (msg->ip_flags & ~(FTF_IP_FLAG_RESERVE | FTF_IP_FLAG_BOOTP))
			log_info("Unknown IP-flags set in BNDUPD (0x%x).",
				 msg->ip_flags);
	} else /* Flags may only not appear if the values are zero. */
		lt->flags &= ~(RESERVED_LEASE | BOOTP_LEASE);

#if defined (DEBUG_LEASE_STATE_TRANSITIONS)
	log_info ("processing state transition for %s: %s to %s",
		  piaddr (lease -> ip_addr),
		  binding_state_print (lease -> binding_state),
		  binding_state_print (msg -> binding_status));
#endif

	/* If we're in normal state, make sure the state transition
	   we got is valid. */
	if (state -> me.state == normal) {
		new_binding_state =
			(normal_binding_state_transition_check
			 (lease, state, msg -> binding_status,
			  msg -> potential_expiry));
		/* XXX if the transition the peer asked for isn't
		   XXX allowed, maybe we should make the transition
		   XXX into potential-conflict at this point. */
	} else {
		new_binding_state =
			(conflict_binding_state_transition_check
			 (lease, state, msg -> binding_status,
			  msg -> potential_expiry));
	}
	if (new_binding_state != msg -> binding_status) {
		char outbuf [100];

		if (snprintf (outbuf, sizeof outbuf,
			  "%s: invalid state transition: %s to %s",
			  piaddr (lease -> ip_addr),
			  binding_state_print (lease -> binding_state),
			  binding_state_print (msg -> binding_status))
					>= sizeof outbuf)
			log_fatal ("%s: impossible outbuf overflow",
				"dhcp_failover_process_bind_update");

		dhcp_failover_send_bind_ack (state, msg,
					     FTR_FATAL_CONFLICT,
					     outbuf);
		goto out;
	}
	if (new_binding_state == FTS_EXPIRED ||
	    new_binding_state == FTS_RELEASED ||
	    new_binding_state == FTS_RESET) {
		lt -> next_binding_state = FTS_FREE;

		/* Mac address affinity.  Assign the lease to
		 * BACKUP state if we are the primary and the
		 * peer is more likely to reallocate this lease
		 * to a returning client.
		 */
		if ((state->i_am == primary) &&
		    !(lt->flags & (RESERVED_LEASE | BOOTP_LEASE)))
			send_to_backup = peer_wants_lease(lt);
	} else {
		lt -> next_binding_state = new_binding_state;
	}
	msg -> binding_status = lt -> next_binding_state;

	/*
	 * If we accept a peer's binding update, then we can't rewind a
	 * lease behind the peer's state.
	 */
	lease->rewind_binding_state = lt->next_binding_state;

	/* Try to install the new information. */
	if (!supersede_lease (lease, lt, 0, 0, 0, 0) ||
	    !write_lease (lease)) {
		message = "database update failed";
	      bad:
		dhcp_failover_send_bind_ack (state, msg, reason, message);
		goto out;
	} else {
		dhcp_failover_queue_ack (state, msg);
	}

	/* If it is probably wise, assign lease to backup state if the peer
	 * is not already hoarding leases.
	 */
	if (send_to_backup && secondary_not_hoarding(state, lease->pool)) {
		lease->next_binding_state = FTS_BACKUP;
		lease->tstp = cur_time;
		lease->starts = cur_time;

		if (!supersede_lease(lease, NULL, 0, 1, 0, 0) ||
		    !write_lease(lease))
			log_error("can't commit lease %s for mac addr "
				  "affinity", piaddr(lease->ip_addr));

		dhcp_failover_send_updates(state);
	}

      out:
	if (lt)
		lease_dereference (&lt, MDL);
	if (lease)
		lease_dereference (&lease, MDL);

	return ISC_R_SUCCESS;
}

/* This was hairy enough I didn't want to do it all in an if statement.
 *
 * Returns: Truth is the secondary is allowed to get more leases based upon
 * MAC address affinity.  False otherwise.
 */
static inline int
secondary_not_hoarding(dhcp_failover_state_t *state, struct pool *p) {
	int total;
	int hold;
	int lts;

	total = p->free_leases + p->backup_leases;

	/* How many leases is one side or the other allowed to "hold"? */
	hold = ((total * state->max_lease_ownership) + 50) / 100;

	/* If we were to send leases (or if the secondary were to send us
	 * leases in the negative direction), how many would that be?
	 */
	lts = (p->free_leases - p->backup_leases) / 2;

