summaryrefslogtreecommitdiff
path: root/ipc/mach_msg.c
blob: 78247a4de068f1148633097387ec7213e99dcdab (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
/*
 * Mach Operating System
 * Copyright (c) 1991,1990,1989 Carnegie Mellon University.
 * Copyright (c) 1993,1994 The University of Utah and
 * the Computer Systems Laboratory (CSL).
 * All rights reserved.
 *
 * Permission to use, copy, modify and distribute this software and its
 * documentation is hereby granted, provided that both the copyright
 * notice and this permission notice appear in all copies of the
 * software, derivative works or modified versions, and any portions
 * thereof, and that both notices appear in supporting documentation.
 *
 * CARNEGIE MELLON, THE UNIVERSITY OF UTAH AND CSL ALLOW FREE USE OF
 * THIS SOFTWARE IN ITS "AS IS" CONDITION, AND DISCLAIM ANY LIABILITY
 * OF ANY KIND FOR ANY DAMAGES WHATSOEVER RESULTING FROM THE USE OF
 * THIS SOFTWARE.
 *
 * Carnegie Mellon requests users of this software to return to
 *
 *  Software Distribution Coordinator  or  Software.Distribution@CS.CMU.EDU
 *  School of Computer Science
 *  Carnegie Mellon University
 *  Pittsburgh PA 15213-3890
 *
 * any improvements or extensions that they make and grant Carnegie Mellon
 * the rights to redistribute these changes.
 */
/*
 */
/*
 *	File:	ipc/mach_msg.c
 *	Author:	Rich Draves
 *	Date:	1989
 *
 *	Exported message traps.  See mach/message.h.
 */

#include <mach/kern_return.h>
#include <mach/port.h>
#include <mach/message.h>
#include <kern/assert.h>
#include <kern/counters.h>
#include <kern/debug.h>
#include <kern/lock.h>
#include <kern/printf.h>
#include <kern/sched_prim.h>
#include <kern/ipc_sched.h>
#include <kern/exception.h>
#include <vm/vm_map.h>
#include <ipc/ipc_kmsg.h>
#include <ipc/ipc_marequest.h>
#include <ipc/ipc_mqueue.h>
#include <ipc/ipc_object.h>
#include <ipc/ipc_notify.h>
#include <ipc/ipc_port.h>
#include <ipc/ipc_pset.h>
#include <ipc/ipc_space.h>
#include <ipc/ipc_thread.h>
#include <ipc/ipc_entry.h>
#include <ipc/mach_msg.h>
#include <machine/locore.h>
#include <machine/pcb.h>

/*
 *	Routine:	mach_msg_send
 *	Purpose:
 *		Send a message.
 *	Conditions:
 *		Nothing locked.
 *	Returns:
 *		MACH_MSG_SUCCESS	Sent the message.
 *		MACH_SEND_MSG_TOO_SMALL	Message smaller than a header.
 *		MACH_SEND_NO_BUFFER	Couldn't allocate buffer.
 *		MACH_SEND_INVALID_DATA	Couldn't copy message data.
 *		MACH_SEND_INVALID_HEADER
 *			Illegal value in the message header bits.
 *		MACH_SEND_INVALID_DEST	The space is dead.
 *		MACH_SEND_INVALID_NOTIFY	Bad notify port.
 *		MACH_SEND_INVALID_DEST	Can't copyin destination port.
 *		MACH_SEND_INVALID_REPLY	Can't copyin reply port.
 *		MACH_SEND_TIMED_OUT	Timeout expired without delivery.
 *		MACH_SEND_INTERRUPTED	Delivery interrupted.
 *		MACH_SEND_NO_NOTIFY	Can't allocate a msg-accepted request.
 *		MACH_SEND_WILL_NOTIFY	Msg-accepted notif. requested.
 *		MACH_SEND_NOTIFY_IN_PROGRESS
 *			This space has already forced a message to this port.
 */

mach_msg_return_t
mach_msg_send(msg, option, send_size, time_out, notify)
	mach_msg_header_t *msg;
	mach_msg_option_t option;
	mach_msg_size_t send_size;
	mach_msg_timeout_t time_out;
	mach_port_t notify;
{
	ipc_space_t space = current_space();
	vm_map_t map = current_map();
	ipc_kmsg_t kmsg;
	mach_msg_return_t mr;

	mr = ipc_kmsg_get(msg, send_size, &kmsg);
	if (mr != MACH_MSG_SUCCESS)
		return mr;

	if (option & MACH_SEND_CANCEL) {
		if (notify == MACH_PORT_NULL)
			mr = MACH_SEND_INVALID_NOTIFY;
		else
			mr = ipc_kmsg_copyin(kmsg, space, map, notify);
	} else
		mr = ipc_kmsg_copyin(kmsg, space, map, MACH_PORT_NULL);
	if (mr != MACH_MSG_SUCCESS) {
		ikm_free(kmsg);
		return mr;
	}

	if (option & MACH_SEND_NOTIFY) {
		mr = ipc_mqueue_send(kmsg, MACH_SEND_TIMEOUT,
				     ((option & MACH_SEND_TIMEOUT) ?
				      time_out : MACH_MSG_TIMEOUT_NONE));
		if (mr == MACH_SEND_TIMED_OUT) {
			ipc_port_t dest = (ipc_port_t)
				kmsg->ikm_header.msgh_remote_port;

			if (notify == MACH_PORT_NULL)
				mr = MACH_SEND_INVALID_NOTIFY;
			else
				mr = ipc_marequest_create(space, dest,
						notify, &kmsg->ikm_marequest);
			if (mr == MACH_MSG_SUCCESS) {
				ipc_mqueue_send_always(kmsg);
				return MACH_SEND_WILL_NOTIFY;
			}
		}
	} else
		mr = ipc_mqueue_send(kmsg, option & MACH_SEND_TIMEOUT,
				     time_out);

	if (mr != MACH_MSG_SUCCESS) {
		mr |= ipc_kmsg_copyout_pseudo(kmsg, space, map);

		assert(kmsg->ikm_marequest == IMAR_NULL);
		(void) ipc_kmsg_put(msg, kmsg, kmsg->ikm_header.msgh_size);
	}

	return mr;
}

/*
 *	Routine:	mach_msg_receive
 *	Purpose:
 *		Receive a message.
 *	Conditions:
 *		Nothing locked.
 *	Returns:
 *		MACH_MSG_SUCCESS	Received a message.
 *		MACH_RCV_INVALID_NAME	The name doesn't denote a right,
 *			or the denoted right is not receive or port set.
 *		MACH_RCV_IN_SET		Receive right is a member of a set.
 *		MACH_RCV_TOO_LARGE	Message wouldn't fit into buffer.
 *		MACH_RCV_TIMED_OUT	Timeout expired without a message.
 *		MACH_RCV_INTERRUPTED	Reception interrupted.
 *		MACH_RCV_PORT_DIED	Port/set died while receiving.
 *		MACH_RCV_PORT_CHANGED	Port moved into set while receiving.
 *		MACH_RCV_INVALID_DATA	Couldn't copy to user buffer.
 *		MACH_RCV_INVALID_NOTIFY	Bad notify port.
 *		MACH_RCV_HEADER_ERROR
 */

mach_msg_return_t
mach_msg_receive(msg, option, rcv_size, rcv_name, time_out, notify)
	mach_msg_header_t *msg;
	mach_msg_option_t option;
	mach_msg_size_t rcv_size;
	mach_port_t rcv_name;
	mach_msg_timeout_t time_out;
	mach_port_t notify;
{
	ipc_thread_t self = current_thread();
	ipc_space_t space = current_space();
	vm_map_t map = current_map();
	ipc_object_t object;
	ipc_mqueue_t mqueue;
	ipc_kmsg_t kmsg;
	mach_port_seqno_t seqno;
	mach_msg_return_t mr;

	mr = ipc_mqueue_copyin(space, rcv_name, &mqueue, &object);
	if (mr != MACH_MSG_SUCCESS)
		return mr;
	/* hold ref for object; mqueue is locked */

