summaryrefslogtreecommitdiffstats
path: root/sys/netinet/tcp_output.c
blob: 21620d3a1920373adb2bda14426577937757b732 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
/*-
 * Copyright (c) 1982, 1986, 1988, 1990, 1993, 1995
 *	The Regents of the University of California.  All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 4. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 *	@(#)tcp_output.c	8.4 (Berkeley) 5/24/95
 */

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

#include "opt_inet.h"
#include "opt_inet6.h"
#include "opt_ipsec.h"
#include "opt_tcpdebug.h"

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/domain.h>
#include <sys/hhook.h>
#include <sys/kernel.h>
#include <sys/lock.h>
#include <sys/mbuf.h>
#include <sys/mutex.h>
#include <sys/protosw.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/sysctl.h>

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

#include <netinet/cc.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
#include <netinet/in_pcb.h>
#include <netinet/ip_var.h>
#include <netinet/ip_options.h>
#ifdef INET6
#include <netinet6/in6_pcb.h>
#include <netinet/ip6.h>
#include <netinet6/ip6_var.h>
#endif
#define	TCPOUTFLAGS
#include <netinet/tcp_fsm.h>
#include <netinet/tcp_seq.h>
#include <netinet/tcp_timer.h>
#include <netinet/tcp_var.h>
#include <netinet/tcpip.h>
#ifdef TCPDEBUG
#include <netinet/tcp_debug.h>
#endif

#ifdef IPSEC
#include <netipsec/ipsec.h>
#endif /*IPSEC*/

#include <machine/in_cksum.h>

#include <security/mac/mac_framework.h>

VNET_DEFINE(int, path_mtu_discovery) = 1;
SYSCTL_VNET_INT(_net_inet_tcp, OID_AUTO, path_mtu_discovery, CTLFLAG_RW,
	&VNET_NAME(path_mtu_discovery), 1,
	"Enable Path MTU Discovery");

VNET_DEFINE(int, tcp_do_tso) = 1;
#define	V_tcp_do_tso		VNET(tcp_do_tso)
SYSCTL_VNET_INT(_net_inet_tcp, OID_AUTO, tso, CTLFLAG_RW,
	&VNET_NAME(tcp_do_tso), 0,
	"Enable TCP Segmentation Offload");

VNET_DEFINE(int, tcp_sendspace) = 1024*32;
#define	V_tcp_sendspace	VNET(tcp_sendspace)
SYSCTL_VNET_INT(_net_inet_tcp, TCPCTL_SENDSPACE, sendspace, CTLFLAG_RW,
	&VNET_NAME(tcp_sendspace), 0, "Initial send socket buffer size");

VNET_DEFINE(int, tcp_do_autosndbuf) = 1;
#define	V_tcp_do_autosndbuf	VNET(tcp_do_autosndbuf)
SYSCTL_VNET_INT(_net_inet_tcp, OID_AUTO, sendbuf_auto, CTLFLAG_RW,
	&VNET_NAME(tcp_do_autosndbuf), 0,
	"Enable automatic send buffer sizing");

VNET_DEFINE(int, tcp_autosndbuf_inc) = 8*1024;
#define	V_tcp_autosndbuf_inc	VNET(tcp_autosndbuf_inc)
SYSCTL_VNET_INT(_net_inet_tcp, OID_AUTO, sendbuf_inc, CTLFLAG_RW,
	&VNET_NAME(tcp_autosndbuf_inc), 0,
	"Incrementor step size of automatic send buffer");

VNET_DEFINE(int, tcp_autosndbuf_max) = 2*1024*1024;
#define	V_tcp_autosndbuf_max	VNET(tcp_autosndbuf_max)
SYSCTL_VNET_INT(_net_inet_tcp, OID_AUTO, sendbuf_max, CTLFLAG_RW,
	&VNET_NAME(tcp_autosndbuf_max), 0,
	"Max size of automatic send buffer");

static void inline	hhook_run_tcp_est_out(struct tcpcb *tp,
			    struct tcphdr *th, struct tcpopt *to,
			    long len, int tso);
static void inline	cc_after_idle(struct tcpcb *tp);

/*
 * Wrapper for the TCP established ouput helper hook.
 */
static void inline
hhook_run_tcp_est_out(struct tcpcb *tp, struct tcphdr *th,
    struct tcpopt *to, long len, int tso)
{
	struct tcp_hhook_data hhook_data;

	if (V_tcp_hhh[HHOOK_TCP_EST_OUT]->hhh_nhooks > 0) {
		hhook_data.tp = tp;
		hhook_data.th = th;
		hhook_data.to = to;
		hhook_data.len = len;
		hhook_data.tso = tso;

		hhook_run_hooks(V_tcp_hhh[HHOOK_TCP_EST_OUT], &hhook_data,
		    tp->osd);
	}
}

/*
 * CC wrapper hook functions
 */
static void inline
cc_after_idle(struct tcpcb *tp)
{
	INP_WLOCK_ASSERT(tp->t_inpcb);

	if (CC_ALGO(tp)->after_idle != NULL)
		CC_ALGO(tp)->after_idle(tp->ccv);
}

/*
 * Tcp output routine: figure out what should be sent and send it.
 */
int
tcp_output(struct tcpcb *tp)
{
	struct socket *so = tp->t_inpcb->inp_socket;
	long len, recwin, sendwin;
	int off, flags, error = 0;	/* Keep compiler happy */
	struct mbuf *m;
	struct ip *ip = NULL;
	struct ipovly *ipov = NULL;
	struct tcphdr *th;
	u_char opt[TCP_MAXOLEN];
	unsigned ipoptlen, optlen, hdrlen;
#ifdef IPSEC
	unsigned ipsec_optlen = 0;
#endif
	int idle, sendalot;
	int sack_rxmit, sack_bytes_rxmt;
	struct sackhole *p;
	int tso;
	struct tcpopt to;
#if 0
	int maxburst = TCP_MAXBURST;
#endif
#ifdef INET6
	struct ip6_hdr *ip6 = NULL;
	int isipv6;

	isipv6 = (tp->t_inpcb->inp_vflag & INP_IPV6) != 0;
#endif

	INP_WLOCK_ASSERT(tp->t_inpcb);

