1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
|
/* $OpenBSD: vm.c,v 1.64 2021/07/16 16:21:22 dv Exp $ */
/*
* Copyright (c) 2015 Mike Larkin <mlarkin@openbsd.org>
*
* Permission to use, copy, modify, and distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
* copyright notice and this permission notice appear in all copies.
*
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
*/
#include <sys/types.h>
#include <sys/ioctl.h>
#include <sys/queue.h>
#include <sys/wait.h>
#include <sys/uio.h>
#include <sys/stat.h>
#include <sys/socket.h>
#include <sys/time.h>
#include <sys/mman.h>
#include <dev/ic/i8253reg.h>
#include <dev/isa/isareg.h>
#include <dev/pci/pcireg.h>
#include <machine/param.h>
#include <machine/psl.h>
#include <machine/pte.h>
#include <machine/specialreg.h>
#include <machine/vmmvar.h>
#include <net/if.h>
#include <errno.h>
#include <event.h>
#include <fcntl.h>
#include <imsg.h>
#include <limits.h>
#include <poll.h>
#include <pthread.h>
#include <stddef.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <util.h>
#include "atomicio.h"
#include "fw_cfg.h"
#include "i8253.h"
#include "i8259.h"
#include "loadfile.h"
#include "mc146818.h"
#include "ns8250.h"
#include "pci.h"
#include "virtio.h"
#include "vmd.h"
#include "vmm.h"
io_fn_t ioports_map[MAX_PORTS];
int run_vm(int, int[][VM_MAX_BASE_PER_DISK], int *,
struct vmop_create_params *, struct vcpu_reg_state *);
void vm_dispatch_vmm(int, short, void *);
void *event_thread(void *);
void *vcpu_run_loop(void *);
int vcpu_exit(struct vm_run_params *);
int vcpu_reset(uint32_t, uint32_t, struct vcpu_reg_state *);
void create_memory_map(struct vm_create_params *);
int alloc_guest_mem(struct vm_create_params *);
int vmm_create_vm(struct vm_create_params *);
void init_emulated_hw(struct vmop_create_params *, int,
int[][VM_MAX_BASE_PER_DISK], int *);
void restore_emulated_hw(struct vm_create_params *, int, int *,
int[][VM_MAX_BASE_PER_DISK],int);
void vcpu_exit_inout(struct vm_run_params *);
int vcpu_exit_eptviolation(struct vm_run_params *);
uint8_t vcpu_exit_pci(struct vm_run_params *);
int vcpu_pic_intr(uint32_t, uint32_t, uint8_t);
int loadfile_bios(gzFile, off_t, struct vcpu_reg_state *);
int send_vm(int, struct vm_create_params *);
int dump_send_header(int);
int dump_vmr(int , struct vm_mem_range *);
int dump_mem(int, struct vm_create_params *);
void restore_vmr(int, struct vm_mem_range *);
void restore_mem(int, struct vm_create_params *);
int restore_vm_params(int, struct vm_create_params *);
void pause_vm(struct vm_create_params *);
void unpause_vm(struct vm_create_params *);
int translate_gva(struct vm_exit*, uint64_t, uint64_t *, int);
static struct vm_mem_range *find_gpa_range(struct vm_create_params *, paddr_t,
size_t);
int con_fd;
struct vmd_vm *current_vm;
extern struct vmd *env;
extern char *__progname;
pthread_mutex_t threadmutex;
pthread_cond_t threadcond;
pthread_cond_t vcpu_run_cond[VMM_MAX_VCPUS_PER_VM];
pthread_mutex_t vcpu_run_mtx[VMM_MAX_VCPUS_PER_VM];
pthread_barrier_t vm_pause_barrier;
pthread_cond_t vcpu_unpause_cond[VMM_MAX_VCPUS_PER_VM];
pthread_mutex_t vcpu_unpause_mtx[VMM_MAX_VCPUS_PER_VM];
uint8_t vcpu_hlt[VMM_MAX_VCPUS_PER_VM];
uint8_t vcpu_done[VMM_MAX_VCPUS_PER_VM];
/*
* Represents a standard register set for an OS to be booted
* as a flat 64 bit address space.
*
* NOT set here are:
* RIP
* RSP
* GDTR BASE
*
* Specific bootloaders should clone this structure and override
* those fields as needed.
*
* Note - CR3 and various bits in CR0 may be overridden by vmm(4) based on
* features of the CPU in use.
*/
static const struct vcpu_reg_state vcpu_init_flat64 = {
.vrs_gprs[VCPU_REGS_RFLAGS] = 0x2,
.vrs_gprs[VCPU_REGS_RIP] = 0x0,
.vrs_gprs[VCPU_REGS_RSP] = 0x0,
.vrs_crs[VCPU_REGS_CR0] = CR0_ET | CR0_PE | CR0_PG,
.vrs_crs[VCPU_REGS_CR3] = PML4_PAGE,
.vrs_crs[VCPU_REGS_CR4] = CR4_PAE | CR4_PSE,
.vrs_crs[VCPU_REGS_PDPTE0] = 0ULL,
.vrs_crs[VCPU_REGS_PDPTE1] = 0ULL,
.vrs_crs[VCPU_REGS_PDPTE2] = 0ULL,
.vrs_crs[VCPU_REGS_PDPTE3] = 0ULL,
.vrs_sregs[VCPU_REGS_CS] = { 0x8, 0xFFFFFFFF, 0xC09F, 0x0},
.vrs_sregs[VCPU_REGS_DS] = { 0x10, 0xFFFFFFFF, 0xC093, 0x0},
.vrs_sregs[VCPU_REGS_ES] = { 0x10, 0xFFFFFFFF, 0xC093, 0x0},
.vrs_sregs[VCPU_REGS_FS] = { 0x10, 0xFFFFFFFF, 0xC093, 0x0},
.vrs_sregs[VCPU_REGS_GS] = { 0x10, 0xFFFFFFFF, 0xC093, 0x0},
.vrs_sregs[VCPU_REGS_SS] = { 0x10, 0xFFFFFFFF, 0xC093, 0x0},
.vrs_gdtr = { 0x0, 0xFFFF, 0x0, 0x0},
.vrs_idtr = { 0x0, 0xFFFF, 0x0, 0x0},
.vrs_sregs[VCPU_REGS_LDTR] = { 0x0, 0xFFFF, 0x0082, 0x0},
.vrs_sregs[VCPU_REGS_TR] = { 0x0, 0xFFFF, 0x008B, 0x0},
.vrs_msrs[VCPU_REGS_EFER] = EFER_LME | EFER_LMA,
.vrs_drs[VCPU_REGS_DR0] = 0x0,
.vrs_drs[VCPU_REGS_DR1] = 0x0,
.vrs_drs[VCPU_REGS_DR2] = 0x0,
.vrs_drs[VCPU_REGS_DR3] = 0x0,
.vrs_drs[VCPU_REGS_DR6] = 0xFFFF0FF0,
.vrs_drs[VCPU_REGS_DR7] = 0x400,
.vrs_msrs[VCPU_REGS_STAR] = 0ULL,
.vrs_msrs[VCPU_REGS_LSTAR] = 0ULL,
.vrs_msrs[VCPU_REGS_CSTAR] = 0ULL,
.vrs_msrs[VCPU_REGS_SFMASK] = 0ULL,
.vrs_msrs[VCPU_REGS_KGSBASE] = 0ULL,
.vrs_msrs[VCPU_REGS_MISC_ENABLE] = 0ULL,
.vrs_crs[VCPU_REGS_XCR0] = XCR0_X87
};
/*
* Represents a standard register set for an BIOS to be booted
* as a flat 16 bit address space.
*/
static const struct vcpu_reg_state vcpu_init_flat16 = {
.vrs_gprs[VCPU_REGS_RFLAGS] = 0x2,
.vrs_gprs[VCPU_REGS_RIP] = 0xFFF0,
.vrs_gprs[VCPU_REGS_RSP] = 0x0,
.vrs_crs[VCPU_REGS_CR0] = 0x60000010,
.vrs_crs[VCPU_REGS_CR3] = 0,
.vrs_sregs[VCPU_REGS_CS] = { 0xF000, 0xFFFF, 0x809F, 0xF0000},
.vrs_sregs[VCPU_REGS_DS] = { 0x0, 0xFFFF, 0x8093, 0x0},
.vrs_sregs[VCPU_REGS_ES] = { 0x0, 0xFFFF, 0x8093, 0x0},
.vrs_sregs[VCPU_REGS_FS] = { 0x0, 0xFFFF, 0x8093, 0x0},
.vrs_sregs[VCPU_REGS_GS] = { 0x0, 0xFFFF, 0x8093, 0x0},
.vrs_sregs[VCPU_REGS_SS] = { 0x0, 0xFFFF, 0x8093, 0x0},
.vrs_gdtr = { 0x0, 0xFFFF, 0x0, 0x0},
.vrs_idtr = { 0x0, 0xFFFF, 0x0, 0x0},
.vrs_sregs[VCPU_REGS_LDTR] = { 0x0, 0xFFFF, 0x0082, 0x0},
.vrs_sregs[VCPU_REGS_TR] = { 0x0, 0xFFFF, 0x008B, 0x0},
.vrs_msrs[VCPU_REGS_EFER] = 0ULL,
.vrs_drs[VCPU_REGS_DR0] = 0x0,
.vrs_drs[VCPU_REGS_DR1] = 0x0,
.vrs_drs[VCPU_REGS_DR2] = 0x0,
.vrs_drs[VCPU_REGS_DR3] = 0x0,
.vrs_drs[VCPU_REGS_DR6] = 0xFFFF0FF0,
.vrs_drs[VCPU_REGS_DR7] = 0x400,
.vrs_msrs[VCPU_REGS_STAR] = 0ULL,
.vrs_msrs[VCPU_REGS_LSTAR] = 0ULL,
.vrs_msrs[VCPU_REGS_CSTAR] = 0ULL,
.vrs_msrs[VCPU_REGS_SFMASK] = 0ULL,
.vrs_msrs[VCPU_REGS_KGSBASE] = 0ULL,
.vrs_crs[VCPU_REGS_XCR0] = XCR0_X87
};
/*
* loadfile_bios
*
* Alternatively to loadfile_elf, this function loads a non-ELF BIOS image
* directly into memory.
