aboutsummaryrefslogtreecommitdiffstats
path: root/lib/Transforms/InstCombine/InstCombineCompares.cpp
blob: 6de92a4842ab387d7e7d5697bedc875171db24d3 (plain) (blame)
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
//===- InstCombineCompares.cpp --------------------------------------------===//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file implements the visitICmp and visitFCmp functions.
//
//===----------------------------------------------------------------------===//

#include "InstCombineInternal.h"
#include "llvm/ADT/APSInt.h"
#include "llvm/ADT/SetVector.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/ConstantFolding.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/IR/ConstantRange.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/GetElementPtrTypeIterator.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/PatternMatch.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/KnownBits.h"

using namespace llvm;
using namespace PatternMatch;

#define DEBUG_TYPE "instcombine"

// How many times is a select replaced by one of its operands?
STATISTIC(NumSel, "Number of select opts");


/// Compute Result = In1+In2, returning true if the result overflowed for this
/// type.
static bool addWithOverflow(APInt &Result, const APInt &In1,
                            const APInt &In2, bool IsSigned = false) {
  bool Overflow;
  if (IsSigned)
    Result = In1.sadd_ov(In2, Overflow);
  else
    Result = In1.uadd_ov(In2, Overflow);

  return Overflow;
}

/// Compute Result = In1-In2, returning true if the result overflowed for this
/// type.
static bool subWithOverflow(APInt &Result, const APInt &In1,
                            const APInt &In2, bool IsSigned = false) {
  bool Overflow;
  if (IsSigned)
    Result = In1.ssub_ov(In2, Overflow);
  else
    Result = In1.usub_ov(In2, Overflow);

  return Overflow;
}

/// Given an icmp instruction, return true if any use of this comparison is a
/// branch on sign bit comparison.
static bool hasBranchUse(ICmpInst &I) {
  for (auto *U : I.users())
    if (isa<BranchInst>(U))
      return true;
  return false;
}

/// Given an exploded icmp instruction, return true if the comparison only
/// checks the sign bit. If it only checks the sign bit, set TrueIfSigned if the
/// result of the comparison is true when the input value is signed.
static bool isSignBitCheck(ICmpInst::Predicate Pred, const APInt &RHS,
                           bool &TrueIfSigned) {
  switch (Pred) {
  case ICmpInst::ICMP_SLT:   // True if LHS s< 0
    TrueIfSigned = true;
    return RHS.isNullValue();
  case ICmpInst::ICMP_SLE:   // True if LHS s<= RHS and RHS == -1
    TrueIfSigned = true;
    return RHS.isAllOnesValue();
  case ICmpInst::ICMP_SGT:   // True if LHS s> -1
    TrueIfSigned = false;
    return RHS.isAllOnesValue();
  case ICmpInst::ICMP_UGT:
    // True if LHS u> RHS and RHS == high-bit-mask - 1
    TrueIfSigned = true;
    return RHS.isMaxSignedValue();
  case ICmpInst::ICMP_UGE:
    // True if LHS u>= RHS and RHS == high-bit-mask (2^7, 2^15, 2^31, etc)
    TrueIfSigned = true;
    return RHS.isSignMask();
  default:
    return false;
  }
}

/// Returns true if the exploded icmp can be expressed as a signed comparison
/// to zero and updates the predicate accordingly.
/// The signedness of the comparison is preserved.
/// TODO: Refactor with decomposeBitTestICmp()?
static bool isSignTest(ICmpInst::Predicate &Pred, const APInt &C) {
  if (!ICmpInst::isSigned(Pred))
    return false;

  if (C.isNullValue())
    return ICmpInst::isRelational(Pred);

  if (C.isOneValue()) {
    if (Pred == ICmpInst::ICMP_SLT) {
      Pred = ICmpInst::ICMP_SLE;
      return true;
    }
  } else if (C.isAllOnesValue()) {
    if (Pred == ICmpInst::ICMP_SGT) {
      Pred = ICmpInst::ICMP_SGE;
      return true;
    }
  }

  return false;
}

/// Given a signed integer type and a set of known zero and one bits, compute
/// the maximum and minimum values that could have the specified known zero and
/// known one bits, returning them in Min/Max.
/// TODO: Move to method on KnownBits struct?
static void computeSignedMinMaxValuesFromKnownBits(const KnownBits &Known,
                                                   APInt &Min, APInt &Max) {
  assert(Known.getBitWidth() == Min.getBitWidth() &&
         Known.getBitWidth() == Max.getBitWidth() &&
         "KnownZero, KnownOne and Min, Max must have equal bitwidth.");
  APInt UnknownBits = ~(Known.Zero|Known.One);

  // The minimum value is when all unknown bits are zeros, EXCEPT for the sign
  // bit if it is unknown.
  Min = Known.One;
  Max = Known.One|UnknownBits;

  if (UnknownBits.isNegative()) { // Sign bit is unknown
    Min.setSignBit();
    Max.clearSignBit();
  }
}

/// Given an unsigned integer type and a set of known zero and one bits, compute
/// the maximum and minimum values that could have the specified known zero and
/// known one bits, returning them in Min/Max.
/// TODO: Move to method on KnownBits struct?
static void computeUnsignedMinMaxValuesFromKnownBits(const KnownBits &Known,
                                                     APInt &Min, APInt &Max) {
  assert(Known.getBitWidth() == Min.getBitWidth() &&
         Known.getBitWidth() == Max.getBitWidth() &&
         "Ty, KnownZero, KnownOne and Min, Max must have equal bitwidth.");
  APInt UnknownBits = ~(Known.Zero|Known.One);

  // The minimum value is when the unknown bits are all zeros.
  Min = Known.One;
  // The maximum value is when the unknown bits are all ones.
  Max = Known.One|UnknownBits;
}

/// This is called when we see this pattern:
///   cmp pred (load (gep GV, ...)), cmpcst
/// where GV is a global variable with a constant initializer. Try to simplify
/// this into some simple computation that does not need the load. For example
/// we can optimize "icmp eq (load (gep "foo", 0, i)), 0" into "icmp eq i, 3".
///
/// If AndCst is non-null, then the loaded value is masked with that constant
/// before doing the comparison. This handles cases like "A[i]&4 == 0".
Instruction *InstCombiner::foldCmpLoadFromIndexedGlobal(GetElementPtrInst *GEP,
                                                        GlobalVariable *GV,
                                                        CmpInst &ICI,
                                                        ConstantInt *AndCst) {
  Constant *Init = GV->getInitializer();
  if (!isa<ConstantArray>(Init) && !isa<ConstantDataArray>(Init))
    return nullptr;

  uint64_t ArrayElementCount = Init->getType()->getArrayNumElements();
  // Don't blow up on huge arrays.
  if (ArrayElementCount > MaxArraySizeForCombine)
    return nullptr;

  // There are many forms of this optimization we can handle, for now, just do
  // the simple index into a single-dimensional array.
  //
  // Require: GEP GV, 0, i {{, constant indices}}
  if (GEP->getNumOperands() < 3 ||
      !isa<ConstantInt>(GEP->getOperand(1)) ||
      !cast<ConstantInt>(GEP->getOperand(1))->isZero() ||
      isa<Constant>(GEP->getOperand(2)))
    return nullptr;

  // Check that indices after the variable are constants and in-range for the
  // type they index.  Collect the indices.  This is typically for arrays of
  // structs.
  SmallVector<unsigned, 4> LaterIndices;

  Type *EltTy = Init->getType()->getArrayElementType();
  for (unsigned i = 3, e = GEP->getNumOperands(); i != e; ++i) {
    ConstantInt *Idx = dyn_cast<ConstantInt>(GEP->getOperand(i));
    if (!Idx) return nullptr;  // Variable index.

    uint64_t IdxVal = Idx->getZExtValue();
    if ((unsigned)IdxVal != IdxVal) return nullptr; // Too large array index.

    if (StructType *STy = dyn_cast<StructType>(EltTy))
      EltTy = STy->getElementType(IdxVal);
    else if (ArrayType *ATy = dyn_cast<ArrayType>(EltTy)) {
      if (IdxVal >= ATy->getNumElements()) return nullptr;
      EltTy = ATy->getElementType();
    } else {
      return nullptr; // Unknown type.
    }

    LaterIndices.push_back(IdxVal);
  }

  enum { Overdefined = -3, Undefined = -2 };

  // Variables for our state machines.

  // FirstTrueElement/SecondTrueElement - Used to emit a comparison of the form
  // "i == 47 | i == 87", where 47 is the first index the condition is true for,
  // and 87 is the second (and last) index.  FirstTrueElement is -2 when
  // undefined, otherwise set to the first true element.  SecondTrueElement is
  // -2 when undefined, -3 when overdefined and >= 0 when that index is true.
  int FirstTrueElement = Undefined, SecondTrueElement = Undefined;

  // FirstFalseElement/SecondFalseElement - Used to emit a comparison of the
  // form "i != 47 & i != 87".  Same state transitions as for true elements.
  int FirstFalseElement = Undefined, SecondFalseElement = Undefined;

  /// TrueRangeEnd/FalseRangeEnd - In conjunction with First*Element, these
  /// define a state machine that triggers for ranges of values that the index
  /// is true or false for.  This triggers on things like "abbbbc"[i] == 'b'.
  /// This is -2 when undefined, -3 when overdefined, and otherwise the last
  /// index in the range (inclusive).  We use -2 for undefined here because we
  /// use relative comparisons and don't want 0-1 to match -1.
  int TrueRangeEnd = Undefined, FalseRangeEnd = Undefined;

  // MagicBitvector - This is a magic bitvector where we set a bit if the
  // comparison is true for element 'i'.  If there are 64 elements or less in
  // the array, this will fully represent all the comparison results.
  uint64_t MagicBitvector = 0;

  // Scan the array and see if one of our patterns matches.
  Constant *CompareRHS = cast<Constant>(ICI.getOperand(1));
  for (unsigned i = 0, e = ArrayElementCount; i != e; ++i) {
    Constant *Elt = Init->getAggregateElement(i);
    if (!Elt) return nullptr;

    // If this is indexing an array of structures, get the structure element.
    if (!LaterIndices.empty())
      Elt = ConstantExpr::getExtractValue(Elt, LaterIndices);

    // If the element is masked, handle it.
    if (AndCst) Elt = ConstantExpr::getAnd(Elt, AndCst);

    // Find out if the comparison would be true or false for the i'th element.
    Constant *C = ConstantFoldCompareInstOperands(ICI.getPredicate(), Elt,
                                                  CompareRHS, DL, &TLI);
    // If the result is undef for this element, ignore it.
    if (isa<UndefValue>(C)) {
      // Extend range state machines to cover this element in case there is an
      // undef in the middle of the range.
      if (TrueRangeEnd == (int)i-1)
        TrueRangeEnd = i;
      if (FalseRangeEnd == (int)i-1)
        FalseRangeEnd = i;
      continue;
    }

    // If we can't compute the result for any of the elements, we have to give
    // up evaluating the entire conditional.
    if (!isa<ConstantInt>(C)) return nullptr;

    // Otherwise, we know if the comparison is true or false for this element,
    // update our state machines.
    bool IsTrueForElt = !cast<ConstantInt>(C)->isZero();

    // State machine for single/double/range index comparison.
    if (IsTrueForElt) {
      // Update the TrueElement state machine.
      if (FirstTrueElement == Undefined)
        FirstTrueElement = TrueRangeEnd = i;  // First true element.
      else {
        // Update double-compare state machine.
        if (SecondTrueElement == Undefined)
          SecondTrueElement = i;
        else
          SecondTrueElement = Overdefined;

        // Update range state machine.
        if (TrueRangeEnd == (int)i-1)
          TrueRangeEnd = i;
        else
          TrueRangeEnd = Overdefined;
      }
    } else {
      // Update the FalseElement state machine.
      if (FirstFalseElement == Undefined)
        FirstFalseElement = FalseRangeEnd = i; // First false element.
      else {
        // Update double-compare state machine.
        if (SecondFalseElement == Undefined)
          SecondFalseElement = i;
        else
          SecondFalseElement = Overdefined;

        // Update range state machine.
        if (FalseRangeEnd == (int)i-1)
          FalseRangeEnd = i;
        else
          FalseRangeEnd = Overdefined;
      }
    }

    // If this element is in range, update our magic bitvector.
    if (i < 64 && IsTrueForElt)
      MagicBitvector |= 1ULL << i;

    // If all of our states become overdefined, bail out early.  Since the
    // predicate is expensive, only check it every 8 elements.  This is only
    // really useful for really huge arrays.
    if ((i & 8) == 0 && i >= 64 && SecondTrueElement == Overdefined &&
        SecondFalseElement == Overdefined && TrueRangeEnd == Overdefined &&
        FalseRangeEnd == Overdefined)
      return nullptr;
  }

  // Now that we've scanned the entire array, emit our new comparison(s).  We
  // order the state machines in complexity of the generated code.
  Value *Idx = GEP->getOperand(2);

  // If the index is larger than the pointer size of the target, truncate the
  // index down like the GEP would do implicitly.  We don't have to do this for
  // an inbounds GEP because the index can't be out of range.
  if (!GEP->isInBounds()) {
    Type *IntPtrTy = DL.getIntPtrType(GEP->getType());
    unsigned PtrSize = IntPtrTy->getIntegerBitWidth();
    if (Idx->getType()->getPrimitiveSizeInBits() > PtrSize)
      Idx = Builder.CreateTrunc(Idx, IntPtrTy);
  }

  // If the comparison is only true for one or two elements, emit direct
  // comparisons.
  if (SecondTrueElement != Overdefined) {
    // None true -> false.
    if (FirstTrueElement == Undefined)
      return replaceInstUsesWith(ICI, Builder.getFalse());

    Value *FirstTrueIdx = ConstantInt::get(Idx->getType(), FirstTrueElement);

    // True for one element -> 'i == 47'.
    if (SecondTrueElement == Undefined)
      return new ICmpInst(ICmpInst::ICMP_EQ, Idx, FirstTrueIdx);

    // True for two elements -> 'i == 47 | i == 72'.
    Value *C1 = Builder.CreateICmpEQ(Idx, FirstTrueIdx);
    Value *SecondTrueIdx = ConstantInt::get(Idx->getType(), SecondTrueElement);
    Value *C2 = Builder.CreateICmpEQ(Idx, SecondTrueIdx);
    return BinaryOperator::CreateOr(C1, C2);
  }

  // If the comparison is only false for one or two elements, emit direct
  // comparisons.
  if (SecondFalseElement != Overdefined) {
    // None false -> true.
    if (FirstFalseElement == Undefined)
      return replaceInstUsesWith(ICI, Builder.getTrue());

    Value *FirstFalseIdx = ConstantInt::get(Idx->getType(), FirstFalseElement);

    // False for one element -> 'i != 47'.
    if (SecondFalseElement == Undefined)
      return new ICmpInst(ICmpInst::ICMP_NE, Idx, FirstFalseIdx);

    // False for two elements -> 'i != 47 & i != 72'.
    Value *C1 = Builder.CreateICmpNE(Idx, FirstFalseIdx);
    Value *SecondFalseIdx = ConstantInt::get(Idx->getType(),SecondFalseElement);
    Value *C2 = Builder.CreateICmpNE(Idx, SecondFalseIdx);
    return BinaryOperator::CreateAnd(C1, C2);
  }

  // If the comparison can be replaced with a range comparison for the elements
  // where it is true, emit the range check.
  if (TrueRangeEnd != Overdefined) {
    assert(TrueRangeEnd != FirstTrueElement && "Should emit single compare");

    // Generate (i-FirstTrue) <u (TrueRangeEnd-FirstTrue+1).
    if (FirstTrueElement) {
      Value *Offs = ConstantInt::get(Idx->getType(), -FirstTrueElement);
      Idx = Builder.CreateAdd(Idx, Offs);
    }

    Value *End = ConstantInt::get(Idx->getType(),
                                  TrueRangeEnd-FirstTrueElement+1);
    return new ICmpInst(ICmpInst::ICMP_ULT, Idx, End);
  }

  // False range check.
  if (FalseRangeEnd != Overdefined) {
    assert(FalseRangeEnd != FirstFalseElement && "Should emit single compare");
    // Generate (i-FirstFalse) >u (FalseRangeEnd-FirstFalse).
    if (FirstFalseElement) {
      Value *Offs = ConstantInt::get(Idx->getType(), -FirstFalseElement);
      Idx = Builder.CreateAdd(Idx, Offs);
    }

    Value *End = ConstantInt::get(Idx->getType(),
                                  FalseRangeEnd-FirstFalseElement);
    return new ICmpInst(ICmpInst::ICMP_UGT, Idx, End);
  }

  // If a magic bitvector captures the entire comparison state
  // of this load, replace it with computation that does:
  //   ((magic_cst >> i) & 1) != 0
  {
    Type *Ty = nullptr;

    // Look for an appropriate type:
    // - The type of Idx if the magic fits
    // - The smallest fitting legal type
    if (ArrayElementCount <= Idx->getType()->getIntegerBitWidth())
      Ty = Idx->getType();
    else
      Ty = DL.getSmallestLegalIntType(Init->getContext(), ArrayElementCount);

    if (Ty) {
      Value *V = Builder.CreateIntCast(Idx, Ty, false);
      V = Builder.CreateLShr(ConstantInt::get(Ty, MagicBitvector), V);
      V = Builder.CreateAnd(ConstantInt::get(Ty, 1), V);
      return new ICmpInst(ICmpInst::ICMP_NE, V, ConstantInt::get(Ty, 0));
    }
  }

  return nullptr;
}

/// Return a value that can be used to compare the *offset* implied by a GEP to
/// zero. For example, if we have &A[i], we want to return 'i' for
/// "icmp ne i, 0". Note that, in general, indices can be complex, and scales
/// are involved. The above expression would also be legal to codegen as
/// "icmp ne (i*4), 0" (assuming A is a pointer to i32).
/// This latter form is less amenable to optimization though, and we are allowed
/// to generate the first by knowing that pointer arithmetic doesn't overflow.
///
/// If we can't emit an optimized form for this expression, this returns null.
///
static Value *evaluateGEPOffsetExpression(User *GEP, InstCombiner &IC,
                                          const DataLayout &DL) {
  gep_type_iterator GTI = gep_type_begin(GEP);

  // Check to see if this gep only has a single variable index.  If so, and if
  // any constant indices are a multiple of its scale, then we can compute this
  // in terms of the scale of the variable index.  For example, if the GEP
  // implies an offset of "12 + i*4", then we can codegen this as "3 + i",
  // because the expression will cross zero at the same point.
  unsigned i, e = GEP->getNumOperands();
  int64_t Offset = 0;
  for (i = 1; i != e; ++i, ++GTI) {
    if (ConstantInt *CI = dyn_cast<ConstantInt>(GEP->getOperand(i))) {
      // Compute the aggregate offset of constant indices.
      if (CI->isZero()) continue;

      // Handle a struct index, which adds its field offset to the pointer.
      if (StructType *STy = GTI.getStructTypeOrNull()) {
        Offset += DL.getStructLayout(STy)->getElementOffset(CI->getZExtValue());
      } else {
        uint64_t Size = DL.getTypeAllocSize(GTI.getIndexedType());
        Offset += Size*CI->getSExtValue();
      }
    } else {
      // Found our variable index.
      break;
    }
  }

  // If there are no variable indices, we must have a constant offset, just
  // evaluate it the general way.
  if (i == e) return nullptr;

  Value *VariableIdx = GEP->getOperand(i);
  // Determine the scale factor of the variable element.  For example, this is
  // 4 if the variable index is into an array of i32.
  uint64_t VariableScale = DL.getTypeAllocSize(GTI.getIndexedType());

  // Verify that there are no other variable indices.  If so, emit the hard way.
  for (++i, ++GTI; i != e; ++i, ++GTI) {
    ConstantInt *CI = dyn_cast<ConstantInt>(GEP->getOperand(i));
    if (!CI) return nullptr;

    // Compute the aggregate offset of constant indices.
    if (CI->isZero()) continue;

    // Handle a struct index, which adds its field offset to the pointer.
    if (StructType *STy = GTI.getStructTypeOrNull()) {
      Offset += DL.getStructLayout(STy)->getElementOffset(CI->getZExtValue());
    } else {
      uint64_t Size = DL.getTypeAllocSize(GTI.getIndexedType());
      Offset += Size*CI->getSExtValue();
    }
  }

  // Okay, we know we have a single variable index, which must be a
  // pointer/array/vector index.  If there is no offset, life is simple, return
  // the index.
  Type *IntPtrTy = DL.getIntPtrType(GEP->getOperand(0)->getType());
  unsigned IntPtrWidth = IntPtrTy->getIntegerBitWidth();
  if (Offset == 0) {
    // Cast to intptrty in case a truncation occurs.  If an extension is needed,
    // we don't need to bother extending: the extension won't affect where the
    // computation crosses zero.
    if (VariableIdx->getType()->getPrimitiveSizeInBits() > IntPtrWidth) {
      VariableIdx = IC.Builder.CreateTrunc(VariableIdx, IntPtrTy);
    }
    return VariableIdx;
  }

  // Otherwise, there is an index.  The computation we will do will be modulo
  // the pointer size, so get it.
  uint64_t PtrSizeMask = ~0ULL >> (64-IntPtrWidth);

  Offset &= PtrSizeMask;
  VariableScale &= PtrSizeMask;

  // To do this transformation, any constant index must be a multiple of the
  // variable scale factor.  For example, we can evaluate "12 + 4*i" as "3 + i",
  // but we can't evaluate "10 + 3*i" in terms of i.  Check that the offset is a
  // multiple of the variable scale.
  int64_t NewOffs = Offset / (int64_t)VariableScale;
  if (Offset != NewOffs*(int64_t)VariableScale)
    return nullptr;

  // Okay, we can do this evaluation.  Start by converting the index to intptr.
  if (VariableIdx->getType() != IntPtrTy)
    VariableIdx = IC.Builder.CreateIntCast(VariableIdx, IntPtrTy,
                                            true /*Signed*/);
  Constant *OffsetVal = ConstantInt::get(IntPtrTy, NewOffs);
  return IC.Builder.CreateAdd(VariableIdx, OffsetVal, "offset");
}

/// Returns true if we can rewrite Start as a GEP with pointer Base
/// and some integer offset. The nodes that need to be re-written
/// for this transformation will be added to Explored.
static bool canRewriteGEPAsOffset(Value *Start, Value *Base,
                                  const DataLayout &DL,
                                  SetVector<Value *> &Explored) {
  SmallVector<Value *, 16> WorkList(1, Start);
  Explored.insert(Base);

  // The following traversal gives us an order which can be used
  // when doing the final transformation. Since in the final
  // transformation we create the PHI replacement instructions first,
  // we don't have to get them in any particular order.
  //
  // However, for other instructions we will have to traverse the
  // operands of an instruction first, which means that we have to
  // do a post-order traversal.
  while (!WorkList.empty()) {
    SetVector<PHINode *> PHIs;

    while (!WorkList.empty()) {
      if (Explored.size() >= 100)
        return false;

      Value *V = WorkList.back();

      if (Explored.count(V) != 0) {
        WorkList.pop_back();
        continue;
      }

      if (!isa<IntToPtrInst>(V) && !isa<PtrToIntInst>(V) &&
          !isa<GetElementPtrInst>(V) && !isa<PHINode>(V))
        // We've found some value that we can't explore which is different from
        // the base. Therefore we can't do this transformation.
        return false;

      if (isa<IntToPtrInst>(V) || isa<PtrToIntInst>(V)) {
        auto *CI = dyn_cast<CastInst>(V);
        if (!CI->isNoopCast(DL))
          return false;

        if (Explored.count(CI->getOperand(0)) == 0)
          WorkList.push_back(CI->getOperand(0));
      }

      if (auto *GEP = dyn_cast<GEPOperator>(V)) {
        // We're limiting the GEP to having one index. This will preserve
        // the original pointer type. We could handle more cases in the
        // future.
        if (GEP->getNumIndices() != 1 || !GEP->isInBounds() ||
            GEP->getType() != Start->getType())
          return false;

        if (Explored.count(GEP->getOperand(0)) == 0)
          WorkList.push_back(GEP->getOperand(0));
      }

      if (WorkList.back() == V) {
        WorkList.pop_back();
        // We've finished visiting this node, mark it as such.
        Explored.insert(V);
      }

      if (auto *PN = dyn_cast<PHINode>(V)) {
        // We cannot transform PHIs on unsplittable basic blocks.
        if (isa<CatchSwitchInst>(PN->getParent()->getTerminator()))
          return false;
        Explored.insert(PN);
        PHIs.insert(PN);
      }
    }

    // Explore the PHI nodes further.
    for (auto *PN : PHIs)
      for (Value *Op : PN->incoming_values())
        if (Explored.count(Op) == 0)
          WorkList.push_back(Op);
  }

