1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
3485
3486
3487
3488
3489
3490
3491
3492
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
3514
3515
3516
3517
3518
3519
3520
3521
3522
3523
3524
3525
3526
3527
3528
3529
3530
3531
3532
3533
3534
3535
3536
3537
3538
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
3552
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
3566
3567
3568
3569
3570
3571
3572
3573
3574
3575
3576
3577
3578
3579
3580
3581
3582
3583
3584
3585
3586
3587
3588
3589
3590
3591
3592
3593
3594
3595
3596
3597
3598
3599
3600
3601
3602
3603
3604
3605
3606
3607
3608
3609
3610
3611
3612
3613
3614
3615
3616
3617
3618
3619
3620
3621
3622
3623
3624
3625
3626
3627
3628
3629
3630
3631
3632
3633
3634
3635
3636
3637
3638
3639
3640
3641
3642
3643
3644
3645
3646
3647
3648
3649
3650
3651
3652
3653
3654
3655
3656
3657
3658
3659
3660
3661
3662
3663
3664
3665
3666
3667
3668
3669
3670
3671
3672
3673
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
3686
3687
3688
3689
3690
3691
3692
3693
3694
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
3706
3707
3708
3709
3710
3711
3712
3713
3714
3715
3716
3717
3718
3719
3720
3721
3722
3723
3724
3725
3726
3727
3728
3729
3730
3731
3732
3733
3734
3735
3736
3737
3738
3739
3740
3741
3742
3743
3744
3745
3746
3747
3748
3749
3750
3751
3752
3753
3754
3755
3756
3757
3758
3759
3760
3761
3762
3763
3764
3765
3766
3767
3768
3769
3770
3771
3772
3773
3774
3775
3776
3777
3778
3779
3780
3781
3782
3783
3784
3785
3786
3787
3788
3789
3790
3791
3792
3793
3794
3795
3796
3797
3798
3799
3800
3801
3802
3803
3804
3805
3806
3807
3808
3809
3810
3811
3812
3813
3814
3815
3816
3817
3818
3819
3820
3821
3822
3823
3824
3825
3826
3827
3828
3829
3830
3831
3832
3833
3834
3835
3836
3837
3838
3839
3840
3841
3842
3843
3844
3845
3846
3847
3848
3849
3850
3851
3852
3853
3854
3855
3856
3857
3858
3859
3860
3861
3862
3863
3864
3865
3866
3867
3868
3869
3870
3871
3872
3873
3874
3875
3876
3877
3878
3879
3880
3881
3882
3883
3884
3885
3886
3887
3888
3889
3890
3891
3892
3893
3894
3895
3896
3897
3898
3899
3900
3901
3902
3903
3904
3905
3906
3907
3908
3909
3910
3911
3912
3913
3914
3915
3916
3917
3918
3919
3920
3921
3922
3923
3924
3925
3926
3927
3928
3929
3930
3931
3932
3933
3934
3935
3936
3937
3938
3939
3940
3941
3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
3953
3954
3955
3956
3957
3958
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
3969
3970
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
3981
3982
3983
3984
3985
3986
3987
3988
3989
3990
3991
3992
3993
3994
3995
3996
3997
3998
3999
4000
4001
4002
4003
4004
4005
4006
4007
4008
4009
4010
4011
4012
4013
4014
4015
4016
4017
4018
4019
4020
4021
4022
4023
4024
4025
4026
4027
4028
4029
4030
4031
4032
4033
4034
4035
4036
4037
4038
4039
4040
4041
4042
4043
4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
4068
4069
4070
4071
4072
4073
4074
4075
4076
4077
4078
4079
4080
4081
4082
4083
4084
4085
4086
4087
4088
4089
4090
4091
4092
4093
4094
4095
4096
4097
4098
4099
4100
4101
4102
4103
4104
4105
4106
4107
4108
4109
4110
4111
4112
4113
4114
4115
4116
4117
4118
4119
4120
4121
4122
4123
4124
4125
4126
4127
4128
4129
4130
4131
4132
4133
4134
4135
4136
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
4162
4163
4164
4165
4166
4167
4168
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
4179
4180
4181
4182
4183
4184
4185
4186
4187
4188
4189
4190
4191
4192
4193
4194
4195
4196
4197
4198
4199
4200
4201
4202
4203
4204
4205
4206
4207
4208
4209
4210
4211
4212
4213
4214
4215
4216
4217
4218
4219
4220
4221
4222
4223
4224
4225
4226
4227
4228
4229
4230
4231
4232
4233
4234
4235
4236
4237
4238
4239
4240
4241
4242
4243
4244
4245
4246
4247
4248
4249
4250
4251
4252
4253
4254
4255
4256
4257
4258
4259
4260
4261
4262
4263
4264
4265
4266
4267
4268
4269
4270
4271
4272
4273
4274
4275
4276
4277
4278
4279
4280
4281
4282
4283
4284
4285
4286
4287
4288
4289
4290
4291
4292
4293
4294
4295
4296
4297
4298
4299
4300
4301
4302
4303
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
4315
4316
4317
4318
4319
4320
4321
4322
4323
4324
4325
4326
4327
4328
4329
4330
4331
4332
4333
4334
4335
4336
4337
4338
4339
4340
4341
4342
4343
4344
4345
4346
4347
4348
4349
4350
4351
4352
4353
4354
4355
4356
4357
4358
4359
4360
4361
4362
4363
4364
4365
4366
4367
4368
4369
4370
4371
4372
4373
4374
4375
4376
4377
4378
4379
4380
4381
4382
4383
4384
4385
4386
4387
4388
4389
4390
4391
4392
4393
4394
4395
4396
4397
4398
4399
4400
4401
4402
4403
4404
4405
4406
4407
4408
4409
4410
4411
4412
4413
4414
4415
4416
4417
4418
4419
4420
4421
4422
4423
4424
4425
4426
4427
4428
4429
4430
4431
4432
4433
4434
4435
4436
4437
4438
4439
4440
4441
4442
4443
4444
4445
4446
4447
4448
4449
4450
4451
4452
4453
4454
4455
4456
4457
4458
4459
4460
4461
4462
4463
4464
4465
4466
4467
4468
4469
4470
4471
4472
4473
4474
4475
4476
4477
4478
4479
4480
4481
4482
4483
4484
4485
4486
4487
4488
4489
4490
4491
4492
4493
4494
4495
4496
4497
4498
4499
4500
4501
4502
4503
4504
4505
4506
4507
4508
4509
4510
4511
4512
4513
4514
4515
4516
4517
4518
4519
4520
4521
4522
4523
4524
4525
4526
4527
4528
4529
4530
4531
4532
4533
4534
4535
4536
4537
4538
4539
4540
4541
4542
4543
4544
4545
4546
4547
4548
4549
4550
4551
4552
4553
4554
4555
4556
4557
4558
4559
4560
4561
4562
4563
4564
4565
4566
4567
4568
4569
4570
4571
4572
4573
4574
4575
4576
4577
4578
4579
4580
4581
4582
4583
4584
4585
4586
4587
4588
4589
4590
4591
4592
4593
4594
4595
4596
4597
4598
4599
4600
4601
4602
4603
4604
4605
4606
4607
4608
4609
4610
4611
4612
4613
4614
4615
4616
4617
4618
4619
4620
4621
4622
4623
4624
4625
4626
4627
4628
4629
4630
4631
4632
4633
4634
4635
4636
4637
4638
4639
4640
4641
4642
4643
4644
4645
4646
4647
4648
4649
4650
4651
4652
4653
4654
4655
4656
4657
4658
4659
4660
4661
4662
4663
4664
4665
4666
4667
4668
4669
4670
4671
4672
4673
4674
4675
4676
4677
4678
4679
4680
4681
4682
4683
4684
4685
4686
4687
4688
4689
4690
4691
4692
4693
4694
4695
4696
4697
4698
4699
4700
4701
4702
4703
4704
4705
4706
4707
4708
4709
4710
4711
4712
4713
4714
4715
4716
4717
4718
4719
4720
4721
4722
4723
4724
4725
4726
4727
4728
4729
4730
4731
4732
4733
4734
4735
4736
4737
4738
4739
4740
4741
4742
4743
4744
4745
4746
4747
4748
4749
4750
4751
4752
4753
4754
4755
4756
4757
4758
4759
4760
4761
4762
4763
4764
4765
4766
4767
4768
4769
4770
4771
4772
4773
4774
4775
4776
4777
4778
4779
4780
4781
4782
4783
4784
4785
4786
4787
4788
4789
4790
4791
4792
4793
4794
4795
4796
4797
4798
4799
4800
4801
4802
4803
4804
4805
4806
4807
4808
4809
4810
4811
4812
4813
4814
4815
4816
4817
4818
4819
4820
4821
4822
4823
4824
4825
4826
4827
4828
4829
4830
4831
4832
4833
4834
4835
4836
4837
4838
4839
4840
4841
4842
4843
4844
4845
4846
4847
4848
4849
4850
4851
4852
4853
4854
4855
4856
4857
4858
4859
4860
4861
4862
4863
4864
4865
4866
4867
4868
4869
4870
4871
4872
4873
4874
4875
4876
4877
4878
4879
4880
4881
4882
4883
4884
4885
4886
4887
4888
4889
4890
4891
4892
4893
4894
4895
4896
4897
4898
4899
4900
4901
4902
4903
4904
4905
4906
4907
4908
4909
4910
4911
4912
4913
4914
4915
4916
4917
4918
4919
4920
4921
4922
4923
4924
4925
4926
4927
4928
4929
4930
4931
4932
4933
4934
4935
4936
4937
4938
4939
4940
4941
4942
4943
4944
4945
4946
4947
4948
4949
4950
4951
4952
4953
4954
4955
4956
4957
4958
4959
4960
4961
4962
4963
4964
4965
4966
4967
4968
4969
4970
4971
4972
4973
4974
4975
4976
4977
4978
4979
4980
4981
4982
4983
4984
4985
4986
4987
4988
4989
4990
4991
4992
4993
4994
4995
4996
4997
4998
4999
5000
5001
5002
5003
5004
5005
5006
5007
5008
5009
5010
5011
5012
5013
5014
5015
5016
5017
5018
5019
5020
5021
5022
5023
5024
5025
5026
5027
5028
5029
5030
5031
5032
5033
5034
5035
5036
5037
5038
5039
5040
5041
5042
5043
5044
5045
5046
5047
5048
5049
5050
5051
5052
5053
5054
5055
5056
5057
5058
5059
5060
5061
5062
5063
5064
5065
5066
5067
5068
5069
5070
5071
5072
5073
5074
5075
5076
5077
5078
5079
5080
5081
5082
5083
5084
5085
5086
5087
5088
5089
5090
5091
5092
5093
5094
5095
5096
5097
5098
5099
5100
5101
5102
5103
5104
5105
5106
5107
5108
5109
5110
5111
5112
5113
5114
5115
5116
5117
5118
5119
5120
5121
5122
5123
5124
5125
5126
5127
5128
5129
5130
5131
5132
5133
5134
5135
5136
5137
5138
5139
5140
5141
5142
5143
5144
5145
5146
5147
5148
5149
5150
5151
5152
5153
5154
5155
5156
5157
5158
5159
5160
5161
5162
5163
5164
5165
5166
5167
5168
5169
5170
5171
5172
5173
5174
5175
5176
5177
5178
5179
5180
5181
5182
5183
5184
5185
5186
5187
5188
5189
5190
5191
5192
5193
5194
5195
5196
5197
5198
5199
5200
5201
5202
5203
5204
5205
5206
5207
5208
5209
5210
5211
5212
5213
5214
5215
5216
5217
5218
5219
5220
5221
5222
5223
5224
5225
5226
5227
5228
5229
5230
5231
5232
5233
5234
5235
5236
5237
5238
5239
5240
5241
5242
5243
5244
5245
5246
5247
5248
5249
5250
5251
5252
5253
5254
5255
5256
5257
5258
5259
5260
5261
5262
5263
5264
5265
5266
5267
5268
5269
5270
5271
5272
5273
5274
5275
5276
5277
5278
5279
5280
5281
5282
5283
5284
5285
5286
5287
5288
5289
5290
5291
5292
5293
5294
5295
5296
5297
5298
5299
5300
5301
5302
5303
5304
5305
5306
5307
5308
5309
5310
5311
5312
5313
5314
5315
5316
5317
5318
5319
5320
5321
5322
5323
5324
5325
5326
5327
5328
5329
5330
5331
5332
5333
5334
5335
5336
5337
5338
5339
5340
5341
5342
5343
5344
5345
5346
5347
5348
5349
5350
5351
5352
5353
5354
5355
5356
5357
5358
5359
5360
5361
5362
5363
5364
5365
5366
5367
5368
5369
5370
5371
5372
5373
5374
5375
5376
5377
5378
5379
5380
5381
5382
5383
5384
5385
5386
5387
5388
5389
5390
5391
5392
5393
5394
5395
5396
5397
5398
5399
5400
5401
5402
5403
5404
5405
5406
5407
5408
5409
5410
5411
5412
5413
5414
5415
5416
5417
5418
5419
5420
5421
5422
5423
5424
5425
5426
5427
5428
5429
5430
5431
5432
5433
5434
5435
5436
5437
5438
5439
5440
5441
5442
5443
5444
5445
5446
5447
5448
5449
5450
5451
5452
5453
5454
5455
5456
5457
5458
5459
5460
5461
5462
5463
5464
5465
5466
5467
5468
5469
5470
5471
5472
5473
5474
5475
5476
5477
5478
5479
5480
5481
5482
5483
5484
5485
5486
5487
5488
5489
5490
5491
5492
5493
5494
5495
5496
5497
5498
5499
5500
5501
5502
5503
5504
5505
5506
5507
5508
5509
5510
5511
5512
5513
5514
5515
5516
5517
5518
5519
5520
5521
5522
5523
5524
5525
5526
5527
5528
5529
5530
5531
5532
5533
5534
5535
5536
5537
5538
5539
5540
5541
5542
5543
5544
5545
5546
5547
5548
5549
5550
5551
5552
5553
5554
5555
5556
5557
5558
5559
5560
5561
5562
5563
5564
5565
5566
5567
5568
5569
5570
5571
5572
5573
5574
5575
5576
5577
5578
5579
5580
5581
5582
5583
5584
5585
5586
5587
5588
5589
5590
5591
5592
5593
5594
5595
5596
5597
5598
5599
5600
5601
5602
5603
5604
5605
5606
5607
5608
5609
5610
5611
5612
5613
5614
5615
5616
5617
5618
5619
5620
5621
5622
5623
5624
5625
5626
5627
5628
5629
5630
5631
5632
5633
5634
5635
5636
5637
5638
5639
5640
5641
5642
5643
5644
5645
5646
5647
5648
5649
5650
5651
5652
5653
5654
5655
5656
5657
5658
5659
5660
5661
5662
5663
5664
5665
5666
5667
5668
5669
5670
5671
5672
5673
5674
5675
5676
5677
5678
5679
5680
5681
5682
5683
5684
5685
5686
5687
5688
5689
5690
5691
5692
5693
5694
5695
5696
5697
5698
5699
5700
5701
5702
5703
5704
5705
5706
5707
5708
5709
5710
5711
5712
5713
5714
5715
5716
5717
5718
5719
5720
5721
5722
5723
5724
5725
5726
5727
5728
5729
5730
5731
5732
5733
5734
5735
5736
5737
5738
5739
5740
5741
5742
5743
5744
5745
5746
5747
5748
5749
5750
5751
5752
5753
5754
5755
5756
5757
5758
5759
5760
5761
5762
5763
5764
5765
5766
5767
5768
5769
5770
5771
5772
5773
5774
5775
5776
5777
5778
5779
5780
5781
5782
5783
5784
5785
5786
5787
5788
5789
5790
5791
5792
5793
5794
5795
5796
5797
5798
5799
5800
5801
5802
5803
5804
5805
5806
5807
5808
5809
5810
5811
5812
5813
5814
5815
5816
5817
5818
5819
5820
5821
5822
5823
5824
5825
5826
5827
5828
5829
5830
5831
5832
5833
5834
5835
5836
5837
5838
5839
5840
5841
5842
5843
5844
5845
5846
5847
5848
5849
5850
5851
5852
5853
5854
5855
5856
5857
5858
5859
5860
5861
5862
5863
|
//===- MemorySanitizer.cpp - detector of uninitialized reads --------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
/// \file
/// This file is a part of MemorySanitizer, a detector of uninitialized
/// reads.
///
/// The algorithm of the tool is similar to Memcheck
/// (http://goo.gl/QKbem). We associate a few shadow bits with every
/// byte of the application memory, poison the shadow of the malloc-ed
/// or alloca-ed memory, load the shadow bits on every memory read,
/// propagate the shadow bits through some of the arithmetic
/// instruction (including MOV), store the shadow bits on every memory
/// write, report a bug on some other instructions (e.g. JMP) if the
/// associated shadow is poisoned.
///
/// But there are differences too. The first and the major one:
/// compiler instrumentation instead of binary instrumentation. This
/// gives us much better register allocation, possible compiler
/// optimizations and a fast start-up. But this brings the major issue
/// as well: msan needs to see all program events, including system
/// calls and reads/writes in system libraries, so we either need to
/// compile *everything* with msan or use a binary translation
/// component (e.g. DynamoRIO) to instrument pre-built libraries.
/// Another difference from Memcheck is that we use 8 shadow bits per
/// byte of application memory and use a direct shadow mapping. This
/// greatly simplifies the instrumentation code and avoids races on
/// shadow updates (Memcheck is single-threaded so races are not a
/// concern there. Memcheck uses 2 shadow bits per byte with a slow
/// path storage that uses 8 bits per byte).
///
/// The default value of shadow is 0, which means "clean" (not poisoned).
///
/// Every module initializer should call __msan_init to ensure that the
/// shadow memory is ready. On error, __msan_warning is called. Since
/// parameters and return values may be passed via registers, we have a
/// specialized thread-local shadow for return values
/// (__msan_retval_tls) and parameters (__msan_param_tls).
///
/// Origin tracking.
///
/// MemorySanitizer can track origins (allocation points) of all uninitialized
/// values. This behavior is controlled with a flag (msan-track-origins) and is
/// disabled by default.
///
/// Origins are 4-byte values created and interpreted by the runtime library.
/// They are stored in a second shadow mapping, one 4-byte value for 4 bytes
/// of application memory. Propagation of origins is basically a bunch of
/// "select" instructions that pick the origin of a dirty argument, if an
/// instruction has one.
///
/// Every 4 aligned, consecutive bytes of application memory have one origin
/// value associated with them. If these bytes contain uninitialized data
/// coming from 2 different allocations, the last store wins. Because of this,
/// MemorySanitizer reports can show unrelated origins, but this is unlikely in
/// practice.
///
/// Origins are meaningless for fully initialized values, so MemorySanitizer
/// avoids storing origin to memory when a fully initialized value is stored.
/// This way it avoids needless overwriting origin of the 4-byte region on
/// a short (i.e. 1 byte) clean store, and it is also good for performance.
///
/// Atomic handling.
///
/// Ideally, every atomic store of application value should update the
/// corresponding shadow location in an atomic way. Unfortunately, atomic store
/// of two disjoint locations can not be done without severe slowdown.
///
/// Therefore, we implement an approximation that may err on the safe side.
/// In this implementation, every atomically accessed location in the program
/// may only change from (partially) uninitialized to fully initialized, but
/// not the other way around. We load the shadow _after_ the application load,
/// and we store the shadow _before_ the app store. Also, we always store clean
/// shadow (if the application store is atomic). This way, if the store-load
/// pair constitutes a happens-before arc, shadow store and load are correctly
/// ordered such that the load will get either the value that was stored, or
/// some later value (which is always clean).
///
/// This does not work very well with Compare-And-Swap (CAS) and
/// Read-Modify-Write (RMW) operations. To follow the above logic, CAS and RMW
/// must store the new shadow before the app operation, and load the shadow
/// after the app operation. Computers don't work this way. Current
/// implementation ignores the load aspect of CAS/RMW, always returning a clean
/// value. It implements the store part as a simple atomic store by storing a
/// clean shadow.
///
/// Instrumenting inline assembly.
///
/// For inline assembly code LLVM has little idea about which memory locations
/// become initialized depending on the arguments. It can be possible to figure
/// out which arguments are meant to point to inputs and outputs, but the
/// actual semantics can be only visible at runtime. In the Linux kernel it's
/// also possible that the arguments only indicate the offset for a base taken
/// from a segment register, so it's dangerous to treat any asm() arguments as
/// pointers. We take a conservative approach generating calls to
/// __msan_instrument_asm_store(ptr, size)
/// , which defer the memory unpoisoning to the runtime library.
/// The latter can perform more complex address checks to figure out whether
/// it's safe to touch the shadow memory.
/// Like with atomic operations, we call __msan_instrument_asm_store() before
/// the assembly call, so that changes to the shadow memory will be seen by
/// other threads together with main memory initialization.
///
/// KernelMemorySanitizer (KMSAN) implementation.
///
/// The major differences between KMSAN and MSan instrumentation are:
/// - KMSAN always tracks the origins and implies msan-keep-going=true;
/// - KMSAN allocates shadow and origin memory for each page separately, so
/// there are no explicit accesses to shadow and origin in the
/// instrumentation.
/// Shadow and origin values for a particular X-byte memory location
/// (X=1,2,4,8) are accessed through pointers obtained via the
/// __msan_metadata_ptr_for_load_X(ptr)
/// __msan_metadata_ptr_for_store_X(ptr)
/// functions. The corresponding functions check that the X-byte accesses
/// are possible and returns the pointers to shadow and origin memory.
/// Arbitrary sized accesses are handled with:
/// __msan_metadata_ptr_for_load_n(ptr, size)
/// __msan_metadata_ptr_for_store_n(ptr, size);
/// Note that the sanitizer code has to deal with how shadow/origin pairs
/// returned by the these functions are represented in different ABIs. In
/// the X86_64 ABI they are returned in RDX:RAX, in PowerPC64 they are
/// returned in r3 and r4, and in the SystemZ ABI they are written to memory
/// pointed to by a hidden parameter.
/// - TLS variables are stored in a single per-task struct. A call to a
/// function __msan_get_context_state() returning a pointer to that struct
/// is inserted into every instrumented function before the entry block;
/// - __msan_warning() takes a 32-bit origin parameter;
/// - local variables are poisoned with __msan_poison_alloca() upon function
/// entry and unpoisoned with __msan_unpoison_alloca() before leaving the
/// function;
/// - the pass doesn't declare any global variables or add global constructors
/// to the translation unit.
///
/// Also, KMSAN currently ignores uninitialized memory passed into inline asm
/// calls, making sure we're on the safe side wrt. possible false positives.
///
/// KernelMemorySanitizer only supports X86_64, SystemZ and PowerPC64 at the
/// moment.
///
//
// FIXME: This sanitizer does not yet handle scalable vectors
//
//===----------------------------------------------------------------------===//
#include "llvm/Transforms/Instrumentation/MemorySanitizer.h"
#include "llvm/ADT/APInt.h"
#include "llvm/ADT/ArrayRef.h"
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/DepthFirstIterator.h"
#include "llvm/ADT/SetVector.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/Analysis/GlobalsModRef.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/Analysis/ValueTracking.h"
#include "llvm/IR/Argument.h"
#include "llvm/IR/AttributeMask.h"
#include "llvm/IR/Attributes.h"
#include "llvm/IR/BasicBlock.h"
#include "llvm/IR/CallingConv.h"
#include "llvm/IR/Constant.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/DerivedTypes.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/GlobalValue.h"
#include "llvm/IR/GlobalVariable.h"
#include "llvm/IR/IRBuilder.h"
#include "llvm/IR/InlineAsm.h"
#include "llvm/IR/InstVisitor.h"
#include "llvm/IR/InstrTypes.h"
#include "llvm/IR/Instruction.h"
#include "llvm/IR/Instructions.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/Intrinsics.h"
#include "llvm/IR/IntrinsicsX86.h"
#include "llvm/IR/MDBuilder.h"
#include "llvm/IR/Module.h"
#include "llvm/IR/Type.h"
#include "llvm/IR/Value.h"
#include "llvm/IR/ValueMap.h"
#include "llvm/Support/Alignment.h"
#include "llvm/Support/AtomicOrdering.h"
#include "llvm/Support/Casting.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/DebugCounter.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/MathExtras.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/TargetParser/Triple.h"
#include "llvm/Transforms/Utils/BasicBlockUtils.h"
#include "llvm/Transforms/Utils/Local.h"
#include "llvm/Transforms/Utils/ModuleUtils.h"
#include <algorithm>
#include <cassert>
#include <cstddef>
#include <cstdint>
#include <memory>
#include <string>
#include <tuple>
using namespace llvm;
#define DEBUG_TYPE "msan"
DEBUG_COUNTER(DebugInsertCheck, "msan-insert-check",
"Controls which checks to insert");
DEBUG_COUNTER(DebugInstrumentInstruction, "msan-instrument-instruction",
"Controls which instruction to instrument");
static const unsigned kOriginSize = 4;
static const Align kMinOriginAlignment = Align(4);
static const Align kShadowTLSAlignment = Align(8);
// These constants must be kept in sync with the ones in msan.h.
static const unsigned kParamTLSSize = 800;
static const unsigned kRetvalTLSSize = 800;
// Accesses sizes are powers of two: 1, 2, 4, 8.
static const size_t kNumberOfAccessSizes = 4;
/// Track origins of uninitialized values.
///
/// Adds a section to MemorySanitizer report that points to the allocation
/// (stack or heap) the uninitialized bits came from originally.
static cl::opt<int> ClTrackOrigins(
"msan-track-origins",
cl::desc("Track origins (allocation sites) of poisoned memory"), cl::Hidden,
cl::init(0));
static cl::opt<bool> ClKeepGoing("msan-keep-going",
cl::desc("keep going after reporting a UMR"),
cl::Hidden, cl::init(false));
static cl::opt<bool>
ClPoisonStack("msan-poison-stack",
cl::desc("poison uninitialized stack variables"), cl::Hidden,
cl::init(true));
static cl::opt<bool> ClPoisonStackWithCall(
"msan-poison-stack-with-call",
cl::desc("poison uninitialized stack variables with a call"), cl::Hidden,
cl::init(false));
static cl::opt<int> ClPoisonStackPattern(
"msan-poison-stack-pattern",
cl::desc("poison uninitialized stack variables with the given pattern"),
cl::Hidden, cl::init(0xff));
static cl::opt<bool>
ClPrintStackNames("msan-print-stack-names",
cl::desc("Print name of local stack variable"),
cl::Hidden, cl::init(true));
static cl::opt<bool> ClPoisonUndef("msan-poison-undef",
cl::desc("poison undef temps"), cl::Hidden,
cl::init(true));
static cl::opt<bool>
ClHandleICmp("msan-handle-icmp",
cl::desc("propagate shadow through ICmpEQ and ICmpNE"),
cl::Hidden, cl::init(true));
static cl::opt<bool>
ClHandleICmpExact("msan-handle-icmp-exact",
cl::desc("exact handling of relational integer ICmp"),
cl::Hidden, cl::init(false));
static cl::opt<bool> ClHandleLifetimeIntrinsics(
"msan-handle-lifetime-intrinsics",
cl::desc(
"when possible, poison scoped variables at the beginning of the scope "
"(slower, but more precise)"),
cl::Hidden, cl::init(true));
// When compiling the Linux kernel, we sometimes see false positives related to
// MSan being unable to understand that inline assembly calls may initialize
// local variables.
// This flag makes the compiler conservatively unpoison every memory location
// passed into an assembly call. Note that this may cause false positives.
// Because it's impossible to figure out the array sizes, we can only unpoison
// the first sizeof(type) bytes for each type* pointer.
static cl::opt<bool> ClHandleAsmConservative(
"msan-handle-asm-conservative",
cl::desc("conservative handling of inline assembly"), cl::Hidden,
cl::init(true));
// This flag controls whether we check the shadow of the address
// operand of load or store. Such bugs are very rare, since load from
// a garbage address typically results in SEGV, but still happen
// (e.g. only lower bits of address are garbage, or the access happens
// early at program startup where malloc-ed memory is more likely to
// be zeroed. As of 2012-08-28 this flag adds 20% slowdown.
static cl::opt<bool> ClCheckAccessAddress(
"msan-check-access-address",
cl::desc("report accesses through a pointer which has poisoned shadow"),
cl::Hidden, cl::init(true));
static cl::opt<bool> ClEagerChecks(
"msan-eager-checks",
cl::desc("check arguments and return values at function call boundaries"),
cl::Hidden, cl::init(false));
static cl::opt<bool> ClDumpStrictInstructions(
"msan-dump-strict-instructions",
cl::desc("print out instructions with default strict semantics"),
cl::Hidden, cl::init(false));
static cl::opt<int> ClInstrumentationWithCallThreshold(
"msan-instrumentation-with-call-threshold",
cl::desc(
"If the function being instrumented requires more than "
"this number of checks and origin stores, use callbacks instead of "
"inline checks (-1 means never use callbacks)."),
cl::Hidden, cl::init(3500));
static cl::opt<bool>
ClEnableKmsan("msan-kernel",
cl::desc("Enable KernelMemorySanitizer instrumentation"),
cl::Hidden, cl::init(false));
static cl::opt<bool>
ClDisableChecks("msan-disable-checks",
cl::desc("Apply no_sanitize to the whole file"), cl::Hidden,
cl::init(false));
static cl::opt<bool>
ClCheckConstantShadow("msan-check-constant-shadow",
cl::desc("Insert checks for constant shadow values"),
cl::Hidden, cl::init(true));
// This is off by default because of a bug in gold:
// https://sourceware.org/bugzilla/show_bug.cgi?id=19002
static cl::opt<bool>
ClWithComdat("msan-with-comdat",
cl::desc("Place MSan constructors in comdat sections"),
cl::Hidden, cl::init(false));
// These options allow to specify custom memory map parameters
// See MemoryMapParams for details.
static cl::opt<uint64_t> ClAndMask("msan-and-mask",
cl::desc("Define custom MSan AndMask"),
cl::Hidden, cl::init(0));
static cl::opt<uint64_t> ClXorMask("msan-xor-mask",
cl::desc("Define custom MSan XorMask"),
cl::Hidden, cl::init(0));
static cl::opt<uint64_t> ClShadowBase("msan-shadow-base",
cl::desc("Define custom MSan ShadowBase"),
cl::Hidden, cl::init(0));
static cl::opt<uint64_t> ClOriginBase("msan-origin-base",
cl::desc("Define custom MSan OriginBase"),
cl::Hidden, cl::init(0));
static cl::opt<int>
ClDisambiguateWarning("msan-disambiguate-warning-threshold",
cl::desc("Define threshold for number of checks per "
"debug location to force origin update."),
cl::Hidden, cl::init(3));
const char kMsanModuleCtorName[] = "msan.module_ctor";
const char kMsanInitName[] = "__msan_init";
namespace {
// Memory map parameters used in application-to-shadow address calculation.