	/* The peer is not hoarding leases if we would send them more leases
	 * (or they would take fewer leases) than the maximum they are allowed
	 * to hold (the negative hold).
	 */
	return(lts > -hold);
}

isc_result_t dhcp_failover_process_bind_ack (dhcp_failover_state_t *state,
					     failover_message_t *msg)
{
	struct lease *lease = NULL;
	struct iaddr ia;
	const char *message = "no memory";
	u_int32_t pot_expire;
	int send_to_backup = ISC_FALSE;
	struct timeval tv;

	ia.len = sizeof msg -> assigned_addr;
	memcpy (ia.iabuf, &msg -> assigned_addr, ia.len);

	if (!find_lease_by_ip_addr (&lease, ia, MDL)) {
		message = "no such lease";
		goto bad;
	}

	/* XXX check for conflicts. */
	if (msg -> options_present & FTB_REJECT_REASON) {
		log_error ("bind update on %s from %s rejected: %.*s",
			   piaddr (ia), state -> name,
			   (int)((msg -> options_present & FTB_MESSAGE)
				 ? msg -> message.count
				 : strlen (dhcp_failover_reject_reason_print
					   (msg -> reject_reason))),
			   (msg -> options_present & FTB_MESSAGE)
			   ? (const char *)(msg -> message.data)
			   : (dhcp_failover_reject_reason_print
			      (msg -> reject_reason)));
		goto unqueue;
	}

	/* Silently discard acks for leases we did not update (or multiple
	 * acks).
	 */
	if (!lease->last_xid)
		goto unqueue;

	if (lease->last_xid != msg->xid) {
		message = "xid mismatch";
		goto bad;
	}

	/* XXX Times may need to be adjusted based on clock skew! */
	if (msg->options_present & FTO_POTENTIAL_EXPIRY)
		pot_expire = msg->potential_expiry;
	else
		pot_expire = lease->tstp;

	/* If the lease was desired to enter a binding state, we set
	 * such a value upon transmitting a bndupd.  We do not clear it
	 * if we receive a bndupd in the meantime (or change the state
	 * of the lease again ourselves), but we do set binding_state
	 * if we get a bndupd.
	 *
	 * So desired_binding_state tells us what we sent a bndupd for,
	 * and binding_state tells us what we have since determined in
	 * the meantime.
	 */
	if (lease->desired_binding_state == FTS_EXPIRED ||
	    lease->desired_binding_state == FTS_RESET ||
	    lease->desired_binding_state == FTS_RELEASED)
	{
		/* It is not a problem to do this directly as we call
		 * supersede_lease immediately after: the lease is requeued
		 * even if its sort order (tsfp) has changed.
		 */
		lease->atsfp = lease->tsfp = pot_expire;
		if ((state->i_am == secondary) &&
		    (lease->flags & RESERVED_LEASE))
			lease->next_binding_state = FTS_BACKUP;
		else
			lease->next_binding_state = FTS_FREE;

		/* Clear this condition for the next go-round. */
		lease->desired_binding_state = lease->next_binding_state;

		/* The peer will have made this state change, so set rewind. */
		lease->rewind_binding_state = lease->next_binding_state;

		supersede_lease(lease, NULL, 0, 0, 0, 0);
		write_lease(lease);

		/* Lease has returned to FREE state from the
		 * transitional states.  If the lease 'belongs'
		 * to a client that would be served by the
		 * peer, process a binding update now to send
		 * the lease to backup state.  But not if we
		 * think we already have.
		 */
		if (state->i_am == primary &&
		    !(lease->flags & (RESERVED_LEASE | BOOTP_LEASE)) &&
		    peer_wants_lease(lease))
			send_to_backup = ISC_TRUE;

		if (!send_to_backup && state->me.state == normal)
			commit_leases();
	} else {
		/* XXX It could be a problem to do this directly if the lease
		 * XXX is sorted by tsfp.
		 */
		lease->atsfp = lease->tsfp = pot_expire;
		if (lease->desired_binding_state != lease->binding_state) {
			lease->next_binding_state =
				lease->desired_binding_state;
			supersede_lease(lease, NULL, 0, 0, 0, 0);
		}
		write_lease(lease);
		/* Commit the lease only after a two-second timeout,
		   so that if we get a bunch of acks in quick
		   succession (e.g., when stealing leases from the
		   secondary), we do not do an immediate commit for
		   each one. */
		tv.tv_sec = cur_time + 2;
		tv.tv_usec = 0;
		add_timeout(&tv, commit_leases_timeout, (void *)0, 0, 0);
	}

      unqueue:
	dhcp_failover_ack_queue_remove (state, lease);

	/* If we are supposed to send an update done after we send
	   this lease, go ahead and send it. */
	if (state -> send_update_done == lease) {
		lease_dereference (&state -> send_update_done, MDL);
		dhcp_failover_send_update_done (state);
	}

