File: tcp.c

package info (click to toggle)
kernel-source-2.0.32 2.0.32-5
  • links: PTS
  • area: main
  • in suites: hamm
  • size: 29,648 kB
  • ctags: 86,850
  • sloc: ansic: 542,141; asm: 26,201; makefile: 3,423; sh: 1,195; perl: 727; tcl: 408; cpp: 277; lisp: 211; awk: 134
file content (2438 lines) | stat: -rw-r--r-- 62,082 bytes parent folder | download | duplicates (2)
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
/*
 * INET		An implementation of the TCP/IP protocol suite for the LINUX
 *		operating system.  INET is implemented using the  BSD Socket
 *		interface as the means of communication with the user level.
 *
 *		Implementation of the Transmission Control Protocol(TCP).
 *
 * Version:	@(#)tcp.c	1.0.16	05/25/93
 *
 * Authors:	Ross Biro, <bir7@leland.Stanford.Edu>
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *		Mark Evans, <evansmp@uhura.aston.ac.uk>
 *		Corey Minyard <wf-rch!minyard@relay.EU.net>
 *		Florian La Roche, <flla@stud.uni-sb.de>
 *		Charles Hedrick, <hedrick@klinzhai.rutgers.edu>
 *		Linus Torvalds, <torvalds@cs.helsinki.fi>
 *		Alan Cox, <gw4pts@gw4pts.ampr.org>
 *		Matthew Dillon, <dillon@apollo.west.oic.com>
 *		Arnt Gulbrandsen, <agulbra@nvg.unit.no>
 *		Jorge Cwik, <jorge@laser.satlink.net>
 *
 * Fixes:
 *		Alan Cox	:	Numerous verify_area() calls
 *		Alan Cox	:	Set the ACK bit on a reset
 *		Alan Cox	:	Stopped it crashing if it closed while
 *					sk->inuse=1 and was trying to connect
 *					(tcp_err()).
 *		Alan Cox	:	All icmp error handling was broken
 *					pointers passed where wrong and the
 *					socket was looked up backwards. Nobody
 *					tested any icmp error code obviously.
 *		Alan Cox	:	tcp_err() now handled properly. It
 *					wakes people on errors. select
 *					behaves and the icmp error race
 *					has gone by moving it into sock.c
 *		Alan Cox	:	tcp_send_reset() fixed to work for
 *					everything not just packets for
 *					unknown sockets.
 *		Alan Cox	:	tcp option processing.
 *		Alan Cox	:	Reset tweaked (still not 100%) [Had
 *					syn rule wrong]
 *		Herp Rosmanith  :	More reset fixes
 *		Alan Cox	:	No longer acks invalid rst frames.
 *					Acking any kind of RST is right out.
 *		Alan Cox	:	Sets an ignore me flag on an rst
 *					receive otherwise odd bits of prattle
 *					escape still
 *		Alan Cox	:	Fixed another acking RST frame bug.
 *					Should stop LAN workplace lockups.
 *		Alan Cox	: 	Some tidyups using the new skb list
 *					facilities
 *		Alan Cox	:	sk->keepopen now seems to work
 *		Alan Cox	:	Pulls options out correctly on accepts
 *		Alan Cox	:	Fixed assorted sk->rqueue->next errors
 *		Alan Cox	:	PSH doesn't end a TCP read. Switched a
 *					bit to skb ops.
 *		Alan Cox	:	Tidied tcp_data to avoid a potential
 *					nasty.
 *		Alan Cox	:	Added some better commenting, as the
 *					tcp is hard to follow
 *		Alan Cox	:	Removed incorrect check for 20 * psh
 *	Michael O'Reilly	:	ack < copied bug fix.
 *	Johannes Stille		:	Misc tcp fixes (not all in yet).
 *		Alan Cox	:	FIN with no memory -> CRASH
 *		Alan Cox	:	Added socket option proto entries.
 *					Also added awareness of them to accept.
 *		Alan Cox	:	Added TCP options (SOL_TCP)
 *		Alan Cox	:	Switched wakeup calls to callbacks,
 *					so the kernel can layer network
 *					sockets.
 *		Alan Cox	:	Use ip_tos/ip_ttl settings.
 *		Alan Cox	:	Handle FIN (more) properly (we hope).
 *		Alan Cox	:	RST frames sent on unsynchronised
 *					state ack error.
 *		Alan Cox	:	Put in missing check for SYN bit.
 *		Alan Cox	:	Added tcp_select_window() aka NET2E
 *					window non shrink trick.
 *		Alan Cox	:	Added a couple of small NET2E timer
 *					fixes
 *		Charles Hedrick :	TCP fixes
 *		Toomas Tamm	:	TCP window fixes
 *		Alan Cox	:	Small URG fix to rlogin ^C ack fight
 *		Charles Hedrick	:	Rewrote most of it to actually work
 *		Linus		:	Rewrote tcp_read() and URG handling
 *					completely
 *		Gerhard Koerting:	Fixed some missing timer handling
 *		Matthew Dillon  :	Reworked TCP machine states as per RFC
 *		Gerhard Koerting:	PC/TCP workarounds
 *		Adam Caldwell	:	Assorted timer/timing errors
 *		Matthew Dillon	:	Fixed another RST bug
 *		Alan Cox	:	Move to kernel side addressing changes.
 *		Alan Cox	:	Beginning work on TCP fastpathing
 *					(not yet usable)
 *		Arnt Gulbrandsen:	Turbocharged tcp_check() routine.
 *		Alan Cox	:	TCP fast path debugging
 *		Alan Cox	:	Window clamping
 *		Michael Riepe	:	Bug in tcp_check()
 *		Matt Dillon	:	More TCP improvements and RST bug fixes
 *		Matt Dillon	:	Yet more small nasties remove from the
 *					TCP code (Be very nice to this man if
 *					tcp finally works 100%) 8)
 *		Alan Cox	:	BSD accept semantics.
 *		Alan Cox	:	Reset on closedown bug.
 *	Peter De Schrijver	:	ENOTCONN check missing in tcp_sendto().
 *		Michael Pall	:	Handle select() after URG properly in
 *					all cases.
 *		Michael Pall	:	Undo the last fix in tcp_read_urg()
 *					(multi URG PUSH broke rlogin).
 *		Michael Pall	:	Fix the multi URG PUSH problem in
 *					tcp_readable(), select() after URG
 *					works now.
 *		Michael Pall	:	recv(...,MSG_OOB) never blocks in the
 *					BSD api.
 *		Alan Cox	:	Changed the semantics of sk->socket to
 *					fix a race and a signal problem with
 *					accept() and async I/O.
 *		Alan Cox	:	Relaxed the rules on tcp_sendto().
 *		Yury Shevchuk	:	Really fixed accept() blocking problem.
 *		Craig I. Hagan  :	Allow for BSD compatible TIME_WAIT for
 *					clients/servers which listen in on
 *					fixed ports.
 *		Alan Cox	:	Cleaned the above up and shrank it to
 *					a sensible code size.
 *		Alan Cox	:	Self connect lockup fix.
 *		Alan Cox	:	No connect to multicast.
 *		Ross Biro	:	Close unaccepted children on master
 *					socket close.
 *		Alan Cox	:	Reset tracing code.
 *		Alan Cox	:	Spurious resets on shutdown.
 *		Alan Cox	:	Giant 15 minute/60 second timer error
 *		Alan Cox	:	Small whoops in selecting before an
 *					accept.
 *		Alan Cox	:	Kept the state trace facility since
 *					it's handy for debugging.
 *		Alan Cox	:	More reset handler fixes.
 *		Alan Cox	:	Started rewriting the code based on
 *					the RFC's for other useful protocol
 *					references see: Comer, KA9Q NOS, and
 *					for a reference on the difference
 *					between specifications and how BSD
 *					works see the 4.4lite source.
 *		A.N.Kuznetsov	:	Don't time wait on completion of tidy
 *					close.
 *		Linus Torvalds	:	Fin/Shutdown & copied_seq changes.
 *		Linus Torvalds	:	Fixed BSD port reuse to work first syn
 *		Alan Cox	:	Reimplemented timers as per the RFC
 *					and using multiple timers for sanity.
 *		Alan Cox	:	Small bug fixes, and a lot of new
 *					comments.
 *		Alan Cox	:	Fixed dual reader crash by locking
 *					the buffers (much like datagram.c)
 *		Alan Cox	:	Fixed stuck sockets in probe. A probe
 *					now gets fed up of retrying without
 *					(even a no space) answer.
 *		Alan Cox	:	Extracted closing code better
 *		Alan Cox	:	Fixed the closing state machine to
 *					resemble the RFC.
 *		Alan Cox	:	More 'per spec' fixes.
 *		Jorge Cwik	:	Even faster checksumming.
 *		Alan Cox	:	tcp_data() doesn't ack illegal PSH
 *					only frames. At least one pc tcp stack
 *					generates them.
 *		Alan Cox	:	Cache last socket.
 *		Alan Cox	:	Per route irtt.
 *		Matt Day	:	Select() match BSD precisely on error
 *		Alan Cox	:	New buffers
 *		Marc Tamsky	:	Various sk->prot->retransmits and
 *					sk->retransmits misupdating fixed.
 *					Fixed tcp_write_timeout: stuck close,
 *					and TCP syn retries gets used now.
 *		Mark Yarvis	:	In tcp_read_wakeup(), don't send an
 *					ack if stat is TCP_CLOSED.
 *		Alan Cox	:	Look up device on a retransmit - routes may
 *					change. Doesn't yet cope with MSS shrink right
 *					but it's a start!
 *		Marc Tamsky	:	Closing in closing fixes.
 *		Mike Shaver	:	RFC1122 verifications.
 *		Alan Cox	:	rcv_saddr errors.
 *		Alan Cox	:	Block double connect().
 *		Alan Cox	:	Small hooks for enSKIP.
 *		Alexey Kuznetsov:	Path MTU discovery.
 *		Alan Cox	:	Support soft errors.
 *		Alan Cox	:	Fix MTU discovery pathological case
 *					when the remote claims no mtu!
 *		Marc Tamsky	:	TCP_CLOSE fix.
 *		Colin (G3TNE)	:	Send a reset on syn ack replies in
 *					window but wrong (fixes NT lpd problems)
 *		Pedro Roque	:	Better TCP window handling, delayed ack.
 *		Joerg Reuter	:	No modification of locked buffers in
 *					tcp_do_retransmit()
 *		Eric Schenk	:	Changed receiver side silly window
 *					avoidance algorithm to BSD style
 *					algorithm. This doubles throughput
 *					against machines running Solaris,
 *					and seems to result in general
 *					improvement.
 *		Eric Schenk	:	Changed receiver side silly window
 *					avoidance algorithm to BSD style
 *					algorithm. This doubles throughput
 *					against machines running Solaris,
 *					and seems to result in general
 *					improvement.
 *	Stefan Magdalinski	:	adjusted tcp_readable() to fix FIONREAD
 *	Willy Konynenberg	:	Transparent proxying support.
 *		Theodore Ts'o	:	Do secure TCP sequence numbers.
 *		David S. Miller	:	New socket lookup architecture for ISS.
 *					This code is dedicated to John Dyson.
 *		Elliot Poger	:	Added support for SO_BINDTODEVICE.
 *					
 * To Fix:
 *		Fast path the code. Two things here - fix the window calculation
 *		so it doesn't iterate over the queue, also spot packets with no funny
 *		options arriving in order and process directly.
 *
 *		Rewrite output state machine to use a single queue.
 *		Speed up input assembly algorithm.
 *		RFC1323 - PAWS and window scaling. PAWS is required for IPv6 so we
 *		could do with it working on IPv4
 *		User settable/learned rtt/max window/mtu
 *
 *		Change the fundamental structure to a single send queue maintained
 *		by TCP (removing the bogus ip stuff [thus fixing mtu drops on
 *		active routes too]). Cut the queue off in tcp_retransmit/
 *		tcp_transmit.
 *		Change the receive queue to assemble as it goes. This lets us
 *		dispose of most of tcp_sequence, half of tcp_ack and chunks of
 *		tcp_data/tcp_read as well as the window shrink crud.
 *		Separate out duplicated code - tcp_alloc_skb, tcp_build_ack
 *		tcp_queue_skb seem obvious routines to extract.
 *
 *		This program is free software; you can redistribute it and/or
 *		modify it under the terms of the GNU General Public License
 *		as published by the Free Software Foundation; either version
 *		2 of the License, or(at your option) any later version.
 *
 * Description of States:
 *
 *	TCP_SYN_SENT		sent a connection request, waiting for ack
 *
 *	TCP_SYN_RECV		received a connection request, sent ack,
 *				waiting for final ack in three-way handshake.
 *
 *	TCP_ESTABLISHED		connection established
 *
 *	TCP_FIN_WAIT1		our side has shutdown, waiting to complete
 *				transmission of remaining buffered data
 *
 *	TCP_FIN_WAIT2		all buffered data sent, waiting for remote
 *				to shutdown
 *
 *	TCP_CLOSING		both sides have shutdown but we still have
 *				data we have to finish sending
 *
 *	TCP_TIME_WAIT		timeout to catch resent junk before entering
 *				closed, can only be entered from FIN_WAIT2
 *				or CLOSING.  Required because the other end
 *				may not have gotten our last ACK causing it
 *				to retransmit the data packet (which we ignore)
 *
 *	TCP_CLOSE_WAIT		remote side has shutdown and is waiting for
 *				us to finish writing our data and to shutdown
 *				(we have to close() to move on to LAST_ACK)
 *
 *	TCP_LAST_ACK		out side has shutdown after remote has
 *				shutdown.  There may still be data in our
 *				buffer that we have to finish sending
 *
 *	TCP_CLOSE		socket is finished
 */