	/*
	 * Determine length of data that should be transmitted,
	 * and flags that will be used.
	 * If there is some data or critical controls (SYN, RST)
	 * to send, then transmit; otherwise, investigate further.
	 */
	idle = (tp->t_flags & TF_LASTIDLE) || (tp->snd_max == tp->snd_una);
	if (idle && ticks - tp->t_rcvtime >= tp->t_rxtcur)
		cc_after_idle(tp);
	tp->t_flags &= ~TF_LASTIDLE;
	if (idle) {
		if (tp->t_flags & TF_MORETOCOME) {
			tp->t_flags |= TF_LASTIDLE;
			idle = 0;
		}
	}
again:
	/*
	 * If we've recently taken a timeout, snd_max will be greater than
	 * snd_nxt.  There may be SACK information that allows us to avoid
	 * resending already delivered data.  Adjust snd_nxt accordingly.
	 */
	if ((tp->t_flags & TF_SACK_PERMIT) &&
	    SEQ_LT(tp->snd_nxt, tp->snd_max))
		tcp_sack_adjust(tp);
	sendalot = 0;
	tso = 0;
	off = tp->snd_nxt - tp->snd_una;
	sendwin = min(tp->snd_wnd, tp->snd_cwnd);

	flags = tcp_outflags[tp->t_state];
	/*
	 * Send any SACK-generated retransmissions.  If we're explicitly trying
	 * to send out new data (when sendalot is 1), bypass this function.
	 * If we retransmit in fast recovery mode, decrement snd_cwnd, since
	 * we're replacing a (future) new transmission with a retransmission
	 * now, and we previously incremented snd_cwnd in tcp_input().
	 */
	/*
	 * Still in sack recovery , reset rxmit flag to zero.
	 */
	sack_rxmit = 0;
	sack_bytes_rxmt = 0;
	len = 0;
	p = NULL;
	if ((tp->t_flags & TF_SACK_PERMIT) && IN_FASTRECOVERY(tp->t_flags) &&
	    (p = tcp_sack_output(tp, &sack_bytes_rxmt))) {
		long cwin;
		
		cwin = min(tp->snd_wnd, tp->snd_cwnd) - sack_bytes_rxmt;
		if (cwin < 0)
			cwin = 0;
		/* Do not retransmit SACK segments beyond snd_recover */
		if (SEQ_GT(p->end, tp->snd_recover)) {
			/*
			 * (At least) part of sack hole extends beyond
			 * snd_recover. Check to see if we can rexmit data
			 * for this hole.
			 */
			if (SEQ_GEQ(p->rxmit, tp->snd_recover)) {
				/*
				 * Can't rexmit any more data for this hole.
				 * That data will be rexmitted in the next
				 * sack recovery episode, when snd_recover
				 * moves past p->rxmit.
				 */
				p = NULL;
				goto after_sack_rexmit;
			} else
				/* Can rexmit part of the current hole */
				len = ((long)ulmin(cwin,
						   tp->snd_recover - p->rxmit));
		} else
			len = ((long)ulmin(cwin, p->end - p->rxmit));
		off = p->rxmit - tp->snd_una;
		KASSERT(off >= 0,("%s: sack block to the left of una : %d",
		    __func__, off));
		if (len > 0) {
			sack_rxmit = 1;
			sendalot = 1;
			TCPSTAT_INC(tcps_sack_rexmits);
			TCPSTAT_ADD(tcps_sack_rexmit_bytes,
			    min(len, tp->t_maxseg));
		}
	}
after_sack_rexmit:
	/*
	 * Get standard flags, and add SYN or FIN if requested by 'hidden'
	 * state flags.
	 */
	if (tp->t_flags & TF_NEEDFIN)
		flags |= TH_FIN;
	if (tp->t_flags & TF_NEEDSYN)
		flags |= TH_SYN;

	SOCKBUF_LOCK(&so->so_snd);
	/*
	 * If in persist timeout with window of 0, send 1 byte.
	 * Otherwise, if window is small but nonzero
	 * and timer expired, we will send what we can
	 * and go to transmit state.
	 */
	if (tp->t_flags & TF_FORCEDATA) {
		if (sendwin == 0) {
			/*
			 * If we still have some data to send, then
			 * clear the FIN bit.  Usually this would
			 * happen below when it realizes that we
			 * aren't sending all the data.  However,
			 * if we have exactly 1 byte of unsent data,
			 * then it won't clear the FIN bit below,
			 * and if we are in persist state, we wind
			 * up sending the packet without recording
			 * that we sent the FIN bit.
			 *
			 * We can't just blindly clear the FIN bit,
			 * because if we don't have any more data
			 * to send then the probe will be the FIN
			 * itself.
			 */
			if (off < so->so_snd.sb_cc)
				flags &= ~TH_FIN;
			sendwin = 1;
		} else {
			tcp_timer_activate(tp, TT_PERSIST, 0);
			tp->t_rxtshift = 0;
		}
	}

	/*
	 * If snd_nxt == snd_max and we have transmitted a FIN, the
	 * offset will be > 0 even if so_snd.sb_cc is 0, resulting in
	 * a negative length.  This can also occur when TCP opens up
	 * its congestion window while receiving additional duplicate
	 * acks after fast-retransmit because TCP will reset snd_nxt
	 * to snd_max after the fast-retransmit.
	 *
	 * In the normal retransmit-FIN-only case, however, snd_nxt will
	 * be set to snd_una, the offset will be 0, and the length may
	 * wind up 0.
	 *
	 * If sack_rxmit is true we are retransmitting from the scoreboard
	 * in which case len is already set.
	 */
	if (sack_rxmit == 0) {
		if (sack_bytes_rxmt == 0)
			len = ((long)ulmin(so->so_snd.sb_cc, sendwin) - off);
		else {
			long cwin;

                        /*
			 * We are inside of a SACK recovery episode and are
			 * sending new data, having retransmitted all the
			 * data possible in the scoreboard.
			 */
			len = ((long)ulmin(so->so_snd.sb_cc, tp->snd_wnd) 
			       - off);
			/*
			 * Don't remove this (len > 0) check !
			 * We explicitly check for len > 0 here (although it 
			 * isn't really necessary), to work around a gcc 
			 * optimization issue - to force gcc to compute
			 * len above. Without this check, the computation
			 * of len is bungled by the optimizer.
			 */
			if (len > 0) {
				cwin = tp->snd_cwnd - 
					(tp->snd_nxt - tp->sack_newdata) -
					sack_bytes_rxmt;
				if (cwin < 0)
					cwin = 0;
				len = lmin(len, cwin);
			}
		}
	}

	/*
	 * Lop off SYN bit if it has already been sent.  However, if this
	 * is SYN-SENT state and if segment contains data and if we don't
	 * know that foreign host supports TAO, suppress sending segment.
	 */
	if ((flags & TH_SYN) && SEQ_GT(tp->snd_nxt, tp->snd_una)) {
		if (tp->t_state != TCPS_SYN_RECEIVED)
			flags &= ~TH_SYN;
		off--, len++;
	}

	/*
	 * Be careful not to send data and/or FIN on SYN segments.
	 * This measure is needed to prevent interoperability problems
	 * with not fully conformant TCP implementations.
	 */
	if ((flags & TH_SYN) && (tp->t_flags & TF_NOOPT)) {
		len = 0;
		flags &= ~TH_FIN;
	}

	if (len < 0) {
		/*
		 * If FIN has been sent but not acked,
		 * but we haven't been called to retransmit,
		 * len will be < 0.  Otherwise, window shrank
		 * after we sent into it.  If window shrank to 0,
		 * cancel pending retransmit, pull snd_nxt back
		 * to (closed) window, and set the persist timer
		 * if it isn't already going.  If the window didn't
		 * close completely, just wait for an ACK.
		 */
		len = 0;
		if (sendwin == 0) {
			tcp_timer_activate(tp, TT_REXMT, 0);
			tp->t_rxtshift = 0;
			tp->snd_nxt = tp->snd_una;
			if (!tcp_timer_active(tp, TT_PERSIST))
				tcp_setpersist(tp);
		}
	}

	/* len will be >= 0 after this point. */
	KASSERT(len >= 0, ("[%s:%d]: len < 0", __func__, __LINE__));