*
* Parameters:
* fp: file of a kernel file to load
* size: uncompressed size of the image
* (out) vrs: register state to set on init for this kernel
*
* Return values:
* 0 if successful
* various error codes returned from read(2) or loadelf functions
*/
int
loadfile_bios(gzFile fp, off_t size, struct vcpu_reg_state *vrs)
{
off_t off;
/* Set up a "flat 16 bit" register state for BIOS */
memcpy(vrs, &vcpu_init_flat16, sizeof(*vrs));
/* Seek to the beginning of the BIOS image */
if (gzseek(fp, 0, SEEK_SET) == -1)
return (-1);
/* The BIOS image must end at 1M */
if ((off = 1048576 - size) < 0)
return (-1);
/* Read BIOS image into memory */
if (mread(fp, off, size) != (size_t)size) {
errno = EIO;
return (-1);
}
log_debug("%s: loaded BIOS image", __func__);
return (0);
}
/*
* start_vm
*
* After forking a new VM process, starts the new VM with the creation
* parameters supplied (in the incoming vm->vm_params field). This
* function performs a basic sanity check on the incoming parameters
* and then performs the following steps to complete the creation of the VM:
*
* 1. validates and create the new VM
* 2. opens the imsg control channel to the parent and drops more privilege
* 3. drops additional privleges by calling pledge(2)
* 4. loads the kernel from the disk image or file descriptor
* 5. runs the VM's VCPU loops.
*
* Parameters:
* vm: The VM data structure that is including the VM create parameters.
* fd: The imsg socket that is connected to the parent process.
*
* Return values:
* 0: success
* !0 : failure - typically an errno indicating the source of the failure
*/
int
start_vm(struct vmd_vm *vm, int fd)
{
struct vmop_create_params *vmc = &vm->vm_params;
struct vm_create_params *vcp = &vmc->vmc_params;
struct vcpu_reg_state vrs;
int nicfds[VMM_MAX_NICS_PER_VM];
int ret;
gzFile fp;
size_t i;
struct vm_rwregs_params vrp;
struct stat sb;
/* Child */
setproctitle("%s", vcp->vcp_name);
log_procinit(vcp->vcp_name);
if (!(vm->vm_state & VM_STATE_RECEIVED))
create_memory_map(vcp);
ret = alloc_guest_mem(vcp);
if (ret) {
errno = ret;
fatal("could not allocate guest memory - exiting");
}
ret = vmm_create_vm(vcp);
current_vm = vm;
/* send back the kernel-generated vm id (0 on error) */
if (write(fd, &vcp->vcp_id, sizeof(vcp->vcp_id)) !=
sizeof(vcp->vcp_id))
fatal("write vcp id");
if (ret) {
errno = ret;
fatal("create vmm ioctl failed - exiting");
}
/*
* pledge in the vm processes:
* stdio - for malloc and basic I/O including events.
* recvfd - for send/recv.
* vmm - for the vmm ioctls and operations.
*/
if (pledge("stdio vmm recvfd", NULL) == -1)
fatal("pledge");
if (vm->vm_state & VM_STATE_RECEIVED) {
ret = read(vm->vm_receive_fd, &vrp, sizeof(vrp));
if (ret != sizeof(vrp)) {
fatal("received incomplete vrp - exiting");
}
vrs = vrp.vrwp_regs;
} else {
/*
* Set up default "flat 64 bit" register state - RIP,
* RSP, and GDT info will be set in bootloader
*/
memcpy(&vrs, &vcpu_init_flat64, sizeof(vrs));
/* Find and open kernel image */
if ((fp = gzdopen(vm->vm_kernel, "r")) == NULL)
fatalx("failed to open kernel - exiting");
/* Load kernel image */
ret = loadfile_elf(fp, vcp, &vrs);
/*
* Try BIOS as a fallback (only if it was provided as an image
* with vm->vm_kernel and the file is not compressed)
*/
if (ret && errno == ENOEXEC && vm->vm_kernel != -1 &&
gzdirect(fp) && (ret = fstat(vm->vm_kernel, &sb)) == 0)
ret = loadfile_bios(fp, sb.st_size, &vrs);
if (ret)
fatal("failed to load kernel or BIOS - exiting");
gzclose(fp);
}
if (vm->vm_kernel != -1)
close(vm->vm_kernel);
con_fd = vm->vm_tty;
if (fcntl(con_fd, F_SETFL, O_NONBLOCK) == -1)
fatal("failed to set nonblocking mode on console");
for (i = 0; i < VMM_MAX_NICS_PER_VM; i++)
nicfds[i] = vm->vm_ifs[i].vif_fd;
event_init();
if (vm->vm_state & VM_STATE_RECEIVED) {
restore_emulated_hw(vcp, vm->vm_receive_fd, nicfds,
vm->vm_disks, vm->vm_cdrom);
restore_mem(vm->vm_receive_fd, vcp);
if (restore_vm_params(vm->vm_receive_fd, vcp))
fatal("restore vm params failed");
unpause_vm(vcp);
}
if (vmm_pipe(vm, fd, vm_dispatch_vmm) == -1)
fatal("setup vm pipe");
/* Execute the vcpu run loop(s) for this VM */
ret = run_vm(vm->vm_cdrom, vm->vm_disks, nicfds, &vm->vm_params, &vrs);
/* Ensure that any in-flight data is written back */
virtio_shutdown(vm);
return (ret);
}
/*
* vm_dispatch_vmm
*
* imsg callback for messages that are received from the vmm parent process.
*/
void
vm_dispatch_vmm(int fd, short event, void *arg)
{
struct vmd_vm *vm = arg;
struct vmop_result vmr;
struct vmop_addr_result var;
struct imsgev *iev = &vm->vm_iev;
struct imsgbuf *ibuf = &iev->ibuf;
struct imsg imsg;
ssize_t n;
int verbose;
if (event & EV_READ) {
if ((n = imsg_read(ibuf)) == -1 && errno != EAGAIN)
fatal("%s: imsg_read", __func__);
if (n == 0)
_exit(0);
}
if (event & EV_WRITE) {
if ((n = msgbuf_write(&ibuf->w)) == -1 && errno != EAGAIN)
fatal("%s: msgbuf_write fd %d", __func__, ibuf->fd);
if (n == 0)
_exit(0);
}
for (;;) {
if ((n = imsg_get(ibuf, &imsg)) == -1)
fatal("%s: imsg_get", __func__);
if (n == 0)
break;
#if DEBUG > 1
log_debug("%s: got imsg %d from %s",
__func__, imsg.hdr.type,
vm->vm_params.vmc_params.vcp_name);
#endif
switch (imsg.hdr.type) {
case IMSG_CTL_VERBOSE:
IMSG_SIZE_CHECK(&imsg, &verbose);
memcpy(&verbose, imsg.data, sizeof(verbose));
log_setverbose(verbose);
break;
case IMSG_VMDOP_VM_SHUTDOWN:
if (vmmci_ctl(VMMCI_SHUTDOWN) == -1)
_exit(0);
break;
case IMSG_VMDOP_VM_REBOOT:
if (vmmci_ctl(VMMCI_REBOOT) == -1)
_exit(0);
break;
case IMSG_VMDOP_PAUSE_VM:
vmr.vmr_result = 0;
vmr.vmr_id = vm->vm_vmid;
pause_vm(&vm->vm_params.vmc_params);
imsg_compose_event(&vm->vm_iev,
IMSG_VMDOP_PAUSE_VM_RESPONSE,
imsg.hdr.peerid, imsg.hdr.pid, -1, &vmr,
sizeof(vmr));
break;
case IMSG_VMDOP_UNPAUSE_VM:
vmr.vmr_result = 0;
vmr.vmr_id = vm->vm_vmid;
unpause_vm(&vm->vm_params.vmc_params);
imsg_compose_event(&vm->vm_iev,
IMSG_VMDOP_UNPAUSE_VM_RESPONSE,
imsg.hdr.peerid, imsg.hdr.pid, -1, &vmr,
sizeof(vmr));
break;
case IMSG_VMDOP_SEND_VM_REQUEST:
vmr.vmr_id = vm->vm_vmid;
vmr.vmr_result = send_vm(imsg.fd,
&vm->vm_params.vmc_params);
imsg_compose_event(&vm->vm_iev,
IMSG_VMDOP_SEND_VM_RESPONSE,
imsg.hdr.peerid, imsg.hdr.pid, -1, &vmr,
sizeof(vmr));
if (!vmr.vmr_result) {
imsg_flush(¤t_vm->vm_iev.ibuf);
_exit(0);
}
break;
case IMSG_VMDOP_PRIV_GET_ADDR_RESPONSE:
IMSG_SIZE_CHECK(&imsg, &var);
memcpy(&var, imsg.data, sizeof(var));
log_debug("%s: received tap addr %s for nic %d",
vm->vm_params.vmc_params.vcp_name,
ether_ntoa((void *)var.var_addr), var.var_nic_idx);
vionet_set_hostmac(vm, var.var_nic_idx, var.var_addr);
break;
default:
fatalx("%s: got invalid imsg %d from %s",
__func__, imsg.hdr.type,
vm->vm_params.vmc_params.vcp_name);
}
imsg_free(&imsg);
}
imsg_event_add(iev);
}
/*
* vm_shutdown
*
* Tell the vmm parent process to shutdown or reboot the VM and exit.