	/* Now that the lease is off the ack queue, consider putting it
	 * back on the update queue for mac address affinity.
	 */
	if (send_to_backup && secondary_not_hoarding(state, lease->pool)) {
		lease->next_binding_state = FTS_BACKUP;
		lease->tstp = lease->starts = cur_time;

		if (!supersede_lease(lease, NULL, 0, 1, 0, 0) ||
		    !write_lease(lease))
			log_error("can't commit lease %s for "
				  "client affinity", piaddr(lease->ip_addr));

		if (state->me.state == normal)
			commit_leases();
	}

	/* If there are updates pending, we've created space to send at
	   least one. */
	dhcp_failover_send_updates (state);

      out:
	lease_dereference (&lease, MDL);
	return ISC_R_SUCCESS;

      bad:
	log_info ("bind update on %s got ack from %s: %s.",
		  piaddr (ia), state -> name, message);
	goto out;
}

isc_result_t dhcp_failover_generate_update_queue (dhcp_failover_state_t *state,
						  int everythingp)
{
	struct shared_network *s;
	struct pool *p;
	struct lease *l;
	int i;
#define FREE_LEASES 0
#define ACTIVE_LEASES 1
#define EXPIRED_LEASES 2
#define ABANDONED_LEASES 3
#define BACKUP_LEASES 4
#define RESERVED_LEASES 5
	LEASE_STRUCT_PTR lptr[RESERVED_LEASES+1];

	/* Loop through each pool in each shared network and call the
	   expiry routine on the pool. */
	for (s = shared_networks; s; s = s -> next) {
	    for (p = s -> pools; p; p = p -> next) {
		if (p->failover_peer != state)
			continue;

		lptr[FREE_LEASES] = &p->free;
		lptr[ACTIVE_LEASES] = &p->active;
		lptr[EXPIRED_LEASES] = &p->expired;
		lptr[ABANDONED_LEASES] = &p->abandoned;
		lptr[BACKUP_LEASES] = &p->backup;
		lptr[RESERVED_LEASES] = &p->reserved;

		for (i = FREE_LEASES; i <= RESERVED_LEASES; i++) {
		    for (l = LEASE_GET_FIRSTP(lptr[i]);
			 l != NULL;
			 l = LEASE_GET_NEXTP(lptr[i], l)) {
			if ((l->flags & ON_QUEUE) == 0 &&
			    (everythingp ||
			     (l->tstp > l->atsfp) ||
			     (i == EXPIRED_LEASES))) {
				l -> desired_binding_state = l -> binding_state;
				dhcp_failover_queue_update (l, 0);
			}
		    }
		}
	    }
	}
	return ISC_R_SUCCESS;
}

isc_result_t
dhcp_failover_process_update_request (dhcp_failover_state_t *state,
				      failover_message_t *msg)
{
	if (state->send_update_done) {
		log_info("Received update request while old update still "
			 "flying!  Silently discarding old request.");
		lease_dereference(&state->send_update_done, MDL);
	}

	/* Generate a fresh update queue. */
	dhcp_failover_generate_update_queue (state, 0);

	state->updxid = msg->xid;

	/* If there's anything on the update queue (there shouldn't be
	   anything on the ack queue), trigger an update done message
	   when we get an ack for that lease. */
	if (state -> update_queue_tail) {
		lease_reference (&state -> send_update_done,
				 state -> update_queue_tail, MDL);
		dhcp_failover_send_updates (state);
		log_info ("Update request from %s: sending update",
			   state -> name);
	} else {
		/* Otherwise, there are no updates to send, so we can
		   just send an UPDDONE message immediately. */
		dhcp_failover_send_update_done (state);
		log_info ("Update request from %s: nothing pending",
			   state -> name);
	}

	return ISC_R_SUCCESS;
}

isc_result_t
dhcp_failover_process_update_request_all (dhcp_failover_state_t *state,
					  failover_message_t *msg)
{
	if (state->send_update_done) {
		log_info("Received update request while old update still "
			 "flying!  Silently discarding old request.");
		lease_dereference(&state->send_update_done, MDL);
	}

	/* Generate a fresh update queue that includes every lease. */
	dhcp_failover_generate_update_queue (state, 1);

	state->updxid = msg->xid;

	if (state -> update_queue_tail) {
		lease_reference (&state -> send_update_done,
				 state -> update_queue_tail, MDL);
		dhcp_failover_send_updates (state);
		log_info ("Update request all from %s: sending update",
			   state -> name);
	} else {
		/* This should really never happen, but it could happen
		   on a server that currently has no leases configured. */
		dhcp_failover_send_update_done (state);
		log_info ("Update request all from %s: nothing pending",
			   state -> name);
	}

	return ISC_R_SUCCESS;
}

isc_result_t
dhcp_failover_process_update_done (dhcp_failover_state_t *state,
				   failover_message_t *msg)
{
	struct timeval tv;

	log_info ("failover peer %s: peer update completed.",
		  state -> name);

	state -> curUPD = 0;

	switch (state -> me.state) {
	      case unknown_state:
	      case partner_down:
	      case normal:
	      case communications_interrupted:
	      case resolution_interrupted:
	      case shut_down:
	      case paused:
	      case recover_done:
	      case startup:
	      case recover_wait:
		break;	/* shouldn't happen. */