	/*
	 *	ipc_mqueue_receive may not return, because if we block
	 *	then our kernel stack may be discarded.  So we save
	 *	state here for mach_msg_receive_continue to pick up.
	 */

	self->ith_msg = msg;
	self->ith_option = option;
	self->ith_rcv_size = rcv_size;
	self->ith_timeout = time_out;
	self->ith_notify = notify;
	self->ith_object = object;
	self->ith_mqueue = mqueue;

	if (option & MACH_RCV_LARGE) {
		mr = ipc_mqueue_receive(mqueue, option & MACH_RCV_TIMEOUT,
					rcv_size, time_out,
					FALSE, mach_msg_receive_continue,
					&kmsg, &seqno);
		/* mqueue is unlocked */
		ipc_object_release(object);
		if (mr != MACH_MSG_SUCCESS) {
			if (mr == MACH_RCV_TOO_LARGE) {
				mach_msg_size_t real_size =
					(mach_msg_size_t) (vm_offset_t) kmsg;

				assert(real_size > rcv_size);

				(void) copyout(&real_size,
					       &msg->msgh_size,
					       sizeof(mach_msg_size_t));
			}

			return mr;
		}

		kmsg->ikm_header.msgh_seqno = seqno;
		assert(kmsg->ikm_header.msgh_size <= rcv_size);
	} else {
		mr = ipc_mqueue_receive(mqueue, option & MACH_RCV_TIMEOUT,
					MACH_MSG_SIZE_MAX, time_out,
					FALSE, mach_msg_receive_continue,
					&kmsg, &seqno);
		/* mqueue is unlocked */
		ipc_object_release(object);
		if (mr != MACH_MSG_SUCCESS)
			return mr;

		kmsg->ikm_header.msgh_seqno = seqno;
		if (kmsg->ikm_header.msgh_size > rcv_size) {
			ipc_kmsg_copyout_dest(kmsg, space);
			(void) ipc_kmsg_put(msg, kmsg, sizeof *msg);
			return MACH_RCV_TOO_LARGE;
		}
	}

	if (option & MACH_RCV_NOTIFY) {
		if (notify == MACH_PORT_NULL)
			mr = MACH_RCV_INVALID_NOTIFY;
		else
			mr = ipc_kmsg_copyout(kmsg, space, map, notify);
	} else
		mr = ipc_kmsg_copyout(kmsg, space, map, MACH_PORT_NULL);
	if (mr != MACH_MSG_SUCCESS) {
		if ((mr &~ MACH_MSG_MASK) == MACH_RCV_BODY_ERROR) {
			(void) ipc_kmsg_put(msg, kmsg,
					    kmsg->ikm_header.msgh_size);
		} else {
			ipc_kmsg_copyout_dest(kmsg, space);
			(void) ipc_kmsg_put(msg, kmsg, sizeof *msg);
		}

		return mr;
	}

	return ipc_kmsg_put(msg, kmsg, kmsg->ikm_header.msgh_size);
}

/*
 *	Routine:	mach_msg_receive_continue
 *	Purpose:
 *		Continue after blocking for a message.
 *	Conditions:
 *		Nothing locked.  We are running on a new kernel stack,
 *		with the receive state saved in the thread.  From here
 *		control goes back to user space.
 */

void
mach_msg_receive_continue(void)
{
	ipc_thread_t self = current_thread();
	ipc_space_t space = current_space();
	vm_map_t map = current_map();
	mach_msg_header_t *msg = self->ith_msg;
	mach_msg_option_t option = self->ith_option;
	mach_msg_size_t rcv_size = self->ith_rcv_size;
	mach_msg_timeout_t time_out = self->ith_timeout;
	mach_port_t notify = self->ith_notify;
	ipc_object_t object = self->ith_object;
	ipc_mqueue_t mqueue = self->ith_mqueue;
	ipc_kmsg_t kmsg;
	mach_port_seqno_t seqno;
	mach_msg_return_t mr;

	if (option & MACH_RCV_LARGE) {
		mr = ipc_mqueue_receive(mqueue, option & MACH_RCV_TIMEOUT,
					rcv_size, time_out,
					TRUE, mach_msg_receive_continue,
					&kmsg, &seqno);
		/* mqueue is unlocked */
		ipc_object_release(object);
		if (mr != MACH_MSG_SUCCESS) {
			if (mr == MACH_RCV_TOO_LARGE) {
				mach_msg_size_t real_size =
					(mach_msg_size_t) (vm_offset_t) kmsg;

				assert(real_size > rcv_size);

				(void) copyout(&real_size,
					       &msg->msgh_size,
					       sizeof(mach_msg_size_t));
			}

			thread_syscall_return(mr);
			/*NOTREACHED*/
		}

		kmsg->ikm_header.msgh_seqno = seqno;
		assert(kmsg->ikm_header.msgh_size <= rcv_size);
	} else {
		mr = ipc_mqueue_receive(mqueue, option & MACH_RCV_TIMEOUT,
					MACH_MSG_SIZE_MAX, time_out,
					TRUE, mach_msg_receive_continue,
					&kmsg, &seqno);
		/* mqueue is unlocked */
		ipc_object_release(object);
		if (mr != MACH_MSG_SUCCESS) {
			thread_syscall_return(mr);
			/*NOTREACHED*/
		}

		kmsg->ikm_header.msgh_seqno = seqno;
		if (kmsg->ikm_header.msgh_size > rcv_size) {
			ipc_kmsg_copyout_dest(kmsg, space);
			(void) ipc_kmsg_put(msg, kmsg, sizeof *msg);
			thread_syscall_return(MACH_RCV_TOO_LARGE);
			/*NOTREACHED*/
		}
	}

	if (option & MACH_RCV_NOTIFY) {
		if (notify == MACH_PORT_NULL)
			mr = MACH_RCV_INVALID_NOTIFY;
		else
			mr = ipc_kmsg_copyout(kmsg, space, map, notify);
	} else
		mr = ipc_kmsg_copyout(kmsg, space, map, MACH_PORT_NULL);
	if (mr != MACH_MSG_SUCCESS) {
		if ((mr &~ MACH_MSG_MASK) == MACH_RCV_BODY_ERROR) {
			(void) ipc_kmsg_put(msg, kmsg,
					    kmsg->ikm_header.msgh_size);
		} else {
			ipc_kmsg_copyout_dest(kmsg, space);
			(void) ipc_kmsg_put(msg, kmsg, sizeof *msg);
		}

		thread_syscall_return(mr);
		/*NOTREACHED*/
	}

	mr = ipc_kmsg_put(msg, kmsg, kmsg->ikm_header.msgh_size);
	thread_syscall_return(mr);
	/*NOTREACHED*/
}