// Offset = (Addr & ~AndMask) ^ XorMask
// Shadow = ShadowBase + Offset
// Origin = OriginBase + Offset
struct MemoryMapParams {
uint64_t AndMask;
uint64_t XorMask;
uint64_t ShadowBase;
uint64_t OriginBase;
};
struct PlatformMemoryMapParams {
const MemoryMapParams *bits32;
const MemoryMapParams *bits64;
};
} // end anonymous namespace
// i386 Linux
static const MemoryMapParams Linux_I386_MemoryMapParams = {
0x000080000000, // AndMask
0, // XorMask (not used)
0, // ShadowBase (not used)
0x000040000000, // OriginBase
};
// x86_64 Linux
static const MemoryMapParams Linux_X86_64_MemoryMapParams = {
0, // AndMask (not used)
0x500000000000, // XorMask
0, // ShadowBase (not used)
0x100000000000, // OriginBase
};
// mips64 Linux
static const MemoryMapParams Linux_MIPS64_MemoryMapParams = {
0, // AndMask (not used)
0x008000000000, // XorMask
0, // ShadowBase (not used)
0x002000000000, // OriginBase
};
// ppc64 Linux
static const MemoryMapParams Linux_PowerPC64_MemoryMapParams = {
0xE00000000000, // AndMask
0x100000000000, // XorMask
0x080000000000, // ShadowBase
0x1C0000000000, // OriginBase
};
// s390x Linux
static const MemoryMapParams Linux_S390X_MemoryMapParams = {
0xC00000000000, // AndMask
0, // XorMask (not used)
0x080000000000, // ShadowBase
0x1C0000000000, // OriginBase
};
// aarch64 Linux
static const MemoryMapParams Linux_AArch64_MemoryMapParams = {
0, // AndMask (not used)
0x0B00000000000, // XorMask
0, // ShadowBase (not used)
0x0200000000000, // OriginBase
};
// loongarch64 Linux
static const MemoryMapParams Linux_LoongArch64_MemoryMapParams = {
0, // AndMask (not used)
0x500000000000, // XorMask
0, // ShadowBase (not used)
0x100000000000, // OriginBase
};
// aarch64 FreeBSD
static const MemoryMapParams FreeBSD_AArch64_MemoryMapParams = {
0x1800000000000, // AndMask
0x0400000000000, // XorMask
0x0200000000000, // ShadowBase
0x0700000000000, // OriginBase
};
// i386 FreeBSD
static const MemoryMapParams FreeBSD_I386_MemoryMapParams = {
0x000180000000, // AndMask
0x000040000000, // XorMask
0x000020000000, // ShadowBase
0x000700000000, // OriginBase
};
// x86_64 FreeBSD
static const MemoryMapParams FreeBSD_X86_64_MemoryMapParams = {
0xc00000000000, // AndMask
0x200000000000, // XorMask
0x100000000000, // ShadowBase
0x380000000000, // OriginBase
};
// x86_64 NetBSD
static const MemoryMapParams NetBSD_X86_64_MemoryMapParams = {
0, // AndMask
0x500000000000, // XorMask
0, // ShadowBase
0x100000000000, // OriginBase
};
static const PlatformMemoryMapParams Linux_X86_MemoryMapParams = {
&Linux_I386_MemoryMapParams,
&Linux_X86_64_MemoryMapParams,
};
static const PlatformMemoryMapParams Linux_MIPS_MemoryMapParams = {
nullptr,
&Linux_MIPS64_MemoryMapParams,
};
static const PlatformMemoryMapParams Linux_PowerPC_MemoryMapParams = {
nullptr,
&Linux_PowerPC64_MemoryMapParams,
};
static const PlatformMemoryMapParams Linux_S390_MemoryMapParams = {
nullptr,
&Linux_S390X_MemoryMapParams,
};
static const PlatformMemoryMapParams Linux_ARM_MemoryMapParams = {
nullptr,
&Linux_AArch64_MemoryMapParams,
};
static const PlatformMemoryMapParams Linux_LoongArch_MemoryMapParams = {
nullptr,
&Linux_LoongArch64_MemoryMapParams,
};
static const PlatformMemoryMapParams FreeBSD_ARM_MemoryMapParams = {
nullptr,
&FreeBSD_AArch64_MemoryMapParams,
};
static const PlatformMemoryMapParams FreeBSD_X86_MemoryMapParams = {
&FreeBSD_I386_MemoryMapParams,
&FreeBSD_X86_64_MemoryMapParams,
};
static const PlatformMemoryMapParams NetBSD_X86_MemoryMapParams = {
nullptr,
&NetBSD_X86_64_MemoryMapParams,
};
namespace {
/// Instrument functions of a module to detect uninitialized reads.
///
/// Instantiating MemorySanitizer inserts the msan runtime library API function
/// declarations into the module if they don't exist already. Instantiating
/// ensures the __msan_init function is in the list of global constructors for
/// the module.
class MemorySanitizer {
public:
MemorySanitizer(Module &M, MemorySanitizerOptions Options)
: CompileKernel(Options.Kernel), TrackOrigins(Options.TrackOrigins),
Recover(Options.Recover), EagerChecks(Options.EagerChecks) {
initializeModule(M);
}
// MSan cannot be moved or copied because of MapParams.
MemorySanitizer(MemorySanitizer &&) = delete;
MemorySanitizer &operator=(MemorySanitizer &&) = delete;
MemorySanitizer(const MemorySanitizer &) = delete;
MemorySanitizer &operator=(const MemorySanitizer &) = delete;
bool sanitizeFunction(Function &F, TargetLibraryInfo &TLI);
private:
friend struct MemorySanitizerVisitor;
friend struct VarArgHelperBase;
friend struct VarArgAMD64Helper;
friend struct VarArgMIPS64Helper;
friend struct VarArgAArch64Helper;
friend struct VarArgPowerPC64Helper;
friend struct VarArgSystemZHelper;
void initializeModule(Module &M);
void initializeCallbacks(Module &M, const TargetLibraryInfo &TLI);
void createKernelApi(Module &M, const TargetLibraryInfo &TLI);
void createUserspaceApi(Module &M, const TargetLibraryInfo &TLI);
template <typename... ArgsTy>
FunctionCallee getOrInsertMsanMetadataFunction(Module &M, StringRef Name,
ArgsTy... Args);
/// True if we're compiling the Linux kernel.
bool CompileKernel;
/// Track origins (allocation points) of uninitialized values.
int TrackOrigins;
bool Recover;
bool EagerChecks;
Triple TargetTriple;
LLVMContext *C;
Type *IntptrTy; ///< Integer type with the size of a ptr in default AS.
Type *OriginTy;
PointerType *PtrTy; ///< Integer type with the size of a ptr in default AS.
// XxxTLS variables represent the per-thread state in MSan and per-task state
// in KMSAN.
// For the userspace these point to thread-local globals. In the kernel land
// they point to the members of a per-task struct obtained via a call to
// __msan_get_context_state().
/// Thread-local shadow storage for function parameters.
Value *ParamTLS;
/// Thread-local origin storage for function parameters.
Value *ParamOriginTLS;
/// Thread-local shadow storage for function return value.
Value *RetvalTLS;
/// Thread-local origin storage for function return value.
Value *RetvalOriginTLS;
/// Thread-local shadow storage for in-register va_arg function.
Value *VAArgTLS;
/// Thread-local shadow storage for in-register va_arg function.
Value *VAArgOriginTLS;
/// Thread-local shadow storage for va_arg overflow area.
Value *VAArgOverflowSizeTLS;
/// Are the instrumentation callbacks set up?
bool CallbacksInitialized = false;
/// The run-time callback to print a warning.
FunctionCallee WarningFn;
// These arrays are indexed by log2(AccessSize).
FunctionCallee MaybeWarningFn[kNumberOfAccessSizes];
FunctionCallee MaybeStoreOriginFn[kNumberOfAccessSizes];
/// Run-time helper that generates a new origin value for a stack
/// allocation.
FunctionCallee MsanSetAllocaOriginWithDescriptionFn;
// No description version
FunctionCallee MsanSetAllocaOriginNoDescriptionFn;
/// Run-time helper that poisons stack on function entry.
FunctionCallee MsanPoisonStackFn;
/// Run-time helper that records a store (or any event) of an
/// uninitialized value and returns an updated origin id encoding this info.
FunctionCallee MsanChainOriginFn;
/// Run-time helper that paints an origin over a region.
FunctionCallee MsanSetOriginFn;
/// MSan runtime replacements for memmove, memcpy and memset.
FunctionCallee MemmoveFn, MemcpyFn, MemsetFn;
/// KMSAN callback for task-local function argument shadow.
StructType *MsanContextStateTy;
FunctionCallee MsanGetContextStateFn;
/// Functions for poisoning/unpoisoning local variables
FunctionCallee MsanPoisonAllocaFn, MsanUnpoisonAllocaFn;
/// Pair of shadow/origin pointers.
Type *MsanMetadata;
/// Each of the MsanMetadataPtrXxx functions returns a MsanMetadata.
FunctionCallee MsanMetadataPtrForLoadN, MsanMetadataPtrForStoreN;
FunctionCallee MsanMetadataPtrForLoad_1_8[4];
FunctionCallee MsanMetadataPtrForStore_1_8[4];
FunctionCallee MsanInstrumentAsmStoreFn;
/// Storage for return values of the MsanMetadataPtrXxx functions.
Value *MsanMetadataAlloca;
/// Helper to choose between different MsanMetadataPtrXxx().
FunctionCallee getKmsanShadowOriginAccessFn(bool isStore, int size);
/// Memory map parameters used in application-to-shadow calculation.
const MemoryMapParams *MapParams;
/// Custom memory map parameters used when -msan-shadow-base or
// -msan-origin-base is provided.
MemoryMapParams CustomMapParams;
MDNode *ColdCallWeights;
/// Branch weights for origin store.
MDNode *OriginStoreWeights;
};
void insertModuleCtor(Module &M) {
getOrCreateSanitizerCtorAndInitFunctions(
M, kMsanModuleCtorName, kMsanInitName,
/*InitArgTypes=*/{},
/*InitArgs=*/{},
// This callback is invoked when the functions are created the first
// time. Hook them into the global ctors list in that case:
[&](Function *Ctor, FunctionCallee) {
if (!ClWithComdat) {
appendToGlobalCtors(M, Ctor, 0);
return;
}
Comdat *MsanCtorComdat = M.getOrInsertComdat(kMsanModuleCtorName);
Ctor->setComdat(MsanCtorComdat);
appendToGlobalCtors(M, Ctor, 0, Ctor);
});
}
template <class T> T getOptOrDefault(const cl::opt<T> &Opt, T Default) {
return (Opt.getNumOccurrences() > 0) ? Opt : Default;
}
} // end anonymous namespace
MemorySanitizerOptions::MemorySanitizerOptions(int TO, bool R, bool K,
bool EagerChecks)
: Kernel(getOptOrDefault(ClEnableKmsan, K)),
TrackOrigins(getOptOrDefault(ClTrackOrigins, Kernel ? 2 : TO)),
Recover(getOptOrDefault(ClKeepGoing, Kernel || R)),
EagerChecks(getOptOrDefault(ClEagerChecks, EagerChecks)) {}
PreservedAnalyses MemorySanitizerPass::run(Module &M,
ModuleAnalysisManager &AM) {
bool Modified = false;
if (!Options.Kernel) {
insertModuleCtor(M);
Modified = true;
}
auto &FAM = AM.getResult<FunctionAnalysisManagerModuleProxy>(M).getManager();
for (Function &F : M) {
if (F.empty())
continue;
MemorySanitizer Msan(*F.getParent(), Options);
Modified |=
Msan.sanitizeFunction(F, FAM.getResult<TargetLibraryAnalysis>(F));
}
if (!Modified)
return PreservedAnalyses::all();
PreservedAnalyses PA = PreservedAnalyses::none();
// GlobalsAA is considered stateless and does not get invalidated unless
// explicitly invalidated; PreservedAnalyses::none() is not enough. Sanitizers
// make changes that require GlobalsAA to be invalidated.
PA.abandon<GlobalsAA>();
return PA;
}
void MemorySanitizerPass::printPipeline(
raw_ostream &OS, function_ref<StringRef(StringRef)> MapClassName2PassName) {
static_cast<PassInfoMixin<MemorySanitizerPass> *>(this)->printPipeline(
OS, MapClassName2PassName);
OS << '<';
if (Options.Recover)
OS << "recover;";
if (Options.Kernel)
OS << "kernel;";
if (Options.EagerChecks)
OS << "eager-checks;";
OS << "track-origins=" << Options.TrackOrigins;
OS << '>';
}
/// Create a non-const global initialized with the given string.
///
/// Creates a writable global for Str so that we can pass it to the
/// run-time lib. Runtime uses first 4 bytes of the string to store the
/// frame ID, so the string needs to be mutable.
static GlobalVariable *createPrivateConstGlobalForString(Module &M,
StringRef Str) {
Constant *StrConst = ConstantDataArray::getString(M.getContext(), Str);
return new GlobalVariable(M, StrConst->getType(), /*isConstant=*/true,
GlobalValue::PrivateLinkage, StrConst, "");
}
template <typename... ArgsTy>
FunctionCallee
MemorySanitizer::getOrInsertMsanMetadataFunction(Module &M, StringRef Name,
ArgsTy... Args) {
if (TargetTriple.getArch() == Triple::systemz) {
// SystemZ ABI: shadow/origin pair is returned via a hidden parameter.
return M.getOrInsertFunction(Name, Type::getVoidTy(*C),
PointerType::get(MsanMetadata, 0),
std::forward<ArgsTy>(Args)...);
}
return M.getOrInsertFunction(Name, MsanMetadata,
std::forward<ArgsTy>(Args)...);
}
/// Create KMSAN API callbacks.
void MemorySanitizer::createKernelApi(Module &M, const TargetLibraryInfo &TLI) {
IRBuilder<> IRB(*C);
// These will be initialized in insertKmsanPrologue().
RetvalTLS = nullptr;
RetvalOriginTLS = nullptr;
ParamTLS = nullptr;
ParamOriginTLS = nullptr;
VAArgTLS = nullptr;
VAArgOriginTLS = nullptr;
VAArgOverflowSizeTLS = nullptr;
WarningFn = M.getOrInsertFunction("__msan_warning",
TLI.getAttrList(C, {0}, /*Signed=*/false),
IRB.getVoidTy(), IRB.getInt32Ty());
// Requests the per-task context state (kmsan_context_state*) from the
// runtime library.
MsanContextStateTy = StructType::get(
ArrayType::get(IRB.getInt64Ty(), kParamTLSSize / 8),
ArrayType::get(IRB.getInt64Ty(), kRetvalTLSSize / 8),
ArrayType::get(IRB.getInt64Ty(), kParamTLSSize / 8),
ArrayType::get(IRB.getInt64Ty(), kParamTLSSize / 8), /* va_arg_origin */
IRB.getInt64Ty(), ArrayType::get(OriginTy, kParamTLSSize / 4), OriginTy,
OriginTy);
MsanGetContextStateFn = M.getOrInsertFunction(
"__msan_get_context_state", PointerType::get(MsanContextStateTy, 0));
MsanMetadata = StructType::get(PointerType::get(IRB.getInt8Ty(), 0),
PointerType::get(IRB.getInt32Ty(), 0));
for (int ind = 0, size = 1; ind < 4; ind++, size <<= 1) {
std::string name_load =
"__msan_metadata_ptr_for_load_" + std::to_string(size);
std::string name_store =
"__msan_metadata_ptr_for_store_" + std::to_string(size);
MsanMetadataPtrForLoad_1_8[ind] = getOrInsertMsanMetadataFunction(
M, name_load, PointerType::get(IRB.getInt8Ty(), 0));
MsanMetadataPtrForStore_1_8[ind] = getOrInsertMsanMetadataFunction(
M, name_store, PointerType::get(IRB.getInt8Ty(), 0));
}
MsanMetadataPtrForLoadN = getOrInsertMsanMetadataFunction(
M, "__msan_metadata_ptr_for_load_n", PointerType::get(IRB.getInt8Ty(), 0),
IRB.getInt64Ty());
MsanMetadataPtrForStoreN = getOrInsertMsanMetadataFunction(
M, "__msan_metadata_ptr_for_store_n",
PointerType::get(IRB.getInt8Ty(), 0), IRB.getInt64Ty());
// Functions for poisoning and unpoisoning memory.
MsanPoisonAllocaFn = M.getOrInsertFunction(
"__msan_poison_alloca", IRB.getVoidTy(), PtrTy, IntptrTy, PtrTy);
MsanUnpoisonAllocaFn = M.getOrInsertFunction(
"__msan_unpoison_alloca", IRB.getVoidTy(), PtrTy, IntptrTy);
}
static Constant *getOrInsertGlobal(Module &M, StringRef Name, Type *Ty) {
return M.getOrInsertGlobal(Name, Ty, [&] {
return new GlobalVariable(M, Ty, false, GlobalVariable::ExternalLinkage,
nullptr, Name, nullptr,
GlobalVariable::InitialExecTLSModel);
});
}
/// Insert declarations for userspace-specific functions and globals.
void MemorySanitizer::createUserspaceApi(Module &M, const TargetLibraryInfo &TLI) {
IRBuilder<> IRB(*C);
// Create the callback.
// FIXME: this function should have "Cold" calling conv,
// which is not yet implemented.
if (TrackOrigins) {
StringRef WarningFnName = Recover ? "__msan_warning_with_origin"
: "__msan_warning_with_origin_noreturn";
WarningFn = M.getOrInsertFunction(WarningFnName,
TLI.getAttrList(C, {0}, /*Signed=*/false),
IRB.getVoidTy(), IRB.getInt32Ty());
} else {
StringRef WarningFnName =
Recover ? "__msan_warning" : "__msan_warning_noreturn";
WarningFn = M.getOrInsertFunction(WarningFnName, IRB.getVoidTy());
}
// Create the global TLS variables.
RetvalTLS =
getOrInsertGlobal(M, "__msan_retval_tls",
ArrayType::get(IRB.getInt64Ty(), kRetvalTLSSize / 8));
RetvalOriginTLS = getOrInsertGlobal(M, "__msan_retval_origin_tls", OriginTy);
ParamTLS =
getOrInsertGlobal(M, "__msan_param_tls",
ArrayType::get(IRB.getInt64Ty(), kParamTLSSize / 8));
ParamOriginTLS =
getOrInsertGlobal(M, "__msan_param_origin_tls",
ArrayType::get(OriginTy, kParamTLSSize / 4));
VAArgTLS =
getOrInsertGlobal(M, "__msan_va_arg_tls",
ArrayType::get(IRB.getInt64Ty(), kParamTLSSize / 8));
VAArgOriginTLS =
getOrInsertGlobal(M, "__msan_va_arg_origin_tls",
ArrayType::get(OriginTy, kParamTLSSize / 4));
VAArgOverflowSizeTLS =
getOrInsertGlobal(M, "__msan_va_arg_overflow_size_tls", IRB.getInt64Ty());
for (size_t AccessSizeIndex = 0; AccessSizeIndex < kNumberOfAccessSizes;
AccessSizeIndex++) {
unsigned AccessSize = 1 << AccessSizeIndex;
std::string FunctionName = "__msan_maybe_warning_" + itostr(AccessSize);
MaybeWarningFn[AccessSizeIndex] = M.getOrInsertFunction(
FunctionName, TLI.getAttrList(C, {0, 1}, /*Signed=*/false),
IRB.getVoidTy(), IRB.getIntNTy(AccessSize * 8), IRB.getInt32Ty());
FunctionName = "__msan_maybe_store_origin_" + itostr(AccessSize);
MaybeStoreOriginFn[AccessSizeIndex] = M.getOrInsertFunction(
FunctionName, TLI.getAttrList(C, {0, 2}, /*Signed=*/false),
IRB.getVoidTy(), IRB.getIntNTy(AccessSize * 8), PtrTy,
IRB.getInt32Ty());
}
MsanSetAllocaOriginWithDescriptionFn =
M.getOrInsertFunction("__msan_set_alloca_origin_with_descr",
IRB.getVoidTy(), PtrTy, IntptrTy, PtrTy, PtrTy);
MsanSetAllocaOriginNoDescriptionFn =
M.getOrInsertFunction("__msan_set_alloca_origin_no_descr",
IRB.getVoidTy(), PtrTy, IntptrTy, PtrTy);
MsanPoisonStackFn = M.getOrInsertFunction("__msan_poison_stack",
IRB.getVoidTy(), PtrTy, IntptrTy);
}
/// Insert extern declaration of runtime-provided functions and globals.
void MemorySanitizer::initializeCallbacks(Module &M, const TargetLibraryInfo &TLI) {
// Only do this once.
if (CallbacksInitialized)
return;
IRBuilder<> IRB(*C);
// Initialize callbacks that are common for kernel and userspace
// instrumentation.
MsanChainOriginFn = M.getOrInsertFunction(
"__msan_chain_origin",
TLI.getAttrList(C, {0}, /*Signed=*/false, /*Ret=*/true), IRB.getInt32Ty(),
IRB.getInt32Ty());
MsanSetOriginFn = M.getOrInsertFunction(
"__msan_set_origin", TLI.getAttrList(C, {2}, /*Signed=*/false),
IRB.getVoidTy(), PtrTy, IntptrTy, IRB.getInt32Ty());
MemmoveFn =
M.getOrInsertFunction("__msan_memmove", PtrTy, PtrTy, PtrTy, IntptrTy);
MemcpyFn =
M.getOrInsertFunction("__msan_memcpy", PtrTy, PtrTy, PtrTy, IntptrTy);
MemsetFn = M.getOrInsertFunction("__msan_memset",
TLI.getAttrList(C, {1}, /*Signed=*/true),
PtrTy, PtrTy, IRB.getInt32Ty(), IntptrTy);
MsanInstrumentAsmStoreFn =
M.getOrInsertFunction("__msan_instrument_asm_store", IRB.getVoidTy(),
PointerType::get(IRB.getInt8Ty(), 0), IntptrTy);
if (CompileKernel) {
createKernelApi(M, TLI);
} else {
createUserspaceApi(M, TLI);
}
CallbacksInitialized = true;
}
FunctionCallee MemorySanitizer::getKmsanShadowOriginAccessFn(bool isStore,
int size) {
FunctionCallee *Fns =
isStore ? MsanMetadataPtrForStore_1_8 : MsanMetadataPtrForLoad_1_8;
switch (size) {
case 1:
return Fns[0];
case 2:
return Fns[1];
case 4:
return Fns[2];
case 8:
return Fns[3];
default:
return nullptr;
}
}
/// Module-level initialization.
///
/// inserts a call to __msan_init to the module's constructor list.
void MemorySanitizer::initializeModule(Module &M) {
auto &DL = M.getDataLayout();
TargetTriple = Triple(M.getTargetTriple());
bool ShadowPassed = ClShadowBase.getNumOccurrences() > 0;
bool OriginPassed = ClOriginBase.getNumOccurrences() > 0;
// Check the overrides first
if (ShadowPassed || OriginPassed) {
CustomMapParams.AndMask = ClAndMask;
CustomMapParams.XorMask = ClXorMask;
CustomMapParams.ShadowBase = ClShadowBase;
CustomMapParams.OriginBase = ClOriginBase;
MapParams = &CustomMapParams;
} else {
switch (TargetTriple.getOS()) {
case Triple::FreeBSD:
switch (TargetTriple.getArch()) {
case Triple::aarch64:
MapParams = FreeBSD_ARM_MemoryMapParams.bits64;
break;
case Triple::x86_64:
MapParams = FreeBSD_X86_MemoryMapParams.bits64;
break;
case Triple::x86:
MapParams = FreeBSD_X86_MemoryMapParams.bits32;
break;
default:
report_fatal_error("unsupported architecture");
}
break;
case Triple::NetBSD:
switch (TargetTriple.getArch()) {
case Triple::x86_64:
MapParams = NetBSD_X86_MemoryMapParams.bits64;
break;
default:
report_fatal_error("unsupported architecture");
}
break;
case Triple::Linux:
switch (TargetTriple.getArch()) {
case Triple::x86_64:
MapParams = Linux_X86_MemoryMapParams.bits64;
break;
case Triple::x86:
MapParams = Linux_X86_MemoryMapParams.bits32;
break;
case Triple::mips64:
case Triple::mips64el:
MapParams = Linux_MIPS_MemoryMapParams.bits64;
break;
case Triple::ppc64:
case Triple::ppc64le:
MapParams = Linux_PowerPC_MemoryMapParams.bits64;
break;
case Triple::systemz:
MapParams = Linux_S390_MemoryMapParams.bits64;
break;
case Triple::aarch64:
case Triple::aarch64_be:
MapParams = Linux_ARM_MemoryMapParams.bits64;
break;
case Triple::loongarch64:
MapParams = Linux_LoongArch_MemoryMapParams.bits64;
break;
default:
report_fatal_error("unsupported architecture");
}
break;
default:
report_fatal_error("unsupported operating system");
}
}
C = &(M.getContext());
IRBuilder<> IRB(*C);
IntptrTy = IRB.getIntPtrTy(DL);
OriginTy = IRB.getInt32Ty();
PtrTy = IRB.getPtrTy();
ColdCallWeights = MDBuilder(*C).createUnlikelyBranchWeights();
OriginStoreWeights = MDBuilder(*C).createUnlikelyBranchWeights();
if (!CompileKernel) {
if (TrackOrigins)
M.getOrInsertGlobal("__msan_track_origins", IRB.getInt32Ty(), [&] {
return new GlobalVariable(
M, IRB.getInt32Ty(), true, GlobalValue::WeakODRLinkage,
IRB.getInt32(TrackOrigins), "__msan_track_origins");
});
if (Recover)
M.getOrInsertGlobal("__msan_keep_going", IRB.getInt32Ty(), [&] {
return new GlobalVariable(M, IRB.getInt32Ty(), true,
GlobalValue::WeakODRLinkage,
IRB.getInt32(Recover), "__msan_keep_going");
});
}
}
namespace {
/// A helper class that handles instrumentation of VarArg
/// functions on a particular platform.
///
/// Implementations are expected to insert the instrumentation
/// necessary to propagate argument shadow through VarArg function
/// calls. Visit* methods are called during an InstVisitor pass over
/// the function, and should avoid creating new basic blocks. A new
/// instance of this class is created for each instrumented function.
struct VarArgHelper {
virtual ~VarArgHelper() = default;
/// Visit a CallBase.
virtual void visitCallBase(CallBase &CB, IRBuilder<> &IRB) = 0;
/// Visit a va_start call.
virtual void visitVAStartInst(VAStartInst &I) = 0;
/// Visit a va_copy call.
virtual void visitVACopyInst(VACopyInst &I) = 0;
/// Finalize function instrumentation.
///
/// This method is called after visiting all interesting (see above)
/// instructions in a function.
virtual void finalizeInstrumentation() = 0;
};
struct MemorySanitizerVisitor;
} // end anonymous namespace
static VarArgHelper *CreateVarArgHelper(Function &Func, MemorySanitizer &Msan,
MemorySanitizerVisitor &Visitor);
static unsigned TypeSizeToSizeIndex(TypeSize TS) {
if (TS.isScalable())
// Scalable types unconditionally take slowpaths.
return kNumberOfAccessSizes;
unsigned TypeSizeFixed = TS.getFixedValue();
if (TypeSizeFixed <= 8)
return 0;
return Log2_32_Ceil((TypeSizeFixed + 7) / 8);
}
namespace {
/// Helper class to attach debug information of the given instruction onto new
/// instructions inserted after.
class NextNodeIRBuilder : public IRBuilder<> {
public:
explicit NextNodeIRBuilder(Instruction *IP) : IRBuilder<>(IP->getNextNode()) {
SetCurrentDebugLocation(IP->getDebugLoc());
}
};
/// This class does all the work for a given function. Store and Load
/// instructions store and load corresponding shadow and origin
/// values. Most instructions propagate shadow from arguments to their
/// return values. Certain instructions (most importantly, BranchInst)
/// test their argument shadow and print reports (with a runtime call) if it's
/// non-zero.
struct MemorySanitizerVisitor : public InstVisitor<MemorySanitizerVisitor> {
Function &F;
MemorySanitizer &MS;
SmallVector<PHINode *, 16> ShadowPHINodes, OriginPHINodes;
ValueMap<Value *, Value *> ShadowMap, OriginMap;
std::unique_ptr<VarArgHelper> VAHelper;
const TargetLibraryInfo *TLI;
Instruction *FnPrologueEnd;
SmallVector<Instruction *, 16> Instructions;
// The following flags disable parts of MSan instrumentation based on
// exclusion list contents and command-line options.
bool InsertChecks;
bool PropagateShadow;
bool PoisonStack;
bool PoisonUndef;
struct ShadowOriginAndInsertPoint {
Value *Shadow;
Value *Origin;
Instruction *OrigIns;
ShadowOriginAndInsertPoint(Value *S, Value *O, Instruction *I)
: Shadow(S), Origin(O), OrigIns(I) {}
};
SmallVector<ShadowOriginAndInsertPoint, 16> InstrumentationList;
DenseMap<const DILocation *, int> LazyWarningDebugLocationCount;
bool InstrumentLifetimeStart = ClHandleLifetimeIntrinsics;
SmallSetVector<AllocaInst *, 16> AllocaSet;
SmallVector<std::pair<IntrinsicInst *, AllocaInst *>, 16> LifetimeStartList;
SmallVector<StoreInst *, 16> StoreList;
int64_t SplittableBlocksCount = 0;
MemorySanitizerVisitor(Function &F, MemorySanitizer &MS,
const TargetLibraryInfo &TLI)
: F(F), MS(MS), VAHelper(CreateVarArgHelper(F, MS, *this)), TLI(&TLI) {
bool SanitizeFunction =
F.hasFnAttribute(Attribute::SanitizeMemory) && !ClDisableChecks;
InsertChecks = SanitizeFunction;
PropagateShadow = SanitizeFunction;
PoisonStack = SanitizeFunction && ClPoisonStack;
PoisonUndef = SanitizeFunction && ClPoisonUndef;
// In the presence of unreachable blocks, we may see Phi nodes with
// incoming nodes from such blocks. Since InstVisitor skips unreachable
// blocks, such nodes will not have any shadow value associated with them.
// It's easier to remove unreachable blocks than deal with missing shadow.
removeUnreachableBlocks(F);
MS.initializeCallbacks(*F.getParent(), TLI);
FnPrologueEnd = IRBuilder<>(F.getEntryBlock().getFirstNonPHI())
.CreateIntrinsic(Intrinsic::donothing, {}, {});
if (MS.CompileKernel) {
IRBuilder<> IRB(FnPrologueEnd);
insertKmsanPrologue(IRB);
}
LLVM_DEBUG(if (!InsertChecks) dbgs()
<< "MemorySanitizer is not inserting checks into '"
<< F.getName() << "'\n");
}
bool instrumentWithCalls(Value *V) {
// Constants likely will be eliminated by follow-up passes.
if (isa<Constant>(V))
return false;
++SplittableBlocksCount;
return ClInstrumentationWithCallThreshold >= 0 &&
SplittableBlocksCount > ClInstrumentationWithCallThreshold;
}
bool isInPrologue(Instruction &I) {
return I.getParent() == FnPrologueEnd->getParent() &&
(&I == FnPrologueEnd || I.comesBefore(FnPrologueEnd));
}
// Creates a new origin and records the stack trace. In general we can call
// this function for any origin manipulation we like. However it will cost
// runtime resources. So use this wisely only if it can provide additional
// information helpful to a user.
Value *updateOrigin(Value *V, IRBuilder<> &IRB) {
if (MS.TrackOrigins <= 1)
return V;
return IRB.CreateCall(MS.MsanChainOriginFn, V);
}
Value *originToIntptr(IRBuilder<> &IRB, Value *Origin) {
const DataLayout &DL = F.getParent()->getDataLayout();
unsigned IntptrSize = DL.getTypeStoreSize(MS.IntptrTy);
if (IntptrSize == kOriginSize)
return Origin;
assert(IntptrSize == kOriginSize * 2);
Origin = IRB.CreateIntCast(Origin, MS.IntptrTy, /* isSigned */ false);
return IRB.CreateOr(Origin, IRB.CreateShl(Origin, kOriginSize * 8));
}
/// Fill memory range with the given origin value.
void paintOrigin(IRBuilder<> &IRB, Value *Origin, Value *OriginPtr,
TypeSize TS, Align Alignment) {
const DataLayout &DL = F.getParent()->getDataLayout();
const Align IntptrAlignment = DL.getABITypeAlign(MS.IntptrTy);
unsigned IntptrSize = DL.getTypeStoreSize(MS.IntptrTy);
assert(IntptrAlignment >= kMinOriginAlignment);
assert(IntptrSize >= kOriginSize);
// Note: The loop based formation works for fixed length vectors too,
// however we prefer to unroll and specialize alignment below.
if (TS.isScalable()) {
Value *Size = IRB.CreateTypeSize(MS.IntptrTy, TS);
Value *RoundUp =
IRB.CreateAdd(Size, ConstantInt::get(MS.IntptrTy, kOriginSize - 1));
Value *End =
IRB.CreateUDiv(RoundUp, ConstantInt::get(MS.IntptrTy, kOriginSize));
auto [InsertPt, Index] =
SplitBlockAndInsertSimpleForLoop(End, &*IRB.GetInsertPoint());
IRB.SetInsertPoint(InsertPt);
Value *GEP = IRB.CreateGEP(MS.OriginTy, OriginPtr, Index);
IRB.CreateAlignedStore(Origin, GEP, kMinOriginAlignment);
return;
}
unsigned Size = TS.getFixedValue();
unsigned Ofs = 0;
Align CurrentAlignment = Alignment;
if (Alignment >= IntptrAlignment && IntptrSize > kOriginSize) {
Value *IntptrOrigin = originToIntptr(IRB, Origin);
Value *IntptrOriginPtr =
IRB.CreatePointerCast(OriginPtr, PointerType::get(MS.IntptrTy, 0));
for (unsigned i = 0; i < Size / IntptrSize; ++i) {
Value *Ptr = i ? IRB.CreateConstGEP1_32(MS.IntptrTy, IntptrOriginPtr, i)
: IntptrOriginPtr;
IRB.CreateAlignedStore(IntptrOrigin, Ptr, CurrentAlignment);
Ofs += IntptrSize / kOriginSize;
CurrentAlignment = IntptrAlignment;
}
}
for (unsigned i = Ofs; i < (Size + kOriginSize - 1) / kOriginSize; ++i) {
Value *GEP =
i ? IRB.CreateConstGEP1_32(MS.OriginTy, OriginPtr, i) : OriginPtr;
IRB.CreateAlignedStore(Origin, GEP, CurrentAlignment);
CurrentAlignment = kMinOriginAlignment;
}
}
void storeOrigin(IRBuilder<> &IRB, Value *Addr, Value *Shadow, Value *Origin,
Value *OriginPtr, Align Alignment) {
const DataLayout &DL = F.getParent()->getDataLayout();
const Align OriginAlignment = std::max(kMinOriginAlignment, Alignment);
TypeSize StoreSize = DL.getTypeStoreSize(Shadow->getType());
Value *ConvertedShadow = convertShadowToScalar(Shadow, IRB);
if (auto *ConstantShadow = dyn_cast<Constant>(ConvertedShadow)) {
if (!ClCheckConstantShadow || ConstantShadow->isZeroValue()) {
// Origin is not needed: value is initialized or const shadow is
// ignored.
return;
}
if (llvm::isKnownNonZero(ConvertedShadow, DL)) {
// Copy origin as the value is definitely uninitialized.
paintOrigin(IRB, updateOrigin(Origin, IRB), OriginPtr, StoreSize,
OriginAlignment);
return;
}
// Fallback to runtime check, which still can be optimized out later.