		/* We got the UPDDONE, so we can go into normal state! */
	      case potential_conflict:
		if (state->partner.state == conflict_done) {
			if (state->i_am == secondary) {
				dhcp_failover_set_state (state, normal);
			} else {
				log_error("Secondary is in conflict_done "
					  "state after conflict resolution, "
					  "this is illegal.");
				dhcp_failover_set_state (state, shut_down);
			}
		} else {
			if (state->i_am == primary)
				dhcp_failover_set_state (state, conflict_done);
			else
				log_error("Spurious update-done message.");
		}

		break;

	      case conflict_done:
		log_error("Spurious update-done message.");
		break;

	      case recover:
		/* Wait for MCLT to expire before moving to recover_done,
		   except that if both peers come up in recover, there is
		   no point in waiting for MCLT to expire - this probably
		   indicates the initial startup of a newly-configured
		   failover pair. */
		if (state -> me.stos + state -> mclt > cur_time &&
		    state -> partner.state != recover &&
		    state -> partner.state != recover_done) {
			dhcp_failover_set_state (state, recover_wait);
#if defined (DEBUG_FAILOVER_TIMING)
			log_info ("add_timeout +%d %s",
				  (int)(cur_time -
					state -> me.stos + state -> mclt),
				  "dhcp_failover_recover_done");
#endif
			tv . tv_sec = (int)(state -> me.stos + state -> mclt);
			tv . tv_usec = 0;
			add_timeout (&tv,
				     dhcp_failover_recover_done,
				     state,
				     (tvref_t)omapi_object_reference,
				     (tvunref_t)
				     omapi_object_dereference);
		} else
			dhcp_failover_recover_done (state);
	}

	return ISC_R_SUCCESS;
}

void dhcp_failover_recover_done (void *sp)
{
	dhcp_failover_state_t *state = sp;

#if defined (DEBUG_FAILOVER_TIMING)
	log_info ("dhcp_failover_recover_done");
#endif

	dhcp_failover_set_state (state, recover_done);
}

#if defined (DEBUG_FAILOVER_MESSAGES)
/* Print hunks of failover messages, doing line breaks as appropriate.
   Note that this assumes syslog is being used, rather than, e.g., the
   Windows NT logging facility, where just dumping the whole message in
   one hunk would be more appropriate. */

void failover_print (char *obuf,
		     unsigned *obufix, unsigned obufmax, const char *s)
{
	int len = strlen (s);

	while (len + *obufix + 1 >= obufmax) {
		log_debug ("%s", obuf);
		if (!*obufix) {
			log_debug ("%s", s);
			*obufix = 0;
			return;
		}
		*obufix = 0;
	}
	strcpy (&obuf [*obufix], s);
	*obufix += len;
}
#endif /* defined (DEBUG_FAILOVER_MESSAGES) */

/* Taken from draft-ietf-dhc-loadb-01.txt: */
/* A "mixing table" of 256 distinct values, in pseudo-random order. */
unsigned char loadb_mx_tbl[256] = {
    251, 175, 119, 215,  81,  14,  79, 191, 103,  49,
    181, 143, 186, 157,   0, 232,  31,  32,  55,  60,
    152,  58,  17, 237, 174,  70, 160, 144, 220,  90,
    57,  223,  59,   3,  18, 140, 111, 166, 203, 196,
    134, 243, 124,  95, 222, 179, 197,  65, 180,  48,
     36,  15, 107,  46, 233, 130, 165,  30, 123, 161,
    209,  23,  97,  16,  40,  91, 219,  61, 100,  10,
    210, 109, 250, 127,  22, 138,  29, 108, 244,  67,
    207,   9, 178, 204,  74,  98, 126, 249, 167, 116,
    34,   77, 193, 200, 121,   5,  20, 113,  71,  35,
    128,  13, 182,  94,  25, 226, 227, 199,  75,  27,
     41, 245, 230, 224,  43, 225, 177,  26, 155, 150,
    212, 142, 218, 115, 241,  73,  88, 105,  39, 114,
     62, 255, 192, 201, 145, 214, 168, 158, 221, 148,
    154, 122,  12,  84,  82, 163,  44, 139, 228, 236,
    205, 242, 217,  11, 187, 146, 159,  64,  86, 239,
    195,  42, 106, 198, 118, 112, 184, 172,  87,   2,
    173, 117, 176, 229, 247, 253, 137, 185,  99, 164,
    102, 147,  45,  66, 231,  52, 141, 211, 194, 206,
    246, 238,  56, 110,  78, 248,  63, 240, 189,  93,
     92,  51,  53, 183,  19, 171,  72,  50,  33, 104,
    101,  69,   8, 252,  83, 120,  76, 135,  85,  54,
    202, 125, 188, 213,  96, 235, 136, 208, 162, 129,
    190, 132, 156,  38,  47,   1,   7, 254,  24,   4,
    216, 131,  89,  21,  28, 133,  37, 153, 149,  80,
    170,  68,   6, 169, 234, 151 };

