1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
|
/*-
* Copyright (c) 2011 NetApp, Inc.
* 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.
*
* THIS SOFTWARE IS PROVIDED BY NETAPP, INC ``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 NETAPP, INC 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.
*
* $FreeBSD$
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/module.h>
#include <sys/sysctl.h>
#include <sys/malloc.h>
#include <sys/pcpu.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/proc.h>
#include <sys/rwlock.h>
#include <sys/sched.h>
#include <sys/smp.h>
#include <sys/systm.h>
#include <vm/vm.h>
#include <vm/vm_object.h>
#include <vm/vm_page.h>
#include <vm/pmap.h>
#include <vm/vm_map.h>
#include <vm/vm_extern.h>
#include <vm/vm_param.h>
#include <machine/cpu.h>
#include <machine/vm.h>
#include <machine/pcb.h>
#include <machine/smp.h>
#include <x86/psl.h>
#include <x86/apicreg.h>
#include <machine/vmparam.h>
#include <machine/vmm.h>
#include <machine/vmm_dev.h>
#include "vmm_ktr.h"
#include "vmm_host.h"
#include "vmm_mem.h"
#include "vmm_util.h"
#include "vatpic.h"
#include "vatpit.h"
#include "vhpet.h"
#include "vioapic.h"
#include "vlapic.h"
#include "vmm_msr.h"
#include "vmm_ipi.h"
#include "vmm_stat.h"
#include "vmm_lapic.h"
#include "io/ppt.h"
#include "io/iommu.h"
struct vlapic;
struct vcpu {
int flags;
enum vcpu_state state;
struct mtx mtx;
int hostcpu; /* host cpuid this vcpu last ran on */
uint64_t guest_msrs[VMM_MSR_NUM];
struct vlapic *vlapic;
int vcpuid;
struct savefpu *guestfpu; /* guest fpu state */
uint64_t guest_xcr0;
void *stats;
struct vm_exit exitinfo;
enum x2apic_state x2apic_state;
int nmi_pending;
int extint_pending;
struct vm_exception exception;
int exception_pending;
};
#define vcpu_lock_init(v) mtx_init(&((v)->mtx), "vcpu lock", 0, MTX_SPIN)
#define vcpu_lock(v) mtx_lock_spin(&((v)->mtx))
#define vcpu_unlock(v) mtx_unlock_spin(&((v)->mtx))
#define vcpu_assert_locked(v) mtx_assert(&((v)->mtx), MA_OWNED)
struct mem_seg {
vm_paddr_t gpa;
size_t len;
boolean_t wired;
vm_object_t object;
};
#define VM_MAX_MEMORY_SEGMENTS 2
struct vm {
void *cookie; /* processor-specific data */
void *iommu; /* iommu-specific data */
struct vhpet *vhpet; /* virtual HPET */
struct vioapic *vioapic; /* virtual ioapic */
struct vatpic *vatpic; /* virtual atpic */
struct vatpit *vatpit; /* virtual atpit */
struct vmspace *vmspace; /* guest's address space */
struct vcpu vcpu[VM_MAXCPU];
int num_mem_segs;
struct mem_seg mem_segs[VM_MAX_MEMORY_SEGMENTS];
char name[VM_MAX_NAMELEN];
/*
* Set of active vcpus.
* An active vcpu is one that has been started implicitly (BSP) or
* explicitly (AP) by sending it a startup ipi.
*/
volatile cpuset_t active_cpus;
struct mtx rendezvous_mtx;
cpuset_t rendezvous_req_cpus;
cpuset_t rendezvous_done_cpus;
void *rendezvous_arg;
vm_rendezvous_func_t rendezvous_func;
int suspend;
volatile cpuset_t suspended_cpus;
};
static int vmm_initialized;
static struct vmm_ops *ops;
#define VMM_INIT(num) (ops != NULL ? (*ops->init)(num) : 0)
#define VMM_CLEANUP() (ops != NULL ? (*ops->cleanup)() : 0)
#define VMM_RESUME() (ops != NULL ? (*ops->resume)() : 0)
#define VMINIT(vm, pmap) (ops != NULL ? (*ops->vminit)(vm, pmap): NULL)
#define VMRUN(vmi, vcpu, rip, pmap, rptr, sptr) \
(ops != NULL ? (*ops->vmrun)(vmi, vcpu, rip, pmap, rptr, sptr) : ENXIO)
#define VMCLEANUP(vmi) (ops != NULL ? (*ops->vmcleanup)(vmi) : NULL)
#define VMSPACE_ALLOC(min, max) \
(ops != NULL ? (*ops->vmspace_alloc)(min, max) : NULL)
#define VMSPACE_FREE(vmspace) \
(ops != NULL ? (*ops->vmspace_free)(vmspace) : ENXIO)
#define VMGETREG(vmi, vcpu, num, retval) \
(ops != NULL ? (*ops->vmgetreg)(vmi, vcpu, num, retval) : ENXIO)
#define VMSETREG(vmi, vcpu, num, val) \
(ops != NULL ? (*ops->vmsetreg)(vmi, vcpu, num, val) : ENXIO)
#define VMGETDESC(vmi, vcpu, num, desc) \
(ops != NULL ? (*ops->vmgetdesc)(vmi, vcpu, num, desc) : ENXIO)
#define VMSETDESC(vmi, vcpu, num, desc) \
(ops != NULL ? (*ops->vmsetdesc)(vmi, vcpu, num, desc) : ENXIO)
#define VMGETCAP(vmi, vcpu, num, retval) \
(ops != NULL ? (*ops->vmgetcap)(vmi, vcpu, num, retval) : ENXIO)
#define VMSETCAP(vmi, vcpu, num, val) \
(ops != NULL ? (*ops->vmsetcap)(vmi, vcpu, num, val) : ENXIO)
#define VLAPIC_INIT(vmi, vcpu) \
(ops != NULL ? (*ops->vlapic_init)(vmi, vcpu) : NULL)
#define VLAPIC_CLEANUP(vmi, vlapic) \
(ops != NULL ? (*ops->vlapic_cleanup)(vmi, vlapic) : NULL)
#define fpu_start_emulating() load_cr0(rcr0() | CR0_TS)
#define fpu_stop_emulating() clts()
static MALLOC_DEFINE(M_VM, "vm", "vm");
CTASSERT(VMM_MSR_NUM <= 64); /* msr_mask can keep track of up to 64 msrs */
/* statistics */
static VMM_STAT(VCPU_TOTAL_RUNTIME, "vcpu total runtime");
SYSCTL_NODE(_hw, OID_AUTO, vmm, CTLFLAG_RW, NULL, NULL);
static int vmm_ipinum;
SYSCTL_INT(_hw_vmm, OID_AUTO, ipinum, CTLFLAG_RD, &vmm_ipinum, 0,
"IPI vector used for vcpu notifications");
static void
vcpu_cleanup(struct vm *vm, int i)
{