  // Make sure that we can do this. Since we can't insert GEPs in a basic
  // block before a PHI node, we can't easily do this transformation if
  // we have PHI node users of transformed instructions.
  for (Value *Val : Explored) {
    for (Value *Use : Val->uses()) {

      auto *PHI = dyn_cast<PHINode>(Use);
      auto *Inst = dyn_cast<Instruction>(Val);

      if (Inst == Base || Inst == PHI || !Inst || !PHI ||
          Explored.count(PHI) == 0)
        continue;

      if (PHI->getParent() == Inst->getParent())
        return false;
    }
  }
  return true;
}

// Sets the appropriate insert point on Builder where we can add
// a replacement Instruction for V (if that is possible).
static void setInsertionPoint(IRBuilder<> &Builder, Value *V,
                              bool Before = true) {
  if (auto *PHI = dyn_cast<PHINode>(V)) {
    Builder.SetInsertPoint(&*PHI->getParent()->getFirstInsertionPt());
    return;
  }
  if (auto *I = dyn_cast<Instruction>(V)) {
    if (!Before)
      I = &*std::next(I->getIterator());
    Builder.SetInsertPoint(I);
    return;
  }
  if (auto *A = dyn_cast<Argument>(V)) {
    // Set the insertion point in the entry block.
    BasicBlock &Entry = A->getParent()->getEntryBlock();
    Builder.SetInsertPoint(&*Entry.getFirstInsertionPt());
    return;
  }
  // Otherwise, this is a constant and we don't need to set a new
  // insertion point.
  assert(isa<Constant>(V) && "Setting insertion point for unknown value!");
}

/// Returns a re-written value of Start as an indexed GEP using Base as a
/// pointer.
static Value *rewriteGEPAsOffset(Value *Start, Value *Base,
                                 const DataLayout &DL,
                                 SetVector<Value *> &Explored) {
  // Perform all the substitutions. This is a bit tricky because we can
  // have cycles in our use-def chains.
  // 1. Create the PHI nodes without any incoming values.
  // 2. Create all the other values.
  // 3. Add the edges for the PHI nodes.
  // 4. Emit GEPs to get the original pointers.
  // 5. Remove the original instructions.
  Type *IndexType = IntegerType::get(
      Base->getContext(), DL.getIndexTypeSizeInBits(Start->getType()));

  DenseMap<Value *, Value *> NewInsts;
  NewInsts[Base] = ConstantInt::getNullValue(IndexType);

  // Create the new PHI nodes, without adding any incoming values.
  for (Value *Val : Explored) {
    if (Val == Base)
      continue;
    // Create empty phi nodes. This avoids cyclic dependencies when creating
    // the remaining instructions.
    if (auto *PHI = dyn_cast<PHINode>(Val))
      NewInsts[PHI] = PHINode::Create(IndexType, PHI->getNumIncomingValues(),
                                      PHI->getName() + ".idx", PHI);
  }
  IRBuilder<> Builder(Base->getContext());

  // Create all the other instructions.
  for (Value *Val : Explored) {

    if (NewInsts.find(Val) != NewInsts.end())
      continue;

    if (auto *CI = dyn_cast<CastInst>(Val)) {
      NewInsts[CI] = NewInsts[CI->getOperand(0)];
      continue;
    }
    if (auto *GEP = dyn_cast<GEPOperator>(Val)) {
      Value *Index = NewInsts[GEP->getOperand(1)] ? NewInsts[GEP->getOperand(1)]
                                                  : GEP->getOperand(1);
      setInsertionPoint(Builder, GEP);
      // Indices might need to be sign extended. GEPs will magically do
      // this, but we need to do it ourselves here.
      if (Index->getType()->getScalarSizeInBits() !=
          NewInsts[GEP->getOperand(0)]->getType()->getScalarSizeInBits()) {
        Index = Builder.CreateSExtOrTrunc(
            Index, NewInsts[GEP->getOperand(0)]->getType(),
            GEP->getOperand(0)->getName() + ".sext");
      }

      auto *Op = NewInsts[GEP->getOperand(0)];
      if (isa<ConstantInt>(Op) && cast<ConstantInt>(Op)->isZero())
        NewInsts[GEP] = Index;
      else
        NewInsts[GEP] = Builder.CreateNSWAdd(
            Op, Index, GEP->getOperand(0)->getName() + ".add");
      continue;
    }
    if (isa<PHINode>(Val))
      continue;

    llvm_unreachable("Unexpected instruction type");
  }

  // Add the incoming values to the PHI nodes.
  for (Value *Val : Explored) {
    if (Val == Base)
      continue;
    // All the instructions have been created, we can now add edges to the
    // phi nodes.
    if (auto *PHI = dyn_cast<PHINode>(Val)) {
      PHINode *NewPhi = static_cast<PHINode *>(NewInsts[PHI]);
      for (unsigned I = 0, E = PHI->getNumIncomingValues(); I < E; ++I) {
        Value *NewIncoming = PHI->getIncomingValue(I);

        if (NewInsts.find(NewIncoming) != NewInsts.end())
          NewIncoming = NewInsts[NewIncoming];

        NewPhi->addIncoming(NewIncoming, PHI->getIncomingBlock(I));
      }
    }
  }

  for (Value *Val : Explored) {
    if (Val == Base)
      continue;

    // Depending on the type, for external users we have to emit
    // a GEP or a GEP + ptrtoint.
    setInsertionPoint(Builder, Val, false);

    // If required, create an inttoptr instruction for Base.
    Value *NewBase = Base;
    if (!Base->getType()->isPointerTy())
      NewBase = Builder.CreateBitOrPointerCast(Base, Start->getType(),
                                               Start->getName() + "to.ptr");

    Value *GEP = Builder.CreateInBoundsGEP(
        Start->getType()->getPointerElementType(), NewBase,
        makeArrayRef(NewInsts[Val]), Val->getName() + ".ptr");

    if (!Val->getType()->isPointerTy()) {
      Value *Cast = Builder.CreatePointerCast(GEP, Val->getType(),
                                              Val->getName() + ".conv");
      GEP = Cast;
    }
    Val->replaceAllUsesWith(GEP);
  }

  return NewInsts[Start];
}

/// Looks through GEPs, IntToPtrInsts and PtrToIntInsts in order to express
/// the input Value as a constant indexed GEP. Returns a pair containing
/// the GEPs Pointer and Index.
static std::pair<Value *, Value *>
getAsConstantIndexedAddress(Value *V, const DataLayout &DL) {
  Type *IndexType = IntegerType::get(V->getContext(),
                                     DL.getIndexTypeSizeInBits(V->getType()));

  Constant *Index = ConstantInt::getNullValue(IndexType);
  while (true) {
    if (GEPOperator *GEP = dyn_cast<GEPOperator>(V)) {
      // We accept only inbouds GEPs here to exclude the possibility of
      // overflow.
      if (!GEP->isInBounds())
        break;
      if (GEP->hasAllConstantIndices() && GEP->getNumIndices() == 1 &&
          GEP->getType() == V->getType()) {
        V = GEP->getOperand(0);
        Constant *GEPIndex = static_cast<Constant *>(GEP->getOperand(1));
        Index = ConstantExpr::getAdd(
            Index, ConstantExpr::getSExtOrBitCast(GEPIndex, IndexType));
        continue;
      }
      break;
    }
    if (auto *CI = dyn_cast<IntToPtrInst>(V)) {
      if (!CI->isNoopCast(DL))
        break;
      V = CI->getOperand(0);
      continue;
    }
    if (auto *CI = dyn_cast<PtrToIntInst>(V)) {
      if (!CI->isNoopCast(DL))
        break;
      V = CI->getOperand(0);
      continue;
    }
    break;
  }
  return {V, Index};
}

/// Converts (CMP GEPLHS, RHS) if this change would make RHS a constant.
/// We can look through PHIs, GEPs and casts in order to determine a common base
/// between GEPLHS and RHS.
static Instruction *transformToIndexedCompare(GEPOperator *GEPLHS, Value *RHS,
                                              ICmpInst::Predicate Cond,
                                              const DataLayout &DL) {
  if (!GEPLHS->hasAllConstantIndices())
    return nullptr;

  // Make sure the pointers have the same type.
  if (GEPLHS->getType() != RHS->getType())
    return nullptr;

  Value *PtrBase, *Index;
  std::tie(PtrBase, Index) = getAsConstantIndexedAddress(GEPLHS, DL);

  // The set of nodes that will take part in this transformation.
  SetVector<Value *> Nodes;

  if (!canRewriteGEPAsOffset(RHS, PtrBase, DL, Nodes))
    return nullptr;

  // We know we can re-write this as
  //  ((gep Ptr, OFFSET1) cmp (gep Ptr, OFFSET2)
  // Since we've only looked through inbouds GEPs we know that we
  // can't have overflow on either side. We can therefore re-write
  // this as:
  //   OFFSET1 cmp OFFSET2
  Value *NewRHS = rewriteGEPAsOffset(RHS, PtrBase, DL, Nodes);

  // RewriteGEPAsOffset has replaced RHS and all of its uses with a re-written
  // GEP having PtrBase as the pointer base, and has returned in NewRHS the
  // offset. Since Index is the offset of LHS to the base pointer, we will now
  // compare the offsets instead of comparing the pointers.
  return new ICmpInst(ICmpInst::getSignedPredicate(Cond), Index, NewRHS);
}

/// Fold comparisons between a GEP instruction and something else. At this point
/// we know that the GEP is on the LHS of the comparison.
Instruction *InstCombiner::foldGEPICmp(GEPOperator *GEPLHS, Value *RHS,
                                       ICmpInst::Predicate Cond,
                                       Instruction &I) {
  // Don't transform signed compares of GEPs into index compares. Even if the
  // GEP is inbounds, the final add of the base pointer can have signed overflow
  // and would change the result of the icmp.
  // e.g. "&foo[0] <s &foo[1]" can't be folded to "true" because "foo" could be
  // the maximum signed value for the pointer type.
  if (ICmpInst::isSigned(Cond))
    return nullptr;

  // Look through bitcasts and addrspacecasts. We do not however want to remove
  // 0 GEPs.
  if (!isa<GetElementPtrInst>(RHS))
    RHS = RHS->stripPointerCasts();

  Value *PtrBase = GEPLHS->getOperand(0);
  if (PtrBase == RHS && GEPLHS->isInBounds()) {
    // ((gep Ptr, OFFSET) cmp Ptr)   ---> (OFFSET cmp 0).
    // This transformation (ignoring the base and scales) is valid because we
    // know pointers can't overflow since the gep is inbounds.  See if we can
    // output an optimized form.
    Value *Offset = evaluateGEPOffsetExpression(GEPLHS, *this, DL);

    // If not, synthesize the offset the hard way.
    if (!Offset)
      Offset = EmitGEPOffset(GEPLHS);
    return new ICmpInst(ICmpInst::getSignedPredicate(Cond), Offset,
                        Constant::getNullValue(Offset->getType()));
  } else if (GEPOperator *GEPRHS = dyn_cast<GEPOperator>(RHS)) {
    // If the base pointers are different, but the indices are the same, just
    // compare the base pointer.
    if (PtrBase != GEPRHS->getOperand(0)) {
      bool IndicesTheSame = GEPLHS->getNumOperands()==GEPRHS->getNumOperands();
      IndicesTheSame &= GEPLHS->getOperand(0)->getType() ==
                        GEPRHS->getOperand(0)->getType();
      if (IndicesTheSame)
        for (unsigned i = 1, e = GEPLHS->getNumOperands(); i != e; ++i)
          if (GEPLHS->getOperand(i) != GEPRHS->getOperand(i)) {
            IndicesTheSame = false;
            break;
          }

      // If all indices are the same, just compare the base pointers.
      if (IndicesTheSame)
        return new ICmpInst(Cond, GEPLHS->getOperand(0), GEPRHS->getOperand(0));

      // If we're comparing GEPs with two base pointers that only differ in type
      // and both GEPs have only constant indices or just one use, then fold
      // the compare with the adjusted indices.
      if (GEPLHS->isInBounds() && GEPRHS->isInBounds() &&
          (GEPLHS->hasAllConstantIndices() || GEPLHS->hasOneUse()) &&
          (GEPRHS->hasAllConstantIndices() || GEPRHS->hasOneUse()) &&
          PtrBase->stripPointerCasts() ==
              GEPRHS->getOperand(0)->stripPointerCasts()) {
        Value *LOffset = EmitGEPOffset(GEPLHS);
        Value *ROffset = EmitGEPOffset(GEPRHS);

        // If we looked through an addrspacecast between different sized address
        // spaces, the LHS and RHS pointers are different sized
        // integers. Truncate to the smaller one.
        Type *LHSIndexTy = LOffset->getType();
        Type *RHSIndexTy = ROffset->getType();
        if (LHSIndexTy != RHSIndexTy) {
          if (LHSIndexTy->getPrimitiveSizeInBits() <
              RHSIndexTy->getPrimitiveSizeInBits()) {
            ROffset = Builder.CreateTrunc(ROffset, LHSIndexTy);
          } else
            LOffset = Builder.CreateTrunc(LOffset, RHSIndexTy);
        }

        Value *Cmp = Builder.CreateICmp(ICmpInst::getSignedPredicate(Cond),
                                        LOffset, ROffset);
        return replaceInstUsesWith(I, Cmp);
      }

      // Otherwise, the base pointers are different and the indices are
      // different. Try convert this to an indexed compare by looking through
      // PHIs/casts.
      return transformToIndexedCompare(GEPLHS, RHS, Cond, DL);
    }

    // If one of the GEPs has all zero indices, recurse.
    if (GEPLHS->hasAllZeroIndices())
      return foldGEPICmp(GEPRHS, GEPLHS->getOperand(0),
                         ICmpInst::getSwappedPredicate(Cond), I);

    // If the other GEP has all zero indices, recurse.
    if (GEPRHS->hasAllZeroIndices())
      return foldGEPICmp(GEPLHS, GEPRHS->getOperand(0), Cond, I);

    bool GEPsInBounds = GEPLHS->isInBounds() && GEPRHS->isInBounds();
    if (GEPLHS->getNumOperands() == GEPRHS->getNumOperands()) {
      // If the GEPs only differ by one index, compare it.
      unsigned NumDifferences = 0;  // Keep track of # differences.
      unsigned DiffOperand = 0;     // The operand that differs.
      for (unsigned i = 1, e = GEPRHS->getNumOperands(); i != e; ++i)
        if (GEPLHS->getOperand(i) != GEPRHS->getOperand(i)) {
          if (GEPLHS->getOperand(i)->getType()->getPrimitiveSizeInBits() !=
                   GEPRHS->getOperand(i)->getType()->getPrimitiveSizeInBits()) {
            // Irreconcilable differences.
            NumDifferences = 2;
            break;
          } else {
            if (NumDifferences++) break;
            DiffOperand = i;
          }
        }

      if (NumDifferences == 0)   // SAME GEP?
        return replaceInstUsesWith(I, // No comparison is needed here.
                             Builder.getInt1(ICmpInst::isTrueWhenEqual(Cond)));

      else if (NumDifferences == 1 && GEPsInBounds) {
        Value *LHSV = GEPLHS->getOperand(DiffOperand);
        Value *RHSV = GEPRHS->getOperand(DiffOperand);
        // Make sure we do a signed comparison here.
        return new ICmpInst(ICmpInst::getSignedPredicate(Cond), LHSV, RHSV);
      }
    }

    // Only lower this if the icmp is the only user of the GEP or if we expect
    // the result to fold to a constant!
    if (GEPsInBounds && (isa<ConstantExpr>(GEPLHS) || GEPLHS->hasOneUse()) &&
        (isa<ConstantExpr>(GEPRHS) || GEPRHS->hasOneUse())) {
      // ((gep Ptr, OFFSET1) cmp (gep Ptr, OFFSET2)  --->  (OFFSET1 cmp OFFSET2)
      Value *L = EmitGEPOffset(GEPLHS);
      Value *R = EmitGEPOffset(GEPRHS);
      return new ICmpInst(ICmpInst::getSignedPredicate(Cond), L, R);
    }
  }

  // Try convert this to an indexed compare by looking through PHIs/casts as a
  // last resort.
  return transformToIndexedCompare(GEPLHS, RHS, Cond, DL);
}

Instruction *InstCombiner::foldAllocaCmp(ICmpInst &ICI,
                                         const AllocaInst *Alloca,
                                         const Value *Other) {
  assert(ICI.isEquality() && "Cannot fold non-equality comparison.");

  // It would be tempting to fold away comparisons between allocas and any
  // pointer not based on that alloca (e.g. an argument). However, even
  // though such pointers cannot alias, they can still compare equal.
  //
  // But LLVM doesn't specify where allocas get their memory, so if the alloca
  // doesn't escape we can argue that it's impossible to guess its value, and we
  // can therefore act as if any such guesses are wrong.
  //
  // The code below checks that the alloca doesn't escape, and that it's only
  // used in a comparison once (the current instruction). The
  // single-comparison-use condition ensures that we're trivially folding all
  // comparisons against the alloca consistently, and avoids the risk of
  // erroneously folding a comparison of the pointer with itself.

  unsigned MaxIter = 32; // Break cycles and bound to constant-time.

  SmallVector<const Use *, 32> Worklist;
  for (const Use &U : Alloca->uses()) {
    if (Worklist.size() >= MaxIter)
      return nullptr;
    Worklist.push_back(&U);
  }

  unsigned NumCmps = 0;
  while (!Worklist.empty()) {
    assert(Worklist.size() <= MaxIter);
    const Use *U = Worklist.pop_back_val();
    const Value *V = U->getUser();
    --MaxIter;

    if (isa<BitCastInst>(V) || isa<GetElementPtrInst>(V) || isa<PHINode>(V) ||
        isa<SelectInst>(V)) {
      // Track the uses.
    } else if (isa<LoadInst>(V)) {
      // Loading from the pointer doesn't escape it.
      continue;
    } else if (const auto *SI = dyn_cast<StoreInst>(V)) {
      // Storing *to* the pointer is fine, but storing the pointer escapes it.
      if (SI->getValueOperand() == U->get())
        return nullptr;
      continue;
    } else if (isa<ICmpInst>(V)) {
      if (NumCmps++)
        return nullptr; // Found more than one cmp.
      continue;
    } else if (const auto *Intrin = dyn_cast<IntrinsicInst>(V)) {
      switch (Intrin->getIntrinsicID()) {
        // These intrinsics don't escape or compare the pointer. Memset is safe
        // because we don't allow ptrtoint. Memcpy and memmove are safe because
        // we don't allow stores, so src cannot point to V.
        case Intrinsic::lifetime_start: case Intrinsic::lifetime_end:
        case Intrinsic::memcpy: case Intrinsic::memmove: case Intrinsic::memset:
          continue;
        default:
          return nullptr;
      }
    } else {
      return nullptr;
    }
    for (const Use &U : V->uses()) {
      if (Worklist.size() >= MaxIter)
        return nullptr;
      Worklist.push_back(&U);
    }
  }

  Type *CmpTy = CmpInst::makeCmpResultType(Other->getType());
  return replaceInstUsesWith(
      ICI,
      ConstantInt::get(CmpTy, !CmpInst::isTrueWhenEqual(ICI.getPredicate())));
}

/// Fold "icmp pred (X+CI), X".
Instruction *InstCombiner::foldICmpAddOpConst(Value *X, ConstantInt *CI,
                                              ICmpInst::Predicate Pred) {
  // From this point on, we know that (X+C <= X) --> (X+C < X) because C != 0,
  // so the values can never be equal.  Similarly for all other "or equals"
  // operators.

  // (X+1) <u X        --> X >u (MAXUINT-1)        --> X == 255
  // (X+2) <u X        --> X >u (MAXUINT-2)        --> X > 253
  // (X+MAXUINT) <u X  --> X >u (MAXUINT-MAXUINT)  --> X != 0
  if (Pred == ICmpInst::ICMP_ULT || Pred == ICmpInst::ICMP_ULE) {
    Value *R =
      ConstantExpr::getSub(ConstantInt::getAllOnesValue(CI->getType()), CI);
    return new ICmpInst(ICmpInst::ICMP_UGT, X, R);
  }

  // (X+1) >u X        --> X <u (0-1)        --> X != 255
  // (X+2) >u X        --> X <u (0-2)        --> X <u 254
  // (X+MAXUINT) >u X  --> X <u (0-MAXUINT)  --> X <u 1  --> X == 0
  if (Pred == ICmpInst::ICMP_UGT || Pred == ICmpInst::ICMP_UGE)
    return new ICmpInst(ICmpInst::ICMP_ULT, X, ConstantExpr::getNeg(CI));

  unsigned BitWidth = CI->getType()->getPrimitiveSizeInBits();
  ConstantInt *SMax = ConstantInt::get(X->getContext(),
                                       APInt::getSignedMaxValue(BitWidth));

  // (X+ 1) <s X       --> X >s (MAXSINT-1)          --> X == 127
  // (X+ 2) <s X       --> X >s (MAXSINT-2)          --> X >s 125
  // (X+MAXSINT) <s X  --> X >s (MAXSINT-MAXSINT)    --> X >s 0
  // (X+MINSINT) <s X  --> X >s (MAXSINT-MINSINT)    --> X >s -1
  // (X+ -2) <s X      --> X >s (MAXSINT- -2)        --> X >s 126
  // (X+ -1) <s X      --> X >s (MAXSINT- -1)        --> X != 127
  if (Pred == ICmpInst::ICMP_SLT || Pred == ICmpInst::ICMP_SLE)
    return new ICmpInst(ICmpInst::ICMP_SGT, X, ConstantExpr::getSub(SMax, CI));

  // (X+ 1) >s X       --> X <s (MAXSINT-(1-1))       --> X != 127
  // (X+ 2) >s X       --> X <s (MAXSINT-(2-1))       --> X <s 126
  // (X+MAXSINT) >s X  --> X <s (MAXSINT-(MAXSINT-1)) --> X <s 1
  // (X+MINSINT) >s X  --> X <s (MAXSINT-(MINSINT-1)) --> X <s -2
  // (X+ -2) >s X      --> X <s (MAXSINT-(-2-1))      --> X <s -126
  // (X+ -1) >s X      --> X <s (MAXSINT-(-1-1))      --> X == -128

  assert(Pred == ICmpInst::ICMP_SGT || Pred == ICmpInst::ICMP_SGE);
  Constant *C = Builder.getInt(CI->getValue() - 1);
  return new ICmpInst(ICmpInst::ICMP_SLT, X, ConstantExpr::getSub(SMax, C));
}

/// Handle "(icmp eq/ne (ashr/lshr AP2, A), AP1)" ->
/// (icmp eq/ne A, Log2(AP2/AP1)) ->
/// (icmp eq/ne A, Log2(AP2) - Log2(AP1)).
Instruction *InstCombiner::foldICmpShrConstConst(ICmpInst &I, Value *A,
                                                 const APInt &AP1,
                                                 const APInt &AP2) {
  assert(I.isEquality() && "Cannot fold icmp gt/lt");

  auto getICmp = [&I](CmpInst::Predicate Pred, Value *LHS, Value *RHS) {
    if (I.getPredicate() == I.ICMP_NE)
      Pred = CmpInst::getInversePredicate(Pred);
    return new ICmpInst(Pred, LHS, RHS);
  };

  // Don't bother doing any work for cases which InstSimplify handles.
  if (AP2.isNullValue())
    return nullptr;

  bool IsAShr = isa<AShrOperator>(I.getOperand(0));
  if (IsAShr) {
    if (AP2.isAllOnesValue())
      return nullptr;
    if (AP2.isNegative() != AP1.isNegative())
      return nullptr;
    if (AP2.sgt(AP1))
      return nullptr;
  }

  if (!AP1)
    // 'A' must be large enough to shift out the highest set bit.
    return getICmp(I.ICMP_UGT, A,
                   ConstantInt::get(A->getType(), AP2.logBase2()));

  if (AP1 == AP2)
    return getICmp(I.ICMP_EQ, A, ConstantInt::getNullValue(A->getType()));

  int Shift;
  if (IsAShr && AP1.isNegative())
    Shift = AP1.countLeadingOnes() - AP2.countLeadingOnes();
  else
    Shift = AP1.countLeadingZeros() - AP2.countLeadingZeros();

  if (Shift > 0) {
    if (IsAShr && AP1 == AP2.ashr(Shift)) {
      // There are multiple solutions if we are comparing against -1 and the LHS
      // of the ashr is not a power of two.
      if (AP1.isAllOnesValue() && !AP2.isPowerOf2())
        return getICmp(I.ICMP_UGE, A, ConstantInt::get(A->getType(), Shift));
      return getICmp(I.ICMP_EQ, A, ConstantInt::get(A->getType(), Shift));
    } else if (AP1 == AP2.lshr(Shift)) {
      return getICmp(I.ICMP_EQ, A, ConstantInt::get(A->getType(), Shift));
    }
  }

  // Shifting const2 will never be equal to const1.
  // FIXME: This should always be handled by InstSimplify?
  auto *TorF = ConstantInt::get(I.getType(), I.getPredicate() == I.ICMP_NE);
  return replaceInstUsesWith(I, TorF);
}

/// Handle "(icmp eq/ne (shl AP2, A), AP1)" ->
/// (icmp eq/ne A, TrailingZeros(AP1) - TrailingZeros(AP2)).
Instruction *InstCombiner::foldICmpShlConstConst(ICmpInst &I, Value *A,
                                                 const APInt &AP1,
                                                 const APInt &AP2) {
  assert(I.isEquality() && "Cannot fold icmp gt/lt");

  auto getICmp = [&I](CmpInst::Predicate Pred, Value *LHS, Value *RHS) {
    if (I.getPredicate() == I.ICMP_NE)
      Pred = CmpInst::getInversePredicate(Pred);
    return new ICmpInst(Pred, LHS, RHS);
  };

  // Don't bother doing any work for cases which InstSimplify handles.
  if (AP2.isNullValue())
    return nullptr;

  unsigned AP2TrailingZeros = AP2.countTrailingZeros();

  if (!AP1 && AP2TrailingZeros != 0)
    return getICmp(
        I.ICMP_UGE, A,
        ConstantInt::get(A->getType(), AP2.getBitWidth() - AP2TrailingZeros));

  if (AP1 == AP2)
    return getICmp(I.ICMP_EQ, A, ConstantInt::getNullValue(A->getType()));

  // Get the distance between the lowest bits that are set.
  int Shift = AP1.countTrailingZeros() - AP2TrailingZeros;

  if (Shift > 0 && AP2.shl(Shift) == AP1)
    return getICmp(I.ICMP_EQ, A, ConstantInt::get(A->getType(), Shift));

  // Shifting const2 will never be equal to const1.
  // FIXME: This should always be handled by InstSimplify?
  auto *TorF = ConstantInt::get(I.getType(), I.getPredicate() == I.ICMP_NE);
  return replaceInstUsesWith(I, TorF);
}

/// The caller has matched a pattern of the form:
///   I = icmp ugt (add (add A, B), CI2), CI1
/// If this is of the form:
///   sum = a + b
///   if (sum+128 >u 255)
/// Then replace it with llvm.sadd.with.overflow.i8.
///
static Instruction *processUGT_ADDCST_ADD(ICmpInst &I, Value *A, Value *B,
                                          ConstantInt *CI2, ConstantInt *CI1,
                                          InstCombiner &IC) {
  // The transformation we're trying to do here is to transform this into an
  // llvm.sadd.with.overflow.  To do this, we have to replace the original add
  // with a narrower add, and discard the add-with-constant that is part of the
  // range check (if we can't eliminate it, this isn't profitable).