/*
 *	Routine:	mach_msg_trap [mach trap]
 *	Purpose:
 *		Possibly send a message; possibly receive a message.
 *	Conditions:
 *		Nothing locked.
 *	Returns:
 *		All of mach_msg_send and mach_msg_receive error codes.
 */

mach_msg_return_t
mach_msg_trap(msg, option, send_size, rcv_size, rcv_name, time_out, notify)
	mach_msg_header_t *msg;
	mach_msg_option_t option;
	mach_msg_size_t send_size;
	mach_msg_size_t rcv_size;
	mach_port_t rcv_name;
	mach_msg_timeout_t time_out;
	mach_port_t notify;
{
	mach_msg_return_t mr;

	/* first check for common cases */

	if (option == (MACH_SEND_MSG|MACH_RCV_MSG)) {
		ipc_thread_t self = current_thread();
		ipc_space_t space = self->task->itk_space;
		ipc_kmsg_t kmsg;
		ipc_port_t dest_port;
		ipc_object_t rcv_object;
		ipc_mqueue_t rcv_mqueue;
		mach_msg_size_t reply_size;

		/*
		 *	This case is divided into ten sections, each
		 *	with a label.  There are five optimized
		 *	sections and six unoptimized sections, which
		 *	do the same thing but handle all possible
		 *	cases and are slower.
		 *
		 *	The five sections for an RPC are
		 *	    1) Get request message into a buffer.
		 *		(fast_get or slow_get)
		 *	    2) Copyin request message and rcv_name.
		 *		(fast_copyin or slow_copyin)
		 *	    3) Enqueue request and dequeue reply.
		 *		(fast_send_receive or
		 *		 slow_send and slow_receive)
		 *	    4) Copyout reply message.
		 *		(fast_copyout or slow_copyout)
		 *	    5) Put reply message to user's buffer.
		 *		(fast_put or slow_put)
		 *
		 *	Keep the locking hierarchy firmly in mind.
		 *	(First spaces, then ports, then port sets,
		 *	then message queues.)  Only a non-blocking
		 *	attempt can be made to acquire locks out of
		 *	order, or acquire two locks on the same level.
		 *	Acquiring two locks on the same level will
		 *	fail if the objects are really the same,
		 *	unless simple locking is disabled.  This is OK,
		 *	because then the extra unlock does nothing.
		 *
		 *	There are two major reasons these RPCs can't use
		 *	ipc_thread_switch, and use slow_send/slow_receive:
		 *		1) Kernel RPCs.
		 *		2) Servers fall behind clients, so
		 *		client doesn't find a blocked server thread and
		 *		server finds waiting messages and can't block.
		 */

	/*
	    fast_get:
	*/
		/*
		 *	optimized ipc_kmsg_get
		 *
		 *	No locks, references, or messages held.
		 *	We must clear ikm_cache before copyinmsg.
		 */

		if ((send_size > IKM_SAVED_MSG_SIZE) ||
		    (send_size < sizeof(mach_msg_header_t)) ||
		    (send_size & 3) ||
		    ((kmsg = ikm_cache()) == IKM_NULL))
			goto slow_get;

		ikm_cache() = IKM_NULL;
		ikm_check_initialized(kmsg, IKM_SAVED_KMSG_SIZE);

		if (copyinmsg(msg, &kmsg->ikm_header,
			      send_size)) {
			ikm_free(kmsg);
			goto slow_get;
		}

		kmsg->ikm_header.msgh_size = send_size;

	    fast_copyin:
		/*
		 *	optimized ipc_kmsg_copyin/ipc_mqueue_copyin
		 *
		 *	We have the request message data in kmsg.
		 *	Must still do copyin, send, receive, etc.
		 *
		 *	If the message isn't simple, we can't combine
		 *	ipc_kmsg_copyin_header and ipc_mqueue_copyin,
		 *	because copyin of the message body might
		 *	affect rcv_name.
		 */

		switch (kmsg->ikm_header.msgh_bits) {
		    case MACH_MSGH_BITS(MACH_MSG_TYPE_COPY_SEND,
					MACH_MSG_TYPE_MAKE_SEND_ONCE): {
			ipc_entry_t table;
			ipc_entry_num_t size;
			ipc_port_t reply_port;

			/* sending a request message */

		    {
			mach_port_index_t index;
			mach_port_gen_t gen;

		    {
			mach_port_t reply_name =
				kmsg->ikm_header.msgh_local_port;

			if (reply_name != rcv_name)
				goto slow_copyin;

			/* optimized ipc_entry_lookup of reply_name */

			index = MACH_PORT_INDEX(reply_name);
			gen = MACH_PORT_GEN(reply_name);
		    }

			is_read_lock(space);
			assert(space->is_active);

			size = space->is_table_size;
			table = space->is_table;

			if (index >= size)
				goto abort_request_copyin;

		    {
			ipc_entry_t entry;
			ipc_entry_bits_t bits;

			entry = &table[index];
			bits = entry->ie_bits;

			/* check generation number and type bit */

			if ((bits & (IE_BITS_GEN_MASK|
				     MACH_PORT_TYPE_RECEIVE)) !=
			    (gen | MACH_PORT_TYPE_RECEIVE))
				goto abort_request_copyin;

			reply_port = (ipc_port_t) entry->ie_object;
			assert(reply_port != IP_NULL);
		    }
		    }

			/* optimized ipc_entry_lookup of dest_name */

		    {
			mach_port_index_t index;
			mach_port_gen_t gen;

		    {
			mach_port_t dest_name =
				kmsg->ikm_header.msgh_remote_port;

			index = MACH_PORT_INDEX(dest_name);
			gen = MACH_PORT_GEN(dest_name);
		    }

			if (index >= size)
				goto abort_request_copyin;

		    {
			ipc_entry_t entry;
			ipc_entry_bits_t bits;

			entry = &table[index];
			bits = entry->ie_bits;

			/* check generation number and type bit */

			if ((bits & (IE_BITS_GEN_MASK|MACH_PORT_TYPE_SEND)) !=
			    (gen | MACH_PORT_TYPE_SEND))
				goto abort_request_copyin;

			assert(IE_BITS_UREFS(bits) > 0);

			dest_port = (ipc_port_t) entry->ie_object;
			assert(dest_port != IP_NULL);
		    }
		    }

			/*
			 *	To do an atomic copyin, need simultaneous
			 *	locks on both ports and the space.  If
			 *	dest_port == reply_port, and simple locking is
			 *	enabled, then we will abort.  Otherwise it's
			 *	OK to unlock twice.
			 */

			ip_lock(dest_port);
			if (!ip_active(dest_port) ||
			    !ip_lock_try(reply_port)) {
				ip_unlock(dest_port);
				goto abort_request_copyin;
			}
			is_read_unlock(space);

			assert(dest_port->ip_srights > 0);
			dest_port->ip_srights++;
			ip_reference(dest_port);

			assert(ip_active(reply_port));
			assert(reply_port->ip_receiver_name ==
			       kmsg->ikm_header.msgh_local_port);
			assert(reply_port->ip_receiver == space);

			reply_port->ip_sorights++;
			ip_reference(reply_port);

			kmsg->ikm_header.msgh_bits =
				MACH_MSGH_BITS(MACH_MSG_TYPE_PORT_SEND,
					       MACH_MSG_TYPE_PORT_SEND_ONCE);
			kmsg->ikm_header.msgh_remote_port =
					(mach_port_t) dest_port;
			kmsg->ikm_header.msgh_local_port =
					(mach_port_t) reply_port;

			/* make sure we can queue to the destination */

			if (dest_port->ip_receiver == ipc_space_kernel) {
				/*
				 * The kernel server has a reference to
				 * the reply port, which it hands back
				 * to us in the reply message.  We do
				 * not need to keep another reference to
				 * it.
				 */
				ip_unlock(reply_port);

				assert(ip_active(dest_port));
				ip_unlock(dest_port);
				goto kernel_send;
			}

			if (dest_port->ip_msgcount >= dest_port->ip_qlimit)
				goto abort_request_send_receive;