struct vcpu *vcpu = &vm->vcpu[i];
VLAPIC_CLEANUP(vm->cookie, vcpu->vlapic);
vmm_stat_free(vcpu->stats);
fpu_save_area_free(vcpu->guestfpu);
}
static void
vcpu_init(struct vm *vm, uint32_t vcpu_id)
{
struct vcpu *vcpu;
vcpu = &vm->vcpu[vcpu_id];
vcpu_lock_init(vcpu);
vcpu->hostcpu = NOCPU;
vcpu->vcpuid = vcpu_id;
vcpu->vlapic = VLAPIC_INIT(vm->cookie, vcpu_id);
vm_set_x2apic_state(vm, vcpu_id, X2APIC_DISABLED);
vcpu->guest_xcr0 = XFEATURE_ENABLED_X87;
vcpu->guestfpu = fpu_save_area_alloc();
fpu_save_area_reset(vcpu->guestfpu);
vcpu->stats = vmm_stat_alloc();
}
struct vm_exit *
vm_exitinfo(struct vm *vm, int cpuid)
{
struct vcpu *vcpu;
if (cpuid < 0 || cpuid >= VM_MAXCPU)
panic("vm_exitinfo: invalid cpuid %d", cpuid);
vcpu = &vm->vcpu[cpuid];
return (&vcpu->exitinfo);
}
static void
vmm_resume(void)
{
VMM_RESUME();
}
static int
vmm_init(void)
{
int error;
vmm_host_state_init();
vmm_ipinum = vmm_ipi_alloc();
if (vmm_ipinum == 0)
vmm_ipinum = IPI_AST;
error = vmm_mem_init();
if (error)
return (error);
if (vmm_is_intel())
ops = &vmm_ops_intel;
else if (vmm_is_amd())
ops = &vmm_ops_amd;
else
return (ENXIO);
vmm_msr_init();
vmm_resume_p = vmm_resume;
return (VMM_INIT(vmm_ipinum));
}
static int
vmm_handler(module_t mod, int what, void *arg)
{
int error;
switch (what) {
case MOD_LOAD:
vmmdev_init();
if (ppt_avail_devices() > 0)
iommu_init();
error = vmm_init();
if (error == 0)
vmm_initialized = 1;
break;
case MOD_UNLOAD:
error = vmmdev_cleanup();
if (error == 0) {
vmm_resume_p = NULL;
iommu_cleanup();
if (vmm_ipinum != IPI_AST)
vmm_ipi_free(vmm_ipinum);
error = VMM_CLEANUP();
/*
* Something bad happened - prevent new
* VMs from being created
*/
if (error)
vmm_initialized = 0;
}
break;
default:
error = 0;
break;
}
return (error);
}
static moduledata_t vmm_kmod = {
"vmm",
vmm_handler,
NULL
};
/*
* vmm initialization has the following dependencies:
*
* - iommu initialization must happen after the pci passthru driver has had
* a chance to attach to any passthru devices (after SI_SUB_CONFIGURE).
*
* - VT-x initialization requires smp_rendezvous() and therefore must happen
* after SMP is fully functional (after SI_SUB_SMP).
*/
DECLARE_MODULE(vmm, vmm_kmod, SI_SUB_SMP + 1, SI_ORDER_ANY);
MODULE_VERSION(vmm, 1);
int
vm_create(const char *name, struct vm **retvm)
{
int i;
struct vm *vm;
struct vmspace *vmspace;
const int BSP = 0;
/*
* If vmm.ko could not be successfully initialized then don't attempt
* to create the virtual machine.
*/
if (!vmm_initialized)
return (ENXIO);
if (name == NULL || strlen(name) >= VM_MAX_NAMELEN)
return (EINVAL);
vmspace = VMSPACE_ALLOC(VM_MIN_ADDRESS, VM_MAXUSER_ADDRESS);
if (vmspace == NULL)
return (ENOMEM);
vm = malloc(sizeof(struct vm), M_VM, M_WAITOK | M_ZERO);
strcpy(vm->name, name);
vm->vmspace = vmspace;
mtx_init(&vm->rendezvous_mtx, "vm rendezvous lock", 0, MTX_DEF);
vm->cookie = VMINIT(vm, vmspace_pmap(vmspace));
vm->vioapic = vioapic_init(vm);
vm->vhpet = vhpet_init(vm);
vm->vatpic = vatpic_init(vm);
vm->vatpit = vatpit_init(vm);
for (i = 0; i < VM_MAXCPU; i++) {
vcpu_init(vm, i);
guest_msrs_init(vm, i);
}
vm_activate_cpu(vm, BSP);
*retvm = vm;
return (0);
}
static void
vm_free_mem_seg(struct vm *vm, struct mem_seg *seg)
{
if (seg->object != NULL)
vmm_mem_free(vm->vmspace, seg->gpa, seg->len);
bzero(seg, sizeof(*seg));
}
void
vm_destroy(struct vm *vm)
{
int i;
ppt_unassign_all(vm);
if (vm->iommu != NULL)
iommu_destroy_domain(vm->iommu);
vatpit_cleanup(vm->vatpit);
vhpet_cleanup(vm->vhpet);
vatpic_cleanup(vm->vatpic);
vioapic_cleanup(vm->vioapic);
for (i = 0; i < vm->num_mem_segs; i++)
vm_free_mem_seg(vm, &vm->mem_segs[i]);
vm->num_mem_segs = 0;
for (i = 0; i < VM_MAXCPU; i++)
vcpu_cleanup(vm, i);
VMSPACE_FREE(vm->vmspace);
VMCLEANUP(vm->cookie);
free(vm, M_VM);
}
const char *
vm_name(struct vm *vm)
{
return (vm->name);
}
int
vm_map_mmio(struct vm *vm, vm_paddr_t gpa, size_t len, vm_paddr_t hpa)
{
vm_object_t obj;
if ((obj = vmm_mmio_alloc(vm->vmspace, gpa, len, hpa)) == NULL)
return (ENOMEM);
else
return (0);
}
int
vm_unmap_mmio(struct vm *vm, vm_paddr_t gpa, size_t len)
{
vmm_mmio_free(vm->vmspace, gpa, len);
return (0);
}
boolean_t
vm_mem_allocated(struct vm *vm, vm_paddr_t gpa)
{
int i;
vm_paddr_t gpabase, gpalimit;
for (i = 0; i < vm->num_mem_segs; i++) {
gpabase = vm->mem_segs[i].gpa;
gpalimit = gpabase + vm->mem_segs[i].len;
if (gpa >= gpabase && gpa < gpalimit)
return (TRUE); /* 'gpa' is regular memory */
}
if (ppt_is_mmio(vm, gpa))
return (TRUE); /* 'gpa' is pci passthru mmio */
return (FALSE);
}
int
vm_malloc(struct vm *vm, vm_paddr_t gpa, size_t len)
{
int available, allocated;
struct mem_seg *seg;
vm_object_t object;
vm_paddr_t g;
if ((gpa & PAGE_MASK) || (len & PAGE_MASK) || len == 0)
return (EINVAL);
available = allocated = 0;
g = gpa;
while (g < gpa + len) {
if (vm_mem_allocated(vm, g))
allocated++;
else
available++;
g += PAGE_SIZE;
}
/*
* If there are some allocated and some available pages in the address
* range then it is an error.