/*
 * RFC1122 status:
 * NOTE: I'm not going to be doing comments in the code for this one except
 * for violations and the like.  tcp.c is just too big... If I say something
 * "does?" or "doesn't?", it means I'm not sure, and will have to hash it out
 * with Alan. -- MS 950903
 *
 * Use of PSH (4.2.2.2)
 *   MAY aggregate data sent without the PSH flag. (does)
 *   MAY queue data received without the PSH flag. (does)
 *   SHOULD collapse successive PSH flags when it packetizes data. (doesn't)
 *   MAY implement PSH on send calls. (doesn't, thus:)
 *     MUST NOT buffer data indefinitely (doesn't [1 second])
 *     MUST set PSH on last segment (does)
 *   MAY pass received PSH to application layer (doesn't)
 *   SHOULD send maximum-sized segment whenever possible. (almost always does)
 *
 * Window Size (4.2.2.3, 4.2.2.16)
 *   MUST treat window size as an unsigned number (does)
 *   SHOULD treat window size as a 32-bit number (does not)
 *   MUST NOT shrink window once it is offered (does not normally)
 *
 * Urgent Pointer (4.2.2.4)
 * **MUST point urgent pointer to last byte of urgent data (not right
 *     after). (doesn't, to be like BSD)
 *   MUST inform application layer asynchronously of incoming urgent
 *     data. (does)
 *   MUST provide application with means of determining the amount of
 *     urgent data pending. (does)
 * **MUST support urgent data sequence of arbitrary length. (doesn't, but
 *   it's sort of tricky to fix, as urg_ptr is a 16-bit quantity)
 *	[Follows BSD 1 byte of urgent data]
 *
 * TCP Options (4.2.2.5)
 *   MUST be able to receive TCP options in any segment. (does)
 *   MUST ignore unsupported options (does)
 *
 * Maximum Segment Size Option (4.2.2.6)
 *   MUST implement both sending and receiving MSS. (does)
 *   SHOULD send an MSS with every SYN where receive MSS != 536 (MAY send
 *     it always). (does, even when MSS == 536, which is legal)
 *   MUST assume MSS == 536 if no MSS received at connection setup (does)
 *   MUST calculate "effective send MSS" correctly:
 *     min(physical_MTU, remote_MSS+20) - sizeof(tcphdr) - sizeof(ipopts)
 *     (does - but allows operator override)
 *
 * TCP Checksum (4.2.2.7)
 *   MUST generate and check TCP checksum. (does)
 *
 * Initial Sequence Number Selection (4.2.2.8)
 *   MUST use the RFC 793 clock selection mechanism.  (doesn't, but it's
 *     OK: RFC 793 specifies a 250KHz clock, while we use 1MHz, which is
 *     necessary for 10Mbps networks - and harder than BSD to spoof!)
 *
 * Simultaneous Open Attempts (4.2.2.10)
 *   MUST support simultaneous open attempts (does)
 *
 * Recovery from Old Duplicate SYN (4.2.2.11)
 *   MUST keep track of active vs. passive open (does)
 *
 * RST segment (4.2.2.12)
 *   SHOULD allow an RST segment to contain data (does, but doesn't do
 *     anything with it, which is standard)
 *
 * Closing a Connection (4.2.2.13)
 *   MUST inform application of whether connection was closed by RST or
 *     normal close. (does)
 *   MAY allow "half-duplex" close (treat connection as closed for the
 *     local app, even before handshake is done). (does)
 *   MUST linger in TIME_WAIT for 2 * MSL (does)
 *
 * Retransmission Timeout (4.2.2.15)
 *   MUST implement Jacobson's slow start and congestion avoidance
 *     stuff. (does)
 *
 * Probing Zero Windows (4.2.2.17)
 *   MUST support probing of zero windows. (does)
 *   MAY keep offered window closed indefinitely. (does)
 *   MUST allow remote window to stay closed indefinitely. (does)
 *
 * Passive Open Calls (4.2.2.18)
 *   MUST NOT let new passive open affect other connections. (doesn't)
 *   MUST support passive opens (LISTENs) concurrently. (does)
 *
 * Time to Live (4.2.2.19)
 *   MUST make TCP TTL configurable. (does - IP_TTL option)
 *
 * Event Processing (4.2.2.20)
 *   SHOULD queue out-of-order segments. (does)
 *   MUST aggregate ACK segments whenever possible. (does but badly)
 *
 * Retransmission Timeout Calculation (4.2.3.1)
 *   MUST implement Karn's algorithm and Jacobson's algorithm for RTO
 *     calculation. (does, or at least explains them in the comments 8*b)
 *  SHOULD initialize RTO to 0 and RTT to 3. (does)
 *
 * When to Send an ACK Segment (4.2.3.2)
 *   SHOULD implement delayed ACK. (does)
 *   MUST keep ACK delay < 0.5 sec. (does)
 *
 * When to Send a Window Update (4.2.3.3)
 *   MUST implement receiver-side SWS. (does)
 *
 * When to Send Data (4.2.3.4)
 *   MUST implement sender-side SWS. (does)
 *   SHOULD implement Nagle algorithm. (does)
 *
 * TCP Connection Failures (4.2.3.5)
 *  MUST handle excessive retransmissions "properly" (see the RFC). (does)
 *   SHOULD inform application layer of soft errors. (does)
 *
 * TCP Keep-Alives (4.2.3.6)
 *   MAY provide keep-alives. (does)
 *   MUST make keep-alives configurable on a per-connection basis. (does)
 *   MUST default to no keep-alives. (does)
 * **MUST make keep-alive interval configurable. (doesn't)
 * **MUST make default keep-alive interval > 2 hours. (doesn't)
 *   MUST NOT interpret failure to ACK keep-alive packet as dead
 *     connection. (doesn't)
 *   SHOULD send keep-alive with no data. (does)
 *
 * TCP Multihoming (4.2.3.7)
 *   MUST get source address from IP layer before sending first
 *     SYN. (does)
 *   MUST use same local address for all segments of a connection. (does)
 *
 * IP Options (4.2.3.8)
 *   MUST ignore unsupported IP options. (does)
 *   MAY support Time Stamp and Record Route. (does)
 *   MUST allow application to specify a source route. (does)
 *   MUST allow received Source Route option to set route for all future
 *     segments on this connection. (does not (security issues))
 *
 * ICMP messages (4.2.3.9)
 *   MUST act on ICMP errors. (does)
 *   MUST slow transmission upon receipt of a Source Quench. (does)
 *   MUST NOT abort connection upon receipt of soft Destination
 *     Unreachables (0, 1, 5), Time Exceededs and Parameter
 *     Problems. (doesn't)
 *   SHOULD report soft Destination Unreachables etc. to the
 *     application. (does)
 *   SHOULD abort connection upon receipt of hard Destination Unreachable
 *     messages (2, 3, 4). (does)
 *
 * Remote Address Validation (4.2.3.10)
 *   MUST reject as an error OPEN for invalid remote IP address. (does)
 *   MUST ignore SYN with invalid source address. (does)
 *   MUST silently discard incoming SYN for broadcast/multicast
 *     address. (does)
 *
 * Asynchronous Reports (4.2.4.1)
 * MUST provide mechanism for reporting soft errors to application
 *     layer. (does)
 *
 * Type of Service (4.2.4.2)
 *   MUST allow application layer to set Type of Service. (does IP_TOS)
 *
 * (Whew. -- MS 950903)
 **/