	/*
	 * Automatic sizing of send socket buffer.  Often the send buffer
	 * size is not optimally adjusted to the actual network conditions
	 * at hand (delay bandwidth product).  Setting the buffer size too
	 * small limits throughput on links with high bandwidth and high
	 * delay (eg. trans-continental/oceanic links).  Setting the
	 * buffer size too big consumes too much real kernel memory,
	 * especially with many connections on busy servers.
	 *
	 * The criteria to step up the send buffer one notch are:
	 *  1. receive window of remote host is larger than send buffer
	 *     (with a fudge factor of 5/4th);
	 *  2. send buffer is filled to 7/8th with data (so we actually
	 *     have data to make use of it);
	 *  3. send buffer fill has not hit maximal automatic size;
	 *  4. our send window (slow start and cogestion controlled) is
	 *     larger than sent but unacknowledged data in send buffer.
	 *
	 * The remote host receive window scaling factor may limit the
	 * growing of the send buffer before it reaches its allowed
	 * maximum.
	 *
	 * It scales directly with slow start or congestion window
	 * and does at most one step per received ACK.  This fast
	 * scaling has the drawback of growing the send buffer beyond
	 * what is strictly necessary to make full use of a given
	 * delay*bandwith product.  However testing has shown this not
	 * to be much of an problem.  At worst we are trading wasting
	 * of available bandwith (the non-use of it) for wasting some
	 * socket buffer memory.
	 *
	 * TODO: Shrink send buffer during idle periods together
	 * with congestion window.  Requires another timer.  Has to
	 * wait for upcoming tcp timer rewrite.
	 */
	if (V_tcp_do_autosndbuf && so->so_snd.sb_flags & SB_AUTOSIZE) {
		if ((tp->snd_wnd / 4 * 5) >= so->so_snd.sb_hiwat &&
		    so->so_snd.sb_cc >= (so->so_snd.sb_hiwat / 8 * 7) &&
		    so->so_snd.sb_cc < V_tcp_autosndbuf_max &&
		    sendwin >= (so->so_snd.sb_cc - (tp->snd_nxt - tp->snd_una))) {
			if (!sbreserve_locked(&so->so_snd,
			    min(so->so_snd.sb_hiwat + V_tcp_autosndbuf_inc,
			     V_tcp_autosndbuf_max), so, curthread))
				so->so_snd.sb_flags &= ~SB_AUTOSIZE;
		}
	}

	/*
	 * Decide if we can use TCP Segmentation Offloading (if supported by
	 * hardware).
	 *
	 * TSO may only be used if we are in a pure bulk sending state.  The
	 * presence of TCP-MD5, SACK retransmits, SACK advertizements and
	 * IP options prevent using TSO.  With TSO the TCP header is the same
	 * (except for the sequence number) for all generated packets.  This
	 * makes it impossible to transmit any options which vary per generated
	 * segment or packet.
	 */
#ifdef IPSEC
	/*
	 * Pre-calculate here as we save another lookup into the darknesses
	 * of IPsec that way and can actually decide if TSO is ok.
	 */
	ipsec_optlen = ipsec_hdrsiz_tcp(tp);
#endif
	if ((tp->t_flags & TF_TSO) && V_tcp_do_tso && len > tp->t_maxseg &&
	    ((tp->t_flags & TF_SIGNATURE) == 0) &&
	    tp->rcv_numsacks == 0 && sack_rxmit == 0 &&
#ifdef IPSEC
	    ipsec_optlen == 0 &&
#endif
	    tp->t_inpcb->inp_options == NULL &&
	    tp->t_inpcb->in6p_options == NULL)
		tso = 1;

	if (sack_rxmit) {
		if (SEQ_LT(p->rxmit + len, tp->snd_una + so->so_snd.sb_cc))
			flags &= ~TH_FIN;
	} else {
		if (SEQ_LT(tp->snd_nxt + len, tp->snd_una + so->so_snd.sb_cc))
			flags &= ~TH_FIN;
	}

	recwin = sbspace(&so->so_rcv);

	/*
	 * Sender silly window avoidance.   We transmit under the following
	 * conditions when len is non-zero:
	 *
	 *	- We have a full segment (or more with TSO)
	 *	- This is the last buffer in a write()/send() and we are
	 *	  either idle or running NODELAY
	 *	- we've timed out (e.g. persist timer)
	 *	- we have more then 1/2 the maximum send window's worth of
	 *	  data (receiver may be limited the window size)
	 *	- we need to retransmit
	 */
	if (len) {
		if (len >= tp->t_maxseg)
			goto send;
		/*
		 * NOTE! on localhost connections an 'ack' from the remote
		 * end may occur synchronously with the output and cause
		 * us to flush a buffer queued with moretocome.  XXX
		 *
		 * note: the len + off check is almost certainly unnecessary.
		 */
		if (!(tp->t_flags & TF_MORETOCOME) &&	/* normal case */
		    (idle || (tp->t_flags & TF_NODELAY)) &&
		    len + off >= so->so_snd.sb_cc &&
		    (tp->t_flags & TF_NOPUSH) == 0) {
			goto send;
		}
		if (tp->t_flags & TF_FORCEDATA)		/* typ. timeout case */
			goto send;
		if (len >= tp->max_sndwnd / 2 && tp->max_sndwnd > 0)
			goto send;
		if (SEQ_LT(tp->snd_nxt, tp->snd_max))	/* retransmit case */
			goto send;
		if (sack_rxmit)
			goto send;
	}

	/*
	 * Compare available window to amount of window
	 * known to peer (as advertised window less
	 * next expected input).  If the difference is at least two
	 * max size segments, or at least 50% of the maximum possible
	 * window, then want to send a window update to peer.
	 * Skip this if the connection is in T/TCP half-open state.
	 * Don't send pure window updates when the peer has closed
	 * the connection and won't ever send more data.
	 */
	if (recwin > 0 && !(tp->t_flags & TF_NEEDSYN) &&
	    !TCPS_HAVERCVDFIN(tp->t_state)) {
		/*
		 * "adv" is the amount we can increase the window,
		 * taking into account that we are limited by
		 * TCP_MAXWIN << tp->rcv_scale.
		 */
		long adv;
		int oldwin;

		adv = min(recwin, (long)TCP_MAXWIN << tp->rcv_scale);
		if (SEQ_GT(tp->rcv_adv, tp->rcv_nxt)) {
			oldwin = (tp->rcv_adv - tp->rcv_nxt);
			adv -= oldwin;
		} else
			oldwin = 0;

		/* 
		 * If the new window size ends up being the same as the old
		 * size when it is scaled, then don't force a window update.
		 */
		if (oldwin >> tp->rcv_scale == (adv + oldwin) >> tp->rcv_scale)
			goto dontupdate;
		if (adv >= (long) (2 * tp->t_maxseg))
			goto send;
		if (2 * adv >= (long) so->so_rcv.sb_hiwat)
			goto send;
	}
dontupdate:

	/*
	 * Send if we owe the peer an ACK, RST, SYN, or urgent data.  ACKNOW
	 * is also a catch-all for the retransmit timer timeout case.
	 */
	if (tp->t_flags & TF_ACKNOW)
		goto send;
	if ((flags & TH_RST) ||
	    ((flags & TH_SYN) && (tp->t_flags & TF_NEEDSYN) == 0))
		goto send;
	if (SEQ_GT(tp->snd_up, tp->snd_una))
		goto send;
	/*
	 * If our state indicates that FIN should be sent
	 * and we have not yet done so, then we need to send.
	 */
	if (flags & TH_FIN &&
	    ((tp->t_flags & TF_SENTFIN) == 0 || tp->snd_nxt == tp->snd_una))
		goto send;
	/*
	 * In SACK, it is possible for tcp_output to fail to send a segment
	 * after the retransmission timer has been turned off.  Make sure
	 * that the retransmission timer is set.
	 */
	if ((tp->t_flags & TF_SACK_PERMIT) &&
	    SEQ_GT(tp->snd_max, tp->snd_una) &&
	    !tcp_timer_active(tp, TT_REXMT) &&
	    !tcp_timer_active(tp, TT_PERSIST)) {
		tcp_timer_activate(tp, TT_REXMT, tp->t_rxtcur);
		goto just_return;
	} 
	/*
	 * TCP window updates are not reliable, rather a polling protocol
	 * using ``persist'' packets is used to insure receipt of window
	 * updates.  The three ``states'' for the output side are:
	 *	idle			not doing retransmits or persists
	 *	persisting		to move a small or zero window
	 *	(re)transmitting	and thereby not persisting
	 *
	 * tcp_timer_active(tp, TT_PERSIST)
	 *	is true when we are in persist state.
	 * (tp->t_flags & TF_FORCEDATA)
	 *	is set when we are called to send a persist packet.
	 * tcp_timer_active(tp, TT_REXMT)
	 *	is set when we are retransmitting
	 * The output side is idle when both timers are zero.
	 *
	 * If send window is too small, there is data to transmit, and no
	 * retransmit or persist is pending, then go to persist state.
	 * If nothing happens soon, send when timer expires:
	 * if window is nonzero, transmit what we can,
	 * otherwise force out a byte.
	 */
	if (so->so_snd.sb_cc && !tcp_timer_active(tp, TT_REXMT) &&
	    !tcp_timer_active(tp, TT_PERSIST)) {
		tp->t_rxtshift = 0;
		tcp_setpersist(tp);
	}

	/*
	 * No reason to send a segment, just return.
	 */
just_return:
	SOCKBUF_UNLOCK(&so->so_snd);
	return (0);

send:
	SOCKBUF_LOCK_ASSERT(&so->so_snd);
	/*
	 * Before ESTABLISHED, force sending of initial options
	 * unless TCP set not to do any options.
	 * NOTE: we assume that the IP/TCP header plus TCP options
	 * always fit in a single mbuf, leaving room for a maximum
	 * link header, i.e.
	 *	max_linkhdr + sizeof (struct tcpiphdr) + optlen <= MCLBYTES
	 */
	optlen = 0;
#ifdef INET6
	if (isipv6)
		hdrlen = sizeof (struct ip6_hdr) + sizeof (struct tcphdr);
	else
#endif
		hdrlen = sizeof (struct tcpiphdr);

	/*
	 * Compute options for segment.
	 * We only have to care about SYN and established connection
	 * segments.  Options for SYN-ACK segments are handled in TCP
	 * syncache.
	 */
	if ((tp->t_flags & TF_NOOPT) == 0) {
		to.to_flags = 0;
		/* Maximum segment size. */
		if (flags & TH_SYN) {
			tp->snd_nxt = tp->iss;
			to.to_mss = tcp_mssopt(&tp->t_inpcb->inp_inc);
			to.to_flags |= TOF_MSS;
		}
		/* Window scaling. */
		if ((flags & TH_SYN) && (tp->t_flags & TF_REQ_SCALE)) {
			to.to_wscale = tp->request_r_scale;
			to.to_flags |= TOF_SCALE;
		}
		/* Timestamps. */
		if ((tp->t_flags & TF_RCVD_TSTMP) ||
		    ((flags & TH_SYN) && (tp->t_flags & TF_REQ_TSTMP))) {
			to.to_tsval = ticks + tp->ts_offset;
			to.to_tsecr = tp->ts_recent;
			to.to_flags |= TOF_TS;
			/* Set receive buffer autosizing timestamp. */
			if (tp->rfbuf_ts == 0 &&
			    (so->so_rcv.sb_flags & SB_AUTOSIZE))
				tp->rfbuf_ts = ticks;
		}
		/* Selective ACK's. */
		if (tp->t_flags & TF_SACK_PERMIT) {
			if (flags & TH_SYN)
				to.to_flags |= TOF_SACKPERM;
			else if (TCPS_HAVEESTABLISHED(tp->t_state) &&
			    (tp->t_flags & TF_SACK_PERMIT) &&
			    tp->rcv_numsacks > 0) {
				to.to_flags |= TOF_SACK;
				to.to_nsacks = tp->rcv_numsacks;
				to.to_sacks = (u_char *)tp->sackblks;
			}
		}
#ifdef TCP_SIGNATURE
		/* TCP-MD5 (RFC2385). */
		if (tp->t_flags & TF_SIGNATURE)
			to.to_flags |= TOF_SIGNATURE;
#endif /* TCP_SIGNATURE */

		/* Processing the options. */
		hdrlen += optlen = tcp_addoptions(&to, opt);
	}

#ifdef INET6
	if (isipv6)
		ipoptlen = ip6_optlen(tp->t_inpcb);
	else
#endif
	if (tp->t_inpcb->inp_options)
		ipoptlen = tp->t_inpcb->inp_options->m_len -
				offsetof(struct ipoption, ipopt_list);
	else
		ipoptlen = 0;
#ifdef IPSEC
	ipoptlen += ipsec_optlen;
#endif

	/*
	 * Adjust data length if insertion of options will
	 * bump the packet length beyond the t_maxopd length.
	 * Clear the FIN bit because we cut off the tail of
	 * the segment.
	 */
	if (len + optlen + ipoptlen > tp->t_maxopd) {
		flags &= ~TH_FIN;

		if (tso) {
			KASSERT(ipoptlen == 0,
			    ("%s: TSO can't do IP options", __func__));

			/*
			 * Limit a burst to IP_MAXPACKET minus IP,
			 * TCP and options length to keep ip->ip_len
			 * from overflowing.
			 */
			if (len > IP_MAXPACKET - hdrlen) {
				len = IP_MAXPACKET - hdrlen;
				sendalot = 1;
			}

			/*
			 * Prevent the last segment from being
			 * fractional unless the send sockbuf can
			 * be emptied.
			 */
			if (sendalot && off + len < so->so_snd.sb_cc) {
				len -= len % (tp->t_maxopd - optlen);
				sendalot = 1;
			}

			/*
			 * Send the FIN in a separate segment
			 * after the bulk sending is done.
			 * We don't trust the TSO implementations
			 * to clear the FIN flag on all but the
			 * last segment.
			 */
			if (tp->t_flags & TF_NEEDFIN)
				sendalot = 1;

		} else {
			len = tp->t_maxopd - optlen - ipoptlen;
			sendalot = 1;
		}
	} else
		tso = 0;