*/
if (allocated && available)
return (EINVAL);
/*
* If the entire address range being requested has already been
* allocated then there isn't anything more to do.
*/
if (allocated && available == 0)
return (0);
if (vm->num_mem_segs >= VM_MAX_MEMORY_SEGMENTS)
return (E2BIG);
seg = &vm->mem_segs[vm->num_mem_segs];
if ((object = vmm_mem_alloc(vm->vmspace, gpa, len)) == NULL)
return (ENOMEM);
seg->gpa = gpa;
seg->len = len;
seg->object = object;
seg->wired = FALSE;
vm->num_mem_segs++;
return (0);
}
static void
vm_gpa_unwire(struct vm *vm)
{
int i, rv;
struct mem_seg *seg;
for (i = 0; i < vm->num_mem_segs; i++) {
seg = &vm->mem_segs[i];
if (!seg->wired)
continue;
rv = vm_map_unwire(&vm->vmspace->vm_map,
seg->gpa, seg->gpa + seg->len,
VM_MAP_WIRE_USER | VM_MAP_WIRE_NOHOLES);
KASSERT(rv == KERN_SUCCESS, ("vm(%s) memory segment "
"%#lx/%ld could not be unwired: %d",
vm_name(vm), seg->gpa, seg->len, rv));
seg->wired = FALSE;
}
}
static int
vm_gpa_wire(struct vm *vm)
{
int i, rv;
struct mem_seg *seg;
for (i = 0; i < vm->num_mem_segs; i++) {
seg = &vm->mem_segs[i];
if (seg->wired)
continue;
/* XXX rlimits? */
rv = vm_map_wire(&vm->vmspace->vm_map,
seg->gpa, seg->gpa + seg->len,
VM_MAP_WIRE_USER | VM_MAP_WIRE_NOHOLES);
if (rv != KERN_SUCCESS)
break;
seg->wired = TRUE;
}
if (i < vm->num_mem_segs) {
/*
* Undo the wiring before returning an error.
*/
vm_gpa_unwire(vm);
return (EAGAIN);
}
return (0);
}
static void
vm_iommu_modify(struct vm *vm, boolean_t map)
{
int i, sz;
vm_paddr_t gpa, hpa;
struct mem_seg *seg;
void *vp, *cookie, *host_domain;
sz = PAGE_SIZE;
host_domain = iommu_host_domain();
for (i = 0; i < vm->num_mem_segs; i++) {
seg = &vm->mem_segs[i];
KASSERT(seg->wired, ("vm(%s) memory segment %#lx/%ld not wired",
vm_name(vm), seg->gpa, seg->len));
gpa = seg->gpa;
while (gpa < seg->gpa + seg->len) {
vp = vm_gpa_hold(vm, gpa, PAGE_SIZE, VM_PROT_WRITE,
&cookie);
KASSERT(vp != NULL, ("vm(%s) could not map gpa %#lx",
vm_name(vm), gpa));
vm_gpa_release(cookie);
hpa = DMAP_TO_PHYS((uintptr_t)vp);
if (map) {
iommu_create_mapping(vm->iommu, gpa, hpa, sz);
iommu_remove_mapping(host_domain, hpa, sz);
} else {
iommu_remove_mapping(vm->iommu, gpa, sz);
iommu_create_mapping(host_domain, hpa, hpa, sz);
}
gpa += PAGE_SIZE;
}
}
/*
* Invalidate the cached translations associated with the domain
* from which pages were removed.
*/
if (map)
iommu_invalidate_tlb(host_domain);
else
iommu_invalidate_tlb(vm->iommu);
}
#define vm_iommu_unmap(vm) vm_iommu_modify((vm), FALSE)
#define vm_iommu_map(vm) vm_iommu_modify((vm), TRUE)
int
vm_unassign_pptdev(struct vm *vm, int bus, int slot, int func)
{
int error;
error = ppt_unassign_device(vm, bus, slot, func);
if (error)
return (error);
if (ppt_assigned_devices(vm) == 0) {
vm_iommu_unmap(vm);
vm_gpa_unwire(vm);
}
return (0);
}
int
vm_assign_pptdev(struct vm *vm, int bus, int slot, int func)
{
int error;
vm_paddr_t maxaddr;
/*
* Virtual machines with pci passthru devices get special treatment:
* - the guest physical memory is wired
* - the iommu is programmed to do the 'gpa' to 'hpa' translation
*
* We need to do this before the first pci passthru device is attached.