static unsigned char loadb_p_hash (const unsigned char *, unsigned);
static unsigned char loadb_p_hash (const unsigned char *key, unsigned len)
{
        unsigned char hash = len;
        int i;
        for(i = len; i > 0;  )
		hash = loadb_mx_tbl [hash ^ (key [--i])];
        return hash;
}

int load_balance_mine (struct packet *packet, dhcp_failover_state_t *state)
{
	struct option_cache *oc;
	struct data_string ds;
	unsigned char hbaix;
	int hm;
	u_int16_t ec;

	ec = ntohs(packet->raw->secs);

	/*
	 * If desired check to see if the secs field may have been byte
	 * swapped.  We assume it has if the high order byte isn't cleared
	 * while the low order byte is cleared.  In this case we swap the
	 * bytes and continue processing.
	 */
	if ((check_secs_byte_order == 1) &&
	    ((ec > 255) && ((ec & 0xff) == 0))) {
		ec = (ec >> 8) | (ec << 8);
	}

	if ((state->load_balance_max_secs == 0) ||
	    (state->load_balance_max_secs < ec)) {
		return (1);
	}

	/* If we don't have a hash bucket array, we can't tell if this
	   one's ours, so we assume it's not. */
	if (!state->hba)
		return (0);

	oc = lookup_option(&dhcp_universe, packet->options,
			   DHO_DHCP_CLIENT_IDENTIFIER);
	memset(&ds, 0, sizeof ds);
	if (oc &&
	    evaluate_option_cache(&ds, packet, NULL, NULL,
				  packet->options, NULL,
				  &global_scope, oc, MDL)) {
		hbaix = loadb_p_hash(ds.data, ds.len);

		data_string_forget(&ds, MDL);
	} else {
		hbaix = loadb_p_hash(packet->raw->chaddr,
				     packet->raw->hlen);
	}

	hm = state->hba[(hbaix >> 3) & 0x1F] & (1 << (hbaix & 0x07));

	if (state->i_am == primary)
		return (hm);
	else
		return (!hm);
}

/* The inverse of load_balance_mine ("load balance theirs").  We can't
 * use the regular load_balance_mine() and invert it because of the case
 * where there might not be an HBA, and we want to indicate false here
 * in this case only.
 */
int
peer_wants_lease(struct lease *lp)
{
	dhcp_failover_state_t *state;
	unsigned char hbaix;
	int hm;

	if (!lp->pool)
		return 0;

	state = lp->pool->failover_peer;

	if (!state || !state->hba)
		return 0;

	if (lp->uid_len)
		hbaix = loadb_p_hash(lp->uid, lp->uid_len);
	else if (lp->hardware_addr.hlen > 1)
		/* Skip the first byte, which is the hardware type, and is
		 * not included during actual load balancing checks above
		 * since it is separate from the packet header chaddr field.
		 * The remainder of the hardware address should be identical
		 * to the chaddr contents.
		 */
		hbaix = loadb_p_hash(lp->hardware_addr.hbuf + 1,
				     lp->hardware_addr.hlen - 1);
	else /* impossible to categorize into LBA */
		return 0;

	hm = state->hba[(hbaix >> 3) & 0x1F] & (1 << (hbaix & 0x07));

	if (state->i_am == primary)
		return !hm;
	else
		return hm;
}

/* This deals with what to do with bind updates when
   we're in the normal state

   Note that tsfp had better be set from the latest bind update
   _before_ this function is called! */

binding_state_t
normal_binding_state_transition_check (struct lease *lease,
				       dhcp_failover_state_t *state,
				       binding_state_t binding_state,
				       u_int32_t tsfp)
{
	binding_state_t new_state;

	/* If there is no transition, it's no problem. */
	if (binding_state == lease -> binding_state)
		return binding_state;

	switch (lease -> binding_state) {
	      case FTS_FREE:
	      case FTS_ABANDONED:
		switch (binding_state) {
		      case FTS_ACTIVE:
		      case FTS_ABANDONED:
		      case FTS_BACKUP:
		      case FTS_EXPIRED:
		      case FTS_RELEASED:
		      case FTS_RESET:
			/* If the lease was free, and our peer is primary,
			   then it can make it active, or abandoned, or
			   backup.    Abandoned is treated like free in
			   this case. */
			if (state -> i_am == secondary)
				return binding_state;