			/* optimized ipc_mqueue_copyin */

			if (reply_port->ip_pset != IPS_NULL)
				goto abort_request_send_receive;

			rcv_object = (ipc_object_t) reply_port;
			io_reference(rcv_object);
			rcv_mqueue = &reply_port->ip_messages;
			imq_lock(rcv_mqueue);
			io_unlock(rcv_object);
			goto fast_send_receive;

		    abort_request_copyin:
			is_read_unlock(space);
			goto slow_copyin;

		    abort_request_send_receive:
			ip_unlock(dest_port);
			ip_unlock(reply_port);
			goto slow_send;
		    }

		    case MACH_MSGH_BITS(MACH_MSG_TYPE_MOVE_SEND_ONCE, 0): {
			ipc_entry_num_t size;
			ipc_entry_t table;

			/* sending a reply message */

		    {
			mach_port_t reply_name =
				kmsg->ikm_header.msgh_local_port;

			if (reply_name != MACH_PORT_NULL)
				goto slow_copyin;
		    }

			is_write_lock(space);
			assert(space->is_active);

			/* optimized ipc_entry_lookup */

			size = space->is_table_size;
			table = space->is_table;

		    {
			ipc_entry_t entry;
			mach_port_gen_t gen;
			mach_port_index_t index;

		    {
			mach_port_t dest_name =
				kmsg->ikm_header.msgh_remote_port;

			index = MACH_PORT_INDEX(dest_name);
			gen = MACH_PORT_GEN(dest_name);
		    }

			if (index >= size)
				goto abort_reply_dest_copyin;

			entry = &table[index];

			/* check generation, collision bit, and type bit */

			if ((entry->ie_bits & (IE_BITS_GEN_MASK|
					       IE_BITS_COLLISION|
					       MACH_PORT_TYPE_SEND_ONCE)) !=
			    (gen | MACH_PORT_TYPE_SEND_ONCE))
				goto abort_reply_dest_copyin;

			/* optimized ipc_right_copyin */

			assert(IE_BITS_TYPE(entry->ie_bits) ==
						MACH_PORT_TYPE_SEND_ONCE);
			assert(IE_BITS_UREFS(entry->ie_bits) == 1);
			assert((entry->ie_bits & IE_BITS_MAREQUEST) == 0);

			if (entry->ie_request != 0)
				goto abort_reply_dest_copyin;

			dest_port = (ipc_port_t) entry->ie_object;
			assert(dest_port != IP_NULL);

			ip_lock(dest_port);
			if (!ip_active(dest_port)) {
				ip_unlock(dest_port);
				goto abort_reply_dest_copyin;
			}

			assert(dest_port->ip_sorights > 0);

			/* optimized ipc_entry_dealloc */

			entry->ie_next = table->ie_next;
			table->ie_next = index;
			entry->ie_bits = gen;
			entry->ie_object = IO_NULL;
		    }

			kmsg->ikm_header.msgh_bits =
				MACH_MSGH_BITS(MACH_MSG_TYPE_PORT_SEND_ONCE,
					       0);
			kmsg->ikm_header.msgh_remote_port =
					(mach_port_t) dest_port;

			/* make sure we can queue to the destination */

			assert(dest_port->ip_receiver != ipc_space_kernel);

			/* optimized ipc_entry_lookup/ipc_mqueue_copyin */

		    {
			ipc_entry_t entry;
			ipc_entry_bits_t bits;

		    {
			mach_port_index_t index;
			mach_port_gen_t gen;

			index = MACH_PORT_INDEX(rcv_name);
			gen = MACH_PORT_GEN(rcv_name);

			if (index >= size)
				goto abort_reply_rcv_copyin;

			entry = &table[index];
			bits = entry->ie_bits;

			/* check generation number */

			if ((bits & IE_BITS_GEN_MASK) != gen)
				goto abort_reply_rcv_copyin;
		    }

			/* check type bits; looking for receive or set */

			if (bits & MACH_PORT_TYPE_PORT_SET) {
				ipc_pset_t rcv_pset;

				rcv_pset = (ipc_pset_t) entry->ie_object;
				assert(rcv_pset != IPS_NULL);

				ips_lock(rcv_pset);
				assert(ips_active(rcv_pset));

				rcv_object = (ipc_object_t) rcv_pset;
				rcv_mqueue = &rcv_pset->ips_messages;
			} else if (bits & MACH_PORT_TYPE_RECEIVE) {
				ipc_port_t rcv_port;

				rcv_port = (ipc_port_t) entry->ie_object;
				assert(rcv_port != IP_NULL);

				if (!ip_lock_try(rcv_port))
					goto abort_reply_rcv_copyin;
				assert(ip_active(rcv_port));

				if (rcv_port->ip_pset != IPS_NULL) {
					ip_unlock(rcv_port);
					goto abort_reply_rcv_copyin;
				}

				rcv_object = (ipc_object_t) rcv_port;
				rcv_mqueue = &rcv_port->ip_messages;
			} else
				goto abort_reply_rcv_copyin;
		    }

			is_write_unlock(space);
			io_reference(rcv_object);
			imq_lock(rcv_mqueue);
			io_unlock(rcv_object);
			goto fast_send_receive;

		    abort_reply_dest_copyin:
			is_write_unlock(space);
			goto slow_copyin;

		    abort_reply_rcv_copyin:
			ip_unlock(dest_port);
			is_write_unlock(space);
			goto slow_send;
		    }

		    default:
			goto slow_copyin;
		}
		/*NOTREACHED*/

	    fast_send_receive:
		/*
		 *	optimized ipc_mqueue_send/ipc_mqueue_receive
		 *
		 *	Finished get/copyin of kmsg and copyin of rcv_name.
		 *	space is unlocked, dest_port is locked,
		 *	we can queue kmsg to dest_port,
		 *	rcv_mqueue is locked, rcv_object holds a ref,
		 *	if rcv_object is a port it isn't in a port set
		 *
		 *	Note that if simple locking is turned off,
		 *	then we could have dest_mqueue == rcv_mqueue
		 *	and not abort when we try to lock dest_mqueue.
		 */

		assert(ip_active(dest_port));
		assert(dest_port->ip_receiver != ipc_space_kernel);
		assert((dest_port->ip_msgcount < dest_port->ip_qlimit) ||
		       (MACH_MSGH_BITS_REMOTE(kmsg->ikm_header.msgh_bits) ==
						MACH_MSG_TYPE_PORT_SEND_ONCE));
		assert((kmsg->ikm_header.msgh_bits &
						MACH_MSGH_BITS_CIRCULAR) == 0);

	    {
		ipc_mqueue_t dest_mqueue;
		ipc_thread_t receiver;

	    {
		ipc_pset_t dest_pset;

		dest_pset = dest_port->ip_pset;
		if (dest_pset == IPS_NULL)
			dest_mqueue = &dest_port->ip_messages;
		else
			dest_mqueue = &dest_pset->ips_messages;
	    }

		if (!imq_lock_try(dest_mqueue)) {
		    abort_send_receive:
			ip_unlock(dest_port);
			imq_unlock(rcv_mqueue);
			ipc_object_release(rcv_object);
			goto slow_send;
		}

		receiver = ipc_thread_queue_first(&dest_mqueue->imq_threads);
		if ((receiver == ITH_NULL) ||
		    (ipc_kmsg_queue_first(&rcv_mqueue->imq_messages)
								!= IKM_NULL)) {
			imq_unlock(dest_mqueue);
			goto abort_send_receive;
		}