*/
__dead void
vm_shutdown(unsigned int cmd)
{
switch (cmd) {
case VMMCI_NONE:
case VMMCI_SHUTDOWN:
(void)imsg_compose_event(¤t_vm->vm_iev,
IMSG_VMDOP_VM_SHUTDOWN, 0, 0, -1, NULL, 0);
break;
case VMMCI_REBOOT:
(void)imsg_compose_event(¤t_vm->vm_iev,
IMSG_VMDOP_VM_REBOOT, 0, 0, -1, NULL, 0);
break;
default:
fatalx("invalid vm ctl command: %d", cmd);
}
imsg_flush(¤t_vm->vm_iev.ibuf);
_exit(0);
}
int
send_vm(int fd, struct vm_create_params *vcp)
{
struct vm_rwregs_params vrp;
struct vm_rwvmparams_params vpp;
struct vmop_create_params *vmc;
struct vm_terminate_params vtp;
unsigned int flags = 0;
unsigned int i;
int ret = 0;
size_t sz;
if (dump_send_header(fd)) {
log_info("%s: failed to send vm dump header", __func__);
goto err;
}
pause_vm(vcp);
vmc = calloc(1, sizeof(struct vmop_create_params));
if (vmc == NULL) {
log_warn("%s: calloc error geting vmc", __func__);
ret = -1;
goto err;
}
flags |= VMOP_CREATE_MEMORY;
memcpy(&vmc->vmc_params, ¤t_vm->vm_params, sizeof(struct
vmop_create_params));
vmc->vmc_flags = flags;
vrp.vrwp_vm_id = vcp->vcp_id;
vrp.vrwp_mask = VM_RWREGS_ALL;
vpp.vpp_mask = VM_RWVMPARAMS_ALL;
vpp.vpp_vm_id = vcp->vcp_id;
sz = atomicio(vwrite, fd, vmc,sizeof(struct vmop_create_params));
if (sz != sizeof(struct vmop_create_params)) {
ret = -1;
goto err;
}
for (i = 0; i < vcp->vcp_ncpus; i++) {
vrp.vrwp_vcpu_id = i;
if ((ret = ioctl(env->vmd_fd, VMM_IOC_READREGS, &vrp))) {
log_warn("%s: readregs failed", __func__);
goto err;
}
sz = atomicio(vwrite, fd, &vrp,
sizeof(struct vm_rwregs_params));
if (sz != sizeof(struct vm_rwregs_params)) {
log_warn("%s: dumping registers failed", __func__);
ret = -1;
goto err;
}
}
if ((ret = i8253_dump(fd)))
goto err;
if ((ret = i8259_dump(fd)))
goto err;
if ((ret = ns8250_dump(fd)))
goto err;
if ((ret = mc146818_dump(fd)))
goto err;
if ((ret = fw_cfg_dump(fd)))
goto err;
if ((ret = pci_dump(fd)))
goto err;
if ((ret = virtio_dump(fd)))
goto err;
if ((ret = dump_mem(fd, vcp)))
goto err;
for (i = 0; i < vcp->vcp_ncpus; i++) {
vpp.vpp_vcpu_id = i;
if ((ret = ioctl(env->vmd_fd, VMM_IOC_READVMPARAMS, &vpp))) {
log_warn("%s: readvmparams failed", __func__);
goto err;
}
sz = atomicio(vwrite, fd, &vpp,
sizeof(struct vm_rwvmparams_params));
if (sz != sizeof(struct vm_rwvmparams_params)) {
log_warn("%s: dumping vm params failed", __func__);
ret = -1;
goto err;
}
}
vtp.vtp_vm_id = vcp->vcp_id;
if (ioctl(env->vmd_fd, VMM_IOC_TERM, &vtp) == -1) {
log_warnx("%s: term IOC error: %d, %d", __func__,
errno, ENOENT);
}
err:
close(fd);
if (ret)
unpause_vm(vcp);
return ret;
}
int
dump_send_header(int fd) {
struct vm_dump_header vmh;
int i;
memcpy(&vmh.vmh_signature, VM_DUMP_SIGNATURE,
sizeof(vmh.vmh_signature));
vmh.vmh_cpuids[0].code = 0x00;
vmh.vmh_cpuids[0].leaf = 0x00;
vmh.vmh_cpuids[1].code = 0x01;
vmh.vmh_cpuids[1].leaf = 0x00;
vmh.vmh_cpuids[2].code = 0x07;
vmh.vmh_cpuids[2].leaf = 0x00;
vmh.vmh_cpuids[3].code = 0x0d;
vmh.vmh_cpuids[3].leaf = 0x00;
vmh.vmh_cpuids[4].code = 0x80000001;
vmh.vmh_cpuids[4].leaf = 0x00;
vmh.vmh_version = VM_DUMP_VERSION;
for (i=0; i < VM_DUMP_HEADER_CPUID_COUNT; i++) {
CPUID_LEAF(vmh.vmh_cpuids[i].code,
vmh.vmh_cpuids[i].leaf,
vmh.vmh_cpuids[i].a,
vmh.vmh_cpuids[i].b,
vmh.vmh_cpuids[i].c,
vmh.vmh_cpuids[i].d);
}
if (atomicio(vwrite, fd, &vmh, sizeof(vmh)) != sizeof(vmh))
return (-1);
return (0);
}
int
dump_mem(int fd, struct vm_create_params *vcp)
{
unsigned int i;
int ret;
struct vm_mem_range *vmr;
for (i = 0; i < vcp->vcp_nmemranges; i++) {
vmr = &vcp->vcp_memranges[i];
ret = dump_vmr(fd, vmr);
if (ret)
return ret;
}
return (0);
}
int
restore_vm_params(int fd, struct vm_create_params *vcp) {
unsigned int i;
struct vm_rwvmparams_params vpp;
for (i = 0; i < vcp->vcp_ncpus; i++) {
if (atomicio(read, fd, &vpp, sizeof(vpp)) != sizeof(vpp)) {
log_warn("%s: error restoring vm params", __func__);
return (-1);
}
vpp.vpp_vm_id = vcp->vcp_id;
vpp.vpp_vcpu_id = i;
if (ioctl(env->vmd_fd, VMM_IOC_WRITEVMPARAMS, &vpp) < 0) {
log_debug("%s: writing vm params failed", __func__);
return (-1);
}
}
return (0);
}
void
restore_mem(int fd, struct vm_create_params *vcp)
{
unsigned int i;
struct vm_mem_range *vmr;
for (i = 0; i < vcp->vcp_nmemranges; i++) {
vmr = &vcp->vcp_memranges[i];
restore_vmr(fd, vmr);
}
}
int
dump_vmr(int fd, struct vm_mem_range *vmr)
{
size_t rem = vmr->vmr_size, read=0;
char buf[PAGE_SIZE];
while (rem > 0) {
if (read_mem(vmr->vmr_gpa + read, buf, PAGE_SIZE)) {
log_warn("failed to read vmr");
return (-1);
}
if (atomicio(vwrite, fd, buf, sizeof(buf)) != sizeof(buf)) {
log_warn("failed to dump vmr");
return (-1);
}
rem = rem - PAGE_SIZE;
read = read + PAGE_SIZE;
}
return (0);
}
void
restore_vmr(int fd, struct vm_mem_range *vmr)
{
size_t rem = vmr->vmr_size, wrote=0;
char buf[PAGE_SIZE];
while (rem > 0) {
if (atomicio(read, fd, buf, sizeof(buf)) != sizeof(buf))
fatal("failed to restore vmr");
if (write_mem(vmr->vmr_gpa + wrote, buf, PAGE_SIZE))
fatal("failed to write vmr");
rem = rem - PAGE_SIZE;
wrote = wrote + PAGE_SIZE;
}
}
void
pause_vm(struct vm_create_params *vcp)
{
unsigned int n;
int ret;
if (current_vm->vm_state & VM_STATE_PAUSED)
return;
current_vm->vm_state |= VM_STATE_PAUSED;
ret = pthread_barrier_init(&vm_pause_barrier, NULL, vcp->vcp_ncpus + 1);
if (ret) {
log_warnx("%s: cannot initialize pause barrier (%d)",
__progname, ret);
return;
}
for (n = 0; n < vcp->vcp_ncpus; n++) {
ret = pthread_cond_broadcast(&vcpu_run_cond[n]);
if (ret) {
log_warnx("%s: can't broadcast vcpu run cond (%d)",
__func__, (int)ret);
return;
}
}
ret = pthread_barrier_wait(&vm_pause_barrier);
if (ret != 0 && ret != PTHREAD_BARRIER_SERIAL_THREAD) {
log_warnx("%s: could not wait on pause barrier (%d)",
__func__, (int)ret);
return;
}
ret = pthread_barrier_destroy(&vm_pause_barrier);
if (ret) {
log_warnx("%s: could not destroy pause barrier (%d)",
__progname, ret);
return;
}
i8253_stop();
mc146818_stop();
ns8250_stop();
virtio_stop(vcp);
}
void
unpause_vm(struct vm_create_params *vcp)
{
unsigned int n;
int ret;
if (!(current_vm->vm_state & VM_STATE_PAUSED))
return;
current_vm->vm_state &= ~VM_STATE_PAUSED;
for (n = 0; n < vcp->vcp_ncpus; n++) {
ret = pthread_cond_broadcast(&vcpu_unpause_cond[n]);
if (ret) {
log_warnx("%s: can't broadcast vcpu unpause cond (%d)",
__func__, (int)ret);
return;
}
}
i8253_start();
mc146818_start();
ns8250_start();
virtio_start(vcp);
}
/*
* vcpu_reset
*
* Requests vmm(4) to reset the VCPUs in the indicated VM to
* the register state provided
*
* Parameters
* vmid: VM ID to reset
* vcpu_id: VCPU ID to reset
* vrs: the register state to initialize
*
* Return values:
* 0: success
* !0 : ioctl to vmm(4) failed (eg, ENOENT if the supplied VM ID is not
* valid)
*/
int
vcpu_reset(uint32_t vmid, uint32_t vcpu_id, struct vcpu_reg_state *vrs)
{
struct vm_resetcpu_params vrp;
memset(&vrp, 0, sizeof(vrp));
vrp.vrp_vm_id = vmid;
vrp.vrp_vcpu_id = vcpu_id;
memcpy(&vrp.vrp_init_state, vrs, sizeof(struct vcpu_reg_state));
log_debug("%s: resetting vcpu %d for vm %d", __func__, vcpu_id, vmid);
if (ioctl(env->vmd_fd, VMM_IOC_RESETCPU, &vrp) == -1)
return (errno);
return (0);
}
/*
* create_memory_map
*
* Sets up the guest physical memory ranges that the VM can access.