	KASSERT(len + hdrlen + ipoptlen <= IP_MAXPACKET,
	    ("%s: len > IP_MAXPACKET", __func__));

/*#ifdef DIAGNOSTIC*/
#ifdef INET6
	if (max_linkhdr + hdrlen > MCLBYTES)
#else
	if (max_linkhdr + hdrlen > MHLEN)
#endif
		panic("tcphdr too big");
/*#endif*/

	/*
	 * This KASSERT is here to catch edge cases at a well defined place.
	 * Before, those had triggered (random) panic conditions further down.
	 */
	KASSERT(len >= 0, ("[%s:%d]: len < 0", __func__, __LINE__));

	/*
	 * Grab a header mbuf, attaching a copy of data to
	 * be transmitted, and initialize the header from
	 * the template for sends on this connection.
	 */
	if (len) {
		struct mbuf *mb;
		u_int moff;

		if ((tp->t_flags & TF_FORCEDATA) && len == 1)
			TCPSTAT_INC(tcps_sndprobe);
		else if (SEQ_LT(tp->snd_nxt, tp->snd_max) || sack_rxmit) {
			tp->t_sndrexmitpack++;
			TCPSTAT_INC(tcps_sndrexmitpack);
			TCPSTAT_ADD(tcps_sndrexmitbyte, len);
		} else {
			TCPSTAT_INC(tcps_sndpack);
			TCPSTAT_ADD(tcps_sndbyte, len);
		}
		MGETHDR(m, M_DONTWAIT, MT_DATA);
		if (m == NULL) {
			SOCKBUF_UNLOCK(&so->so_snd);
			error = ENOBUFS;
			goto out;
		}
#ifdef INET6
		if (MHLEN < hdrlen + max_linkhdr) {
			MCLGET(m, M_DONTWAIT);
			if ((m->m_flags & M_EXT) == 0) {
				SOCKBUF_UNLOCK(&so->so_snd);
				m_freem(m);
				error = ENOBUFS;
				goto out;
			}
		}
#endif
		m->m_data += max_linkhdr;
		m->m_len = hdrlen;

		/*
		 * Start the m_copy functions from the closest mbuf
		 * to the offset in the socket buffer chain.
		 */
		mb = sbsndptr(&so->so_snd, off, len, &moff);

		if (len <= MHLEN - hdrlen - max_linkhdr) {
			m_copydata(mb, moff, (int)len,
			    mtod(m, caddr_t) + hdrlen);
			m->m_len += len;
		} else {
			m->m_next = m_copy(mb, moff, (int)len);
			if (m->m_next == NULL) {
				SOCKBUF_UNLOCK(&so->so_snd);
				(void) m_free(m);
				error = ENOBUFS;
				goto out;
			}
		}

		/*
		 * If we're sending everything we've got, set PUSH.
		 * (This will keep happy those implementations which only
		 * give data to the user when a buffer fills or
		 * a PUSH comes in.)
		 */
		if (off + len == so->so_snd.sb_cc)
			flags |= TH_PUSH;
		SOCKBUF_UNLOCK(&so->so_snd);
	} else {
		SOCKBUF_UNLOCK(&so->so_snd);
		if (tp->t_flags & TF_ACKNOW)
			TCPSTAT_INC(tcps_sndacks);
		else if (flags & (TH_SYN|TH_FIN|TH_RST))
			TCPSTAT_INC(tcps_sndctrl);
		else if (SEQ_GT(tp->snd_up, tp->snd_una))
			TCPSTAT_INC(tcps_sndurg);
		else
			TCPSTAT_INC(tcps_sndwinup);

		MGETHDR(m, M_DONTWAIT, MT_DATA);
		if (m == NULL) {
			error = ENOBUFS;
			goto out;
		}
#ifdef INET6
		if (isipv6 && (MHLEN < hdrlen + max_linkhdr) &&
		    MHLEN >= hdrlen) {
			MH_ALIGN(m, hdrlen);
		} else
#endif
		m->m_data += max_linkhdr;
		m->m_len = hdrlen;
	}
	SOCKBUF_UNLOCK_ASSERT(&so->so_snd);
	m->m_pkthdr.rcvif = (struct ifnet *)0;
#ifdef MAC
	mac_inpcb_create_mbuf(tp->t_inpcb, m);
#endif
#ifdef INET6
	if (isipv6) {
		ip6 = mtod(m, struct ip6_hdr *);
		th = (struct tcphdr *)(ip6 + 1);
		tcpip_fillheaders(tp->t_inpcb, ip6, th);
	} else
#endif /* INET6 */
	{
		ip = mtod(m, struct ip *);
		ipov = (struct ipovly *)ip;
		th = (struct tcphdr *)(ip + 1);
		tcpip_fillheaders(tp->t_inpcb, ip, th);
	}

	/*
	 * Fill in fields, remembering maximum advertised
	 * window for use in delaying messages about window sizes.
	 * If resending a FIN, be sure not to use a new sequence number.
	 */
	if (flags & TH_FIN && tp->t_flags & TF_SENTFIN &&
	    tp->snd_nxt == tp->snd_max)
		tp->snd_nxt--;
	/*
	 * If we are starting a connection, send ECN setup
	 * SYN packet. If we are on a retransmit, we may
	 * resend those bits a number of times as per
	 * RFC 3168.
	 */
	if (tp->t_state == TCPS_SYN_SENT && V_tcp_do_ecn) {
		if (tp->t_rxtshift >= 1) {
			if (tp->t_rxtshift <= V_tcp_ecn_maxretries)
				flags |= TH_ECE|TH_CWR;
		} else
			flags |= TH_ECE|TH_CWR;
	}
	
	if (tp->t_state == TCPS_ESTABLISHED &&
	    (tp->t_flags & TF_ECN_PERMIT)) {
		/*
		 * If the peer has ECN, mark data packets with
		 * ECN capable transmission (ECT).
		 * Ignore pure ack packets, retransmissions and window probes.
		 */
		if (len > 0 && SEQ_GEQ(tp->snd_nxt, tp->snd_max) &&
		    !((tp->t_flags & TF_FORCEDATA) && len == 1)) {
#ifdef INET6
			if (isipv6)
				ip6->ip6_flow |= htonl(IPTOS_ECN_ECT0 << 20);
			else
#endif
				ip->ip_tos |= IPTOS_ECN_ECT0;
			TCPSTAT_INC(tcps_ecn_ect0);
		}
		
		/*
		 * Reply with proper ECN notifications.
		 */
		if (tp->t_flags & TF_ECN_SND_CWR) {
			flags |= TH_CWR;
			tp->t_flags &= ~TF_ECN_SND_CWR;
		} 
		if (tp->t_flags & TF_ECN_SND_ECE)
			flags |= TH_ECE;
	}
	