#include <linux/config.h>
#include <linux/types.h>
#include <linux/fcntl.h>
#include <linux/random.h>

#include <net/icmp.h>
#include <net/tcp.h>

#include <asm/segment.h>

unsigned long seq_offset;
struct tcp_mib	tcp_statistics;

/* This is for sockets with full identity only.  Sockets here will always
 * be without wildcards and will have the following invariant:
 *          TCP_ESTABLISHED <= sk->state < TCP_CLOSE
 */
struct sock *tcp_established_hash[TCP_HTABLE_SIZE];

/* All sockets in TCP_LISTEN state will be in here.  This is the only table
 * where wildcard'd TCP sockets can exist.  Hash function here is just local
 * port number.  XXX Fix or we'll lose with thousands of IP aliases...
 */
struct sock *tcp_listening_hash[TCP_LHTABLE_SIZE];

/* Ok, let's try this, I give up, we do need a local binding
 * TCP hash as well as the others for fast bind/connect.
 */
struct sock *tcp_bound_hash[TCP_BHTABLE_SIZE];

extern struct sock *tcp_v4_lookup(u32 saddr, u16 sport, u32 daddr, u16 dport);

static int tcp_v4_verify_bind(struct sock *sk, unsigned short snum)
{
	struct sock *sk2;
	int retval = 0, sk_reuse = sk->reuse;

	SOCKHASH_LOCK();
	sk2 = tcp_bound_hash[tcp_bhashfn(snum)];
	for(; sk2 != NULL; sk2 = sk2->bind_next) {
		if((sk2->num == snum) && (sk2 != sk)) {
			unsigned char state = sk2->state;
			int sk2_reuse = sk2->reuse;

			/* Two sockets can be bound to the same port if they're
			 * bound to different interfaces... */
			if (sk->bound_device != sk2->bound_device)
				continue;

			if(!sk2->rcv_saddr || !sk->rcv_saddr) {
				if((!sk2_reuse)			||
				   (!sk_reuse)			||
				   (state == TCP_LISTEN)) {
					retval = 1;
					break;
				}
			} else if(sk2->rcv_saddr == sk->rcv_saddr) {
				if((!sk_reuse)			||
				   (!sk2_reuse)			||
				   (state == TCP_LISTEN)) {
					retval = 1;
					break;
				}
			}
		}
	}
	SOCKHASH_UNLOCK();

	return retval;
}

static __inline__ int tcp_lport_inuse(int num)
{
	struct sock *sk = tcp_bound_hash[tcp_bhashfn(num)];

	for(; sk != NULL; sk = sk->bind_next) {
		if(sk->num == num)
			return 1;
	}
	return 0;
}

/* Find a "good" local port, this is family independant.
 * There are several strategies working in unison here to
 * get the best possible performance.  The current socket
 * load is kept track of, if it is zero there is a strong
 * likely hood that there is a zero length chain we will
 * find with a small amount of searching, else the load is
 * what we shoot for for when the chains all have at least
 * one entry.  The base helps us walk the chains in an
 * order such that a good chain is found as quickly as possible.  -DaveM
 */
unsigned short tcp_good_socknum(void)
{
	static int start = PROT_SOCK;
	static int binding_contour = 0;
	int best = 0;
	int size = 32767; /* a big num. */
	int retval = 0, i, end, bc;

	SOCKHASH_LOCK();
	i = tcp_bhashfn(start);
	end = i + TCP_BHTABLE_SIZE;
	bc = binding_contour;
	do {
		struct sock *sk = tcp_bound_hash[i&(TCP_BHTABLE_SIZE-1)];
		if(!sk) {
			/* find the smallest value no smaller than start
			 * that has this hash value.
			 */
			retval = tcp_bhashnext(start-1,i&(TCP_BHTABLE_SIZE-1));

			/* Check for decreasing load. */
			if (bc != 0)
				binding_contour = 0;
			goto done;
		} else {
			int j = 0;
			do { sk = sk->bind_next; } while (++j < size && sk);
			if (j < size) {
				best = i&(TCP_BHTABLE_SIZE-1);
				size = j;
				if (bc && size <= bc) {
					i = best;
					goto verify;
				}
			}
		}
	} while(++i != end);
	i = best;

	/* Socket load is increasing, adjust our load average. */
	binding_contour = size;
verify:
	if (size < binding_contour)
		binding_contour = size;

	retval = tcp_bhashnext(start-1,i);

	best = retval;	/* mark the starting point to avoid infinite loops */
	while(tcp_lport_inuse(retval)) {
		retval = tcp_bhashnext(retval,i);
		if (retval > 32767)	/* Upper bound */
			retval = tcp_bhashnext(PROT_SOCK,i);
		if (retval == best) {
			/* This hash chain is full. No answer. */
			retval = 0;
			break;
		}
	}

done:
	start = (retval + 1);
	if (start > 32767 || start < PROT_SOCK)
		start = PROT_SOCK;
	SOCKHASH_UNLOCK();

	return retval;
}

void tcp_v4_hash(struct sock *sk)
{
	unsigned char state;

	SOCKHASH_LOCK();
	state = sk->state;
	if(state != TCP_CLOSE || !sk->dead) {
		struct sock **skp;

		if(state == TCP_LISTEN)
			skp = &tcp_listening_hash[tcp_sk_listen_hashfn(sk)];
		else
			skp = &tcp_established_hash[tcp_sk_hashfn(sk)];

		if((sk->next = *skp) != NULL)
			(*skp)->pprev = &sk->next;
		*skp = sk;
		sk->pprev = skp;
		tcp_sk_bindify(sk);
	}
	SOCKHASH_UNLOCK();
}

void tcp_v4_unhash(struct sock *sk)
{
	SOCKHASH_LOCK();
	if(sk->pprev) {
		if(sk->next)
			sk->next->pprev = sk->pprev;
		*sk->pprev = sk->next;
		sk->pprev = NULL;
		tcp_sk_unbindify(sk);
	}
	SOCKHASH_UNLOCK();
}

void tcp_v4_rehash(struct sock *sk)
{
	unsigned char state;

	SOCKHASH_LOCK();
	state = sk->state;
	if(sk->pprev) {
		if(sk->next)
			sk->next->pprev = sk->pprev;
		*sk->pprev = sk->next;
		sk->pprev = NULL;
		tcp_sk_unbindify(sk);
	}
	if(state != TCP_CLOSE || !sk->dead) {
		struct sock **skp;

		if(state == TCP_LISTEN)
			skp = &tcp_listening_hash[tcp_sk_listen_hashfn(sk)];
		else
			skp = &tcp_established_hash[tcp_sk_hashfn(sk)];

		if((sk->next = *skp) != NULL)
			(*skp)->pprev = &sk->next;
		*skp = sk;
		sk->pprev = skp;
		tcp_sk_bindify(sk);
	}
	SOCKHASH_UNLOCK();
}

static void tcp_close(struct sock *sk, unsigned long timeout);

/*
 *	Find someone to 'accept'. Must be called with
 *	the socket locked or with interrupts disabled
 */

static struct sk_buff *tcp_find_established(struct sock *s)
{
	struct sk_buff *p=skb_peek(&s->receive_queue);
	if(p==NULL)
		return NULL;
	do
	{
		if(p->sk->state == TCP_ESTABLISHED || p->sk->state >= TCP_FIN_WAIT1)
			return p;
		p=p->next;
	}
	while(p!=(struct sk_buff *)&s->receive_queue);
	return NULL;
}

/*
 *	This routine closes sockets which have been at least partially
 *	opened, but not yet accepted. Currently it is only called by
 *	tcp_close, and timeout mirrors the value there.
 */

static void tcp_close_pending (struct sock *sk)
{
	struct sk_buff *skb;

	while ((skb = skb_dequeue(&sk->receive_queue)) != NULL)
	{
		tcp_close(skb->sk, 0);
		kfree_skb(skb, FREE_READ);
	}
	return;
}