}
TypeSize TypeSizeInBits = DL.getTypeSizeInBits(ConvertedShadow->getType());
unsigned SizeIndex = TypeSizeToSizeIndex(TypeSizeInBits);
if (instrumentWithCalls(ConvertedShadow) &&
SizeIndex < kNumberOfAccessSizes && !MS.CompileKernel) {
FunctionCallee Fn = MS.MaybeStoreOriginFn[SizeIndex];
Value *ConvertedShadow2 =
IRB.CreateZExt(ConvertedShadow, IRB.getIntNTy(8 * (1 << SizeIndex)));
CallBase *CB = IRB.CreateCall(Fn, {ConvertedShadow2, Addr, Origin});
CB->addParamAttr(0, Attribute::ZExt);
CB->addParamAttr(2, Attribute::ZExt);
} else {
Value *Cmp = convertToBool(ConvertedShadow, IRB, "_mscmp");
Instruction *CheckTerm = SplitBlockAndInsertIfThen(
Cmp, &*IRB.GetInsertPoint(), false, MS.OriginStoreWeights);
IRBuilder<> IRBNew(CheckTerm);
paintOrigin(IRBNew, updateOrigin(Origin, IRBNew), OriginPtr, StoreSize,
OriginAlignment);
}
}
void materializeStores() {
for (StoreInst *SI : StoreList) {
IRBuilder<> IRB(SI);
Value *Val = SI->getValueOperand();
Value *Addr = SI->getPointerOperand();
Value *Shadow = SI->isAtomic() ? getCleanShadow(Val) : getShadow(Val);
Value *ShadowPtr, *OriginPtr;
Type *ShadowTy = Shadow->getType();
const Align Alignment = SI->getAlign();
const Align OriginAlignment = std::max(kMinOriginAlignment, Alignment);
std::tie(ShadowPtr, OriginPtr) =
getShadowOriginPtr(Addr, IRB, ShadowTy, Alignment, /*isStore*/ true);
StoreInst *NewSI = IRB.CreateAlignedStore(Shadow, ShadowPtr, Alignment);
LLVM_DEBUG(dbgs() << " STORE: " << *NewSI << "\n");
(void)NewSI;
if (SI->isAtomic())
SI->setOrdering(addReleaseOrdering(SI->getOrdering()));
if (MS.TrackOrigins && !SI->isAtomic())
storeOrigin(IRB, Addr, Shadow, getOrigin(Val), OriginPtr,
OriginAlignment);
}
}
// Returns true if Debug Location corresponds to multiple warnings.
bool shouldDisambiguateWarningLocation(const DebugLoc &DebugLoc) {
if (MS.TrackOrigins < 2)
return false;
if (LazyWarningDebugLocationCount.empty())
for (const auto &I : InstrumentationList)
++LazyWarningDebugLocationCount[I.OrigIns->getDebugLoc()];
return LazyWarningDebugLocationCount[DebugLoc] >= ClDisambiguateWarning;
}
/// Helper function to insert a warning at IRB's current insert point.
void insertWarningFn(IRBuilder<> &IRB, Value *Origin) {
if (!Origin)
Origin = (Value *)IRB.getInt32(0);
assert(Origin->getType()->isIntegerTy());
if (shouldDisambiguateWarningLocation(IRB.getCurrentDebugLocation())) {
// Try to create additional origin with debug info of the last origin
// instruction. It may provide additional information to the user.
if (Instruction *OI = dyn_cast_or_null<Instruction>(Origin)) {
assert(MS.TrackOrigins);
auto NewDebugLoc = OI->getDebugLoc();
// Origin update with missing or the same debug location provides no
// additional value.
if (NewDebugLoc && NewDebugLoc != IRB.getCurrentDebugLocation()) {
// Insert update just before the check, so we call runtime only just
// before the report.
IRBuilder<> IRBOrigin(&*IRB.GetInsertPoint());
IRBOrigin.SetCurrentDebugLocation(NewDebugLoc);
Origin = updateOrigin(Origin, IRBOrigin);
}
}
}
if (MS.CompileKernel || MS.TrackOrigins)
IRB.CreateCall(MS.WarningFn, Origin)->setCannotMerge();
else
IRB.CreateCall(MS.WarningFn)->setCannotMerge();
// FIXME: Insert UnreachableInst if !MS.Recover?
// This may invalidate some of the following checks and needs to be done
// at the very end.
}
void materializeOneCheck(IRBuilder<> &IRB, Value *ConvertedShadow,
Value *Origin) {
const DataLayout &DL = F.getParent()->getDataLayout();
TypeSize TypeSizeInBits = DL.getTypeSizeInBits(ConvertedShadow->getType());
unsigned SizeIndex = TypeSizeToSizeIndex(TypeSizeInBits);
if (instrumentWithCalls(ConvertedShadow) &&
SizeIndex < kNumberOfAccessSizes && !MS.CompileKernel) {
FunctionCallee Fn = MS.MaybeWarningFn[SizeIndex];
Value *ConvertedShadow2 =
IRB.CreateZExt(ConvertedShadow, IRB.getIntNTy(8 * (1 << SizeIndex)));
CallBase *CB = IRB.CreateCall(
Fn, {ConvertedShadow2,
MS.TrackOrigins && Origin ? Origin : (Value *)IRB.getInt32(0)});
CB->addParamAttr(0, Attribute::ZExt);
CB->addParamAttr(1, Attribute::ZExt);
} else {
Value *Cmp = convertToBool(ConvertedShadow, IRB, "_mscmp");
Instruction *CheckTerm = SplitBlockAndInsertIfThen(
Cmp, &*IRB.GetInsertPoint(),
/* Unreachable */ !MS.Recover, MS.ColdCallWeights);
IRB.SetInsertPoint(CheckTerm);
insertWarningFn(IRB, Origin);
LLVM_DEBUG(dbgs() << " CHECK: " << *Cmp << "\n");
}
}
void materializeInstructionChecks(
ArrayRef<ShadowOriginAndInsertPoint> InstructionChecks) {
const DataLayout &DL = F.getParent()->getDataLayout();
// Disable combining in some cases. TrackOrigins checks each shadow to pick
// correct origin.
bool Combine = !MS.TrackOrigins;
Instruction *Instruction = InstructionChecks.front().OrigIns;
Value *Shadow = nullptr;
for (const auto &ShadowData : InstructionChecks) {
assert(ShadowData.OrigIns == Instruction);
IRBuilder<> IRB(Instruction);
Value *ConvertedShadow = ShadowData.Shadow;
if (auto *ConstantShadow = dyn_cast<Constant>(ConvertedShadow)) {
if (!ClCheckConstantShadow || ConstantShadow->isZeroValue()) {
// Skip, value is initialized or const shadow is ignored.
continue;
}
if (llvm::isKnownNonZero(ConvertedShadow, DL)) {
// Report as the value is definitely uninitialized.
insertWarningFn(IRB, ShadowData.Origin);
if (!MS.Recover)
return; // Always fail and stop here, not need to check the rest.
// Skip entire instruction,
continue;
}
// Fallback to runtime check, which still can be optimized out later.
}
if (!Combine) {
materializeOneCheck(IRB, ConvertedShadow, ShadowData.Origin);
continue;
}
if (!Shadow) {
Shadow = ConvertedShadow;
continue;
}
Shadow = convertToBool(Shadow, IRB, "_mscmp");
ConvertedShadow = convertToBool(ConvertedShadow, IRB, "_mscmp");
Shadow = IRB.CreateOr(Shadow, ConvertedShadow, "_msor");
}
if (Shadow) {
assert(Combine);
IRBuilder<> IRB(Instruction);
materializeOneCheck(IRB, Shadow, nullptr);
}
}
void materializeChecks() {
#ifndef NDEBUG
// For assert below.
SmallPtrSet<Instruction *, 16> Done;
#endif
for (auto I = InstrumentationList.begin();
I != InstrumentationList.end();) {
auto OrigIns = I->OrigIns;
// Checks are grouped by the original instruction. We call all
// `insertShadowCheck` for an instruction at once.
assert(Done.insert(OrigIns).second);
auto J = std::find_if(I + 1, InstrumentationList.end(),
[OrigIns](const ShadowOriginAndInsertPoint &R) {
return OrigIns != R.OrigIns;
});
// Process all checks of instruction at once.
materializeInstructionChecks(ArrayRef<ShadowOriginAndInsertPoint>(I, J));
I = J;
}
LLVM_DEBUG(dbgs() << "DONE:\n" << F);
}
// Returns the last instruction in the new prologue
void insertKmsanPrologue(IRBuilder<> &IRB) {
Value *ContextState = IRB.CreateCall(MS.MsanGetContextStateFn, {});
Constant *Zero = IRB.getInt32(0);
MS.ParamTLS = IRB.CreateGEP(MS.MsanContextStateTy, ContextState,
{Zero, IRB.getInt32(0)}, "param_shadow");
MS.RetvalTLS = IRB.CreateGEP(MS.MsanContextStateTy, ContextState,
{Zero, IRB.getInt32(1)}, "retval_shadow");
MS.VAArgTLS = IRB.CreateGEP(MS.MsanContextStateTy, ContextState,
{Zero, IRB.getInt32(2)}, "va_arg_shadow");
MS.VAArgOriginTLS = IRB.CreateGEP(MS.MsanContextStateTy, ContextState,
{Zero, IRB.getInt32(3)}, "va_arg_origin");
MS.VAArgOverflowSizeTLS =
IRB.CreateGEP(MS.MsanContextStateTy, ContextState,
{Zero, IRB.getInt32(4)}, "va_arg_overflow_size");
MS.ParamOriginTLS = IRB.CreateGEP(MS.MsanContextStateTy, ContextState,
{Zero, IRB.getInt32(5)}, "param_origin");
MS.RetvalOriginTLS =
IRB.CreateGEP(MS.MsanContextStateTy, ContextState,
{Zero, IRB.getInt32(6)}, "retval_origin");
if (MS.TargetTriple.getArch() == Triple::systemz)
MS.MsanMetadataAlloca = IRB.CreateAlloca(MS.MsanMetadata, 0u);
}
/// Add MemorySanitizer instrumentation to a function.
bool runOnFunction() {
// Iterate all BBs in depth-first order and create shadow instructions
// for all instructions (where applicable).
// For PHI nodes we create dummy shadow PHIs which will be finalized later.
for (BasicBlock *BB : depth_first(FnPrologueEnd->getParent()))
visit(*BB);
// `visit` above only collects instructions. Process them after iterating
// CFG to avoid requirement on CFG transformations.
for (Instruction *I : Instructions)
InstVisitor<MemorySanitizerVisitor>::visit(*I);
// Finalize PHI nodes.
for (PHINode *PN : ShadowPHINodes) {
PHINode *PNS = cast<PHINode>(getShadow(PN));
PHINode *PNO = MS.TrackOrigins ? cast<PHINode>(getOrigin(PN)) : nullptr;
size_t NumValues = PN->getNumIncomingValues();
for (size_t v = 0; v < NumValues; v++) {
PNS->addIncoming(getShadow(PN, v), PN->getIncomingBlock(v));
if (PNO)
PNO->addIncoming(getOrigin(PN, v), PN->getIncomingBlock(v));
}
}
VAHelper->finalizeInstrumentation();
// Poison llvm.lifetime.start intrinsics, if we haven't fallen back to
// instrumenting only allocas.
if (InstrumentLifetimeStart) {
for (auto Item : LifetimeStartList) {
instrumentAlloca(*Item.second, Item.first);
AllocaSet.remove(Item.second);
}
}
// Poison the allocas for which we didn't instrument the corresponding
// lifetime intrinsics.
for (AllocaInst *AI : AllocaSet)
instrumentAlloca(*AI);
// Insert shadow value checks.
materializeChecks();
// Delayed instrumentation of StoreInst.
// This may not add new address checks.
materializeStores();
return true;
}
/// Compute the shadow type that corresponds to a given Value.
Type *getShadowTy(Value *V) { return getShadowTy(V->getType()); }
/// Compute the shadow type that corresponds to a given Type.
Type *getShadowTy(Type *OrigTy) {
if (!OrigTy->isSized()) {
return nullptr;
}
// For integer type, shadow is the same as the original type.
// This may return weird-sized types like i1.
if (IntegerType *IT = dyn_cast<IntegerType>(OrigTy))
return IT;
const DataLayout &DL = F.getParent()->getDataLayout();
if (VectorType *VT = dyn_cast<VectorType>(OrigTy)) {
uint32_t EltSize = DL.getTypeSizeInBits(VT->getElementType());
return VectorType::get(IntegerType::get(*MS.C, EltSize),
VT->getElementCount());
}
if (ArrayType *AT = dyn_cast<ArrayType>(OrigTy)) {
return ArrayType::get(getShadowTy(AT->getElementType()),
AT->getNumElements());
}
if (StructType *ST = dyn_cast<StructType>(OrigTy)) {
SmallVector<Type *, 4> Elements;
for (unsigned i = 0, n = ST->getNumElements(); i < n; i++)
Elements.push_back(getShadowTy(ST->getElementType(i)));
StructType *Res = StructType::get(*MS.C, Elements, ST->isPacked());
LLVM_DEBUG(dbgs() << "getShadowTy: " << *ST << " ===> " << *Res << "\n");
return Res;
}
uint32_t TypeSize = DL.getTypeSizeInBits(OrigTy);
return IntegerType::get(*MS.C, TypeSize);
}
/// Extract combined shadow of struct elements as a bool
Value *collapseStructShadow(StructType *Struct, Value *Shadow,
IRBuilder<> &IRB) {
Value *FalseVal = IRB.getIntN(/* width */ 1, /* value */ 0);
Value *Aggregator = FalseVal;
for (unsigned Idx = 0; Idx < Struct->getNumElements(); Idx++) {
// Combine by ORing together each element's bool shadow
Value *ShadowItem = IRB.CreateExtractValue(Shadow, Idx);
Value *ShadowBool = convertToBool(ShadowItem, IRB);
if (Aggregator != FalseVal)
Aggregator = IRB.CreateOr(Aggregator, ShadowBool);
else
Aggregator = ShadowBool;
}
return Aggregator;
}
// Extract combined shadow of array elements
Value *collapseArrayShadow(ArrayType *Array, Value *Shadow,
IRBuilder<> &IRB) {
if (!Array->getNumElements())
return IRB.getIntN(/* width */ 1, /* value */ 0);
Value *FirstItem = IRB.CreateExtractValue(Shadow, 0);
Value *Aggregator = convertShadowToScalar(FirstItem, IRB);
for (unsigned Idx = 1; Idx < Array->getNumElements(); Idx++) {
Value *ShadowItem = IRB.CreateExtractValue(Shadow, Idx);
Value *ShadowInner = convertShadowToScalar(ShadowItem, IRB);
Aggregator = IRB.CreateOr(Aggregator, ShadowInner);
}
return Aggregator;
}
/// Convert a shadow value to it's flattened variant. The resulting
/// shadow may not necessarily have the same bit width as the input
/// value, but it will always be comparable to zero.
Value *convertShadowToScalar(Value *V, IRBuilder<> &IRB) {
if (StructType *Struct = dyn_cast<StructType>(V->getType()))
return collapseStructShadow(Struct, V, IRB);
if (ArrayType *Array = dyn_cast<ArrayType>(V->getType()))
return collapseArrayShadow(Array, V, IRB);
if (isa<VectorType>(V->getType())) {
if (isa<ScalableVectorType>(V->getType()))
return convertShadowToScalar(IRB.CreateOrReduce(V), IRB);
unsigned BitWidth =
V->getType()->getPrimitiveSizeInBits().getFixedValue();
return IRB.CreateBitCast(V, IntegerType::get(*MS.C, BitWidth));
}
return V;
}
// Convert a scalar value to an i1 by comparing with 0
Value *convertToBool(Value *V, IRBuilder<> &IRB, const Twine &name = "") {
Type *VTy = V->getType();
if (!VTy->isIntegerTy())
return convertToBool(convertShadowToScalar(V, IRB), IRB, name);
if (VTy->getIntegerBitWidth() == 1)
// Just converting a bool to a bool, so do nothing.
return V;
return IRB.CreateICmpNE(V, ConstantInt::get(VTy, 0), name);
}
Type *ptrToIntPtrType(Type *PtrTy) const {
if (VectorType *VectTy = dyn_cast<VectorType>(PtrTy)) {
return VectorType::get(ptrToIntPtrType(VectTy->getElementType()),
VectTy->getElementCount());
}
assert(PtrTy->isIntOrPtrTy());
return MS.IntptrTy;
}
Type *getPtrToShadowPtrType(Type *IntPtrTy, Type *ShadowTy) const {
if (VectorType *VectTy = dyn_cast<VectorType>(IntPtrTy)) {
return VectorType::get(
getPtrToShadowPtrType(VectTy->getElementType(), ShadowTy),
VectTy->getElementCount());
}
assert(IntPtrTy == MS.IntptrTy);
return PointerType::get(*MS.C, 0);
}
Constant *constToIntPtr(Type *IntPtrTy, uint64_t C) const {
if (VectorType *VectTy = dyn_cast<VectorType>(IntPtrTy)) {
return ConstantVector::getSplat(
VectTy->getElementCount(), constToIntPtr(VectTy->getElementType(), C));
}
assert(IntPtrTy == MS.IntptrTy);
return ConstantInt::get(MS.IntptrTy, C);
}
/// Compute the integer shadow offset that corresponds to a given
/// application address.
///
/// Offset = (Addr & ~AndMask) ^ XorMask
/// Addr can be a ptr or <N x ptr>. In both cases ShadowTy the shadow type of
/// a single pointee.
/// Returns <shadow_ptr, origin_ptr> or <<N x shadow_ptr>, <N x origin_ptr>>.
Value *getShadowPtrOffset(Value *Addr, IRBuilder<> &IRB) {
Type *IntptrTy = ptrToIntPtrType(Addr->getType());
Value *OffsetLong = IRB.CreatePointerCast(Addr, IntptrTy);
if (uint64_t AndMask = MS.MapParams->AndMask)
OffsetLong = IRB.CreateAnd(OffsetLong, constToIntPtr(IntptrTy, ~AndMask));
if (uint64_t XorMask = MS.MapParams->XorMask)
OffsetLong = IRB.CreateXor(OffsetLong, constToIntPtr(IntptrTy, XorMask));
return OffsetLong;
}
/// Compute the shadow and origin addresses corresponding to a given
/// application address.
///
/// Shadow = ShadowBase + Offset
/// Origin = (OriginBase + Offset) & ~3ULL
/// Addr can be a ptr or <N x ptr>. In both cases ShadowTy the shadow type of
/// a single pointee.
/// Returns <shadow_ptr, origin_ptr> or <<N x shadow_ptr>, <N x origin_ptr>>.
std::pair<Value *, Value *>
getShadowOriginPtrUserspace(Value *Addr, IRBuilder<> &IRB, Type *ShadowTy,
MaybeAlign Alignment) {
VectorType *VectTy = dyn_cast<VectorType>(Addr->getType());
if (!VectTy) {
assert(Addr->getType()->isPointerTy());
} else {
assert(VectTy->getElementType()->isPointerTy());
}
Type *IntptrTy = ptrToIntPtrType(Addr->getType());
Value *ShadowOffset = getShadowPtrOffset(Addr, IRB);
Value *ShadowLong = ShadowOffset;
if (uint64_t ShadowBase = MS.MapParams->ShadowBase) {
ShadowLong =
IRB.CreateAdd(ShadowLong, constToIntPtr(IntptrTy, ShadowBase));
}
Value *ShadowPtr = IRB.CreateIntToPtr(
ShadowLong, getPtrToShadowPtrType(IntptrTy, ShadowTy));
Value *OriginPtr = nullptr;
if (MS.TrackOrigins) {
Value *OriginLong = ShadowOffset;
uint64_t OriginBase = MS.MapParams->OriginBase;
if (OriginBase != 0)
OriginLong =
IRB.CreateAdd(OriginLong, constToIntPtr(IntptrTy, OriginBase));
if (!Alignment || *Alignment < kMinOriginAlignment) {
uint64_t Mask = kMinOriginAlignment.value() - 1;
OriginLong = IRB.CreateAnd(OriginLong, constToIntPtr(IntptrTy, ~Mask));
}
OriginPtr = IRB.CreateIntToPtr(
OriginLong, getPtrToShadowPtrType(IntptrTy, MS.OriginTy));
}
return std::make_pair(ShadowPtr, OriginPtr);
}
template <typename... ArgsTy>
Value *createMetadataCall(IRBuilder<> &IRB, FunctionCallee Callee,
ArgsTy... Args) {
if (MS.TargetTriple.getArch() == Triple::systemz) {
IRB.CreateCall(Callee,
{MS.MsanMetadataAlloca, std::forward<ArgsTy>(Args)...});
return IRB.CreateLoad(MS.MsanMetadata, MS.MsanMetadataAlloca);
}
return IRB.CreateCall(Callee, {std::forward<ArgsTy>(Args)...});
}
std::pair<Value *, Value *> getShadowOriginPtrKernelNoVec(Value *Addr,
IRBuilder<> &IRB,
Type *ShadowTy,
bool isStore) {
Value *ShadowOriginPtrs;
const DataLayout &DL = F.getParent()->getDataLayout();
TypeSize Size = DL.getTypeStoreSize(ShadowTy);
FunctionCallee Getter = MS.getKmsanShadowOriginAccessFn(isStore, Size);
Value *AddrCast =
IRB.CreatePointerCast(Addr, PointerType::get(IRB.getInt8Ty(), 0));
if (Getter) {
ShadowOriginPtrs = createMetadataCall(IRB, Getter, AddrCast);
} else {
Value *SizeVal = ConstantInt::get(MS.IntptrTy, Size);
ShadowOriginPtrs = createMetadataCall(
IRB,
isStore ? MS.MsanMetadataPtrForStoreN : MS.MsanMetadataPtrForLoadN,
AddrCast, SizeVal);
}
Value *ShadowPtr = IRB.CreateExtractValue(ShadowOriginPtrs, 0);
ShadowPtr = IRB.CreatePointerCast(ShadowPtr, PointerType::get(ShadowTy, 0));
Value *OriginPtr = IRB.CreateExtractValue(ShadowOriginPtrs, 1);
return std::make_pair(ShadowPtr, OriginPtr);
}
/// Addr can be a ptr or <N x ptr>. In both cases ShadowTy the shadow type of
/// a single pointee.
/// Returns <shadow_ptr, origin_ptr> or <<N x shadow_ptr>, <N x origin_ptr>>.
std::pair<Value *, Value *> getShadowOriginPtrKernel(Value *Addr,
IRBuilder<> &IRB,
Type *ShadowTy,
bool isStore) {
VectorType *VectTy = dyn_cast<VectorType>(Addr->getType());
if (!VectTy) {
assert(Addr->getType()->isPointerTy());
return getShadowOriginPtrKernelNoVec(Addr, IRB, ShadowTy, isStore);
}
// TODO: Support callbacs with vectors of addresses.
unsigned NumElements = cast<FixedVectorType>(VectTy)->getNumElements();
Value *ShadowPtrs = ConstantInt::getNullValue(
FixedVectorType::get(IRB.getPtrTy(), NumElements));
Value *OriginPtrs = nullptr;
if (MS.TrackOrigins)
OriginPtrs = ConstantInt::getNullValue(
FixedVectorType::get(IRB.getPtrTy(), NumElements));
for (unsigned i = 0; i < NumElements; ++i) {
Value *OneAddr =
IRB.CreateExtractElement(Addr, ConstantInt::get(IRB.getInt32Ty(), i));
auto [ShadowPtr, OriginPtr] =
getShadowOriginPtrKernelNoVec(OneAddr, IRB, ShadowTy, isStore);
ShadowPtrs = IRB.CreateInsertElement(
ShadowPtrs, ShadowPtr, ConstantInt::get(IRB.getInt32Ty(), i));
if (MS.TrackOrigins)
OriginPtrs = IRB.CreateInsertElement(
OriginPtrs, OriginPtr, ConstantInt::get(IRB.getInt32Ty(), i));
}
return {ShadowPtrs, OriginPtrs};
}
std::pair<Value *, Value *> getShadowOriginPtr(Value *Addr, IRBuilder<> &IRB,
Type *ShadowTy,
MaybeAlign Alignment,
bool isStore) {
if (MS.CompileKernel)
return getShadowOriginPtrKernel(Addr, IRB, ShadowTy, isStore);
return getShadowOriginPtrUserspace(Addr, IRB, ShadowTy, Alignment);
}
/// Compute the shadow address for a given function argument.
///
/// Shadow = ParamTLS+ArgOffset.
Value *getShadowPtrForArgument(IRBuilder<> &IRB, int ArgOffset) {
Value *Base = IRB.CreatePointerCast(MS.ParamTLS, MS.IntptrTy);
if (ArgOffset)
Base = IRB.CreateAdd(Base, ConstantInt::get(MS.IntptrTy, ArgOffset));
return IRB.CreateIntToPtr(Base, IRB.getPtrTy(0), "_msarg");
}
/// Compute the origin address for a given function argument.
Value *getOriginPtrForArgument(IRBuilder<> &IRB, int ArgOffset) {
if (!MS.TrackOrigins)
return nullptr;
Value *Base = IRB.CreatePointerCast(MS.ParamOriginTLS, MS.IntptrTy);
if (ArgOffset)
Base = IRB.CreateAdd(Base, ConstantInt::get(MS.IntptrTy, ArgOffset));
return IRB.CreateIntToPtr(Base, IRB.getPtrTy(0), "_msarg_o");
}
/// Compute the shadow address for a retval.
Value *getShadowPtrForRetval(IRBuilder<> &IRB) {
return IRB.CreatePointerCast(MS.RetvalTLS, IRB.getPtrTy(0), "_msret");
}
/// Compute the origin address for a retval.
Value *getOriginPtrForRetval() {
// We keep a single origin for the entire retval. Might be too optimistic.
return MS.RetvalOriginTLS;
}
/// Set SV to be the shadow value for V.
void setShadow(Value *V, Value *SV) {
assert(!ShadowMap.count(V) && "Values may only have one shadow");
ShadowMap[V] = PropagateShadow ? SV : getCleanShadow(V);
}
/// Set Origin to be the origin value for V.
void setOrigin(Value *V, Value *Origin) {
if (!MS.TrackOrigins)
return;
assert(!OriginMap.count(V) && "Values may only have one origin");
LLVM_DEBUG(dbgs() << "ORIGIN: " << *V << " ==> " << *Origin << "\n");
OriginMap[V] = Origin;
}
Constant *getCleanShadow(Type *OrigTy) {
Type *ShadowTy = getShadowTy(OrigTy);
if (!ShadowTy)
return nullptr;
return Constant::getNullValue(ShadowTy);
}
/// Create a clean shadow value for a given value.
///
/// Clean shadow (all zeroes) means all bits of the value are defined
/// (initialized).
Constant *getCleanShadow(Value *V) { return getCleanShadow(V->getType()); }
/// Create a dirty shadow of a given shadow type.
Constant *getPoisonedShadow(Type *ShadowTy) {
assert(ShadowTy);
if (isa<IntegerType>(ShadowTy) || isa<VectorType>(ShadowTy))
return Constant::getAllOnesValue(ShadowTy);
if (ArrayType *AT = dyn_cast<ArrayType>(ShadowTy)) {
SmallVector<Constant *, 4> Vals(AT->getNumElements(),
getPoisonedShadow(AT->getElementType()));
return ConstantArray::get(AT, Vals);
}
if (StructType *ST = dyn_cast<StructType>(ShadowTy)) {
SmallVector<Constant *, 4> Vals;
for (unsigned i = 0, n = ST->getNumElements(); i < n; i++)
Vals.push_back(getPoisonedShadow(ST->getElementType(i)));
return ConstantStruct::get(ST, Vals);
}
llvm_unreachable("Unexpected shadow type");
}
/// Create a dirty shadow for a given value.
Constant *getPoisonedShadow(Value *V) {
Type *ShadowTy = getShadowTy(V);
if (!ShadowTy)
return nullptr;
return getPoisonedShadow(ShadowTy);
}
/// Create a clean (zero) origin.
Value *getCleanOrigin() { return Constant::getNullValue(MS.OriginTy); }
/// Get the shadow value for a given Value.
///
/// This function either returns the value set earlier with setShadow,
/// or extracts if from ParamTLS (for function arguments).
Value *getShadow(Value *V) {
if (Instruction *I = dyn_cast<Instruction>(V)) {
if (!PropagateShadow || I->getMetadata(LLVMContext::MD_nosanitize))
return getCleanShadow(V);
// For instructions the shadow is already stored in the map.
Value *Shadow = ShadowMap[V];
if (!Shadow) {
LLVM_DEBUG(dbgs() << "No shadow: " << *V << "\n" << *(I->getParent()));
(void)I;
assert(Shadow && "No shadow for a value");
}
return Shadow;
}
if (UndefValue *U = dyn_cast<UndefValue>(V)) {
Value *AllOnes = (PropagateShadow && PoisonUndef) ? getPoisonedShadow(V)
: getCleanShadow(V);
LLVM_DEBUG(dbgs() << "Undef: " << *U << " ==> " << *AllOnes << "\n");
(void)U;
return AllOnes;
}
if (Argument *A = dyn_cast<Argument>(V)) {
// For arguments we compute the shadow on demand and store it in the map.
Value *&ShadowPtr = ShadowMap[V];
if (ShadowPtr)
return ShadowPtr;
Function *F = A->getParent();
IRBuilder<> EntryIRB(FnPrologueEnd);
unsigned ArgOffset = 0;
const DataLayout &DL = F->getParent()->getDataLayout();
for (auto &FArg : F->args()) {
if (!FArg.getType()->isSized() || FArg.getType()->isScalableTy()) {
LLVM_DEBUG(dbgs() << (FArg.getType()->isScalableTy()
? "vscale not fully supported\n"
: "Arg is not sized\n"));
if (A == &FArg) {
ShadowPtr = getCleanShadow(V);
setOrigin(A, getCleanOrigin());
break;
}
continue;
}
unsigned Size = FArg.hasByValAttr()
? DL.getTypeAllocSize(FArg.getParamByValType())
: DL.getTypeAllocSize(FArg.getType());
if (A == &FArg) {
bool Overflow = ArgOffset + Size > kParamTLSSize;
if (FArg.hasByValAttr()) {
// ByVal pointer itself has clean shadow. We copy the actual
// argument shadow to the underlying memory.
// Figure out maximal valid memcpy alignment.
const Align ArgAlign = DL.getValueOrABITypeAlignment(
FArg.getParamAlign(), FArg.getParamByValType());
Value *CpShadowPtr, *CpOriginPtr;
std::tie(CpShadowPtr, CpOriginPtr) =
getShadowOriginPtr(V, EntryIRB, EntryIRB.getInt8Ty(), ArgAlign,
/*isStore*/ true);
if (!PropagateShadow || Overflow) {
// ParamTLS overflow.