*
* Parameters:
* vcp: VM create parameters describing the VM whose memory map
* is being created
*
* Return values:
* nothing
*/
void
create_memory_map(struct vm_create_params *vcp)
{
size_t len, mem_bytes, mem_mb;
mem_mb = vcp->vcp_memranges[0].vmr_size;
vcp->vcp_nmemranges = 0;
if (mem_mb < 1 || mem_mb > VMM_MAX_VM_MEM_SIZE)
return;
mem_bytes = mem_mb * 1024 * 1024;
/* First memory region: 0 - LOWMEM_KB (DOS low mem) */
len = LOWMEM_KB * 1024;
vcp->vcp_memranges[0].vmr_gpa = 0x0;
vcp->vcp_memranges[0].vmr_size = len;
mem_bytes -= len;
/*
* Second memory region: LOWMEM_KB - 1MB.
*
* N.B. - Normally ROMs or parts of video RAM are mapped here.
* We have to add this region, because some systems
* unconditionally write to 0xb8000 (VGA RAM), and
* we need to make sure that vmm(4) permits accesses
* to it. So allocate guest memory for it.
*/
len = 0x100000 - LOWMEM_KB * 1024;
vcp->vcp_memranges[1].vmr_gpa = LOWMEM_KB * 1024;
vcp->vcp_memranges[1].vmr_size = len;
mem_bytes -= len;
/* Make sure that we do not place physical memory into MMIO ranges. */
if (mem_bytes > VMM_PCI_MMIO_BAR_BASE - 0x100000)
len = VMM_PCI_MMIO_BAR_BASE - 0x100000;
else
len = mem_bytes;
/* Third memory region: 1MB - (1MB + len) */
vcp->vcp_memranges[2].vmr_gpa = 0x100000;
vcp->vcp_memranges[2].vmr_size = len;
mem_bytes -= len;
if (mem_bytes > 0) {
/* Fourth memory region for the remaining memory (if any) */
vcp->vcp_memranges[3].vmr_gpa = VMM_PCI_MMIO_BAR_END + 1;
vcp->vcp_memranges[3].vmr_size = mem_bytes;
vcp->vcp_nmemranges = 4;
} else
vcp->vcp_nmemranges = 3;
}
/*
* alloc_guest_mem
*
* Allocates memory for the guest.
* Instead of doing a single allocation with one mmap(), we allocate memory
* separately for every range for the following reasons:
* - ASLR for the individual ranges
* - to reduce memory consumption in the UVM subsystem: if vmm(4) had to
* map the single mmap'd userspace memory to the individual guest physical
* memory ranges, the underlying amap of the single mmap'd range would have
* to allocate per-page reference counters. The reason is that the
* individual guest physical ranges would reference the single mmap'd region
* only partially. However, if every guest physical range has its own
* corresponding mmap'd userspace allocation, there are no partial
* references: every guest physical range fully references an mmap'd
* range => no per-page reference counters have to be allocated.
*
* Return values:
* 0: success
* !0: failure - errno indicating the source of the failure
*/
int
alloc_guest_mem(struct vm_create_params *vcp)
{
void *p;
int ret;
size_t i, j;
struct vm_mem_range *vmr;
for (i = 0; i < vcp->vcp_nmemranges; i++) {
vmr = &vcp->vcp_memranges[i];
p = mmap(NULL, vmr->vmr_size, PROT_READ | PROT_WRITE,
MAP_PRIVATE | MAP_ANON, -1, 0);
if (p == MAP_FAILED) {
ret = errno;
for (j = 0; j < i; j++) {
vmr = &vcp->vcp_memranges[j];
munmap((void *)vmr->vmr_va, vmr->vmr_size);
}
return (ret);
}
vmr->vmr_va = (vaddr_t)p;
}
return (0);
}
/*
* vmm_create_vm
*
* Requests vmm(4) to create a new VM using the supplied creation
* parameters. This operation results in the creation of the in-kernel
* structures for the VM, but does not start the VM's vcpu(s).
*
* Parameters:
* vcp: vm_create_params struct containing the VM's desired creation
* configuration
*
* Return values:
* 0: success
* !0 : ioctl to vmm(4) failed
*/
int
vmm_create_vm(struct vm_create_params *vcp)
{
/* Sanity check arguments */
if (vcp->vcp_ncpus > VMM_MAX_VCPUS_PER_VM)
return (EINVAL);
if (vcp->vcp_nmemranges == 0 ||
vcp->vcp_nmemranges > VMM_MAX_MEM_RANGES)
return (EINVAL);
if (vcp->vcp_ndisks > VMM_MAX_DISKS_PER_VM)
return (EINVAL);
if (vcp->vcp_nnics > VMM_MAX_NICS_PER_VM)
return (EINVAL);
if (ioctl(env->vmd_fd, VMM_IOC_CREATE, vcp) == -1)
return (errno);
return (0);
}
/*
* init_emulated_hw
*
* Initializes the userspace hardware emulation
*/
void
init_emulated_hw(struct vmop_create_params *vmc, int child_cdrom,
int child_disks[][VM_MAX_BASE_PER_DISK], int *child_taps)
{
struct vm_create_params *vcp = &vmc->vmc_params;
int i;
uint64_t memlo, memhi;
/* Calculate memory size for NVRAM registers */
memlo = memhi = 0;
if (vcp->vcp_nmemranges > 2)
memlo = vcp->vcp_memranges[2].vmr_size - 15 * 0x100000;
if (vcp->vcp_nmemranges > 3)
memhi = vcp->vcp_memranges[3].vmr_size;
/* Reset the IO port map */
memset(&ioports_map, 0, sizeof(io_fn_t) * MAX_PORTS);
/* Init i8253 PIT */
i8253_init(vcp->vcp_id);
ioports_map[TIMER_CTRL] = vcpu_exit_i8253;
ioports_map[TIMER_BASE + TIMER_CNTR0] = vcpu_exit_i8253;
ioports_map[TIMER_BASE + TIMER_CNTR1] = vcpu_exit_i8253;
ioports_map[TIMER_BASE + TIMER_CNTR2] = vcpu_exit_i8253;
ioports_map[PCKBC_AUX] = vcpu_exit_i8253_misc;
/* Init mc146818 RTC */
mc146818_init(vcp->vcp_id, memlo, memhi);
ioports_map[IO_RTC] = vcpu_exit_mc146818;
ioports_map[IO_RTC + 1] = vcpu_exit_mc146818;
/* Init master and slave PICs */
i8259_init();
ioports_map[IO_ICU1] = vcpu_exit_i8259;
ioports_map[IO_ICU1 + 1] = vcpu_exit_i8259;
ioports_map[IO_ICU2] = vcpu_exit_i8259;
ioports_map[IO_ICU2 + 1] = vcpu_exit_i8259;
ioports_map[ELCR0] = vcpu_exit_elcr;
ioports_map[ELCR1] = vcpu_exit_elcr;
/* Init ns8250 UART */
ns8250_init(con_fd, vcp->vcp_id);
for (i = COM1_DATA; i <= COM1_SCR; i++)
ioports_map[i] = vcpu_exit_com;
/* Init QEMU fw_cfg interface */
fw_cfg_init(vmc);
ioports_map[FW_CFG_IO_SELECT] = vcpu_exit_fw_cfg;
ioports_map[FW_CFG_IO_DATA] = vcpu_exit_fw_cfg;
ioports_map[FW_CFG_IO_DMA_ADDR_HIGH] = vcpu_exit_fw_cfg_dma;
ioports_map[FW_CFG_IO_DMA_ADDR_LOW] = vcpu_exit_fw_cfg_dma;
/* Initialize PCI */
for (i = VMM_PCI_IO_BAR_BASE; i <= VMM_PCI_IO_BAR_END; i++)
ioports_map[i] = vcpu_exit_pci;
ioports_map[PCI_MODE1_ADDRESS_REG] = vcpu_exit_pci;
ioports_map[PCI_MODE1_DATA_REG] = vcpu_exit_pci;
ioports_map[PCI_MODE1_DATA_REG + 1] = vcpu_exit_pci;
ioports_map[PCI_MODE1_DATA_REG + 2] = vcpu_exit_pci;
ioports_map[PCI_MODE1_DATA_REG + 3] = vcpu_exit_pci;
pci_init();
/* Initialize virtio devices */
virtio_init(current_vm, child_cdrom, child_disks, child_taps);
}
/*
* restore_emulated_hw
*
* Restores the userspace hardware emulation from fd
*/
void
restore_emulated_hw(struct vm_create_params *vcp, int fd,
int *child_taps, int child_disks[][VM_MAX_BASE_PER_DISK], int child_cdrom)
{
/* struct vm_create_params *vcp = &vmc->vmc_params; */
int i;
memset(&ioports_map, 0, sizeof(io_fn_t) * MAX_PORTS);
/* Init i8253 PIT */
i8253_restore(fd, vcp->vcp_id);
ioports_map[TIMER_CTRL] = vcpu_exit_i8253;
ioports_map[TIMER_BASE + TIMER_CNTR0] = vcpu_exit_i8253;
ioports_map[TIMER_BASE + TIMER_CNTR1] = vcpu_exit_i8253;
ioports_map[TIMER_BASE + TIMER_CNTR2] = vcpu_exit_i8253;
/* Init master and slave PICs */
i8259_restore(fd);
ioports_map[IO_ICU1] = vcpu_exit_i8259;
ioports_map[IO_ICU1 + 1] = vcpu_exit_i8259;
ioports_map[IO_ICU2] = vcpu_exit_i8259;