  // In order to eliminate the add-with-constant, the compare can be its only
  // use.
  Instruction *AddWithCst = cast<Instruction>(I.getOperand(0));
  if (!AddWithCst->hasOneUse())
    return nullptr;

  // If CI2 is 2^7, 2^15, 2^31, then it might be an sadd.with.overflow.
  if (!CI2->getValue().isPowerOf2())
    return nullptr;
  unsigned NewWidth = CI2->getValue().countTrailingZeros();
  if (NewWidth != 7 && NewWidth != 15 && NewWidth != 31)
    return nullptr;

  // The width of the new add formed is 1 more than the bias.
  ++NewWidth;

  // Check to see that CI1 is an all-ones value with NewWidth bits.
  if (CI1->getBitWidth() == NewWidth ||
      CI1->getValue() != APInt::getLowBitsSet(CI1->getBitWidth(), NewWidth))
    return nullptr;

  // This is only really a signed overflow check if the inputs have been
  // sign-extended; check for that condition. For example, if CI2 is 2^31 and
  // the operands of the add are 64 bits wide, we need at least 33 sign bits.
  unsigned NeededSignBits = CI1->getBitWidth() - NewWidth + 1;
  if (IC.ComputeNumSignBits(A, 0, &I) < NeededSignBits ||
      IC.ComputeNumSignBits(B, 0, &I) < NeededSignBits)
    return nullptr;

  // In order to replace the original add with a narrower
  // llvm.sadd.with.overflow, the only uses allowed are the add-with-constant
  // and truncates that discard the high bits of the add.  Verify that this is
  // the case.
  Instruction *OrigAdd = cast<Instruction>(AddWithCst->getOperand(0));
  for (User *U : OrigAdd->users()) {
    if (U == AddWithCst)
      continue;

    // Only accept truncates for now.  We would really like a nice recursive
    // predicate like SimplifyDemandedBits, but which goes downwards the use-def
    // chain to see which bits of a value are actually demanded.  If the
    // original add had another add which was then immediately truncated, we
    // could still do the transformation.
    TruncInst *TI = dyn_cast<TruncInst>(U);
    if (!TI || TI->getType()->getPrimitiveSizeInBits() > NewWidth)
      return nullptr;
  }

  // If the pattern matches, truncate the inputs to the narrower type and
  // use the sadd_with_overflow intrinsic to efficiently compute both the
  // result and the overflow bit.
  Type *NewType = IntegerType::get(OrigAdd->getContext(), NewWidth);
  Value *F = Intrinsic::getDeclaration(I.getModule(),
                                       Intrinsic::sadd_with_overflow, NewType);

  InstCombiner::BuilderTy &Builder = IC.Builder;

  // Put the new code above the original add, in case there are any uses of the
  // add between the add and the compare.
  Builder.SetInsertPoint(OrigAdd);

  Value *TruncA = Builder.CreateTrunc(A, NewType, A->getName() + ".trunc");
  Value *TruncB = Builder.CreateTrunc(B, NewType, B->getName() + ".trunc");
  CallInst *Call = Builder.CreateCall(F, {TruncA, TruncB}, "sadd");
  Value *Add = Builder.CreateExtractValue(Call, 0, "sadd.result");
  Value *ZExt = Builder.CreateZExt(Add, OrigAdd->getType());

  // The inner add was the result of the narrow add, zero extended to the
  // wider type.  Replace it with the result computed by the intrinsic.
  IC.replaceInstUsesWith(*OrigAdd, ZExt);

  // The original icmp gets replaced with the overflow value.
  return ExtractValueInst::Create(Call, 1, "sadd.overflow");
}

// Handle (icmp sgt smin(PosA, B) 0) -> (icmp sgt B 0)
Instruction *InstCombiner::foldICmpWithZero(ICmpInst &Cmp) {
  CmpInst::Predicate Pred = Cmp.getPredicate();
  Value *X = Cmp.getOperand(0);

  if (match(Cmp.getOperand(1), m_Zero()) && Pred == ICmpInst::ICMP_SGT) {
    Value *A, *B;
    SelectPatternResult SPR = matchSelectPattern(X, A, B);
    if (SPR.Flavor == SPF_SMIN) {
      if (isKnownPositive(A, DL, 0, &AC, &Cmp, &DT))
        return new ICmpInst(Pred, B, Cmp.getOperand(1));
      if (isKnownPositive(B, DL, 0, &AC, &Cmp, &DT))
        return new ICmpInst(Pred, A, Cmp.getOperand(1));
    }
  }
  return nullptr;
}

// Fold icmp Pred X, C.
Instruction *InstCombiner::foldICmpWithConstant(ICmpInst &Cmp) {
  CmpInst::Predicate Pred = Cmp.getPredicate();
  Value *X = Cmp.getOperand(0);

  const APInt *C;
  if (!match(Cmp.getOperand(1), m_APInt(C)))
    return nullptr;

  Value *A = nullptr, *B = nullptr;

  // Match the following pattern, which is a common idiom when writing
  // overflow-safe integer arithmetic functions. The source performs an addition
  // in wider type and explicitly checks for overflow using comparisons against
  // INT_MIN and INT_MAX. Simplify by using the sadd_with_overflow intrinsic.
  //
  // TODO: This could probably be generalized to handle other overflow-safe
  // operations if we worked out the formulas to compute the appropriate magic
  // constants.
  //
  // sum = a + b
  // if (sum+128 >u 255)  ...  -> llvm.sadd.with.overflow.i8
  {
    ConstantInt *CI2; // I = icmp ugt (add (add A, B), CI2), CI
    if (Pred == ICmpInst::ICMP_UGT &&
        match(X, m_Add(m_Add(m_Value(A), m_Value(B)), m_ConstantInt(CI2))))
      if (Instruction *Res = processUGT_ADDCST_ADD(
              Cmp, A, B, CI2, cast<ConstantInt>(Cmp.getOperand(1)), *this))
        return Res;
  }

  // FIXME: Use m_APInt to allow folds for splat constants.
  ConstantInt *CI = dyn_cast<ConstantInt>(Cmp.getOperand(1));
  if (!CI)
    return nullptr;

  // Canonicalize icmp instructions based on dominating conditions.
  BasicBlock *Parent = Cmp.getParent();
  BasicBlock *Dom = Parent->getSinglePredecessor();
  auto *BI = Dom ? dyn_cast<BranchInst>(Dom->getTerminator()) : nullptr;
  ICmpInst::Predicate Pred2;
  BasicBlock *TrueBB, *FalseBB;
  ConstantInt *CI2;
  if (BI && match(BI, m_Br(m_ICmp(Pred2, m_Specific(X), m_ConstantInt(CI2)),
                           TrueBB, FalseBB)) &&
      TrueBB != FalseBB) {
    ConstantRange CR =
        ConstantRange::makeAllowedICmpRegion(Pred, CI->getValue());
    ConstantRange DominatingCR =
        (Parent == TrueBB)
            ? ConstantRange::makeExactICmpRegion(Pred2, CI2->getValue())
            : ConstantRange::makeExactICmpRegion(
                  CmpInst::getInversePredicate(Pred2), CI2->getValue());
    ConstantRange Intersection = DominatingCR.intersectWith(CR);
    ConstantRange Difference = DominatingCR.difference(CR);
    if (Intersection.isEmptySet())
      return replaceInstUsesWith(Cmp, Builder.getFalse());
    if (Difference.isEmptySet())
      return replaceInstUsesWith(Cmp, Builder.getTrue());

    // If this is a normal comparison, it demands all bits. If it is a sign
    // bit comparison, it only demands the sign bit.
    bool UnusedBit;
    bool IsSignBit = isSignBitCheck(Pred, CI->getValue(), UnusedBit);

    // Canonicalizing a sign bit comparison that gets used in a branch,
    // pessimizes codegen by generating branch on zero instruction instead
    // of a test and branch. So we avoid canonicalizing in such situations
    // because test and branch instruction has better branch displacement
    // than compare and branch instruction.
    if (Cmp.isEquality() || (IsSignBit && hasBranchUse(Cmp)))
      return nullptr;

    if (auto *AI = Intersection.getSingleElement())
      return new ICmpInst(ICmpInst::ICMP_EQ, X, Builder.getInt(*AI));
    if (auto *AD = Difference.getSingleElement())
      return new ICmpInst(ICmpInst::ICMP_NE, X, Builder.getInt(*AD));
  }

  return nullptr;
}

/// Fold icmp (trunc X, Y), C.
Instruction *InstCombiner::foldICmpTruncConstant(ICmpInst &Cmp,
                                                 TruncInst *Trunc,
                                                 const APInt &C) {
  ICmpInst::Predicate Pred = Cmp.getPredicate();
  Value *X = Trunc->getOperand(0);
  if (C.isOneValue() && C.getBitWidth() > 1) {
    // icmp slt trunc(signum(V)) 1 --> icmp slt V, 1
    Value *V = nullptr;
    if (Pred == ICmpInst::ICMP_SLT && match(X, m_Signum(m_Value(V))))
      return new ICmpInst(ICmpInst::ICMP_SLT, V,
                          ConstantInt::get(V->getType(), 1));
  }

  if (Cmp.isEquality() && Trunc->hasOneUse()) {
    // Simplify icmp eq (trunc x to i8), 42 -> icmp eq x, 42|highbits if all
    // of the high bits truncated out of x are known.
    unsigned DstBits = Trunc->getType()->getScalarSizeInBits(),
             SrcBits = X->getType()->getScalarSizeInBits();
    KnownBits Known = computeKnownBits(X, 0, &Cmp);

    // If all the high bits are known, we can do this xform.
    if ((Known.Zero | Known.One).countLeadingOnes() >= SrcBits - DstBits) {
      // Pull in the high bits from known-ones set.
      APInt NewRHS = C.zext(SrcBits);
      NewRHS |= Known.One & APInt::getHighBitsSet(SrcBits, SrcBits - DstBits);
      return new ICmpInst(Pred, X, ConstantInt::get(X->getType(), NewRHS));
    }
  }

  return nullptr;
}

/// Fold icmp (xor X, Y), C.
Instruction *InstCombiner::foldICmpXorConstant(ICmpInst &Cmp,
                                               BinaryOperator *Xor,
                                               const APInt &C) {
  Value *X = Xor->getOperand(0);
  Value *Y = Xor->getOperand(1);
  const APInt *XorC;
  if (!match(Y, m_APInt(XorC)))
    return nullptr;

  // If this is a comparison that tests the signbit (X < 0) or (x > -1),
  // fold the xor.
  ICmpInst::Predicate Pred = Cmp.getPredicate();
  bool TrueIfSigned = false;
  if (isSignBitCheck(Cmp.getPredicate(), C, TrueIfSigned)) {

    // If the sign bit of the XorCst is not set, there is no change to
    // the operation, just stop using the Xor.
    if (!XorC->isNegative()) {
      Cmp.setOperand(0, X);
      Worklist.Add(Xor);
      return &Cmp;
    }

    // Emit the opposite comparison.
    if (TrueIfSigned)
      return new ICmpInst(ICmpInst::ICMP_SGT, X,
                          ConstantInt::getAllOnesValue(X->getType()));
    else
      return new ICmpInst(ICmpInst::ICMP_SLT, X,
                          ConstantInt::getNullValue(X->getType()));
  }

  if (Xor->hasOneUse()) {
    // (icmp u/s (xor X SignMask), C) -> (icmp s/u X, (xor C SignMask))
    if (!Cmp.isEquality() && XorC->isSignMask()) {
      Pred = Cmp.isSigned() ? Cmp.getUnsignedPredicate()
                            : Cmp.getSignedPredicate();
      return new ICmpInst(Pred, X, ConstantInt::get(X->getType(), C ^ *XorC));
    }

    // (icmp u/s (xor X ~SignMask), C) -> (icmp s/u X, (xor C ~SignMask))
    if (!Cmp.isEquality() && XorC->isMaxSignedValue()) {
      Pred = Cmp.isSigned() ? Cmp.getUnsignedPredicate()
                            : Cmp.getSignedPredicate();
      Pred = Cmp.getSwappedPredicate(Pred);
      return new ICmpInst(Pred, X, ConstantInt::get(X->getType(), C ^ *XorC));
    }
  }

  // (icmp ugt (xor X, C), ~C) -> (icmp ult X, C)
  //   iff -C is a power of 2
  if (Pred == ICmpInst::ICMP_UGT && *XorC == ~C && (C + 1).isPowerOf2())
    return new ICmpInst(ICmpInst::ICMP_ULT, X, Y);

  // (icmp ult (xor X, C), -C) -> (icmp uge X, C)
  //   iff -C is a power of 2
  if (Pred == ICmpInst::ICMP_ULT && *XorC == -C && C.isPowerOf2())
    return new ICmpInst(ICmpInst::ICMP_UGE, X, Y);

  return nullptr;
}

/// Fold icmp (and (sh X, Y), C2), C1.
Instruction *InstCombiner::foldICmpAndShift(ICmpInst &Cmp, BinaryOperator *And,
                                            const APInt &C1, const APInt &C2) {
  BinaryOperator *Shift = dyn_cast<BinaryOperator>(And->getOperand(0));
  if (!Shift || !Shift->isShift())
    return nullptr;

  // If this is: (X >> C3) & C2 != C1 (where any shift and any compare could
  // exist), turn it into (X & (C2 << C3)) != (C1 << C3). This happens a LOT in
  // code produced by the clang front-end, for bitfield access.
  // This seemingly simple opportunity to fold away a shift turns out to be
  // rather complicated. See PR17827 for details.
  unsigned ShiftOpcode = Shift->getOpcode();
  bool IsShl = ShiftOpcode == Instruction::Shl;
  const APInt *C3;
  if (match(Shift->getOperand(1), m_APInt(C3))) {
    bool CanFold = false;
    if (ShiftOpcode == Instruction::Shl) {
      // For a left shift, we can fold if the comparison is not signed. We can
      // also fold a signed comparison if the mask value and comparison value
      // are not negative. These constraints may not be obvious, but we can
      // prove that they are correct using an SMT solver.
      if (!Cmp.isSigned() || (!C2.isNegative() && !C1.isNegative()))
        CanFold = true;
    } else {
      bool IsAshr = ShiftOpcode == Instruction::AShr;
      // For a logical right shift, we can fold if the comparison is not signed.
      // We can also fold a signed comparison if the shifted mask value and the
      // shifted comparison value are not negative. These constraints may not be
      // obvious, but we can prove that they are correct using an SMT solver.
      // For an arithmetic shift right we can do the same, if we ensure
      // the And doesn't use any bits being shifted in. Normally these would
      // be turned into lshr by SimplifyDemandedBits, but not if there is an
      // additional user.
      if (!IsAshr || (C2.shl(*C3).lshr(*C3) == C2)) {
        if (!Cmp.isSigned() ||
            (!C2.shl(*C3).isNegative() && !C1.shl(*C3).isNegative()))
          CanFold = true;
      }
    }

    if (CanFold) {
      APInt NewCst = IsShl ? C1.lshr(*C3) : C1.shl(*C3);
      APInt SameAsC1 = IsShl ? NewCst.shl(*C3) : NewCst.lshr(*C3);
      // Check to see if we are shifting out any of the bits being compared.
      if (SameAsC1 != C1) {
        // If we shifted bits out, the fold is not going to work out. As a
        // special case, check to see if this means that the result is always
        // true or false now.
        if (Cmp.getPredicate() == ICmpInst::ICMP_EQ)
          return replaceInstUsesWith(Cmp, ConstantInt::getFalse(Cmp.getType()));
        if (Cmp.getPredicate() == ICmpInst::ICMP_NE)
          return replaceInstUsesWith(Cmp, ConstantInt::getTrue(Cmp.getType()));
      } else {
        Cmp.setOperand(1, ConstantInt::get(And->getType(), NewCst));
        APInt NewAndCst = IsShl ? C2.lshr(*C3) : C2.shl(*C3);
        And->setOperand(1, ConstantInt::get(And->getType(), NewAndCst));
        And->setOperand(0, Shift->getOperand(0));
        Worklist.Add(Shift); // Shift is dead.
        return &Cmp;
      }
    }
  }

  // Turn ((X >> Y) & C2) == 0  into  (X & (C2 << Y)) == 0.  The latter is
  // preferable because it allows the C2 << Y expression to be hoisted out of a
  // loop if Y is invariant and X is not.
  if (Shift->hasOneUse() && C1.isNullValue() && Cmp.isEquality() &&
      !Shift->isArithmeticShift() && !isa<Constant>(Shift->getOperand(0))) {
    // Compute C2 << Y.
    Value *NewShift =
        IsShl ? Builder.CreateLShr(And->getOperand(1), Shift->getOperand(1))
              : Builder.CreateShl(And->getOperand(1), Shift->getOperand(1));

    // Compute X & (C2 << Y).
    Value *NewAnd = Builder.CreateAnd(Shift->getOperand(0), NewShift);
    Cmp.setOperand(0, NewAnd);
    return &Cmp;
  }

  return nullptr;
}

/// Fold icmp (and X, C2), C1.
Instruction *InstCombiner::foldICmpAndConstConst(ICmpInst &Cmp,
                                                 BinaryOperator *And,
                                                 const APInt &C1) {
  const APInt *C2;
  if (!match(And->getOperand(1), m_APInt(C2)))
    return nullptr;

  if (!And->hasOneUse())
    return nullptr;

  // If the LHS is an 'and' of a truncate and we can widen the and/compare to
  // the input width without changing the value produced, eliminate the cast:
  //
  // icmp (and (trunc W), C2), C1 -> icmp (and W, C2'), C1'
  //
  // We can do this transformation if the constants do not have their sign bits
  // set or if it is an equality comparison. Extending a relational comparison
  // when we're checking the sign bit would not work.
  Value *W;
  if (match(And->getOperand(0), m_OneUse(m_Trunc(m_Value(W)))) &&
      (Cmp.isEquality() || (!C1.isNegative() && !C2->isNegative()))) {
    // TODO: Is this a good transform for vectors? Wider types may reduce
    // throughput. Should this transform be limited (even for scalars) by using
    // shouldChangeType()?
    if (!Cmp.getType()->isVectorTy()) {
      Type *WideType = W->getType();
      unsigned WideScalarBits = WideType->getScalarSizeInBits();
      Constant *ZextC1 = ConstantInt::get(WideType, C1.zext(WideScalarBits));
      Constant *ZextC2 = ConstantInt::get(WideType, C2->zext(WideScalarBits));
      Value *NewAnd = Builder.CreateAnd(W, ZextC2, And->getName());
      return new ICmpInst(Cmp.getPredicate(), NewAnd, ZextC1);
    }
  }

  if (Instruction *I = foldICmpAndShift(Cmp, And, C1, *C2))
    return I;

  // (icmp pred (and (or (lshr A, B), A), 1), 0) -->
  // (icmp pred (and A, (or (shl 1, B), 1), 0))
  //
  // iff pred isn't signed
  if (!Cmp.isSigned() && C1.isNullValue() && And->getOperand(0)->hasOneUse() &&
      match(And->getOperand(1), m_One())) {
    Constant *One = cast<Constant>(And->getOperand(1));
    Value *Or = And->getOperand(0);
    Value *A, *B, *LShr;
    if (match(Or, m_Or(m_Value(LShr), m_Value(A))) &&
        match(LShr, m_LShr(m_Specific(A), m_Value(B)))) {
      unsigned UsesRemoved = 0;
      if (And->hasOneUse())
        ++UsesRemoved;
      if (Or->hasOneUse())
        ++UsesRemoved;
      if (LShr->hasOneUse())
        ++UsesRemoved;

      // Compute A & ((1 << B) | 1)
      Value *NewOr = nullptr;
      if (auto *C = dyn_cast<Constant>(B)) {
        if (UsesRemoved >= 1)
          NewOr = ConstantExpr::getOr(ConstantExpr::getNUWShl(One, C), One);
      } else {
        if (UsesRemoved >= 3)
          NewOr = Builder.CreateOr(Builder.CreateShl(One, B, LShr->getName(),
                                                     /*HasNUW=*/true),
                                   One, Or->getName());
      }
      if (NewOr) {
        Value *NewAnd = Builder.CreateAnd(A, NewOr, And->getName());
        Cmp.setOperand(0, NewAnd);
        return &Cmp;
      }
    }
  }

  return nullptr;
}

/// Fold icmp (and X, Y), C.
Instruction *InstCombiner::foldICmpAndConstant(ICmpInst &Cmp,
                                               BinaryOperator *And,
                                               const APInt &C) {
  if (Instruction *I = foldICmpAndConstConst(Cmp, And, C))
    return I;

  // TODO: These all require that Y is constant too, so refactor with the above.