			/* Otherwise, it can't legitimately do any sort of
			   state transition.   Because the lease was free,
			   and the error has already been made, we allow the
			   peer to change its state anyway, but log a warning
			   message in hopes that the error will be fixed. */
		      case FTS_FREE: /* for compiler */
			new_state = binding_state;
			goto out;

		      default:
			log_fatal ("Impossible case at %s:%d.", MDL);
			return FTS_RESET;
		}
	      case FTS_ACTIVE:
		/* The secondary can't change the state of an active
		   lease. */
		if (state -> i_am == primary) {
			/* Except that the client may send the DHCPRELEASE
			   to the secondary.  We also allow for when the
                           secondary gets a DECLINE and the primary does not.*/
                        if ((binding_state == FTS_RELEASED) ||
                            (binding_state == FTS_ABANDONED))
                                return binding_state;

			new_state = lease -> binding_state;
			goto out;
		}

		/* So this is only for transitions made by the primary: */
		switch (binding_state) {
		      case FTS_FREE:
		      case FTS_BACKUP:
			/* Can't set a lease to free or backup until the
			   peer agrees that it's expired. */
			if (tsfp > cur_time) {
				new_state = lease -> binding_state;
				goto out;
			}
			return binding_state;

		      case FTS_EXPIRED:
			/* XXX 65 should be the clock skew between the peers
			   XXX plus a fudge factor.   This code will result
			   XXX in problems if MCLT is really short or the
			   XXX max-lease-time is really short (less than the
			   XXX fudge factor. */
			if (lease -> ends - 65 > cur_time) {
				new_state = lease -> binding_state;
				goto out;
			}

		      case FTS_RELEASED:
		      case FTS_ABANDONED:
		      case FTS_RESET:
		      case FTS_ACTIVE:
			return binding_state;

		      default:
			log_fatal ("Impossible case at %s:%d.", MDL);
			return FTS_RESET;
		}
		break;
	      case FTS_EXPIRED:
		switch (binding_state) {
		      case FTS_BACKUP:
		      case FTS_FREE:
			/* Can't set a lease to free or backup until the
			   peer agrees that it's expired. */
			if (tsfp > cur_time) {
				new_state = lease -> binding_state;
				goto out;
			}
			return binding_state;

		      case FTS_ACTIVE:
		      case FTS_RELEASED:
		      case FTS_ABANDONED:
		      case FTS_RESET:
		      case FTS_EXPIRED:
			return binding_state;

		      default:
			log_fatal ("Impossible case at %s:%d.", MDL);
			return FTS_RESET;
		}
	      case FTS_RELEASED:
		switch (binding_state) {
		      case FTS_FREE:
		      case FTS_BACKUP:

			/* These are invalid state transitions - should we
			   prevent them? */
		      case FTS_EXPIRED:
		      case FTS_ABANDONED:
		      case FTS_RESET:
		      case FTS_ACTIVE:
		      case FTS_RELEASED:
			return binding_state;

		      default:
			log_fatal ("Impossible case at %s:%d.", MDL);
			return FTS_RESET;
		}
	      case FTS_RESET:
		switch (binding_state) {
		      case FTS_FREE:
		      case FTS_BACKUP:
			/* Can't set a lease to free or backup until the
			   peer agrees that it's expired. */
			if (tsfp > cur_time) {
				new_state = lease -> binding_state;
				goto out;
			}
			return binding_state;

		      case FTS_ACTIVE:
		      case FTS_EXPIRED:
		      case FTS_RELEASED:
		      case FTS_ABANDONED:
		      case FTS_RESET:
			return binding_state;

		      default:
			log_fatal ("Impossible case at %s:%d.", MDL);
			return FTS_RESET;
		}
	      case FTS_BACKUP:
		switch (binding_state) {
		      case FTS_ACTIVE:
		      case FTS_ABANDONED:
		      case FTS_EXPIRED:
		      case FTS_RELEASED:
		      case FTS_RESET:
			/* If the lease was in backup, and our peer
			   is secondary, then it can make it active
			   or abandoned. */
			if (state -> i_am == primary)
				return binding_state;

			/* Either the primary or the secondary can
			   reasonably move a lease from the backup
			   state to the free state. */
		      case FTS_FREE:
			return binding_state;

		      case FTS_BACKUP:
			new_state = lease -> binding_state;
			goto out;

		      default:
			log_fatal ("Impossible case at %s:%d.", MDL);
			return FTS_RESET;
		}

	      default:
		log_fatal ("Impossible case at %s:%d.", MDL);
		return FTS_RESET;
	}
      out:
	return new_state;
}

/* Determine whether the state transition is okay when we're potentially
   in conflict with the peer. */
binding_state_t
conflict_binding_state_transition_check (struct lease *lease,
					 dhcp_failover_state_t *state,
					 binding_state_t binding_state,
					 u_int32_t tsfp)
{
	binding_state_t new_state;