/*
 *	Enter the time wait state.
 */

void tcp_time_wait(struct sock *sk)
{
	tcp_set_state(sk,TCP_TIME_WAIT);
	sk->shutdown = SHUTDOWN_MASK;
	if (!sk->dead)
		sk->state_change(sk);
	tcp_reset_msl_timer(sk, TIME_CLOSE, TCP_TIMEWAIT_LEN);
}


/*
 * This routine is called by the ICMP module when it gets some
 * sort of error condition.  If err < 0 then the socket should
 * be closed and the error returned to the user.  If err > 0
 * it's just the icmp type << 8 | icmp code.  After adjustment
 * header points to the first 8 bytes of the tcp header.  We need
 * to find the appropriate port.
 */

void tcp_err(int type, int code, unsigned char *header, __u32 daddr,
	__u32 saddr, struct inet_protocol *protocol, int len)
{
	struct tcphdr *th = (struct tcphdr *)header;
	struct sock *sk;

	/*
	 *	This one is _WRONG_. FIXME urgently.
	 */
#ifndef CONFIG_NO_PATH_MTU_DISCOVERY
	struct iphdr *iph=(struct iphdr *)(header-sizeof(struct iphdr));
#endif
	th =(struct tcphdr *)header;
	
	if (len < 8)	/* NOT sizeof(struct tcphdr) */
		return;
		
	sk = tcp_v4_lookup(daddr, th->dest, saddr, th->source);
	if (sk == NULL)
		return;

	if (type == ICMP_SOURCE_QUENCH)
	{
		/*
		 * FIXME:
		 * Follow BSD for now and just reduce cong_window to 1 again.
		 * It is possible that we just want to reduce the
		 * window by 1/2, or that we want to reduce ssthresh by 1/2
		 * here as well.
		 */
		sk->cong_window = 1;
		sk->cong_count = 0;
		sk->high_seq = sk->sent_seq;
		return;
	}

	if (type == ICMP_PARAMETERPROB)
	{
		sk->err=EPROTO;
		sk->error_report(sk);
	}

#ifndef CONFIG_NO_PATH_MTU_DISCOVERY
	if (type == ICMP_DEST_UNREACH && code == ICMP_FRAG_NEEDED)
	{
		struct rtable * rt;
		/*
		 * Ugly trick to pass MTU to protocol layer.
		 * Really we should add argument "info" to error handler.
		 */
		unsigned short new_mtu = ntohs(iph->id);

		if ((rt = sk->ip_route_cache) != NULL)
			if (rt->rt_mtu > new_mtu)
				rt->rt_mtu = new_mtu;

		/*
		 *	FIXME::
		 *	Not the nicest of fixes: Lose a MTU update if the socket is
		 *	locked this instant. Not the right answer but will be best
		 *	for the production fix. Make 2.1 work right!
		 */
		 
		if (sk->mtu > new_mtu - sizeof(struct iphdr) - sizeof(struct tcphdr)
			&& new_mtu > sizeof(struct iphdr)+sizeof(struct tcphdr) && !sk->users)
			sk->mtu = new_mtu - sizeof(struct iphdr) - sizeof(struct tcphdr);

		return;
	}
#endif

	/*
	 * If we've already connected we will keep trying
	 * until we time out, or the user gives up.
	 */

	if(code<=NR_ICMP_UNREACH)
	{
		if(icmp_err_convert[code].fatal || sk->state == TCP_SYN_SENT || sk->state == TCP_SYN_RECV)
		{
			sk->err = icmp_err_convert[code].errno;
			if (sk->state == TCP_SYN_SENT || sk->state == TCP_SYN_RECV)
			{
				tcp_statistics.TcpAttemptFails++;
				tcp_set_state(sk,TCP_CLOSE);
				sk->error_report(sk);		/* Wake people up to see the error (see connect in sock.c) */
			}
		}
		else	/* Only an error on timeout */
			sk->err_soft = icmp_err_convert[code].errno;
	}
}


/*
 *	Walk down the receive queue counting readable data until we hit the end or we find a gap
 *	in the received data queue (ie a frame missing that needs sending to us). Not
 *	sorting using two queues as data arrives makes life so much harder.
 */

static int tcp_readable(struct sock *sk)
{
	unsigned long counted;
	unsigned long amount;
	struct sk_buff *skb;
	int sum;
	unsigned long flags;

	if(sk && sk->debug)
	  	printk("tcp_readable: %p - ",sk);

	save_flags(flags);
	cli();
	if (sk == NULL || (skb = skb_peek(&sk->receive_queue)) == NULL)
	{
		restore_flags(flags);
	  	if(sk && sk->debug)
	  		printk("empty\n");
	  	return(0);
	}

	counted = sk->copied_seq;	/* Where we are at the moment */
	amount = 0;

	/*
	 *	Do until a push or until we are out of data.
	 */

	do
	{
		if (before(counted, skb->seq))	 	/* Found a hole so stops here */
			break;
		sum = skb->len - (counted - skb->seq);	/* Length - header but start from where we are up to (avoid overlaps) */
		if (skb->h.th->syn)
			sum++;
		if (sum > 0)
		{					/* Add it up, move on */
			amount += sum;
			if (skb->h.th->syn)
				amount--;
			counted += sum;
		}
		/*
		 * Don't count urg data ... but do it in the right place!
		 * Consider: "old_data (ptr is here) URG PUSH data"
		 * The old code would stop at the first push because
		 * it counted the urg (amount==1) and then does amount--
		 * *after* the loop.  This means tcp_readable() always
		 * returned zero if any URG PUSH was in the queue, even
		 * though there was normal data available. If we subtract
		 * the urg data right here, we even get it to work for more
		 * than one URG PUSH skb without normal data.
		 * This means that select() finally works now with urg data
		 * in the queue.  Note that rlogin was never affected
		 * because it doesn't use select(); it uses two processes
		 * and a blocking read().  And the queue scan in tcp_read()
		 * was correct.  Mike <pall@rz.uni-karlsruhe.de>
		 */
		if (skb->h.th->urg)
			amount--;	/* don't count urg data */
/*		if (amount && skb->h.th->psh) break;*/
		skb = skb->next;
	}
	while(skb != (struct sk_buff *)&sk->receive_queue);

	restore_flags(flags);
	if(sk->debug)
	  	printk("got %lu bytes.\n",amount);
	return(amount);
}

/*
 * LISTEN is a special case for select..
 */
static int tcp_listen_select(struct sock *sk, int sel_type, select_table *wait)
{
	if (sel_type == SEL_IN) {
		struct sk_buff * skb;

		lock_sock(sk);
		skb = tcp_find_established(sk);
		release_sock(sk);
		if (skb)
			return 1;
		select_wait(sk->sleep,wait);
		return 0;
	}
	return 0;
}


/*
 *	Wait for a TCP event.
 *
 *	Note that we don't need to lock the socket, as the upper select layers
 *	take care of normal races (between the test and the event) and we don't
 *	go look at any of the socket buffers directly.
 */
static int tcp_select(struct sock *sk, int sel_type, select_table *wait)
{
	if (sk->state == TCP_LISTEN)
		return tcp_listen_select(sk, sel_type, wait);

	switch(sel_type) {
	case SEL_IN:
		if (sk->err)
			return 1;
		if (sk->state == TCP_SYN_SENT || sk->state == TCP_SYN_RECV)
			break;

		if (sk->shutdown & RCV_SHUTDOWN)
			return 1;

		if (sk->acked_seq == sk->copied_seq)
			break;

		if (sk->urg_seq != sk->copied_seq ||
		    sk->acked_seq != sk->copied_seq+1 ||
		    sk->urginline || !sk->urg_data)
			return 1;
		break;

	case SEL_OUT:
		if (sk->err)
			return 1;
		if (sk->shutdown & SEND_SHUTDOWN)
			return 0;
		if (sk->state == TCP_SYN_SENT || sk->state == TCP_SYN_RECV)
			break;
		if (sk->wmem_alloc*2 > sk->sndbuf)
			break;
		return 1;

	case SEL_EX:
		if (sk->urg_data)
			return 1;
		break;
	}
	select_wait(sk->sleep, wait);
	return 0;
}

int tcp_ioctl(struct sock *sk, int cmd, unsigned long arg)
{
	int err;
	switch(cmd)
	{

		case TIOCINQ:
#ifdef FIXME	/* FIXME: */
		case FIONREAD:
#endif
		{
			unsigned long amount;

			if (sk->state == TCP_LISTEN)
				return(-EINVAL);

			lock_sock(sk);
			amount = tcp_readable(sk);
			release_sock(sk);
			err=verify_area(VERIFY_WRITE,(void *)arg, sizeof(int));
			if(err)
				return err;
			put_user(amount, (int *)arg);
			return(0);
		}
		case SIOCATMARK:
		{
			int answ = sk->urg_data && sk->urg_seq == sk->copied_seq;

			err = verify_area(VERIFY_WRITE,(void *) arg, sizeof(int));
			if (err)
				return err;
			put_user(answ,(int *) arg);
			return(0);
		}
		case TIOCOUTQ:
		{
			unsigned long amount;

			if (sk->state == TCP_LISTEN) return(-EINVAL);
			amount = sock_wspace(sk);
			err=verify_area(VERIFY_WRITE,(void *)arg, sizeof(int));
			if(err)
				return err;
			put_user(amount, (int *)arg);
			return(0);
		}
		default:
			return(-EINVAL);
	}
}