  // Try to optimize things like "A[i] & 42 == 0" to index computations.
  Value *X = And->getOperand(0);
  Value *Y = And->getOperand(1);
  if (auto *LI = dyn_cast<LoadInst>(X))
    if (auto *GEP = dyn_cast<GetElementPtrInst>(LI->getOperand(0)))
      if (auto *GV = dyn_cast<GlobalVariable>(GEP->getOperand(0)))
        if (GV->isConstant() && GV->hasDefinitiveInitializer() &&
            !LI->isVolatile() && isa<ConstantInt>(Y)) {
          ConstantInt *C2 = cast<ConstantInt>(Y);
          if (Instruction *Res = foldCmpLoadFromIndexedGlobal(GEP, GV, Cmp, C2))
            return Res;
        }

  if (!Cmp.isEquality())
    return nullptr;

  // X & -C == -C -> X >  u ~C
  // X & -C != -C -> X <= u ~C
  //   iff C is a power of 2
  if (Cmp.getOperand(1) == Y && (-C).isPowerOf2()) {
    auto NewPred = Cmp.getPredicate() == CmpInst::ICMP_EQ ? CmpInst::ICMP_UGT
                                                          : CmpInst::ICMP_ULE;
    return new ICmpInst(NewPred, X, SubOne(cast<Constant>(Cmp.getOperand(1))));
  }

  // (X & C2) == 0 -> (trunc X) >= 0
  // (X & C2) != 0 -> (trunc X) <  0
  //   iff C2 is a power of 2 and it masks the sign bit of a legal integer type.
  const APInt *C2;
  if (And->hasOneUse() && C.isNullValue() && match(Y, m_APInt(C2))) {
    int32_t ExactLogBase2 = C2->exactLogBase2();
    if (ExactLogBase2 != -1 && DL.isLegalInteger(ExactLogBase2 + 1)) {
      Type *NTy = IntegerType::get(Cmp.getContext(), ExactLogBase2 + 1);
      if (And->getType()->isVectorTy())
        NTy = VectorType::get(NTy, And->getType()->getVectorNumElements());
      Value *Trunc = Builder.CreateTrunc(X, NTy);
      auto NewPred = Cmp.getPredicate() == CmpInst::ICMP_EQ ? CmpInst::ICMP_SGE
                                                            : CmpInst::ICMP_SLT;
      return new ICmpInst(NewPred, Trunc, Constant::getNullValue(NTy));
    }
  }

  return nullptr;
}

/// Fold icmp (or X, Y), C.
Instruction *InstCombiner::foldICmpOrConstant(ICmpInst &Cmp, BinaryOperator *Or,
                                              const APInt &C) {
  ICmpInst::Predicate Pred = Cmp.getPredicate();
  if (C.isOneValue()) {
    // icmp slt signum(V) 1 --> icmp slt V, 1
    Value *V = nullptr;
    if (Pred == ICmpInst::ICMP_SLT && match(Or, m_Signum(m_Value(V))))
      return new ICmpInst(ICmpInst::ICMP_SLT, V,
                          ConstantInt::get(V->getType(), 1));
  }

  // X | C == C --> X <=u C
  // X | C != C --> X  >u C
  //   iff C+1 is a power of 2 (C is a bitmask of the low bits)
  if (Cmp.isEquality() && Cmp.getOperand(1) == Or->getOperand(1) &&
      (C + 1).isPowerOf2()) {
    Pred = (Pred == CmpInst::ICMP_EQ) ? CmpInst::ICMP_ULE : CmpInst::ICMP_UGT;
    return new ICmpInst(Pred, Or->getOperand(0), Or->getOperand(1));
  }

  if (!Cmp.isEquality() || !C.isNullValue() || !Or->hasOneUse())
    return nullptr;

  Value *P, *Q;
  if (match(Or, m_Or(m_PtrToInt(m_Value(P)), m_PtrToInt(m_Value(Q))))) {
    // Simplify icmp eq (or (ptrtoint P), (ptrtoint Q)), 0
    // -> and (icmp eq P, null), (icmp eq Q, null).
    Value *CmpP =
        Builder.CreateICmp(Pred, P, ConstantInt::getNullValue(P->getType()));
    Value *CmpQ =
        Builder.CreateICmp(Pred, Q, ConstantInt::getNullValue(Q->getType()));
    auto BOpc = Pred == CmpInst::ICMP_EQ ? Instruction::And : Instruction::Or;
    return BinaryOperator::Create(BOpc, CmpP, CmpQ);
  }

  // Are we using xors to bitwise check for a pair of (in)equalities? Convert to
  // a shorter form that has more potential to be folded even further.
  Value *X1, *X2, *X3, *X4;
  if (match(Or->getOperand(0), m_OneUse(m_Xor(m_Value(X1), m_Value(X2)))) &&
      match(Or->getOperand(1), m_OneUse(m_Xor(m_Value(X3), m_Value(X4))))) {
    // ((X1 ^ X2) || (X3 ^ X4)) == 0 --> (X1 == X2) && (X3 == X4)
    // ((X1 ^ X2) || (X3 ^ X4)) != 0 --> (X1 != X2) || (X3 != X4)
    Value *Cmp12 = Builder.CreateICmp(Pred, X1, X2);
    Value *Cmp34 = Builder.CreateICmp(Pred, X3, X4);
    auto BOpc = Pred == CmpInst::ICMP_EQ ? Instruction::And : Instruction::Or;
    return BinaryOperator::Create(BOpc, Cmp12, Cmp34);
  }

  return nullptr;
}

/// Fold icmp (mul X, Y), C.
Instruction *InstCombiner::foldICmpMulConstant(ICmpInst &Cmp,
                                               BinaryOperator *Mul,
                                               const APInt &C) {
  const APInt *MulC;
  if (!match(Mul->getOperand(1), m_APInt(MulC)))
    return nullptr;

  // If this is a test of the sign bit and the multiply is sign-preserving with
  // a constant operand, use the multiply LHS operand instead.
  ICmpInst::Predicate Pred = Cmp.getPredicate();
  if (isSignTest(Pred, C) && Mul->hasNoSignedWrap()) {
    if (MulC->isNegative())
      Pred = ICmpInst::getSwappedPredicate(Pred);
    return new ICmpInst(Pred, Mul->getOperand(0),
                        Constant::getNullValue(Mul->getType()));
  }

  return nullptr;
}

/// Fold icmp (shl 1, Y), C.
static Instruction *foldICmpShlOne(ICmpInst &Cmp, Instruction *Shl,
                                   const APInt &C) {
  Value *Y;
  if (!match(Shl, m_Shl(m_One(), m_Value(Y))))
    return nullptr;

  Type *ShiftType = Shl->getType();
  unsigned TypeBits = C.getBitWidth();
  bool CIsPowerOf2 = C.isPowerOf2();
  ICmpInst::Predicate Pred = Cmp.getPredicate();
  if (Cmp.isUnsigned()) {
    // (1 << Y) pred C -> Y pred Log2(C)
    if (!CIsPowerOf2) {
      // (1 << Y) <  30 -> Y <= 4
      // (1 << Y) <= 30 -> Y <= 4
      // (1 << Y) >= 30 -> Y >  4
      // (1 << Y) >  30 -> Y >  4
      if (Pred == ICmpInst::ICMP_ULT)
        Pred = ICmpInst::ICMP_ULE;
      else if (Pred == ICmpInst::ICMP_UGE)
        Pred = ICmpInst::ICMP_UGT;
    }

    // (1 << Y) >= 2147483648 -> Y >= 31 -> Y == 31
    // (1 << Y) <  2147483648 -> Y <  31 -> Y != 31
    unsigned CLog2 = C.logBase2();
    if (CLog2 == TypeBits - 1) {
      if (Pred == ICmpInst::ICMP_UGE)
        Pred = ICmpInst::ICMP_EQ;
      else if (Pred == ICmpInst::ICMP_ULT)
        Pred = ICmpInst::ICMP_NE;
    }
    return new ICmpInst(Pred, Y, ConstantInt::get(ShiftType, CLog2));
  } else if (Cmp.isSigned()) {
    Constant *BitWidthMinusOne = ConstantInt::get(ShiftType, TypeBits - 1);
    if (C.isAllOnesValue()) {
      // (1 << Y) <= -1 -> Y == 31
      if (Pred == ICmpInst::ICMP_SLE)
        return new ICmpInst(ICmpInst::ICMP_EQ, Y, BitWidthMinusOne);

      // (1 << Y) >  -1 -> Y != 31
      if (Pred == ICmpInst::ICMP_SGT)
        return new ICmpInst(ICmpInst::ICMP_NE, Y, BitWidthMinusOne);
    } else if (!C) {
      // (1 << Y) <  0 -> Y == 31
      // (1 << Y) <= 0 -> Y == 31
      if (Pred == ICmpInst::ICMP_SLT || Pred == ICmpInst::ICMP_SLE)
        return new ICmpInst(ICmpInst::ICMP_EQ, Y, BitWidthMinusOne);

      // (1 << Y) >= 0 -> Y != 31
      // (1 << Y) >  0 -> Y != 31
      if (Pred == ICmpInst::ICMP_SGT || Pred == ICmpInst::ICMP_SGE)
        return new ICmpInst(ICmpInst::ICMP_NE, Y, BitWidthMinusOne);
    }
  } else if (Cmp.isEquality() && CIsPowerOf2) {
    return new ICmpInst(Pred, Y, ConstantInt::get(ShiftType, C.logBase2()));
  }

  return nullptr;
}

/// Fold icmp (shl X, Y), C.
Instruction *InstCombiner::foldICmpShlConstant(ICmpInst &Cmp,
                                               BinaryOperator *Shl,
                                               const APInt &C) {
  const APInt *ShiftVal;
  if (Cmp.isEquality() && match(Shl->getOperand(0), m_APInt(ShiftVal)))
    return foldICmpShlConstConst(Cmp, Shl->getOperand(1), C, *ShiftVal);

  const APInt *ShiftAmt;
  if (!match(Shl->getOperand(1), m_APInt(ShiftAmt)))
    return foldICmpShlOne(Cmp, Shl, C);

  // Check that the shift amount is in range. If not, don't perform undefined
  // shifts. When the shift is visited, it will be simplified.
  unsigned TypeBits = C.getBitWidth();
  if (ShiftAmt->uge(TypeBits))
    return nullptr;

  ICmpInst::Predicate Pred = Cmp.getPredicate();
  Value *X = Shl->getOperand(0);
  Type *ShType = Shl->getType();

  // NSW guarantees that we are only shifting out sign bits from the high bits,
  // so we can ASHR the compare constant without needing a mask and eliminate
  // the shift.
  if (Shl->hasNoSignedWrap()) {
    if (Pred == ICmpInst::ICMP_SGT) {
      // icmp Pred (shl nsw X, ShiftAmt), C --> icmp Pred X, (C >>s ShiftAmt)
      APInt ShiftedC = C.ashr(*ShiftAmt);
      return new ICmpInst(Pred, X, ConstantInt::get(ShType, ShiftedC));
    }
    if ((Pred == ICmpInst::ICMP_EQ || Pred == ICmpInst::ICMP_NE) &&
        C.ashr(*ShiftAmt).shl(*ShiftAmt) == C) {
      APInt ShiftedC = C.ashr(*ShiftAmt);
      return new ICmpInst(Pred, X, ConstantInt::get(ShType, ShiftedC));
    }
    if (Pred == ICmpInst::ICMP_SLT) {
      // SLE is the same as above, but SLE is canonicalized to SLT, so convert:
      // (X << S) <=s C is equiv to X <=s (C >> S) for all C
      // (X << S) <s (C + 1) is equiv to X <s (C >> S) + 1 if C <s SMAX
      // (X << S) <s C is equiv to X <s ((C - 1) >> S) + 1 if C >s SMIN
      assert(!C.isMinSignedValue() && "Unexpected icmp slt");
      APInt ShiftedC = (C - 1).ashr(*ShiftAmt) + 1;
      return new ICmpInst(Pred, X, ConstantInt::get(ShType, ShiftedC));
    }
    // If this is a signed comparison to 0 and the shift is sign preserving,
    // use the shift LHS operand instead; isSignTest may change 'Pred', so only
    // do that if we're sure to not continue on in this function.
    if (isSignTest(Pred, C))
      return new ICmpInst(Pred, X, Constant::getNullValue(ShType));
  }

  // NUW guarantees that we are only shifting out zero bits from the high bits,
  // so we can LSHR the compare constant without needing a mask and eliminate
  // the shift.
  if (Shl->hasNoUnsignedWrap()) {
    if (Pred == ICmpInst::ICMP_UGT) {
      // icmp Pred (shl nuw X, ShiftAmt), C --> icmp Pred X, (C >>u ShiftAmt)
      APInt ShiftedC = C.lshr(*ShiftAmt);
      return new ICmpInst(Pred, X, ConstantInt::get(ShType, ShiftedC));
    }
    if ((Pred == ICmpInst::ICMP_EQ || Pred == ICmpInst::ICMP_NE) &&
        C.lshr(*ShiftAmt).shl(*ShiftAmt) == C) {
      APInt ShiftedC = C.lshr(*ShiftAmt);
      return new ICmpInst(Pred, X, ConstantInt::get(ShType, ShiftedC));
    }
    if (Pred == ICmpInst::ICMP_ULT) {
      // ULE is the same as above, but ULE is canonicalized to ULT, so convert:
      // (X << S) <=u C is equiv to X <=u (C >> S) for all C
      // (X << S) <u (C + 1) is equiv to X <u (C >> S) + 1 if C <u ~0u
      // (X << S) <u C is equiv to X <u ((C - 1) >> S) + 1 if C >u 0
      assert(C.ugt(0) && "ult 0 should have been eliminated");
      APInt ShiftedC = (C - 1).lshr(*ShiftAmt) + 1;
      return new ICmpInst(Pred, X, ConstantInt::get(ShType, ShiftedC));
    }
  }

  if (Cmp.isEquality() && Shl->hasOneUse()) {
    // Strength-reduce the shift into an 'and'.
    Constant *Mask = ConstantInt::get(
        ShType,
        APInt::getLowBitsSet(TypeBits, TypeBits - ShiftAmt->getZExtValue()));
    Value *And = Builder.CreateAnd(X, Mask, Shl->getName() + ".mask");
    Constant *LShrC = ConstantInt::get(ShType, C.lshr(*ShiftAmt));
    return new ICmpInst(Pred, And, LShrC);
  }

  // Otherwise, if this is a comparison of the sign bit, simplify to and/test.
  bool TrueIfSigned = false;
  if (Shl->hasOneUse() && isSignBitCheck(Pred, C, TrueIfSigned)) {
    // (X << 31) <s 0  --> (X & 1) != 0
    Constant *Mask = ConstantInt::get(
        ShType,
        APInt::getOneBitSet(TypeBits, TypeBits - ShiftAmt->getZExtValue() - 1));
    Value *And = Builder.CreateAnd(X, Mask, Shl->getName() + ".mask");
    return new ICmpInst(TrueIfSigned ? ICmpInst::ICMP_NE : ICmpInst::ICMP_EQ,
                        And, Constant::getNullValue(ShType));
  }

  // Transform (icmp pred iM (shl iM %v, N), C)
  // -> (icmp pred i(M-N) (trunc %v iM to i(M-N)), (trunc (C>>N))
  // Transform the shl to a trunc if (trunc (C>>N)) has no loss and M-N.
  // This enables us to get rid of the shift in favor of a trunc that may be
  // free on the target. It has the additional benefit of comparing to a
  // smaller constant that may be more target-friendly.
  unsigned Amt = ShiftAmt->getLimitedValue(TypeBits - 1);
  if (Shl->hasOneUse() && Amt != 0 && C.countTrailingZeros() >= Amt &&
      DL.isLegalInteger(TypeBits - Amt)) {
    Type *TruncTy = IntegerType::get(Cmp.getContext(), TypeBits - Amt);
    if (ShType->isVectorTy())
      TruncTy = VectorType::get(TruncTy, ShType->getVectorNumElements());
    Constant *NewC =
        ConstantInt::get(TruncTy, C.ashr(*ShiftAmt).trunc(TypeBits - Amt));
    return new ICmpInst(Pred, Builder.CreateTrunc(X, TruncTy), NewC);
  }

  return nullptr;
}

/// Fold icmp ({al}shr X, Y), C.
Instruction *InstCombiner::foldICmpShrConstant(ICmpInst &Cmp,
                                               BinaryOperator *Shr,
                                               const APInt &C) {
  // An exact shr only shifts out zero bits, so:
  // icmp eq/ne (shr X, Y), 0 --> icmp eq/ne X, 0
  Value *X = Shr->getOperand(0);
  CmpInst::Predicate Pred = Cmp.getPredicate();
  if (Cmp.isEquality() && Shr->isExact() && Shr->hasOneUse() &&
      C.isNullValue())
    return new ICmpInst(Pred, X, Cmp.getOperand(1));

  const APInt *ShiftVal;
  if (Cmp.isEquality() && match(Shr->getOperand(0), m_APInt(ShiftVal)))
    return foldICmpShrConstConst(Cmp, Shr->getOperand(1), C, *ShiftVal);

  const APInt *ShiftAmt;
  if (!match(Shr->getOperand(1), m_APInt(ShiftAmt)))
    return nullptr;

  // Check that the shift amount is in range. If not, don't perform undefined
  // shifts. When the shift is visited it will be simplified.
  unsigned TypeBits = C.getBitWidth();
  unsigned ShAmtVal = ShiftAmt->getLimitedValue(TypeBits);
  if (ShAmtVal >= TypeBits || ShAmtVal == 0)
    return nullptr;

  bool IsAShr = Shr->getOpcode() == Instruction::AShr;
  bool IsExact = Shr->isExact();
  Type *ShrTy = Shr->getType();
  // TODO: If we could guarantee that InstSimplify would handle all of the
  // constant-value-based preconditions in the folds below, then we could assert
  // those conditions rather than checking them. This is difficult because of
  // undef/poison (PR34838).
  if (IsAShr) {
    if (Pred == CmpInst::ICMP_SLT || (Pred == CmpInst::ICMP_SGT && IsExact)) {
      // icmp slt (ashr X, ShAmtC), C --> icmp slt X, (C << ShAmtC)
      // icmp sgt (ashr exact X, ShAmtC), C --> icmp sgt X, (C << ShAmtC)
      APInt ShiftedC = C.shl(ShAmtVal);
      if (ShiftedC.ashr(ShAmtVal) == C)
        return new ICmpInst(Pred, X, ConstantInt::get(ShrTy, ShiftedC));
    }
    if (Pred == CmpInst::ICMP_SGT) {
      // icmp sgt (ashr X, ShAmtC), C --> icmp sgt X, ((C + 1) << ShAmtC) - 1
      APInt ShiftedC = (C + 1).shl(ShAmtVal) - 1;
      if (!C.isMaxSignedValue() && !(C + 1).shl(ShAmtVal).isMinSignedValue() &&
          (ShiftedC + 1).ashr(ShAmtVal) == (C + 1))
        return new ICmpInst(Pred, X, ConstantInt::get(ShrTy, ShiftedC));
    }
  } else {
    if (Pred == CmpInst::ICMP_ULT || (Pred == CmpInst::ICMP_UGT && IsExact)) {
      // icmp ult (lshr X, ShAmtC), C --> icmp ult X, (C << ShAmtC)
      // icmp ugt (lshr exact X, ShAmtC), C --> icmp ugt X, (C << ShAmtC)
      APInt ShiftedC = C.shl(ShAmtVal);
      if (ShiftedC.lshr(ShAmtVal) == C)
        return new ICmpInst(Pred, X, ConstantInt::get(ShrTy, ShiftedC));
    }
    if (Pred == CmpInst::ICMP_UGT) {
      // icmp ugt (lshr X, ShAmtC), C --> icmp ugt X, ((C + 1) << ShAmtC) - 1
      APInt ShiftedC = (C + 1).shl(ShAmtVal) - 1;
      if ((ShiftedC + 1).lshr(ShAmtVal) == (C + 1))
        return new ICmpInst(Pred, X, ConstantInt::get(ShrTy, ShiftedC));
    }
  }

  if (!Cmp.isEquality())
    return nullptr;

  // Handle equality comparisons of shift-by-constant.

  // If the comparison constant changes with the shift, the comparison cannot
  // succeed (bits of the comparison constant cannot match the shifted value).
  // This should be known by InstSimplify and already be folded to true/false.
  assert(((IsAShr && C.shl(ShAmtVal).ashr(ShAmtVal) == C) ||
          (!IsAShr && C.shl(ShAmtVal).lshr(ShAmtVal) == C)) &&
         "Expected icmp+shr simplify did not occur.");

  // If the bits shifted out are known zero, compare the unshifted value:
  //  (X & 4) >> 1 == 2  --> (X & 4) == 4.
  if (Shr->isExact())
    return new ICmpInst(Pred, X, ConstantInt::get(ShrTy, C << ShAmtVal));

  if (Shr->hasOneUse()) {
    // Canonicalize the shift into an 'and':
    // icmp eq/ne (shr X, ShAmt), C --> icmp eq/ne (and X, HiMask), (C << ShAmt)
    APInt Val(APInt::getHighBitsSet(TypeBits, TypeBits - ShAmtVal));
    Constant *Mask = ConstantInt::get(ShrTy, Val);
    Value *And = Builder.CreateAnd(X, Mask, Shr->getName() + ".mask");
    return new ICmpInst(Pred, And, ConstantInt::get(ShrTy, C << ShAmtVal));
  }

  return nullptr;
}

/// Fold icmp (udiv X, Y), C.
Instruction *InstCombiner::foldICmpUDivConstant(ICmpInst &Cmp,
                                                BinaryOperator *UDiv,
                                                const APInt &C) {
  const APInt *C2;
  if (!match(UDiv->getOperand(0), m_APInt(C2)))
    return nullptr;

  assert(*C2 != 0 && "udiv 0, X should have been simplified already.");

  // (icmp ugt (udiv C2, Y), C) -> (icmp ule Y, C2/(C+1))
  Value *Y = UDiv->getOperand(1);
  if (Cmp.getPredicate() == ICmpInst::ICMP_UGT) {
    assert(!C.isMaxValue() &&
           "icmp ugt X, UINT_MAX should have been simplified already.");
    return new ICmpInst(ICmpInst::ICMP_ULE, Y,
                        ConstantInt::get(Y->getType(), C2->udiv(C + 1)));
  }

  // (icmp ult (udiv C2, Y), C) -> (icmp ugt Y, C2/C)
  if (Cmp.getPredicate() == ICmpInst::ICMP_ULT) {
    assert(C != 0 && "icmp ult X, 0 should have been simplified already.");
    return new ICmpInst(ICmpInst::ICMP_UGT, Y,
                        ConstantInt::get(Y->getType(), C2->udiv(C)));
  }

  return nullptr;
}

/// Fold icmp ({su}div X, Y), C.
Instruction *InstCombiner::foldICmpDivConstant(ICmpInst &Cmp,
                                               BinaryOperator *Div,
                                               const APInt &C) {
  // Fold: icmp pred ([us]div X, C2), C -> range test
  // Fold this div into the comparison, producing a range check.
  // Determine, based on the divide type, what the range is being
  // checked.  If there is an overflow on the low or high side, remember
  // it, otherwise compute the range [low, hi) bounding the new value.
  // See: InsertRangeTest above for the kinds of replacements possible.
  const APInt *C2;
  if (!match(Div->getOperand(1), m_APInt(C2)))
    return nullptr;

  // FIXME: If the operand types don't match the type of the divide
  // then don't attempt this transform. The code below doesn't have the
  // logic to deal with a signed divide and an unsigned compare (and
  // vice versa). This is because (x /s C2) <s C  produces different
  // results than (x /s C2) <u C or (x /u C2) <s C or even
  // (x /u C2) <u C.  Simply casting the operands and result won't
  // work. :(  The if statement below tests that condition and bails
  // if it finds it.
  bool DivIsSigned = Div->getOpcode() == Instruction::SDiv;
  if (!Cmp.isEquality() && DivIsSigned != Cmp.isSigned())
    return nullptr;

  // The ProdOV computation fails on divide by 0 and divide by -1. Cases with
  // INT_MIN will also fail if the divisor is 1. Although folds of all these
  // division-by-constant cases should be present, we can not assert that they
  // have happened before we reach this icmp instruction.
  if (C2->isNullValue() || C2->isOneValue() ||
      (DivIsSigned && C2->isAllOnesValue()))
    return nullptr;

  // Compute Prod = C * C2. We are essentially solving an equation of
  // form X / C2 = C. We solve for X by multiplying C2 and C.
  // By solving for X, we can turn this into a range check instead of computing
  // a divide.
  APInt Prod = C * *C2;

  // Determine if the product overflows by seeing if the product is not equal to
  // the divide. Make sure we do the same kind of divide as in the LHS
  // instruction that we're folding.
  bool ProdOV = (DivIsSigned ? Prod.sdiv(*C2) : Prod.udiv(*C2)) != C;

  ICmpInst::Predicate Pred = Cmp.getPredicate();

  // If the division is known to be exact, then there is no remainder from the
  // divide, so the covered range size is unit, otherwise it is the divisor.
  APInt RangeSize = Div->isExact() ? APInt(C2->getBitWidth(), 1) : *C2;

  // Figure out the interval that is being checked.  For example, a comparison
  // like "X /u 5 == 0" is really checking that X is in the interval [0, 5).
  // Compute this interval based on the constants involved and the signedness of
  // the compare/divide.  This computes a half-open interval, keeping track of
  // whether either value in the interval overflows.  After analysis each
  // overflow variable is set to 0 if it's corresponding bound variable is valid
  // -1 if overflowed off the bottom end, or +1 if overflowed off the top end.
  int LoOverflow = 0, HiOverflow = 0;
  APInt LoBound, HiBound;

  if (!DivIsSigned) {  // udiv
    // e.g. X/5 op 3  --> [15, 20)
    LoBound = Prod;
    HiOverflow = LoOverflow = ProdOV;
    if (!HiOverflow) {
      // If this is not an exact divide, then many values in the range collapse
      // to the same result value.
      HiOverflow = addWithOverflow(HiBound, LoBound, RangeSize, false);
    }
  } else if (C2->isStrictlyPositive()) { // Divisor is > 0.
    if (C.isNullValue()) {       // (X / pos) op 0
      // Can't overflow.  e.g.  X/2 op 0 --> [-1, 2)
      LoBound = -(RangeSize - 1);
      HiBound = RangeSize;
    } else if (C.isStrictlyPositive()) {   // (X / pos) op pos
      LoBound = Prod;     // e.g.   X/5 op 3 --> [15, 20)
      HiOverflow = LoOverflow = ProdOV;
      if (!HiOverflow)
        HiOverflow = addWithOverflow(HiBound, Prod, RangeSize, true);
    } else {                       // (X / pos) op neg
      // e.g. X/5 op -3  --> [-15-4, -15+1) --> [-19, -14)
      HiBound = Prod + 1;
      LoOverflow = HiOverflow = ProdOV ? -1 : 0;
      if (!LoOverflow) {
        APInt DivNeg = -RangeSize;
        LoOverflow = addWithOverflow(LoBound, HiBound, DivNeg, true) ? -1 : 0;
      }
    }
  } else if (C2->isNegative()) { // Divisor is < 0.
    if (Div->isExact())
      RangeSize.negate();
    if (C.isNullValue()) { // (X / neg) op 0
      // e.g. X/-5 op 0  --> [-4, 5)
      LoBound = RangeSize + 1;
      HiBound = -RangeSize;
      if (HiBound == *C2) {        // -INTMIN = INTMIN
        HiOverflow = 1;            // [INTMIN+1, overflow)
        HiBound = APInt();         // e.g. X/INTMIN = 0 --> X > INTMIN
      }
    } else if (C.isStrictlyPositive()) {   // (X / neg) op pos
      // e.g. X/-5 op 3  --> [-19, -14)
      HiBound = Prod + 1;
      HiOverflow = LoOverflow = ProdOV ? -1 : 0;
      if (!LoOverflow)
        LoOverflow = addWithOverflow(LoBound, HiBound, RangeSize, true) ? -1:0;
    } else {                       // (X / neg) op neg
      LoBound = Prod;       // e.g. X/-5 op -3  --> [15, 20)
      LoOverflow = HiOverflow = ProdOV;
      if (!HiOverflow)
        HiOverflow = subWithOverflow(HiBound, Prod, RangeSize, true);
    }