		/*
		 *	There is a receiver thread waiting, and
		 *	there is no reply message for us to pick up.
		 *	We have hope of hand-off, so save state.
		 */

		self->ith_msg = msg;
		self->ith_rcv_size = rcv_size;
		self->ith_object = rcv_object;
		self->ith_mqueue = rcv_mqueue;

		if ((receiver->swap_func == (void (*)()) mach_msg_continue) &&
		    thread_handoff(self, mach_msg_continue, receiver)) {
			assert(current_thread() == receiver);

			/*
			 *	We can use the optimized receive code,
			 *	because the receiver is using no options.
			 */
		} else if ((receiver->swap_func ==
				(void (*)()) exception_raise_continue) &&
			   thread_handoff(self, mach_msg_continue, receiver)) {
			counter(c_mach_msg_trap_block_exc++);
			assert(current_thread() == receiver);

			/*
			 *	We are a reply message coming back through
			 *	the optimized exception-handling path.
			 *	Finish with rcv_mqueue and dest_mqueue,
			 *	and then jump to exception code with
			 *	dest_port still locked.  We don't bother
			 *	with a sequence number in this case.
			 */

			ipc_thread_enqueue_macro(
				&rcv_mqueue->imq_threads, self);
			self->ith_state = MACH_RCV_IN_PROGRESS;
			self->ith_msize = MACH_MSG_SIZE_MAX;
			imq_unlock(rcv_mqueue);

			ipc_thread_rmqueue_first_macro(
				&dest_mqueue->imq_threads, receiver);
			imq_unlock(dest_mqueue);

			exception_raise_continue_fast(dest_port, kmsg);
			/*NOTREACHED*/
			return MACH_MSG_SUCCESS;
		} else if ((send_size <= receiver->ith_msize) &&
			   thread_handoff(self, mach_msg_continue, receiver)) {
			assert(current_thread() == receiver);

			if ((receiver->swap_func ==
				(void (*)()) mach_msg_receive_continue) &&
			    ((receiver->ith_option & MACH_RCV_NOTIFY) == 0)) {
				/*
				 *	We can still use the optimized code.
				 */
			} else {
				counter(c_mach_msg_trap_block_slow++);
				/*
				 *	We are running as the receiver,
				 *	but we can't use the optimized code.
				 *	Finish send/receive processing.
				 */

				dest_port->ip_msgcount++;
				ip_unlock(dest_port);

				ipc_thread_enqueue_macro(
					&rcv_mqueue->imq_threads, self);
				self->ith_state = MACH_RCV_IN_PROGRESS;
				self->ith_msize = MACH_MSG_SIZE_MAX;
				imq_unlock(rcv_mqueue);

				ipc_thread_rmqueue_first_macro(
					&dest_mqueue->imq_threads, receiver);
				receiver->ith_state = MACH_MSG_SUCCESS;
				receiver->ith_kmsg = kmsg;
				receiver->ith_seqno = dest_port->ip_seqno++;
				imq_unlock(dest_mqueue);

				/*
				 *	Call the receiver's continuation.
				 */

				receiver->wait_result = THREAD_AWAKENED;
				(*receiver->swap_func)();
				/*NOTREACHED*/
				return MACH_MSG_SUCCESS;
			}
		} else {
			/*
			 *	The receiver can't accept the message,
			 *	or we can't switch to the receiver.
			 */

			imq_unlock(dest_mqueue);
			goto abort_send_receive;
		}
		counter(c_mach_msg_trap_block_fast++);

		/*
		 *	Safe to unlock dest_port now that we are
		 *	committed to this path, because we hold
		 *	dest_mqueue locked.  We never bother changing
		 *	dest_port->ip_msgcount.
		 */

		ip_unlock(dest_port);

		/*
		 *	We need to finish preparing self for its
		 *	time asleep in rcv_mqueue.
		 */

		ipc_thread_enqueue_macro(&rcv_mqueue->imq_threads, self);
		self->ith_state = MACH_RCV_IN_PROGRESS;
		self->ith_msize = MACH_MSG_SIZE_MAX;
		imq_unlock(rcv_mqueue);

		/*
		 *	Finish extracting receiver from dest_mqueue.
		 */

		ipc_thread_rmqueue_first_macro(
			&dest_mqueue->imq_threads, receiver);
		kmsg->ikm_header.msgh_seqno = dest_port->ip_seqno++;
		imq_unlock(dest_mqueue);

		/*
		 *	We don't have to do any post-dequeue processing of
		 *	the message.  We never incremented ip_msgcount, we
		 *	know it has no msg-accepted request, and blocked
		 *	senders aren't a worry because we found the port
		 *	with a receiver waiting.
		 */

		self = receiver;
		space = self->task->itk_space;

		msg = self->ith_msg;
		rcv_size = self->ith_rcv_size;
		rcv_object = self->ith_object;

		/* inline ipc_object_release */
		io_lock(rcv_object);
		io_release(rcv_object);
		io_check_unlock(rcv_object);
	    }

	    fast_copyout:
		/*
		 *	Nothing locked and no references held, except
		 *	we have kmsg with msgh_seqno filled in.  Must
		 *	still check against rcv_size and do
		 *	ipc_kmsg_copyout/ipc_kmsg_put.
		 */

		assert((ipc_port_t) kmsg->ikm_header.msgh_remote_port
						== dest_port);

		reply_size = kmsg->ikm_header.msgh_size;
		if (rcv_size < reply_size)
			goto slow_copyout;

		/* optimized ipc_kmsg_copyout/ipc_kmsg_copyout_header */

		switch (kmsg->ikm_header.msgh_bits) {
		    case MACH_MSGH_BITS(MACH_MSG_TYPE_PORT_SEND,
					MACH_MSG_TYPE_PORT_SEND_ONCE): {
			ipc_port_t reply_port =
				(ipc_port_t) kmsg->ikm_header.msgh_local_port;
			mach_port_t dest_name, reply_name;

			/* receiving a request message */

			if (!IP_VALID(reply_port))
				goto slow_copyout;

			is_write_lock(space);
			assert(space->is_active);

			/*
			 *	To do an atomic copyout, need simultaneous
			 *	locks on both ports and the space.  If
			 *	dest_port == reply_port, and simple locking is
			 *	enabled, then we will abort.  Otherwise it's
			 *	OK to unlock twice.
			 */

			ip_lock(dest_port);
			if (!ip_active(dest_port) ||
			    !ip_lock_try(reply_port))
				goto abort_request_copyout;

			if (!ip_active(reply_port)) {
				ip_unlock(reply_port);
				goto abort_request_copyout;
			}

			assert(reply_port->ip_sorights > 0);
			ip_unlock(reply_port);

		    {
			ipc_entry_t table;
			ipc_entry_t entry;
			mach_port_index_t index;

			/* optimized ipc_entry_get */

			table = space->is_table;
			index = table->ie_next;

			if (index == 0)
				goto abort_request_copyout;

			entry = &table[index];
			table->ie_next = entry->ie_next;
			entry->ie_request = 0;

		    {
			mach_port_gen_t gen;

			assert((entry->ie_bits &~ IE_BITS_GEN_MASK) == 0);
			gen = entry->ie_bits + IE_BITS_GEN_ONE;

			reply_name = MACH_PORT_MAKE(index, gen);