ioports_map[IO_ICU2 + 1] = vcpu_exit_i8259;
/* Init ns8250 UART */
ns8250_restore(fd, con_fd, vcp->vcp_id);
for (i = COM1_DATA; i <= COM1_SCR; i++)
ioports_map[i] = vcpu_exit_com;
/* Init mc146818 RTC */
mc146818_restore(fd, vcp->vcp_id);
ioports_map[IO_RTC] = vcpu_exit_mc146818;
ioports_map[IO_RTC + 1] = vcpu_exit_mc146818;
/* Init QEMU fw_cfg interface */
fw_cfg_restore(fd);
ioports_map[FW_CFG_IO_SELECT] = vcpu_exit_fw_cfg;
ioports_map[FW_CFG_IO_DATA] = vcpu_exit_fw_cfg;
ioports_map[FW_CFG_IO_DMA_ADDR_HIGH] = vcpu_exit_fw_cfg_dma;
ioports_map[FW_CFG_IO_DMA_ADDR_LOW] = vcpu_exit_fw_cfg_dma;
/* Initialize PCI */
for (i = VMM_PCI_IO_BAR_BASE; i <= VMM_PCI_IO_BAR_END; i++)
ioports_map[i] = vcpu_exit_pci;
ioports_map[PCI_MODE1_ADDRESS_REG] = vcpu_exit_pci;
ioports_map[PCI_MODE1_DATA_REG] = vcpu_exit_pci;
ioports_map[PCI_MODE1_DATA_REG + 1] = vcpu_exit_pci;
ioports_map[PCI_MODE1_DATA_REG + 2] = vcpu_exit_pci;
ioports_map[PCI_MODE1_DATA_REG + 3] = vcpu_exit_pci;
pci_restore(fd);
virtio_restore(fd, current_vm, child_cdrom, child_disks, child_taps);
}
/*
* run_vm
*
* Runs the VM whose creation parameters are specified in vcp
*
* Parameters:
* child_cdrom: previously-opened child ISO disk file descriptor
* child_disks: previously-opened child VM disk file file descriptors
* child_taps: previously-opened child tap file descriptors
* vmc: vmop_create_params struct containing the VM's desired creation
* configuration
* vrs: VCPU register state to initialize
*
* Return values:
* 0: the VM exited normally
* !0 : the VM exited abnormally or failed to start
*/
int
run_vm(int child_cdrom, int child_disks[][VM_MAX_BASE_PER_DISK],
int *child_taps, struct vmop_create_params *vmc,
struct vcpu_reg_state *vrs)
{
struct vm_create_params *vcp = &vmc->vmc_params;
struct vm_rwregs_params vregsp;
uint8_t evdone = 0;
size_t i;
int ret;
pthread_t *tid, evtid;
struct vm_run_params **vrp;
void *exit_status;
if (vcp == NULL)
return (EINVAL);
if (child_cdrom == -1 && strlen(vcp->vcp_cdrom))
return (EINVAL);
if (child_disks == NULL && vcp->vcp_ndisks != 0)
return (EINVAL);
if (child_taps == NULL && vcp->vcp_nnics != 0)
return (EINVAL);
if (vcp->vcp_ncpus > VMM_MAX_VCPUS_PER_VM)
return (EINVAL);
if (vcp->vcp_ndisks > VMM_MAX_DISKS_PER_VM)
return (EINVAL);
if (vcp->vcp_nnics > VMM_MAX_NICS_PER_VM)
return (EINVAL);
if (vcp->vcp_nmemranges == 0 ||
vcp->vcp_nmemranges > VMM_MAX_MEM_RANGES)
return (EINVAL);
tid = calloc(vcp->vcp_ncpus, sizeof(pthread_t));
vrp = calloc(vcp->vcp_ncpus, sizeof(struct vm_run_params *));
if (tid == NULL || vrp == NULL) {
log_warn("%s: memory allocation error - exiting.",
__progname);
return (ENOMEM);
}
log_debug("%s: initializing hardware for vm %s", __func__,
vcp->vcp_name);
if (!(current_vm->vm_state & VM_STATE_RECEIVED))
init_emulated_hw(vmc, child_cdrom, child_disks, child_taps);
ret = pthread_mutex_init(&threadmutex, NULL);
if (ret) {
log_warn("%s: could not initialize thread state mutex",
__func__);
return (ret);
}
ret = pthread_cond_init(&threadcond, NULL);
if (ret) {
log_warn("%s: could not initialize thread state "
"condition variable", __func__);
return (ret);
}
mutex_lock(&threadmutex);
log_debug("%s: starting vcpu threads for vm %s", __func__,
vcp->vcp_name);
/*
* Create and launch one thread for each VCPU. These threads may
* migrate between PCPUs over time; the need to reload CPU state
* in such situations is detected and performed by vmm(4) in the
* kernel.
*/
for (i = 0 ; i < vcp->vcp_ncpus; i++) {
vrp[i] = malloc(sizeof(struct vm_run_params));
if (vrp[i] == NULL) {
log_warn("%s: memory allocation error - "
"exiting.", __progname);
/* caller will exit, so skip freeing */
return (ENOMEM);
}
vrp[i]->vrp_exit = malloc(sizeof(struct vm_exit));
if (vrp[i]->vrp_exit == NULL) {
log_warn("%s: memory allocation error - "
"exiting.", __progname);
/* caller will exit, so skip freeing */
return (ENOMEM);
}
vrp[i]->vrp_vm_id = vcp->vcp_id;
vrp[i]->vrp_vcpu_id = i;
if (vcpu_reset(vcp->vcp_id, i, vrs)) {
log_warnx("%s: cannot reset VCPU %zu - exiting.",
__progname, i);
return (EIO);
}
/* once more because reset_cpu changes regs */
if (current_vm->vm_state & VM_STATE_RECEIVED) {
vregsp.vrwp_vm_id = vcp->vcp_id;
vregsp.vrwp_vcpu_id = i;
vregsp.vrwp_regs = *vrs;
vregsp.vrwp_mask = VM_RWREGS_ALL;
if ((ret = ioctl(env->vmd_fd, VMM_IOC_WRITEREGS,
&vregsp)) == -1) {
log_warn("%s: writeregs failed", __func__);
return (ret);
}
}
ret = pthread_cond_init(&vcpu_run_cond[i], NULL);
if (ret) {
log_warnx("%s: cannot initialize cond var (%d)",
__progname, ret);
return (ret);
}
ret = pthread_mutex_init(&vcpu_run_mtx[i], NULL);
if (ret) {
log_warnx("%s: cannot initialize mtx (%d)",
__progname, ret);
return (ret);
}
ret = pthread_cond_init(&vcpu_unpause_cond[i], NULL);
if (ret) {
log_warnx("%s: cannot initialize unpause var (%d)",
__progname, ret);
return (ret);
}
ret = pthread_mutex_init(&vcpu_unpause_mtx[i], NULL);
if (ret) {
log_warnx("%s: cannot initialize unpause mtx (%d)",
__progname, ret);
return (ret);
}
vcpu_hlt[i] = 0;
/* Start each VCPU run thread at vcpu_run_loop */
ret = pthread_create(&tid[i], NULL, vcpu_run_loop, vrp[i]);
if (ret) {
/* caller will _exit after this return */
ret = errno;
log_warn("%s: could not create vcpu thread %zu",
__func__, i);
return (ret);
}
}
log_debug("%s: waiting on events for VM %s", __func__, vcp->vcp_name);
ret = pthread_create(&evtid, NULL, event_thread, &evdone);
if (ret) {
errno = ret;
log_warn("%s: could not create event thread", __func__);
return (ret);
}
for (;;) {
ret = pthread_cond_wait(&threadcond, &threadmutex);
if (ret) {
log_warn("%s: waiting on thread state condition "
"variable failed", __func__);
return (ret);
}
/*
* Did a VCPU thread exit with an error? => return the first one
*/
for (i = 0; i < vcp->vcp_ncpus; i++) {
if (vcpu_done[i] == 0)
continue;
if (pthread_join(tid[i], &exit_status)) {
log_warn("%s: failed to join thread %zd - "
"exiting", __progname, i);
return (EIO);
}
ret = (intptr_t)exit_status;
}
/* Did the event thread exit? => return with an error */
if (evdone) {
if (pthread_join(evtid, &exit_status)) {
log_warn("%s: failed to join event thread - "
"exiting", __progname);
return (EIO);
}
log_warnx("%s: vm %d event thread exited "
"unexpectedly", __progname, vcp->vcp_id);
return (EIO);
}
/* Did all VCPU threads exit successfully? => return */
for (i = 0; i < vcp->vcp_ncpus; i++) {
if (vcpu_done[i] == 0)
break;
}
if (i == vcp->vcp_ncpus)
return (ret);
/* Some more threads to wait for, start over */
}
return (ret);
}
void *
event_thread(void *arg)
{
uint8_t *donep = arg;
intptr_t ret;
ret = event_dispatch();
mutex_lock(&threadmutex);
*donep = 1;
pthread_cond_signal(&threadcond);
mutex_unlock(&threadmutex);
return (void *)ret;
}
/*
* vcpu_run_loop
*
* Runs a single VCPU until vmm(4) requires help handling an exit,
* or the VM terminates.