	/*
	 * If we are doing retransmissions, then snd_nxt will
	 * not reflect the first unsent octet.  For ACK only
	 * packets, we do not want the sequence number of the
	 * retransmitted packet, we want the sequence number
	 * of the next unsent octet.  So, if there is no data
	 * (and no SYN or FIN), use snd_max instead of snd_nxt
	 * when filling in ti_seq.  But if we are in persist
	 * state, snd_max might reflect one byte beyond the
	 * right edge of the window, so use snd_nxt in that
	 * case, since we know we aren't doing a retransmission.
	 * (retransmit and persist are mutually exclusive...)
	 */
	if (sack_rxmit == 0) {
		if (len || (flags & (TH_SYN|TH_FIN)) ||
		    tcp_timer_active(tp, TT_PERSIST))
			th->th_seq = htonl(tp->snd_nxt);
		else
			th->th_seq = htonl(tp->snd_max);
	} else {
		th->th_seq = htonl(p->rxmit);
		p->rxmit += len;
		tp->sackhint.sack_bytes_rexmit += len;
	}
	th->th_ack = htonl(tp->rcv_nxt);
	if (optlen) {
		bcopy(opt, th + 1, optlen);
		th->th_off = (sizeof (struct tcphdr) + optlen) >> 2;
	}
	th->th_flags = flags;
	/*
	 * Calculate receive window.  Don't shrink window,
	 * but avoid silly window syndrome.
	 */
	if (recwin < (long)(so->so_rcv.sb_hiwat / 4) &&
	    recwin < (long)tp->t_maxseg)
		recwin = 0;
	if (SEQ_GT(tp->rcv_adv, tp->rcv_nxt) &&
	    recwin < (long)(tp->rcv_adv - tp->rcv_nxt))
		recwin = (long)(tp->rcv_adv - tp->rcv_nxt);
	if (recwin > (long)TCP_MAXWIN << tp->rcv_scale)
		recwin = (long)TCP_MAXWIN << tp->rcv_scale;

	/*
	 * According to RFC1323 the window field in a SYN (i.e., a <SYN>
	 * or <SYN,ACK>) segment itself is never scaled.  The <SYN,ACK>
	 * case is handled in syncache.
	 */
	if (flags & TH_SYN)
		th->th_win = htons((u_short)
				(min(sbspace(&so->so_rcv), TCP_MAXWIN)));
	else
		th->th_win = htons((u_short)(recwin >> tp->rcv_scale));

	/*
	 * Adjust the RXWIN0SENT flag - indicate that we have advertised
	 * a 0 window.  This may cause the remote transmitter to stall.  This
	 * flag tells soreceive() to disable delayed acknowledgements when
	 * draining the buffer.  This can occur if the receiver is attempting
	 * to read more data than can be buffered prior to transmitting on
	 * the connection.
	 */
	if (th->th_win == 0) {
		tp->t_sndzerowin++;
		tp->t_flags |= TF_RXWIN0SENT;
	} else
		tp->t_flags &= ~TF_RXWIN0SENT;
	if (SEQ_GT(tp->snd_up, tp->snd_nxt)) {
		th->th_urp = htons((u_short)(tp->snd_up - tp->snd_nxt));
		th->th_flags |= TH_URG;
	} else
		/*
		 * If no urgent pointer to send, then we pull
		 * the urgent pointer to the left edge of the send window
		 * so that it doesn't drift into the send window on sequence
		 * number wraparound.
		 */
		tp->snd_up = tp->snd_una;		/* drag it along */

#ifdef TCP_SIGNATURE
	if (tp->t_flags & TF_SIGNATURE) {
		int sigoff = to.to_signature - opt;
		tcp_signature_compute(m, 0, len, optlen,
		    (u_char *)(th + 1) + sigoff, IPSEC_DIR_OUTBOUND);
	}
#endif

	/*
	 * Put TCP length in extended header, and then
	 * checksum extended header and data.
	 */
	m->m_pkthdr.len = hdrlen + len; /* in6_cksum() need this */
#ifdef INET6
	if (isipv6)
		/*
		 * ip6_plen is not need to be filled now, and will be filled
		 * in ip6_output.
		 */
		th->th_sum = in6_cksum(m, IPPROTO_TCP, sizeof(struct ip6_hdr),
				       sizeof(struct tcphdr) + optlen + len);
	else
#endif /* INET6 */
	{
		m->m_pkthdr.csum_flags = CSUM_TCP;
		m->m_pkthdr.csum_data = offsetof(struct tcphdr, th_sum);
		th->th_sum = in_pseudo(ip->ip_src.s_addr, ip->ip_dst.s_addr,
		    htons(sizeof(struct tcphdr) + IPPROTO_TCP + len + optlen));

		/* IP version must be set here for ipv4/ipv6 checking later */
		KASSERT(ip->ip_v == IPVERSION,
		    ("%s: IP version incorrect: %d", __func__, ip->ip_v));
	}

	/*
	 * Enable TSO and specify the size of the segments.
	 * The TCP pseudo header checksum is always provided.
	 * XXX: Fixme: This is currently not the case for IPv6.
	 */
	if (tso) {
		KASSERT(len > tp->t_maxopd - optlen,
		    ("%s: len <= tso_segsz", __func__));
		m->m_pkthdr.csum_flags |= CSUM_TSO;
		m->m_pkthdr.tso_segsz = tp->t_maxopd - optlen;
	}

#ifdef IPSEC
	KASSERT(len + hdrlen + ipoptlen - ipsec_optlen == m_length(m, NULL),
	    ("%s: mbuf chain shorter than expected: %ld + %u + %u - %u != %u",
	    __func__, len, hdrlen, ipoptlen, ipsec_optlen, m_length(m, NULL)));
#else
	KASSERT(len + hdrlen + ipoptlen == m_length(m, NULL),
	    ("%s: mbuf chain shorter than expected: %ld + %u + %u != %u",
	    __func__, len, hdrlen, ipoptlen, m_length(m, NULL)));
#endif

	/*
	 * In transmit state, time the transmission and arrange for
	 * the retransmit.  In persist state, just set snd_max.
	 */
	if ((tp->t_flags & TF_FORCEDATA) == 0 || 
	    !tcp_timer_active(tp, TT_PERSIST)) {
		tcp_seq startseq = tp->snd_nxt;

		/*
		 * Advance snd_nxt over sequence space of this segment.
		 */
		if (flags & (TH_SYN|TH_FIN)) {
			if (flags & TH_SYN)
				tp->snd_nxt++;
			if (flags & TH_FIN) {
				tp->snd_nxt++;
				tp->t_flags |= TF_SENTFIN;
			}
		}
		if (sack_rxmit)
			goto timer;
		tp->snd_nxt += len;
		if (SEQ_GT(tp->snd_nxt, tp->snd_max)) {
			tp->snd_max = tp->snd_nxt;
			/*
			 * Time this transmission if not a retransmission and
			 * not currently timing anything.
			 */
			if (tp->t_rtttime == 0) {
				tp->t_rtttime = ticks;
				tp->t_rtseq = startseq;
				TCPSTAT_INC(tcps_segstimed);
			}
		}