EntryIRB.CreateMemSet(
CpShadowPtr, Constant::getNullValue(EntryIRB.getInt8Ty()),
Size, ArgAlign);
} else {
Value *Base = getShadowPtrForArgument(EntryIRB, ArgOffset);
const Align CopyAlign = std::min(ArgAlign, kShadowTLSAlignment);
Value *Cpy = EntryIRB.CreateMemCpy(CpShadowPtr, CopyAlign, Base,
CopyAlign, Size);
LLVM_DEBUG(dbgs() << " ByValCpy: " << *Cpy << "\n");
(void)Cpy;
if (MS.TrackOrigins) {
Value *OriginPtr =
getOriginPtrForArgument(EntryIRB, ArgOffset);
// FIXME: OriginSize should be:
// alignTo(V % kMinOriginAlignment + Size, kMinOriginAlignment)
unsigned OriginSize = alignTo(Size, kMinOriginAlignment);
EntryIRB.CreateMemCpy(
CpOriginPtr,
/* by getShadowOriginPtr */ kMinOriginAlignment, OriginPtr,
/* by origin_tls[ArgOffset] */ kMinOriginAlignment,
OriginSize);
}
}
}
if (!PropagateShadow || Overflow || FArg.hasByValAttr() ||
(MS.EagerChecks && FArg.hasAttribute(Attribute::NoUndef))) {
ShadowPtr = getCleanShadow(V);
setOrigin(A, getCleanOrigin());
} else {
// Shadow over TLS
Value *Base = getShadowPtrForArgument(EntryIRB, ArgOffset);
ShadowPtr = EntryIRB.CreateAlignedLoad(getShadowTy(&FArg), Base,
kShadowTLSAlignment);
if (MS.TrackOrigins) {
Value *OriginPtr =
getOriginPtrForArgument(EntryIRB, ArgOffset);
setOrigin(A, EntryIRB.CreateLoad(MS.OriginTy, OriginPtr));
}
}
LLVM_DEBUG(dbgs()
<< " ARG: " << FArg << " ==> " << *ShadowPtr << "\n");
break;
}
ArgOffset += alignTo(Size, kShadowTLSAlignment);
}
assert(ShadowPtr && "Could not find shadow for an argument");
return ShadowPtr;
}
// For everything else the shadow is zero.
return getCleanShadow(V);
}
/// Get the shadow for i-th argument of the instruction I.
Value *getShadow(Instruction *I, int i) {
return getShadow(I->getOperand(i));
}
/// Get the origin for a value.
Value *getOrigin(Value *V) {
if (!MS.TrackOrigins)
return nullptr;
if (!PropagateShadow || isa<Constant>(V) || isa<InlineAsm>(V))
return getCleanOrigin();
assert((isa<Instruction>(V) || isa<Argument>(V)) &&
"Unexpected value type in getOrigin()");
if (Instruction *I = dyn_cast<Instruction>(V)) {
if (I->getMetadata(LLVMContext::MD_nosanitize))
return getCleanOrigin();
}
Value *Origin = OriginMap[V];
assert(Origin && "Missing origin");
return Origin;
}
/// Get the origin for i-th argument of the instruction I.
Value *getOrigin(Instruction *I, int i) {
return getOrigin(I->getOperand(i));
}
/// Remember the place where a shadow check should be inserted.
///
/// This location will be later instrumented with a check that will print a
/// UMR warning in runtime if the shadow value is not 0.
void insertShadowCheck(Value *Shadow, Value *Origin, Instruction *OrigIns) {
assert(Shadow);
if (!InsertChecks)
return;
if (!DebugCounter::shouldExecute(DebugInsertCheck)) {
LLVM_DEBUG(dbgs() << "Skipping check of " << *Shadow << " before "
<< *OrigIns << "\n");
return;
}
#ifndef NDEBUG
Type *ShadowTy = Shadow->getType();
assert((isa<IntegerType>(ShadowTy) || isa<VectorType>(ShadowTy) ||
isa<StructType>(ShadowTy) || isa<ArrayType>(ShadowTy)) &&
"Can only insert checks for integer, vector, and aggregate shadow "
"types");
#endif
InstrumentationList.push_back(
ShadowOriginAndInsertPoint(Shadow, Origin, OrigIns));
}
/// Remember the place where a shadow check should be inserted.
///
/// This location will be later instrumented with a check that will print a
/// UMR warning in runtime if the value is not fully defined.
void insertShadowCheck(Value *Val, Instruction *OrigIns) {
assert(Val);
Value *Shadow, *Origin;
if (ClCheckConstantShadow) {
Shadow = getShadow(Val);
if (!Shadow)
return;
Origin = getOrigin(Val);
} else {
Shadow = dyn_cast_or_null<Instruction>(getShadow(Val));
if (!Shadow)
return;
Origin = dyn_cast_or_null<Instruction>(getOrigin(Val));
}
insertShadowCheck(Shadow, Origin, OrigIns);
}
AtomicOrdering addReleaseOrdering(AtomicOrdering a) {
switch (a) {
case AtomicOrdering::NotAtomic:
return AtomicOrdering::NotAtomic;
case AtomicOrdering::Unordered:
case AtomicOrdering::Monotonic:
case AtomicOrdering::Release:
return AtomicOrdering::Release;
case AtomicOrdering::Acquire:
case AtomicOrdering::AcquireRelease:
return AtomicOrdering::AcquireRelease;
case AtomicOrdering::SequentiallyConsistent:
return AtomicOrdering::SequentiallyConsistent;
}
llvm_unreachable("Unknown ordering");
}
Value *makeAddReleaseOrderingTable(IRBuilder<> &IRB) {
constexpr int NumOrderings = (int)AtomicOrderingCABI::seq_cst + 1;
uint32_t OrderingTable[NumOrderings] = {};
OrderingTable[(int)AtomicOrderingCABI::relaxed] =
OrderingTable[(int)AtomicOrderingCABI::release] =
(int)AtomicOrderingCABI::release;
OrderingTable[(int)AtomicOrderingCABI::consume] =
OrderingTable[(int)AtomicOrderingCABI::acquire] =
OrderingTable[(int)AtomicOrderingCABI::acq_rel] =
(int)AtomicOrderingCABI::acq_rel;
OrderingTable[(int)AtomicOrderingCABI::seq_cst] =
(int)AtomicOrderingCABI::seq_cst;
return ConstantDataVector::get(IRB.getContext(), OrderingTable);
}
AtomicOrdering addAcquireOrdering(AtomicOrdering a) {
switch (a) {
case AtomicOrdering::NotAtomic:
return AtomicOrdering::NotAtomic;
case AtomicOrdering::Unordered:
case AtomicOrdering::Monotonic:
case AtomicOrdering::Acquire:
return AtomicOrdering::Acquire;
case AtomicOrdering::Release:
case AtomicOrdering::AcquireRelease:
return AtomicOrdering::AcquireRelease;
case AtomicOrdering::SequentiallyConsistent:
return AtomicOrdering::SequentiallyConsistent;
}
llvm_unreachable("Unknown ordering");
}
Value *makeAddAcquireOrderingTable(IRBuilder<> &IRB) {
constexpr int NumOrderings = (int)AtomicOrderingCABI::seq_cst + 1;
uint32_t OrderingTable[NumOrderings] = {};
OrderingTable[(int)AtomicOrderingCABI::relaxed] =
OrderingTable[(int)AtomicOrderingCABI::acquire] =
OrderingTable[(int)AtomicOrderingCABI::consume] =
(int)AtomicOrderingCABI::acquire;
OrderingTable[(int)AtomicOrderingCABI::release] =
OrderingTable[(int)AtomicOrderingCABI::acq_rel] =
(int)AtomicOrderingCABI::acq_rel;
OrderingTable[(int)AtomicOrderingCABI::seq_cst] =
(int)AtomicOrderingCABI::seq_cst;
return ConstantDataVector::get(IRB.getContext(), OrderingTable);
}
// ------------------- Visitors.
using InstVisitor<MemorySanitizerVisitor>::visit;
void visit(Instruction &I) {
if (I.getMetadata(LLVMContext::MD_nosanitize))
return;
// Don't want to visit if we're in the prologue
if (isInPrologue(I))
return;
if (!DebugCounter::shouldExecute(DebugInstrumentInstruction)) {
LLVM_DEBUG(dbgs() << "Skipping instruction: " << I << "\n");
// We still need to set the shadow and origin to clean values.
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
return;
}
Instructions.push_back(&I);
}
/// Instrument LoadInst
///
/// Loads the corresponding shadow and (optionally) origin.
/// Optionally, checks that the load address is fully defined.
void visitLoadInst(LoadInst &I) {
assert(I.getType()->isSized() && "Load type must have size");
assert(!I.getMetadata(LLVMContext::MD_nosanitize));
NextNodeIRBuilder IRB(&I);
Type *ShadowTy = getShadowTy(&I);
Value *Addr = I.getPointerOperand();
Value *ShadowPtr = nullptr, *OriginPtr = nullptr;
const Align Alignment = I.getAlign();
if (PropagateShadow) {
std::tie(ShadowPtr, OriginPtr) =
getShadowOriginPtr(Addr, IRB, ShadowTy, Alignment, /*isStore*/ false);
setShadow(&I,
IRB.CreateAlignedLoad(ShadowTy, ShadowPtr, Alignment, "_msld"));
} else {
setShadow(&I, getCleanShadow(&I));
}
if (ClCheckAccessAddress)
insertShadowCheck(I.getPointerOperand(), &I);
if (I.isAtomic())
I.setOrdering(addAcquireOrdering(I.getOrdering()));
if (MS.TrackOrigins) {
if (PropagateShadow) {
const Align OriginAlignment = std::max(kMinOriginAlignment, Alignment);
setOrigin(
&I, IRB.CreateAlignedLoad(MS.OriginTy, OriginPtr, OriginAlignment));
} else {
setOrigin(&I, getCleanOrigin());
}
}
}
/// Instrument StoreInst
///
/// Stores the corresponding shadow and (optionally) origin.
/// Optionally, checks that the store address is fully defined.
void visitStoreInst(StoreInst &I) {
StoreList.push_back(&I);
if (ClCheckAccessAddress)
insertShadowCheck(I.getPointerOperand(), &I);
}
void handleCASOrRMW(Instruction &I) {
assert(isa<AtomicRMWInst>(I) || isa<AtomicCmpXchgInst>(I));
IRBuilder<> IRB(&I);
Value *Addr = I.getOperand(0);
Value *Val = I.getOperand(1);
Value *ShadowPtr = getShadowOriginPtr(Addr, IRB, getShadowTy(Val), Align(1),
/*isStore*/ true)
.first;
if (ClCheckAccessAddress)
insertShadowCheck(Addr, &I);
// Only test the conditional argument of cmpxchg instruction.
// The other argument can potentially be uninitialized, but we can not
// detect this situation reliably without possible false positives.
if (isa<AtomicCmpXchgInst>(I))
insertShadowCheck(Val, &I);
IRB.CreateStore(getCleanShadow(Val), ShadowPtr);
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
}
void visitAtomicRMWInst(AtomicRMWInst &I) {
handleCASOrRMW(I);
I.setOrdering(addReleaseOrdering(I.getOrdering()));
}
void visitAtomicCmpXchgInst(AtomicCmpXchgInst &I) {
handleCASOrRMW(I);
I.setSuccessOrdering(addReleaseOrdering(I.getSuccessOrdering()));
}
// Vector manipulation.
void visitExtractElementInst(ExtractElementInst &I) {
insertShadowCheck(I.getOperand(1), &I);
IRBuilder<> IRB(&I);
setShadow(&I, IRB.CreateExtractElement(getShadow(&I, 0), I.getOperand(1),
"_msprop"));
setOrigin(&I, getOrigin(&I, 0));
}
void visitInsertElementInst(InsertElementInst &I) {
insertShadowCheck(I.getOperand(2), &I);
IRBuilder<> IRB(&I);
auto *Shadow0 = getShadow(&I, 0);
auto *Shadow1 = getShadow(&I, 1);
setShadow(&I, IRB.CreateInsertElement(Shadow0, Shadow1, I.getOperand(2),
"_msprop"));
setOriginForNaryOp(I);
}
void visitShuffleVectorInst(ShuffleVectorInst &I) {
IRBuilder<> IRB(&I);
auto *Shadow0 = getShadow(&I, 0);
auto *Shadow1 = getShadow(&I, 1);
setShadow(&I, IRB.CreateShuffleVector(Shadow0, Shadow1, I.getShuffleMask(),
"_msprop"));
setOriginForNaryOp(I);
}
// Casts.
void visitSExtInst(SExtInst &I) {
IRBuilder<> IRB(&I);
setShadow(&I, IRB.CreateSExt(getShadow(&I, 0), I.getType(), "_msprop"));
setOrigin(&I, getOrigin(&I, 0));
}
void visitZExtInst(ZExtInst &I) {
IRBuilder<> IRB(&I);
setShadow(&I, IRB.CreateZExt(getShadow(&I, 0), I.getType(), "_msprop"));
setOrigin(&I, getOrigin(&I, 0));
}
void visitTruncInst(TruncInst &I) {
IRBuilder<> IRB(&I);
setShadow(&I, IRB.CreateTrunc(getShadow(&I, 0), I.getType(), "_msprop"));
setOrigin(&I, getOrigin(&I, 0));
}
void visitBitCastInst(BitCastInst &I) {
// Special case: if this is the bitcast (there is exactly 1 allowed) between
// a musttail call and a ret, don't instrument. New instructions are not
// allowed after a musttail call.
if (auto *CI = dyn_cast<CallInst>(I.getOperand(0)))
if (CI->isMustTailCall())
return;
IRBuilder<> IRB(&I);
setShadow(&I, IRB.CreateBitCast(getShadow(&I, 0), getShadowTy(&I)));
setOrigin(&I, getOrigin(&I, 0));
}
void visitPtrToIntInst(PtrToIntInst &I) {
IRBuilder<> IRB(&I);
setShadow(&I, IRB.CreateIntCast(getShadow(&I, 0), getShadowTy(&I), false,
"_msprop_ptrtoint"));
setOrigin(&I, getOrigin(&I, 0));
}
void visitIntToPtrInst(IntToPtrInst &I) {
IRBuilder<> IRB(&I);
setShadow(&I, IRB.CreateIntCast(getShadow(&I, 0), getShadowTy(&I), false,
"_msprop_inttoptr"));
setOrigin(&I, getOrigin(&I, 0));
}
void visitFPToSIInst(CastInst &I) { handleShadowOr(I); }
void visitFPToUIInst(CastInst &I) { handleShadowOr(I); }
void visitSIToFPInst(CastInst &I) { handleShadowOr(I); }
void visitUIToFPInst(CastInst &I) { handleShadowOr(I); }
void visitFPExtInst(CastInst &I) { handleShadowOr(I); }
void visitFPTruncInst(CastInst &I) { handleShadowOr(I); }
/// Propagate shadow for bitwise AND.
///
/// This code is exact, i.e. if, for example, a bit in the left argument
/// is defined and 0, then neither the value not definedness of the
/// corresponding bit in B don't affect the resulting shadow.
void visitAnd(BinaryOperator &I) {
IRBuilder<> IRB(&I);
// "And" of 0 and a poisoned value results in unpoisoned value.
// 1&1 => 1; 0&1 => 0; p&1 => p;
// 1&0 => 0; 0&0 => 0; p&0 => 0;
// 1&p => p; 0&p => 0; p&p => p;
// S = (S1 & S2) | (V1 & S2) | (S1 & V2)
Value *S1 = getShadow(&I, 0);
Value *S2 = getShadow(&I, 1);
Value *V1 = I.getOperand(0);
Value *V2 = I.getOperand(1);
if (V1->getType() != S1->getType()) {
V1 = IRB.CreateIntCast(V1, S1->getType(), false);
V2 = IRB.CreateIntCast(V2, S2->getType(), false);
}
Value *S1S2 = IRB.CreateAnd(S1, S2);
Value *V1S2 = IRB.CreateAnd(V1, S2);
Value *S1V2 = IRB.CreateAnd(S1, V2);
setShadow(&I, IRB.CreateOr({S1S2, V1S2, S1V2}));
setOriginForNaryOp(I);
}
void visitOr(BinaryOperator &I) {
IRBuilder<> IRB(&I);
// "Or" of 1 and a poisoned value results in unpoisoned value.
// 1|1 => 1; 0|1 => 1; p|1 => 1;
// 1|0 => 1; 0|0 => 0; p|0 => p;
// 1|p => 1; 0|p => p; p|p => p;
// S = (S1 & S2) | (~V1 & S2) | (S1 & ~V2)
Value *S1 = getShadow(&I, 0);
Value *S2 = getShadow(&I, 1);
Value *V1 = IRB.CreateNot(I.getOperand(0));
Value *V2 = IRB.CreateNot(I.getOperand(1));
if (V1->getType() != S1->getType()) {
V1 = IRB.CreateIntCast(V1, S1->getType(), false);
V2 = IRB.CreateIntCast(V2, S2->getType(), false);
}
Value *S1S2 = IRB.CreateAnd(S1, S2);
Value *V1S2 = IRB.CreateAnd(V1, S2);
Value *S1V2 = IRB.CreateAnd(S1, V2);
setShadow(&I, IRB.CreateOr({S1S2, V1S2, S1V2}));
setOriginForNaryOp(I);
}
/// Default propagation of shadow and/or origin.
///
/// This class implements the general case of shadow propagation, used in all
/// cases where we don't know and/or don't care about what the operation
/// actually does. It converts all input shadow values to a common type
/// (extending or truncating as necessary), and bitwise OR's them.
///
/// This is much cheaper than inserting checks (i.e. requiring inputs to be
/// fully initialized), and less prone to false positives.
///
/// This class also implements the general case of origin propagation. For a
/// Nary operation, result origin is set to the origin of an argument that is
/// not entirely initialized. If there is more than one such arguments, the
/// rightmost of them is picked. It does not matter which one is picked if all
/// arguments are initialized.
template <bool CombineShadow> class Combiner {
Value *Shadow = nullptr;
Value *Origin = nullptr;
IRBuilder<> &IRB;
MemorySanitizerVisitor *MSV;
public:
Combiner(MemorySanitizerVisitor *MSV, IRBuilder<> &IRB)
: IRB(IRB), MSV(MSV) {}
/// Add a pair of shadow and origin values to the mix.
Combiner &Add(Value *OpShadow, Value *OpOrigin) {
if (CombineShadow) {
assert(OpShadow);
if (!Shadow)
Shadow = OpShadow;
else {
OpShadow = MSV->CreateShadowCast(IRB, OpShadow, Shadow->getType());
Shadow = IRB.CreateOr(Shadow, OpShadow, "_msprop");
}
}
if (MSV->MS.TrackOrigins) {
assert(OpOrigin);
if (!Origin) {
Origin = OpOrigin;
} else {
Constant *ConstOrigin = dyn_cast<Constant>(OpOrigin);
// No point in adding something that might result in 0 origin value.
if (!ConstOrigin || !ConstOrigin->isNullValue()) {
Value *Cond = MSV->convertToBool(OpShadow, IRB);
Origin = IRB.CreateSelect(Cond, OpOrigin, Origin);
}
}
}
return *this;
}
/// Add an application value to the mix.
Combiner &Add(Value *V) {
Value *OpShadow = MSV->getShadow(V);
Value *OpOrigin = MSV->MS.TrackOrigins ? MSV->getOrigin(V) : nullptr;
return Add(OpShadow, OpOrigin);
}
/// Set the current combined values as the given instruction's shadow
/// and origin.
void Done(Instruction *I) {
if (CombineShadow) {
assert(Shadow);
Shadow = MSV->CreateShadowCast(IRB, Shadow, MSV->getShadowTy(I));
MSV->setShadow(I, Shadow);
}
if (MSV->MS.TrackOrigins) {
assert(Origin);
MSV->setOrigin(I, Origin);
}
}
};
using ShadowAndOriginCombiner = Combiner<true>;
using OriginCombiner = Combiner<false>;
/// Propagate origin for arbitrary operation.
void setOriginForNaryOp(Instruction &I) {
if (!MS.TrackOrigins)
return;
IRBuilder<> IRB(&I);
OriginCombiner OC(this, IRB);
for (Use &Op : I.operands())
OC.Add(Op.get());
OC.Done(&I);
}
size_t VectorOrPrimitiveTypeSizeInBits(Type *Ty) {
assert(!(Ty->isVectorTy() && Ty->getScalarType()->isPointerTy()) &&
"Vector of pointers is not a valid shadow type");
return Ty->isVectorTy() ? cast<FixedVectorType>(Ty)->getNumElements() *
Ty->getScalarSizeInBits()
: Ty->getPrimitiveSizeInBits();
}
/// Cast between two shadow types, extending or truncating as
/// necessary.
Value *CreateShadowCast(IRBuilder<> &IRB, Value *V, Type *dstTy,
bool Signed = false) {
Type *srcTy = V->getType();
if (srcTy == dstTy)
return V;
size_t srcSizeInBits = VectorOrPrimitiveTypeSizeInBits(srcTy);
size_t dstSizeInBits = VectorOrPrimitiveTypeSizeInBits(dstTy);
if (srcSizeInBits > 1 && dstSizeInBits == 1)
return IRB.CreateICmpNE(V, getCleanShadow(V));
if (dstTy->isIntegerTy() && srcTy->isIntegerTy())
return IRB.CreateIntCast(V, dstTy, Signed);
if (dstTy->isVectorTy() && srcTy->isVectorTy() &&
cast<VectorType>(dstTy)->getElementCount() ==
cast<VectorType>(srcTy)->getElementCount())
return IRB.CreateIntCast(V, dstTy, Signed);
Value *V1 = IRB.CreateBitCast(V, Type::getIntNTy(*MS.C, srcSizeInBits));
Value *V2 =
IRB.CreateIntCast(V1, Type::getIntNTy(*MS.C, dstSizeInBits), Signed);
return IRB.CreateBitCast(V2, dstTy);
// TODO: handle struct types.
}
/// Cast an application value to the type of its own shadow.
Value *CreateAppToShadowCast(IRBuilder<> &IRB, Value *V) {
Type *ShadowTy = getShadowTy(V);
if (V->getType() == ShadowTy)
return V;
if (V->getType()->isPtrOrPtrVectorTy())
return IRB.CreatePtrToInt(V, ShadowTy);
else
return IRB.CreateBitCast(V, ShadowTy);
}
/// Propagate shadow for arbitrary operation.
void handleShadowOr(Instruction &I) {
IRBuilder<> IRB(&I);
ShadowAndOriginCombiner SC(this, IRB);
for (Use &Op : I.operands())
SC.Add(Op.get());
SC.Done(&I);
}
void visitFNeg(UnaryOperator &I) { handleShadowOr(I); }
// Handle multiplication by constant.
//
// Handle a special case of multiplication by constant that may have one or
// more zeros in the lower bits. This makes corresponding number of lower bits
// of the result zero as well. We model it by shifting the other operand
// shadow left by the required number of bits. Effectively, we transform
// (X * (A * 2**B)) to ((X << B) * A) and instrument (X << B) as (Sx << B).
// We use multiplication by 2**N instead of shift to cover the case of
// multiplication by 0, which may occur in some elements of a vector operand.
void handleMulByConstant(BinaryOperator &I, Constant *ConstArg,
Value *OtherArg) {
Constant *ShadowMul;
Type *Ty = ConstArg->getType();
if (auto *VTy = dyn_cast<VectorType>(Ty)) {
unsigned NumElements = cast<FixedVectorType>(VTy)->getNumElements();
Type *EltTy = VTy->getElementType();
SmallVector<Constant *, 16> Elements;
for (unsigned Idx = 0; Idx < NumElements; ++Idx) {
if (ConstantInt *Elt =
dyn_cast<ConstantInt>(ConstArg->getAggregateElement(Idx))) {
const APInt &V = Elt->getValue();
APInt V2 = APInt(V.getBitWidth(), 1) << V.countr_zero();
Elements.push_back(ConstantInt::get(EltTy, V2));
} else {
Elements.push_back(ConstantInt::get(EltTy, 1));
}
}
ShadowMul = ConstantVector::get(Elements);
} else {
if (ConstantInt *Elt = dyn_cast<ConstantInt>(ConstArg)) {
const APInt &V = Elt->getValue();
APInt V2 = APInt(V.getBitWidth(), 1) << V.countr_zero();
ShadowMul = ConstantInt::get(Ty, V2);
} else {
ShadowMul = ConstantInt::get(Ty, 1);
}
}
IRBuilder<> IRB(&I);
setShadow(&I,
IRB.CreateMul(getShadow(OtherArg), ShadowMul, "msprop_mul_cst"));
setOrigin(&I, getOrigin(OtherArg));
}
void visitMul(BinaryOperator &I) {
Constant *constOp0 = dyn_cast<Constant>(I.getOperand(0));
Constant *constOp1 = dyn_cast<Constant>(I.getOperand(1));
if (constOp0 && !constOp1)
handleMulByConstant(I, constOp0, I.getOperand(1));
else if (constOp1 && !constOp0)
handleMulByConstant(I, constOp1, I.getOperand(0));
else
handleShadowOr(I);
}
void visitFAdd(BinaryOperator &I) { handleShadowOr(I); }
void visitFSub(BinaryOperator &I) { handleShadowOr(I); }
void visitFMul(BinaryOperator &I) { handleShadowOr(I); }
void visitAdd(BinaryOperator &I) { handleShadowOr(I); }
void visitSub(BinaryOperator &I) { handleShadowOr(I); }
void visitXor(BinaryOperator &I) { handleShadowOr(I); }
void handleIntegerDiv(Instruction &I) {
IRBuilder<> IRB(&I);
// Strict on the second argument.
insertShadowCheck(I.getOperand(1), &I);
setShadow(&I, getShadow(&I, 0));
setOrigin(&I, getOrigin(&I, 0));
}
void visitUDiv(BinaryOperator &I) { handleIntegerDiv(I); }
void visitSDiv(BinaryOperator &I) { handleIntegerDiv(I); }
void visitURem(BinaryOperator &I) { handleIntegerDiv(I); }
void visitSRem(BinaryOperator &I) { handleIntegerDiv(I); }
// Floating point division is side-effect free. We can not require that the
// divisor is fully initialized and must propagate shadow. See PR37523.
void visitFDiv(BinaryOperator &I) { handleShadowOr(I); }
void visitFRem(BinaryOperator &I) { handleShadowOr(I); }
/// Instrument == and != comparisons.
///
/// Sometimes the comparison result is known even if some of the bits of the
/// arguments are not.
void handleEqualityComparison(ICmpInst &I) {
IRBuilder<> IRB(&I);
Value *A = I.getOperand(0);
Value *B = I.getOperand(1);
Value *Sa = getShadow(A);
Value *Sb = getShadow(B);
// Get rid of pointers and vectors of pointers.
// For ints (and vectors of ints), types of A and Sa match,
// and this is a no-op.
A = IRB.CreatePointerCast(A, Sa->getType());
B = IRB.CreatePointerCast(B, Sb->getType());
// A == B <==> (C = A^B) == 0
// A != B <==> (C = A^B) != 0
// Sc = Sa | Sb
Value *C = IRB.CreateXor(A, B);
Value *Sc = IRB.CreateOr(Sa, Sb);
// Now dealing with i = (C == 0) comparison (or C != 0, does not matter now)
// Result is defined if one of the following is true
// * there is a defined 1 bit in C
// * C is fully defined
// Si = !(C & ~Sc) && Sc
Value *Zero = Constant::getNullValue(Sc->getType());
Value *MinusOne = Constant::getAllOnesValue(Sc->getType());
Value *LHS = IRB.CreateICmpNE(Sc, Zero);
Value *RHS =
IRB.CreateICmpEQ(IRB.CreateAnd(IRB.CreateXor(Sc, MinusOne), C), Zero);
Value *Si = IRB.CreateAnd(LHS, RHS);
Si->setName("_msprop_icmp");
setShadow(&I, Si);
setOriginForNaryOp(I);
}
/// Build the lowest possible value of V, taking into account V's
/// uninitialized bits.
Value *getLowestPossibleValue(IRBuilder<> &IRB, Value *A, Value *Sa,
bool isSigned) {
if (isSigned) {
// Split shadow into sign bit and other bits.
Value *SaOtherBits = IRB.CreateLShr(IRB.CreateShl(Sa, 1), 1);
Value *SaSignBit = IRB.CreateXor(Sa, SaOtherBits);
// Maximise the undefined shadow bit, minimize other undefined bits.
return IRB.CreateOr(IRB.CreateAnd(A, IRB.CreateNot(SaOtherBits)),
SaSignBit);
} else {
// Minimize undefined bits.
return IRB.CreateAnd(A, IRB.CreateNot(Sa));
}
}
/// Build the highest possible value of V, taking into account V's
/// uninitialized bits.
Value *getHighestPossibleValue(IRBuilder<> &IRB, Value *A, Value *Sa,
bool isSigned) {
if (isSigned) {
// Split shadow into sign bit and other bits.
Value *SaOtherBits = IRB.CreateLShr(IRB.CreateShl(Sa, 1), 1);
Value *SaSignBit = IRB.CreateXor(Sa, SaOtherBits);
// Minimise the undefined shadow bit, maximise other undefined bits.
return IRB.CreateOr(IRB.CreateAnd(A, IRB.CreateNot(SaSignBit)),
SaOtherBits);
} else {
// Maximize undefined bits.
return IRB.CreateOr(A, Sa);
}
}
/// Instrument relational comparisons.
///
/// This function does exact shadow propagation for all relational
/// comparisons of integers, pointers and vectors of those.
/// FIXME: output seems suboptimal when one of the operands is a constant
void handleRelationalComparisonExact(ICmpInst &I) {
IRBuilder<> IRB(&I);
Value *A = I.getOperand(0);
Value *B = I.getOperand(1);
Value *Sa = getShadow(A);
Value *Sb = getShadow(B);
// Get rid of pointers and vectors of pointers.
// For ints (and vectors of ints), types of A and Sa match,
// and this is a no-op.
A = IRB.CreatePointerCast(A, Sa->getType());
B = IRB.CreatePointerCast(B, Sb->getType());
// Let [a0, a1] be the interval of possible values of A, taking into account
// its undefined bits. Let [b0, b1] be the interval of possible values of B.
// Then (A cmp B) is defined iff (a0 cmp b1) == (a1 cmp b0).
bool IsSigned = I.isSigned();
Value *S1 = IRB.CreateICmp(I.getPredicate(),
getLowestPossibleValue(IRB, A, Sa, IsSigned),
getHighestPossibleValue(IRB, B, Sb, IsSigned));
Value *S2 = IRB.CreateICmp(I.getPredicate(),
getHighestPossibleValue(IRB, A, Sa, IsSigned),
getLowestPossibleValue(IRB, B, Sb, IsSigned));
Value *Si = IRB.CreateXor(S1, S2);
setShadow(&I, Si);
setOriginForNaryOp(I);
}
/// Instrument signed relational comparisons.
///
/// Handle sign bit tests: x<0, x>=0, x<=-1, x>-1 by propagating the highest
/// bit of the shadow. Everything else is delegated to handleShadowOr().
void handleSignedRelationalComparison(ICmpInst &I) {
Constant *constOp;
Value *op = nullptr;
CmpInst::Predicate pre;
if ((constOp = dyn_cast<Constant>(I.getOperand(1)))) {
op = I.getOperand(0);
pre = I.getPredicate();
} else if ((constOp = dyn_cast<Constant>(I.getOperand(0)))) {
op = I.getOperand(1);
pre = I.getSwappedPredicate();
} else {
handleShadowOr(I);
return;
}
if ((constOp->isNullValue() &&
(pre == CmpInst::ICMP_SLT || pre == CmpInst::ICMP_SGE)) ||
(constOp->isAllOnesValue() &&
(pre == CmpInst::ICMP_SGT || pre == CmpInst::ICMP_SLE))) {
IRBuilder<> IRB(&I);
Value *Shadow = IRB.CreateICmpSLT(getShadow(op), getCleanShadow(op),
"_msprop_icmp_s");
setShadow(&I, Shadow);
setOrigin(&I, getOrigin(op));
} else {
handleShadowOr(I);
}
}
void visitICmpInst(ICmpInst &I) {
if (!ClHandleICmp) {
handleShadowOr(I);
return;
}
if (I.isEquality()) {
handleEqualityComparison(I);
return;
}
assert(I.isRelational());
if (ClHandleICmpExact) {
handleRelationalComparisonExact(I);
return;
}
if (I.isSigned()) {
handleSignedRelationalComparison(I);
return;
}
assert(I.isUnsigned());
if ((isa<Constant>(I.getOperand(0)) || isa<Constant>(I.getOperand(1)))) {
handleRelationalComparisonExact(I);
return;
}
handleShadowOr(I);
}
void visitFCmpInst(FCmpInst &I) { handleShadowOr(I); }
void handleShift(BinaryOperator &I) {
IRBuilder<> IRB(&I);
// If any of the S2 bits are poisoned, the whole thing is poisoned.
// Otherwise perform the same shift on S1.
Value *S1 = getShadow(&I, 0);
Value *S2 = getShadow(&I, 1);
Value *S2Conv =
IRB.CreateSExt(IRB.CreateICmpNE(S2, getCleanShadow(S2)), S2->getType());
Value *V2 = I.getOperand(1);
Value *Shift = IRB.CreateBinOp(I.getOpcode(), S1, V2);
setShadow(&I, IRB.CreateOr(Shift, S2Conv));
setOriginForNaryOp(I);
}
void visitShl(BinaryOperator &I) { handleShift(I); }
void visitAShr(BinaryOperator &I) { handleShift(I); }
void visitLShr(BinaryOperator &I) { handleShift(I); }
void handleFunnelShift(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
// If any of the S2 bits are poisoned, the whole thing is poisoned.
// Otherwise perform the same shift on S0 and S1.