/*
 *	This routine computes a TCP checksum.
 *
 *	Modified January 1995 from a go-faster DOS routine by
 *	Jorge Cwik <jorge@laser.satlink.net>
 */
#undef DEBUG_TCP_CHECK
void tcp_send_check(struct tcphdr *th, unsigned long saddr,
		unsigned long daddr, int len, struct sk_buff *skb)
{
#ifdef DEBUG_TCP_CHECK
	u16 check;
#endif
	th->check = 0;
	th->check = tcp_check(th, len, saddr, daddr,
		csum_partial((char *)th,sizeof(*th),skb->csum));

#ifdef DEBUG_TCP_CHECK
	check = th->check;
	th->check = 0;
	th->check = tcp_check(th, len, saddr, daddr,
		csum_partial((char *)th,len,0));
	if (check != th->check) {
		static int count = 0;
		if (++count < 10) {
			printk("Checksum %x (%x) from %p\n", th->check, check,
				(&th)[-1]);
			printk("TCP=<off:%d a:%d s:%d f:%d>\n", th->doff*4, th->ack, th->syn, th->fin);
		}
	}
#endif
}


/*
 *	This routine builds a generic TCP header.
 */

static inline int tcp_build_header(struct tcphdr *th, struct sock *sk, int push)
{
	memcpy(th,(void *) &(sk->dummy_th), sizeof(*th));
	th->psh = (push == 0) ? 1 : 0;
	th->seq = htonl(sk->write_seq);
	th->ack_seq = htonl(sk->acked_seq);
	th->window = htons(tcp_select_window(sk));

	return(sizeof(*th));
}

/*
 *	Wait for a socket to get into the connected state
 */
static void wait_for_tcp_connect(struct sock * sk)
{
	release_sock(sk);
	cli();
	if (sk->state != TCP_ESTABLISHED && sk->state != TCP_CLOSE_WAIT && sk->err == 0)
    	{
		interruptible_sleep_on(sk->sleep);
	}
	sti();
	lock_sock(sk);
}

static inline int tcp_memory_free(struct sock *sk)
{
	return sk->wmem_alloc < sk->sndbuf;
}

/*
 *	Wait for more memory for a socket
 */
static void wait_for_tcp_memory(struct sock * sk)
{
	release_sock(sk);
	if (!tcp_memory_free(sk)) {
		struct wait_queue wait = { current, NULL };

		sk->socket->flags &= ~SO_NOSPACE;
		add_wait_queue(sk->sleep, &wait);
		for (;;) {
			if (current->signal & ~current->blocked)
				break;
			current->state = TASK_INTERRUPTIBLE;
			if (tcp_memory_free(sk))
				break;
			if (sk->shutdown & SEND_SHUTDOWN)
				break;
			if (sk->err)
				break;
			schedule();
		}
		current->state = TASK_RUNNING;
		remove_wait_queue(sk->sleep, &wait);
	}
	lock_sock(sk);
}


/*
 *	This routine copies from a user buffer into a socket,
 *	and starts the transmit system.
 */

static int do_tcp_sendmsg(struct sock *sk,
	int iovlen, struct iovec *iov,
	int len, int nonblock, int flags)
{
	int copied = 0;
	struct device *dev = NULL;

	/*
	 *	Wait for a connection to finish.
	 */
	while (sk->state != TCP_ESTABLISHED && sk->state != TCP_CLOSE_WAIT)
	{
		if (sk->err)
			return sock_error(sk);

		if (sk->state != TCP_SYN_SENT && sk->state != TCP_SYN_RECV)
		{
			if (sk->keepopen)
				send_sig(SIGPIPE, current, 0);
			return -EPIPE;
		}

		if (nonblock)
			return -EAGAIN;

		if (current->signal & ~current->blocked)
			return -ERESTARTSYS;

		wait_for_tcp_connect(sk);
	}

	/*
	 *	Ok commence sending
	 */

	while (--iovlen >= 0)
	{
		int seglen=iov->iov_len;
		unsigned char * from=iov->iov_base;
		iov++;

		while(seglen > 0)
		{
			int copy, delay;
			int tmp;
			struct sk_buff *skb;

			/*
			 * Stop on errors
			 */
			if (sk->err)
			{
				if (copied)
					return copied;
				return sock_error(sk);
			}

			/*
			 *	Make sure that we are established.
			 */
			if (sk->shutdown & SEND_SHUTDOWN)
			{
				if (copied)
					return copied;
				send_sig(SIGPIPE,current,0);
				return -EPIPE;
			}

			/*
			 * The following code can result in copy <= if sk->mss is ever
			 * decreased.  It shouldn't be.  sk->mss is min(sk->mtu, sk->max_window).
			 * sk->mtu is constant once SYN processing is finished.  I.e. we
			 * had better not get here until we've seen his SYN and at least one
			 * valid ack.  (The SYN sets sk->mtu and the ack sets sk->max_window.)
			 * But ESTABLISHED should guarantee that.  sk->max_window is by definition
			 * non-decreasing.  Note that any ioctl to set user_mss must be done
			 * before the exchange of SYN's.  If the initial ack from the other
			 * end has a window of 0, max_window and thus mss will both be 0.
			 */

			/*
			 *	Now we need to check if we have a half built packet.
			 */
#ifndef CONFIG_NO_PATH_MTU_DISCOVERY
			/*
			 *	Really, we should rebuild all the queues...
			 *	It's difficult. Temporary hack is to send all
			 *	queued segments with allowed fragmentation.
			 */
			{
				/*
				 *	new_mss may be zero. That indicates
				 *	we don't have a window estimate for
				 *	the remote box yet. 
				 *		-- AC
				 */
				
				int new_mss = min(sk->mtu, sk->max_window);
				if (new_mss && new_mss < sk->mss)
				{
					tcp_send_partial(sk);
					sk->mss = new_mss;
				}
			}
#endif

			/*
			 *	If there is a partly filled frame we can fill
			 *	out.
			 */
			if ((skb = tcp_dequeue_partial(sk)) != NULL)
			{
				int tcp_size;

				tcp_size = skb->tail - (unsigned char *)(skb->h.th + 1);

				/* Add more stuff to the end of skb->len */
				if (!(flags & MSG_OOB))
				{
					copy = min(sk->mss - tcp_size, seglen);
					
					/*
					 *	Now we may find the frame is as big, or too
					 *	big for our MSS. Thats all fine. It means the
					 *	MSS shrank (from an ICMP) after we allocated 
					 *	this frame.
					 */
					 
					if (copy <= 0)
					{
						/*
						 *	Send the now forced complete frame out. 
						 *
						 *	Note for 2.1: The MSS reduce code ought to
						 *	flush any frames in partial that are now
						 *	full sized. Not serious, potential tiny
						 *	performance hit.
						 */
						tcp_send_skb(sk,skb);
						/*
						 *	Get a new buffer and try again.
						 */
						continue;
					}
					/*
					 *	Otherwise continue to fill the buffer.
					 */
					tcp_size += copy;
					memcpy_fromfs(skb_put(skb,copy), from, copy);
					skb->csum = csum_partial(skb->tail - tcp_size, tcp_size, 0);
					from += copy;
					copied += copy;
					len -= copy;
					sk->write_seq += copy;
					seglen -= copy;
				}
				/* If we have a full packet or a new OOB
				 * message, we have to force this packet out.
				 */
				if (tcp_size >= sk->mss || (flags & MSG_OOB))
					tcp_send_skb(sk, skb);
				else
					tcp_enqueue_partial(skb, sk);
				continue;
			}

		/*
		 * We also need to worry about the window.
	 	 * If window < 1/2 the maximum window we've seen from this
	 	 *   host, don't use it.  This is sender side
	 	 *   silly window prevention, as specified in RFC1122.
	 	 *   (Note that this is different than earlier versions of
		 *   SWS prevention, e.g. RFC813.).  What we actually do is
		 *   use the whole MSS.  Since the results in the right
		 *   edge of the packet being outside the window, it will
		 *   be queued for later rather than sent.
		 */

			copy = sk->window_seq - sk->write_seq;
			if (copy <= 0 || copy < (sk->max_window >> 1) || copy > sk->mss)
				copy = sk->mss;
			if (copy > seglen)
				copy = seglen;
			if (copy <= 0)
			{
				printk(KERN_CRIT "TCP: **bug**: copy=%d, sk->mss=%d\n", copy, sk->mss);
		  		return -EFAULT;
			}

			/*
			 *	We should really check the window here also.
			 */

			delay = 0;
			tmp = copy + sk->prot->max_header + 15;
			/* If won't fill the current packet, and it's not an OOB message,
			 * then we might want to delay to allow data in the later parts
			 * of iov to fill this packet out. Note that if we aren't
			 * Nagling or there are no packets currently out then the top
			 * level code in tcp_sendmsg() will force any partial packets out
			 * after we finish building the largest packets this write allows.
			 */
			if (copy < sk->mss && !(flags & MSG_OOB)) {
				tmp = tmp - copy + sk->mtu + 128;
				delay = 1;
			}
			skb = sock_wmalloc(sk, tmp, 0, GFP_KERNEL);

			/*
			 *	If we didn't get any memory, we need to sleep.
			 */

			if (skb == NULL)
			{
				sk->socket->flags |= SO_NOSPACE;
				if (nonblock)
				{
					if (copied)
						return copied;
					return -EAGAIN;
				}

				if (current->signal & ~current->blocked)
				{
					if (copied)
						return copied;
					return -ERESTARTSYS;
				}

				wait_for_tcp_memory(sk);
				continue;
			}

			skb->sk = sk;
			skb->free = 0;
			skb->localroute = sk->localroute|(flags&MSG_DONTROUTE);