	/* If there is no transition, it's no problem. */
	if (binding_state == lease -> binding_state)
		new_state = binding_state;
	else {
		switch (lease -> binding_state) {
			/* If we think the lease is not in use, then the
			   state into which the partner put it is just fine,
			   whatever it is. */
		      case FTS_FREE:
		      case FTS_ABANDONED:
		      case FTS_EXPIRED:
		      case FTS_RELEASED:
		      case FTS_RESET:
		      case FTS_BACKUP:
			new_state = binding_state;
			break;

			/* If we think the lease *is* in use, then we're not
			   going to take the partner's change if the partner
			   thinks it's free. */
		      case FTS_ACTIVE:
			switch (binding_state) {
			      case FTS_FREE:
			      case FTS_BACKUP:
				new_state = lease -> binding_state;
				break;

			      case FTS_EXPIRED:
				/* If we don't agree about expiry, it's
				 * invalid.  65 should allow for max
				 * clock skew (60) plus some fudge.
				 * XXX: should we refetch cur_time?
				 */
				if ((lease->ends - 65) > cur_time)
					new_state = lease->binding_state;
				else
					new_state = binding_state;
				break;

				/* RELEASED, RESET, and ABANDONED indicate
				 * that our partner has information about
				 * this lease that we did not witness.  Our
				 * partner wins.
				 */
			      case FTS_RELEASED:
			      case FTS_RESET:
			      case FTS_ABANDONED:
				new_state = binding_state;
				break;

			      default:
				log_fatal ("Impossible case at %s:%d.", MDL);
				return FTS_RESET;
			}
			break;

		      default:
			log_fatal ("Impossible case at %s:%d.", MDL);
			return FTS_RESET;
		}
	}
	return new_state;
}

/* We can reallocate a lease under the following circumstances:

   (1) It belongs to us - it's FTS_FREE, and we're primary, or it's
       FTS_BACKUP, and we're secondary.
   (2) We're in partner_down, and the lease is not active, and we
       can be sure that the other server didn't make it active.
       We can only be sure that the server didn't make it active
       when we are in the partner_down state and one of the following
       two conditions holds:
       (a) in the case that the time sent from the peer is earlier than
           the time we entered the partner_down state, at least MCLT has
	   gone by since we entered partner_down, or
       (b) in the case that the time sent from the peer is later than
           the time when we entered partner_down, the current time is
	   later than the time sent from the peer by at least MCLT. */

int lease_mine_to_reallocate (struct lease *lease)
{
	dhcp_failover_state_t *peer;

	if (lease && lease->pool &&
	    (peer = lease->pool->failover_peer)) {
		/*
		 * In addition to the normal rules governing wether a server
		 * is allowed to operate changes on a lease, the server is
		 * allowed to operate on a lease from the standpoint of the
		 * most conservative guess of the peer's state for this lease.
		 */
		switch (lease->binding_state) {
		      case FTS_ACTIVE:
			/* ACTIVE leases may not be reallocated. */
			return 0;

		      case FTS_FREE:
		      case FTS_ABANDONED:
			/* FREE leases may only be allocated by the primary,
			 * unless the secondary is acting in partner_down
			 * state and stos+mclt or tsfp+mclt has expired,
			 * whichever is greater.
			 *
			 * ABANDONED are treated the same as FREE for all
			 * purposes here.  Note that servers will only try
			 * for ABANDONED leases as a last resort anyway.
			 */
			if (peer -> i_am == primary)
				return 1;

			return(peer->service_state == service_partner_down &&
			       ((lease->tsfp < peer->me.stos) ?
				(peer->me.stos + peer->mclt < cur_time) :
				(lease->tsfp + peer->mclt < cur_time)));

		      case FTS_RELEASED:
		      case FTS_EXPIRED:
			/*
			 * These leases are generally untouchable until the
			 * peer acknowledges their state change.  However, as
			 * this is impossible if the peer is offline, the
			 * failover protocol permits an 'optimization' to
			 * rewind the lease to a previous state that the server
			 * is allowed to operate on, if that was the state that
			 * was last acknowledged by the peer.
			 *
			 * So if a lease was free, was allocated by this
			 * server, and expired without ever being transmitted
			 * to the peer, it can be returned to free and given
			 * to any new client legally.
			 */
			if ((peer->i_am == primary) &&
			    (lease->rewind_binding_state == FTS_FREE))
				return 1;
			if ((peer->i_am == secondary) &&
			    (lease->rewind_binding_state == FTS_BACKUP))
				return 1;