Value *S0 = getShadow(&I, 0);
Value *S1 = getShadow(&I, 1);
Value *S2 = getShadow(&I, 2);
Value *S2Conv =
IRB.CreateSExt(IRB.CreateICmpNE(S2, getCleanShadow(S2)), S2->getType());
Value *V2 = I.getOperand(2);
Function *Intrin = Intrinsic::getDeclaration(
I.getModule(), I.getIntrinsicID(), S2Conv->getType());
Value *Shift = IRB.CreateCall(Intrin, {S0, S1, V2});
setShadow(&I, IRB.CreateOr(Shift, S2Conv));
setOriginForNaryOp(I);
}
/// Instrument llvm.memmove
///
/// At this point we don't know if llvm.memmove will be inlined or not.
/// If we don't instrument it and it gets inlined,
/// our interceptor will not kick in and we will lose the memmove.
/// If we instrument the call here, but it does not get inlined,
/// we will memove the shadow twice: which is bad in case
/// of overlapping regions. So, we simply lower the intrinsic to a call.
///
/// Similar situation exists for memcpy and memset.
void visitMemMoveInst(MemMoveInst &I) {
getShadow(I.getArgOperand(1)); // Ensure shadow initialized
IRBuilder<> IRB(&I);
IRB.CreateCall(MS.MemmoveFn,
{I.getArgOperand(0), I.getArgOperand(1),
IRB.CreateIntCast(I.getArgOperand(2), MS.IntptrTy, false)});
I.eraseFromParent();
}
/// Instrument memcpy
///
/// Similar to memmove: avoid copying shadow twice. This is somewhat
/// unfortunate as it may slowdown small constant memcpys.
/// FIXME: consider doing manual inline for small constant sizes and proper
/// alignment.
///
/// Note: This also handles memcpy.inline, which promises no calls to external
/// functions as an optimization. However, with instrumentation enabled this
/// is difficult to promise; additionally, we know that the MSan runtime
/// exists and provides __msan_memcpy(). Therefore, we assume that with
/// instrumentation it's safe to turn memcpy.inline into a call to
/// __msan_memcpy(). Should this be wrong, such as when implementing memcpy()
/// itself, instrumentation should be disabled with the no_sanitize attribute.
void visitMemCpyInst(MemCpyInst &I) {
getShadow(I.getArgOperand(1)); // Ensure shadow initialized
IRBuilder<> IRB(&I);
IRB.CreateCall(MS.MemcpyFn,
{I.getArgOperand(0), I.getArgOperand(1),
IRB.CreateIntCast(I.getArgOperand(2), MS.IntptrTy, false)});
I.eraseFromParent();
}
// Same as memcpy.
void visitMemSetInst(MemSetInst &I) {
IRBuilder<> IRB(&I);
IRB.CreateCall(
MS.MemsetFn,
{I.getArgOperand(0),
IRB.CreateIntCast(I.getArgOperand(1), IRB.getInt32Ty(), false),
IRB.CreateIntCast(I.getArgOperand(2), MS.IntptrTy, false)});
I.eraseFromParent();
}
void visitVAStartInst(VAStartInst &I) { VAHelper->visitVAStartInst(I); }
void visitVACopyInst(VACopyInst &I) { VAHelper->visitVACopyInst(I); }
/// Handle vector store-like intrinsics.
///
/// Instrument intrinsics that look like a simple SIMD store: writes memory,
/// has 1 pointer argument and 1 vector argument, returns void.
bool handleVectorStoreIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Addr = I.getArgOperand(0);
Value *Shadow = getShadow(&I, 1);
Value *ShadowPtr, *OriginPtr;
// We don't know the pointer alignment (could be unaligned SSE store!).
// Have to assume to worst case.
std::tie(ShadowPtr, OriginPtr) = getShadowOriginPtr(
Addr, IRB, Shadow->getType(), Align(1), /*isStore*/ true);
IRB.CreateAlignedStore(Shadow, ShadowPtr, Align(1));
if (ClCheckAccessAddress)
insertShadowCheck(Addr, &I);
// FIXME: factor out common code from materializeStores
if (MS.TrackOrigins)
IRB.CreateStore(getOrigin(&I, 1), OriginPtr);
return true;
}
/// Handle vector load-like intrinsics.
///
/// Instrument intrinsics that look like a simple SIMD load: reads memory,
/// has 1 pointer argument, returns a vector.
bool handleVectorLoadIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Addr = I.getArgOperand(0);
Type *ShadowTy = getShadowTy(&I);
Value *ShadowPtr = nullptr, *OriginPtr = nullptr;
if (PropagateShadow) {
// We don't know the pointer alignment (could be unaligned SSE load!).
// Have to assume to worst case.
const Align Alignment = Align(1);
std::tie(ShadowPtr, OriginPtr) =
getShadowOriginPtr(Addr, IRB, ShadowTy, Alignment, /*isStore*/ false);
setShadow(&I,
IRB.CreateAlignedLoad(ShadowTy, ShadowPtr, Alignment, "_msld"));
} else {
setShadow(&I, getCleanShadow(&I));
}
if (ClCheckAccessAddress)
insertShadowCheck(Addr, &I);
if (MS.TrackOrigins) {
if (PropagateShadow)
setOrigin(&I, IRB.CreateLoad(MS.OriginTy, OriginPtr));
else
setOrigin(&I, getCleanOrigin());
}
return true;
}
/// Handle (SIMD arithmetic)-like intrinsics.
///
/// Instrument intrinsics with any number of arguments of the same type,
/// equal to the return type. The type should be simple (no aggregates or
/// pointers; vectors are fine).
/// Caller guarantees that this intrinsic does not access memory.
bool maybeHandleSimpleNomemIntrinsic(IntrinsicInst &I) {
Type *RetTy = I.getType();
if (!(RetTy->isIntOrIntVectorTy() || RetTy->isFPOrFPVectorTy() ||
RetTy->isX86_MMXTy()))
return false;
unsigned NumArgOperands = I.arg_size();
for (unsigned i = 0; i < NumArgOperands; ++i) {
Type *Ty = I.getArgOperand(i)->getType();
if (Ty != RetTy)
return false;
}
IRBuilder<> IRB(&I);
ShadowAndOriginCombiner SC(this, IRB);
for (unsigned i = 0; i < NumArgOperands; ++i)
SC.Add(I.getArgOperand(i));
SC.Done(&I);
return true;
}
/// Heuristically instrument unknown intrinsics.
///
/// The main purpose of this code is to do something reasonable with all
/// random intrinsics we might encounter, most importantly - SIMD intrinsics.
/// We recognize several classes of intrinsics by their argument types and
/// ModRefBehaviour and apply special instrumentation when we are reasonably
/// sure that we know what the intrinsic does.
///
/// We special-case intrinsics where this approach fails. See llvm.bswap
/// handling as an example of that.
bool handleUnknownIntrinsic(IntrinsicInst &I) {
unsigned NumArgOperands = I.arg_size();
if (NumArgOperands == 0)
return false;
if (NumArgOperands == 2 && I.getArgOperand(0)->getType()->isPointerTy() &&
I.getArgOperand(1)->getType()->isVectorTy() &&
I.getType()->isVoidTy() && !I.onlyReadsMemory()) {
// This looks like a vector store.
return handleVectorStoreIntrinsic(I);
}
if (NumArgOperands == 1 && I.getArgOperand(0)->getType()->isPointerTy() &&
I.getType()->isVectorTy() && I.onlyReadsMemory()) {
// This looks like a vector load.
return handleVectorLoadIntrinsic(I);
}
if (I.doesNotAccessMemory())
if (maybeHandleSimpleNomemIntrinsic(I))
return true;
// FIXME: detect and handle SSE maskstore/maskload
return false;
}
void handleInvariantGroup(IntrinsicInst &I) {
setShadow(&I, getShadow(&I, 0));
setOrigin(&I, getOrigin(&I, 0));
}
void handleLifetimeStart(IntrinsicInst &I) {
if (!PoisonStack)
return;
AllocaInst *AI = llvm::findAllocaForValue(I.getArgOperand(1));
if (!AI)
InstrumentLifetimeStart = false;
LifetimeStartList.push_back(std::make_pair(&I, AI));
}
void handleBswap(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Op = I.getArgOperand(0);
Type *OpType = Op->getType();
Function *BswapFunc = Intrinsic::getDeclaration(
F.getParent(), Intrinsic::bswap, ArrayRef(&OpType, 1));
setShadow(&I, IRB.CreateCall(BswapFunc, getShadow(Op)));
setOrigin(&I, getOrigin(Op));
}
void handleCountZeroes(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Src = I.getArgOperand(0);
// Set the Output shadow based on input Shadow
Value *BoolShadow = IRB.CreateIsNotNull(getShadow(Src), "_mscz_bs");
// If zero poison is requested, mix in with the shadow
Constant *IsZeroPoison = cast<Constant>(I.getOperand(1));
if (!IsZeroPoison->isZeroValue()) {
Value *BoolZeroPoison = IRB.CreateIsNull(Src, "_mscz_bzp");
BoolShadow = IRB.CreateOr(BoolShadow, BoolZeroPoison, "_mscz_bs");
}
Value *OutputShadow =
IRB.CreateSExt(BoolShadow, getShadowTy(Src), "_mscz_os");
setShadow(&I, OutputShadow);
setOriginForNaryOp(I);
}
// Instrument vector convert intrinsic.
//
// This function instruments intrinsics like cvtsi2ss:
// %Out = int_xxx_cvtyyy(%ConvertOp)
// or
// %Out = int_xxx_cvtyyy(%CopyOp, %ConvertOp)
// Intrinsic converts \p NumUsedElements elements of \p ConvertOp to the same
// number \p Out elements, and (if has 2 arguments) copies the rest of the
// elements from \p CopyOp.
// In most cases conversion involves floating-point value which may trigger a
// hardware exception when not fully initialized. For this reason we require
// \p ConvertOp[0:NumUsedElements] to be fully initialized and trap otherwise.
// We copy the shadow of \p CopyOp[NumUsedElements:] to \p
// Out[NumUsedElements:]. This means that intrinsics without \p CopyOp always
// return a fully initialized value.
void handleVectorConvertIntrinsic(IntrinsicInst &I, int NumUsedElements,
bool HasRoundingMode = false) {
IRBuilder<> IRB(&I);
Value *CopyOp, *ConvertOp;
assert((!HasRoundingMode ||
isa<ConstantInt>(I.getArgOperand(I.arg_size() - 1))) &&
"Invalid rounding mode");
switch (I.arg_size() - HasRoundingMode) {
case 2:
CopyOp = I.getArgOperand(0);
ConvertOp = I.getArgOperand(1);
break;
case 1:
ConvertOp = I.getArgOperand(0);
CopyOp = nullptr;
break;
default:
llvm_unreachable("Cvt intrinsic with unsupported number of arguments.");
}
// The first *NumUsedElements* elements of ConvertOp are converted to the
// same number of output elements. The rest of the output is copied from
// CopyOp, or (if not available) filled with zeroes.
// Combine shadow for elements of ConvertOp that are used in this operation,
// and insert a check.
// FIXME: consider propagating shadow of ConvertOp, at least in the case of
// int->any conversion.
Value *ConvertShadow = getShadow(ConvertOp);
Value *AggShadow = nullptr;
if (ConvertOp->getType()->isVectorTy()) {
AggShadow = IRB.CreateExtractElement(
ConvertShadow, ConstantInt::get(IRB.getInt32Ty(), 0));
for (int i = 1; i < NumUsedElements; ++i) {
Value *MoreShadow = IRB.CreateExtractElement(
ConvertShadow, ConstantInt::get(IRB.getInt32Ty(), i));
AggShadow = IRB.CreateOr(AggShadow, MoreShadow);
}
} else {
AggShadow = ConvertShadow;
}
assert(AggShadow->getType()->isIntegerTy());
insertShadowCheck(AggShadow, getOrigin(ConvertOp), &I);
// Build result shadow by zero-filling parts of CopyOp shadow that come from
// ConvertOp.
if (CopyOp) {
assert(CopyOp->getType() == I.getType());
assert(CopyOp->getType()->isVectorTy());
Value *ResultShadow = getShadow(CopyOp);
Type *EltTy = cast<VectorType>(ResultShadow->getType())->getElementType();
for (int i = 0; i < NumUsedElements; ++i) {
ResultShadow = IRB.CreateInsertElement(
ResultShadow, ConstantInt::getNullValue(EltTy),
ConstantInt::get(IRB.getInt32Ty(), i));
}
setShadow(&I, ResultShadow);
setOrigin(&I, getOrigin(CopyOp));
} else {
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
}
}
// Given a scalar or vector, extract lower 64 bits (or less), and return all
// zeroes if it is zero, and all ones otherwise.
Value *Lower64ShadowExtend(IRBuilder<> &IRB, Value *S, Type *T) {
if (S->getType()->isVectorTy())
S = CreateShadowCast(IRB, S, IRB.getInt64Ty(), /* Signed */ true);
assert(S->getType()->getPrimitiveSizeInBits() <= 64);
Value *S2 = IRB.CreateICmpNE(S, getCleanShadow(S));
return CreateShadowCast(IRB, S2, T, /* Signed */ true);
}
// Given a vector, extract its first element, and return all
// zeroes if it is zero, and all ones otherwise.
Value *LowerElementShadowExtend(IRBuilder<> &IRB, Value *S, Type *T) {
Value *S1 = IRB.CreateExtractElement(S, (uint64_t)0);
Value *S2 = IRB.CreateICmpNE(S1, getCleanShadow(S1));
return CreateShadowCast(IRB, S2, T, /* Signed */ true);
}
Value *VariableShadowExtend(IRBuilder<> &IRB, Value *S) {
Type *T = S->getType();
assert(T->isVectorTy());
Value *S2 = IRB.CreateICmpNE(S, getCleanShadow(S));
return IRB.CreateSExt(S2, T);
}
// Instrument vector shift intrinsic.
//
// This function instruments intrinsics like int_x86_avx2_psll_w.
// Intrinsic shifts %In by %ShiftSize bits.
// %ShiftSize may be a vector. In that case the lower 64 bits determine shift
// size, and the rest is ignored. Behavior is defined even if shift size is
// greater than register (or field) width.
void handleVectorShiftIntrinsic(IntrinsicInst &I, bool Variable) {
assert(I.arg_size() == 2);
IRBuilder<> IRB(&I);
// If any of the S2 bits are poisoned, the whole thing is poisoned.
// Otherwise perform the same shift on S1.
Value *S1 = getShadow(&I, 0);
Value *S2 = getShadow(&I, 1);
Value *S2Conv = Variable ? VariableShadowExtend(IRB, S2)
: Lower64ShadowExtend(IRB, S2, getShadowTy(&I));
Value *V1 = I.getOperand(0);
Value *V2 = I.getOperand(1);
Value *Shift = IRB.CreateCall(I.getFunctionType(), I.getCalledOperand(),
{IRB.CreateBitCast(S1, V1->getType()), V2});
Shift = IRB.CreateBitCast(Shift, getShadowTy(&I));
setShadow(&I, IRB.CreateOr(Shift, S2Conv));
setOriginForNaryOp(I);
}
// Get an X86_MMX-sized vector type.
Type *getMMXVectorTy(unsigned EltSizeInBits) {
const unsigned X86_MMXSizeInBits = 64;
assert(EltSizeInBits != 0 && (X86_MMXSizeInBits % EltSizeInBits) == 0 &&
"Illegal MMX vector element size");
return FixedVectorType::get(IntegerType::get(*MS.C, EltSizeInBits),
X86_MMXSizeInBits / EltSizeInBits);
}
// Returns a signed counterpart for an (un)signed-saturate-and-pack
// intrinsic.
Intrinsic::ID getSignedPackIntrinsic(Intrinsic::ID id) {
switch (id) {
case Intrinsic::x86_sse2_packsswb_128:
case Intrinsic::x86_sse2_packuswb_128:
return Intrinsic::x86_sse2_packsswb_128;
case Intrinsic::x86_sse2_packssdw_128:
case Intrinsic::x86_sse41_packusdw:
return Intrinsic::x86_sse2_packssdw_128;
case Intrinsic::x86_avx2_packsswb:
case Intrinsic::x86_avx2_packuswb:
return Intrinsic::x86_avx2_packsswb;
case Intrinsic::x86_avx2_packssdw:
case Intrinsic::x86_avx2_packusdw:
return Intrinsic::x86_avx2_packssdw;
case Intrinsic::x86_mmx_packsswb:
case Intrinsic::x86_mmx_packuswb:
return Intrinsic::x86_mmx_packsswb;
case Intrinsic::x86_mmx_packssdw:
return Intrinsic::x86_mmx_packssdw;
default:
llvm_unreachable("unexpected intrinsic id");
}
}
// Instrument vector pack intrinsic.
//
// This function instruments intrinsics like x86_mmx_packsswb, that
// packs elements of 2 input vectors into half as many bits with saturation.
// Shadow is propagated with the signed variant of the same intrinsic applied
// to sext(Sa != zeroinitializer), sext(Sb != zeroinitializer).
// EltSizeInBits is used only for x86mmx arguments.
void handleVectorPackIntrinsic(IntrinsicInst &I, unsigned EltSizeInBits = 0) {
assert(I.arg_size() == 2);
bool isX86_MMX = I.getOperand(0)->getType()->isX86_MMXTy();
IRBuilder<> IRB(&I);
Value *S1 = getShadow(&I, 0);
Value *S2 = getShadow(&I, 1);
assert(isX86_MMX || S1->getType()->isVectorTy());
// SExt and ICmpNE below must apply to individual elements of input vectors.
// In case of x86mmx arguments, cast them to appropriate vector types and
// back.
Type *T = isX86_MMX ? getMMXVectorTy(EltSizeInBits) : S1->getType();
if (isX86_MMX) {
S1 = IRB.CreateBitCast(S1, T);
S2 = IRB.CreateBitCast(S2, T);
}
Value *S1_ext =
IRB.CreateSExt(IRB.CreateICmpNE(S1, Constant::getNullValue(T)), T);
Value *S2_ext =
IRB.CreateSExt(IRB.CreateICmpNE(S2, Constant::getNullValue(T)), T);
if (isX86_MMX) {
Type *X86_MMXTy = Type::getX86_MMXTy(*MS.C);
S1_ext = IRB.CreateBitCast(S1_ext, X86_MMXTy);
S2_ext = IRB.CreateBitCast(S2_ext, X86_MMXTy);
}
Function *ShadowFn = Intrinsic::getDeclaration(
F.getParent(), getSignedPackIntrinsic(I.getIntrinsicID()));
Value *S =
IRB.CreateCall(ShadowFn, {S1_ext, S2_ext}, "_msprop_vector_pack");
if (isX86_MMX)
S = IRB.CreateBitCast(S, getShadowTy(&I));
setShadow(&I, S);
setOriginForNaryOp(I);
}
// Instrument sum-of-absolute-differences intrinsic.
void handleVectorSadIntrinsic(IntrinsicInst &I) {
const unsigned SignificantBitsPerResultElement = 16;
bool isX86_MMX = I.getOperand(0)->getType()->isX86_MMXTy();
Type *ResTy = isX86_MMX ? IntegerType::get(*MS.C, 64) : I.getType();
unsigned ZeroBitsPerResultElement =
ResTy->getScalarSizeInBits() - SignificantBitsPerResultElement;
IRBuilder<> IRB(&I);
auto *Shadow0 = getShadow(&I, 0);
auto *Shadow1 = getShadow(&I, 1);
Value *S = IRB.CreateOr(Shadow0, Shadow1);
S = IRB.CreateBitCast(S, ResTy);
S = IRB.CreateSExt(IRB.CreateICmpNE(S, Constant::getNullValue(ResTy)),
ResTy);
S = IRB.CreateLShr(S, ZeroBitsPerResultElement);
S = IRB.CreateBitCast(S, getShadowTy(&I));
setShadow(&I, S);
setOriginForNaryOp(I);
}
// Instrument multiply-add intrinsic.
void handleVectorPmaddIntrinsic(IntrinsicInst &I,
unsigned EltSizeInBits = 0) {
bool isX86_MMX = I.getOperand(0)->getType()->isX86_MMXTy();
Type *ResTy = isX86_MMX ? getMMXVectorTy(EltSizeInBits * 2) : I.getType();
IRBuilder<> IRB(&I);
auto *Shadow0 = getShadow(&I, 0);
auto *Shadow1 = getShadow(&I, 1);
Value *S = IRB.CreateOr(Shadow0, Shadow1);
S = IRB.CreateBitCast(S, ResTy);
S = IRB.CreateSExt(IRB.CreateICmpNE(S, Constant::getNullValue(ResTy)),
ResTy);
S = IRB.CreateBitCast(S, getShadowTy(&I));
setShadow(&I, S);
setOriginForNaryOp(I);
}
// Instrument compare-packed intrinsic.
// Basically, an or followed by sext(icmp ne 0) to end up with all-zeros or
// all-ones shadow.
void handleVectorComparePackedIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Type *ResTy = getShadowTy(&I);
auto *Shadow0 = getShadow(&I, 0);
auto *Shadow1 = getShadow(&I, 1);
Value *S0 = IRB.CreateOr(Shadow0, Shadow1);
Value *S = IRB.CreateSExt(
IRB.CreateICmpNE(S0, Constant::getNullValue(ResTy)), ResTy);
setShadow(&I, S);
setOriginForNaryOp(I);
}
// Instrument compare-scalar intrinsic.
// This handles both cmp* intrinsics which return the result in the first
// element of a vector, and comi* which return the result as i32.
void handleVectorCompareScalarIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
auto *Shadow0 = getShadow(&I, 0);
auto *Shadow1 = getShadow(&I, 1);
Value *S0 = IRB.CreateOr(Shadow0, Shadow1);
Value *S = LowerElementShadowExtend(IRB, S0, getShadowTy(&I));
setShadow(&I, S);
setOriginForNaryOp(I);
}
// Instrument generic vector reduction intrinsics
// by ORing together all their fields.
void handleVectorReduceIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *S = IRB.CreateOrReduce(getShadow(&I, 0));
setShadow(&I, S);
setOrigin(&I, getOrigin(&I, 0));
}
// Instrument vector.reduce.or intrinsic.
// Valid (non-poisoned) set bits in the operand pull low the
// corresponding shadow bits.
void handleVectorReduceOrIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *OperandShadow = getShadow(&I, 0);
Value *OperandUnsetBits = IRB.CreateNot(I.getOperand(0));
Value *OperandUnsetOrPoison = IRB.CreateOr(OperandUnsetBits, OperandShadow);
// Bit N is clean if any field's bit N is 1 and unpoison
Value *OutShadowMask = IRB.CreateAndReduce(OperandUnsetOrPoison);
// Otherwise, it is clean if every field's bit N is unpoison
Value *OrShadow = IRB.CreateOrReduce(OperandShadow);
Value *S = IRB.CreateAnd(OutShadowMask, OrShadow);
setShadow(&I, S);
setOrigin(&I, getOrigin(&I, 0));
}
// Instrument vector.reduce.and intrinsic.
// Valid (non-poisoned) unset bits in the operand pull down the
// corresponding shadow bits.
void handleVectorReduceAndIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *OperandShadow = getShadow(&I, 0);
Value *OperandSetOrPoison = IRB.CreateOr(I.getOperand(0), OperandShadow);
// Bit N is clean if any field's bit N is 0 and unpoison
Value *OutShadowMask = IRB.CreateAndReduce(OperandSetOrPoison);
// Otherwise, it is clean if every field's bit N is unpoison
Value *OrShadow = IRB.CreateOrReduce(OperandShadow);
Value *S = IRB.CreateAnd(OutShadowMask, OrShadow);
setShadow(&I, S);
setOrigin(&I, getOrigin(&I, 0));
}
void handleStmxcsr(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Addr = I.getArgOperand(0);
Type *Ty = IRB.getInt32Ty();
Value *ShadowPtr =
getShadowOriginPtr(Addr, IRB, Ty, Align(1), /*isStore*/ true).first;
IRB.CreateStore(getCleanShadow(Ty), ShadowPtr);
if (ClCheckAccessAddress)
insertShadowCheck(Addr, &I);
}
void handleLdmxcsr(IntrinsicInst &I) {
if (!InsertChecks)
return;
IRBuilder<> IRB(&I);
Value *Addr = I.getArgOperand(0);
Type *Ty = IRB.getInt32Ty();
const Align Alignment = Align(1);
Value *ShadowPtr, *OriginPtr;
std::tie(ShadowPtr, OriginPtr) =
getShadowOriginPtr(Addr, IRB, Ty, Alignment, /*isStore*/ false);
if (ClCheckAccessAddress)
insertShadowCheck(Addr, &I);
Value *Shadow = IRB.CreateAlignedLoad(Ty, ShadowPtr, Alignment, "_ldmxcsr");
Value *Origin = MS.TrackOrigins ? IRB.CreateLoad(MS.OriginTy, OriginPtr)
: getCleanOrigin();
insertShadowCheck(Shadow, Origin, &I);
}
void handleMaskedExpandLoad(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Ptr = I.getArgOperand(0);
Value *Mask = I.getArgOperand(1);
Value *PassThru = I.getArgOperand(2);
if (ClCheckAccessAddress) {
insertShadowCheck(Ptr, &I);
insertShadowCheck(Mask, &I);
}
if (!PropagateShadow) {
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
return;
}
Type *ShadowTy = getShadowTy(&I);
Type *ElementShadowTy = cast<VectorType>(ShadowTy)->getElementType();
auto [ShadowPtr, OriginPtr] =
getShadowOriginPtr(Ptr, IRB, ElementShadowTy, {}, /*isStore*/ false);
Value *Shadow = IRB.CreateMaskedExpandLoad(
ShadowTy, ShadowPtr, Mask, getShadow(PassThru), "_msmaskedexpload");
setShadow(&I, Shadow);
// TODO: Store origins.
setOrigin(&I, getCleanOrigin());
}
void handleMaskedCompressStore(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Values = I.getArgOperand(0);
Value *Ptr = I.getArgOperand(1);
Value *Mask = I.getArgOperand(2);
if (ClCheckAccessAddress) {
insertShadowCheck(Ptr, &I);
insertShadowCheck(Mask, &I);
}
Value *Shadow = getShadow(Values);
Type *ElementShadowTy =
getShadowTy(cast<VectorType>(Values->getType())->getElementType());
auto [ShadowPtr, OriginPtrs] =
getShadowOriginPtr(Ptr, IRB, ElementShadowTy, {}, /*isStore*/ true);
IRB.CreateMaskedCompressStore(Shadow, ShadowPtr, Mask);
// TODO: Store origins.
}
void handleMaskedGather(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Ptrs = I.getArgOperand(0);
const Align Alignment(
cast<ConstantInt>(I.getArgOperand(1))->getZExtValue());
Value *Mask = I.getArgOperand(2);
Value *PassThru = I.getArgOperand(3);
Type *PtrsShadowTy = getShadowTy(Ptrs);
if (ClCheckAccessAddress) {
insertShadowCheck(Mask, &I);
Value *MaskedPtrShadow = IRB.CreateSelect(
Mask, getShadow(Ptrs), Constant::getNullValue((PtrsShadowTy)),
"_msmaskedptrs");
insertShadowCheck(MaskedPtrShadow, getOrigin(Ptrs), &I);
}
if (!PropagateShadow) {
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
return;
}
Type *ShadowTy = getShadowTy(&I);
Type *ElementShadowTy = cast<VectorType>(ShadowTy)->getElementType();
auto [ShadowPtrs, OriginPtrs] = getShadowOriginPtr(
Ptrs, IRB, ElementShadowTy, Alignment, /*isStore*/ false);
Value *Shadow =
IRB.CreateMaskedGather(ShadowTy, ShadowPtrs, Alignment, Mask,
getShadow(PassThru), "_msmaskedgather");
setShadow(&I, Shadow);
// TODO: Store origins.
setOrigin(&I, getCleanOrigin());
}
void handleMaskedScatter(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Values = I.getArgOperand(0);
Value *Ptrs = I.getArgOperand(1);
const Align Alignment(
cast<ConstantInt>(I.getArgOperand(2))->getZExtValue());
Value *Mask = I.getArgOperand(3);
Type *PtrsShadowTy = getShadowTy(Ptrs);
if (ClCheckAccessAddress) {
insertShadowCheck(Mask, &I);
Value *MaskedPtrShadow = IRB.CreateSelect(
Mask, getShadow(Ptrs), Constant::getNullValue((PtrsShadowTy)),
"_msmaskedptrs");
insertShadowCheck(MaskedPtrShadow, getOrigin(Ptrs), &I);
}
Value *Shadow = getShadow(Values);
Type *ElementShadowTy =
getShadowTy(cast<VectorType>(Values->getType())->getElementType());
auto [ShadowPtrs, OriginPtrs] = getShadowOriginPtr(
Ptrs, IRB, ElementShadowTy, Alignment, /*isStore*/ true);
IRB.CreateMaskedScatter(Shadow, ShadowPtrs, Alignment, Mask);
// TODO: Store origin.
}
void handleMaskedStore(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *V = I.getArgOperand(0);
Value *Ptr = I.getArgOperand(1);
const Align Alignment(
cast<ConstantInt>(I.getArgOperand(2))->getZExtValue());
Value *Mask = I.getArgOperand(3);
Value *Shadow = getShadow(V);
if (ClCheckAccessAddress) {
insertShadowCheck(Ptr, &I);
insertShadowCheck(Mask, &I);
}
Value *ShadowPtr;
Value *OriginPtr;
std::tie(ShadowPtr, OriginPtr) = getShadowOriginPtr(
Ptr, IRB, Shadow->getType(), Alignment, /*isStore*/ true);
IRB.CreateMaskedStore(Shadow, ShadowPtr, Alignment, Mask);
if (!MS.TrackOrigins)
return;
auto &DL = F.getParent()->getDataLayout();
paintOrigin(IRB, getOrigin(V), OriginPtr,
DL.getTypeStoreSize(Shadow->getType()),
std::max(Alignment, kMinOriginAlignment));
}
void handleMaskedLoad(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Ptr = I.getArgOperand(0);
const Align Alignment(
cast<ConstantInt>(I.getArgOperand(1))->getZExtValue());
Value *Mask = I.getArgOperand(2);
Value *PassThru = I.getArgOperand(3);
if (ClCheckAccessAddress) {
insertShadowCheck(Ptr, &I);
insertShadowCheck(Mask, &I);
}
if (!PropagateShadow) {
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
return;
}
Type *ShadowTy = getShadowTy(&I);
Value *ShadowPtr, *OriginPtr;
std::tie(ShadowPtr, OriginPtr) =
getShadowOriginPtr(Ptr, IRB, ShadowTy, Alignment, /*isStore*/ false);
setShadow(&I, IRB.CreateMaskedLoad(ShadowTy, ShadowPtr, Alignment, Mask,
getShadow(PassThru), "_msmaskedld"));
if (!MS.TrackOrigins)
return;
// Choose between PassThru's and the loaded value's origins.
Value *MaskedPassThruShadow = IRB.CreateAnd(
getShadow(PassThru), IRB.CreateSExt(IRB.CreateNeg(Mask), ShadowTy));
Value *NotNull = convertToBool(MaskedPassThruShadow, IRB, "_mscmp");
Value *PtrOrigin = IRB.CreateLoad(MS.OriginTy, OriginPtr);
Value *Origin = IRB.CreateSelect(NotNull, getOrigin(PassThru), PtrOrigin);
setOrigin(&I, Origin);
}
// Instrument BMI / BMI2 intrinsics.
// All of these intrinsics are Z = I(X, Y)
// where the types of all operands and the result match, and are either i32 or
// i64. The following instrumentation happens to work for all of them:
// Sz = I(Sx, Y) | (sext (Sy != 0))
void handleBmiIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Type *ShadowTy = getShadowTy(&I);
// If any bit of the mask operand is poisoned, then the whole thing is.
Value *SMask = getShadow(&I, 1);
SMask = IRB.CreateSExt(IRB.CreateICmpNE(SMask, getCleanShadow(ShadowTy)),
ShadowTy);
// Apply the same intrinsic to the shadow of the first operand.
Value *S = IRB.CreateCall(I.getCalledFunction(),
{getShadow(&I, 0), I.getOperand(1)});
S = IRB.CreateOr(SMask, S);
setShadow(&I, S);
setOriginForNaryOp(I);
}
SmallVector<int, 8> getPclmulMask(unsigned Width, bool OddElements) {
SmallVector<int, 8> Mask;
for (unsigned X = OddElements ? 1 : 0; X < Width; X += 2) {
Mask.append(2, X);
}
return Mask;
}
// Instrument pclmul intrinsics.
// These intrinsics operate either on odd or on even elements of the input
// vectors, depending on the constant in the 3rd argument, ignoring the rest.