			/* optimized ipc_right_copyout */

			entry->ie_bits = gen | (MACH_PORT_TYPE_SEND_ONCE | 1);
		    }

			assert(MACH_PORT_VALID(reply_name));
			entry->ie_object = (ipc_object_t) reply_port;
			is_write_unlock(space);
		    }

			/* optimized ipc_object_copyout_dest */

			assert(dest_port->ip_srights > 0);
			ip_release(dest_port);

			if (dest_port->ip_receiver == space)
				dest_name = dest_port->ip_receiver_name;
			else
				dest_name = MACH_PORT_NULL;

			if ((--dest_port->ip_srights == 0) &&
			    (dest_port->ip_nsrequest != IP_NULL)) {
				ipc_port_t nsrequest;
				mach_port_mscount_t mscount;

				/* a rather rare case */

				nsrequest = dest_port->ip_nsrequest;
				mscount = dest_port->ip_mscount;
				dest_port->ip_nsrequest = IP_NULL;
				ip_unlock(dest_port);

				ipc_notify_no_senders(nsrequest, mscount);
			} else
				ip_unlock(dest_port);

			kmsg->ikm_header.msgh_bits =
				MACH_MSGH_BITS(MACH_MSG_TYPE_PORT_SEND_ONCE,
					       MACH_MSG_TYPE_PORT_SEND);
			kmsg->ikm_header.msgh_remote_port = reply_name;
			kmsg->ikm_header.msgh_local_port = dest_name;
			goto fast_put;

		    abort_request_copyout:
			ip_unlock(dest_port);
			is_write_unlock(space);
			goto slow_copyout;
		    }

		    case MACH_MSGH_BITS(MACH_MSG_TYPE_PORT_SEND_ONCE, 0): {
			mach_port_t dest_name;

			/* receiving a reply message */

			ip_lock(dest_port);
			if (!ip_active(dest_port))
				goto slow_copyout;

			/* optimized ipc_object_copyout_dest */

			assert(dest_port->ip_sorights > 0);

			if (dest_port->ip_receiver == space) {
				ip_release(dest_port);
				dest_port->ip_sorights--;
				dest_name = dest_port->ip_receiver_name;
				ip_unlock(dest_port);
			} else {
				ip_unlock(dest_port);

				ipc_notify_send_once(dest_port);
				dest_name = MACH_PORT_NULL;
			}

			kmsg->ikm_header.msgh_bits =
				MACH_MSGH_BITS(0,
					       MACH_MSG_TYPE_PORT_SEND_ONCE);
			kmsg->ikm_header.msgh_remote_port = MACH_PORT_NULL;
			kmsg->ikm_header.msgh_local_port = dest_name;
			goto fast_put;
		    }

		    case MACH_MSGH_BITS_COMPLEX|
			 MACH_MSGH_BITS(MACH_MSG_TYPE_PORT_SEND_ONCE, 0): {
			mach_port_t dest_name;

			/* receiving a complex reply message */

			ip_lock(dest_port);
			if (!ip_active(dest_port))
				goto slow_copyout;

			/* optimized ipc_object_copyout_dest */

			assert(dest_port->ip_sorights > 0);

			if (dest_port->ip_receiver == space) {
				ip_release(dest_port);
				dest_port->ip_sorights--;
				dest_name = dest_port->ip_receiver_name;
				ip_unlock(dest_port);
			} else {
				ip_unlock(dest_port);

				ipc_notify_send_once(dest_port);
				dest_name = MACH_PORT_NULL;
			}

			kmsg->ikm_header.msgh_bits =
				MACH_MSGH_BITS_COMPLEX |
				MACH_MSGH_BITS(0,
					       MACH_MSG_TYPE_PORT_SEND_ONCE);
			kmsg->ikm_header.msgh_remote_port = MACH_PORT_NULL;
			kmsg->ikm_header.msgh_local_port = dest_name;

			mr = ipc_kmsg_copyout_body(
				(vm_offset_t) (&kmsg->ikm_header + 1),
				(vm_offset_t) &kmsg->ikm_header
					+ kmsg->ikm_header.msgh_size,
				space,
				current_map());

			if (mr != MACH_MSG_SUCCESS) {
				(void) ipc_kmsg_put(msg, kmsg,
					kmsg->ikm_header.msgh_size);
				return mr | MACH_RCV_BODY_ERROR;
			}
			goto fast_put;
		    }

		    default:
			goto slow_copyout;
		}
		/*NOTREACHED*/

	    fast_put:
		/*
		 *	We have the reply message data in kmsg,
		 *	and the reply message size in reply_size.
		 *	Just need to copy it out to the user and free kmsg.
		 *	We must check ikm_cache after copyoutmsg.
		 */

		ikm_check_initialized(kmsg, kmsg->ikm_size);

		if ((kmsg->ikm_size != IKM_SAVED_KMSG_SIZE) ||
		    copyoutmsg(&kmsg->ikm_header, msg,
			       reply_size) ||
		    (ikm_cache() != IKM_NULL))
			goto slow_put;

		ikm_cache() = kmsg;
		thread_syscall_return(MACH_MSG_SUCCESS);
		/*NOTREACHED*/
		return MACH_MSG_SUCCESS; /* help for the compiler */

		/*
		 *	The slow path has a few non-register temporary
		 *	variables used only for call-by-reference.
		 */

	    {
		ipc_kmsg_t temp_kmsg;
		mach_port_seqno_t temp_seqno;
		ipc_object_t temp_rcv_object;
		ipc_mqueue_t temp_rcv_mqueue;

	    slow_get:
		/*
		 *	No locks, references, or messages held.
		 *	Still have to get the request, send it,
		 *	receive reply, etc.
		 */

		mr = ipc_kmsg_get(msg, send_size, &temp_kmsg);
		if (mr != MACH_MSG_SUCCESS) {
			thread_syscall_return(mr);
			/*NOTREACHED*/
		}
		kmsg = temp_kmsg;

		/* try to get back on optimized path */
		goto fast_copyin;

	    slow_copyin:
		/*
		 *	We have the message data in kmsg, but
		 *	we still need to copyin, send it,
		 *	receive a reply, and do copyout.
		 */

		mr = ipc_kmsg_copyin(kmsg, space, current_map(),
				     MACH_PORT_NULL);
		if (mr != MACH_MSG_SUCCESS) {
			ikm_free(kmsg);
			thread_syscall_return(mr);
			/*NOTREACHED*/
		}

		/* try to get back on optimized path */

		if (kmsg->ikm_header.msgh_bits & MACH_MSGH_BITS_CIRCULAR)
			goto slow_send;

		dest_port = (ipc_port_t) kmsg->ikm_header.msgh_remote_port;
		assert(IP_VALID(dest_port));

		ip_lock(dest_port);
		if (dest_port->ip_receiver == ipc_space_kernel) {
			assert(ip_active(dest_port));
			ip_unlock(dest_port);
			goto kernel_send;
		}

		if (ip_active(dest_port) &&
		    ((dest_port->ip_msgcount < dest_port->ip_qlimit) ||
		     (MACH_MSGH_BITS_REMOTE(kmsg->ikm_header.msgh_bits) ==
					MACH_MSG_TYPE_PORT_SEND_ONCE)))
		{
		    /*
		     *	Try an optimized ipc_mqueue_copyin.
		     *	It will work if this is a request message.
		     */

		    ipc_port_t reply_port;

		    reply_port = (ipc_port_t)
					kmsg->ikm_header.msgh_local_port;
		    if (IP_VALID(reply_port)) {
			if (ip_lock_try(reply_port)) {
			    if (ip_active(reply_port) &&
				reply_port->ip_receiver == space &&
				reply_port->ip_receiver_name == rcv_name &&
				reply_port->ip_pset == IPS_NULL)
			    {
				/* Grab a reference to the reply port. */
				rcv_object = (ipc_object_t) reply_port;
				io_reference(rcv_object);
				rcv_mqueue = &reply_port->ip_messages;
				imq_lock(rcv_mqueue);
				io_unlock(rcv_object);
				goto fast_send_receive;
			    }
			    ip_unlock(reply_port);
			}
		    }
		}

		ip_unlock(dest_port);
		goto slow_send;

	    kernel_send:
		/*
		 *	Special case: send message to kernel services.
		 *	The request message has been copied into the
		 *	kmsg.  Nothing is locked.
		 */