			/*
			 * FIXME: we need to optimize this.
			 * Perhaps some hints here would be good.
			 */

			tmp = sk->prot->build_header(skb, sk->saddr, sk->daddr, &dev,
				 IPPROTO_TCP, sk->opt, skb->truesize,sk->ip_tos,sk->ip_ttl,&sk->ip_route_cache);
			if (tmp < 0 )
			{
				sock_wfree(sk, skb);
				if (copied)
					return(copied);
				return(tmp);
			}
#ifndef CONFIG_NO_PATH_MTU_DISCOVERY
			skb->ip_hdr->frag_off |= htons(IP_DF);
#endif
			skb->dev = dev;
			skb->h.th =(struct tcphdr *)skb_put(skb,sizeof(struct tcphdr));
			tmp = tcp_build_header(skb->h.th, sk, seglen-copy);
			if (tmp < 0)
			{
				sock_wfree(sk, skb);
				if (copied)
					return(copied);
				return(tmp);
			}

			if (flags & MSG_OOB)
			{
				skb->h.th->urg = 1;
				skb->h.th->urg_ptr = ntohs(copy);
			}

			skb->csum = csum_partial_copy_fromuser(from,
				skb_put(skb,copy), copy, 0);

			from += copy;
			copied += copy;
			len -= copy;
			seglen -= copy;
			skb->free = 0;
			sk->write_seq += copy;

			if (delay)
			{
				tcp_enqueue_partial(skb, sk);
				continue;
			}
			tcp_send_skb(sk, skb);
		}
	}
	sk->err = 0;

	return copied;
}


static int tcp_sendmsg(struct sock *sk, struct msghdr *msg,
	  int len, int nonblock, int flags)
{
	int retval = -EINVAL;

	/*
	 *	Do sanity checking for sendmsg/sendto/send
	 */

	if (flags & ~(MSG_OOB|MSG_DONTROUTE))
		goto out;
	if (msg->msg_name) {
		struct sockaddr_in *addr=(struct sockaddr_in *)msg->msg_name;

		if (msg->msg_namelen < sizeof(*addr))
			goto out;
		if (addr->sin_family && addr->sin_family != AF_INET)
			goto out;
		retval = -ENOTCONN;
		if(sk->state == TCP_CLOSE)
			goto out;
		retval = -EISCONN;
		if (addr->sin_port != sk->dummy_th.dest)
			goto out;
		if (addr->sin_addr.s_addr != sk->daddr)
			goto out;
	}

	lock_sock(sk);
	retval = do_tcp_sendmsg(sk, msg->msg_iovlen, msg->msg_iov, len, nonblock, flags);

/*
 *	Nagle's rule. Turn Nagle off with TCP_NODELAY for highly
 *	interactive fast network servers. It's meant to be on and
 *	it really improves the throughput though not the echo time
 *	on my slow slip link - Alan
 *
 *	If not nagling we can send on the before case too..
 */

	if (sk->partial) {
		if (!sk->packets_out ||
		    (sk->nonagle && before(sk->write_seq , sk->window_seq))) {
	  		tcp_send_partial(sk);
	  	}
	}

	release_sock(sk);

out:
	return retval;
}


/*
 *	Send an ack if one is backlogged at this point.
 */

void tcp_read_wakeup(struct sock *sk)
{
	if (!sk->ack_backlog)
		return;

	/*
	 * If we're closed, don't send an ack, or we'll get a RST
	 * from the closed destination.
	 */
	if ((sk->state == TCP_CLOSE) || (sk->state == TCP_TIME_WAIT))
		return;

	tcp_send_ack(sk);
}


/*
 *	Handle reading urgent data. BSD has very simple semantics for
 *	this, no blocking and very strange errors 8)
 */

static int tcp_recv_urg(struct sock * sk, int nonblock,
	     struct msghdr *msg, int len, int flags, int *addr_len)
{
	/*
	 *	No URG data to read
	 */
	if (sk->urginline || !sk->urg_data || sk->urg_data == URG_READ)
		return -EINVAL;	/* Yes this is right ! */

	if (sk->err)
		return sock_error(sk);

	if (sk->state == TCP_CLOSE || sk->done)
	{
		if (!sk->done)
		{
			sk->done = 1;
			return 0;
		}
		return -ENOTCONN;
	}

	if (sk->shutdown & RCV_SHUTDOWN)
	{
		sk->done = 1;
		return 0;
	}
	lock_sock(sk);
	if (sk->urg_data & URG_VALID)
	{
		char c = sk->urg_data;
		if (!(flags & MSG_PEEK))
			sk->urg_data = URG_READ;
		memcpy_toiovec(msg->msg_iov, &c, 1);
		if(msg->msg_name)
		{
			struct sockaddr_in *sin=(struct sockaddr_in *)msg->msg_name;
			sin->sin_family=AF_INET;
			sin->sin_addr.s_addr=sk->daddr;
			sin->sin_port=sk->dummy_th.dest;
		}
		if(addr_len)
			*addr_len=sizeof(struct sockaddr_in);
		release_sock(sk);
		return 1;
	}
	release_sock(sk);

	/*
	 * Fixed the recv(..., MSG_OOB) behaviour.  BSD docs and
	 * the available implementations agree in this case:
	 * this call should never block, independent of the
	 * blocking state of the socket.
	 * Mike <pall@rz.uni-karlsruhe.de>
	 */
	return -EAGAIN;
}

/*
 *	Release a skb if it is no longer needed. This routine
 *	must be called with interrupts disabled or with the
 *	socket locked so that the sk_buff queue operation is ok.
 */

static inline void tcp_eat_skb(struct sock *sk, struct sk_buff * skb)
{
	skb->sk = sk;
	__skb_unlink(skb, &sk->receive_queue);
	kfree_skb(skb, FREE_READ);
}

/*
 *	FIXME:
 *	This routine frees used buffers.
 *	It should consider sending an ACK to let the
 *	other end know we now have a bigger window.
 */

static void cleanup_rbuf(struct sock *sk)
{
	/*
	 * NOTE! The socket must be locked, so that we don't get
	 * a messed-up receive queue.
	 */
	while (!skb_queue_empty(&sk->receive_queue)) {
		struct sk_buff *skb = sk->receive_queue.next;
		if (!skb->used || skb->users)
			break;
		tcp_eat_skb(sk, skb);
	}

	/*
	 * Tell the world if we raised the window.
	 */
	if (tcp_raise_window(sk))
		tcp_send_ack(sk);
}


/*
 *	This routine copies from a sock struct into the user buffer.
 */

static int tcp_recvmsg(struct sock *sk, struct msghdr *msg,
	int len, int nonblock, int flags, int *addr_len)
{
	struct wait_queue wait = { current, NULL };
	int copied = 0;
	u32 peek_seq;
	volatile u32 *seq;	/* So gcc doesn't overoptimise */
	unsigned long used;

	/*
	 *	This error should be checked.
	 */

	if (sk->state == TCP_LISTEN)
		return -ENOTCONN;

	/*
	 *	Urgent data needs to be handled specially.
	 */

	if (flags & MSG_OOB)
		return tcp_recv_urg(sk, nonblock, msg, len, flags, addr_len);

	/*
	 *	Copying sequence to update. This is volatile to handle
	 *	the multi-reader case neatly (memcpy_to/fromfs might be
	 *	inline and thus not flush cached variables otherwise).
	 */

	peek_seq = sk->copied_seq;
	seq = &sk->copied_seq;
	if (flags & MSG_PEEK)
		seq = &peek_seq;

	add_wait_queue(sk->sleep, &wait);
	lock_sock(sk);
	while (len > 0)
	{
		struct sk_buff * skb;
		u32 offset;

		/*
		 * Are we at urgent data? Stop if we have read anything.
		 */

		if (copied && sk->urg_data && sk->urg_seq == *seq)
			break;

		/*
		 * We need to check signals first, to get correct SIGURG
		 * handling.
		 */
		if (current->signal & ~current->blocked) {
			if (copied)
				break;
			copied = -ERESTARTSYS;
			if (nonblock)
				copied = -EAGAIN;
			break;
		}

		/*
		 *	Next get a buffer.
		 */

		current->state = TASK_INTERRUPTIBLE;

		skb = sk->receive_queue.next;
		while (skb != (struct sk_buff *)&sk->receive_queue)
		{
			if (before(*seq, skb->seq))
				break;
			offset = *seq - skb->seq;
			if (skb->h.th->syn)
				offset--;
			if (offset < skb->len)
				goto found_ok_skb;
			if (skb->h.th->fin)
				goto found_fin_ok;
			if (!(flags & MSG_PEEK))
				skb->used = 1;
			skb = skb->next;
		}

		if (copied)
			break;

		if (sk->err && !(flags&MSG_PEEK))
		{
			copied = sock_error(sk);
			break;
		}

		if (sk->state == TCP_CLOSE)
		{
			if (!sk->done)
			{
				sk->done = 1;
				break;
			}
			copied = -ENOTCONN;
			break;
		}

		if (sk->shutdown & RCV_SHUTDOWN)
		{
			sk->done = 1;
			break;
		}

		if (nonblock)
		{
			copied = -EAGAIN;
			break;
		}

		cleanup_rbuf(sk);
		release_sock(sk);
		sk->socket->flags |= SO_WAITDATA;
		schedule();
		sk->socket->flags &= ~SO_WAITDATA;
		lock_sock(sk);
		continue;

	found_ok_skb:
		/*
		 *	Lock the buffer. We can be fairly relaxed as
		 *	an interrupt will never steal a buffer we are
		 *	using unless I've missed something serious in
		 *	tcp_data.
		 */

		skb->users++;

		/*
		 *	Ok so how much can we use ?
		 */

		used = skb->len - offset;
		if (len < used)
			used = len;
		/*
		 *	Do we have urgent data here?
		 */

		if (sk->urg_data)
		{
			u32 urg_offset = sk->urg_seq - *seq;
			if (urg_offset < used)
			{
				if (!urg_offset)
				{
					if (!sk->urginline)
					{
						++*seq;
						offset++;
						used--;
					}
				}
				else
					used = urg_offset;
			}
		}