// Replace the unused elements with copies of the used ones, ex:
// (0, 1, 2, 3) -> (0, 0, 2, 2) (even case)
// or
// (0, 1, 2, 3) -> (1, 1, 3, 3) (odd case)
// and then apply the usual shadow combining logic.
void handlePclmulIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
unsigned Width =
cast<FixedVectorType>(I.getArgOperand(0)->getType())->getNumElements();
assert(isa<ConstantInt>(I.getArgOperand(2)) &&
"pclmul 3rd operand must be a constant");
unsigned Imm = cast<ConstantInt>(I.getArgOperand(2))->getZExtValue();
Value *Shuf0 = IRB.CreateShuffleVector(getShadow(&I, 0),
getPclmulMask(Width, Imm & 0x01));
Value *Shuf1 = IRB.CreateShuffleVector(getShadow(&I, 1),
getPclmulMask(Width, Imm & 0x10));
ShadowAndOriginCombiner SOC(this, IRB);
SOC.Add(Shuf0, getOrigin(&I, 0));
SOC.Add(Shuf1, getOrigin(&I, 1));
SOC.Done(&I);
}
// Instrument _mm_*_sd|ss intrinsics
void handleUnarySdSsIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
unsigned Width =
cast<FixedVectorType>(I.getArgOperand(0)->getType())->getNumElements();
Value *First = getShadow(&I, 0);
Value *Second = getShadow(&I, 1);
// First element of second operand, remaining elements of first operand
SmallVector<int, 16> Mask;
Mask.push_back(Width);
for (unsigned i = 1; i < Width; i++)
Mask.push_back(i);
Value *Shadow = IRB.CreateShuffleVector(First, Second, Mask);
setShadow(&I, Shadow);
setOriginForNaryOp(I);
}
void handleVtestIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Shadow0 = getShadow(&I, 0);
Value *Shadow1 = getShadow(&I, 1);
Value *Or = IRB.CreateOr(Shadow0, Shadow1);
Value *NZ = IRB.CreateICmpNE(Or, Constant::getNullValue(Or->getType()));
Value *Scalar = convertShadowToScalar(NZ, IRB);
Value *Shadow = IRB.CreateZExt(Scalar, getShadowTy(&I));
setShadow(&I, Shadow);
setOriginForNaryOp(I);
}
void handleBinarySdSsIntrinsic(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
unsigned Width =
cast<FixedVectorType>(I.getArgOperand(0)->getType())->getNumElements();
Value *First = getShadow(&I, 0);
Value *Second = getShadow(&I, 1);
Value *OrShadow = IRB.CreateOr(First, Second);
// First element of both OR'd together, remaining elements of first operand
SmallVector<int, 16> Mask;
Mask.push_back(Width);
for (unsigned i = 1; i < Width; i++)
Mask.push_back(i);
Value *Shadow = IRB.CreateShuffleVector(First, OrShadow, Mask);
setShadow(&I, Shadow);
setOriginForNaryOp(I);
}
// Instrument abs intrinsic.
// handleUnknownIntrinsic can't handle it because of the last
// is_int_min_poison argument which does not match the result type.
void handleAbsIntrinsic(IntrinsicInst &I) {
assert(I.getType()->isIntOrIntVectorTy());
assert(I.getArgOperand(0)->getType() == I.getType());
// FIXME: Handle is_int_min_poison.
IRBuilder<> IRB(&I);
setShadow(&I, getShadow(&I, 0));
setOrigin(&I, getOrigin(&I, 0));
}
void handleIsFpClass(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Shadow = getShadow(&I, 0);
setShadow(&I, IRB.CreateICmpNE(Shadow, getCleanShadow(Shadow)));
setOrigin(&I, getOrigin(&I, 0));
}
void handleArithmeticWithOverflow(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *Shadow0 = getShadow(&I, 0);
Value *Shadow1 = getShadow(&I, 1);
Value *ShadowElt0 = IRB.CreateOr(Shadow0, Shadow1);
Value *ShadowElt1 =
IRB.CreateICmpNE(ShadowElt0, getCleanShadow(ShadowElt0));
Value *Shadow = PoisonValue::get(getShadowTy(&I));
Shadow = IRB.CreateInsertValue(Shadow, ShadowElt0, 0);
Shadow = IRB.CreateInsertValue(Shadow, ShadowElt1, 1);
setShadow(&I, Shadow);
setOriginForNaryOp(I);
}
void visitIntrinsicInst(IntrinsicInst &I) {
switch (I.getIntrinsicID()) {
case Intrinsic::uadd_with_overflow:
case Intrinsic::sadd_with_overflow:
case Intrinsic::usub_with_overflow:
case Intrinsic::ssub_with_overflow:
case Intrinsic::umul_with_overflow:
case Intrinsic::smul_with_overflow:
handleArithmeticWithOverflow(I);
break;
case Intrinsic::abs:
handleAbsIntrinsic(I);
break;
case Intrinsic::is_fpclass:
handleIsFpClass(I);
break;
case Intrinsic::lifetime_start:
handleLifetimeStart(I);
break;
case Intrinsic::launder_invariant_group:
case Intrinsic::strip_invariant_group:
handleInvariantGroup(I);
break;
case Intrinsic::bswap:
handleBswap(I);
break;
case Intrinsic::ctlz:
case Intrinsic::cttz:
handleCountZeroes(I);
break;
case Intrinsic::masked_compressstore:
handleMaskedCompressStore(I);
break;
case Intrinsic::masked_expandload:
handleMaskedExpandLoad(I);
break;
case Intrinsic::masked_gather:
handleMaskedGather(I);
break;
case Intrinsic::masked_scatter:
handleMaskedScatter(I);
break;
case Intrinsic::masked_store:
handleMaskedStore(I);
break;
case Intrinsic::masked_load:
handleMaskedLoad(I);
break;
case Intrinsic::vector_reduce_and:
handleVectorReduceAndIntrinsic(I);
break;
case Intrinsic::vector_reduce_or:
handleVectorReduceOrIntrinsic(I);
break;
case Intrinsic::vector_reduce_add:
case Intrinsic::vector_reduce_xor:
case Intrinsic::vector_reduce_mul:
handleVectorReduceIntrinsic(I);
break;
case Intrinsic::x86_sse_stmxcsr:
handleStmxcsr(I);
break;
case Intrinsic::x86_sse_ldmxcsr:
handleLdmxcsr(I);
break;
case Intrinsic::x86_avx512_vcvtsd2usi64:
case Intrinsic::x86_avx512_vcvtsd2usi32:
case Intrinsic::x86_avx512_vcvtss2usi64:
case Intrinsic::x86_avx512_vcvtss2usi32:
case Intrinsic::x86_avx512_cvttss2usi64:
case Intrinsic::x86_avx512_cvttss2usi:
case Intrinsic::x86_avx512_cvttsd2usi64:
case Intrinsic::x86_avx512_cvttsd2usi:
case Intrinsic::x86_avx512_cvtusi2ss:
case Intrinsic::x86_avx512_cvtusi642sd:
case Intrinsic::x86_avx512_cvtusi642ss:
handleVectorConvertIntrinsic(I, 1, true);
break;
case Intrinsic::x86_sse2_cvtsd2si64:
case Intrinsic::x86_sse2_cvtsd2si:
case Intrinsic::x86_sse2_cvtsd2ss:
case Intrinsic::x86_sse2_cvttsd2si64:
case Intrinsic::x86_sse2_cvttsd2si:
case Intrinsic::x86_sse_cvtss2si64:
case Intrinsic::x86_sse_cvtss2si:
case Intrinsic::x86_sse_cvttss2si64:
case Intrinsic::x86_sse_cvttss2si:
handleVectorConvertIntrinsic(I, 1);
break;
case Intrinsic::x86_sse_cvtps2pi:
case Intrinsic::x86_sse_cvttps2pi:
handleVectorConvertIntrinsic(I, 2);
break;
case Intrinsic::x86_avx512_psll_w_512:
case Intrinsic::x86_avx512_psll_d_512:
case Intrinsic::x86_avx512_psll_q_512:
case Intrinsic::x86_avx512_pslli_w_512:
case Intrinsic::x86_avx512_pslli_d_512:
case Intrinsic::x86_avx512_pslli_q_512:
case Intrinsic::x86_avx512_psrl_w_512:
case Intrinsic::x86_avx512_psrl_d_512:
case Intrinsic::x86_avx512_psrl_q_512:
case Intrinsic::x86_avx512_psra_w_512:
case Intrinsic::x86_avx512_psra_d_512:
case Intrinsic::x86_avx512_psra_q_512:
case Intrinsic::x86_avx512_psrli_w_512:
case Intrinsic::x86_avx512_psrli_d_512:
case Intrinsic::x86_avx512_psrli_q_512:
case Intrinsic::x86_avx512_psrai_w_512:
case Intrinsic::x86_avx512_psrai_d_512:
case Intrinsic::x86_avx512_psrai_q_512:
case Intrinsic::x86_avx512_psra_q_256:
case Intrinsic::x86_avx512_psra_q_128:
case Intrinsic::x86_avx512_psrai_q_256:
case Intrinsic::x86_avx512_psrai_q_128:
case Intrinsic::x86_avx2_psll_w:
case Intrinsic::x86_avx2_psll_d:
case Intrinsic::x86_avx2_psll_q:
case Intrinsic::x86_avx2_pslli_w:
case Intrinsic::x86_avx2_pslli_d:
case Intrinsic::x86_avx2_pslli_q:
case Intrinsic::x86_avx2_psrl_w:
case Intrinsic::x86_avx2_psrl_d:
case Intrinsic::x86_avx2_psrl_q:
case Intrinsic::x86_avx2_psra_w:
case Intrinsic::x86_avx2_psra_d:
case Intrinsic::x86_avx2_psrli_w:
case Intrinsic::x86_avx2_psrli_d:
case Intrinsic::x86_avx2_psrli_q:
case Intrinsic::x86_avx2_psrai_w:
case Intrinsic::x86_avx2_psrai_d:
case Intrinsic::x86_sse2_psll_w:
case Intrinsic::x86_sse2_psll_d:
case Intrinsic::x86_sse2_psll_q:
case Intrinsic::x86_sse2_pslli_w:
case Intrinsic::x86_sse2_pslli_d:
case Intrinsic::x86_sse2_pslli_q:
case Intrinsic::x86_sse2_psrl_w:
case Intrinsic::x86_sse2_psrl_d:
case Intrinsic::x86_sse2_psrl_q:
case Intrinsic::x86_sse2_psra_w:
case Intrinsic::x86_sse2_psra_d:
case Intrinsic::x86_sse2_psrli_w:
case Intrinsic::x86_sse2_psrli_d:
case Intrinsic::x86_sse2_psrli_q:
case Intrinsic::x86_sse2_psrai_w:
case Intrinsic::x86_sse2_psrai_d:
case Intrinsic::x86_mmx_psll_w:
case Intrinsic::x86_mmx_psll_d:
case Intrinsic::x86_mmx_psll_q:
case Intrinsic::x86_mmx_pslli_w:
case Intrinsic::x86_mmx_pslli_d:
case Intrinsic::x86_mmx_pslli_q:
case Intrinsic::x86_mmx_psrl_w:
case Intrinsic::x86_mmx_psrl_d:
case Intrinsic::x86_mmx_psrl_q:
case Intrinsic::x86_mmx_psra_w:
case Intrinsic::x86_mmx_psra_d:
case Intrinsic::x86_mmx_psrli_w:
case Intrinsic::x86_mmx_psrli_d:
case Intrinsic::x86_mmx_psrli_q:
case Intrinsic::x86_mmx_psrai_w:
case Intrinsic::x86_mmx_psrai_d:
handleVectorShiftIntrinsic(I, /* Variable */ false);
break;
case Intrinsic::x86_avx2_psllv_d:
case Intrinsic::x86_avx2_psllv_d_256:
case Intrinsic::x86_avx512_psllv_d_512:
case Intrinsic::x86_avx2_psllv_q:
case Intrinsic::x86_avx2_psllv_q_256:
case Intrinsic::x86_avx512_psllv_q_512:
case Intrinsic::x86_avx2_psrlv_d:
case Intrinsic::x86_avx2_psrlv_d_256:
case Intrinsic::x86_avx512_psrlv_d_512:
case Intrinsic::x86_avx2_psrlv_q:
case Intrinsic::x86_avx2_psrlv_q_256:
case Intrinsic::x86_avx512_psrlv_q_512:
case Intrinsic::x86_avx2_psrav_d:
case Intrinsic::x86_avx2_psrav_d_256:
case Intrinsic::x86_avx512_psrav_d_512:
case Intrinsic::x86_avx512_psrav_q_128:
case Intrinsic::x86_avx512_psrav_q_256:
case Intrinsic::x86_avx512_psrav_q_512:
handleVectorShiftIntrinsic(I, /* Variable */ true);
break;
case Intrinsic::x86_sse2_packsswb_128:
case Intrinsic::x86_sse2_packssdw_128:
case Intrinsic::x86_sse2_packuswb_128:
case Intrinsic::x86_sse41_packusdw:
case Intrinsic::x86_avx2_packsswb:
case Intrinsic::x86_avx2_packssdw:
case Intrinsic::x86_avx2_packuswb:
case Intrinsic::x86_avx2_packusdw:
handleVectorPackIntrinsic(I);
break;
case Intrinsic::x86_mmx_packsswb:
case Intrinsic::x86_mmx_packuswb:
handleVectorPackIntrinsic(I, 16);
break;
case Intrinsic::x86_mmx_packssdw:
handleVectorPackIntrinsic(I, 32);
break;
case Intrinsic::x86_mmx_psad_bw:
case Intrinsic::x86_sse2_psad_bw:
case Intrinsic::x86_avx2_psad_bw:
handleVectorSadIntrinsic(I);
break;
case Intrinsic::x86_sse2_pmadd_wd:
case Intrinsic::x86_avx2_pmadd_wd:
case Intrinsic::x86_ssse3_pmadd_ub_sw_128:
case Intrinsic::x86_avx2_pmadd_ub_sw:
handleVectorPmaddIntrinsic(I);
break;
case Intrinsic::x86_ssse3_pmadd_ub_sw:
handleVectorPmaddIntrinsic(I, 8);
break;
case Intrinsic::x86_mmx_pmadd_wd:
handleVectorPmaddIntrinsic(I, 16);
break;
case Intrinsic::x86_sse_cmp_ss:
case Intrinsic::x86_sse2_cmp_sd:
case Intrinsic::x86_sse_comieq_ss:
case Intrinsic::x86_sse_comilt_ss:
case Intrinsic::x86_sse_comile_ss:
case Intrinsic::x86_sse_comigt_ss:
case Intrinsic::x86_sse_comige_ss:
case Intrinsic::x86_sse_comineq_ss:
case Intrinsic::x86_sse_ucomieq_ss:
case Intrinsic::x86_sse_ucomilt_ss:
case Intrinsic::x86_sse_ucomile_ss:
case Intrinsic::x86_sse_ucomigt_ss:
case Intrinsic::x86_sse_ucomige_ss:
case Intrinsic::x86_sse_ucomineq_ss:
case Intrinsic::x86_sse2_comieq_sd:
case Intrinsic::x86_sse2_comilt_sd:
case Intrinsic::x86_sse2_comile_sd:
case Intrinsic::x86_sse2_comigt_sd:
case Intrinsic::x86_sse2_comige_sd:
case Intrinsic::x86_sse2_comineq_sd:
case Intrinsic::x86_sse2_ucomieq_sd:
case Intrinsic::x86_sse2_ucomilt_sd:
case Intrinsic::x86_sse2_ucomile_sd:
case Intrinsic::x86_sse2_ucomigt_sd:
case Intrinsic::x86_sse2_ucomige_sd:
case Intrinsic::x86_sse2_ucomineq_sd:
handleVectorCompareScalarIntrinsic(I);
break;
case Intrinsic::x86_avx_cmp_pd_256:
case Intrinsic::x86_avx_cmp_ps_256:
case Intrinsic::x86_sse2_cmp_pd:
case Intrinsic::x86_sse_cmp_ps:
handleVectorComparePackedIntrinsic(I);
break;
case Intrinsic::x86_bmi_bextr_32:
case Intrinsic::x86_bmi_bextr_64:
case Intrinsic::x86_bmi_bzhi_32:
case Intrinsic::x86_bmi_bzhi_64:
case Intrinsic::x86_bmi_pdep_32:
case Intrinsic::x86_bmi_pdep_64:
case Intrinsic::x86_bmi_pext_32:
case Intrinsic::x86_bmi_pext_64:
handleBmiIntrinsic(I);
break;
case Intrinsic::x86_pclmulqdq:
case Intrinsic::x86_pclmulqdq_256:
case Intrinsic::x86_pclmulqdq_512:
handlePclmulIntrinsic(I);
break;
case Intrinsic::x86_sse41_round_sd:
case Intrinsic::x86_sse41_round_ss:
handleUnarySdSsIntrinsic(I);
break;
case Intrinsic::x86_sse2_max_sd:
case Intrinsic::x86_sse_max_ss:
case Intrinsic::x86_sse2_min_sd:
case Intrinsic::x86_sse_min_ss:
handleBinarySdSsIntrinsic(I);
break;
case Intrinsic::x86_avx_vtestc_pd:
case Intrinsic::x86_avx_vtestc_pd_256:
case Intrinsic::x86_avx_vtestc_ps:
case Intrinsic::x86_avx_vtestc_ps_256:
case Intrinsic::x86_avx_vtestnzc_pd:
case Intrinsic::x86_avx_vtestnzc_pd_256:
case Intrinsic::x86_avx_vtestnzc_ps:
case Intrinsic::x86_avx_vtestnzc_ps_256:
case Intrinsic::x86_avx_vtestz_pd:
case Intrinsic::x86_avx_vtestz_pd_256:
case Intrinsic::x86_avx_vtestz_ps:
case Intrinsic::x86_avx_vtestz_ps_256:
case Intrinsic::x86_avx_ptestc_256:
case Intrinsic::x86_avx_ptestnzc_256:
case Intrinsic::x86_avx_ptestz_256:
case Intrinsic::x86_sse41_ptestc:
case Intrinsic::x86_sse41_ptestnzc:
case Intrinsic::x86_sse41_ptestz:
handleVtestIntrinsic(I);
break;
case Intrinsic::fshl:
case Intrinsic::fshr:
handleFunnelShift(I);
break;
case Intrinsic::is_constant:
// The result of llvm.is.constant() is always defined.
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
break;
default:
if (!handleUnknownIntrinsic(I))
visitInstruction(I);
break;
}
}
void visitLibAtomicLoad(CallBase &CB) {
// Since we use getNextNode here, we can't have CB terminate the BB.
assert(isa<CallInst>(CB));
IRBuilder<> IRB(&CB);
Value *Size = CB.getArgOperand(0);
Value *SrcPtr = CB.getArgOperand(1);
Value *DstPtr = CB.getArgOperand(2);
Value *Ordering = CB.getArgOperand(3);
// Convert the call to have at least Acquire ordering to make sure
// the shadow operations aren't reordered before it.
Value *NewOrdering =
IRB.CreateExtractElement(makeAddAcquireOrderingTable(IRB), Ordering);
CB.setArgOperand(3, NewOrdering);
NextNodeIRBuilder NextIRB(&CB);
Value *SrcShadowPtr, *SrcOriginPtr;
std::tie(SrcShadowPtr, SrcOriginPtr) =
getShadowOriginPtr(SrcPtr, NextIRB, NextIRB.getInt8Ty(), Align(1),
/*isStore*/ false);
Value *DstShadowPtr =
getShadowOriginPtr(DstPtr, NextIRB, NextIRB.getInt8Ty(), Align(1),
/*isStore*/ true)
.first;
NextIRB.CreateMemCpy(DstShadowPtr, Align(1), SrcShadowPtr, Align(1), Size);
if (MS.TrackOrigins) {
Value *SrcOrigin = NextIRB.CreateAlignedLoad(MS.OriginTy, SrcOriginPtr,
kMinOriginAlignment);
Value *NewOrigin = updateOrigin(SrcOrigin, NextIRB);
NextIRB.CreateCall(MS.MsanSetOriginFn, {DstPtr, Size, NewOrigin});
}
}
void visitLibAtomicStore(CallBase &CB) {
IRBuilder<> IRB(&CB);
Value *Size = CB.getArgOperand(0);
Value *DstPtr = CB.getArgOperand(2);
Value *Ordering = CB.getArgOperand(3);
// Convert the call to have at least Release ordering to make sure
// the shadow operations aren't reordered after it.
Value *NewOrdering =
IRB.CreateExtractElement(makeAddReleaseOrderingTable(IRB), Ordering);
CB.setArgOperand(3, NewOrdering);
Value *DstShadowPtr =
getShadowOriginPtr(DstPtr, IRB, IRB.getInt8Ty(), Align(1),
/*isStore*/ true)
.first;
// Atomic store always paints clean shadow/origin. See file header.
IRB.CreateMemSet(DstShadowPtr, getCleanShadow(IRB.getInt8Ty()), Size,
Align(1));
}
void visitCallBase(CallBase &CB) {
assert(!CB.getMetadata(LLVMContext::MD_nosanitize));
if (CB.isInlineAsm()) {
// For inline asm (either a call to asm function, or callbr instruction),
// do the usual thing: check argument shadow and mark all outputs as
// clean. Note that any side effects of the inline asm that are not
// immediately visible in its constraints are not handled.
if (ClHandleAsmConservative)
visitAsmInstruction(CB);
else
visitInstruction(CB);
return;
}
LibFunc LF;
if (TLI->getLibFunc(CB, LF)) {
// libatomic.a functions need to have special handling because there isn't
// a good way to intercept them or compile the library with
// instrumentation.
switch (LF) {
case LibFunc_atomic_load:
if (!isa<CallInst>(CB)) {
llvm::errs() << "MSAN -- cannot instrument invoke of libatomic load."
"Ignoring!\n";
break;
}
visitLibAtomicLoad(CB);
return;
case LibFunc_atomic_store:
visitLibAtomicStore(CB);
return;
default:
break;
}
}
if (auto *Call = dyn_cast<CallInst>(&CB)) {
assert(!isa<IntrinsicInst>(Call) && "intrinsics are handled elsewhere");
// We are going to insert code that relies on the fact that the callee
// will become a non-readonly function after it is instrumented by us. To
// prevent this code from being optimized out, mark that function
// non-readonly in advance.
// TODO: We can likely do better than dropping memory() completely here.
AttributeMask B;
B.addAttribute(Attribute::Memory).addAttribute(Attribute::Speculatable);
Call->removeFnAttrs(B);
if (Function *Func = Call->getCalledFunction()) {
Func->removeFnAttrs(B);
}
maybeMarkSanitizerLibraryCallNoBuiltin(Call, TLI);
}
IRBuilder<> IRB(&CB);
bool MayCheckCall = MS.EagerChecks;
if (Function *Func = CB.getCalledFunction()) {
// __sanitizer_unaligned_{load,store} functions may be called by users
// and always expects shadows in the TLS. So don't check them.
MayCheckCall &= !Func->getName().starts_with("__sanitizer_unaligned_");
}
unsigned ArgOffset = 0;
LLVM_DEBUG(dbgs() << " CallSite: " << CB << "\n");
for (const auto &[i, A] : llvm::enumerate(CB.args())) {
if (!A->getType()->isSized()) {
LLVM_DEBUG(dbgs() << "Arg " << i << " is not sized: " << CB << "\n");
continue;
}
if (A->getType()->isScalableTy()) {
LLVM_DEBUG(dbgs() << "Arg " << i << " is vscale: " << CB << "\n");
// Handle as noundef, but don't reserve tls slots.
insertShadowCheck(A, &CB);
continue;
}
unsigned Size = 0;
const DataLayout &DL = F.getParent()->getDataLayout();
bool ByVal = CB.paramHasAttr(i, Attribute::ByVal);
bool NoUndef = CB.paramHasAttr(i, Attribute::NoUndef);
bool EagerCheck = MayCheckCall && !ByVal && NoUndef;
if (EagerCheck) {
insertShadowCheck(A, &CB);
Size = DL.getTypeAllocSize(A->getType());
} else {
Value *Store = nullptr;
// Compute the Shadow for arg even if it is ByVal, because
// in that case getShadow() will copy the actual arg shadow to
// __msan_param_tls.
Value *ArgShadow = getShadow(A);
Value *ArgShadowBase = getShadowPtrForArgument(IRB, ArgOffset);
LLVM_DEBUG(dbgs() << " Arg#" << i << ": " << *A
<< " Shadow: " << *ArgShadow << "\n");
if (ByVal) {
// ByVal requires some special handling as it's too big for a single
// load
assert(A->getType()->isPointerTy() &&
"ByVal argument is not a pointer!");
Size = DL.getTypeAllocSize(CB.getParamByValType(i));
if (ArgOffset + Size > kParamTLSSize)
break;
const MaybeAlign ParamAlignment(CB.getParamAlign(i));
MaybeAlign Alignment = std::nullopt;
if (ParamAlignment)
Alignment = std::min(*ParamAlignment, kShadowTLSAlignment);
Value *AShadowPtr, *AOriginPtr;
std::tie(AShadowPtr, AOriginPtr) =
getShadowOriginPtr(A, IRB, IRB.getInt8Ty(), Alignment,
/*isStore*/ false);
if (!PropagateShadow) {
Store = IRB.CreateMemSet(ArgShadowBase,
Constant::getNullValue(IRB.getInt8Ty()),
Size, Alignment);
} else {
Store = IRB.CreateMemCpy(ArgShadowBase, Alignment, AShadowPtr,
Alignment, Size);
if (MS.TrackOrigins) {
Value *ArgOriginBase = getOriginPtrForArgument(IRB, ArgOffset);
// FIXME: OriginSize should be:
// alignTo(A % kMinOriginAlignment + Size, kMinOriginAlignment)
unsigned OriginSize = alignTo(Size, kMinOriginAlignment);
IRB.CreateMemCpy(
ArgOriginBase,
/* by origin_tls[ArgOffset] */ kMinOriginAlignment,
AOriginPtr,
/* by getShadowOriginPtr */ kMinOriginAlignment, OriginSize);
}
}
} else {
// Any other parameters mean we need bit-grained tracking of uninit
// data
Size = DL.getTypeAllocSize(A->getType());
if (ArgOffset + Size > kParamTLSSize)
break;
Store = IRB.CreateAlignedStore(ArgShadow, ArgShadowBase,
kShadowTLSAlignment);
Constant *Cst = dyn_cast<Constant>(ArgShadow);
if (MS.TrackOrigins && !(Cst && Cst->isNullValue())) {
IRB.CreateStore(getOrigin(A),
getOriginPtrForArgument(IRB, ArgOffset));
}
}
(void)Store;
assert(Store != nullptr);
LLVM_DEBUG(dbgs() << " Param:" << *Store << "\n");
}
assert(Size != 0);
ArgOffset += alignTo(Size, kShadowTLSAlignment);
}
LLVM_DEBUG(dbgs() << " done with call args\n");
FunctionType *FT = CB.getFunctionType();
if (FT->isVarArg()) {
VAHelper->visitCallBase(CB, IRB);
}
// Now, get the shadow for the RetVal.
if (!CB.getType()->isSized())
return;
// Don't emit the epilogue for musttail call returns.
if (isa<CallInst>(CB) && cast<CallInst>(CB).isMustTailCall())
return;
if (MayCheckCall && CB.hasRetAttr(Attribute::NoUndef)) {
setShadow(&CB, getCleanShadow(&CB));
setOrigin(&CB, getCleanOrigin());
return;
}
IRBuilder<> IRBBefore(&CB);
// Until we have full dynamic coverage, make sure the retval shadow is 0.
Value *Base = getShadowPtrForRetval(IRBBefore);
IRBBefore.CreateAlignedStore(getCleanShadow(&CB), Base,
kShadowTLSAlignment);
BasicBlock::iterator NextInsn;
if (isa<CallInst>(CB)) {
NextInsn = ++CB.getIterator();
assert(NextInsn != CB.getParent()->end());
} else {
BasicBlock *NormalDest = cast<InvokeInst>(CB).getNormalDest();
if (!NormalDest->getSinglePredecessor()) {
// FIXME: this case is tricky, so we are just conservative here.
// Perhaps we need to split the edge between this BB and NormalDest,
// but a naive attempt to use SplitEdge leads to a crash.
setShadow(&CB, getCleanShadow(&CB));
setOrigin(&CB, getCleanOrigin());
return;
}
// FIXME: NextInsn is likely in a basic block that has not been visited
// yet. Anything inserted there will be instrumented by MSan later!
NextInsn = NormalDest->getFirstInsertionPt();
assert(NextInsn != NormalDest->end() &&
"Could not find insertion point for retval shadow load");
}
IRBuilder<> IRBAfter(&*NextInsn);
Value *RetvalShadow = IRBAfter.CreateAlignedLoad(
getShadowTy(&CB), getShadowPtrForRetval(IRBAfter),
kShadowTLSAlignment, "_msret");
setShadow(&CB, RetvalShadow);
if (MS.TrackOrigins)
setOrigin(&CB, IRBAfter.CreateLoad(MS.OriginTy,
getOriginPtrForRetval()));
}
bool isAMustTailRetVal(Value *RetVal) {
if (auto *I = dyn_cast<BitCastInst>(RetVal)) {
RetVal = I->getOperand(0);
}
if (auto *I = dyn_cast<CallInst>(RetVal)) {
return I->isMustTailCall();
}
return false;
}
void visitReturnInst(ReturnInst &I) {
IRBuilder<> IRB(&I);
Value *RetVal = I.getReturnValue();
if (!RetVal)
return;
// Don't emit the epilogue for musttail call returns.
if (isAMustTailRetVal(RetVal))
return;
Value *ShadowPtr = getShadowPtrForRetval(IRB);
bool HasNoUndef = F.hasRetAttribute(Attribute::NoUndef);
bool StoreShadow = !(MS.EagerChecks && HasNoUndef);
// FIXME: Consider using SpecialCaseList to specify a list of functions that
// must always return fully initialized values. For now, we hardcode "main".
bool EagerCheck = (MS.EagerChecks && HasNoUndef) || (F.getName() == "main");
Value *Shadow = getShadow(RetVal);
bool StoreOrigin = true;
if (EagerCheck) {
insertShadowCheck(RetVal, &I);
Shadow = getCleanShadow(RetVal);
StoreOrigin = false;
}
// The caller may still expect information passed over TLS if we pass our
// check
if (StoreShadow) {
IRB.CreateAlignedStore(Shadow, ShadowPtr, kShadowTLSAlignment);
if (MS.TrackOrigins && StoreOrigin)
IRB.CreateStore(getOrigin(RetVal), getOriginPtrForRetval());
}
}
void visitPHINode(PHINode &I) {
IRBuilder<> IRB(&I);
if (!PropagateShadow) {
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
return;
}
ShadowPHINodes.push_back(&I);
setShadow(&I, IRB.CreatePHI(getShadowTy(&I), I.getNumIncomingValues(),
"_msphi_s"));
if (MS.TrackOrigins)
setOrigin(
&I, IRB.CreatePHI(MS.OriginTy, I.getNumIncomingValues(), "_msphi_o"));
}
Value *getLocalVarIdptr(AllocaInst &I) {
ConstantInt *IntConst =
ConstantInt::get(Type::getInt32Ty((*F.getParent()).getContext()), 0);
return new GlobalVariable(*F.getParent(), IntConst->getType(),
/*isConstant=*/false, GlobalValue::PrivateLinkage,
IntConst);
}
Value *getLocalVarDescription(AllocaInst &I) {
return createPrivateConstGlobalForString(*F.getParent(), I.getName());
}
void poisonAllocaUserspace(AllocaInst &I, IRBuilder<> &IRB, Value *Len) {
if (PoisonStack && ClPoisonStackWithCall) {
IRB.CreateCall(MS.MsanPoisonStackFn, {&I, Len});
} else {
Value *ShadowBase, *OriginBase;
std::tie(ShadowBase, OriginBase) = getShadowOriginPtr(
&I, IRB, IRB.getInt8Ty(), Align(1), /*isStore*/ true);
Value *PoisonValue = IRB.getInt8(PoisonStack ? ClPoisonStackPattern : 0);
IRB.CreateMemSet(ShadowBase, PoisonValue, Len, I.getAlign());
}
if (PoisonStack && MS.TrackOrigins) {
Value *Idptr = getLocalVarIdptr(I);
if (ClPrintStackNames) {
Value *Descr = getLocalVarDescription(I);
IRB.CreateCall(MS.MsanSetAllocaOriginWithDescriptionFn,
{&I, Len, Idptr, Descr});
} else {
IRB.CreateCall(MS.MsanSetAllocaOriginNoDescriptionFn, {&I, Len, Idptr});
}
}
}
void poisonAllocaKmsan(AllocaInst &I, IRBuilder<> &IRB, Value *Len) {
Value *Descr = getLocalVarDescription(I);
if (PoisonStack) {
IRB.CreateCall(MS.MsanPoisonAllocaFn, {&I, Len, Descr});
} else {
IRB.CreateCall(MS.MsanUnpoisonAllocaFn, {&I, Len});
}
}
void instrumentAlloca(AllocaInst &I, Instruction *InsPoint = nullptr) {
if (!InsPoint)
InsPoint = &I;
NextNodeIRBuilder IRB(InsPoint);
const DataLayout &DL = F.getParent()->getDataLayout();
TypeSize TS = DL.getTypeAllocSize(I.getAllocatedType());
Value *Len = IRB.CreateTypeSize(MS.IntptrTy, TS);
if (I.isArrayAllocation())
Len = IRB.CreateMul(Len,
IRB.CreateZExtOrTrunc(I.getArraySize(), MS.IntptrTy));
if (MS.CompileKernel)
poisonAllocaKmsan(I, IRB, Len);
else
poisonAllocaUserspace(I, IRB, Len);
}
void visitAllocaInst(AllocaInst &I) {
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
// We'll get to this alloca later unless it's poisoned at the corresponding
// llvm.lifetime.start.