	    {
		ipc_port_t	reply_port;

		/*
		 * Perform the kernel function.
		 */

		kmsg = ipc_kobject_server(kmsg);
		if (kmsg == IKM_NULL) {
			/*
			 * No reply.  Take the
			 * slow receive path.
			 */
			goto slow_get_rcv_port;
		}

		/*
		 * Check that:
		 *	the reply port is alive
		 *	we hold the receive right
		 *	the name has not changed.
		 *	the port is not in a set
		 * If any of these are not true,
		 * we cannot directly receive the reply
		 * message.
		 */
		reply_port = (ipc_port_t) kmsg->ikm_header.msgh_remote_port;
		ip_lock(reply_port);

		if ((!ip_active(reply_port)) ||
		    (reply_port->ip_receiver != space) ||
		    (reply_port->ip_receiver_name != rcv_name) ||
		    (reply_port->ip_pset != IPS_NULL))
		{
			ip_unlock(reply_port);
			ipc_mqueue_send_always(kmsg);
			goto slow_get_rcv_port;
		}

		rcv_mqueue = &reply_port->ip_messages;
		imq_lock(rcv_mqueue);
		/* keep port locked, and don`t change ref count yet */

		/*
		 * If there are messages on the port
		 * or other threads waiting for a message,
		 * we cannot directly receive the reply.
		 */
		if ((ipc_thread_queue_first(&rcv_mqueue->imq_threads)
			!= ITH_NULL) ||
		    (ipc_kmsg_queue_first(&rcv_mqueue->imq_messages)
			!= IKM_NULL))
		{
			imq_unlock(rcv_mqueue);
			ip_unlock(reply_port);
			ipc_mqueue_send_always(kmsg);
			goto slow_get_rcv_port;
		}

		/*
		 * We can directly receive this reply.
		 * Since the kernel reply never blocks,
		 * it holds no message_accepted request.
		 * Since there were no messages queued
		 * on the reply port, there should be
		 * no threads blocked waiting to send.
		 */

		assert(kmsg->ikm_marequest == IMAR_NULL);
		assert(ipc_thread_queue_first(&reply_port->ip_blocked)
				== ITH_NULL);

		dest_port = reply_port;
		kmsg->ikm_header.msgh_seqno = dest_port->ip_seqno++;
		imq_unlock(rcv_mqueue);

		/*
		 * inline ipc_object_release.
		 * Port is still locked.
		 * Reference count was not incremented.
		 */
		ip_check_unlock(reply_port);

		/* copy out the kernel reply */
		goto fast_copyout;
	    }

	    slow_send:
		/*
		 *	Nothing is locked.  We have acquired kmsg, but
		 *	we still need to send it and receive a reply.
		 */

		mr = ipc_mqueue_send(kmsg, MACH_MSG_OPTION_NONE,
				     MACH_MSG_TIMEOUT_NONE);
		if (mr != MACH_MSG_SUCCESS) {
			mr |= ipc_kmsg_copyout_pseudo(kmsg, space,
						      current_map());

			assert(kmsg->ikm_marequest == IMAR_NULL);
			(void) ipc_kmsg_put(msg, kmsg,
					    kmsg->ikm_header.msgh_size);
			thread_syscall_return(mr);
			/*NOTREACHED*/
		}

	    slow_get_rcv_port:
		/*
		 * We have sent the message.  Copy in the receive port.
		 */
		mr = ipc_mqueue_copyin(space, rcv_name,
				       &temp_rcv_mqueue, &temp_rcv_object);
		if (mr != MACH_MSG_SUCCESS) {
			thread_syscall_return(mr);
			/*NOTREACHED*/
		}
		rcv_mqueue = temp_rcv_mqueue;
		rcv_object = temp_rcv_object;
		/* hold ref for rcv_object; rcv_mqueue is locked */

	/*
	    slow_receive:
	*/
		/*
		 *	Now we have sent the request and copied in rcv_name,
		 *	so rcv_mqueue is locked and hold ref for rcv_object.
		 *	Just receive a reply and try to get back to fast path.
		 *
		 *	ipc_mqueue_receive may not return, because if we block
		 *	then our kernel stack may be discarded.  So we save
		 *	state here for mach_msg_continue to pick up.
		 */

		self->ith_msg = msg;
		self->ith_rcv_size = rcv_size;
		self->ith_object = rcv_object;
		self->ith_mqueue = rcv_mqueue;

		mr = ipc_mqueue_receive(rcv_mqueue,
					MACH_MSG_OPTION_NONE,
					MACH_MSG_SIZE_MAX,
					MACH_MSG_TIMEOUT_NONE,
					FALSE, mach_msg_continue,
		       			&temp_kmsg, &temp_seqno);
		/* rcv_mqueue is unlocked */
		ipc_object_release(rcv_object);
		if (mr != MACH_MSG_SUCCESS) {
			thread_syscall_return(mr);
			/*NOTREACHED*/
		}

		(kmsg = temp_kmsg)->ikm_header.msgh_seqno = temp_seqno;
		dest_port = (ipc_port_t) kmsg->ikm_header.msgh_remote_port;
		goto fast_copyout;

	    slow_copyout:
		/*
		 *	Nothing locked and no references held, except
		 *	we have kmsg with msgh_seqno filled in.  Must
		 *	still check against rcv_size and do
		 *	ipc_kmsg_copyout/ipc_kmsg_put.
		 */

		reply_size = kmsg->ikm_header.msgh_size;
		if (rcv_size < reply_size) {
			ipc_kmsg_copyout_dest(kmsg, space);
			(void) ipc_kmsg_put(msg, kmsg, sizeof *msg);
			thread_syscall_return(MACH_RCV_TOO_LARGE);
			/*NOTREACHED*/
		}

		mr = ipc_kmsg_copyout(kmsg, space, current_map(),
				      MACH_PORT_NULL);
		if (mr != MACH_MSG_SUCCESS) {
			if ((mr &~ MACH_MSG_MASK) == MACH_RCV_BODY_ERROR) {
				(void) ipc_kmsg_put(msg, kmsg,
						kmsg->ikm_header.msgh_size);
			} else {
				ipc_kmsg_copyout_dest(kmsg, space);
				(void) ipc_kmsg_put(msg, kmsg, sizeof *msg);
			}

			thread_syscall_return(mr);
			/*NOTREACHED*/
		}

		/* try to get back on optimized path */

		goto fast_put;

	    slow_put:
		mr = ipc_kmsg_put(msg, kmsg, reply_size);
		thread_syscall_return(mr);
		/*NOTREACHED*/
	    }
	} else if (option == MACH_SEND_MSG) {
		ipc_space_t space = current_space();
		vm_map_t map = current_map();
		ipc_kmsg_t kmsg;

		mr = ipc_kmsg_get(msg, send_size, &kmsg);
		if (mr != MACH_MSG_SUCCESS)
			return mr;