    // Dividing by a negative swaps the condition.  LT <-> GT
    Pred = ICmpInst::getSwappedPredicate(Pred);
  }

  Value *X = Div->getOperand(0);
  switch (Pred) {
    default: llvm_unreachable("Unhandled icmp opcode!");
    case ICmpInst::ICMP_EQ:
      if (LoOverflow && HiOverflow)
        return replaceInstUsesWith(Cmp, Builder.getFalse());
      if (HiOverflow)
        return new ICmpInst(DivIsSigned ? ICmpInst::ICMP_SGE :
                            ICmpInst::ICMP_UGE, X,
                            ConstantInt::get(Div->getType(), LoBound));
      if (LoOverflow)
        return new ICmpInst(DivIsSigned ? ICmpInst::ICMP_SLT :
                            ICmpInst::ICMP_ULT, X,
                            ConstantInt::get(Div->getType(), HiBound));
      return replaceInstUsesWith(
          Cmp, insertRangeTest(X, LoBound, HiBound, DivIsSigned, true));
    case ICmpInst::ICMP_NE:
      if (LoOverflow && HiOverflow)
        return replaceInstUsesWith(Cmp, Builder.getTrue());
      if (HiOverflow)
        return new ICmpInst(DivIsSigned ? ICmpInst::ICMP_SLT :
                            ICmpInst::ICMP_ULT, X,
                            ConstantInt::get(Div->getType(), LoBound));
      if (LoOverflow)
        return new ICmpInst(DivIsSigned ? ICmpInst::ICMP_SGE :
                            ICmpInst::ICMP_UGE, X,
                            ConstantInt::get(Div->getType(), HiBound));
      return replaceInstUsesWith(Cmp,
                                 insertRangeTest(X, LoBound, HiBound,
                                                 DivIsSigned, false));
    case ICmpInst::ICMP_ULT:
    case ICmpInst::ICMP_SLT:
      if (LoOverflow == +1)   // Low bound is greater than input range.
        return replaceInstUsesWith(Cmp, Builder.getTrue());
      if (LoOverflow == -1)   // Low bound is less than input range.
        return replaceInstUsesWith(Cmp, Builder.getFalse());
      return new ICmpInst(Pred, X, ConstantInt::get(Div->getType(), LoBound));
    case ICmpInst::ICMP_UGT:
    case ICmpInst::ICMP_SGT:
      if (HiOverflow == +1)       // High bound greater than input range.
        return replaceInstUsesWith(Cmp, Builder.getFalse());
      if (HiOverflow == -1)       // High bound less than input range.
        return replaceInstUsesWith(Cmp, Builder.getTrue());
      if (Pred == ICmpInst::ICMP_UGT)
        return new ICmpInst(ICmpInst::ICMP_UGE, X,
                            ConstantInt::get(Div->getType(), HiBound));
      return new ICmpInst(ICmpInst::ICMP_SGE, X,
                          ConstantInt::get(Div->getType(), HiBound));
  }

  return nullptr;
}

/// Fold icmp (sub X, Y), C.
Instruction *InstCombiner::foldICmpSubConstant(ICmpInst &Cmp,
                                               BinaryOperator *Sub,
                                               const APInt &C) {
  Value *X = Sub->getOperand(0), *Y = Sub->getOperand(1);
  ICmpInst::Predicate Pred = Cmp.getPredicate();

  // The following transforms are only worth it if the only user of the subtract
  // is the icmp.
  if (!Sub->hasOneUse())
    return nullptr;

  if (Sub->hasNoSignedWrap()) {
    // (icmp sgt (sub nsw X, Y), -1) -> (icmp sge X, Y)
    if (Pred == ICmpInst::ICMP_SGT && C.isAllOnesValue())
      return new ICmpInst(ICmpInst::ICMP_SGE, X, Y);

    // (icmp sgt (sub nsw X, Y), 0) -> (icmp sgt X, Y)
    if (Pred == ICmpInst::ICMP_SGT && C.isNullValue())
      return new ICmpInst(ICmpInst::ICMP_SGT, X, Y);

    // (icmp slt (sub nsw X, Y), 0) -> (icmp slt X, Y)
    if (Pred == ICmpInst::ICMP_SLT && C.isNullValue())
      return new ICmpInst(ICmpInst::ICMP_SLT, X, Y);

    // (icmp slt (sub nsw X, Y), 1) -> (icmp sle X, Y)
    if (Pred == ICmpInst::ICMP_SLT && C.isOneValue())
      return new ICmpInst(ICmpInst::ICMP_SLE, X, Y);
  }

  const APInt *C2;
  if (!match(X, m_APInt(C2)))
    return nullptr;

  // C2 - Y <u C -> (Y | (C - 1)) == C2
  //   iff (C2 & (C - 1)) == C - 1 and C is a power of 2
  if (Pred == ICmpInst::ICMP_ULT && C.isPowerOf2() &&
      (*C2 & (C - 1)) == (C - 1))
    return new ICmpInst(ICmpInst::ICMP_EQ, Builder.CreateOr(Y, C - 1), X);

  // C2 - Y >u C -> (Y | C) != C2
  //   iff C2 & C == C and C + 1 is a power of 2
  if (Pred == ICmpInst::ICMP_UGT && (C + 1).isPowerOf2() && (*C2 & C) == C)
    return new ICmpInst(ICmpInst::ICMP_NE, Builder.CreateOr(Y, C), X);

  return nullptr;
}

/// Fold icmp (add X, Y), C.
Instruction *InstCombiner::foldICmpAddConstant(ICmpInst &Cmp,
                                               BinaryOperator *Add,
                                               const APInt &C) {
  Value *Y = Add->getOperand(1);
  const APInt *C2;
  if (Cmp.isEquality() || !match(Y, m_APInt(C2)))
    return nullptr;

  // Fold icmp pred (add X, C2), C.
  Value *X = Add->getOperand(0);
  Type *Ty = Add->getType();
  CmpInst::Predicate Pred = Cmp.getPredicate();

  // If the add does not wrap, we can always adjust the compare by subtracting
  // the constants. Equality comparisons are handled elsewhere. SGE/SLE are
  // canonicalized to SGT/SLT.
  if (Add->hasNoSignedWrap() &&
      (Pred == ICmpInst::ICMP_SGT || Pred == ICmpInst::ICMP_SLT)) {
    bool Overflow;
    APInt NewC = C.ssub_ov(*C2, Overflow);
    // If there is overflow, the result must be true or false.
    // TODO: Can we assert there is no overflow because InstSimplify always
    // handles those cases?
    if (!Overflow)
      // icmp Pred (add nsw X, C2), C --> icmp Pred X, (C - C2)
      return new ICmpInst(Pred, X, ConstantInt::get(Ty, NewC));
  }

  auto CR = ConstantRange::makeExactICmpRegion(Pred, C).subtract(*C2);
  const APInt &Upper = CR.getUpper();
  const APInt &Lower = CR.getLower();
  if (Cmp.isSigned()) {
    if (Lower.isSignMask())
      return new ICmpInst(ICmpInst::ICMP_SLT, X, ConstantInt::get(Ty, Upper));
    if (Upper.isSignMask())
      return new ICmpInst(ICmpInst::ICMP_SGE, X, ConstantInt::get(Ty, Lower));
  } else {
    if (Lower.isMinValue())
      return new ICmpInst(ICmpInst::ICMP_ULT, X, ConstantInt::get(Ty, Upper));
    if (Upper.isMinValue())
      return new ICmpInst(ICmpInst::ICMP_UGE, X, ConstantInt::get(Ty, Lower));
  }

  if (!Add->hasOneUse())
    return nullptr;

  // X+C <u C2 -> (X & -C2) == C
  //   iff C & (C2-1) == 0
  //       C2 is a power of 2
  if (Pred == ICmpInst::ICMP_ULT && C.isPowerOf2() && (*C2 & (C - 1)) == 0)
    return new ICmpInst(ICmpInst::ICMP_EQ, Builder.CreateAnd(X, -C),
                        ConstantExpr::getNeg(cast<Constant>(Y)));

  // X+C >u C2 -> (X & ~C2) != C
  //   iff C & C2 == 0
  //       C2+1 is a power of 2
  if (Pred == ICmpInst::ICMP_UGT && (C + 1).isPowerOf2() && (*C2 & C) == 0)
    return new ICmpInst(ICmpInst::ICMP_NE, Builder.CreateAnd(X, ~C),
                        ConstantExpr::getNeg(cast<Constant>(Y)));

  return nullptr;
}

bool InstCombiner::matchThreeWayIntCompare(SelectInst *SI, Value *&LHS,
                                           Value *&RHS, ConstantInt *&Less,
                                           ConstantInt *&Equal,
                                           ConstantInt *&Greater) {
  // TODO: Generalize this to work with other comparison idioms or ensure
  // they get canonicalized into this form.

  // select i1 (a == b), i32 Equal, i32 (select i1 (a < b), i32 Less, i32
  // Greater), where Equal, Less and Greater are placeholders for any three
  // constants.
  ICmpInst::Predicate PredA, PredB;
  if (match(SI->getTrueValue(), m_ConstantInt(Equal)) &&
      match(SI->getCondition(), m_ICmp(PredA, m_Value(LHS), m_Value(RHS))) &&
      PredA == ICmpInst::ICMP_EQ &&
      match(SI->getFalseValue(),
            m_Select(m_ICmp(PredB, m_Specific(LHS), m_Specific(RHS)),
                     m_ConstantInt(Less), m_ConstantInt(Greater))) &&
      PredB == ICmpInst::ICMP_SLT) {
    return true;
  }
  return false;
}

Instruction *InstCombiner::foldICmpSelectConstant(ICmpInst &Cmp,
                                                  SelectInst *Select,
                                                  ConstantInt *C) {

  assert(C && "Cmp RHS should be a constant int!");
  // If we're testing a constant value against the result of a three way
  // comparison, the result can be expressed directly in terms of the
  // original values being compared.  Note: We could possibly be more
  // aggressive here and remove the hasOneUse test. The original select is
  // really likely to simplify or sink when we remove a test of the result.
  Value *OrigLHS, *OrigRHS;
  ConstantInt *C1LessThan, *C2Equal, *C3GreaterThan;
  if (Cmp.hasOneUse() &&
      matchThreeWayIntCompare(Select, OrigLHS, OrigRHS, C1LessThan, C2Equal,
                              C3GreaterThan)) {
    assert(C1LessThan && C2Equal && C3GreaterThan);

    bool TrueWhenLessThan =
        ConstantExpr::getCompare(Cmp.getPredicate(), C1LessThan, C)
            ->isAllOnesValue();
    bool TrueWhenEqual =
        ConstantExpr::getCompare(Cmp.getPredicate(), C2Equal, C)
            ->isAllOnesValue();
    bool TrueWhenGreaterThan =
        ConstantExpr::getCompare(Cmp.getPredicate(), C3GreaterThan, C)
            ->isAllOnesValue();

    // This generates the new instruction that will replace the original Cmp
    // Instruction. Instead of enumerating the various combinations when
    // TrueWhenLessThan, TrueWhenEqual and TrueWhenGreaterThan are true versus
    // false, we rely on chaining of ORs and future passes of InstCombine to
    // simplify the OR further (i.e. a s< b || a == b becomes a s<= b).

    // When none of the three constants satisfy the predicate for the RHS (C),
    // the entire original Cmp can be simplified to a false.
    Value *Cond = Builder.getFalse();
    if (TrueWhenLessThan)
      Cond = Builder.CreateOr(Cond, Builder.CreateICmp(ICmpInst::ICMP_SLT, OrigLHS, OrigRHS));
    if (TrueWhenEqual)
      Cond = Builder.CreateOr(Cond, Builder.CreateICmp(ICmpInst::ICMP_EQ, OrigLHS, OrigRHS));
    if (TrueWhenGreaterThan)
      Cond = Builder.CreateOr(Cond, Builder.CreateICmp(ICmpInst::ICMP_SGT, OrigLHS, OrigRHS));

    return replaceInstUsesWith(Cmp, Cond);
  }
  return nullptr;
}

Instruction *InstCombiner::foldICmpBitCastConstant(ICmpInst &Cmp,
                                                   BitCastInst *Bitcast,
                                                   const APInt &C) {
  // Folding: icmp <pred> iN X, C
  //  where X = bitcast <M x iK> (shufflevector <M x iK> %vec, undef, SC)) to iN
  //    and C is a splat of a K-bit pattern
  //    and SC is a constant vector = <C', C', C', ..., C'>
  // Into:
  //   %E = extractelement <M x iK> %vec, i32 C'
  //   icmp <pred> iK %E, trunc(C)
  if (!Bitcast->getType()->isIntegerTy() ||
      !Bitcast->getSrcTy()->isIntOrIntVectorTy())
    return nullptr;

  Value *BCIOp = Bitcast->getOperand(0);
  Value *Vec = nullptr;     // 1st vector arg of the shufflevector
  Constant *Mask = nullptr; // Mask arg of the shufflevector
  if (match(BCIOp,
            m_ShuffleVector(m_Value(Vec), m_Undef(), m_Constant(Mask)))) {
    // Check whether every element of Mask is the same constant
    if (auto *Elem = dyn_cast_or_null<ConstantInt>(Mask->getSplatValue())) {
      auto *VecTy = cast<VectorType>(BCIOp->getType());
      auto *EltTy = cast<IntegerType>(VecTy->getElementType());
      auto Pred = Cmp.getPredicate();
      if (C.isSplat(EltTy->getBitWidth())) {
        // Fold the icmp based on the value of C
        // If C is M copies of an iK sized bit pattern,
        // then:
        //   =>  %E = extractelement <N x iK> %vec, i32 Elem
        //       icmp <pred> iK %SplatVal, <pattern>
        Value *Extract = Builder.CreateExtractElement(Vec, Elem);
        Value *NewC = ConstantInt::get(EltTy, C.trunc(EltTy->getBitWidth()));
        return new ICmpInst(Pred, Extract, NewC);
      }
    }
  }
  return nullptr;
}

/// Try to fold integer comparisons with a constant operand: icmp Pred X, C
/// where X is some kind of instruction.
Instruction *InstCombiner::foldICmpInstWithConstant(ICmpInst &Cmp) {
  const APInt *C;
  if (!match(Cmp.getOperand(1), m_APInt(C)))
    return nullptr;

  if (auto *BO = dyn_cast<BinaryOperator>(Cmp.getOperand(0))) {
    switch (BO->getOpcode()) {
    case Instruction::Xor:
      if (Instruction *I = foldICmpXorConstant(Cmp, BO, *C))
        return I;
      break;
    case Instruction::And:
      if (Instruction *I = foldICmpAndConstant(Cmp, BO, *C))
        return I;
      break;
    case Instruction::Or:
      if (Instruction *I = foldICmpOrConstant(Cmp, BO, *C))
        return I;
      break;
    case Instruction::Mul:
      if (Instruction *I = foldICmpMulConstant(Cmp, BO, *C))
        return I;
      break;
    case Instruction::Shl:
      if (Instruction *I = foldICmpShlConstant(Cmp, BO, *C))
        return I;
      break;
    case Instruction::LShr:
    case Instruction::AShr:
      if (Instruction *I = foldICmpShrConstant(Cmp, BO, *C))
        return I;
      break;
    case Instruction::UDiv:
      if (Instruction *I = foldICmpUDivConstant(Cmp, BO, *C))
        return I;
      LLVM_FALLTHROUGH;
    case Instruction::SDiv:
      if (Instruction *I = foldICmpDivConstant(Cmp, BO, *C))
        return I;
      break;
    case Instruction::Sub:
      if (Instruction *I = foldICmpSubConstant(Cmp, BO, *C))
        return I;
      break;
    case Instruction::Add:
      if (Instruction *I = foldICmpAddConstant(Cmp, BO, *C))
        return I;
      break;
    default:
      break;
    }
    // TODO: These folds could be refactored to be part of the above calls.
    if (Instruction *I = foldICmpBinOpEqualityWithConstant(Cmp, BO, *C))
      return I;
  }

  // Match against CmpInst LHS being instructions other than binary operators.

  if (auto *SI = dyn_cast<SelectInst>(Cmp.getOperand(0))) {
    // For now, we only support constant integers while folding the
    // ICMP(SELECT)) pattern. We can extend this to support vector of integers
    // similar to the cases handled by binary ops above.
    if (ConstantInt *ConstRHS = dyn_cast<ConstantInt>(Cmp.getOperand(1)))
      if (Instruction *I = foldICmpSelectConstant(Cmp, SI, ConstRHS))
        return I;
  }

  if (auto *TI = dyn_cast<TruncInst>(Cmp.getOperand(0))) {
    if (Instruction *I = foldICmpTruncConstant(Cmp, TI, *C))
      return I;
  }

  if (auto *BCI = dyn_cast<BitCastInst>(Cmp.getOperand(0))) {
    if (Instruction *I = foldICmpBitCastConstant(Cmp, BCI, *C))
      return I;
  }

  if (Instruction *I = foldICmpIntrinsicWithConstant(Cmp, *C))
    return I;

  return nullptr;
}

/// Fold an icmp equality instruction with binary operator LHS and constant RHS:
/// icmp eq/ne BO, C.
Instruction *InstCombiner::foldICmpBinOpEqualityWithConstant(ICmpInst &Cmp,
                                                             BinaryOperator *BO,
                                                             const APInt &C) {
  // TODO: Some of these folds could work with arbitrary constants, but this
  // function is limited to scalar and vector splat constants.
  if (!Cmp.isEquality())
    return nullptr;

  ICmpInst::Predicate Pred = Cmp.getPredicate();
  bool isICMP_NE = Pred == ICmpInst::ICMP_NE;
  Constant *RHS = cast<Constant>(Cmp.getOperand(1));
  Value *BOp0 = BO->getOperand(0), *BOp1 = BO->getOperand(1);

  switch (BO->getOpcode()) {
  case Instruction::SRem:
    // If we have a signed (X % (2^c)) == 0, turn it into an unsigned one.
    if (C.isNullValue() && BO->hasOneUse()) {
      const APInt *BOC;
      if (match(BOp1, m_APInt(BOC)) && BOC->sgt(1) && BOC->isPowerOf2()) {
        Value *NewRem = Builder.CreateURem(BOp0, BOp1, BO->getName());
        return new ICmpInst(Pred, NewRem,
                            Constant::getNullValue(BO->getType()));
      }
    }
    break;
  case Instruction::Add: {
    // Replace ((add A, B) != C) with (A != C-B) if B & C are constants.
    const APInt *BOC;
    if (match(BOp1, m_APInt(BOC))) {
      if (BO->hasOneUse()) {
        Constant *SubC = ConstantExpr::getSub(RHS, cast<Constant>(BOp1));
        return new ICmpInst(Pred, BOp0, SubC);
      }
    } else if (C.isNullValue()) {
      // Replace ((add A, B) != 0) with (A != -B) if A or B is
      // efficiently invertible, or if the add has just this one use.
      if (Value *NegVal = dyn_castNegVal(BOp1))
        return new ICmpInst(Pred, BOp0, NegVal);
      if (Value *NegVal = dyn_castNegVal(BOp0))
        return new ICmpInst(Pred, NegVal, BOp1);
      if (BO->hasOneUse()) {
        Value *Neg = Builder.CreateNeg(BOp1);
        Neg->takeName(BO);
        return new ICmpInst(Pred, BOp0, Neg);
      }
    }
    break;
  }
  case Instruction::Xor:
    if (BO->hasOneUse()) {
      if (Constant *BOC = dyn_cast<Constant>(BOp1)) {
        // For the xor case, we can xor two constants together, eliminating
        // the explicit xor.
        return new ICmpInst(Pred, BOp0, ConstantExpr::getXor(RHS, BOC));
      } else if (C.isNullValue()) {
        // Replace ((xor A, B) != 0) with (A != B)
        return new ICmpInst(Pred, BOp0, BOp1);
      }
    }
    break;
  case Instruction::Sub:
    if (BO->hasOneUse()) {
      const APInt *BOC;
      if (match(BOp0, m_APInt(BOC))) {
        // Replace ((sub BOC, B) != C) with (B != BOC-C).
        Constant *SubC = ConstantExpr::getSub(cast<Constant>(BOp0), RHS);
        return new ICmpInst(Pred, BOp1, SubC);
      } else if (C.isNullValue()) {
        // Replace ((sub A, B) != 0) with (A != B).
        return new ICmpInst(Pred, BOp0, BOp1);
      }
    }
    break;
  case Instruction::Or: {
    const APInt *BOC;
    if (match(BOp1, m_APInt(BOC)) && BO->hasOneUse() && RHS->isAllOnesValue()) {
      // Comparing if all bits outside of a constant mask are set?
      // Replace (X | C) == -1 with (X & ~C) == ~C.
      // This removes the -1 constant.
      Constant *NotBOC = ConstantExpr::getNot(cast<Constant>(BOp1));
      Value *And = Builder.CreateAnd(BOp0, NotBOC);
      return new ICmpInst(Pred, And, NotBOC);
    }
    break;
  }
  case Instruction::And: {
    const APInt *BOC;
    if (match(BOp1, m_APInt(BOC))) {
      // If we have ((X & C) == C), turn it into ((X & C) != 0).
      if (C == *BOC && C.isPowerOf2())
        return new ICmpInst(isICMP_NE ? ICmpInst::ICMP_EQ : ICmpInst::ICMP_NE,
                            BO, Constant::getNullValue(RHS->getType()));

      // Don't perform the following transforms if the AND has multiple uses
      if (!BO->hasOneUse())
        break;

      // Replace (and X, (1 << size(X)-1) != 0) with x s< 0
      if (BOC->isSignMask()) {
        Constant *Zero = Constant::getNullValue(BOp0->getType());
        auto NewPred = isICMP_NE ? ICmpInst::ICMP_SLT : ICmpInst::ICMP_SGE;
        return new ICmpInst(NewPred, BOp0, Zero);
      }

      // ((X & ~7) == 0) --> X < 8
      if (C.isNullValue() && (~(*BOC) + 1).isPowerOf2()) {
        Constant *NegBOC = ConstantExpr::getNeg(cast<Constant>(BOp1));
        auto NewPred = isICMP_NE ? ICmpInst::ICMP_UGE : ICmpInst::ICMP_ULT;
        return new ICmpInst(NewPred, BOp0, NegBOC);
      }
    }
    break;
  }
  case Instruction::Mul:
    if (C.isNullValue() && BO->hasNoSignedWrap()) {
      const APInt *BOC;
      if (match(BOp1, m_APInt(BOC)) && !BOC->isNullValue()) {
        // The trivial case (mul X, 0) is handled by InstSimplify.
        // General case : (mul X, C) != 0 iff X != 0
        //                (mul X, C) == 0 iff X == 0
        return new ICmpInst(Pred, BOp0, Constant::getNullValue(RHS->getType()));
      }
    }
    break;
  case Instruction::UDiv:
    if (C.isNullValue()) {
      // (icmp eq/ne (udiv A, B), 0) -> (icmp ugt/ule i32 B, A)
      auto NewPred = isICMP_NE ? ICmpInst::ICMP_ULE : ICmpInst::ICMP_UGT;
      return new ICmpInst(NewPred, BOp1, BOp0);
    }
    break;
  default:
    break;
  }
  return nullptr;
}

/// Fold an icmp with LLVM intrinsic and constant operand: icmp Pred II, C.
Instruction *InstCombiner::foldICmpIntrinsicWithConstant(ICmpInst &Cmp,
                                                         const APInt &C) {
  IntrinsicInst *II = dyn_cast<IntrinsicInst>(Cmp.getOperand(0));
  if (!II || !Cmp.isEquality())
    return nullptr;