*
* Parameters:
* arg: vcpu_run_params for the VCPU being run by this thread
*
* Return values:
* NULL: the VCPU shutdown properly
* !NULL: error processing VCPU run, or the VCPU shutdown abnormally
*/
void *
vcpu_run_loop(void *arg)
{
struct vm_run_params *vrp = (struct vm_run_params *)arg;
intptr_t ret = 0;
int irq;
uint32_t n;
vrp->vrp_continue = 0;
n = vrp->vrp_vcpu_id;
for (;;) {
ret = pthread_mutex_lock(&vcpu_run_mtx[n]);
if (ret) {
log_warnx("%s: can't lock vcpu run mtx (%d)",
__func__, (int)ret);
return ((void *)ret);
}
/* If we are halted and need to pause, pause */
if (vcpu_hlt[n] && (current_vm->vm_state & VM_STATE_PAUSED)) {
ret = pthread_barrier_wait(&vm_pause_barrier);
if (ret != 0 && ret != PTHREAD_BARRIER_SERIAL_THREAD) {
log_warnx("%s: could not wait on pause barrier (%d)",
__func__, (int)ret);
return ((void *)ret);
}
ret = pthread_mutex_lock(&vcpu_unpause_mtx[n]);
if (ret) {
log_warnx("%s: can't lock vcpu unpause mtx (%d)",
__func__, (int)ret);
return ((void *)ret);
}
ret = pthread_cond_wait(&vcpu_unpause_cond[n],
&vcpu_unpause_mtx[n]);
if (ret) {
log_warnx(
"%s: can't wait on unpause cond (%d)",
__func__, (int)ret);
break;
}
ret = pthread_mutex_unlock(&vcpu_unpause_mtx[n]);
if (ret) {
log_warnx("%s: can't unlock unpause mtx (%d)",
__func__, (int)ret);
break;
}
}
/* If we are halted and not paused, wait */
if (vcpu_hlt[n]) {
ret = pthread_cond_wait(&vcpu_run_cond[n],
&vcpu_run_mtx[n]);
if (ret) {
log_warnx(
"%s: can't wait on cond (%d)",
__func__, (int)ret);
(void)pthread_mutex_unlock(
&vcpu_run_mtx[n]);
break;
}
}
ret = pthread_mutex_unlock(&vcpu_run_mtx[n]);
if (ret) {
log_warnx("%s: can't unlock mutex on cond (%d)",
__func__, (int)ret);
break;
}
if (vrp->vrp_irqready && i8259_is_pending()) {
irq = i8259_ack();
vrp->vrp_irq = irq;
} else
vrp->vrp_irq = 0xFFFF;
/* Still more pending? */
if (i8259_is_pending()) {
/*
* XXX can probably avoid ioctls here by providing intr
* in vrp
*/
if (vcpu_pic_intr(vrp->vrp_vm_id,
vrp->vrp_vcpu_id, 1)) {
fatal("can't set INTR");
}
} else {
if (vcpu_pic_intr(vrp->vrp_vm_id,
vrp->vrp_vcpu_id, 0)) {
fatal("can't clear INTR");
}
}
if (ioctl(env->vmd_fd, VMM_IOC_RUN, vrp) == -1) {
/* If run ioctl failed, exit */
ret = errno;
log_warn("%s: vm %d / vcpu %d run ioctl failed",
__func__, vrp->vrp_vm_id, n);
break;
}
/* If the VM is terminating, exit normally */
if (vrp->vrp_exit_reason == VM_EXIT_TERMINATED) {
ret = (intptr_t)NULL;
break;
}
if (vrp->vrp_exit_reason != VM_EXIT_NONE) {
/*
* vmm(4) needs help handling an exit, handle in
* vcpu_exit.
*/
ret = vcpu_exit(vrp);
if (ret)
break;
}
}
mutex_lock(&threadmutex);
vcpu_done[n] = 1;
pthread_cond_signal(&threadcond);
mutex_unlock(&threadmutex);
return ((void *)ret);
}
int
vcpu_pic_intr(uint32_t vm_id, uint32_t vcpu_id, uint8_t intr)
{
struct vm_intr_params vip;
memset(&vip, 0, sizeof(vip));
vip.vip_vm_id = vm_id;
vip.vip_vcpu_id = vcpu_id; /* XXX always 0? */
vip.vip_intr = intr;
if (ioctl(env->vmd_fd, VMM_IOC_INTR, &vip) == -1)
return (errno);
return (0);
}
/*
* vcpu_exit_pci
*
* Handle all I/O to the emulated PCI subsystem.
*
* Parameters:
* vrp: vcpu run paramters containing guest state for this exit
*
* Return value:
* Interrupt to inject to the guest VM, or 0xFF if no interrupt should
* be injected.
*/
uint8_t
vcpu_exit_pci(struct vm_run_params *vrp)
{
struct vm_exit *vei = vrp->vrp_exit;
uint8_t intr;
intr = 0xFF;
switch (vei->vei.vei_port) {
case PCI_MODE1_ADDRESS_REG:
pci_handle_address_reg(vrp);
break;
case PCI_MODE1_DATA_REG:
case PCI_MODE1_DATA_REG + 1:
case PCI_MODE1_DATA_REG + 2:
case PCI_MODE1_DATA_REG + 3:
pci_handle_data_reg(vrp);
break;
case VMM_PCI_IO_BAR_BASE ... VMM_PCI_IO_BAR_END:
intr = pci_handle_io(vrp);
break;
default:
log_warnx("%s: unknown PCI register 0x%llx",
__progname, (uint64_t)vei->vei.vei_port);
break;
}
return (intr);
}
/*
* vcpu_exit_inout
*
* Handle all I/O exits that need to be emulated in vmd. This includes the
* i8253 PIT, the com1 ns8250 UART, and the MC146818 RTC/NVRAM device.
*
* Parameters:
* vrp: vcpu run parameters containing guest state for this exit
*/
void
vcpu_exit_inout(struct vm_run_params *vrp)
{
struct vm_exit *vei = vrp->vrp_exit;
uint8_t intr = 0xFF;
if (ioports_map[vei->vei.vei_port] != NULL)
intr = ioports_map[vei->vei.vei_port](vrp);
else if (vei->vei.vei_dir == VEI_DIR_IN)
set_return_data(vei, 0xFFFFFFFF);
if (intr != 0xFF)
vcpu_assert_pic_irq(vrp->vrp_vm_id, vrp->vrp_vcpu_id, intr);
}
/*
* vcpu_exit_eptviolation
*
* handle an EPT Violation
*
* Parameters:
* vrp: vcpu run parameters containing guest state for this exit
*
* Return values:
* 0: no action required
* EAGAIN: a protection fault occured, kill the vm.
*/
int
vcpu_exit_eptviolation(struct vm_run_params *vrp)
{
struct vm_exit *ve = vrp->vrp_exit;
/*
* vmd may be exiting to vmd to handle a pending interrupt
* but last exit type may have been VMX_EXIT_EPT_VIOLATION,
* check the fault_type to ensure we really are processing
* a VMX_EXIT_EPT_VIOLATION.