AllocaSet.insert(&I);
}
void visitSelectInst(SelectInst &I) {
// a = select b, c, d
Value *B = I.getCondition();
Value *C = I.getTrueValue();
Value *D = I.getFalseValue();
handleSelectLikeInst(I, B, C, D);
}
void handleSelectLikeInst(Instruction &I, Value *B, Value *C, Value *D) {
IRBuilder<> IRB(&I);
Value *Sb = getShadow(B);
Value *Sc = getShadow(C);
Value *Sd = getShadow(D);
Value *Ob = MS.TrackOrigins ? getOrigin(B) : nullptr;
Value *Oc = MS.TrackOrigins ? getOrigin(C) : nullptr;
Value *Od = MS.TrackOrigins ? getOrigin(D) : nullptr;
// Result shadow if condition shadow is 0.
Value *Sa0 = IRB.CreateSelect(B, Sc, Sd);
Value *Sa1;
if (I.getType()->isAggregateType()) {
// To avoid "sign extending" i1 to an arbitrary aggregate type, we just do
// an extra "select". This results in much more compact IR.
// Sa = select Sb, poisoned, (select b, Sc, Sd)
Sa1 = getPoisonedShadow(getShadowTy(I.getType()));
} else {
// Sa = select Sb, [ (c^d) | Sc | Sd ], [ b ? Sc : Sd ]
// If Sb (condition is poisoned), look for bits in c and d that are equal
// and both unpoisoned.
// If !Sb (condition is unpoisoned), simply pick one of Sc and Sd.
// Cast arguments to shadow-compatible type.
C = CreateAppToShadowCast(IRB, C);
D = CreateAppToShadowCast(IRB, D);
// Result shadow if condition shadow is 1.
Sa1 = IRB.CreateOr({IRB.CreateXor(C, D), Sc, Sd});
}
Value *Sa = IRB.CreateSelect(Sb, Sa1, Sa0, "_msprop_select");
setShadow(&I, Sa);
if (MS.TrackOrigins) {
// Origins are always i32, so any vector conditions must be flattened.
// FIXME: consider tracking vector origins for app vectors?
if (B->getType()->isVectorTy()) {
B = convertToBool(B, IRB);
Sb = convertToBool(Sb, IRB);
}
// a = select b, c, d
// Oa = Sb ? Ob : (b ? Oc : Od)
setOrigin(&I, IRB.CreateSelect(Sb, Ob, IRB.CreateSelect(B, Oc, Od)));
}
}
void visitLandingPadInst(LandingPadInst &I) {
// Do nothing.
// See https://github.com/google/sanitizers/issues/504
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
}
void visitCatchSwitchInst(CatchSwitchInst &I) {
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
}
void visitFuncletPadInst(FuncletPadInst &I) {
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
}
void visitGetElementPtrInst(GetElementPtrInst &I) { handleShadowOr(I); }
void visitExtractValueInst(ExtractValueInst &I) {
IRBuilder<> IRB(&I);
Value *Agg = I.getAggregateOperand();
LLVM_DEBUG(dbgs() << "ExtractValue: " << I << "\n");
Value *AggShadow = getShadow(Agg);
LLVM_DEBUG(dbgs() << " AggShadow: " << *AggShadow << "\n");
Value *ResShadow = IRB.CreateExtractValue(AggShadow, I.getIndices());
LLVM_DEBUG(dbgs() << " ResShadow: " << *ResShadow << "\n");
setShadow(&I, ResShadow);
setOriginForNaryOp(I);
}
void visitInsertValueInst(InsertValueInst &I) {
IRBuilder<> IRB(&I);
LLVM_DEBUG(dbgs() << "InsertValue: " << I << "\n");
Value *AggShadow = getShadow(I.getAggregateOperand());
Value *InsShadow = getShadow(I.getInsertedValueOperand());
LLVM_DEBUG(dbgs() << " AggShadow: " << *AggShadow << "\n");
LLVM_DEBUG(dbgs() << " InsShadow: " << *InsShadow << "\n");
Value *Res = IRB.CreateInsertValue(AggShadow, InsShadow, I.getIndices());
LLVM_DEBUG(dbgs() << " Res: " << *Res << "\n");
setShadow(&I, Res);
setOriginForNaryOp(I);
}
void dumpInst(Instruction &I) {
if (CallInst *CI = dyn_cast<CallInst>(&I)) {
errs() << "ZZZ call " << CI->getCalledFunction()->getName() << "\n";
} else {
errs() << "ZZZ " << I.getOpcodeName() << "\n";
}
errs() << "QQQ " << I << "\n";
}
void visitResumeInst(ResumeInst &I) {
LLVM_DEBUG(dbgs() << "Resume: " << I << "\n");
// Nothing to do here.
}
void visitCleanupReturnInst(CleanupReturnInst &CRI) {
LLVM_DEBUG(dbgs() << "CleanupReturn: " << CRI << "\n");
// Nothing to do here.
}
void visitCatchReturnInst(CatchReturnInst &CRI) {
LLVM_DEBUG(dbgs() << "CatchReturn: " << CRI << "\n");
// Nothing to do here.
}
void instrumentAsmArgument(Value *Operand, Type *ElemTy, Instruction &I,
IRBuilder<> &IRB, const DataLayout &DL,
bool isOutput) {
// For each assembly argument, we check its value for being initialized.
// If the argument is a pointer, we assume it points to a single element
// of the corresponding type (or to a 8-byte word, if the type is unsized).
// Each such pointer is instrumented with a call to the runtime library.
Type *OpType = Operand->getType();
// Check the operand value itself.
insertShadowCheck(Operand, &I);
if (!OpType->isPointerTy() || !isOutput) {
assert(!isOutput);
return;
}
if (!ElemTy->isSized())
return;
auto Size = DL.getTypeStoreSize(ElemTy);
Value *SizeVal = IRB.CreateTypeSize(MS.IntptrTy, Size);
if (MS.CompileKernel) {
IRB.CreateCall(MS.MsanInstrumentAsmStoreFn, {Operand, SizeVal});
} else {
// ElemTy, derived from elementtype(), does not encode the alignment of
// the pointer. Conservatively assume that the shadow memory is unaligned.
// When Size is large, avoid StoreInst as it would expand to many
// instructions.
auto [ShadowPtr, _] =
getShadowOriginPtrUserspace(Operand, IRB, IRB.getInt8Ty(), Align(1));
if (Size <= 32)
IRB.CreateAlignedStore(getCleanShadow(ElemTy), ShadowPtr, Align(1));
else
IRB.CreateMemSet(ShadowPtr, ConstantInt::getNullValue(IRB.getInt8Ty()),
SizeVal, Align(1));
}
}
/// Get the number of output arguments returned by pointers.
int getNumOutputArgs(InlineAsm *IA, CallBase *CB) {
int NumRetOutputs = 0;
int NumOutputs = 0;
Type *RetTy = cast<Value>(CB)->getType();
if (!RetTy->isVoidTy()) {
// Register outputs are returned via the CallInst return value.
auto *ST = dyn_cast<StructType>(RetTy);
if (ST)
NumRetOutputs = ST->getNumElements();
else
NumRetOutputs = 1;
}
InlineAsm::ConstraintInfoVector Constraints = IA->ParseConstraints();
for (const InlineAsm::ConstraintInfo &Info : Constraints) {
switch (Info.Type) {
case InlineAsm::isOutput:
NumOutputs++;
break;
default:
break;
}
}
return NumOutputs - NumRetOutputs;
}
void visitAsmInstruction(Instruction &I) {
// Conservative inline assembly handling: check for poisoned shadow of
// asm() arguments, then unpoison the result and all the memory locations
// pointed to by those arguments.
// An inline asm() statement in C++ contains lists of input and output
// arguments used by the assembly code. These are mapped to operands of the
// CallInst as follows:
// - nR register outputs ("=r) are returned by value in a single structure
// (SSA value of the CallInst);
// - nO other outputs ("=m" and others) are returned by pointer as first
// nO operands of the CallInst;
// - nI inputs ("r", "m" and others) are passed to CallInst as the
// remaining nI operands.
// The total number of asm() arguments in the source is nR+nO+nI, and the
// corresponding CallInst has nO+nI+1 operands (the last operand is the
// function to be called).
const DataLayout &DL = F.getParent()->getDataLayout();
CallBase *CB = cast<CallBase>(&I);
IRBuilder<> IRB(&I);
InlineAsm *IA = cast<InlineAsm>(CB->getCalledOperand());
int OutputArgs = getNumOutputArgs(IA, CB);
// The last operand of a CallInst is the function itself.
int NumOperands = CB->getNumOperands() - 1;
// Check input arguments. Doing so before unpoisoning output arguments, so
// that we won't overwrite uninit values before checking them.
for (int i = OutputArgs; i < NumOperands; i++) {
Value *Operand = CB->getOperand(i);
instrumentAsmArgument(Operand, CB->getParamElementType(i), I, IRB, DL,
/*isOutput*/ false);
}
// Unpoison output arguments. This must happen before the actual InlineAsm
// call, so that the shadow for memory published in the asm() statement
// remains valid.
for (int i = 0; i < OutputArgs; i++) {
Value *Operand = CB->getOperand(i);
instrumentAsmArgument(Operand, CB->getParamElementType(i), I, IRB, DL,
/*isOutput*/ true);
}
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
}
void visitFreezeInst(FreezeInst &I) {
// Freeze always returns a fully defined value.
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
}
void visitInstruction(Instruction &I) {
// Everything else: stop propagating and check for poisoned shadow.
if (ClDumpStrictInstructions)
dumpInst(I);
LLVM_DEBUG(dbgs() << "DEFAULT: " << I << "\n");
for (size_t i = 0, n = I.getNumOperands(); i < n; i++) {
Value *Operand = I.getOperand(i);
if (Operand->getType()->isSized())
insertShadowCheck(Operand, &I);
}
setShadow(&I, getCleanShadow(&I));
setOrigin(&I, getCleanOrigin());
}
};
struct VarArgHelperBase : public VarArgHelper {
Function &F;
MemorySanitizer &MS;
MemorySanitizerVisitor &MSV;
SmallVector<CallInst *, 16> VAStartInstrumentationList;
const unsigned VAListTagSize;
VarArgHelperBase(Function &F, MemorySanitizer &MS,
MemorySanitizerVisitor &MSV, unsigned VAListTagSize)
: F(F), MS(MS), MSV(MSV), VAListTagSize(VAListTagSize) {}
Value *getShadowAddrForVAArgument(IRBuilder<> &IRB, unsigned ArgOffset) {
Value *Base = IRB.CreatePointerCast(MS.VAArgTLS, MS.IntptrTy);
return IRB.CreateAdd(Base, ConstantInt::get(MS.IntptrTy, ArgOffset));
}
/// Compute the shadow address for a given va_arg.
Value *getShadowPtrForVAArgument(Type *Ty, IRBuilder<> &IRB,
unsigned ArgOffset) {
Value *Base = IRB.CreatePointerCast(MS.VAArgTLS, MS.IntptrTy);
Base = IRB.CreateAdd(Base, ConstantInt::get(MS.IntptrTy, ArgOffset));
return IRB.CreateIntToPtr(Base, PointerType::get(MSV.getShadowTy(Ty), 0),
"_msarg_va_s");
}
/// Compute the shadow address for a given va_arg.
Value *getShadowPtrForVAArgument(Type *Ty, IRBuilder<> &IRB,
unsigned ArgOffset, unsigned ArgSize) {
// Make sure we don't overflow __msan_va_arg_tls.
if (ArgOffset + ArgSize > kParamTLSSize)
return nullptr;
return getShadowPtrForVAArgument(Ty, IRB, ArgOffset);
}
/// Compute the origin address for a given va_arg.
Value *getOriginPtrForVAArgument(IRBuilder<> &IRB, int ArgOffset) {
Value *Base = IRB.CreatePointerCast(MS.VAArgOriginTLS, MS.IntptrTy);
// getOriginPtrForVAArgument() is always called after
// getShadowPtrForVAArgument(), so __msan_va_arg_origin_tls can never
// overflow.
Base = IRB.CreateAdd(Base, ConstantInt::get(MS.IntptrTy, ArgOffset));
return IRB.CreateIntToPtr(Base, PointerType::get(MS.OriginTy, 0),
"_msarg_va_o");
}
void CleanUnusedTLS(IRBuilder<> &IRB, Value *ShadowBase,
unsigned BaseOffset) {
// The tails of __msan_va_arg_tls is not large enough to fit full
// value shadow, but it will be copied to backup anyway. Make it
// clean.
if (BaseOffset >= kParamTLSSize)
return;
Value *TailSize =
ConstantInt::getSigned(IRB.getInt32Ty(), kParamTLSSize - BaseOffset);
IRB.CreateMemSet(ShadowBase, ConstantInt::getNullValue(IRB.getInt8Ty()),
TailSize, Align(8));
}
void unpoisonVAListTagForInst(IntrinsicInst &I) {
IRBuilder<> IRB(&I);
Value *VAListTag = I.getArgOperand(0);
const Align Alignment = Align(8);
auto [ShadowPtr, OriginPtr] = MSV.getShadowOriginPtr(
VAListTag, IRB, IRB.getInt8Ty(), Alignment, /*isStore*/ true);
// Unpoison the whole __va_list_tag.
IRB.CreateMemSet(ShadowPtr, Constant::getNullValue(IRB.getInt8Ty()),
VAListTagSize, Alignment, false);
}
void visitVAStartInst(VAStartInst &I) override {
if (F.getCallingConv() == CallingConv::Win64)
return;
VAStartInstrumentationList.push_back(&I);
unpoisonVAListTagForInst(I);
}
void visitVACopyInst(VACopyInst &I) override {
if (F.getCallingConv() == CallingConv::Win64)
return;
unpoisonVAListTagForInst(I);
}
};
/// AMD64-specific implementation of VarArgHelper.
struct VarArgAMD64Helper : public VarArgHelperBase {
// An unfortunate workaround for asymmetric lowering of va_arg stuff.
// See a comment in visitCallBase for more details.
static const unsigned AMD64GpEndOffset = 48; // AMD64 ABI Draft 0.99.6 p3.5.7
static const unsigned AMD64FpEndOffsetSSE = 176;
// If SSE is disabled, fp_offset in va_list is zero.
static const unsigned AMD64FpEndOffsetNoSSE = AMD64GpEndOffset;
unsigned AMD64FpEndOffset;
AllocaInst *VAArgTLSCopy = nullptr;
AllocaInst *VAArgTLSOriginCopy = nullptr;
Value *VAArgOverflowSize = nullptr;
enum ArgKind { AK_GeneralPurpose, AK_FloatingPoint, AK_Memory };
VarArgAMD64Helper(Function &F, MemorySanitizer &MS,
MemorySanitizerVisitor &MSV)
: VarArgHelperBase(F, MS, MSV, /*VAListTagSize=*/24) {
AMD64FpEndOffset = AMD64FpEndOffsetSSE;
for (const auto &Attr : F.getAttributes().getFnAttrs()) {
if (Attr.isStringAttribute() &&
(Attr.getKindAsString() == "target-features")) {
if (Attr.getValueAsString().contains("-sse"))
AMD64FpEndOffset = AMD64FpEndOffsetNoSSE;
break;
}
}
}
ArgKind classifyArgument(Value *arg) {
// A very rough approximation of X86_64 argument classification rules.
Type *T = arg->getType();
if (T->isX86_FP80Ty())
return AK_Memory;
if (T->isFPOrFPVectorTy() || T->isX86_MMXTy())
return AK_FloatingPoint;
if (T->isIntegerTy() && T->getPrimitiveSizeInBits() <= 64)
return AK_GeneralPurpose;
if (T->isPointerTy())
return AK_GeneralPurpose;
return AK_Memory;
}
// For VarArg functions, store the argument shadow in an ABI-specific format
// that corresponds to va_list layout.
// We do this because Clang lowers va_arg in the frontend, and this pass
// only sees the low level code that deals with va_list internals.
// A much easier alternative (provided that Clang emits va_arg instructions)
// would have been to associate each live instance of va_list with a copy of
// MSanParamTLS, and extract shadow on va_arg() call in the argument list
// order.
void visitCallBase(CallBase &CB, IRBuilder<> &IRB) override {
unsigned GpOffset = 0;
unsigned FpOffset = AMD64GpEndOffset;
unsigned OverflowOffset = AMD64FpEndOffset;
const DataLayout &DL = F.getParent()->getDataLayout();
for (const auto &[ArgNo, A] : llvm::enumerate(CB.args())) {
bool IsFixed = ArgNo < CB.getFunctionType()->getNumParams();
bool IsByVal = CB.paramHasAttr(ArgNo, Attribute::ByVal);
if (IsByVal) {
// ByVal arguments always go to the overflow area.
// Fixed arguments passed through the overflow area will be stepped
// over by va_start, so don't count them towards the offset.
if (IsFixed)
continue;
assert(A->getType()->isPointerTy());
Type *RealTy = CB.getParamByValType(ArgNo);
uint64_t ArgSize = DL.getTypeAllocSize(RealTy);
uint64_t AlignedSize = alignTo(ArgSize, 8);
unsigned BaseOffset = OverflowOffset;
Value *ShadowBase =
getShadowPtrForVAArgument(RealTy, IRB, OverflowOffset);
Value *OriginBase = nullptr;
if (MS.TrackOrigins)
OriginBase = getOriginPtrForVAArgument(IRB, OverflowOffset);
OverflowOffset += AlignedSize;
if (OverflowOffset > kParamTLSSize) {
CleanUnusedTLS(IRB, ShadowBase, BaseOffset);
continue; // We have no space to copy shadow there.
}
Value *ShadowPtr, *OriginPtr;
std::tie(ShadowPtr, OriginPtr) =
MSV.getShadowOriginPtr(A, IRB, IRB.getInt8Ty(), kShadowTLSAlignment,
/*isStore*/ false);
IRB.CreateMemCpy(ShadowBase, kShadowTLSAlignment, ShadowPtr,
kShadowTLSAlignment, ArgSize);
if (MS.TrackOrigins)
IRB.CreateMemCpy(OriginBase, kShadowTLSAlignment, OriginPtr,
kShadowTLSAlignment, ArgSize);
} else {
ArgKind AK = classifyArgument(A);
if (AK == AK_GeneralPurpose && GpOffset >= AMD64GpEndOffset)
AK = AK_Memory;
if (AK == AK_FloatingPoint && FpOffset >= AMD64FpEndOffset)
AK = AK_Memory;
Value *ShadowBase, *OriginBase = nullptr;
switch (AK) {
case AK_GeneralPurpose:
ShadowBase = getShadowPtrForVAArgument(A->getType(), IRB, GpOffset);
if (MS.TrackOrigins)
OriginBase = getOriginPtrForVAArgument(IRB, GpOffset);
GpOffset += 8;
assert(GpOffset <= kParamTLSSize);
break;
case AK_FloatingPoint:
ShadowBase = getShadowPtrForVAArgument(A->getType(), IRB, FpOffset);
if (MS.TrackOrigins)
OriginBase = getOriginPtrForVAArgument(IRB, FpOffset);
FpOffset += 16;
assert(FpOffset <= kParamTLSSize);
break;
case AK_Memory:
if (IsFixed)
continue;
uint64_t ArgSize = DL.getTypeAllocSize(A->getType());
uint64_t AlignedSize = alignTo(ArgSize, 8);
unsigned BaseOffset = OverflowOffset;
ShadowBase =
getShadowPtrForVAArgument(A->getType(), IRB, OverflowOffset);
if (MS.TrackOrigins) {
OriginBase = getOriginPtrForVAArgument(IRB, OverflowOffset);
}
OverflowOffset += AlignedSize;
if (OverflowOffset > kParamTLSSize) {
// We have no space to copy shadow there.
CleanUnusedTLS(IRB, ShadowBase, BaseOffset);
continue;
}
}
// Take fixed arguments into account for GpOffset and FpOffset,
// but don't actually store shadows for them.
// TODO(glider): don't call get*PtrForVAArgument() for them.
if (IsFixed)
continue;
Value *Shadow = MSV.getShadow(A);
IRB.CreateAlignedStore(Shadow, ShadowBase, kShadowTLSAlignment);
if (MS.TrackOrigins) {
Value *Origin = MSV.getOrigin(A);
TypeSize StoreSize = DL.getTypeStoreSize(Shadow->getType());
MSV.paintOrigin(IRB, Origin, OriginBase, StoreSize,
std::max(kShadowTLSAlignment, kMinOriginAlignment));
}
}
}
Constant *OverflowSize =
ConstantInt::get(IRB.getInt64Ty(), OverflowOffset - AMD64FpEndOffset);
IRB.CreateStore(OverflowSize, MS.VAArgOverflowSizeTLS);
}
void finalizeInstrumentation() override {
assert(!VAArgOverflowSize && !VAArgTLSCopy &&
"finalizeInstrumentation called twice");
if (!VAStartInstrumentationList.empty()) {
// If there is a va_start in this function, make a backup copy of
// va_arg_tls somewhere in the function entry block.
IRBuilder<> IRB(MSV.FnPrologueEnd);
VAArgOverflowSize =
IRB.CreateLoad(IRB.getInt64Ty(), MS.VAArgOverflowSizeTLS);
Value *CopySize = IRB.CreateAdd(
ConstantInt::get(MS.IntptrTy, AMD64FpEndOffset), VAArgOverflowSize);
VAArgTLSCopy = IRB.CreateAlloca(Type::getInt8Ty(*MS.C), CopySize);
VAArgTLSCopy->setAlignment(kShadowTLSAlignment);
IRB.CreateMemSet(VAArgTLSCopy, Constant::getNullValue(IRB.getInt8Ty()),
CopySize, kShadowTLSAlignment, false);
Value *SrcSize = IRB.CreateBinaryIntrinsic(
Intrinsic::umin, CopySize,
ConstantInt::get(MS.IntptrTy, kParamTLSSize));
IRB.CreateMemCpy(VAArgTLSCopy, kShadowTLSAlignment, MS.VAArgTLS,
kShadowTLSAlignment, SrcSize);
if (MS.TrackOrigins) {
VAArgTLSOriginCopy = IRB.CreateAlloca(Type::getInt8Ty(*MS.C), CopySize);
VAArgTLSOriginCopy->setAlignment(kShadowTLSAlignment);
IRB.CreateMemCpy(VAArgTLSOriginCopy, kShadowTLSAlignment,
MS.VAArgOriginTLS, kShadowTLSAlignment, SrcSize);
}
}
// Instrument va_start.
// Copy va_list shadow from the backup copy of the TLS contents.
for (size_t i = 0, n = VAStartInstrumentationList.size(); i < n; i++) {
CallInst *OrigInst = VAStartInstrumentationList[i];
NextNodeIRBuilder IRB(OrigInst);
Value *VAListTag = OrigInst->getArgOperand(0);
Type *RegSaveAreaPtrTy = PointerType::getUnqual(*MS.C); // i64*
Value *RegSaveAreaPtrPtr = IRB.CreateIntToPtr(
IRB.CreateAdd(IRB.CreatePtrToInt(VAListTag, MS.IntptrTy),
ConstantInt::get(MS.IntptrTy, 16)),
PointerType::get(RegSaveAreaPtrTy, 0));
Value *RegSaveAreaPtr =
IRB.CreateLoad(RegSaveAreaPtrTy, RegSaveAreaPtrPtr);
Value *RegSaveAreaShadowPtr, *RegSaveAreaOriginPtr;
const Align Alignment = Align(16);
std::tie(RegSaveAreaShadowPtr, RegSaveAreaOriginPtr) =
MSV.getShadowOriginPtr(RegSaveAreaPtr, IRB, IRB.getInt8Ty(),
Alignment, /*isStore*/ true);
IRB.CreateMemCpy(RegSaveAreaShadowPtr, Alignment, VAArgTLSCopy, Alignment,
AMD64FpEndOffset);
if (MS.TrackOrigins)
IRB.CreateMemCpy(RegSaveAreaOriginPtr, Alignment, VAArgTLSOriginCopy,
Alignment, AMD64FpEndOffset);
Type *OverflowArgAreaPtrTy = PointerType::getUnqual(*MS.C); // i64*
Value *OverflowArgAreaPtrPtr = IRB.CreateIntToPtr(
IRB.CreateAdd(IRB.CreatePtrToInt(VAListTag, MS.IntptrTy),
ConstantInt::get(MS.IntptrTy, 8)),
PointerType::get(OverflowArgAreaPtrTy, 0));
Value *OverflowArgAreaPtr =
IRB.CreateLoad(OverflowArgAreaPtrTy, OverflowArgAreaPtrPtr);
Value *OverflowArgAreaShadowPtr, *OverflowArgAreaOriginPtr;
std::tie(OverflowArgAreaShadowPtr, OverflowArgAreaOriginPtr) =
MSV.getShadowOriginPtr(OverflowArgAreaPtr, IRB, IRB.getInt8Ty(),
Alignment, /*isStore*/ true);
Value *SrcPtr = IRB.CreateConstGEP1_32(IRB.getInt8Ty(), VAArgTLSCopy,
AMD64FpEndOffset);
IRB.CreateMemCpy(OverflowArgAreaShadowPtr, Alignment, SrcPtr, Alignment,
VAArgOverflowSize);
if (MS.TrackOrigins) {
SrcPtr = IRB.CreateConstGEP1_32(IRB.getInt8Ty(), VAArgTLSOriginCopy,
AMD64FpEndOffset);
IRB.CreateMemCpy(OverflowArgAreaOriginPtr, Alignment, SrcPtr, Alignment,
VAArgOverflowSize);
}
}
}
};
/// MIPS64-specific implementation of VarArgHelper.
/// NOTE: This is also used for LoongArch64.
struct VarArgMIPS64Helper : public VarArgHelperBase {
AllocaInst *VAArgTLSCopy = nullptr;
Value *VAArgSize = nullptr;
VarArgMIPS64Helper(Function &F, MemorySanitizer &MS,
MemorySanitizerVisitor &MSV)
: VarArgHelperBase(F, MS, MSV, /*VAListTagSize=*/8) {}
void visitCallBase(CallBase &CB, IRBuilder<> &IRB) override {
unsigned VAArgOffset = 0;
const DataLayout &DL = F.getParent()->getDataLayout();
for (Value *A :
llvm::drop_begin(CB.args(), CB.getFunctionType()->getNumParams())) {
Triple TargetTriple(F.getParent()->getTargetTriple());
Value *Base;
uint64_t ArgSize = DL.getTypeAllocSize(A->getType());
if (TargetTriple.getArch() == Triple::mips64) {
// Adjusting the shadow for argument with size < 8 to match the
// placement of bits in big endian system
if (ArgSize < 8)
VAArgOffset += (8 - ArgSize);
}
Base = getShadowPtrForVAArgument(A->getType(), IRB, VAArgOffset, ArgSize);
VAArgOffset += ArgSize;
VAArgOffset = alignTo(VAArgOffset, 8);
if (!Base)
continue;
IRB.CreateAlignedStore(MSV.getShadow(A), Base, kShadowTLSAlignment);
}
Constant *TotalVAArgSize = ConstantInt::get(IRB.getInt64Ty(), VAArgOffset);
// Here using VAArgOverflowSizeTLS as VAArgSizeTLS to avoid creation of
// a new class member i.e. it is the total size of all VarArgs.
IRB.CreateStore(TotalVAArgSize, MS.VAArgOverflowSizeTLS);
}
void finalizeInstrumentation() override {
assert(!VAArgSize && !VAArgTLSCopy &&
"finalizeInstrumentation called twice");
IRBuilder<> IRB(MSV.FnPrologueEnd);
VAArgSize = IRB.CreateLoad(IRB.getInt64Ty(), MS.VAArgOverflowSizeTLS);
Value *CopySize =
IRB.CreateAdd(ConstantInt::get(MS.IntptrTy, 0), VAArgSize);
if (!VAStartInstrumentationList.empty()) {
// If there is a va_start in this function, make a backup copy of
// va_arg_tls somewhere in the function entry block.
VAArgTLSCopy = IRB.CreateAlloca(Type::getInt8Ty(*MS.C), CopySize);
VAArgTLSCopy->setAlignment(kShadowTLSAlignment);
IRB.CreateMemSet(VAArgTLSCopy, Constant::getNullValue(IRB.getInt8Ty()),
CopySize, kShadowTLSAlignment, false);
Value *SrcSize = IRB.CreateBinaryIntrinsic(
Intrinsic::umin, CopySize,
ConstantInt::get(MS.IntptrTy, kParamTLSSize));
IRB.CreateMemCpy(VAArgTLSCopy, kShadowTLSAlignment, MS.VAArgTLS,
kShadowTLSAlignment, SrcSize);
}
// Instrument va_start.
// Copy va_list shadow from the backup copy of the TLS contents.
for (size_t i = 0, n = VAStartInstrumentationList.size(); i < n; i++) {
CallInst *OrigInst = VAStartInstrumentationList[i];
NextNodeIRBuilder IRB(OrigInst);
Value *VAListTag = OrigInst->getArgOperand(0);
Type *RegSaveAreaPtrTy = PointerType::getUnqual(*MS.C); // i64*
Value *RegSaveAreaPtrPtr =
IRB.CreateIntToPtr(IRB.CreatePtrToInt(VAListTag, MS.IntptrTy),
PointerType::get(RegSaveAreaPtrTy, 0));
Value *RegSaveAreaPtr =
IRB.CreateLoad(RegSaveAreaPtrTy, RegSaveAreaPtrPtr);
Value *RegSaveAreaShadowPtr, *RegSaveAreaOriginPtr;
const Align Alignment = Align(8);
std::tie(RegSaveAreaShadowPtr, RegSaveAreaOriginPtr) =
MSV.getShadowOriginPtr(RegSaveAreaPtr, IRB, IRB.getInt8Ty(),
Alignment, /*isStore*/ true);
IRB.CreateMemCpy(RegSaveAreaShadowPtr, Alignment, VAArgTLSCopy, Alignment,
CopySize);
}
}
};
/// AArch64-specific implementation of VarArgHelper.
struct VarArgAArch64Helper : public VarArgHelperBase {
static const unsigned kAArch64GrArgSize = 64;
static const unsigned kAArch64VrArgSize = 128;
static const unsigned AArch64GrBegOffset = 0;
static const unsigned AArch64GrEndOffset = kAArch64GrArgSize;
// Make VR space aligned to 16 bytes.
static const unsigned AArch64VrBegOffset = AArch64GrEndOffset;
static const unsigned AArch64VrEndOffset =
AArch64VrBegOffset + kAArch64VrArgSize;
static const unsigned AArch64VAEndOffset = AArch64VrEndOffset;
AllocaInst *VAArgTLSCopy = nullptr;
Value *VAArgOverflowSize = nullptr;
enum ArgKind { AK_GeneralPurpose, AK_FloatingPoint, AK_Memory };
VarArgAArch64Helper(Function &F, MemorySanitizer &MS,
MemorySanitizerVisitor &MSV)
: VarArgHelperBase(F, MS, MSV, /*VAListTagSize=*/32) {}
// A very rough approximation of aarch64 argument classification rules.
std::pair<ArgKind, uint64_t> classifyArgument(Type *T) {
if (T->isIntOrPtrTy() && T->getPrimitiveSizeInBits() <= 64)
return {AK_GeneralPurpose, 1};
if (T->isFloatingPointTy() && T->getPrimitiveSizeInBits() <= 128)
return {AK_FloatingPoint, 1};
if (T->isArrayTy()) {
auto R = classifyArgument(T->getArrayElementType());
R.second *= T->getScalarType()->getArrayNumElements();
return R;
}
if (const FixedVectorType *FV = dyn_cast<FixedVectorType>(T)) {
auto R = classifyArgument(FV->getScalarType());
R.second *= FV->getNumElements();
return R;
}
LLVM_DEBUG(errs() << "Unknown vararg type: " << *T << "\n");
return {AK_Memory, 0};
}
// The instrumentation stores the argument shadow in a non ABI-specific
// format because it does not know which argument is named (since Clang,
// like x86_64 case, lowers the va_args in the frontend and this pass only
// sees the low level code that deals with va_list internals).