  // Handle icmp {eq|ne} <intrinsic>, Constant.
  Type *Ty = II->getType();
  switch (II->getIntrinsicID()) {
  case Intrinsic::bswap:
    Worklist.Add(II);
    Cmp.setOperand(0, II->getArgOperand(0));
    Cmp.setOperand(1, ConstantInt::get(Ty, C.byteSwap()));
    return &Cmp;

  case Intrinsic::ctlz:
  case Intrinsic::cttz:
    // ctz(A) == bitwidth(A)  ->  A == 0 and likewise for !=
    if (C == C.getBitWidth()) {
      Worklist.Add(II);
      Cmp.setOperand(0, II->getArgOperand(0));
      Cmp.setOperand(1, ConstantInt::getNullValue(Ty));
      return &Cmp;
    }
    break;

  case Intrinsic::ctpop: {
    // popcount(A) == 0  ->  A == 0 and likewise for !=
    // popcount(A) == bitwidth(A)  ->  A == -1 and likewise for !=
    bool IsZero = C.isNullValue();
    if (IsZero || C == C.getBitWidth()) {
      Worklist.Add(II);
      Cmp.setOperand(0, II->getArgOperand(0));
      auto *NewOp =
          IsZero ? Constant::getNullValue(Ty) : Constant::getAllOnesValue(Ty);
      Cmp.setOperand(1, NewOp);
      return &Cmp;
    }
    break;
  }
  default:
    break;
  }

  return nullptr;
}

/// Handle icmp with constant (but not simple integer constant) RHS.
Instruction *InstCombiner::foldICmpInstWithConstantNotInt(ICmpInst &I) {
  Value *Op0 = I.getOperand(0), *Op1 = I.getOperand(1);
  Constant *RHSC = dyn_cast<Constant>(Op1);
  Instruction *LHSI = dyn_cast<Instruction>(Op0);
  if (!RHSC || !LHSI)
    return nullptr;

  switch (LHSI->getOpcode()) {
  case Instruction::GetElementPtr:
    // icmp pred GEP (P, int 0, int 0, int 0), null -> icmp pred P, null
    if (RHSC->isNullValue() &&
        cast<GetElementPtrInst>(LHSI)->hasAllZeroIndices())
      return new ICmpInst(
          I.getPredicate(), LHSI->getOperand(0),
          Constant::getNullValue(LHSI->getOperand(0)->getType()));
    break;
  case Instruction::PHI:
    // Only fold icmp into the PHI if the phi and icmp are in the same
    // block.  If in the same block, we're encouraging jump threading.  If
    // not, we are just pessimizing the code by making an i1 phi.
    if (LHSI->getParent() == I.getParent())
      if (Instruction *NV = foldOpIntoPhi(I, cast<PHINode>(LHSI)))
        return NV;
    break;
  case Instruction::Select: {
    // If either operand of the select is a constant, we can fold the
    // comparison into the select arms, which will cause one to be
    // constant folded and the select turned into a bitwise or.
    Value *Op1 = nullptr, *Op2 = nullptr;
    ConstantInt *CI = nullptr;
    if (Constant *C = dyn_cast<Constant>(LHSI->getOperand(1))) {
      Op1 = ConstantExpr::getICmp(I.getPredicate(), C, RHSC);
      CI = dyn_cast<ConstantInt>(Op1);
    }
    if (Constant *C = dyn_cast<Constant>(LHSI->getOperand(2))) {
      Op2 = ConstantExpr::getICmp(I.getPredicate(), C, RHSC);
      CI = dyn_cast<ConstantInt>(Op2);
    }

    // We only want to perform this transformation if it will not lead to
    // additional code. This is true if either both sides of the select
    // fold to a constant (in which case the icmp is replaced with a select
    // which will usually simplify) or this is the only user of the
    // select (in which case we are trading a select+icmp for a simpler
    // select+icmp) or all uses of the select can be replaced based on
    // dominance information ("Global cases").
    bool Transform = false;
    if (Op1 && Op2)
      Transform = true;
    else if (Op1 || Op2) {
      // Local case
      if (LHSI->hasOneUse())
        Transform = true;
      // Global cases
      else if (CI && !CI->isZero())
        // When Op1 is constant try replacing select with second operand.
        // Otherwise Op2 is constant and try replacing select with first
        // operand.
        Transform =
            replacedSelectWithOperand(cast<SelectInst>(LHSI), &I, Op1 ? 2 : 1);
    }
    if (Transform) {
      if (!Op1)
        Op1 = Builder.CreateICmp(I.getPredicate(), LHSI->getOperand(1), RHSC,
                                 I.getName());
      if (!Op2)
        Op2 = Builder.CreateICmp(I.getPredicate(), LHSI->getOperand(2), RHSC,
                                 I.getName());
      return SelectInst::Create(LHSI->getOperand(0), Op1, Op2);
    }
    break;
  }
  case Instruction::IntToPtr:
    // icmp pred inttoptr(X), null -> icmp pred X, 0
    if (RHSC->isNullValue() &&
        DL.getIntPtrType(RHSC->getType()) == LHSI->getOperand(0)->getType())
      return new ICmpInst(
          I.getPredicate(), LHSI->getOperand(0),
          Constant::getNullValue(LHSI->getOperand(0)->getType()));
    break;

  case Instruction::Load:
    // Try to optimize things like "A[i] > 4" to index computations.
    if (GetElementPtrInst *GEP =
            dyn_cast<GetElementPtrInst>(LHSI->getOperand(0))) {
      if (GlobalVariable *GV = dyn_cast<GlobalVariable>(GEP->getOperand(0)))
        if (GV->isConstant() && GV->hasDefinitiveInitializer() &&
            !cast<LoadInst>(LHSI)->isVolatile())
          if (Instruction *Res = foldCmpLoadFromIndexedGlobal(GEP, GV, I))
            return Res;
    }
    break;
  }

  return nullptr;
}

/// Some comparisons can be simplified.
/// In this case, we are looking for comparisons that look like
/// a check for a lossy truncation.
/// Folds:
///   x & (-1 >> y) SrcPred x    to    x DstPred (-1 >> y)
/// The Mask can be a constant, too.
/// For some predicates, the operands are commutative.
/// For others, x can only be on a specific side.
static Value *foldICmpWithLowBitMaskedVal(ICmpInst &I,
                                          InstCombiner::BuilderTy &Builder) {
  ICmpInst::Predicate SrcPred;
  Value *X, *M;
  auto m_Mask = m_CombineOr(m_LShr(m_AllOnes(), m_Value()), m_LowBitMask());
  if (!match(&I, m_c_ICmp(SrcPred,
                          m_c_And(m_CombineAnd(m_Mask, m_Value(M)), m_Value(X)),
                          m_Deferred(X))))
    return nullptr;

  ICmpInst::Predicate DstPred;
  switch (SrcPred) {
  case ICmpInst::Predicate::ICMP_EQ:
    //  x & (-1 >> y) == x    ->    x u<= (-1 >> y)
    DstPred = ICmpInst::Predicate::ICMP_ULE;
    break;
  case ICmpInst::Predicate::ICMP_NE:
    //  x & (-1 >> y) != x    ->    x u> (-1 >> y)
    DstPred = ICmpInst::Predicate::ICMP_UGT;
    break;
  case ICmpInst::Predicate::ICMP_UGT:
    //  x u> x & (-1 >> y)    ->    x u> (-1 >> y)
    assert(X == I.getOperand(0) && "instsimplify took care of commut. variant");
    DstPred = ICmpInst::Predicate::ICMP_UGT;
    break;
  case ICmpInst::Predicate::ICMP_UGE:
    //  x & (-1 >> y) u>= x    ->    x u<= (-1 >> y)
    assert(X == I.getOperand(1) && "instsimplify took care of commut. variant");
    DstPred = ICmpInst::Predicate::ICMP_ULE;
    break;
  case ICmpInst::Predicate::ICMP_ULT:
    //  x & (-1 >> y) u< x    ->    x u> (-1 >> y)
    assert(X == I.getOperand(1) && "instsimplify took care of commut. variant");
    DstPred = ICmpInst::Predicate::ICMP_UGT;
    break;
  case ICmpInst::Predicate::ICMP_ULE:
    //  x u<= x & (-1 >> y)    ->    x u<= (-1 >> y)
    assert(X == I.getOperand(0) && "instsimplify took care of commut. variant");
    DstPred = ICmpInst::Predicate::ICMP_ULE;
    break;
  case ICmpInst::Predicate::ICMP_SGT:
    //  x s> x & (-1 >> y)    ->    x s> (-1 >> y)
    if (X != I.getOperand(0)) // X must be on LHS of comparison!
      return nullptr;         // Ignore the other case.
    DstPred = ICmpInst::Predicate::ICMP_SGT;
    break;
  case ICmpInst::Predicate::ICMP_SGE:
    //  x & (-1 >> y) s>= x    ->    x s<= (-1 >> y)
    if (X != I.getOperand(1)) // X must be on RHS of comparison!
      return nullptr;         // Ignore the other case.
    DstPred = ICmpInst::Predicate::ICMP_SLE;
    break;
  case ICmpInst::Predicate::ICMP_SLT:
    //  x & (-1 >> y) s< x    ->    x s> (-1 >> y)
    if (X != I.getOperand(1)) // X must be on RHS of comparison!
      return nullptr;         // Ignore the other case.
    DstPred = ICmpInst::Predicate::ICMP_SGT;
    break;
  case ICmpInst::Predicate::ICMP_SLE:
    //  x s<= x & (-1 >> y)    ->    x s<= (-1 >> y)
    if (X != I.getOperand(0)) // X must be on LHS of comparison!
      return nullptr;         // Ignore the other case.
    DstPred = ICmpInst::Predicate::ICMP_SLE;
    break;
  default:
    llvm_unreachable("All possible folds are handled.");
  }

  return Builder.CreateICmp(DstPred, X, M);
}

/// Some comparisons can be simplified.
/// In this case, we are looking for comparisons that look like
/// a check for a lossy signed truncation.
/// Folds:   (MaskedBits is a constant.)
///   ((%x << MaskedBits) a>> MaskedBits) SrcPred %x
/// Into:
///   (add %x, (1 << (KeptBits-1))) DstPred (1 << KeptBits)
/// Where  KeptBits = bitwidth(%x) - MaskedBits
static Value *
foldICmpWithTruncSignExtendedVal(ICmpInst &I,
                                 InstCombiner::BuilderTy &Builder) {
  ICmpInst::Predicate SrcPred;
  Value *X;
  const APInt *C0, *C1; // FIXME: non-splats, potentially with undef.
  // We are ok with 'shl' having multiple uses, but 'ashr' must be one-use.
  if (!match(&I, m_c_ICmp(SrcPred,
                          m_OneUse(m_AShr(m_Shl(m_Value(X), m_APInt(C0)),
                                          m_APInt(C1))),
                          m_Deferred(X))))
    return nullptr;

  // Potential handling of non-splats: for each element:
  //  * if both are undef, replace with constant 0.
  //    Because (1<<0) is OK and is 1, and ((1<<0)>>1) is also OK and is 0.
  //  * if both are not undef, and are different, bailout.
  //  * else, only one is undef, then pick the non-undef one.

  // The shift amount must be equal.
  if (*C0 != *C1)
    return nullptr;
  const APInt &MaskedBits = *C0;
  assert(MaskedBits != 0 && "shift by zero should be folded away already.");

  ICmpInst::Predicate DstPred;
  switch (SrcPred) {
  case ICmpInst::Predicate::ICMP_EQ:
    // ((%x << MaskedBits) a>> MaskedBits) == %x
    //   =>
    // (add %x, (1 << (KeptBits-1))) u< (1 << KeptBits)
    DstPred = ICmpInst::Predicate::ICMP_ULT;
    break;
  case ICmpInst::Predicate::ICMP_NE:
    // ((%x << MaskedBits) a>> MaskedBits) != %x
    //   =>
    // (add %x, (1 << (KeptBits-1))) u>= (1 << KeptBits)
    DstPred = ICmpInst::Predicate::ICMP_UGE;
    break;
  // FIXME: are more folds possible?
  default:
    return nullptr;
  }

  auto *XType = X->getType();
  const unsigned XBitWidth = XType->getScalarSizeInBits();
  const APInt BitWidth = APInt(XBitWidth, XBitWidth);
  assert(BitWidth.ugt(MaskedBits) && "shifts should leave some bits untouched");

  // KeptBits = bitwidth(%x) - MaskedBits
  const APInt KeptBits = BitWidth - MaskedBits;
  assert(KeptBits.ugt(0) && KeptBits.ult(BitWidth) && "unreachable");
  // ICmpCst = (1 << KeptBits)
  const APInt ICmpCst = APInt(XBitWidth, 1).shl(KeptBits);
  assert(ICmpCst.isPowerOf2());
  // AddCst = (1 << (KeptBits-1))
  const APInt AddCst = ICmpCst.lshr(1);
  assert(AddCst.ult(ICmpCst) && AddCst.isPowerOf2());

  // T0 = add %x, AddCst
  Value *T0 = Builder.CreateAdd(X, ConstantInt::get(XType, AddCst));
  // T1 = T0 DstPred ICmpCst
  Value *T1 = Builder.CreateICmp(DstPred, T0, ConstantInt::get(XType, ICmpCst));

  return T1;
}

/// Try to fold icmp (binop), X or icmp X, (binop).
/// TODO: A large part of this logic is duplicated in InstSimplify's
/// simplifyICmpWithBinOp(). We should be able to share that and avoid the code
/// duplication.
Instruction *InstCombiner::foldICmpBinOp(ICmpInst &I) {
  Value *Op0 = I.getOperand(0), *Op1 = I.getOperand(1);

  // Special logic for binary operators.
  BinaryOperator *BO0 = dyn_cast<BinaryOperator>(Op0);
  BinaryOperator *BO1 = dyn_cast<BinaryOperator>(Op1);
  if (!BO0 && !BO1)
    return nullptr;

  const CmpInst::Predicate Pred = I.getPredicate();
  bool NoOp0WrapProblem = false, NoOp1WrapProblem = false;
  if (BO0 && isa<OverflowingBinaryOperator>(BO0))
    NoOp0WrapProblem =
        ICmpInst::isEquality(Pred) ||
        (CmpInst::isUnsigned(Pred) && BO0->hasNoUnsignedWrap()) ||
        (CmpInst::isSigned(Pred) && BO0->hasNoSignedWrap());
  if (BO1 && isa<OverflowingBinaryOperator>(BO1))
    NoOp1WrapProblem =
        ICmpInst::isEquality(Pred) ||
        (CmpInst::isUnsigned(Pred) && BO1->hasNoUnsignedWrap()) ||
        (CmpInst::isSigned(Pred) && BO1->hasNoSignedWrap());

  // Analyze the case when either Op0 or Op1 is an add instruction.
  // Op0 = A + B (or A and B are null); Op1 = C + D (or C and D are null).
  Value *A = nullptr, *B = nullptr, *C = nullptr, *D = nullptr;
  if (BO0 && BO0->getOpcode() == Instruction::Add) {
    A = BO0->getOperand(0);
    B = BO0->getOperand(1);
  }
  if (BO1 && BO1->getOpcode() == Instruction::Add) {
    C = BO1->getOperand(0);
    D = BO1->getOperand(1);
  }

  // icmp (X+Y), X -> icmp Y, 0 for equalities or if there is no overflow.
  if ((A == Op1 || B == Op1) && NoOp0WrapProblem)
    return new ICmpInst(Pred, A == Op1 ? B : A,
                        Constant::getNullValue(Op1->getType()));

  // icmp X, (X+Y) -> icmp 0, Y for equalities or if there is no overflow.
  if ((C == Op0 || D == Op0) && NoOp1WrapProblem)
    return new ICmpInst(Pred, Constant::getNullValue(Op0->getType()),
                        C == Op0 ? D : C);

  // icmp (X+Y), (X+Z) -> icmp Y, Z for equalities or if there is no overflow.
  if (A && C && (A == C || A == D || B == C || B == D) && NoOp0WrapProblem &&
      NoOp1WrapProblem &&
      // Try not to increase register pressure.
      BO0->hasOneUse() && BO1->hasOneUse()) {
    // Determine Y and Z in the form icmp (X+Y), (X+Z).
    Value *Y, *Z;
    if (A == C) {
      // C + B == C + D  ->  B == D
      Y = B;
      Z = D;
    } else if (A == D) {
      // D + B == C + D  ->  B == C
      Y = B;
      Z = C;
    } else if (B == C) {
      // A + C == C + D  ->  A == D
      Y = A;
      Z = D;
    } else {
      assert(B == D);
      // A + D == C + D  ->  A == C
      Y = A;
      Z = C;
    }
    return new ICmpInst(Pred, Y, Z);
  }

  // icmp slt (X + -1), Y -> icmp sle X, Y
  if (A && NoOp0WrapProblem && Pred == CmpInst::ICMP_SLT &&
      match(B, m_AllOnes()))
    return new ICmpInst(CmpInst::ICMP_SLE, A, Op1);

  // icmp sge (X + -1), Y -> icmp sgt X, Y
  if (A && NoOp0WrapProblem && Pred == CmpInst::ICMP_SGE &&
      match(B, m_AllOnes()))
    return new ICmpInst(CmpInst::ICMP_SGT, A, Op1);

  // icmp sle (X + 1), Y -> icmp slt X, Y
  if (A && NoOp0WrapProblem && Pred == CmpInst::ICMP_SLE && match(B, m_One()))
    return new ICmpInst(CmpInst::ICMP_SLT, A, Op1);

  // icmp sgt (X + 1), Y -> icmp sge X, Y
  if (A && NoOp0WrapProblem && Pred == CmpInst::ICMP_SGT && match(B, m_One()))
    return new ICmpInst(CmpInst::ICMP_SGE, A, Op1);

  // icmp sgt X, (Y + -1) -> icmp sge X, Y
  if (C && NoOp1WrapProblem && Pred == CmpInst::ICMP_SGT &&
      match(D, m_AllOnes()))
    return new ICmpInst(CmpInst::ICMP_SGE, Op0, C);

  // icmp sle X, (Y + -1) -> icmp slt X, Y
  if (C && NoOp1WrapProblem && Pred == CmpInst::ICMP_SLE &&
      match(D, m_AllOnes()))
    return new ICmpInst(CmpInst::ICMP_SLT, Op0, C);

  // icmp sge X, (Y + 1) -> icmp sgt X, Y
  if (C && NoOp1WrapProblem && Pred == CmpInst::ICMP_SGE && match(D, m_One()))
    return new ICmpInst(CmpInst::ICMP_SGT, Op0, C);

  // icmp slt X, (Y + 1) -> icmp sle X, Y
  if (C && NoOp1WrapProblem && Pred == CmpInst::ICMP_SLT && match(D, m_One()))
    return new ICmpInst(CmpInst::ICMP_SLE, Op0, C);

  // TODO: The subtraction-related identities shown below also hold, but
  // canonicalization from (X -nuw 1) to (X + -1) means that the combinations
  // wouldn't happen even if they were implemented.
  //
  // icmp ult (X - 1), Y -> icmp ule X, Y
  // icmp uge (X - 1), Y -> icmp ugt X, Y
  // icmp ugt X, (Y - 1) -> icmp uge X, Y
  // icmp ule X, (Y - 1) -> icmp ult X, Y

  // icmp ule (X + 1), Y -> icmp ult X, Y
  if (A && NoOp0WrapProblem && Pred == CmpInst::ICMP_ULE && match(B, m_One()))
    return new ICmpInst(CmpInst::ICMP_ULT, A, Op1);

  // icmp ugt (X + 1), Y -> icmp uge X, Y
  if (A && NoOp0WrapProblem && Pred == CmpInst::ICMP_UGT && match(B, m_One()))
    return new ICmpInst(CmpInst::ICMP_UGE, A, Op1);

  // icmp uge X, (Y + 1) -> icmp ugt X, Y
  if (C && NoOp1WrapProblem && Pred == CmpInst::ICMP_UGE && match(D, m_One()))
    return new ICmpInst(CmpInst::ICMP_UGT, Op0, C);

  // icmp ult X, (Y + 1) -> icmp ule X, Y
  if (C && NoOp1WrapProblem && Pred == CmpInst::ICMP_ULT && match(D, m_One()))
    return new ICmpInst(CmpInst::ICMP_ULE, Op0, C);

  // if C1 has greater magnitude than C2:
  //  icmp (X + C1), (Y + C2) -> icmp (X + C3), Y
  //  s.t. C3 = C1 - C2
  //
  // if C2 has greater magnitude than C1:
  //  icmp (X + C1), (Y + C2) -> icmp X, (Y + C3)
  //  s.t. C3 = C2 - C1
  if (A && C && NoOp0WrapProblem && NoOp1WrapProblem &&
      (BO0->hasOneUse() || BO1->hasOneUse()) && !I.isUnsigned())
    if (ConstantInt *C1 = dyn_cast<ConstantInt>(B))
      if (ConstantInt *C2 = dyn_cast<ConstantInt>(D)) {
        const APInt &AP1 = C1->getValue();
        const APInt &AP2 = C2->getValue();
        if (AP1.isNegative() == AP2.isNegative()) {
          APInt AP1Abs = C1->getValue().abs();
          APInt AP2Abs = C2->getValue().abs();
          if (AP1Abs.uge(AP2Abs)) {
            ConstantInt *C3 = Builder.getInt(AP1 - AP2);
            Value *NewAdd = Builder.CreateNSWAdd(A, C3);
            return new ICmpInst(Pred, NewAdd, C);
          } else {
            ConstantInt *C3 = Builder.getInt(AP2 - AP1);
            Value *NewAdd = Builder.CreateNSWAdd(C, C3);
            return new ICmpInst(Pred, A, NewAdd);
          }
        }
      }

  // Analyze the case when either Op0 or Op1 is a sub instruction.
  // Op0 = A - B (or A and B are null); Op1 = C - D (or C and D are null).
  A = nullptr;
  B = nullptr;
  C = nullptr;
  D = nullptr;
  if (BO0 && BO0->getOpcode() == Instruction::Sub) {
    A = BO0->getOperand(0);
    B = BO0->getOperand(1);
  }
  if (BO1 && BO1->getOpcode() == Instruction::Sub) {
    C = BO1->getOperand(0);
    D = BO1->getOperand(1);
  }

  // icmp (X-Y), X -> icmp 0, Y for equalities or if there is no overflow.
  if (A == Op1 && NoOp0WrapProblem)
    return new ICmpInst(Pred, Constant::getNullValue(Op1->getType()), B);
  // icmp X, (X-Y) -> icmp Y, 0 for equalities or if there is no overflow.
  if (C == Op0 && NoOp1WrapProblem)
    return new ICmpInst(Pred, D, Constant::getNullValue(Op0->getType()));

  // (A - B) >u A --> A <u B
  if (A == Op1 && Pred == ICmpInst::ICMP_UGT)
    return new ICmpInst(ICmpInst::ICMP_ULT, A, B);
  // C <u (C - D) --> C <u D
  if (C == Op0 && Pred == ICmpInst::ICMP_ULT)
    return new ICmpInst(ICmpInst::ICMP_ULT, C, D);

  // icmp (Y-X), (Z-X) -> icmp Y, Z for equalities or if there is no overflow.
  if (B && D && B == D && NoOp0WrapProblem && NoOp1WrapProblem &&
      // Try not to increase register pressure.
      BO0->hasOneUse() && BO1->hasOneUse())
    return new ICmpInst(Pred, A, C);
  // icmp (X-Y), (X-Z) -> icmp Z, Y for equalities or if there is no overflow.
  if (A && C && A == C && NoOp0WrapProblem && NoOp1WrapProblem &&
      // Try not to increase register pressure.
      BO0->hasOneUse() && BO1->hasOneUse())
    return new ICmpInst(Pred, D, B);

  // icmp (0-X) < cst --> x > -cst
  if (NoOp0WrapProblem && ICmpInst::isSigned(Pred)) {
    Value *X;
    if (match(BO0, m_Neg(m_Value(X))))
      if (Constant *RHSC = dyn_cast<Constant>(Op1))
        if (RHSC->isNotMinSignedValue())
          return new ICmpInst(I.getSwappedPredicate(), X,
                              ConstantExpr::getNeg(RHSC));
  }