*/
if (ve->vee.vee_fault_type == VEE_FAULT_PROTECT) {
log_debug("%s: EPT Violation: rip=0x%llx",
__progname, vrp->vrp_exit->vrs.vrs_gprs[VCPU_REGS_RIP]);
return (EAGAIN);
}
return (0);
}
/*
* vcpu_exit
*
* Handle a vcpu exit. This function is called when it is determined that
* vmm(4) requires the assistance of vmd to support a particular guest
* exit type (eg, accessing an I/O port or device). Guest state is contained
* in 'vrp', and will be resent to vmm(4) on exit completion.
*
* Upon conclusion of handling the exit, the function determines if any
* interrupts should be injected into the guest, and asserts the proper
* IRQ line whose interrupt should be vectored.
*
* Parameters:
* vrp: vcpu run parameters containing guest state for this exit
*
* Return values:
* 0: the exit was handled successfully
* 1: an error occurred (eg, unknown exit reason passed in 'vrp')
*/
int
vcpu_exit(struct vm_run_params *vrp)
{
int ret;
switch (vrp->vrp_exit_reason) {
case VMX_EXIT_INT_WINDOW:
case SVM_VMEXIT_VINTR:
case VMX_EXIT_CPUID:
case VMX_EXIT_EXTINT:
case SVM_VMEXIT_INTR:
case SVM_VMEXIT_NPF:
case SVM_VMEXIT_MSR:
case SVM_VMEXIT_CPUID:
/*
* We may be exiting to vmd to handle a pending interrupt but
* at the same time the last exit type may have been one of
* these. In this case, there's nothing extra to be done
* here (and falling through to the default case below results
* in more vmd log spam).
*/
break;
case VMX_EXIT_EPT_VIOLATION:
ret = vcpu_exit_eptviolation(vrp);
if (ret)
return (ret);
break;
case VMX_EXIT_IO:
case SVM_VMEXIT_IOIO:
vcpu_exit_inout(vrp);
break;
case VMX_EXIT_HLT:
case SVM_VMEXIT_HLT:
ret = pthread_mutex_lock(&vcpu_run_mtx[vrp->vrp_vcpu_id]);
if (ret) {
log_warnx("%s: can't lock vcpu mutex (%d)",
__func__, ret);
return (ret);
}
vcpu_hlt[vrp->vrp_vcpu_id] = 1;
ret = pthread_mutex_unlock(&vcpu_run_mtx[vrp->vrp_vcpu_id]);
if (ret) {
log_warnx("%s: can't unlock vcpu mutex (%d)",
__func__, ret);
return (ret);
}
break;
case VMX_EXIT_TRIPLE_FAULT:
case SVM_VMEXIT_SHUTDOWN:
/* reset VM */
return (EAGAIN);
default:
log_debug("%s: unknown exit reason 0x%x",
__progname, vrp->vrp_exit_reason);
}
vrp->vrp_continue = 1;
return (0);
}
/*
* find_gpa_range
*
* Search for a contiguous guest physical mem range.
*
* Parameters:
* vcp: VM create parameters that contain the memory map to search in
* gpa: the starting guest physical address
* len: the length of the memory range
*
* Return values:
* NULL: on failure if there is no memory range as described by the parameters
* Pointer to vm_mem_range that contains the start of the range otherwise.
*/
static struct vm_mem_range *
find_gpa_range(struct vm_create_params *vcp, paddr_t gpa, size_t len)
{
size_t i, n;
struct vm_mem_range *vmr;
/* Find the first vm_mem_range that contains gpa */
for (i = 0; i < vcp->vcp_nmemranges; i++) {
vmr = &vcp->vcp_memranges[i];
if (vmr->vmr_gpa + vmr->vmr_size >= gpa)
break;
}
/* No range found. */
if (i == vcp->vcp_nmemranges)
return (NULL);
/*
* vmr may cover the range [gpa, gpa + len) only partly. Make
* sure that the following vm_mem_ranges are contiguous and
* cover the rest.
*/
n = vmr->vmr_size - (gpa - vmr->vmr_gpa);
if (len < n)
len = 0;
else
len -= n;
gpa = vmr->vmr_gpa + vmr->vmr_size;
for (i = i + 1; len != 0 && i < vcp->vcp_nmemranges; i++) {
vmr = &vcp->vcp_memranges[i];
if (gpa != vmr->vmr_gpa)
return (NULL);
if (len <= vmr->vmr_size)
len = 0;
else
len -= vmr->vmr_size;
gpa = vmr->vmr_gpa + vmr->vmr_size;
}
if (len != 0)
return (NULL);
return (vmr);
}
void *
vaddr_mem(paddr_t gpa, size_t len)
{
struct vm_create_params *vcp = ¤t_vm->vm_params.vmc_params;
size_t i;
struct vm_mem_range *vmr;
paddr_t gpend = gpa + len;
/* Find the first vm_mem_range that contains gpa */
for (i = 0; i < vcp->vcp_nmemranges; i++) {
vmr = &vcp->vcp_memranges[i];
if (gpa < vmr->vmr_gpa)
continue;
if (gpend >= vmr->vmr_gpa + vmr->vmr_size)
continue;
return ((char *)vmr->vmr_va + (gpa - vmr->vmr_gpa));
}
return (NULL);
}
/*
* write_mem
*
* Copies data from 'buf' into the guest VM's memory at paddr 'dst'.
*
* Parameters:
* dst: the destination paddr_t in the guest VM
* buf: data to copy (or NULL to zero the data)
* len: number of bytes to copy
*
* Return values:
* 0: success
* EINVAL: if the guest physical memory range [dst, dst + len) does not
* exist in the guest.
*/
int
write_mem(paddr_t dst, const void *buf, size_t len)
{
const char *from = buf;
char *to;
size_t n, off;
struct vm_mem_range *vmr;
vmr = find_gpa_range(¤t_vm->vm_params.vmc_params, dst, len);
if (vmr == NULL) {
errno = EINVAL;
log_warn("%s: failed - invalid memory range dst = 0x%lx, "
"len = 0x%zx", __func__, dst, len);
return (EINVAL);
}
off = dst - vmr->vmr_gpa;
while (len != 0) {
n = vmr->vmr_size - off;
if (len < n)
n = len;
to = (char *)vmr->vmr_va + off;
if (buf == NULL)
memset(to, 0, n);
else {
memcpy(to, from, n);
from += n;
}
len -= n;
off = 0;
vmr++;
}
return (0);
}
/*
* read_mem
*
* Reads memory at guest paddr 'src' into 'buf'.
*
* Parameters:
* src: the source paddr_t in the guest VM to read from.
* buf: destination (local) buffer
* len: number of bytes to read
*
* Return values:
* 0: success
* EINVAL: if the guest physical memory range [dst, dst + len) does not
* exist in the guest.
*/
int
read_mem(paddr_t src, void *buf, size_t len)
{
char *from, *to = buf;
size_t n, off;
struct vm_mem_range *vmr;
vmr = find_gpa_range(¤t_vm->vm_params.vmc_params, src, len);
if (vmr == NULL) {
errno = EINVAL;
log_warn("%s: failed - invalid memory range src = 0x%lx, "
"len = 0x%zx", __func__, src, len);
return (EINVAL);
}
off = src - vmr->vmr_gpa;
while (len != 0) {
n = vmr->vmr_size - off;
if (len < n)
n = len;
from = (char *)vmr->vmr_va + off;
memcpy(to, from, n);
to += n;
len -= n;
off = 0;
vmr++;
}
return (0);
}
int
iovec_mem(paddr_t src, size_t len, struct iovec *iov, int iovcnt)
{
size_t n, off;
struct vm_mem_range *vmr;
int niov = 0;
vmr = find_gpa_range(¤t_vm->vm_params.vmc_params, src, len);
if (vmr == NULL) {
errno = EINVAL;
return (-1);
}
off = src - vmr->vmr_gpa;
while (len > 0) {
if (niov == iovcnt) {
errno = ENOMEM;
return (-1);
}
n = vmr->vmr_size - off;
if (len < n)
n = len;
iov[niov].iov_base = (char *)vmr->vmr_va + off;
iov[niov].iov_len = n;
niov++;
len -= n;
off = 0;
vmr++;
}
return (niov);
}
/*
* vcpu_assert_pic_irq
*
* Injects the specified IRQ on the supplied vcpu/vm
*
* Parameters:
* vm_id: VM ID to inject to
* vcpu_id: VCPU ID to inject to
* irq: IRQ to inject
*/
void
vcpu_assert_pic_irq(uint32_t vm_id, uint32_t vcpu_id, int irq)
{
int ret;
i8259_assert_irq(irq);
if (i8259_is_pending()) {
if (vcpu_pic_intr(vm_id, vcpu_id, 1))
fatalx("%s: can't assert INTR", __func__);
ret = pthread_mutex_lock(&vcpu_run_mtx[vcpu_id]);
if (ret)
fatalx("%s: can't lock vcpu mtx (%d)", __func__, ret);
vcpu_hlt[vcpu_id] = 0;
ret = pthread_cond_signal(&vcpu_run_cond[vcpu_id]);
if (ret)
fatalx("%s: can't signal (%d)", __func__, ret);
ret = pthread_mutex_unlock(&vcpu_run_mtx[vcpu_id]);
if (ret)
fatalx("%s: can't unlock vcpu mtx (%d)", __func__, ret);
}
}
/*
* vcpu_deassert_pic_irq
*
* Clears the specified IRQ on the supplied vcpu/vm
*
* Parameters:
* vm_id: VM ID to clear in
* vcpu_id: VCPU ID to clear in
* irq: IRQ to clear
*/
void
vcpu_deassert_pic_irq(uint32_t vm_id, uint32_t vcpu_id, int irq)
{
i8259_deassert_irq(irq);
if (!i8259_is_pending()) {
if (vcpu_pic_intr(vm_id, vcpu_id, 0))
fatalx("%s: can't deassert INTR for vm_id %d, "
"vcpu_id %d", __func__, vm_id, vcpu_id);
}
}
/*
* fd_hasdata
*
* Determines if data can be read from a file descriptor.