// The first seven GR registers are saved in the first 56 bytes of the
// va_arg tls arra, followed by the first 8 FP/SIMD registers, and then
// the remaining arguments.
// Using constant offset within the va_arg TLS array allows fast copy
// in the finalize instrumentation.
void visitCallBase(CallBase &CB, IRBuilder<> &IRB) override {
unsigned GrOffset = AArch64GrBegOffset;
unsigned VrOffset = AArch64VrBegOffset;
unsigned OverflowOffset = AArch64VAEndOffset;
const DataLayout &DL = F.getParent()->getDataLayout();
for (const auto &[ArgNo, A] : llvm::enumerate(CB.args())) {
bool IsFixed = ArgNo < CB.getFunctionType()->getNumParams();
auto [AK, RegNum] = classifyArgument(A->getType());
if (AK == AK_GeneralPurpose &&
(GrOffset + RegNum * 8) > AArch64GrEndOffset)
AK = AK_Memory;
if (AK == AK_FloatingPoint &&
(VrOffset + RegNum * 16) > AArch64VrEndOffset)
AK = AK_Memory;
Value *Base;
switch (AK) {
case AK_GeneralPurpose:
Base = getShadowPtrForVAArgument(A->getType(), IRB, GrOffset);
GrOffset += 8 * RegNum;
break;
case AK_FloatingPoint:
Base = getShadowPtrForVAArgument(A->getType(), IRB, VrOffset);
VrOffset += 16 * RegNum;
break;
case AK_Memory:
// Don't count fixed arguments in the overflow area - va_start will
// skip right over them.
if (IsFixed)
continue;
uint64_t ArgSize = DL.getTypeAllocSize(A->getType());
uint64_t AlignedSize = alignTo(ArgSize, 8);
unsigned BaseOffset = OverflowOffset;
Base = getShadowPtrForVAArgument(A->getType(), IRB, BaseOffset);
OverflowOffset += AlignedSize;
if (OverflowOffset > kParamTLSSize) {
// We have no space to copy shadow there.
CleanUnusedTLS(IRB, Base, BaseOffset);
continue;
}
break;
}
// Count Gp/Vr fixed arguments to their respective offsets, but don't
// bother to actually store a shadow.
if (IsFixed)
continue;
IRB.CreateAlignedStore(MSV.getShadow(A), Base, kShadowTLSAlignment);
}
Constant *OverflowSize =
ConstantInt::get(IRB.getInt64Ty(), OverflowOffset - AArch64VAEndOffset);
IRB.CreateStore(OverflowSize, MS.VAArgOverflowSizeTLS);
}
// Retrieve a va_list field of 'void*' size.
Value *getVAField64(IRBuilder<> &IRB, Value *VAListTag, int offset) {
Value *SaveAreaPtrPtr = IRB.CreateIntToPtr(
IRB.CreateAdd(IRB.CreatePtrToInt(VAListTag, MS.IntptrTy),
ConstantInt::get(MS.IntptrTy, offset)),
PointerType::get(*MS.C, 0));
return IRB.CreateLoad(Type::getInt64Ty(*MS.C), SaveAreaPtrPtr);
}
// Retrieve a va_list field of 'int' size.
Value *getVAField32(IRBuilder<> &IRB, Value *VAListTag, int offset) {
Value *SaveAreaPtr = IRB.CreateIntToPtr(
IRB.CreateAdd(IRB.CreatePtrToInt(VAListTag, MS.IntptrTy),
ConstantInt::get(MS.IntptrTy, offset)),
PointerType::get(*MS.C, 0));
Value *SaveArea32 = IRB.CreateLoad(IRB.getInt32Ty(), SaveAreaPtr);
return IRB.CreateSExt(SaveArea32, MS.IntptrTy);
}
void finalizeInstrumentation() override {
assert(!VAArgOverflowSize && !VAArgTLSCopy &&
"finalizeInstrumentation called twice");
if (!VAStartInstrumentationList.empty()) {
// If there is a va_start in this function, make a backup copy of
// va_arg_tls somewhere in the function entry block.
IRBuilder<> IRB(MSV.FnPrologueEnd);
VAArgOverflowSize =
IRB.CreateLoad(IRB.getInt64Ty(), MS.VAArgOverflowSizeTLS);
Value *CopySize = IRB.CreateAdd(
ConstantInt::get(MS.IntptrTy, AArch64VAEndOffset), VAArgOverflowSize);
VAArgTLSCopy = IRB.CreateAlloca(Type::getInt8Ty(*MS.C), CopySize);
VAArgTLSCopy->setAlignment(kShadowTLSAlignment);
IRB.CreateMemSet(VAArgTLSCopy, Constant::getNullValue(IRB.getInt8Ty()),
CopySize, kShadowTLSAlignment, false);
Value *SrcSize = IRB.CreateBinaryIntrinsic(
Intrinsic::umin, CopySize,
ConstantInt::get(MS.IntptrTy, kParamTLSSize));
IRB.CreateMemCpy(VAArgTLSCopy, kShadowTLSAlignment, MS.VAArgTLS,
kShadowTLSAlignment, SrcSize);
}
Value *GrArgSize = ConstantInt::get(MS.IntptrTy, kAArch64GrArgSize);
Value *VrArgSize = ConstantInt::get(MS.IntptrTy, kAArch64VrArgSize);
// Instrument va_start, copy va_list shadow from the backup copy of
// the TLS contents.
for (size_t i = 0, n = VAStartInstrumentationList.size(); i < n; i++) {
CallInst *OrigInst = VAStartInstrumentationList[i];
NextNodeIRBuilder IRB(OrigInst);
Value *VAListTag = OrigInst->getArgOperand(0);
// The variadic ABI for AArch64 creates two areas to save the incoming
// argument registers (one for 64-bit general register xn-x7 and another
// for 128-bit FP/SIMD vn-v7).
// We need then to propagate the shadow arguments on both regions
// 'va::__gr_top + va::__gr_offs' and 'va::__vr_top + va::__vr_offs'.
// The remaining arguments are saved on shadow for 'va::stack'.
// One caveat is it requires only to propagate the non-named arguments,
// however on the call site instrumentation 'all' the arguments are
// saved. So to copy the shadow values from the va_arg TLS array
// we need to adjust the offset for both GR and VR fields based on
// the __{gr,vr}_offs value (since they are stores based on incoming
// named arguments).
Type *RegSaveAreaPtrTy = IRB.getPtrTy();
// Read the stack pointer from the va_list.
Value *StackSaveAreaPtr =
IRB.CreateIntToPtr(getVAField64(IRB, VAListTag, 0), RegSaveAreaPtrTy);
// Read both the __gr_top and __gr_off and add them up.
Value *GrTopSaveAreaPtr = getVAField64(IRB, VAListTag, 8);
Value *GrOffSaveArea = getVAField32(IRB, VAListTag, 24);
Value *GrRegSaveAreaPtr = IRB.CreateIntToPtr(
IRB.CreateAdd(GrTopSaveAreaPtr, GrOffSaveArea), RegSaveAreaPtrTy);
// Read both the __vr_top and __vr_off and add them up.
Value *VrTopSaveAreaPtr = getVAField64(IRB, VAListTag, 16);
Value *VrOffSaveArea = getVAField32(IRB, VAListTag, 28);
Value *VrRegSaveAreaPtr = IRB.CreateIntToPtr(
IRB.CreateAdd(VrTopSaveAreaPtr, VrOffSaveArea), RegSaveAreaPtrTy);
// It does not know how many named arguments is being used and, on the
// callsite all the arguments were saved. Since __gr_off is defined as
// '0 - ((8 - named_gr) * 8)', the idea is to just propagate the variadic
// argument by ignoring the bytes of shadow from named arguments.
Value *GrRegSaveAreaShadowPtrOff =
IRB.CreateAdd(GrArgSize, GrOffSaveArea);
Value *GrRegSaveAreaShadowPtr =
MSV.getShadowOriginPtr(GrRegSaveAreaPtr, IRB, IRB.getInt8Ty(),
Align(8), /*isStore*/ true)
.first;
Value *GrSrcPtr =
IRB.CreateInBoundsPtrAdd(VAArgTLSCopy, GrRegSaveAreaShadowPtrOff);
Value *GrCopySize = IRB.CreateSub(GrArgSize, GrRegSaveAreaShadowPtrOff);
IRB.CreateMemCpy(GrRegSaveAreaShadowPtr, Align(8), GrSrcPtr, Align(8),
GrCopySize);
// Again, but for FP/SIMD values.
Value *VrRegSaveAreaShadowPtrOff =
IRB.CreateAdd(VrArgSize, VrOffSaveArea);
Value *VrRegSaveAreaShadowPtr =
MSV.getShadowOriginPtr(VrRegSaveAreaPtr, IRB, IRB.getInt8Ty(),
Align(8), /*isStore*/ true)
.first;
Value *VrSrcPtr = IRB.CreateInBoundsPtrAdd(
IRB.CreateInBoundsPtrAdd(VAArgTLSCopy,
IRB.getInt32(AArch64VrBegOffset)),
VrRegSaveAreaShadowPtrOff);
Value *VrCopySize = IRB.CreateSub(VrArgSize, VrRegSaveAreaShadowPtrOff);
IRB.CreateMemCpy(VrRegSaveAreaShadowPtr, Align(8), VrSrcPtr, Align(8),
VrCopySize);
// And finally for remaining arguments.
Value *StackSaveAreaShadowPtr =
MSV.getShadowOriginPtr(StackSaveAreaPtr, IRB, IRB.getInt8Ty(),
Align(16), /*isStore*/ true)
.first;
Value *StackSrcPtr = IRB.CreateInBoundsPtrAdd(
VAArgTLSCopy, IRB.getInt32(AArch64VAEndOffset));
IRB.CreateMemCpy(StackSaveAreaShadowPtr, Align(16), StackSrcPtr,
Align(16), VAArgOverflowSize);
}
}
};
/// PowerPC64-specific implementation of VarArgHelper.
struct VarArgPowerPC64Helper : public VarArgHelperBase {
AllocaInst *VAArgTLSCopy = nullptr;
Value *VAArgSize = nullptr;
VarArgPowerPC64Helper(Function &F, MemorySanitizer &MS,
MemorySanitizerVisitor &MSV)
: VarArgHelperBase(F, MS, MSV, /*VAListTagSize=*/8) {}
void visitCallBase(CallBase &CB, IRBuilder<> &IRB) override {
// For PowerPC, we need to deal with alignment of stack arguments -
// they are mostly aligned to 8 bytes, but vectors and i128 arrays
// are aligned to 16 bytes, byvals can be aligned to 8 or 16 bytes,
// For that reason, we compute current offset from stack pointer (which is
// always properly aligned), and offset for the first vararg, then subtract
// them.
unsigned VAArgBase;
Triple TargetTriple(F.getParent()->getTargetTriple());
// Parameter save area starts at 48 bytes from frame pointer for ABIv1,
// and 32 bytes for ABIv2. This is usually determined by target
// endianness, but in theory could be overridden by function attribute.
if (TargetTriple.getArch() == Triple::ppc64)
VAArgBase = 48;
else
VAArgBase = 32;
unsigned VAArgOffset = VAArgBase;
const DataLayout &DL = F.getParent()->getDataLayout();
for (const auto &[ArgNo, A] : llvm::enumerate(CB.args())) {
bool IsFixed = ArgNo < CB.getFunctionType()->getNumParams();
bool IsByVal = CB.paramHasAttr(ArgNo, Attribute::ByVal);
if (IsByVal) {
assert(A->getType()->isPointerTy());
Type *RealTy = CB.getParamByValType(ArgNo);
uint64_t ArgSize = DL.getTypeAllocSize(RealTy);
Align ArgAlign = CB.getParamAlign(ArgNo).value_or(Align(8));
if (ArgAlign < 8)
ArgAlign = Align(8);
VAArgOffset = alignTo(VAArgOffset, ArgAlign);
if (!IsFixed) {
Value *Base = getShadowPtrForVAArgument(
RealTy, IRB, VAArgOffset - VAArgBase, ArgSize);
if (Base) {
Value *AShadowPtr, *AOriginPtr;
std::tie(AShadowPtr, AOriginPtr) =
MSV.getShadowOriginPtr(A, IRB, IRB.getInt8Ty(),
kShadowTLSAlignment, /*isStore*/ false);
IRB.CreateMemCpy(Base, kShadowTLSAlignment, AShadowPtr,
kShadowTLSAlignment, ArgSize);
}
}
VAArgOffset += alignTo(ArgSize, Align(8));
} else {
Value *Base;
uint64_t ArgSize = DL.getTypeAllocSize(A->getType());
Align ArgAlign = Align(8);
if (A->getType()->isArrayTy()) {
// Arrays are aligned to element size, except for long double
// arrays, which are aligned to 8 bytes.
Type *ElementTy = A->getType()->getArrayElementType();
if (!ElementTy->isPPC_FP128Ty())
ArgAlign = Align(DL.getTypeAllocSize(ElementTy));
} else if (A->getType()->isVectorTy()) {
// Vectors are naturally aligned.
ArgAlign = Align(ArgSize);
}
if (ArgAlign < 8)
ArgAlign = Align(8);
VAArgOffset = alignTo(VAArgOffset, ArgAlign);
if (DL.isBigEndian()) {
// Adjusting the shadow for argument with size < 8 to match the
// placement of bits in big endian system
if (ArgSize < 8)
VAArgOffset += (8 - ArgSize);
}
if (!IsFixed) {
Base = getShadowPtrForVAArgument(A->getType(), IRB,
VAArgOffset - VAArgBase, ArgSize);
if (Base)
IRB.CreateAlignedStore(MSV.getShadow(A), Base, kShadowTLSAlignment);
}
VAArgOffset += ArgSize;
VAArgOffset = alignTo(VAArgOffset, Align(8));
}
if (IsFixed)
VAArgBase = VAArgOffset;
}
Constant *TotalVAArgSize =
ConstantInt::get(IRB.getInt64Ty(), VAArgOffset - VAArgBase);
// Here using VAArgOverflowSizeTLS as VAArgSizeTLS to avoid creation of
// a new class member i.e. it is the total size of all VarArgs.
IRB.CreateStore(TotalVAArgSize, MS.VAArgOverflowSizeTLS);
}
void finalizeInstrumentation() override {
assert(!VAArgSize && !VAArgTLSCopy &&
"finalizeInstrumentation called twice");
IRBuilder<> IRB(MSV.FnPrologueEnd);
VAArgSize = IRB.CreateLoad(IRB.getInt64Ty(), MS.VAArgOverflowSizeTLS);
Value *CopySize =
IRB.CreateAdd(ConstantInt::get(MS.IntptrTy, 0), VAArgSize);
if (!VAStartInstrumentationList.empty()) {
// If there is a va_start in this function, make a backup copy of
// va_arg_tls somewhere in the function entry block.
VAArgTLSCopy = IRB.CreateAlloca(Type::getInt8Ty(*MS.C), CopySize);
VAArgTLSCopy->setAlignment(kShadowTLSAlignment);
IRB.CreateMemSet(VAArgTLSCopy, Constant::getNullValue(IRB.getInt8Ty()),
CopySize, kShadowTLSAlignment, false);
Value *SrcSize = IRB.CreateBinaryIntrinsic(
Intrinsic::umin, CopySize,
ConstantInt::get(MS.IntptrTy, kParamTLSSize));
IRB.CreateMemCpy(VAArgTLSCopy, kShadowTLSAlignment, MS.VAArgTLS,
kShadowTLSAlignment, SrcSize);
}
// Instrument va_start.
// Copy va_list shadow from the backup copy of the TLS contents.
for (size_t i = 0, n = VAStartInstrumentationList.size(); i < n; i++) {
CallInst *OrigInst = VAStartInstrumentationList[i];
NextNodeIRBuilder IRB(OrigInst);
Value *VAListTag = OrigInst->getArgOperand(0);
Type *RegSaveAreaPtrTy = PointerType::getUnqual(*MS.C); // i64*
Value *RegSaveAreaPtrPtr =
IRB.CreateIntToPtr(IRB.CreatePtrToInt(VAListTag, MS.IntptrTy),
PointerType::get(RegSaveAreaPtrTy, 0));
Value *RegSaveAreaPtr =
IRB.CreateLoad(RegSaveAreaPtrTy, RegSaveAreaPtrPtr);
Value *RegSaveAreaShadowPtr, *RegSaveAreaOriginPtr;
const Align Alignment = Align(8);
std::tie(RegSaveAreaShadowPtr, RegSaveAreaOriginPtr) =
MSV.getShadowOriginPtr(RegSaveAreaPtr, IRB, IRB.getInt8Ty(),
Alignment, /*isStore*/ true);
IRB.CreateMemCpy(RegSaveAreaShadowPtr, Alignment, VAArgTLSCopy, Alignment,
CopySize);
}
}
};
/// SystemZ-specific implementation of VarArgHelper.
struct VarArgSystemZHelper : public VarArgHelperBase {
static const unsigned SystemZGpOffset = 16;
static const unsigned SystemZGpEndOffset = 56;
static const unsigned SystemZFpOffset = 128;
static const unsigned SystemZFpEndOffset = 160;
static const unsigned SystemZMaxVrArgs = 8;
static const unsigned SystemZRegSaveAreaSize = 160;
static const unsigned SystemZOverflowOffset = 160;
static const unsigned SystemZVAListTagSize = 32;
static const unsigned SystemZOverflowArgAreaPtrOffset = 16;
static const unsigned SystemZRegSaveAreaPtrOffset = 24;
bool IsSoftFloatABI;
AllocaInst *VAArgTLSCopy = nullptr;
AllocaInst *VAArgTLSOriginCopy = nullptr;
Value *VAArgOverflowSize = nullptr;
enum class ArgKind {
GeneralPurpose,
FloatingPoint,
Vector,
Memory,
Indirect,
};
enum class ShadowExtension { None, Zero, Sign };
VarArgSystemZHelper(Function &F, MemorySanitizer &MS,
MemorySanitizerVisitor &MSV)
: VarArgHelperBase(F, MS, MSV, SystemZVAListTagSize),
IsSoftFloatABI(F.getFnAttribute("use-soft-float").getValueAsBool()) {}
ArgKind classifyArgument(Type *T) {
// T is a SystemZABIInfo::classifyArgumentType() output, and there are
// only a few possibilities of what it can be. In particular, enums, single
// element structs and large types have already been taken care of.
// Some i128 and fp128 arguments are converted to pointers only in the
// back end.
if (T->isIntegerTy(128) || T->isFP128Ty())
return ArgKind::Indirect;
if (T->isFloatingPointTy())
return IsSoftFloatABI ? ArgKind::GeneralPurpose : ArgKind::FloatingPoint;
if (T->isIntegerTy() || T->isPointerTy())
return ArgKind::GeneralPurpose;
if (T->isVectorTy())
return ArgKind::Vector;
return ArgKind::Memory;
}
ShadowExtension getShadowExtension(const CallBase &CB, unsigned ArgNo) {
// ABI says: "One of the simple integer types no more than 64 bits wide.
// ... If such an argument is shorter than 64 bits, replace it by a full
// 64-bit integer representing the same number, using sign or zero
// extension". Shadow for an integer argument has the same type as the
// argument itself, so it can be sign or zero extended as well.
bool ZExt = CB.paramHasAttr(ArgNo, Attribute::ZExt);
bool SExt = CB.paramHasAttr(ArgNo, Attribute::SExt);
if (ZExt) {
assert(!SExt);
return ShadowExtension::Zero;
}
if (SExt) {
assert(!ZExt);
return ShadowExtension::Sign;
}
return ShadowExtension::None;
}
void visitCallBase(CallBase &CB, IRBuilder<> &IRB) override {
unsigned GpOffset = SystemZGpOffset;
unsigned FpOffset = SystemZFpOffset;
unsigned VrIndex = 0;
unsigned OverflowOffset = SystemZOverflowOffset;
const DataLayout &DL = F.getParent()->getDataLayout();
for (const auto &[ArgNo, A] : llvm::enumerate(CB.args())) {
bool IsFixed = ArgNo < CB.getFunctionType()->getNumParams();
// SystemZABIInfo does not produce ByVal parameters.
assert(!CB.paramHasAttr(ArgNo, Attribute::ByVal));
Type *T = A->getType();
ArgKind AK = classifyArgument(T);
if (AK == ArgKind::Indirect) {
T = PointerType::get(T, 0);
AK = ArgKind::GeneralPurpose;
}
if (AK == ArgKind::GeneralPurpose && GpOffset >= SystemZGpEndOffset)
AK = ArgKind::Memory;
if (AK == ArgKind::FloatingPoint && FpOffset >= SystemZFpEndOffset)
AK = ArgKind::Memory;
if (AK == ArgKind::Vector && (VrIndex >= SystemZMaxVrArgs || !IsFixed))
AK = ArgKind::Memory;
Value *ShadowBase = nullptr;
Value *OriginBase = nullptr;
ShadowExtension SE = ShadowExtension::None;
switch (AK) {
case ArgKind::GeneralPurpose: {
// Always keep track of GpOffset, but store shadow only for varargs.
uint64_t ArgSize = 8;
if (GpOffset + ArgSize <= kParamTLSSize) {
if (!IsFixed) {
SE = getShadowExtension(CB, ArgNo);
uint64_t GapSize = 0;
if (SE == ShadowExtension::None) {
uint64_t ArgAllocSize = DL.getTypeAllocSize(T);
assert(ArgAllocSize <= ArgSize);
GapSize = ArgSize - ArgAllocSize;
}
ShadowBase = getShadowAddrForVAArgument(IRB, GpOffset + GapSize);
if (MS.TrackOrigins)
OriginBase = getOriginPtrForVAArgument(IRB, GpOffset + GapSize);
}
GpOffset += ArgSize;
} else {
GpOffset = kParamTLSSize;
}
break;
}
case ArgKind::FloatingPoint: {
// Always keep track of FpOffset, but store shadow only for varargs.
uint64_t ArgSize = 8;
if (FpOffset + ArgSize <= kParamTLSSize) {
if (!IsFixed) {
// PoP says: "A short floating-point datum requires only the
// left-most 32 bit positions of a floating-point register".
// Therefore, in contrast to AK_GeneralPurpose and AK_Memory,
// don't extend shadow and don't mind the gap.
ShadowBase = getShadowAddrForVAArgument(IRB, FpOffset);
if (MS.TrackOrigins)
OriginBase = getOriginPtrForVAArgument(IRB, FpOffset);
}
FpOffset += ArgSize;
} else {
FpOffset = kParamTLSSize;
}
break;
}
case ArgKind::Vector: {
// Keep track of VrIndex. No need to store shadow, since vector varargs
// go through AK_Memory.
assert(IsFixed);
VrIndex++;
break;
}
case ArgKind::Memory: {
// Keep track of OverflowOffset and store shadow only for varargs.
// Ignore fixed args, since we need to copy only the vararg portion of
// the overflow area shadow.
if (!IsFixed) {
uint64_t ArgAllocSize = DL.getTypeAllocSize(T);
uint64_t ArgSize = alignTo(ArgAllocSize, 8);
if (OverflowOffset + ArgSize <= kParamTLSSize) {
SE = getShadowExtension(CB, ArgNo);
uint64_t GapSize =
SE == ShadowExtension::None ? ArgSize - ArgAllocSize : 0;
ShadowBase =
getShadowAddrForVAArgument(IRB, OverflowOffset + GapSize);
if (MS.TrackOrigins)
OriginBase =
getOriginPtrForVAArgument(IRB, OverflowOffset + GapSize);
OverflowOffset += ArgSize;
} else {
OverflowOffset = kParamTLSSize;
}
}
break;
}
case ArgKind::Indirect:
llvm_unreachable("Indirect must be converted to GeneralPurpose");
}
if (ShadowBase == nullptr)
continue;
Value *Shadow = MSV.getShadow(A);
if (SE != ShadowExtension::None)
Shadow = MSV.CreateShadowCast(IRB, Shadow, IRB.getInt64Ty(),
/*Signed*/ SE == ShadowExtension::Sign);
ShadowBase = IRB.CreateIntToPtr(
ShadowBase, PointerType::get(Shadow->getType(), 0), "_msarg_va_s");
IRB.CreateStore(Shadow, ShadowBase);
if (MS.TrackOrigins) {
Value *Origin = MSV.getOrigin(A);
TypeSize StoreSize = DL.getTypeStoreSize(Shadow->getType());
MSV.paintOrigin(IRB, Origin, OriginBase, StoreSize,
kMinOriginAlignment);
}
}
Constant *OverflowSize = ConstantInt::get(
IRB.getInt64Ty(), OverflowOffset - SystemZOverflowOffset);
IRB.CreateStore(OverflowSize, MS.VAArgOverflowSizeTLS);
}
void copyRegSaveArea(IRBuilder<> &IRB, Value *VAListTag) {
Type *RegSaveAreaPtrTy = PointerType::getUnqual(*MS.C); // i64*
Value *RegSaveAreaPtrPtr = IRB.CreateIntToPtr(
IRB.CreateAdd(
IRB.CreatePtrToInt(VAListTag, MS.IntptrTy),
ConstantInt::get(MS.IntptrTy, SystemZRegSaveAreaPtrOffset)),
PointerType::get(RegSaveAreaPtrTy, 0));
Value *RegSaveAreaPtr = IRB.CreateLoad(RegSaveAreaPtrTy, RegSaveAreaPtrPtr);
Value *RegSaveAreaShadowPtr, *RegSaveAreaOriginPtr;
const Align Alignment = Align(8);
std::tie(RegSaveAreaShadowPtr, RegSaveAreaOriginPtr) =
MSV.getShadowOriginPtr(RegSaveAreaPtr, IRB, IRB.getInt8Ty(), Alignment,
/*isStore*/ true);
// TODO(iii): copy only fragments filled by visitCallBase()
// TODO(iii): support packed-stack && !use-soft-float
// For use-soft-float functions, it is enough to copy just the GPRs.
unsigned RegSaveAreaSize =
IsSoftFloatABI ? SystemZGpEndOffset : SystemZRegSaveAreaSize;
IRB.CreateMemCpy(RegSaveAreaShadowPtr, Alignment, VAArgTLSCopy, Alignment,
RegSaveAreaSize);
if (MS.TrackOrigins)
IRB.CreateMemCpy(RegSaveAreaOriginPtr, Alignment, VAArgTLSOriginCopy,
Alignment, RegSaveAreaSize);
}
// FIXME: This implementation limits OverflowOffset to kParamTLSSize, so we
// don't know real overflow size and can't clear shadow beyond kParamTLSSize.
void copyOverflowArea(IRBuilder<> &IRB, Value *VAListTag) {
Type *OverflowArgAreaPtrTy = PointerType::getUnqual(*MS.C); // i64*
Value *OverflowArgAreaPtrPtr = IRB.CreateIntToPtr(
IRB.CreateAdd(
IRB.CreatePtrToInt(VAListTag, MS.IntptrTy),
ConstantInt::get(MS.IntptrTy, SystemZOverflowArgAreaPtrOffset)),
PointerType::get(OverflowArgAreaPtrTy, 0));
Value *OverflowArgAreaPtr =
IRB.CreateLoad(OverflowArgAreaPtrTy, OverflowArgAreaPtrPtr);
Value *OverflowArgAreaShadowPtr, *OverflowArgAreaOriginPtr;
const Align Alignment = Align(8);
std::tie(OverflowArgAreaShadowPtr, OverflowArgAreaOriginPtr) =
MSV.getShadowOriginPtr(OverflowArgAreaPtr, IRB, IRB.getInt8Ty(),
Alignment, /*isStore*/ true);
Value *SrcPtr = IRB.CreateConstGEP1_32(IRB.getInt8Ty(), VAArgTLSCopy,
SystemZOverflowOffset);
IRB.CreateMemCpy(OverflowArgAreaShadowPtr, Alignment, SrcPtr, Alignment,
VAArgOverflowSize);
if (MS.TrackOrigins) {
SrcPtr = IRB.CreateConstGEP1_32(IRB.getInt8Ty(), VAArgTLSOriginCopy,
SystemZOverflowOffset);
IRB.CreateMemCpy(OverflowArgAreaOriginPtr, Alignment, SrcPtr, Alignment,
VAArgOverflowSize);
}
}
void finalizeInstrumentation() override {
assert(!VAArgOverflowSize && !VAArgTLSCopy &&
"finalizeInstrumentation called twice");
if (!VAStartInstrumentationList.empty()) {
// If there is a va_start in this function, make a backup copy of
// va_arg_tls somewhere in the function entry block.
IRBuilder<> IRB(MSV.FnPrologueEnd);
VAArgOverflowSize =
IRB.CreateLoad(IRB.getInt64Ty(), MS.VAArgOverflowSizeTLS);
Value *CopySize =
IRB.CreateAdd(ConstantInt::get(MS.IntptrTy, SystemZOverflowOffset),
VAArgOverflowSize);
VAArgTLSCopy = IRB.CreateAlloca(Type::getInt8Ty(*MS.C), CopySize);
VAArgTLSCopy->setAlignment(kShadowTLSAlignment);
IRB.CreateMemSet(VAArgTLSCopy, Constant::getNullValue(IRB.getInt8Ty()),
CopySize, kShadowTLSAlignment, false);
Value *SrcSize = IRB.CreateBinaryIntrinsic(
Intrinsic::umin, CopySize,
ConstantInt::get(MS.IntptrTy, kParamTLSSize));
IRB.CreateMemCpy(VAArgTLSCopy, kShadowTLSAlignment, MS.VAArgTLS,
kShadowTLSAlignment, SrcSize);
if (MS.TrackOrigins) {
VAArgTLSOriginCopy = IRB.CreateAlloca(Type::getInt8Ty(*MS.C), CopySize);
VAArgTLSOriginCopy->setAlignment(kShadowTLSAlignment);
IRB.CreateMemCpy(VAArgTLSOriginCopy, kShadowTLSAlignment,
MS.VAArgOriginTLS, kShadowTLSAlignment, SrcSize);
}
}
// Instrument va_start.
// Copy va_list shadow from the backup copy of the TLS contents.
for (size_t VaStartNo = 0, VaStartNum = VAStartInstrumentationList.size();
VaStartNo < VaStartNum; VaStartNo++) {
CallInst *OrigInst = VAStartInstrumentationList[VaStartNo];
NextNodeIRBuilder IRB(OrigInst);
Value *VAListTag = OrigInst->getArgOperand(0);
copyRegSaveArea(IRB, VAListTag);
copyOverflowArea(IRB, VAListTag);
}
}
};
// Loongarch64 is not a MIPS, but the current vargs calling convention matches
// the MIPS.
using VarArgLoongArch64Helper = VarArgMIPS64Helper;
/// A no-op implementation of VarArgHelper.
struct VarArgNoOpHelper : public VarArgHelper {
VarArgNoOpHelper(Function &F, MemorySanitizer &MS,
MemorySanitizerVisitor &MSV) {}
void visitCallBase(CallBase &CB, IRBuilder<> &IRB) override {}
void visitVAStartInst(VAStartInst &I) override {}
void visitVACopyInst(VACopyInst &I) override {}
void finalizeInstrumentation() override {}
};
} // end anonymous namespace
static VarArgHelper *CreateVarArgHelper(Function &Func, MemorySanitizer &Msan,
MemorySanitizerVisitor &Visitor) {
// VarArg handling is only implemented on AMD64. False positives are possible
// on other platforms.
Triple TargetTriple(Func.getParent()->getTargetTriple());
if (TargetTriple.getArch() == Triple::x86_64)
return new VarArgAMD64Helper(Func, Msan, Visitor);
else if (TargetTriple.isMIPS64())
return new VarArgMIPS64Helper(Func, Msan, Visitor);
else if (TargetTriple.getArch() == Triple::aarch64)
return new VarArgAArch64Helper(Func, Msan, Visitor);
else if (TargetTriple.getArch() == Triple::ppc64 ||
TargetTriple.getArch() == Triple::ppc64le)
return new VarArgPowerPC64Helper(Func, Msan, Visitor);
else if (TargetTriple.getArch() == Triple::systemz)
return new VarArgSystemZHelper(Func, Msan, Visitor);
else if (TargetTriple.isLoongArch64())
return new VarArgLoongArch64Helper(Func, Msan, Visitor);
else
return new VarArgNoOpHelper(Func, Msan, Visitor);
}
bool MemorySanitizer::sanitizeFunction(Function &F, TargetLibraryInfo &TLI) {
if (!CompileKernel && F.getName() == kMsanModuleCtorName)
return false;
if (F.hasFnAttribute(Attribute::DisableSanitizerInstrumentation))
return false;
MemorySanitizerVisitor Visitor(F, *this, TLI);
// Clear out memory attributes.
AttributeMask B;
B.addAttribute(Attribute::Memory).addAttribute(Attribute::Speculatable);
F.removeFnAttrs(B);
return Visitor.runOnFunction();
}
|