  BinaryOperator *SRem = nullptr;
  // icmp (srem X, Y), Y
  if (BO0 && BO0->getOpcode() == Instruction::SRem && Op1 == BO0->getOperand(1))
    SRem = BO0;
  // icmp Y, (srem X, Y)
  else if (BO1 && BO1->getOpcode() == Instruction::SRem &&
           Op0 == BO1->getOperand(1))
    SRem = BO1;
  if (SRem) {
    // We don't check hasOneUse to avoid increasing register pressure because
    // the value we use is the same value this instruction was already using.
    switch (SRem == BO0 ? ICmpInst::getSwappedPredicate(Pred) : Pred) {
    default:
      break;
    case ICmpInst::ICMP_EQ:
      return replaceInstUsesWith(I, ConstantInt::getFalse(I.getType()));
    case ICmpInst::ICMP_NE:
      return replaceInstUsesWith(I, ConstantInt::getTrue(I.getType()));
    case ICmpInst::ICMP_SGT:
    case ICmpInst::ICMP_SGE:
      return new ICmpInst(ICmpInst::ICMP_SGT, SRem->getOperand(1),
                          Constant::getAllOnesValue(SRem->getType()));
    case ICmpInst::ICMP_SLT:
    case ICmpInst::ICMP_SLE:
      return new ICmpInst(ICmpInst::ICMP_SLT, SRem->getOperand(1),
                          Constant::getNullValue(SRem->getType()));
    }
  }

  if (BO0 && BO1 && BO0->getOpcode() == BO1->getOpcode() && BO0->hasOneUse() &&
      BO1->hasOneUse() && BO0->getOperand(1) == BO1->getOperand(1)) {
    switch (BO0->getOpcode()) {
    default:
      break;
    case Instruction::Add:
    case Instruction::Sub:
    case Instruction::Xor: {
      if (I.isEquality()) // a+x icmp eq/ne b+x --> a icmp b
        return new ICmpInst(Pred, BO0->getOperand(0), BO1->getOperand(0));

      const APInt *C;
      if (match(BO0->getOperand(1), m_APInt(C))) {
        // icmp u/s (a ^ signmask), (b ^ signmask) --> icmp s/u a, b
        if (C->isSignMask()) {
          ICmpInst::Predicate NewPred =
              I.isSigned() ? I.getUnsignedPredicate() : I.getSignedPredicate();
          return new ICmpInst(NewPred, BO0->getOperand(0), BO1->getOperand(0));
        }

        // icmp u/s (a ^ maxsignval), (b ^ maxsignval) --> icmp s/u' a, b
        if (BO0->getOpcode() == Instruction::Xor && C->isMaxSignedValue()) {
          ICmpInst::Predicate NewPred =
              I.isSigned() ? I.getUnsignedPredicate() : I.getSignedPredicate();
          NewPred = I.getSwappedPredicate(NewPred);
          return new ICmpInst(NewPred, BO0->getOperand(0), BO1->getOperand(0));
        }
      }
      break;
    }
    case Instruction::Mul: {
      if (!I.isEquality())
        break;

      const APInt *C;
      if (match(BO0->getOperand(1), m_APInt(C)) && !C->isNullValue() &&
          !C->isOneValue()) {
        // icmp eq/ne (X * C), (Y * C) --> icmp (X & Mask), (Y & Mask)
        // Mask = -1 >> count-trailing-zeros(C).
        if (unsigned TZs = C->countTrailingZeros()) {
          Constant *Mask = ConstantInt::get(
              BO0->getType(),
              APInt::getLowBitsSet(C->getBitWidth(), C->getBitWidth() - TZs));
          Value *And1 = Builder.CreateAnd(BO0->getOperand(0), Mask);
          Value *And2 = Builder.CreateAnd(BO1->getOperand(0), Mask);
          return new ICmpInst(Pred, And1, And2);
        }
        // If there are no trailing zeros in the multiplier, just eliminate
        // the multiplies (no masking is needed):
        // icmp eq/ne (X * C), (Y * C) --> icmp eq/ne X, Y
        return new ICmpInst(Pred, BO0->getOperand(0), BO1->getOperand(0));
      }
      break;
    }
    case Instruction::UDiv:
    case Instruction::LShr:
      if (I.isSigned() || !BO0->isExact() || !BO1->isExact())
        break;
      return new ICmpInst(Pred, BO0->getOperand(0), BO1->getOperand(0));

    case Instruction::SDiv:
      if (!I.isEquality() || !BO0->isExact() || !BO1->isExact())
        break;
      return new ICmpInst(Pred, BO0->getOperand(0), BO1->getOperand(0));

    case Instruction::AShr:
      if (!BO0->isExact() || !BO1->isExact())
        break;
      return new ICmpInst(Pred, BO0->getOperand(0), BO1->getOperand(0));

    case Instruction::Shl: {
      bool NUW = BO0->hasNoUnsignedWrap() && BO1->hasNoUnsignedWrap();
      bool NSW = BO0->hasNoSignedWrap() && BO1->hasNoSignedWrap();
      if (!NUW && !NSW)
        break;
      if (!NSW && I.isSigned())
        break;
      return new ICmpInst(Pred, BO0->getOperand(0), BO1->getOperand(0));
    }
    }
  }

  if (BO0) {
    // Transform  A & (L - 1) `ult` L --> L != 0
    auto LSubOne = m_Add(m_Specific(Op1), m_AllOnes());
    auto BitwiseAnd = m_c_And(m_Value(), LSubOne);

    if (match(BO0, BitwiseAnd) && Pred == ICmpInst::ICMP_ULT) {
      auto *Zero = Constant::getNullValue(BO0->getType());
      return new ICmpInst(ICmpInst::ICMP_NE, Op1, Zero);
    }
  }

  if (Value *V = foldICmpWithLowBitMaskedVal(I, Builder))
    return replaceInstUsesWith(I, V);

  if (Value *V = foldICmpWithTruncSignExtendedVal(I, Builder))
    return replaceInstUsesWith(I, V);

  return nullptr;
}

/// Fold icmp Pred min|max(X, Y), X.
static Instruction *foldICmpWithMinMax(ICmpInst &Cmp) {
  ICmpInst::Predicate Pred = Cmp.getPredicate();
  Value *Op0 = Cmp.getOperand(0);
  Value *X = Cmp.getOperand(1);

  // Canonicalize minimum or maximum operand to LHS of the icmp.
  if (match(X, m_c_SMin(m_Specific(Op0), m_Value())) ||
      match(X, m_c_SMax(m_Specific(Op0), m_Value())) ||
      match(X, m_c_UMin(m_Specific(Op0), m_Value())) ||
      match(X, m_c_UMax(m_Specific(Op0), m_Value()))) {
    std::swap(Op0, X);
    Pred = Cmp.getSwappedPredicate();
  }

  Value *Y;
  if (match(Op0, m_c_SMin(m_Specific(X), m_Value(Y)))) {
    // smin(X, Y)  == X --> X s<= Y
    // smin(X, Y) s>= X --> X s<= Y
    if (Pred == CmpInst::ICMP_EQ || Pred == CmpInst::ICMP_SGE)
      return new ICmpInst(ICmpInst::ICMP_SLE, X, Y);

    // smin(X, Y) != X --> X s> Y
    // smin(X, Y) s< X --> X s> Y
    if (Pred == CmpInst::ICMP_NE || Pred == CmpInst::ICMP_SLT)
      return new ICmpInst(ICmpInst::ICMP_SGT, X, Y);

    // These cases should be handled in InstSimplify:
    // smin(X, Y) s<= X --> true
    // smin(X, Y) s> X --> false
    return nullptr;
  }

  if (match(Op0, m_c_SMax(m_Specific(X), m_Value(Y)))) {
    // smax(X, Y)  == X --> X s>= Y
    // smax(X, Y) s<= X --> X s>= Y
    if (Pred == CmpInst::ICMP_EQ || Pred == CmpInst::ICMP_SLE)
      return new ICmpInst(ICmpInst::ICMP_SGE, X, Y);

    // smax(X, Y) != X --> X s< Y
    // smax(X, Y) s> X --> X s< Y
    if (Pred == CmpInst::ICMP_NE || Pred == CmpInst::ICMP_SGT)
      return new ICmpInst(ICmpInst::ICMP_SLT, X, Y);

    // These cases should be handled in InstSimplify:
    // smax(X, Y) s>= X --> true
    // smax(X, Y) s< X --> false
    return nullptr;
  }

  if (match(Op0, m_c_UMin(m_Specific(X), m_Value(Y)))) {
    // umin(X, Y)  == X --> X u<= Y
    // umin(X, Y) u>= X --> X u<= Y
    if (Pred == CmpInst::ICMP_EQ || Pred == CmpInst::ICMP_UGE)
      return new ICmpInst(ICmpInst::ICMP_ULE, X, Y);

    // umin(X, Y) != X --> X u> Y
    // umin(X, Y) u< X --> X u> Y
    if (Pred == CmpInst::ICMP_NE || Pred == CmpInst::ICMP_ULT)
      return new ICmpInst(ICmpInst::ICMP_UGT, X, Y);

    // These cases should be handled in InstSimplify:
    // umin(X, Y) u<= X --> true
    // umin(X, Y) u> X --> false
    return nullptr;
  }

  if (match(Op0, m_c_UMax(m_Specific(X), m_Value(Y)))) {
    // umax(X, Y)  == X --> X u>= Y
    // umax(X, Y) u<= X --> X u>= Y
    if (Pred == CmpInst::ICMP_EQ || Pred == CmpInst::ICMP_ULE)
      return new ICmpInst(ICmpInst::ICMP_UGE, X, Y);

    // umax(X, Y) != X --> X u< Y
    // umax(X, Y) u> X --> X u< Y
    if (Pred == CmpInst::ICMP_NE || Pred == CmpInst::ICMP_UGT)
      return new ICmpInst(ICmpInst::ICMP_ULT, X, Y);

    // These cases should be handled in InstSimplify:
    // umax(X, Y) u>= X --> true
    // umax(X, Y) u< X --> false
    return nullptr;
  }

  return nullptr;
}

Instruction *InstCombiner::foldICmpEquality(ICmpInst &I) {
  if (!I.isEquality())
    return nullptr;

  Value *Op0 = I.getOperand(0), *Op1 = I.getOperand(1);
  const CmpInst::Predicate Pred = I.getPredicate();
  Value *A, *B, *C, *D;
  if (match(Op0, m_Xor(m_Value(A), m_Value(B)))) {
    if (A == Op1 || B == Op1) { // (A^B) == A  ->  B == 0
      Value *OtherVal = A == Op1 ? B : A;
      return new ICmpInst(Pred, OtherVal, Constant::getNullValue(A->getType()));
    }

    if (match(Op1, m_Xor(m_Value(C), m_Value(D)))) {
      // A^c1 == C^c2 --> A == C^(c1^c2)
      ConstantInt *C1, *C2;
      if (match(B, m_ConstantInt(C1)) && match(D, m_ConstantInt(C2)) &&
          Op1->hasOneUse()) {
        Constant *NC = Builder.getInt(C1->getValue() ^ C2->getValue());
        Value *Xor = Builder.CreateXor(C, NC);
        return new ICmpInst(Pred, A, Xor);
      }

      // A^B == A^D -> B == D
      if (A == C)
        return new ICmpInst(Pred, B, D);
      if (A == D)
        return new ICmpInst(Pred, B, C);
      if (B == C)
        return new ICmpInst(Pred, A, D);
      if (B == D)
        return new ICmpInst(Pred, A, C);
    }
  }

  if (match(Op1, m_Xor(m_Value(A), m_Value(B))) && (A == Op0 || B == Op0)) {
    // A == (A^B)  ->  B == 0
    Value *OtherVal = A == Op0 ? B : A;
    return new ICmpInst(Pred, OtherVal, Constant::getNullValue(A->getType()));
  }

  // (X&Z) == (Y&Z) -> (X^Y) & Z == 0
  if (match(Op0, m_OneUse(m_And(m_Value(A), m_Value(B)))) &&
      match(Op1, m_OneUse(m_And(m_Value(C), m_Value(D))))) {
    Value *X = nullptr, *Y = nullptr, *Z = nullptr;

    if (A == C) {
      X = B;
      Y = D;
      Z = A;
    } else if (A == D) {
      X = B;
      Y = C;
      Z = A;
    } else if (B == C) {
      X = A;
      Y = D;
      Z = B;
    } else if (B == D) {
      X = A;
      Y = C;
      Z = B;
    }

    if (X) { // Build (X^Y) & Z
      Op1 = Builder.CreateXor(X, Y);
      Op1 = Builder.CreateAnd(Op1, Z);
      I.setOperand(0, Op1);
      I.setOperand(1, Constant::getNullValue(Op1->getType()));
      return &I;
    }
  }

  // Transform (zext A) == (B & (1<<X)-1) --> A == (trunc B)
  // and       (B & (1<<X)-1) == (zext A) --> A == (trunc B)
  ConstantInt *Cst1;
  if ((Op0->hasOneUse() && match(Op0, m_ZExt(m_Value(A))) &&
       match(Op1, m_And(m_Value(B), m_ConstantInt(Cst1)))) ||
      (Op1->hasOneUse() && match(Op0, m_And(m_Value(B), m_ConstantInt(Cst1))) &&
       match(Op1, m_ZExt(m_Value(A))))) {
    APInt Pow2 = Cst1->getValue() + 1;
    if (Pow2.isPowerOf2() && isa<IntegerType>(A->getType()) &&
        Pow2.logBase2() == cast<IntegerType>(A->getType())->getBitWidth())
      return new ICmpInst(Pred, A, Builder.CreateTrunc(B, A->getType()));
  }

  // (A >> C) == (B >> C) --> (A^B) u< (1 << C)
  // For lshr and ashr pairs.
  if ((match(Op0, m_OneUse(m_LShr(m_Value(A), m_ConstantInt(Cst1)))) &&
       match(Op1, m_OneUse(m_LShr(m_Value(B), m_Specific(Cst1))))) ||
      (match(Op0, m_OneUse(m_AShr(m_Value(A), m_ConstantInt(Cst1)))) &&
       match(Op1, m_OneUse(m_AShr(m_Value(B), m_Specific(Cst1)))))) {
    unsigned TypeBits = Cst1->getBitWidth();
    unsigned ShAmt = (unsigned)Cst1->getLimitedValue(TypeBits);
    if (ShAmt < TypeBits && ShAmt != 0) {
      ICmpInst::Predicate NewPred =
          Pred == ICmpInst::ICMP_NE ? ICmpInst::ICMP_UGE : ICmpInst::ICMP_ULT;
      Value *Xor = Builder.CreateXor(A, B, I.getName() + ".unshifted");
      APInt CmpVal = APInt::getOneBitSet(TypeBits, ShAmt);
      return new ICmpInst(NewPred, Xor, Builder.getInt(CmpVal));
    }
  }

  // (A << C) == (B << C) --> ((A^B) & (~0U >> C)) == 0
  if (match(Op0, m_OneUse(m_Shl(m_Value(A), m_ConstantInt(Cst1)))) &&
      match(Op1, m_OneUse(m_Shl(m_Value(B), m_Specific(Cst1))))) {
    unsigned TypeBits = Cst1->getBitWidth();
    unsigned ShAmt = (unsigned)Cst1->getLimitedValue(TypeBits);
    if (ShAmt < TypeBits && ShAmt != 0) {
      Value *Xor = Builder.CreateXor(A, B, I.getName() + ".unshifted");
      APInt AndVal = APInt::getLowBitsSet(TypeBits, TypeBits - ShAmt);
      Value *And = Builder.CreateAnd(Xor, Builder.getInt(AndVal),
                                      I.getName() + ".mask");
      return new ICmpInst(Pred, And, Constant::getNullValue(Cst1->getType()));
    }
  }

  // Transform "icmp eq (trunc (lshr(X, cst1)), cst" to
  // "icmp (and X, mask), cst"
  uint64_t ShAmt = 0;
  if (Op0->hasOneUse() &&
      match(Op0, m_Trunc(m_OneUse(m_LShr(m_Value(A), m_ConstantInt(ShAmt))))) &&
      match(Op1, m_ConstantInt(Cst1)) &&
      // Only do this when A has multiple uses.  This is most important to do
      // when it exposes other optimizations.
      !A->hasOneUse()) {
    unsigned ASize = cast<IntegerType>(A->getType())->getPrimitiveSizeInBits();

    if (ShAmt < ASize) {
      APInt MaskV =
          APInt::getLowBitsSet(ASize, Op0->getType()->getPrimitiveSizeInBits());
      MaskV <<= ShAmt;

      APInt CmpV = Cst1->getValue().zext(ASize);
      CmpV <<= ShAmt;

      Value *Mask = Builder.CreateAnd(A, Builder.getInt(MaskV));
      return new ICmpInst(Pred, Mask, Builder.getInt(CmpV));
    }
  }

  // If both operands are byte-swapped or bit-reversed, just compare the
  // original values.
  // TODO: Move this to a function similar to foldICmpIntrinsicWithConstant()
  // and handle more intrinsics.
  if ((match(Op0, m_BSwap(m_Value(A))) && match(Op1, m_BSwap(m_Value(B)))) ||
      (match(Op0, m_BitReverse(m_Value(A))) &&
       match(Op1, m_BitReverse(m_Value(B)))))
    return new ICmpInst(Pred, A, B);

  return nullptr;
}

/// Handle icmp (cast x to y), (cast/cst). We only handle extending casts so
/// far.
Instruction *InstCombiner::foldICmpWithCastAndCast(ICmpInst &ICmp) {
  const CastInst *LHSCI = cast<CastInst>(ICmp.getOperand(0));
  Value *LHSCIOp        = LHSCI->getOperand(0);
  Type *SrcTy     = LHSCIOp->getType();
  Type *DestTy    = LHSCI->getType();
  Value *RHSCIOp;

  // Turn icmp (ptrtoint x), (ptrtoint/c) into a compare of the input if the
  // integer type is the same size as the pointer type.
  const auto& CompatibleSizes = [&](Type* SrcTy, Type* DestTy) -> bool {
    if (isa<VectorType>(SrcTy)) {
      SrcTy = cast<VectorType>(SrcTy)->getElementType();
      DestTy = cast<VectorType>(DestTy)->getElementType();
    }
    return DL.getPointerTypeSizeInBits(SrcTy) == DestTy->getIntegerBitWidth();
  };
  if (LHSCI->getOpcode() == Instruction::PtrToInt &&
      CompatibleSizes(SrcTy, DestTy)) {
    Value *RHSOp = nullptr;
    if (auto *RHSC = dyn_cast<PtrToIntOperator>(ICmp.getOperand(1))) {
      Value *RHSCIOp = RHSC->getOperand(0);
      if (RHSCIOp->getType()->getPointerAddressSpace() ==
          LHSCIOp->getType()->getPointerAddressSpace()) {
        RHSOp = RHSC->getOperand(0);
        // If the pointer types don't match, insert a bitcast.
        if (LHSCIOp->getType() != RHSOp->getType())
          RHSOp = Builder.CreateBitCast(RHSOp, LHSCIOp->getType());
      }
    } else if (auto *RHSC = dyn_cast<Constant>(ICmp.getOperand(1))) {
      RHSOp = ConstantExpr::getIntToPtr(RHSC, SrcTy);
    }

    if (RHSOp)
      return new ICmpInst(ICmp.getPredicate(), LHSCIOp, RHSOp);
  }

  // The code below only handles extension cast instructions, so far.
  // Enforce this.
  if (LHSCI->getOpcode() != Instruction::ZExt &&
      LHSCI->getOpcode() != Instruction::SExt)
    return nullptr;

  bool isSignedExt = LHSCI->getOpcode() == Instruction::SExt;
  bool isSignedCmp = ICmp.isSigned();

  if (auto *CI = dyn_cast<CastInst>(ICmp.getOperand(1))) {
    // Not an extension from the same type?
    RHSCIOp = CI->getOperand(0);
    if (RHSCIOp->getType() != LHSCIOp->getType())
      return nullptr;

    // If the signedness of the two casts doesn't agree (i.e. one is a sext
    // and the other is a zext), then we can't handle this.
    if (CI->getOpcode() != LHSCI->getOpcode())
      return nullptr;

    // Deal with equality cases early.
    if (ICmp.isEquality())
      return new ICmpInst(ICmp.getPredicate(), LHSCIOp, RHSCIOp);

    // A signed comparison of sign extended values simplifies into a
    // signed comparison.
    if (isSignedCmp && isSignedExt)
      return new ICmpInst(ICmp.getPredicate(), LHSCIOp, RHSCIOp);

    // The other three cases all fold into an unsigned comparison.
    return new ICmpInst(ICmp.getUnsignedPredicate(), LHSCIOp, RHSCIOp);
  }

  // If we aren't dealing with a constant on the RHS, exit early.
  auto *C = dyn_cast<Constant>(ICmp.getOperand(1));
  if (!C)
    return nullptr;

  // Compute the constant that would happen if we truncated to SrcTy then
  // re-extended to DestTy.
  Constant *Res1 = ConstantExpr::getTrunc(C, SrcTy);
  Constant *Res2 = ConstantExpr::getCast(LHSCI->getOpcode(), Res1, DestTy);

  // If the re-extended constant didn't change...
  if (Res2 == C) {
    // Deal with equality cases early.
    if (ICmp.isEquality())
      return new ICmpInst(ICmp.getPredicate(), LHSCIOp, Res1);

    // A signed comparison of sign extended values simplifies into a
    // signed comparison.
    if (isSignedExt && isSignedCmp)
      return new ICmpInst(ICmp.getPredicate(), LHSCIOp, Res1);

    // The other three cases all fold into an unsigned comparison.
    return new ICmpInst(ICmp.getUnsignedPredicate(), LHSCIOp, Res1);
  }

  // The re-extended constant changed, partly changed (in the case of a vector),
  // or could not be determined to be equal (in the case of a constant
  // expression), so the constant cannot be represented in the shorter type.
  // Consequently, we cannot emit a simple comparison.
  // All the cases that fold to true or false will have already been handled
  // by SimplifyICmpInst, so only deal with the tricky case.

  if (isSignedCmp || !isSignedExt || !isa<ConstantInt>(C))
    return nullptr;

  // Evaluate the comparison for LT (we invert for GT below). LE and GE cases
  // should have been folded away previously and not enter in here.