*
* Parameters:
* fd: the fd to check
*
* Return values:
* 1 if data can be read from an fd, or 0 otherwise.
*/
int
fd_hasdata(int fd)
{
struct pollfd pfd[1];
int nready, hasdata = 0;
pfd[0].fd = fd;
pfd[0].events = POLLIN;
nready = poll(pfd, 1, 0);
if (nready == -1)
log_warn("checking file descriptor for data failed");
else if (nready == 1 && pfd[0].revents & POLLIN)
hasdata = 1;
return (hasdata);
}
/*
* mutex_lock
*
* Wrapper function for pthread_mutex_lock that does error checking and that
* exits on failure
*/
void
mutex_lock(pthread_mutex_t *m)
{
int ret;
ret = pthread_mutex_lock(m);
if (ret) {
errno = ret;
fatal("could not acquire mutex");
}
}
/*
* mutex_unlock
*
* Wrapper function for pthread_mutex_unlock that does error checking and that
* exits on failure
*/
void
mutex_unlock(pthread_mutex_t *m)
{
int ret;
ret = pthread_mutex_unlock(m);
if (ret) {
errno = ret;
fatal("could not release mutex");
}
}
/*
* set_return_data
*
* Utility function for manipulating register data in vm exit info structs. This
* function ensures that the data is copied to the vei->vei.vei_data field with
* the proper size for the operation being performed.
*
* Parameters:
* vei: exit information
* data: return data
*/
void
set_return_data(struct vm_exit *vei, uint32_t data)
{
switch (vei->vei.vei_size) {
case 1:
vei->vei.vei_data &= ~0xFF;
vei->vei.vei_data |= (uint8_t)data;
break;
case 2:
vei->vei.vei_data &= ~0xFFFF;
vei->vei.vei_data |= (uint16_t)data;
break;
case 4:
vei->vei.vei_data = data;
break;
}
}
/*
* get_input_data
*
* Utility function for manipulating register data in vm exit info
* structs. This function ensures that the data is copied from the
* vei->vei.vei_data field with the proper size for the operation being
* performed.
*
* Parameters:
* vei: exit information
* data: location to store the result
*/
void
get_input_data(struct vm_exit *vei, uint32_t *data)
{
switch (vei->vei.vei_size) {
case 1:
*data &= 0xFFFFFF00;
*data |= (uint8_t)vei->vei.vei_data;
break;
case 2:
*data &= 0xFFFF0000;
*data |= (uint16_t)vei->vei.vei_data;
break;
case 4:
*data = vei->vei.vei_data;
break;
default:
log_warnx("%s: invalid i/o size %d", __func__,
vei->vei.vei_size);
}
}
/*
* translate_gva
*
* Translates a guest virtual address to a guest physical address by walking
* the currently active page table (if needed).
*
* Note - this function can possibly alter the supplied VCPU state.
* Specifically, it may inject exceptions depending on the current VCPU
* configuration, and may alter %cr2 on #PF. Consequently, this function
* should only be used as part of instruction emulation.
*
* Parameters:
* exit: The VCPU this translation should be performed for (guest MMU settings
* are gathered from this VCPU)
* va: virtual address to translate
* pa: pointer to paddr_t variable that will receive the translated physical
* address. 'pa' is unchanged on error.
* mode: one of PROT_READ, PROT_WRITE, PROT_EXEC indicating the mode in which
* the address should be translated
*
* Return values:
* 0: the address was successfully translated - 'pa' contains the physical
* address currently mapped by 'va'.
* EFAULT: the PTE for 'VA' is unmapped. A #PF will be injected in this case
* and %cr2 set in the vcpu structure.
* EINVAL: an error occurred reading paging table structures
*/
int
translate_gva(struct vm_exit* exit, uint64_t va, uint64_t* pa, int mode)
{
int level, shift, pdidx;
uint64_t pte, pt_paddr, pte_paddr, mask, low_mask, high_mask;
uint64_t shift_width, pte_size;
struct vcpu_reg_state *vrs;
vrs = &exit->vrs;
if (!pa)
return (EINVAL);
if (!(vrs->vrs_crs[VCPU_REGS_CR0] & CR0_PG)) {
log_debug("%s: unpaged, va=pa=0x%llx", __func__, va);
*pa = va;
return (0);
}
pt_paddr = vrs->vrs_crs[VCPU_REGS_CR3];
log_debug("%s: guest %%cr0=0x%llx, %%cr3=0x%llx", __func__,
vrs->vrs_crs[VCPU_REGS_CR0], vrs->vrs_crs[VCPU_REGS_CR3]);
if (vrs->vrs_crs[VCPU_REGS_CR0] & CR0_PE) {
if (vrs->vrs_crs[VCPU_REGS_CR4] & CR4_PAE) {
pte_size = sizeof(uint64_t);
shift_width = 9;
if (vrs->vrs_msrs[VCPU_REGS_EFER] & EFER_LMA) {
/* 4 level paging */
level = 4;
mask = L4_MASK;
shift = L4_SHIFT;
} else {
/* 32 bit with PAE paging */
level = 3;
mask = L3_MASK;
shift = L3_SHIFT;
}
} else {
/* 32 bit paging */
level = 2;
shift_width = 10;
mask = 0xFFC00000;
shift = 22;
pte_size = sizeof(uint32_t);
}
} else
return (EINVAL);
/* XXX: Check for R bit in segment selector and set A bit */
for (;level > 0; level--) {
pdidx = (va & mask) >> shift;
pte_paddr = (pt_paddr) + (pdidx * pte_size);
log_debug("%s: read pte level %d @ GPA 0x%llx", __func__,
level, pte_paddr);
if (read_mem(pte_paddr, &pte, pte_size)) {
log_warn("%s: failed to read pte", __func__);
return (EFAULT);
}
log_debug("%s: PTE @ 0x%llx = 0x%llx", __func__, pte_paddr,
pte);
/* XXX: Set CR2 */
if (!(pte & PG_V))
return (EFAULT);
/* XXX: Check for SMAP */
if ((mode == PROT_WRITE) && !(pte & PG_RW))
return (EPERM);
if ((exit->cpl > 0) && !(pte & PG_u))
return (EPERM);
pte = pte | PG_U;
if (mode == PROT_WRITE)
pte = pte | PG_M;
if (write_mem(pte_paddr, &pte, pte_size)) {
log_warn("%s: failed to write back flags to pte",
__func__);
return (EIO);
}
/* XXX: EINVAL if in 32bit and PG_PS is 1 but CR4.PSE is 0 */
if (pte & PG_PS)
break;
if (level > 1) {
pt_paddr = pte & PG_FRAME;
shift -= shift_width;
mask = mask >> shift_width;
}
}
low_mask = (1 << shift) - 1;
high_mask = (((uint64_t)1ULL << ((pte_size * 8) - 1)) - 1) ^ low_mask;
*pa = (pte & high_mask) | (va & low_mask);
log_debug("%s: final GPA for GVA 0x%llx = 0x%llx\n", __func__, va, *pa);
return (0);
}
/*
* vm_pipe_init
*
* Initialize a vm_dev_pipe, setting up its file descriptors and its
* event structure with the given callback.
*
* Parameters:
* p: pointer to vm_dev_pipe struct to initizlize
* cb: callback to use for READ events on the read end of the pipe
*/
void
vm_pipe_init(struct vm_dev_pipe *p, void (*cb)(int, short, void *))
{
int ret;
int fds[2];
memset(p, 0, sizeof(struct vm_dev_pipe));
ret = pipe(fds);
if (ret)
fatal("failed to create vm_dev_pipe pipe");
p->read = fds[0];
p->write = fds[1];
event_set(&p->read_ev, p->read, EV_READ | EV_PERSIST, cb, NULL);
}
/*
* vm_pipe_send
*
* Send a message to an emulated device vie the provided vm_dev_pipe.
*
* Parameters:
* p: pointer to initialized vm_dev_pipe
* msg: message to send in the channel
*/
void
vm_pipe_send(struct vm_dev_pipe *p, enum pipe_msg_type msg)
{
size_t n;
n = write(p->write, &msg, sizeof(msg));
if (n != sizeof(msg))
fatal("failed to write to device pipe");
}
/*
* vm_pipe_recv
*
* Receive a message for an emulated device via the provided vm_dev_pipe.
* Returns the message value, otherwise will exit on failure.
*
* Parameters:
* p: pointer to initialized vm_dev_pipe
*
* Return values:
* a value of enum pipe_msg_type or fatal exit on read(2) error
*/
enum pipe_msg_type
vm_pipe_recv(struct vm_dev_pipe *p)
{
size_t n;
enum pipe_msg_type msg;
n = read(p->read, &msg, sizeof(msg));
if (n != sizeof(msg))
fatal("failed to read from device pipe");
return msg;
}
|