		/*
		 * Set retransmit timer if not currently set,
		 * and not doing a pure ack or a keep-alive probe.
		 * Initial value for retransmit timer is smoothed
		 * round-trip time + 2 * round-trip time variance.
		 * Initialize shift counter which is used for backoff
		 * of retransmit time.
		 */
timer:
		if (!tcp_timer_active(tp, TT_REXMT) &&
		    ((sack_rxmit && tp->snd_nxt != tp->snd_max) ||
		     (tp->snd_nxt != tp->snd_una))) {
			if (tcp_timer_active(tp, TT_PERSIST)) {
				tcp_timer_activate(tp, TT_PERSIST, 0);
				tp->t_rxtshift = 0;
			}
			tcp_timer_activate(tp, TT_REXMT, tp->t_rxtcur);
		}
	} else {
		/*
		 * Persist case, update snd_max but since we are in
		 * persist mode (no window) we do not update snd_nxt.
		 */
		int xlen = len;
		if (flags & TH_SYN)
			++xlen;
		if (flags & TH_FIN) {
			++xlen;
			tp->t_flags |= TF_SENTFIN;
		}
		if (SEQ_GT(tp->snd_nxt + xlen, tp->snd_max))
			tp->snd_max = tp->snd_nxt + len;
	}

	/* Run HHOOK_TCP_ESTABLISHED_OUT helper hooks. */
	hhook_run_tcp_est_out(tp, th, &to, len, tso);

#ifdef TCPDEBUG
	/*
	 * Trace.
	 */
	if (so->so_options & SO_DEBUG) {
		u_short save = 0;
#ifdef INET6
		if (!isipv6)
#endif
		{
			save = ipov->ih_len;
			ipov->ih_len = htons(m->m_pkthdr.len /* - hdrlen + (th->th_off << 2) */);
		}
		tcp_trace(TA_OUTPUT, tp->t_state, tp, mtod(m, void *), th, 0);
#ifdef INET6
		if (!isipv6)
#endif
		ipov->ih_len = save;
	}
#endif /* TCPDEBUG */

	/*
	 * Fill in IP length and desired time to live and
	 * send to IP level.  There should be a better way
	 * to handle ttl and tos; we could keep them in
	 * the template, but need a way to checksum without them.
	 */
	/*
	 * m->m_pkthdr.len should have been set before cksum calcuration,
	 * because in6_cksum() need it.
	 */
#ifdef INET6
	if (isipv6) {
		/*
		 * we separately set hoplimit for every segment, since the
		 * user might want to change the value via setsockopt.
		 * Also, desired default hop limit might be changed via
		 * Neighbor Discovery.
		 */
		ip6->ip6_hlim = in6_selecthlim(tp->t_inpcb, NULL);

		/* TODO: IPv6 IP6TOS_ECT bit on */
		error = ip6_output(m,
			    tp->t_inpcb->in6p_outputopts, NULL,
			    ((so->so_options & SO_DONTROUTE) ?
			    IP_ROUTETOIF : 0), NULL, NULL, tp->t_inpcb);
	}
#endif /* INET6 */
#if defined(INET) && defined(INET6)
	else
#endif
#ifdef INET
    {
	ip->ip_len = m->m_pkthdr.len;
#ifdef INET6
	if (tp->t_inpcb->inp_vflag & INP_IPV6PROTO)
		ip->ip_ttl = in6_selecthlim(tp->t_inpcb, NULL);
#endif /* INET6 */
	/*
	 * If we do path MTU discovery, then we set DF on every packet.
	 * This might not be the best thing to do according to RFC3390
	 * Section 2. However the tcp hostcache migitates the problem
	 * so it affects only the first tcp connection with a host.
	 *
	 * NB: Don't set DF on small MTU/MSS to have a safe fallback.
	 */
	if (V_path_mtu_discovery && tp->t_maxopd > V_tcp_minmss)
		ip->ip_off |= IP_DF;

	error = ip_output(m, tp->t_inpcb->inp_options, NULL,
	    ((so->so_options & SO_DONTROUTE) ? IP_ROUTETOIF : 0), 0,
	    tp->t_inpcb);
    }
#endif /* INET */
	if (error) {

		/*
		 * We know that the packet was lost, so back out the
		 * sequence number advance, if any.
		 *
		 * If the error is EPERM the packet got blocked by the
		 * local firewall.  Normally we should terminate the
		 * connection but the blocking may have been spurious
		 * due to a firewall reconfiguration cycle.  So we treat
		 * it like a packet loss and let the retransmit timer and
		 * timeouts do their work over time.
		 * XXX: It is a POLA question whether calling tcp_drop right
		 * away would be the really correct behavior instead.
		 */
		if (((tp->t_flags & TF_FORCEDATA) == 0 ||
		    !tcp_timer_active(tp, TT_PERSIST)) &&
		    ((flags & TH_SYN) == 0) &&
		    (error != EPERM)) {
			if (sack_rxmit) {
				p->rxmit -= len;
				tp->sackhint.sack_bytes_rexmit -= len;
				KASSERT(tp->sackhint.sack_bytes_rexmit >= 0,
				    ("sackhint bytes rtx >= 0"));
			} else
				tp->snd_nxt -= len;
		}
out:
		SOCKBUF_UNLOCK_ASSERT(&so->so_snd);	/* Check gotos. */
		switch (error) {
		case EPERM:
			tp->t_softerror = error;
			return (error);
		case ENOBUFS:
	                if (!tcp_timer_active(tp, TT_REXMT) &&
			    !tcp_timer_active(tp, TT_PERSIST))
	                        tcp_timer_activate(tp, TT_REXMT, tp->t_rxtcur);
			tp->snd_cwnd = tp->t_maxseg;
			return (0);
		case EMSGSIZE:
			/*
			 * For some reason the interface we used initially
			 * to send segments changed to another or lowered
			 * its MTU.
			 *
			 * tcp_mtudisc() will find out the new MTU and as
			 * its last action, initiate retransmission, so it
			 * is important to not do so here.
			 *
			 * If TSO was active we either got an interface
			 * without TSO capabilits or TSO was turned off.
			 * Disable it for this connection as too and
			 * immediatly retry with MSS sized segments generated
			 * by this function.
			 */
			if (tso)
				tp->t_flags &= ~TF_TSO;
			tcp_mtudisc(tp->t_inpcb, 0);
			return (0);
		case EHOSTDOWN:
		case EHOSTUNREACH:
		case ENETDOWN:
		case ENETUNREACH:
			if (TCPS_HAVERCVDSYN(tp->t_state)) {
				tp->t_softerror = error;
				return (0);
			}
			/* FALLTHROUGH */
		default:
			return (error);
		}
	}
	TCPSTAT_INC(tcps_sndtotal);

	/*
	 * Data sent (as far as we can tell).
	 * If this advertises a larger window than any other segment,
	 * then remember the size of the advertised window.
	 * Any pending ACK has now been sent.
	 */
	if (recwin >= 0 && SEQ_GT(tp->rcv_nxt + recwin, tp->rcv_adv))
		tp->rcv_adv = tp->rcv_nxt + recwin;
	tp->last_ack_sent = tp->rcv_nxt;
	tp->t_flags &= ~(TF_ACKNOW | TF_DELACK);
	if (tcp_timer_active(tp, TT_DELACK))
		tcp_timer_activate(tp, TT_DELACK, 0);
#if 0
	/*
	 * This completely breaks TCP if newreno is turned on.  What happens
	 * is that if delayed-acks are turned on on the receiver, this code
	 * on the transmitter effectively destroys the TCP window, forcing
	 * it to four packets (1.5Kx4 = 6K window).
	 */
	if (sendalot && --maxburst)
		goto again;
#endif
	if (sendalot)
		goto again;
	return (0);
}

void
tcp_setpersist(struct tcpcb *tp)
{
	int t = ((tp->t_srtt >> 2) + tp->t_rttvar) >> 1;
	int tt;

	tp->t_flags &= ~TF_PREVVALID;
	if (tcp_timer_active(tp, TT_REXMT))
		panic("tcp_setpersist: retransmit pending");
	/*
	 * Start/restart persistance timer.
	 */
	TCPT_RANGESET(tt, t * tcp_backoff[tp->t_rxtshift],
		      TCPTV_PERSMIN, TCPTV_PERSMAX);
	tcp_timer_activate(tp, TT_PERSIST, tt);
	if (tp->t_rxtshift < TCP_MAXRXTSHIFT)
		tp->t_rxtshift++;
}