		/*
		 *	Copy it - We _MUST_ update *seq first so that we
		 *	don't ever double read when we have dual readers
		 */

		*seq += used;

		/*
		 *	This memcpy_tofs can sleep. If it sleeps and we
		 *	do a second read it relies on the skb->users to avoid
		 *	a crash when cleanup_rbuf() gets called.
		 */

		memcpy_toiovec(msg->msg_iov,((unsigned char *)skb->h.th) +
			skb->h.th->doff*4 + offset, used);
		copied += used;
		len -= used;

		/*
		 *	We now will not sleep again until we are finished
		 *	with skb. Sorry if you are doing the SMP port
		 *	but you'll just have to fix it neatly ;)
		 */

		skb->users --;

		if (after(sk->copied_seq,sk->urg_seq))
			sk->urg_data = 0;
		if (used + offset < skb->len)
			continue;

		/*
		 *	Process the FIN.
		 */

		if (skb->h.th->fin)
			goto found_fin_ok;
		if (flags & MSG_PEEK)
			continue;
		skb->used = 1;
		if (!skb->users)
			tcp_eat_skb(sk, skb);
		continue;

	found_fin_ok:
		++*seq;
		if (flags & MSG_PEEK)
			break;

		/*
		 *	All is done
		 */

		skb->used = 1;
		sk->shutdown |= RCV_SHUTDOWN;
		break;

	}

	if(copied>0 && msg->msg_name)
	{
		struct sockaddr_in *sin=(struct sockaddr_in *)msg->msg_name;
		sin->sin_family=AF_INET;
		sin->sin_addr.s_addr=sk->daddr;
		sin->sin_port=sk->dummy_th.dest;
	}
	if(addr_len)
		*addr_len=sizeof(struct sockaddr_in);

	remove_wait_queue(sk->sleep, &wait);
	current->state = TASK_RUNNING;

	/* Clean up data we have read: This will do ACK frames */
	cleanup_rbuf(sk);
	release_sock(sk);
	return copied;
}



/*
 *	State processing on a close. This implements the state shift for
 *	sending our FIN frame. Note that we only send a FIN for some
 *	states. A shutdown() may have already sent the FIN, or we may be
 *	closed.
 */

static int tcp_close_state(struct sock *sk, int dead)
{
	int ns=TCP_CLOSE;
	int send_fin=0;
	switch(sk->state)
	{
		case TCP_SYN_SENT:	/* No SYN back, no FIN needed */
			break;
		case TCP_SYN_RECV:
		case TCP_ESTABLISHED:	/* Closedown begin */
			ns=TCP_FIN_WAIT1;
			send_fin=1;
			break;
		case TCP_FIN_WAIT1:	/* Already closing, or FIN sent: no change */
		case TCP_FIN_WAIT2:
		case TCP_CLOSING:
			ns=sk->state;
			break;
		case TCP_CLOSE:
		case TCP_LISTEN:
			break;
		case TCP_LAST_ACK:	/* Could have shutdown() then close().
					   Be careful not to send double fin. */
			ns=TCP_LAST_ACK;
			break;
		case TCP_CLOSE_WAIT:	/* They have FIN'd us. We send our FIN and
					   wait only for the ACK */
			ns=TCP_LAST_ACK;
			send_fin=1;
	}

	tcp_set_state(sk,ns);

	/*
	 *	This is a (useful) BSD violating of the RFC. There is a
	 *	problem with TCP as specified in that the other end could
	 *	keep a socket open forever with no application left this end.
	 *	We use a 3 minute timeout (about the same as BSD) then kill
	 *	our end. If they send after that then tough - BUT: long enough
	 *	that we won't make the old 4*rto = almost no time - whoops
	 *	reset mistake.
	 */
	if(dead && ns==TCP_FIN_WAIT2)
	{
		int timer_active=del_timer(&sk->timer);
		if(timer_active)
			add_timer(&sk->timer);
		else
			tcp_reset_msl_timer(sk, TIME_CLOSE, TCP_FIN_TIMEOUT);
	}

	return send_fin;
}

/*
 *	Shutdown the sending side of a connection. Much like close except
 *	that we don't receive shut down or set sk->dead.
 */

void tcp_shutdown(struct sock *sk, int how)
{
	/*
	 *	We need to grab some memory, and put together a FIN,
	 *	and then put it into the queue to be sent.
	 *		Tim MacKenzie(tym@dibbler.cs.monash.edu.au) 4 Dec '92.
	 */

	if (!(how & SEND_SHUTDOWN))
		return;

	/*
	 *	If we've already sent a FIN, or it's a closed state
	 */

	if (sk->state == TCP_FIN_WAIT1 ||
	    sk->state == TCP_FIN_WAIT2 ||
	    sk->state == TCP_CLOSING ||
	    sk->state == TCP_LAST_ACK ||
	    sk->state == TCP_TIME_WAIT ||
	    sk->state == TCP_CLOSE ||
	    sk->state == TCP_LISTEN
	  )
	{
		return;
	}
	lock_sock(sk);

	/*
	 * flag that the sender has shutdown
	 */

	sk->shutdown |= SEND_SHUTDOWN;

	/*
	 *  Clear out any half completed packets.
	 */

	if (sk->partial)
		tcp_send_partial(sk);

	/*
	 *	FIN if needed
	 */

	if (tcp_close_state(sk,0))
		tcp_send_fin(sk);

	release_sock(sk);
}


/*
 *	Return 1 if we still have things to send in our buffers.
 */

static inline int closing(struct sock * sk)
{
	switch (sk->state) {
		case TCP_FIN_WAIT1:
		case TCP_CLOSING:
		case TCP_LAST_ACK:
			return 1;
	}
	return 0;
}


static void tcp_close(struct sock *sk, unsigned long timeout)
{
	struct sk_buff *skb;

	/*
	 * We need to grab some memory, and put together a FIN,
	 * and then put it into the queue to be sent.
	 */

	lock_sock(sk);

	if(sk->state == TCP_LISTEN)
	{
		/* Special case */
		tcp_set_state(sk, TCP_CLOSE);
		tcp_close_pending(sk);
		release_sock(sk);
		sk->dead = 1;
		tcp_v4_unhash(sk);
		return;
	}

	sk->keepopen = 1;
	sk->shutdown = SHUTDOWN_MASK;

	if (!sk->dead)
	  	sk->state_change(sk);

	/*
	 *  We need to flush the recv. buffs.  We do this only on the
	 *  descriptor close, not protocol-sourced closes, because the
	 *  reader process may not have drained the data yet!
	 */

	while((skb=skb_dequeue(&sk->receive_queue))!=NULL)
		kfree_skb(skb, FREE_READ);

	/*
	 *	Get rid off any half-completed packets.
	 */

	if (sk->partial)
		tcp_send_partial(sk);

	/*
	 *	Timeout is not the same thing - however the code likes
	 *	to send both the same way (sigh).
	 */

	if (tcp_close_state(sk,1)==1)
	{
		tcp_send_fin(sk);
	}

	if (timeout) {
		cli();
		release_sock(sk);
		current->timeout = timeout;
		while(closing(sk) && current->timeout)
		{
			interruptible_sleep_on(sk->sleep);
			if (current->signal & ~current->blocked)
			{
				break;
			}
		}
		current->timeout=0;
		lock_sock(sk);
		sti();
	}

	/* Now that the socket is dead, if we are in the FIN_WAIT2 state
	 * we may need to set up a timer.
	 */
	if (sk->state==TCP_FIN_WAIT2)
	{
		int timer_active=del_timer(&sk->timer);
		if(timer_active)
			add_timer(&sk->timer);
		else
			tcp_reset_msl_timer(sk, TIME_CLOSE, TCP_FIN_TIMEOUT);
	}

	sk->dead = 1;
	release_sock(sk);

	if(sk->state == TCP_CLOSE)
		tcp_v4_unhash(sk);
}


/*
 * Wait for a incoming connection, avoid race
 * conditions. This must be called with the socket
 * locked.
 */
static struct sk_buff * wait_for_connect(struct sock * sk)
{
	struct wait_queue wait = { current, NULL };
	struct sk_buff * skb = NULL;

	add_wait_queue(sk->sleep, &wait);
	for (;;) {
		current->state = TASK_INTERRUPTIBLE;
		release_sock(sk);
		schedule();
		lock_sock(sk);
		skb = tcp_find_established(sk);
		if (skb)
			break;
		if (current->signal & ~current->blocked)
			break;
	}
	remove_wait_queue(sk->sleep, &wait);
	return skb;
}

/*
 *	This will accept the next outstanding connection.
 *
 *	Be careful about race conditions here - this is subtle.
 */

static struct sock *tcp_accept(struct sock *sk, int flags)
{
	int error;
	struct sk_buff *skb;
	struct sock *newsk = NULL;

  /*
   * We need to make sure that this socket is listening,
   * and that it has something pending.
   */

	error = EINVAL;
	if (sk->state != TCP_LISTEN)
		goto no_listen;

	lock_sock(sk);

	skb = tcp_find_established(sk);
	if (skb) {
got_new_connect:
		__skb_unlink(skb, &sk->receive_queue);
		newsk = skb->sk;
		kfree_skb(skb, FREE_READ);
		sk->ack_backlog--;
		error = 0;
out:
		release_sock(sk);
no_listen:
		sk->err = error;
		return newsk;
	}

	error = EAGAIN;
	if (flags & O_NONBLOCK)
		goto out;
	skb = wait_for_connect(sk);
	if (skb)
		goto got_new_connect;
	error = ERESTARTSYS;
	goto out;
}