*/
if (ppt_assigned_devices(vm) == 0) {
KASSERT(vm->iommu == NULL,
("vm_assign_pptdev: iommu must be NULL"));
maxaddr = vmm_mem_maxaddr();
vm->iommu = iommu_create_domain(maxaddr);
error = vm_gpa_wire(vm);
if (error)
return (error);
vm_iommu_map(vm);
}
error = ppt_assign_device(vm, bus, slot, func);
return (error);
}
void *
vm_gpa_hold(struct vm *vm, vm_paddr_t gpa, size_t len, int reqprot,
void **cookie)
{
int count, pageoff;
vm_page_t m;
pageoff = gpa & PAGE_MASK;
if (len > PAGE_SIZE - pageoff)
panic("vm_gpa_hold: invalid gpa/len: 0x%016lx/%lu", gpa, len);
count = vm_fault_quick_hold_pages(&vm->vmspace->vm_map,
trunc_page(gpa), PAGE_SIZE, reqprot, &m, 1);
if (count == 1) {
*cookie = m;
return ((void *)(PHYS_TO_DMAP(VM_PAGE_TO_PHYS(m)) + pageoff));
} else {
*cookie = NULL;
return (NULL);
}
}
void
vm_gpa_release(void *cookie)
{
vm_page_t m = cookie;
vm_page_lock(m);
vm_page_unhold(m);
vm_page_unlock(m);
}
int
vm_gpabase2memseg(struct vm *vm, vm_paddr_t gpabase,
struct vm_memory_segment *seg)
{
int i;
for (i = 0; i < vm->num_mem_segs; i++) {
if (gpabase == vm->mem_segs[i].gpa) {
seg->gpa = vm->mem_segs[i].gpa;
seg->len = vm->mem_segs[i].len;
seg->wired = vm->mem_segs[i].wired;
return (0);
}
}
return (-1);
}
int
vm_get_memobj(struct vm *vm, vm_paddr_t gpa, size_t len,
vm_offset_t *offset, struct vm_object **object)
{
int i;
size_t seg_len;
vm_paddr_t seg_gpa;
vm_object_t seg_obj;
for (i = 0; i < vm->num_mem_segs; i++) {
if ((seg_obj = vm->mem_segs[i].object) == NULL)
continue;
seg_gpa = vm->mem_segs[i].gpa;
seg_len = vm->mem_segs[i].len;
if (gpa >= seg_gpa && gpa < seg_gpa + seg_len) {
*offset = gpa - seg_gpa;
*object = seg_obj;
vm_object_reference(seg_obj);
return (0);
}
}
return (EINVAL);
}
int
vm_get_register(struct vm *vm, int vcpu, int reg, uint64_t *retval)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (reg >= VM_REG_LAST)
return (EINVAL);
return (VMGETREG(vm->cookie, vcpu, reg, retval));
}
int
vm_set_register(struct vm *vm, int vcpu, int reg, uint64_t val)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (reg >= VM_REG_LAST)
return (EINVAL);
return (VMSETREG(vm->cookie, vcpu, reg, val));
}
static boolean_t
is_descriptor_table(int reg)
{
switch (reg) {
case VM_REG_GUEST_IDTR:
case VM_REG_GUEST_GDTR:
return (TRUE);
default:
return (FALSE);
}
}
static boolean_t
is_segment_register(int reg)
{
switch (reg) {
case VM_REG_GUEST_ES:
case VM_REG_GUEST_CS:
case VM_REG_GUEST_SS:
case VM_REG_GUEST_DS:
case VM_REG_GUEST_FS:
case VM_REG_GUEST_GS:
case VM_REG_GUEST_TR:
case VM_REG_GUEST_LDTR:
return (TRUE);
default:
return (FALSE);
}
}
int
vm_get_seg_desc(struct vm *vm, int vcpu, int reg,
struct seg_desc *desc)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (!is_segment_register(reg) && !is_descriptor_table(reg))
return (EINVAL);
return (VMGETDESC(vm->cookie, vcpu, reg, desc));
}
int
vm_set_seg_desc(struct vm *vm, int vcpu, int reg,
struct seg_desc *desc)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (!is_segment_register(reg) && !is_descriptor_table(reg))
return (EINVAL);
return (VMSETDESC(vm->cookie, vcpu, reg, desc));
}
static void
restore_guest_fpustate(struct vcpu *vcpu)
{
/* flush host state to the pcb */
fpuexit(curthread);
/* restore guest FPU state */
fpu_stop_emulating();
fpurestore(vcpu->guestfpu);
/* restore guest XCR0 if XSAVE is enabled in the host */
if (rcr4() & CR4_XSAVE)
load_xcr(0, vcpu->guest_xcr0);
/*
* The FPU is now "dirty" with the guest's state so turn on emulation
* to trap any access to the FPU by the host.
*/
fpu_start_emulating();
}
static void
save_guest_fpustate(struct vcpu *vcpu)
{
if ((rcr0() & CR0_TS) == 0)
panic("fpu emulation not enabled in host!");
/* save guest XCR0 and restore host XCR0 */
if (rcr4() & CR4_XSAVE) {
vcpu->guest_xcr0 = rxcr(0);
load_xcr(0, vmm_get_host_xcr0());
}
/* save guest FPU state */
fpu_stop_emulating();
fpusave(vcpu->guestfpu);
fpu_start_emulating();
}
static VMM_STAT(VCPU_IDLE_TICKS, "number of ticks vcpu was idle");
static int
vcpu_set_state_locked(struct vcpu *vcpu, enum vcpu_state newstate,
bool from_idle)
{
int error;
vcpu_assert_locked(vcpu);
/*
* State transitions from the vmmdev_ioctl() must always begin from
* the VCPU_IDLE state. This guarantees that there is only a single
* ioctl() operating on a vcpu at any point.