/*
 * Insert TCP options according to the supplied parameters to the place
 * optp in a consistent way.  Can handle unaligned destinations.
 *
 * The order of the option processing is crucial for optimal packing and
 * alignment for the scarce option space.
 *
 * The optimal order for a SYN/SYN-ACK segment is:
 *   MSS (4) + NOP (1) + Window scale (3) + SACK permitted (2) +
 *   Timestamp (10) + Signature (18) = 38 bytes out of a maximum of 40.
 *
 * The SACK options should be last.  SACK blocks consume 8*n+2 bytes.
 * So a full size SACK blocks option is 34 bytes (with 4 SACK blocks).
 * At minimum we need 10 bytes (to generate 1 SACK block).  If both
 * TCP Timestamps (12 bytes) and TCP Signatures (18 bytes) are present,
 * we only have 10 bytes for SACK options (40 - (12 + 18)).
 */
int
tcp_addoptions(struct tcpopt *to, u_char *optp)
{
	u_int mask, optlen = 0;

	for (mask = 1; mask < TOF_MAXOPT; mask <<= 1) {
		if ((to->to_flags & mask) != mask)
			continue;
		if (optlen == TCP_MAXOLEN)
			break;
		switch (to->to_flags & mask) {
		case TOF_MSS:
			while (optlen % 4) {
				optlen += TCPOLEN_NOP;
				*optp++ = TCPOPT_NOP;
			}
			if (TCP_MAXOLEN - optlen < TCPOLEN_MAXSEG)
				continue;
			optlen += TCPOLEN_MAXSEG;
			*optp++ = TCPOPT_MAXSEG;
			*optp++ = TCPOLEN_MAXSEG;
			to->to_mss = htons(to->to_mss);
			bcopy((u_char *)&to->to_mss, optp, sizeof(to->to_mss));
			optp += sizeof(to->to_mss);
			break;
		case TOF_SCALE:
			while (!optlen || optlen % 2 != 1) {
				optlen += TCPOLEN_NOP;
				*optp++ = TCPOPT_NOP;
			}
			if (TCP_MAXOLEN - optlen < TCPOLEN_WINDOW)
				continue;
			optlen += TCPOLEN_WINDOW;
			*optp++ = TCPOPT_WINDOW;
			*optp++ = TCPOLEN_WINDOW;
			*optp++ = to->to_wscale;
			break;
		case TOF_SACKPERM:
			while (optlen % 2) {
				optlen += TCPOLEN_NOP;
				*optp++ = TCPOPT_NOP;
			}
			if (TCP_MAXOLEN - optlen < TCPOLEN_SACK_PERMITTED)
				continue;
			optlen += TCPOLEN_SACK_PERMITTED;
			*optp++ = TCPOPT_SACK_PERMITTED;
			*optp++ = TCPOLEN_SACK_PERMITTED;
			break;
		case TOF_TS:
			while (!optlen || optlen % 4 != 2) {
				optlen += TCPOLEN_NOP;
				*optp++ = TCPOPT_NOP;
			}
			if (TCP_MAXOLEN - optlen < TCPOLEN_TIMESTAMP)
				continue;
			optlen += TCPOLEN_TIMESTAMP;
			*optp++ = TCPOPT_TIMESTAMP;
			*optp++ = TCPOLEN_TIMESTAMP;
			to->to_tsval = htonl(to->to_tsval);
			to->to_tsecr = htonl(to->to_tsecr);
			bcopy((u_char *)&to->to_tsval, optp, sizeof(to->to_tsval));
			optp += sizeof(to->to_tsval);
			bcopy((u_char *)&to->to_tsecr, optp, sizeof(to->to_tsecr));
			optp += sizeof(to->to_tsecr);
			break;
		case TOF_SIGNATURE:
			{
			int siglen = TCPOLEN_SIGNATURE - 2;

			while (!optlen || optlen % 4 != 2) {
				optlen += TCPOLEN_NOP;
				*optp++ = TCPOPT_NOP;
			}
			if (TCP_MAXOLEN - optlen < TCPOLEN_SIGNATURE)
				continue;
			optlen += TCPOLEN_SIGNATURE;
			*optp++ = TCPOPT_SIGNATURE;
			*optp++ = TCPOLEN_SIGNATURE;
			to->to_signature = optp;
			while (siglen--)
				 *optp++ = 0;
			break;
			}
		case TOF_SACK:
			{
			int sackblks = 0;
			struct sackblk *sack = (struct sackblk *)to->to_sacks;
			tcp_seq sack_seq;

			while (!optlen || optlen % 4 != 2) {
				optlen += TCPOLEN_NOP;
				*optp++ = TCPOPT_NOP;
			}
			if (TCP_MAXOLEN - optlen < TCPOLEN_SACKHDR + TCPOLEN_SACK)
				continue;
			optlen += TCPOLEN_SACKHDR;
			*optp++ = TCPOPT_SACK;
			sackblks = min(to->to_nsacks,
					(TCP_MAXOLEN - optlen) / TCPOLEN_SACK);
			*optp++ = TCPOLEN_SACKHDR + sackblks * TCPOLEN_SACK;
			while (sackblks--) {
				sack_seq = htonl(sack->start);
				bcopy((u_char *)&sack_seq, optp, sizeof(sack_seq));
				optp += sizeof(sack_seq);
				sack_seq = htonl(sack->end);
				bcopy((u_char *)&sack_seq, optp, sizeof(sack_seq));
				optp += sizeof(sack_seq);
				optlen += TCPOLEN_SACK;
				sack++;
			}
			TCPSTAT_INC(tcps_sack_send_blocks);
			break;
			}
		default:
			panic("%s: unknown TCP option type", __func__);
			break;
		}
	}

	/* Terminate and pad TCP options to a 4 byte boundary. */
	if (optlen % 4) {
		optlen += TCPOLEN_EOL;
		*optp++ = TCPOPT_EOL;
	}
	/*
	 * According to RFC 793 (STD0007):
	 *   "The content of the header beyond the End-of-Option option
	 *    must be header padding (i.e., zero)."
	 *   and later: "The padding is composed of zeros."
	 */
	while (optlen % 4) {
		optlen += TCPOLEN_PAD;
		*optp++ = TCPOPT_PAD;
	}

	KASSERT(optlen <= TCP_MAXOLEN, ("%s: TCP options too long", __func__));
	return (optlen);
}
OpenPOWER on IntegriCloud