/*
 * Check that a TCP address is unique, don't allow multiple
 * connects to/from the same address
 */
static int tcp_unique_address(u32 saddr, u16 snum, u32 daddr, u16 dnum)
{
	int retval = 1, hashent = tcp_hashfn(saddr, snum, daddr, dnum);
	struct sock * sk;

	/* Make sure we are allowed to connect here.
	 * But freeze the hash while we snoop around.
	 */
	SOCKHASH_LOCK();
	sk = tcp_established_hash[hashent];
	for (; sk != NULL; sk = sk->next) {
		if(sk->daddr		== daddr		&& /* remote address */
		   sk->dummy_th.dest	== dnum			&& /* remote port */
		   sk->num		== snum			&& /* local port */
		   sk->saddr		== saddr) {		   /* local address */
			retval = 0;
			break;
		}
	}
	SOCKHASH_UNLOCK();
	return retval;
}


/*
 *	This will initiate an outgoing connection.
 */

static int tcp_connect(struct sock *sk, struct sockaddr_in *usin, int addr_len)
{
	struct sk_buff *buff;
	struct device *dev=NULL;
	unsigned char *ptr;
	int tmp;
	int atype;
	struct tcphdr *t1;
	struct rtable *rt;

	if (sk->state != TCP_CLOSE)
		return(-EISCONN);

	/*
	 *	Don't allow a double connect.
	 */

	if(sk->daddr)
		return -EINVAL;

	if (addr_len < 8)
		return(-EINVAL);

	if (usin->sin_family && usin->sin_family != AF_INET)
		return(-EAFNOSUPPORT);

  	/*
  	 *	connect() to INADDR_ANY means loopback (BSD'ism).
  	 */

  	if (usin->sin_addr.s_addr==INADDR_ANY)
		usin->sin_addr.s_addr=ip_my_addr();

	/*
	 *	Don't want a TCP connection going to a broadcast address
	 */

	if ((atype=ip_chk_addr(usin->sin_addr.s_addr)) == IS_BROADCAST || atype==IS_MULTICAST)
		return -ENETUNREACH;

	if (!tcp_unique_address(sk->saddr, sk->num, usin->sin_addr.s_addr, usin->sin_port))
		return -EADDRNOTAVAIL;

	lock_sock(sk);
	sk->daddr = usin->sin_addr.s_addr;

	sk->rcv_ack_cnt = 1;
	sk->err = 0;
	sk->dummy_th.dest = usin->sin_port;

	buff = sock_wmalloc(sk,MAX_SYN_SIZE,0, GFP_KERNEL);
	if (buff == NULL)
	{
		release_sock(sk);
		return(-ENOMEM);
	}
	buff->sk = sk;
	buff->free = 0;
	buff->localroute = sk->localroute;

	/* If this socket is bound to a particular device, make sure we use it. */
	dev = sk->bound_device;

	/*
	 *	Put in the IP header and routing stuff.
	 */

	tmp = sk->prot->build_header(buff, sk->saddr, sk->daddr, &dev,
		IPPROTO_TCP, sk->opt, MAX_SYN_SIZE,sk->ip_tos,sk->ip_ttl,&sk->ip_route_cache);
	if (tmp < 0)
	{
		sock_wfree(sk, buff);
		release_sock(sk);
		return(-ENETUNREACH);
	}
	if ((rt = sk->ip_route_cache) != NULL && !sk->saddr)
		sk->saddr = rt->rt_src;
	sk->rcv_saddr = sk->saddr;

	/*
	 * Set up our outgoing TCP sequence number
	 */
	sk->write_seq = secure_tcp_sequence_number(sk->saddr, sk->daddr,
						   sk->dummy_th.source,
						   usin->sin_port);
	sk->window_seq = sk->write_seq;
	sk->rcv_ack_seq = sk->write_seq -1;

	t1 = (struct tcphdr *) skb_put(buff,sizeof(struct tcphdr));

	memcpy(t1,(void *)&(sk->dummy_th), sizeof(*t1));
	buff->seq = sk->write_seq++;
	t1->seq = htonl(buff->seq);
	sk->sent_seq = sk->write_seq;
	buff->end_seq = sk->write_seq;
	t1->ack = 0;
	t1->window = 2;
	t1->syn = 1;
	t1->doff = 6;
	/* use 512 or whatever user asked for */

	if(rt!=NULL && (rt->rt_flags&RTF_WINDOW))
		sk->window_clamp=rt->rt_window;
	else
		sk->window_clamp=0;

	if (sk->user_mss)
		sk->mtu = sk->user_mss;
	else if (rt)
		sk->mtu = rt->rt_mtu - sizeof(struct iphdr) - sizeof(struct tcphdr);
	else
		sk->mtu = 576 - sizeof(struct iphdr) - sizeof(struct tcphdr);

	/*
	 *	but not bigger than device MTU
	 */

	sk->mtu = min(sk->mtu, dev->mtu - sizeof(struct iphdr) - sizeof(struct tcphdr));

	/* Must check it here, just to be absolutely safe.  If we end up
	 * with an sk->mtu of zero, we can thus end up with an sk->mss
	 * of zero, which causes us to bomb out in tcp_do_sendmsg. -DaveM
	 */
	if(sk->mtu < 32)
		sk->mtu = 32;	/* Sanity limit */

	/*
	 *	Put in the TCP options to say MTU.
	 */

	ptr = skb_put(buff,4);
	ptr[0] = 2;
	ptr[1] = 4;
	ptr[2] = (sk->mtu) >> 8;
	ptr[3] = (sk->mtu) & 0xff;
	buff->csum = csum_partial(ptr, 4, 0);
	tcp_send_check(t1, sk->saddr, sk->daddr,
		  sizeof(struct tcphdr) + 4, buff);

	tcp_set_state(sk,TCP_SYN_SENT);

	/* Socket identity change complete, no longer
	 * in TCP_CLOSE, so rehash.
	 */
	tcp_v4_rehash(sk);

	if(rt&&rt->rt_flags&RTF_IRTT)
		sk->rto = rt->rt_irtt;
	else
		sk->rto = TCP_TIMEOUT_INIT;
	sk->delack_timer.function = tcp_delack_timer;
	sk->delack_timer.data = (unsigned long) sk;
	sk->retransmit_timer.function = tcp_retransmit_timer;
	sk->retransmit_timer.data = (unsigned long)sk;
	sk->retransmits = 0;
	sk->prot->queue_xmit(sk, dev, buff, 0);
	tcp_reset_xmit_timer(sk, TIME_WRITE, sk->rto);
	tcp_statistics.TcpActiveOpens++;
	tcp_statistics.TcpOutSegs++;

	release_sock(sk);
	return(0);
}

/*
 *	Socket option code for TCP.
 */

int tcp_setsockopt(struct sock *sk, int level, int optname, char *optval, int optlen)
{
	int val,err;

	if(level!=SOL_TCP)
		return ip_setsockopt(sk,level,optname,optval,optlen);

  	if (optval == NULL)
  		return(-EINVAL);

  	err=verify_area(VERIFY_READ, optval, sizeof(int));
  	if(err)
  		return err;

  	val = get_user((int *)optval);

	switch(optname)
	{
		case TCP_MAXSEG:
/*
 * values greater than interface MTU won't take effect.  however at
 * the point when this call is done we typically don't yet know
 * which interface is going to be used
 */
	  		if(val<1||val>MAX_WINDOW)
				return -EINVAL;
			sk->user_mss=val;
			return 0;
		case TCP_NODELAY:
			sk->nonagle=(val==0)?0:1;
			return 0;
		default:
			return(-ENOPROTOOPT);
	}
}

int tcp_getsockopt(struct sock *sk, int level, int optname, char *optval, int *optlen)
{
	int val,err;

	if(level!=SOL_TCP)
		return ip_getsockopt(sk,level,optname,optval,optlen);

	switch(optname)
	{
		case TCP_MAXSEG:
			val=sk->user_mss;
			break;
		case TCP_NODELAY:
			val=sk->nonagle;
			break;
		default:
			return(-ENOPROTOOPT);
	}
	err=verify_area(VERIFY_WRITE, optlen, sizeof(int));
	if(err)
  		return err;
  	put_user(sizeof(int),(int *) optlen);

  	err=verify_area(VERIFY_WRITE, optval, sizeof(int));
  	if(err)
  		return err;
  	put_user(val,(int *)optval);

  	return(0);
}


struct proto tcp_prot = {
	(struct sock *)&tcp_prot,	/* sklist_next */
	(struct sock *)&tcp_prot,	/* sklist_prev */
	tcp_close,			/* close */
	ip_build_header,		/* build_header */
	tcp_connect,			/* connect */
	tcp_accept,			/* accept */
	ip_queue_xmit,			/* queue_xmit */
	tcp_retransmit,			/* retransmit */
	tcp_write_wakeup,		/* write_wakeup */
	tcp_read_wakeup,		/* read_wakeup */
	tcp_rcv,			/* rcv */
	tcp_select,			/* select */
	tcp_ioctl,			/* ioctl */
	NULL,				/* init */
	tcp_shutdown,			/* shutdown */
	tcp_setsockopt,			/* setsockopt */
	tcp_getsockopt,			/* getsockopt */
	tcp_sendmsg,			/* sendmsg */
	tcp_recvmsg,			/* recvmsg */
	NULL,				/* bind */
	tcp_v4_hash,			/* hash */
	tcp_v4_unhash,			/* unhash */
	tcp_v4_rehash,			/* rehash */
	tcp_good_socknum,		/* good_socknum */
	tcp_v4_verify_bind,		/* verify_bind */
	128,				/* max_header */
	0,				/* retransmits */
	"TCP",				/* name */
	0,				/* inuse */
	0				/* highestinuse */
};