*/
if (from_idle) {
while (vcpu->state != VCPU_IDLE)
msleep_spin(&vcpu->state, &vcpu->mtx, "vmstat", hz);
} else {
KASSERT(vcpu->state != VCPU_IDLE, ("invalid transition from "
"vcpu idle state"));
}
if (vcpu->state == VCPU_RUNNING) {
KASSERT(vcpu->hostcpu == curcpu, ("curcpu %d and hostcpu %d "
"mismatch for running vcpu", curcpu, vcpu->hostcpu));
} else {
KASSERT(vcpu->hostcpu == NOCPU, ("Invalid hostcpu %d for a "
"vcpu that is not running", vcpu->hostcpu));
}
/*
* The following state transitions are allowed:
* IDLE -> FROZEN -> IDLE
* FROZEN -> RUNNING -> FROZEN
* FROZEN -> SLEEPING -> FROZEN
*/
switch (vcpu->state) {
case VCPU_IDLE:
case VCPU_RUNNING:
case VCPU_SLEEPING:
error = (newstate != VCPU_FROZEN);
break;
case VCPU_FROZEN:
error = (newstate == VCPU_FROZEN);
break;
default:
error = 1;
break;
}
if (error)
return (EBUSY);
vcpu->state = newstate;
if (newstate == VCPU_RUNNING)
vcpu->hostcpu = curcpu;
else
vcpu->hostcpu = NOCPU;
if (newstate == VCPU_IDLE)
wakeup(&vcpu->state);
return (0);
}
static void
vcpu_require_state(struct vm *vm, int vcpuid, enum vcpu_state newstate)
{
int error;
if ((error = vcpu_set_state(vm, vcpuid, newstate, false)) != 0)
panic("Error %d setting state to %d\n", error, newstate);
}
static void
vcpu_require_state_locked(struct vcpu *vcpu, enum vcpu_state newstate)
{
int error;
if ((error = vcpu_set_state_locked(vcpu, newstate, false)) != 0)
panic("Error %d setting state to %d", error, newstate);
}
static void
vm_set_rendezvous_func(struct vm *vm, vm_rendezvous_func_t func)
{
KASSERT(mtx_owned(&vm->rendezvous_mtx), ("rendezvous_mtx not locked"));
/*
* Update 'rendezvous_func' and execute a write memory barrier to
* ensure that it is visible across all host cpus. This is not needed
* for correctness but it does ensure that all the vcpus will notice
* that the rendezvous is requested immediately.
*/
vm->rendezvous_func = func;
wmb();
}
#define RENDEZVOUS_CTR0(vm, vcpuid, fmt) \
do { \
if (vcpuid >= 0) \
VCPU_CTR0(vm, vcpuid, fmt); \
else \
VM_CTR0(vm, fmt); \
} while (0)
static void
vm_handle_rendezvous(struct vm *vm, int vcpuid)
{
KASSERT(vcpuid == -1 || (vcpuid >= 0 && vcpuid < VM_MAXCPU),
("vm_handle_rendezvous: invalid vcpuid %d", vcpuid));
mtx_lock(&vm->rendezvous_mtx);
while (vm->rendezvous_func != NULL) {
/* 'rendezvous_req_cpus' must be a subset of 'active_cpus' */
CPU_AND(&vm->rendezvous_req_cpus, &vm->active_cpus);
if (vcpuid != -1 &&
CPU_ISSET(vcpuid, &vm->rendezvous_req_cpus) &&
!CPU_ISSET(vcpuid, &vm->rendezvous_done_cpus)) {
VCPU_CTR0(vm, vcpuid, "Calling rendezvous func");
(*vm->rendezvous_func)(vm, vcpuid, vm->rendezvous_arg);
CPU_SET(vcpuid, &vm->rendezvous_done_cpus);
}
if (CPU_CMP(&vm->rendezvous_req_cpus,
&vm->rendezvous_done_cpus) == 0) {
VCPU_CTR0(vm, vcpuid, "Rendezvous completed");
vm_set_rendezvous_func(vm, NULL);
wakeup(&vm->rendezvous_func);
break;
}
RENDEZVOUS_CTR0(vm, vcpuid, "Wait for rendezvous completion");
mtx_sleep(&vm->rendezvous_func, &vm->rendezvous_mtx, 0,
"vmrndv", 0);
}
mtx_unlock(&vm->rendezvous_mtx);
}
/*
* Emulate a guest 'hlt' by sleeping until the vcpu is ready to run.
*/
static int
vm_handle_hlt(struct vm *vm, int vcpuid, bool intr_disabled, bool *retu)
{
struct vcpu *vcpu;
const char *wmesg;
int t;
vcpu = &vm->vcpu[vcpuid];
vcpu_lock(vcpu);
while (1) {
/*
* Do a final check for pending NMI or interrupts before
* really putting this thread to sleep. Also check for
* software events that would cause this vcpu to wakeup.
*
* These interrupts/events could have happened after the
* vcpu returned from VMRUN() and before it acquired the
* vcpu lock above.
*/
if (vm->rendezvous_func != NULL || vm->suspend)
break;
if (vm_nmi_pending(vm, vcpuid))
break;
if (!intr_disabled) {
if (vm_extint_pending(vm, vcpuid) ||
vlapic_pending_intr(vcpu->vlapic, NULL)) {
break;
}
}
if (vlapic_enabled(vcpu->vlapic))
wmesg = "vmidle";
else
wmesg = "vmhalt";
t = ticks;
vcpu_require_state_locked(vcpu, VCPU_SLEEPING);
msleep_spin(vcpu, &vcpu->mtx, wmesg, 0);
vcpu_require_state_locked(vcpu, VCPU_FROZEN);
vmm_stat_incr(vm, vcpuid, VCPU_IDLE_TICKS, ticks - t);
}
vcpu_unlock(vcpu);
return (0);
}
static int
vm_handle_paging(struct vm *vm, int vcpuid, bool *retu)
{
int rv, ftype;
struct vm_map *map;
struct vcpu *vcpu;
struct vm_exit *vme;
vcpu = &vm->vcpu[vcpuid];
vme = &vcpu->exitinfo;
ftype = vme->u.paging.fault_type;
KASSERT(ftype == VM_PROT_READ ||
ftype == VM_PROT_WRITE || ftype == VM_PROT_EXECUTE,
("vm_handle_paging: invalid fault_type %d", ftype));
if (ftype == VM_PROT_READ || ftype == VM_PROT_WRITE) {
rv = pmap_emulate_accessed_dirty(vmspace_pmap(vm->vmspace),
vme->u.paging.gpa, ftype);
if (rv == 0)
goto done;
}
map = &vm->vmspace->vm_map;
rv = vm_fault(map, vme->u.paging.gpa, ftype, VM_FAULT_NORMAL);
VCPU_CTR3(vm, vcpuid, "vm_handle_paging rv = %d, gpa = %#lx, "
"ftype = %d", rv, vme->u.paging.gpa, ftype);
if (rv != KERN_SUCCESS)
return (EFAULT);
done:
/* restart execution at the faulting instruction */
vme->inst_length = 0;
return (0);
}
static int
vm_handle_inst_emul(struct vm *vm, int vcpuid, bool *retu)
{
struct vie *vie;
struct vcpu *vcpu;
struct vm_exit *vme;
int error, inst_length;
uint64_t rip, gla, gpa, cr3;
enum vie_cpu_mode cpu_mode;
enum vie_paging_mode paging_mode;
mem_region_read_t mread;
mem_region_write_t mwrite;
vcpu = &vm->vcpu[vcpuid];
vme = &vcpu->exitinfo;
rip = vme->rip;
inst_length = vme->inst_length;
gla = vme->u.inst_emul.gla;
gpa = vme->u.inst_emul.gpa;
cr3 = vme->u.inst_emul.cr3;
cpu_mode = vme->u.inst_emul.cpu_mode;
paging_mode = vme->u.inst_emul.paging_mode;
vie = &vme->u.inst_emul.vie;
vie_init(vie);
/* Fetch, decode and emulate the faulting instruction */
if (vmm_fetch_instruction(vm, vcpuid, rip, inst_length, cr3,
paging_mode, vie) != 0)
return (EFAULT);
if (vmm_decode_instruction(vm, vcpuid, gla, cpu_mode, vie) != 0)
return (EFAULT);
/* return to userland unless this is an in-kernel emulated device */
if (gpa >= DEFAULT_APIC_BASE && gpa < DEFAULT_APIC_BASE + PAGE_SIZE) {
mread = lapic_mmio_read;
mwrite = lapic_mmio_write;
} else if (gpa >= VIOAPIC_BASE && gpa < VIOAPIC_BASE + VIOAPIC_SIZE) {
mread = vioapic_mmio_read;
mwrite = vioapic_mmio_write;
} else if (gpa >= VHPET_BASE && gpa < VHPET_BASE + VHPET_SIZE) {
mread = vhpet_mmio_read;
mwrite = vhpet_mmio_write;
} else {
*retu = true;
return (0);
}
error = vmm_emulate_instruction(vm, vcpuid, gpa, vie, mread, mwrite,
retu);
return (error);
}
static int
vm_handle_suspend(struct vm *vm, int vcpuid, bool *retu)
{
int i, done;
struct vcpu *vcpu;
done = 0;
vcpu = &vm->vcpu[vcpuid];
CPU_SET_ATOMIC(vcpuid, &vm->suspended_cpus);
/*
* Wait until all 'active_cpus' have suspended themselves.
*
* Since a VM may be suspended at any time including when one or
* more vcpus are doing a rendezvous we need to call the rendezvous
* handler while we are waiting to prevent a deadlock.
*/
vcpu_lock(vcpu);
while (1) {
if (CPU_CMP(&vm->suspended_cpus, &vm->active_cpus) == 0) {
VCPU_CTR0(vm, vcpuid, "All vcpus suspended");
break;
}
if (vm->rendezvous_func == NULL) {
VCPU_CTR0(vm, vcpuid, "Sleeping during suspend");
vcpu_require_state_locked(vcpu, VCPU_SLEEPING);
msleep_spin(vcpu, &vcpu->mtx, "vmsusp", hz);
vcpu_require_state_locked(vcpu, VCPU_FROZEN);
} else {
VCPU_CTR0(vm, vcpuid, "Rendezvous during suspend");
vcpu_unlock(vcpu);
vm_handle_rendezvous(vm, vcpuid);
vcpu_lock(vcpu);
}
}
vcpu_unlock(vcpu);
/*
* Wakeup the other sleeping vcpus and return to userspace.
*/
for (i = 0; i < VM_MAXCPU; i++) {
if (CPU_ISSET(i, &vm->suspended_cpus)) {
vcpu_notify_event(vm, i, false);
}
}
*retu = true;
return (0);
}
int
vm_suspend(struct vm *vm, enum vm_suspend_how how)
{
int i;
if (how <= VM_SUSPEND_NONE || how >= VM_SUSPEND_LAST)
return (EINVAL);
if (atomic_cmpset_int(&vm->suspend, 0, how) == 0) {
VM_CTR2(vm, "virtual machine already suspended %d/%d",
vm->suspend, how);
return (EALREADY);
}
VM_CTR1(vm, "virtual machine successfully suspended %d", how);
/*
* Notify all active vcpus that they are now suspended.
*/
for (i = 0; i < VM_MAXCPU; i++) {
if (CPU_ISSET(i, &vm->active_cpus))
vcpu_notify_event(vm, i, false);
}
return (0);
}
void
vm_exit_suspended(struct vm *vm, int vcpuid, uint64_t rip)
{
struct vm_exit *vmexit;
KASSERT(vm->suspend > VM_SUSPEND_NONE && vm->suspend < VM_SUSPEND_LAST,
("vm_exit_suspended: invalid suspend type %d", vm->suspend));
vmexit = vm_exitinfo(vm, vcpuid);
vmexit->rip = rip;
vmexit->inst_length = 0;
vmexit->exitcode = VM_EXITCODE_SUSPENDED;
vmexit->u.suspended.how = vm->suspend;
}
int
vm_run(struct vm *vm, struct vm_run *vmrun)
{
int error, vcpuid;
struct vcpu *vcpu;
struct pcb *pcb;
uint64_t tscval, rip;
struct vm_exit *vme;
bool retu, intr_disabled;
pmap_t pmap;
void *rptr, *sptr;
vcpuid = vmrun->cpuid;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
rptr = &vm->rendezvous_func;
sptr = &vm->suspend;
pmap = vmspace_pmap(vm->vmspace);
vcpu = &vm->vcpu[vcpuid];
vme = &vcpu->exitinfo;
rip = vmrun->rip;
restart:
critical_enter();
KASSERT(!CPU_ISSET(curcpu, &pmap->pm_active),
("vm_run: absurd pm_active"));
tscval = rdtsc();
pcb = PCPU_GET(curpcb);
set_pcb_flags(pcb, PCB_FULL_IRET);
restore_guest_msrs(vm, vcpuid);
restore_guest_fpustate(vcpu);
vcpu_require_state(vm, vcpuid, VCPU_RUNNING);
error = VMRUN(vm->cookie, vcpuid, rip, pmap, rptr, sptr);
vcpu_require_state(vm, vcpuid, VCPU_FROZEN);
save_guest_fpustate(vcpu);
restore_host_msrs(vm, vcpuid);
vmm_stat_incr(vm, vcpuid, VCPU_TOTAL_RUNTIME, rdtsc() - tscval);
critical_exit();
if (error == 0) {
retu = false;
switch (vme->exitcode) {
case VM_EXITCODE_SUSPENDED:
error = vm_handle_suspend(vm, vcpuid, &retu);
break;
case VM_EXITCODE_IOAPIC_EOI:
vioapic_process_eoi(vm, vcpuid,
vme->u.ioapic_eoi.vector);
break;
case VM_EXITCODE_RENDEZVOUS:
vm_handle_rendezvous(vm, vcpuid);
error = 0;
break;
case VM_EXITCODE_HLT:
intr_disabled = ((vme->u.hlt.rflags & PSL_I) == 0);