		mr = ipc_kmsg_copyin(kmsg, space, map, MACH_PORT_NULL);
		if (mr != MACH_MSG_SUCCESS) {
			ikm_free(kmsg);
			return mr;
		}

		mr = ipc_mqueue_send(kmsg, MACH_MSG_OPTION_NONE,
				     MACH_MSG_TIMEOUT_NONE);
		if (mr != MACH_MSG_SUCCESS) {
			mr |= ipc_kmsg_copyout_pseudo(kmsg, space, map);

			assert(kmsg->ikm_marequest == IMAR_NULL);
			(void) ipc_kmsg_put(msg, kmsg,
					    kmsg->ikm_header.msgh_size);
		}

		return mr;
	} else if (option == MACH_RCV_MSG) {
		ipc_thread_t self = current_thread();
		ipc_space_t space = current_space();
		vm_map_t map = current_map();
		ipc_object_t object;
		ipc_mqueue_t mqueue;
		ipc_kmsg_t kmsg;
		mach_port_seqno_t seqno;

		mr = ipc_mqueue_copyin(space, rcv_name, &mqueue, &object);
		if (mr != MACH_MSG_SUCCESS)
			return mr;
		/* hold ref for object; mqueue is locked */

		/*
		 *	ipc_mqueue_receive may not return, because if we block
		 *	then our kernel stack may be discarded.  So we save
		 *	state here for mach_msg_continue to pick up.
		 */

		self->ith_msg = msg;
		self->ith_rcv_size = rcv_size;
		self->ith_object = object;
		self->ith_mqueue = mqueue;

		mr = ipc_mqueue_receive(mqueue,
					MACH_MSG_OPTION_NONE,
					MACH_MSG_SIZE_MAX,
					MACH_MSG_TIMEOUT_NONE,
					FALSE, mach_msg_continue,
					&kmsg, &seqno);
		/* mqueue is unlocked */
		ipc_object_release(object);
		if (mr != MACH_MSG_SUCCESS)
			return mr;

		kmsg->ikm_header.msgh_seqno = seqno;
		if (rcv_size < kmsg->ikm_header.msgh_size) {
			ipc_kmsg_copyout_dest(kmsg, space);
			(void) ipc_kmsg_put(msg, kmsg, sizeof *msg);
			return MACH_RCV_TOO_LARGE;
		}

		mr = ipc_kmsg_copyout(kmsg, space, map, MACH_PORT_NULL);
		if (mr != MACH_MSG_SUCCESS) {
			if ((mr &~ MACH_MSG_MASK) == MACH_RCV_BODY_ERROR) {
				(void) ipc_kmsg_put(msg, kmsg,
						kmsg->ikm_header.msgh_size);
			} else {
				ipc_kmsg_copyout_dest(kmsg, space);
				(void) ipc_kmsg_put(msg, kmsg, sizeof *msg);
			}

			return mr;
		}

		return ipc_kmsg_put(msg, kmsg, kmsg->ikm_header.msgh_size);
	} else if (option == MACH_MSG_OPTION_NONE) {
		/*
		 *	We can measure the "null mach_msg_trap"
		 *	(syscall entry and thread_syscall_return exit)
		 *	with this path.
		 */

		thread_syscall_return(MACH_MSG_SUCCESS);
		/*NOTREACHED*/
	}

	if (option & MACH_SEND_MSG) {
		mr = mach_msg_send(msg, option, send_size,
				   time_out, notify);
		if (mr != MACH_MSG_SUCCESS)
			return mr;
	}

	if (option & MACH_RCV_MSG) {
		mr = mach_msg_receive(msg, option, rcv_size, rcv_name,
				      time_out, notify);
		if (mr != MACH_MSG_SUCCESS)
			return mr;
	}

	return MACH_MSG_SUCCESS;
}

/*
 *	Routine:	mach_msg_continue
 *	Purpose:
 *		Continue after blocking for a message.
 *	Conditions:
 *		Nothing locked.  We are running on a new kernel stack,
 *		with the receive state saved in the thread.  From here
 *		control goes back to user space.
 */

void
mach_msg_continue(void)
{
	ipc_thread_t thread = current_thread();
	task_t task = thread->task;
	ipc_space_t space = task->itk_space;
	vm_map_t map = task->map;
	mach_msg_header_t *msg = thread->ith_msg;
	mach_msg_size_t rcv_size = thread->ith_rcv_size;
	ipc_object_t object = thread->ith_object;
	ipc_mqueue_t mqueue = thread->ith_mqueue;
	ipc_kmsg_t kmsg;
	mach_port_seqno_t seqno;
	mach_msg_return_t mr;

	mr = ipc_mqueue_receive(mqueue, MACH_MSG_OPTION_NONE,
				MACH_MSG_SIZE_MAX, MACH_MSG_TIMEOUT_NONE,
				TRUE, mach_msg_continue, &kmsg, &seqno);
	/* mqueue is unlocked */
	ipc_object_release(object);
	if (mr != MACH_MSG_SUCCESS) {
		thread_syscall_return(mr);
		/*NOTREACHED*/
	}

	kmsg->ikm_header.msgh_seqno = seqno;
	if (kmsg->ikm_header.msgh_size > rcv_size) {
		ipc_kmsg_copyout_dest(kmsg, space);
		(void) ipc_kmsg_put(msg, kmsg, sizeof *msg);
		thread_syscall_return(MACH_RCV_TOO_LARGE);
		/*NOTREACHED*/
	}

	mr = ipc_kmsg_copyout(kmsg, space, map, MACH_PORT_NULL);
	if (mr != MACH_MSG_SUCCESS) {
		if ((mr &~ MACH_MSG_MASK) == MACH_RCV_BODY_ERROR) {
			(void) ipc_kmsg_put(msg, kmsg,
					kmsg->ikm_header.msgh_size);
		} else {
			ipc_kmsg_copyout_dest(kmsg, space);
			(void) ipc_kmsg_put(msg, kmsg, sizeof *msg);
		}

		thread_syscall_return(mr);
		/*NOTREACHED*/
	}

	mr = ipc_kmsg_put(msg, kmsg, kmsg->ikm_header.msgh_size);
	thread_syscall_return(mr);
	/*NOTREACHED*/
}

/*
 *	Routine:	mach_msg_interrupt
 *	Purpose:
 *		Attempts to force a thread waiting at mach_msg_continue or
 *		mach_msg_receive_continue into a clean point.  Returns TRUE
 *		if this was possible.
 *	Conditions:
 *		Nothing locked.  The thread must NOT be runnable.
 */

boolean_t
mach_msg_interrupt(thread)
	thread_t thread;
{
	ipc_mqueue_t mqueue;

	assert((thread->swap_func == (void (*)()) mach_msg_continue) ||
	       (thread->swap_func == (void (*)()) mach_msg_receive_continue));

	mqueue = thread->ith_mqueue;
	imq_lock(mqueue);
	if (thread->ith_state != MACH_RCV_IN_PROGRESS) {
		/*
		 *	The thread is no longer waiting for a message.
		 *	It may have a message sitting in ith_kmsg.
		 *	We can't clean this up.
		 */

		imq_unlock(mqueue);
		return FALSE;
	}
	ipc_thread_rmqueue(&mqueue->imq_threads, thread);
	imq_unlock(mqueue);

	ipc_object_release(thread->ith_object);

	thread_set_syscall_return(thread, MACH_RCV_INTERRUPTED);
	thread->swap_func = thread_exception_return;
	return TRUE;
}