  // We're performing an unsigned comp with a sign extended value.
  // This is true if the input is >= 0. [aka >s -1]
  Constant *NegOne = Constant::getAllOnesValue(SrcTy);
  Value *Result = Builder.CreateICmpSGT(LHSCIOp, NegOne, ICmp.getName());

  // Finally, return the value computed.
  if (ICmp.getPredicate() == ICmpInst::ICMP_ULT)
    return replaceInstUsesWith(ICmp, Result);

  assert(ICmp.getPredicate() == ICmpInst::ICMP_UGT && "ICmp should be folded!");
  return BinaryOperator::CreateNot(Result);
}

bool InstCombiner::OptimizeOverflowCheck(OverflowCheckFlavor OCF, Value *LHS,
                                         Value *RHS, Instruction &OrigI,
                                         Value *&Result, Constant *&Overflow) {
  if (OrigI.isCommutative() && isa<Constant>(LHS) && !isa<Constant>(RHS))
    std::swap(LHS, RHS);

  auto SetResult = [&](Value *OpResult, Constant *OverflowVal, bool ReuseName) {
    Result = OpResult;
    Overflow = OverflowVal;
    if (ReuseName)
      Result->takeName(&OrigI);
    return true;
  };

  // If the overflow check was an add followed by a compare, the insertion point
  // may be pointing to the compare.  We want to insert the new instructions
  // before the add in case there are uses of the add between the add and the
  // compare.
  Builder.SetInsertPoint(&OrigI);

  switch (OCF) {
  case OCF_INVALID:
    llvm_unreachable("bad overflow check kind!");

  case OCF_UNSIGNED_ADD: {
    OverflowResult OR = computeOverflowForUnsignedAdd(LHS, RHS, &OrigI);
    if (OR == OverflowResult::NeverOverflows)
      return SetResult(Builder.CreateNUWAdd(LHS, RHS), Builder.getFalse(),
                       true);

    if (OR == OverflowResult::AlwaysOverflows)
      return SetResult(Builder.CreateAdd(LHS, RHS), Builder.getTrue(), true);

    // Fall through uadd into sadd
    LLVM_FALLTHROUGH;
  }
  case OCF_SIGNED_ADD: {
    // X + 0 -> {X, false}
    if (match(RHS, m_Zero()))
      return SetResult(LHS, Builder.getFalse(), false);

    // We can strength reduce this signed add into a regular add if we can prove
    // that it will never overflow.
    if (OCF == OCF_SIGNED_ADD)
      if (willNotOverflowSignedAdd(LHS, RHS, OrigI))
        return SetResult(Builder.CreateNSWAdd(LHS, RHS), Builder.getFalse(),
                         true);
    break;
  }

  case OCF_UNSIGNED_SUB:
  case OCF_SIGNED_SUB: {
    // X - 0 -> {X, false}
    if (match(RHS, m_Zero()))
      return SetResult(LHS, Builder.getFalse(), false);

    if (OCF == OCF_SIGNED_SUB) {
      if (willNotOverflowSignedSub(LHS, RHS, OrigI))
        return SetResult(Builder.CreateNSWSub(LHS, RHS), Builder.getFalse(),
                         true);
    } else {
      if (willNotOverflowUnsignedSub(LHS, RHS, OrigI))
        return SetResult(Builder.CreateNUWSub(LHS, RHS), Builder.getFalse(),
                         true);
    }
    break;
  }

  case OCF_UNSIGNED_MUL: {
    OverflowResult OR = computeOverflowForUnsignedMul(LHS, RHS, &OrigI);
    if (OR == OverflowResult::NeverOverflows)
      return SetResult(Builder.CreateNUWMul(LHS, RHS), Builder.getFalse(),
                       true);
    if (OR == OverflowResult::AlwaysOverflows)
      return SetResult(Builder.CreateMul(LHS, RHS), Builder.getTrue(), true);
    LLVM_FALLTHROUGH;
  }
  case OCF_SIGNED_MUL:
    // X * undef -> undef
    if (isa<UndefValue>(RHS))
      return SetResult(RHS, UndefValue::get(Builder.getInt1Ty()), false);

    // X * 0 -> {0, false}
    if (match(RHS, m_Zero()))
      return SetResult(RHS, Builder.getFalse(), false);

    // X * 1 -> {X, false}
    if (match(RHS, m_One()))
      return SetResult(LHS, Builder.getFalse(), false);

    if (OCF == OCF_SIGNED_MUL)
      if (willNotOverflowSignedMul(LHS, RHS, OrigI))
        return SetResult(Builder.CreateNSWMul(LHS, RHS), Builder.getFalse(),
                         true);
    break;
  }

  return false;
}

/// Recognize and process idiom involving test for multiplication
/// overflow.
///
/// The caller has matched a pattern of the form:
///   I = cmp u (mul(zext A, zext B), V
/// The function checks if this is a test for overflow and if so replaces
/// multiplication with call to 'mul.with.overflow' intrinsic.
///
/// \param I Compare instruction.
/// \param MulVal Result of 'mult' instruction.  It is one of the arguments of
///               the compare instruction.  Must be of integer type.
/// \param OtherVal The other argument of compare instruction.
/// \returns Instruction which must replace the compare instruction, NULL if no
///          replacement required.
static Instruction *processUMulZExtIdiom(ICmpInst &I, Value *MulVal,
                                         Value *OtherVal, InstCombiner &IC) {
  // Don't bother doing this transformation for pointers, don't do it for
  // vectors.
  if (!isa<IntegerType>(MulVal->getType()))
    return nullptr;

  assert(I.getOperand(0) == MulVal || I.getOperand(1) == MulVal);
  assert(I.getOperand(0) == OtherVal || I.getOperand(1) == OtherVal);
  auto *MulInstr = dyn_cast<Instruction>(MulVal);
  if (!MulInstr)
    return nullptr;
  assert(MulInstr->getOpcode() == Instruction::Mul);

  auto *LHS = cast<ZExtOperator>(MulInstr->getOperand(0)),
       *RHS = cast<ZExtOperator>(MulInstr->getOperand(1));
  assert(LHS->getOpcode() == Instruction::ZExt);
  assert(RHS->getOpcode() == Instruction::ZExt);
  Value *A = LHS->getOperand(0), *B = RHS->getOperand(0);

  // Calculate type and width of the result produced by mul.with.overflow.
  Type *TyA = A->getType(), *TyB = B->getType();
  unsigned WidthA = TyA->getPrimitiveSizeInBits(),
           WidthB = TyB->getPrimitiveSizeInBits();
  unsigned MulWidth;
  Type *MulType;
  if (WidthB > WidthA) {
    MulWidth = WidthB;
    MulType = TyB;
  } else {
    MulWidth = WidthA;
    MulType = TyA;
  }

  // In order to replace the original mul with a narrower mul.with.overflow,
  // all uses must ignore upper bits of the product.  The number of used low
  // bits must be not greater than the width of mul.with.overflow.
  if (MulVal->hasNUsesOrMore(2))
    for (User *U : MulVal->users()) {
      if (U == &I)
        continue;
      if (TruncInst *TI = dyn_cast<TruncInst>(U)) {
        // Check if truncation ignores bits above MulWidth.
        unsigned TruncWidth = TI->getType()->getPrimitiveSizeInBits();
        if (TruncWidth > MulWidth)
          return nullptr;
      } else if (BinaryOperator *BO = dyn_cast<BinaryOperator>(U)) {
        // Check if AND ignores bits above MulWidth.
        if (BO->getOpcode() != Instruction::And)
          return nullptr;
        if (ConstantInt *CI = dyn_cast<ConstantInt>(BO->getOperand(1))) {
          const APInt &CVal = CI->getValue();
          if (CVal.getBitWidth() - CVal.countLeadingZeros() > MulWidth)
            return nullptr;
        } else {
          // In this case we could have the operand of the binary operation
          // being defined in another block, and performing the replacement
          // could break the dominance relation.
          return nullptr;
        }
      } else {
        // Other uses prohibit this transformation.
        return nullptr;
      }
    }

  // Recognize patterns
  switch (I.getPredicate()) {
  case ICmpInst::ICMP_EQ:
  case ICmpInst::ICMP_NE:
    // Recognize pattern:
    //   mulval = mul(zext A, zext B)
    //   cmp eq/neq mulval, zext trunc mulval
    if (ZExtInst *Zext = dyn_cast<ZExtInst>(OtherVal))
      if (Zext->hasOneUse()) {
        Value *ZextArg = Zext->getOperand(0);
        if (TruncInst *Trunc = dyn_cast<TruncInst>(ZextArg))
          if (Trunc->getType()->getPrimitiveSizeInBits() == MulWidth)
            break; //Recognized
      }

    // Recognize pattern:
    //   mulval = mul(zext A, zext B)
    //   cmp eq/neq mulval, and(mulval, mask), mask selects low MulWidth bits.
    ConstantInt *CI;
    Value *ValToMask;
    if (match(OtherVal, m_And(m_Value(ValToMask), m_ConstantInt(CI)))) {
      if (ValToMask != MulVal)
        return nullptr;
      const APInt &CVal = CI->getValue() + 1;
      if (CVal.isPowerOf2()) {
        unsigned MaskWidth = CVal.logBase2();
        if (MaskWidth == MulWidth)
          break; // Recognized
      }
    }
    return nullptr;

  case ICmpInst::ICMP_UGT:
    // Recognize pattern:
    //   mulval = mul(zext A, zext B)
    //   cmp ugt mulval, max
    if (ConstantInt *CI = dyn_cast<ConstantInt>(OtherVal)) {
      APInt MaxVal = APInt::getMaxValue(MulWidth);
      MaxVal = MaxVal.zext(CI->getBitWidth());
      if (MaxVal.eq(CI->getValue()))
        break; // Recognized
    }
    return nullptr;

  case ICmpInst::ICMP_UGE:
    // Recognize pattern:
    //   mulval = mul(zext A, zext B)
    //   cmp uge mulval, max+1
    if (ConstantInt *CI = dyn_cast<ConstantInt>(OtherVal)) {
      APInt MaxVal = APInt::getOneBitSet(CI->getBitWidth(), MulWidth);
      if (MaxVal.eq(CI->getValue()))
        break; // Recognized
    }
    return nullptr;

  case ICmpInst::ICMP_ULE:
    // Recognize pattern:
    //   mulval = mul(zext A, zext B)
    //   cmp ule mulval, max
    if (ConstantInt *CI = dyn_cast<ConstantInt>(OtherVal)) {
      APInt MaxVal = APInt::getMaxValue(MulWidth);
      MaxVal = MaxVal.zext(CI->getBitWidth());
      if (MaxVal.eq(CI->getValue()))
        break; // Recognized
    }
    return nullptr;

  case ICmpInst::ICMP_ULT:
    // Recognize pattern:
    //   mulval = mul(zext A, zext B)
    //   cmp ule mulval, max + 1
    if (ConstantInt *CI = dyn_cast<ConstantInt>(OtherVal)) {
      APInt MaxVal = APInt::getOneBitSet(CI->getBitWidth(), MulWidth);
      if (MaxVal.eq(CI->getValue()))
        break; // Recognized
    }
    return nullptr;

  default:
    return nullptr;
  }

  InstCombiner::BuilderTy &Builder = IC.Builder;
  Builder.SetInsertPoint(MulInstr);

  // Replace: mul(zext A, zext B) --> mul.with.overflow(A, B)
  Value *MulA = A, *MulB = B;
  if (WidthA < MulWidth)
    MulA = Builder.CreateZExt(A, MulType);
  if (WidthB < MulWidth)
    MulB = Builder.CreateZExt(B, MulType);
  Value *F = Intrinsic::getDeclaration(I.getModule(),
                                       Intrinsic::umul_with_overflow, MulType);
  CallInst *Call = Builder.CreateCall(F, {MulA, MulB}, "umul");
  IC.Worklist.Add(MulInstr);

  // If there are uses of mul result other than the comparison, we know that
  // they are truncation or binary AND. Change them to use result of
  // mul.with.overflow and adjust properly mask/size.
  if (MulVal->hasNUsesOrMore(2)) {
    Value *Mul = Builder.CreateExtractValue(Call, 0, "umul.value");
    for (auto UI = MulVal->user_begin(), UE = MulVal->user_end(); UI != UE;) {
      User *U = *UI++;
      if (U == &I || U == OtherVal)
        continue;
      if (TruncInst *TI = dyn_cast<TruncInst>(U)) {
        if (TI->getType()->getPrimitiveSizeInBits() == MulWidth)
          IC.replaceInstUsesWith(*TI, Mul);
        else
          TI->setOperand(0, Mul);
      } else if (BinaryOperator *BO = dyn_cast<BinaryOperator>(U)) {
        assert(BO->getOpcode() == Instruction::And);
        // Replace (mul & mask) --> zext (mul.with.overflow & short_mask)
        ConstantInt *CI = cast<ConstantInt>(BO->getOperand(1));
        APInt ShortMask = CI->getValue().trunc(MulWidth);
        Value *ShortAnd = Builder.CreateAnd(Mul, ShortMask);
        Instruction *Zext =
            cast<Instruction>(Builder.CreateZExt(ShortAnd, BO->getType()));
        IC.Worklist.Add(Zext);
        IC.replaceInstUsesWith(*BO, Zext);
      } else {
        llvm_unreachable("Unexpected Binary operation");
      }
      IC.Worklist.Add(cast<Instruction>(U));
    }
  }
  if (isa<Instruction>(OtherVal))
    IC.Worklist.Add(cast<Instruction>(OtherVal));

  // The original icmp gets replaced with the overflow value, maybe inverted
  // depending on predicate.
  bool Inverse = false;
  switch (I.getPredicate()) {
  case ICmpInst::ICMP_NE:
    break;
  case ICmpInst::ICMP_EQ:
    Inverse = true;
    break;
  case ICmpInst::ICMP_UGT:
  case ICmpInst::ICMP_UGE:
    if (I.getOperand(0) == MulVal)
      break;
    Inverse = true;
    break;
  case ICmpInst::ICMP_ULT:
  case ICmpInst::ICMP_ULE:
    if (I.getOperand(1) == MulVal)
      break;
    Inverse = true;
    break;
  default:
    llvm_unreachable("Unexpected predicate");
  }
  if (Inverse) {
    Value *Res = Builder.CreateExtractValue(Call, 1);
    return BinaryOperator::CreateNot(Res);
  }

  return ExtractValueInst::Create(Call, 1);
}

/// When performing a comparison against a constant, it is possible that not all
/// the bits in the LHS are demanded. This helper method computes the mask that
/// IS demanded.
static APInt getDemandedBitsLHSMask(ICmpInst &I, unsigned BitWidth) {
  const APInt *RHS;
  if (!match(I.getOperand(1), m_APInt(RHS)))
    return APInt::getAllOnesValue(BitWidth);

  // If this is a normal comparison, it demands all bits. If it is a sign bit
  // comparison, it only demands the sign bit.
  bool UnusedBit;
  if (isSignBitCheck(I.getPredicate(), *RHS, UnusedBit))
    return APInt::getSignMask(BitWidth);

  switch (I.getPredicate()) {
  // For a UGT comparison, we don't care about any bits that
  // correspond to the trailing ones of the comparand.  The value of these
  // bits doesn't impact the outcome of the comparison, because any value
  // greater than the RHS must differ in a bit higher than these due to carry.
  case ICmpInst::ICMP_UGT:
    return APInt::getBitsSetFrom(BitWidth, RHS->countTrailingOnes());

  // Similarly, for a ULT comparison, we don't care about the trailing zeros.
  // Any value less than the RHS must differ in a higher bit because of carries.
  case ICmpInst::ICMP_ULT:
    return APInt::getBitsSetFrom(BitWidth, RHS->countTrailingZeros());

  default:
    return APInt::getAllOnesValue(BitWidth);
  }
}

/// Check if the order of \p Op0 and \p Op1 as operands in an ICmpInst
/// should be swapped.
/// The decision is based on how many times these two operands are reused
/// as subtract operands and their positions in those instructions.
/// The rationale is that several architectures use the same instruction for
/// both subtract and cmp. Thus, it is better if the order of those operands
/// match.
/// \return true if Op0 and Op1 should be swapped.
static bool swapMayExposeCSEOpportunities(const Value *Op0, const Value *Op1) {
  // Filter out pointer values as those cannot appear directly in subtract.
  // FIXME: we may want to go through inttoptrs or bitcasts.
  if (Op0->getType()->isPointerTy())
    return false;
  // If a subtract already has the same operands as a compare, swapping would be
  // bad. If a subtract has the same operands as a compare but in reverse order,
  // then swapping is good.
  int GoodToSwap = 0;
  for (const User *U : Op0->users()) {
    if (match(U, m_Sub(m_Specific(Op1), m_Specific(Op0))))
      GoodToSwap++;
    else if (match(U, m_Sub(m_Specific(Op0), m_Specific(Op1))))
      GoodToSwap--;
  }
  return GoodToSwap > 0;
}

/// Check that one use is in the same block as the definition and all
/// other uses are in blocks dominated by a given block.
///
/// \param DI Definition
/// \param UI Use
/// \param DB Block that must dominate all uses of \p DI outside
///           the parent block
/// \return true when \p UI is the only use of \p DI in the parent block
/// and all other uses of \p DI are in blocks dominated by \p DB.
///
bool InstCombiner::dominatesAllUses(const Instruction *DI,
                                    const Instruction *UI,
                                    const BasicBlock *DB) const {
  assert(DI && UI && "Instruction not defined\n");
  // Ignore incomplete definitions.
  if (!DI->getParent())
    return false;
  // DI and UI must be in the same block.
  if (DI->getParent() != UI->getParent())
    return false;
  // Protect from self-referencing blocks.
  if (DI->getParent() == DB)
    return false;
  for (const User *U : DI->users()) {
    auto *Usr = cast<Instruction>(U);
    if (Usr != UI && !DT.dominates(DB, Usr->getParent()))
      return false;
  }
  return true;
}

/// Return true when the instruction sequence within a block is select-cmp-br.
static bool isChainSelectCmpBranch(const SelectInst *SI) {
  const BasicBlock *BB = SI->getParent();
  if (!BB)
    return false;
  auto *BI = dyn_cast_or_null<BranchInst>(BB->getTerminator());
  if (!BI || BI->getNumSuccessors() != 2)
    return false;
  auto *IC = dyn_cast<ICmpInst>(BI->getCondition());
  if (!IC || (IC->getOperand(0) != SI && IC->getOperand(1) != SI))
    return false;
  return true;
}

/// True when a select result is replaced by one of its operands
/// in select-icmp sequence. This will eventually result in the elimination
/// of the select.
///
/// \param SI    Select instruction
/// \param Icmp  Compare instruction
/// \param SIOpd Operand that replaces the select
///
/// Notes:
/// - The replacement is global and requires dominator information
/// - The caller is responsible for the actual replacement
///
/// Example:
///
/// entry:
///  %4 = select i1 %3, %C* %0, %C* null
///  %5 = icmp eq %C* %4, null
///  br i1 %5, label %9, label %7
///  ...
///  ; <label>:7                                       ; preds = %entry
///  %8 = getelementptr inbounds %C* %4, i64 0, i32 0
///  ...
///
/// can be transformed to
///
///  %5 = icmp eq %C* %0, null
///  %6 = select i1 %3, i1 %5, i1 true
///  br i1 %6, label %9, label %7
///  ...
///  ; <label>:7                                       ; preds = %entry
///  %8 = getelementptr inbounds %C* %0, i64 0, i32 0  // replace by %0!
///
/// Similar when the first operand of the select is a constant or/and
/// the compare is for not equal rather than equal.
///
/// NOTE: The function is only called when the select and compare constants
/// are equal, the optimization can work only for EQ predicates. This is not a
/// major restriction since a NE compare should be 'normalized' to an equal
/// compare, which usually happens in the combiner and test case
/// select-cmp-br.ll checks for it.
bool InstCombiner::replacedSelectWithOperand(SelectInst *SI,
                                             const ICmpInst *Icmp,
                                             const unsigned SIOpd) {
  assert((SIOpd == 1 || SIOpd == 2) && "Invalid select operand!");
  if (isChainSelectCmpBranch(SI) && Icmp->getPredicate() == ICmpInst::ICMP_EQ) {
    BasicBlock *Succ = SI->getParent()->getTerminator()->getSuccessor(1);
    // The check for the single predecessor is not the best that can be
    // done. But it protects efficiently against cases like when SI's
    // home block has two successors, Succ and Succ1, and Succ1 predecessor
    // of Succ. Then SI can't be replaced by SIOpd because the use that gets
    // replaced can be reached on either path. So the uniqueness check
    // guarantees that the path all uses of SI (outside SI's parent) are on
    // is disjoint from all other paths out of SI. But that information
    // is more expensive to compute, and the trade-off here is in favor
    // of compile-time. It should also be noticed that we check for a single
    // predecessor and not only uniqueness. This to handle the situation when
    // Succ and Succ1 points to the same basic block.
    if (Succ->getSinglePredecessor() && dominatesAllUses(SI, Icmp, Succ)) {
      NumSel++;
      SI->replaceUsesOutsideBlock(SI->getOperand(SIOpd), SI->getParent());
      return true;
    }
  }
  return false;
}

/// Try to fold the comparison based on range information we can get by checking
/// whether bits are known to be zero or one in the inputs.
Instruction *InstCombiner::foldICmpUsingKnownBits(ICmpInst &I) {
  Value *Op0 = I.getOperand(0), *Op1 = I.getOperand(1);
  Type *Ty = Op0->getType();
  ICmpInst::Predicate Pred = I.getPredicate();

  // Get scalar or pointer size.
  unsigned BitWidth = Ty->isIntOrIntVectorTy()
                          ? Ty->getScalarSizeInBits()
                          : DL.getIndexTypeSizeInBits(Ty->getScalarType());

  if (!BitWidth)
    return nullptr;

  KnownBits Op0Known(BitWidth);
  KnownBits Op1Known(BitWidth);

  if (SimplifyDemandedBits(&I, 0,
                           getDemandedBitsLHSMask(I, BitWidth),
                           Op0Known, 0))
    return &I;

  if (SimplifyDemandedBits(&I, 1, APInt::getAllOnesValue(BitWidth),
                           Op1Known, 0))
    return &I;

  // Given the known and unknown bits, compute a range that the LHS could be
  // in.  Compute the Min, Max and RHS values based on the known bits. For the
  // EQ and NE we use unsigned values.
  APInt Op0Min(BitWidth, 0), Op0Max(BitWidth, 0);
  APInt Op1Min(BitWidth, 0), Op1Max(BitWidth, 0);
  if (I.isSigned()) {
    computeSignedMinMaxValuesFromKnownBits(Op0Known, Op0Min, Op0Max);
    computeSignedMinMaxValuesFromKnownBits(Op1Known, Op1Min, Op1Max);
  } else {
    computeUnsignedMinMaxValuesFromKnownBits(Op0Known, Op0Min, Op0Max);
    computeUnsignedMinMaxValuesFromKnownBits(Op1Known, Op1Min, Op1Max);
  }

  // If Min and Max are known to be the same, then SimplifyDemandedBits figured
  // out that the LHS or RHS is a constant. Constant fold this now, so that
  // code below can assume that Min != Max.
  if (!isa<Constant>(Op0) && Op0Min == Op0Max)
    return new ICmpInst(Pred, ConstantExpr::getIntegerValue(Ty, Op0Min), Op1);
  if (!isa<Constant>(Op1) && Op1Min == Op1Max)
    return new ICmpInst(Pred, Op0, ConstantExpr::getIntegerValue(Ty, Op1Min));

  // Based on the range information we know about the LHS, see if we can
  // simplify this comparison.  For example, (x&4) < 8 is always true.
  switch (Pred) {
  default:
    llvm_unreachable("Unknown icmp opcode!");
  case ICmpInst::ICMP_EQ:
  case ICmpInst::ICMP_NE: {
    if (Op0Max.ult(Op1Min) || Op0Min.ugt(Op1Max)) {
      return Pred == CmpInst::ICMP_EQ
                 ? replaceInstUsesWith(I, ConstantInt::getFalse(I.getType()))
                 : replaceInstUsesWith(I, ConstantInt::getTrue(I.getType()));
    }

    // If all bits are known zero except for one, then we know at most one bit
    // is set. If the comparison is against zero, then this is a check to see if
    // *that* bit is set.
    APInt Op0KnownZeroInverted = ~Op0Known.Zero;
    if (Op1Known.isZero()) {
      // If the LHS is an AND with the same constant, look through it.
      Value *LHS = nullptr;
      const APInt *LHSC;
      if (!match(Op0, m_And(m_Value(LHS), m_APInt(LHSC))) ||
          *LHSC != Op0KnownZeroInverted)
        LHS = Op0;

      Value *X;
      if (match(LHS, m_Shl(m_One(), m_Value(X)))) {
        APInt ValToCheck = Op0KnownZeroInverted;
        Type *XTy = X->getType();
        if (ValToCheck.isPowerOf2()) {
          // ((1 << X) & 8) == 0 -> X != 3
          // ((1 << X) & 8) != 0 -> X == 3
          auto *CmpC = ConstantInt::get(XTy, ValToCheck.countTrailingZeros());
          auto NewPred = ICmpInst::getInversePredicate(Pred);
          return new ICmpInst(NewPred, X, CmpC);
        } else if ((++ValToCheck).isPowerOf2()) {
          // ((1 << X) & 7) == 0 -> X >= 3
          // ((1 << X) & 7) != 0 -> X  < 3
          auto *CmpC = ConstantInt::get(XTy, ValToCheck.countTrailingZeros());
          auto NewPred =
              Pred == CmpInst::ICMP_EQ ? CmpInst::ICMP_UGE : CmpInst::ICMP_ULT;
          return new ICmpInst(NewPred, X, CmpC);
        }
      }

      // Check if the LHS is 8 >>u x and the result is a power of 2 like 1.
      const APInt *CI;
      if (Op0KnownZeroInverted.isOneValue() &&
          match(LHS, m_LShr(m_Power2(CI), m_Value(X)))) {
        // ((8 >>u X) & 1) == 0 -> X != 3
        // ((8 >>u X) & 1) != 0 -> X == 3
        unsigned CmpVal = CI->countTrailingZeros();
        auto NewPred = ICmpInst::getInversePredicate(Pred);
        return new ICmpInst(NewPred, X, ConstantInt::get(X->getType(), CmpVal));
      }
    }
    break;
  }
  case ICmpInst::ICMP_ULT: {
    if (Op0Max.ult(Op1Min)) // A <u B -> true if max(A) < min(B)
      return replaceInstUsesWith(I, ConstantInt::getTrue(I.getType()));
    if (Op0Min.uge(Op1Max)) // A <u B -> false if min(A) >= max(B)
      return replaceInstUsesWith(I, ConstantInt::getFalse(I.getType()));
    if (Op1Min == Op0Max) // A <u B -> A != B if max(A) == min(B)
      return new ICmpInst(ICmpInst::ICMP_NE, Op0, Op1);

    const APInt *CmpC;
    if (match(Op1, m_APInt(CmpC))) {
      // A <u C -> A == C-1 if min(A)+1 == C
      if (*CmpC == Op0Min + 1)
        return new ICmpInst(ICmpInst::ICMP_EQ, Op0,
                            ConstantInt::get(Op1->getType(), *CmpC - 1));
      // X <u C --> X == 0, if the number of zero bits in the bottom of X
      // exceeds the log2 of C.
      if (Op0Known.countMinTrailingZeros() >= CmpC->ceilLogBase2())
        return new ICmpInst(ICmpInst::ICMP_EQ, Op0,
                            Constant::getNullValue(Op1->getType()));
    }
    break;
  }
  case ICmpInst::ICMP_UGT: {
    if (Op0Min.ugt(Op1Max)) // A >u B -> true if min(A) > max(B)
      return replaceInstUsesWith(I, ConstantInt::getTrue(I.getType()));
    if (Op0Max.ule(Op1Min)) // A >u B -> false if max(A) <= max(B)
      return replaceInstUsesWith(I, ConstantInt::getFalse(I.getType()));
    if (Op1Max == Op0Min) // A >u B -> A != B if min(A) == max(B)
      return new ICmpInst(ICmpInst::ICMP_NE, Op0, Op1);

    const APInt *CmpC;
    if (match(Op1, m_APInt(CmpC))) {
      // A >u C -> A == C+1 if max(a)-1 == C
      if (*CmpC == Op0Max - 1)
        return new ICmpInst(ICmpInst::ICMP_EQ, Op0,
                            ConstantInt::get(Op1->getType(), *CmpC + 1));
      // X >u C --> X != 0, if the number of zero bits in the bottom of X
      // exceeds the log2 of C.
      if (Op0Known.countMinTrailingZeros() >= CmpC->getActiveBits())