error = vm_handle_hlt(vm, vcpuid, intr_disabled, &retu);
break;
case VM_EXITCODE_PAGING:
error = vm_handle_paging(vm, vcpuid, &retu);
break;
case VM_EXITCODE_INST_EMUL:
error = vm_handle_inst_emul(vm, vcpuid, &retu);
break;
default:
retu = true; /* handled in userland */
break;
}
}
if (error == 0 && retu == false) {
rip = vme->rip + vme->inst_length;
goto restart;
}
/* copy the exit information */
bcopy(vme, &vmrun->vm_exit, sizeof(struct vm_exit));
return (error);
}
int
vm_inject_exception(struct vm *vm, int vcpuid, struct vm_exception *exception)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
if (exception->vector < 0 || exception->vector >= 32)
return (EINVAL);
vcpu = &vm->vcpu[vcpuid];
if (vcpu->exception_pending) {
VCPU_CTR2(vm, vcpuid, "Unable to inject exception %d due to "
"pending exception %d", exception->vector,
vcpu->exception.vector);
return (EBUSY);
}
vcpu->exception_pending = 1;
vcpu->exception = *exception;
VCPU_CTR1(vm, vcpuid, "Exception %d pending", exception->vector);
return (0);
}
int
vm_exception_pending(struct vm *vm, int vcpuid, struct vm_exception *exception)
{
struct vcpu *vcpu;
int pending;
KASSERT(vcpuid >= 0 && vcpuid < VM_MAXCPU, ("invalid vcpu %d", vcpuid));
vcpu = &vm->vcpu[vcpuid];
pending = vcpu->exception_pending;
if (pending) {
vcpu->exception_pending = 0;
*exception = vcpu->exception;
VCPU_CTR1(vm, vcpuid, "Exception %d delivered",
exception->vector);
}
return (pending);
}
static void
vm_inject_fault(struct vm *vm, int vcpuid, struct vm_exception *exception)
{
struct vm_exit *vmexit;
int error;
error = vm_inject_exception(vm, vcpuid, exception);
KASSERT(error == 0, ("vm_inject_exception error %d", error));
/*
* A fault-like exception allows the instruction to be restarted
* after the exception handler returns.
*
* By setting the inst_length to 0 we ensure that the instruction
* pointer remains at the faulting instruction.
*/
vmexit = vm_exitinfo(vm, vcpuid);
vmexit->inst_length = 0;
}
void
vm_inject_gp(struct vm *vm, int vcpuid)
{
struct vm_exception gpf = {
.vector = IDT_GP,
.error_code_valid = 1,
.error_code = 0
};
vm_inject_fault(vm, vcpuid, &gpf);
}
void
vm_inject_ud(struct vm *vm, int vcpuid)
{
struct vm_exception udf = {
.vector = IDT_UD,
.error_code_valid = 0
};
vm_inject_fault(vm, vcpuid, &udf);
}
static VMM_STAT(VCPU_NMI_COUNT, "number of NMIs delivered to vcpu");
int
vm_inject_nmi(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
vcpu = &vm->vcpu[vcpuid];
vcpu->nmi_pending = 1;
vcpu_notify_event(vm, vcpuid, false);
return (0);
}
int
vm_nmi_pending(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_nmi_pending: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
return (vcpu->nmi_pending);
}
void
vm_nmi_clear(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_nmi_pending: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
if (vcpu->nmi_pending == 0)
panic("vm_nmi_clear: inconsistent nmi_pending state");
vcpu->nmi_pending = 0;
vmm_stat_incr(vm, vcpuid, VCPU_NMI_COUNT, 1);
}
static VMM_STAT(VCPU_EXTINT_COUNT, "number of ExtINTs delivered to vcpu");
int
vm_inject_extint(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
vcpu = &vm->vcpu[vcpuid];
vcpu->extint_pending = 1;
vcpu_notify_event(vm, vcpuid, false);
return (0);
}
int
vm_extint_pending(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_extint_pending: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
return (vcpu->extint_pending);
}
void
vm_extint_clear(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_extint_pending: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
if (vcpu->extint_pending == 0)
panic("vm_extint_clear: inconsistent extint_pending state");
vcpu->extint_pending = 0;
vmm_stat_incr(vm, vcpuid, VCPU_EXTINT_COUNT, 1);
}
int
vm_get_capability(struct vm *vm, int vcpu, int type, int *retval)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (type < 0 || type >= VM_CAP_MAX)
return (EINVAL);
return (VMGETCAP(vm->cookie, vcpu, type, retval));
}
int
vm_set_capability(struct vm *vm, int vcpu, int type, int val)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (type < 0 || type >= VM_CAP_MAX)
return (EINVAL);
return (VMSETCAP(vm->cookie, vcpu, type, val));
}
uint64_t *
vm_guest_msrs(struct vm *vm, int cpu)
{
return (vm->vcpu[cpu].guest_msrs);
}
struct vlapic *
vm_lapic(struct vm *vm, int cpu)
{
return (vm->vcpu[cpu].vlapic);
}
struct vioapic *
vm_ioapic(struct vm *vm)
{
return (vm->vioapic);
}
struct vhpet *
vm_hpet(struct vm *vm)
{
return (vm->vhpet);
}
boolean_t
vmm_is_pptdev(int bus, int slot, int func)
{
int found, i, n;
int b, s, f;
char *val, *cp, *cp2;
/*
* XXX
* The length of an environment variable is limited to 128 bytes which
* puts an upper limit on the number of passthru devices that may be
* specified using a single environment variable.
*
* Work around this by scanning multiple environment variable
* names instead of a single one - yuck!