			/* FALL THROUGH (released, expired, reset) */
		      case FTS_RESET:
			/*
			 * Released, expired, and reset leases go onto the
			 * 'expired' queue all together.  Upon entry into
			 * partner-down state, this queue of leases has their
			 * tsfp values modified to equal stos+mclt, the point
			 * at which the server is allowed to remove them from
			 * these transitional states.
			 *
			 * Note that although tsfp has been possibly extended
			 * past the actual tsfp we received from the peer, we
			 * don't have to take any special action.  Since tsfp
			 * will be equal to the current time when the lease
			 * transitions to free, tsfp will not be used to grant
			 * lease-times longer than the MCLT to clients, which
			 * is the only danger for this sort of modification.
			 */
			return((peer->service_state == service_partner_down) &&
			       (lease->tsfp < cur_time));

		      case FTS_BACKUP:
			/* Only the secondary may allocate BACKUP leases,
			 * unless in partner_down state in which case at
			 * least TSFP+MCLT or STOS+MCLT must have expired,
			 * whichever is greater.
			 */
			if (peer->i_am == secondary)
				return 1;

			return((peer->service_state == service_partner_down) &&
			       ((lease->tsfp < peer->me.stos) ?
				(peer->me.stos + peer->mclt < cur_time) :
				(lease->tsfp + peer->mclt < cur_time)));

		      default:
			/* All lease states appear above. */
			log_fatal("Impossible case at %s:%d.", MDL);
			break;
		}
		return 0;
	}
	if (lease)
		return(lease->binding_state == FTS_FREE ||
		       lease->binding_state == FTS_BACKUP);
	else
		return 0;
}

static isc_result_t failover_message_reference (failover_message_t **mp,
						failover_message_t *m,
						const char *file, int line)
{
	*mp = m;
	m -> refcnt++;
	return ISC_R_SUCCESS;
}

static isc_result_t failover_message_dereference (failover_message_t **mp,
						  const char *file, int line)
{
	failover_message_t *m;
	m = (*mp);
	m -> refcnt--;
	if (m -> refcnt == 0) {
		if (m -> next)
			failover_message_dereference (&m -> next,
						      file, line);
		if (m -> chaddr.data)
			dfree (m -> chaddr.data, file, line);
		if (m -> client_identifier.data)
			dfree (m -> client_identifier.data, file, line);
		if (m -> hba.data)
			dfree (m -> hba.data, file, line);
		if (m -> message.data)
			dfree (m -> message.data, file, line);
		if (m -> relationship_name.data)
			dfree (m -> relationship_name.data, file, line);
		if (m -> reply_options.data)
			dfree (m -> reply_options.data, file, line);
		if (m -> request_options.data)
			dfree (m -> request_options.data, file, line);
		if (m -> vendor_class.data)
			dfree (m -> vendor_class.data, file, line);
		if (m -> vendor_options.data)
			dfree (m -> vendor_options.data, file, line);
		if (m -> ddns.data)
			dfree (m -> ddns.data, file, line);
		dfree (*mp, file, line);
	}
	*mp = 0;
	return ISC_R_SUCCESS;
}

OMAPI_OBJECT_ALLOC (dhcp_failover_state, dhcp_failover_state_t,
		    dhcp_type_failover_state)
OMAPI_OBJECT_ALLOC (dhcp_failover_listener, dhcp_failover_listener_t,
		    dhcp_type_failover_listener)
OMAPI_OBJECT_ALLOC (dhcp_failover_link, dhcp_failover_link_t,
		    dhcp_type_failover_link)
#endif /* defined (FAILOVER_PROTOCOL) */

const char *binding_state_print (enum failover_state state)
{
	switch (state) {
	      case FTS_FREE:
		return "free";
		break;

	      case FTS_ACTIVE:
		return "active";
		break;

	      case FTS_EXPIRED:
		return "expired";
		break;

	      case FTS_RELEASED:
		return "released";
		break;

	      case FTS_ABANDONED:
		return "abandoned";
		break;

	      case FTS_RESET:
		return "reset";
		break;

	      case FTS_BACKUP:
		return "backup";
		break;

	      default:
		return "unknown";
		break;
	}
}


/*!
 * \brief Given a char pointer, return always return a printable value
 *
 * This function is intended to be used in within log statements, such that
 * its invocation only occurs if the logging level is enabled.
 *
 * \param value pointer the character to print
 *
 * \return If value is null, returns the string "<none>", if it contains
 * non-printable bytes, returns the string "<unsuitable for printing>",
 * otherwise it returns a const pointer to value
 */
static const char *printable(const char* value) {
	const char *print_value = "<none>";
	if (value) {
		if ((strlen (value) <= 64) &&
		     db_printable((unsigned char*)value)) {
			print_value = value;
		}
                else {
                        print_value = "<unsuitable for printing>";
		}
	}

	return (print_value);
}

/*!
 * \brief Remove information from a prior use of a lease
 *
 * Remove information from a lease that is not germain to lease affinity
 *
 * \param lease the lease to scrub
 */
void scrub_lease(struct lease* lease, const char *file, int line) {
	log_debug ("%s(%d):scrubbing lease for %s, hostname: %s", file, line,
		   piaddr(lease->ip_addr), printable(lease->client_hostname));

        if (lease->client_hostname) {
                dfree (lease->client_hostname, MDL);
                lease->client_hostname = (char *)0;
        }
}