*/
const char *names[] = { "pptdevs", "pptdevs2", "pptdevs3", NULL };
/* set pptdevs="1/2/3 4/5/6 7/8/9 10/11/12" */
found = 0;
for (i = 0; names[i] != NULL && !found; i++) {
cp = val = getenv(names[i]);
while (cp != NULL && *cp != '\0') {
if ((cp2 = strchr(cp, ' ')) != NULL)
*cp2 = '\0';
n = sscanf(cp, "%d/%d/%d", &b, &s, &f);
if (n == 3 && bus == b && slot == s && func == f) {
found = 1;
break;
}
if (cp2 != NULL)
*cp2++ = ' ';
cp = cp2;
}
freeenv(val);
}
return (found);
}
void *
vm_iommu_domain(struct vm *vm)
{
return (vm->iommu);
}
int
vcpu_set_state(struct vm *vm, int vcpuid, enum vcpu_state newstate,
bool from_idle)
{
int error;
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_set_run_state: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
vcpu_lock(vcpu);
error = vcpu_set_state_locked(vcpu, newstate, from_idle);
vcpu_unlock(vcpu);
return (error);
}
enum vcpu_state
vcpu_get_state(struct vm *vm, int vcpuid, int *hostcpu)
{
struct vcpu *vcpu;
enum vcpu_state state;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_get_run_state: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
vcpu_lock(vcpu);
state = vcpu->state;
if (hostcpu != NULL)
*hostcpu = vcpu->hostcpu;
vcpu_unlock(vcpu);
return (state);
}
void
vm_activate_cpu(struct vm *vm, int vcpuid)
{
KASSERT(vcpuid >= 0 && vcpuid < VM_MAXCPU,
("vm_activate_cpu: invalid vcpuid %d", vcpuid));
KASSERT(!CPU_ISSET(vcpuid, &vm->active_cpus),
("vm_activate_cpu: vcpuid %d is already active", vcpuid));
VCPU_CTR0(vm, vcpuid, "activated");
CPU_SET_ATOMIC(vcpuid, &vm->active_cpus);
}
cpuset_t
vm_active_cpus(struct vm *vm)
{
return (vm->active_cpus);
}
void *
vcpu_stats(struct vm *vm, int vcpuid)
{
return (vm->vcpu[vcpuid].stats);
}
int
vm_get_x2apic_state(struct vm *vm, int vcpuid, enum x2apic_state *state)
{
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
*state = vm->vcpu[vcpuid].x2apic_state;
return (0);
}
int
vm_set_x2apic_state(struct vm *vm, int vcpuid, enum x2apic_state state)
{
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
if (state >= X2APIC_STATE_LAST)
return (EINVAL);
vm->vcpu[vcpuid].x2apic_state = state;
vlapic_set_x2apic_state(vm, vcpuid, state);
return (0);
}
/*
* This function is called to ensure that a vcpu "sees" a pending event
* as soon as possible:
* - If the vcpu thread is sleeping then it is woken up.
* - If the vcpu is running on a different host_cpu then an IPI will be directed
* to the host_cpu to cause the vcpu to trap into the hypervisor.
*/
void
vcpu_notify_event(struct vm *vm, int vcpuid, bool lapic_intr)
{
int hostcpu;
struct vcpu *vcpu;
vcpu = &vm->vcpu[vcpuid];
vcpu_lock(vcpu);
hostcpu = vcpu->hostcpu;
if (vcpu->state == VCPU_RUNNING) {
KASSERT(hostcpu != NOCPU, ("vcpu running on invalid hostcpu"));
if (hostcpu != curcpu) {
if (lapic_intr) {
vlapic_post_intr(vcpu->vlapic, hostcpu,
vmm_ipinum);
} else {
ipi_cpu(hostcpu, vmm_ipinum);
}
} else {
/*
* If the 'vcpu' is running on 'curcpu' then it must
* be sending a notification to itself (e.g. SELF_IPI).
* The pending event will be picked up when the vcpu
* transitions back to guest context.
*/
}
} else {
KASSERT(hostcpu == NOCPU, ("vcpu state %d not consistent "
"with hostcpu %d", vcpu->state, hostcpu));
if (vcpu->state == VCPU_SLEEPING)
wakeup_one(vcpu);
}
vcpu_unlock(vcpu);
}
struct vmspace *
vm_get_vmspace(struct vm *vm)
{
return (vm->vmspace);
}
int
vm_apicid2vcpuid(struct vm *vm, int apicid)
{
/*
* XXX apic id is assumed to be numerically identical to vcpu id
*/
return (apicid);
}
void
vm_smp_rendezvous(struct vm *vm, int vcpuid, cpuset_t dest,
vm_rendezvous_func_t func, void *arg)
{
int i;
/*
* Enforce that this function is called without any locks
*/
WITNESS_WARN(WARN_PANIC, NULL, "vm_smp_rendezvous");
KASSERT(vcpuid == -1 || (vcpuid >= 0 && vcpuid < VM_MAXCPU),
("vm_smp_rendezvous: invalid vcpuid %d", vcpuid));
restart:
mtx_lock(&vm->rendezvous_mtx);
if (vm->rendezvous_func != NULL) {
/*
* If a rendezvous is already in progress then we need to
* call the rendezvous handler in case this 'vcpuid' is one
* of the targets of the rendezvous.
*/
RENDEZVOUS_CTR0(vm, vcpuid, "Rendezvous already in progress");
mtx_unlock(&vm->rendezvous_mtx);
vm_handle_rendezvous(vm, vcpuid);
goto restart;
}
KASSERT(vm->rendezvous_func == NULL, ("vm_smp_rendezvous: previous "
"rendezvous is still in progress"));
RENDEZVOUS_CTR0(vm, vcpuid, "Initiating rendezvous");
vm->rendezvous_req_cpus = dest;
CPU_ZERO(&vm->rendezvous_done_cpus);
vm->rendezvous_arg = arg;
vm_set_rendezvous_func(vm, func);
mtx_unlock(&vm->rendezvous_mtx);
/*
* Wake up any sleeping vcpus and trigger a VM-exit in any running
* vcpus so they handle the rendezvous as soon as possible.
*/
for (i = 0; i < VM_MAXCPU; i++) {
if (CPU_ISSET(i, &dest))
vcpu_notify_event(vm, i, false);
}
vm_handle_rendezvous(vm, vcpuid);
}
struct vatpic *
vm_atpic(struct vm *vm)
{
return (vm->vatpic);
}
struct vatpit *
vm_atpit(struct vm *vm)
{
return (vm->vatpit);
}
|