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
|
/*
* Copyright © 2014 Connor Abbott
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice (including the next
* paragraph) shall be included in all copies or substantial portions of the
* Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
* IN THE SOFTWARE.
*
* Authors:
* Connor Abbott (cwabbott0@gmail.com)
*
*/
#ifndef NIR_H
#define NIR_H
#include "util/hash_table.h"
#include "compiler/glsl/list.h"
#include "GL/gl.h" /* GLenum */
#include "util/list.h"
#include "util/ralloc.h"
#include "util/set.h"
#include "util/bitscan.h"
#include "util/bitset.h"
#include "util/macros.h"
#include "compiler/nir_types.h"
#include "compiler/shader_enums.h"
#include "compiler/shader_info.h"
#include <stdio.h>
#ifndef NDEBUG
#include "util/debug.h"
#endif /* NDEBUG */
#include "nir_opcodes.h"
#if defined(_WIN32) && !defined(snprintf)
#define snprintf _snprintf
#endif
#ifdef __cplusplus
extern "C" {
#endif
#define NIR_FALSE 0u
#define NIR_TRUE (~0u)
#define NIR_MAX_VEC_COMPONENTS 4
typedef uint8_t nir_component_mask_t;
/** Defines a cast function
*
* This macro defines a cast function from in_type to out_type where
* out_type is some structure type that contains a field of type out_type.
*
* Note that you have to be a bit careful as the generated cast function
* destroys constness.
*/
#define NIR_DEFINE_CAST(name, in_type, out_type, field, \
type_field, type_value) \
static inline out_type * \
name(const in_type *parent) \
{ \
assert(parent && parent->type_field == type_value); \
return exec_node_data(out_type, parent, field); \
}
struct nir_function;
struct nir_shader;
struct nir_instr;
struct nir_builder;
/**
* Description of built-in state associated with a uniform
*
* \sa nir_variable::state_slots
*/
typedef struct {
gl_state_index16 tokens[STATE_LENGTH];
int swizzle;
} nir_state_slot;
typedef enum {
nir_var_shader_in = (1 << 0),
nir_var_shader_out = (1 << 1),
nir_var_global = (1 << 2),
nir_var_local = (1 << 3),
nir_var_uniform = (1 << 4),
nir_var_shader_storage = (1 << 5),
nir_var_system_value = (1 << 6),
nir_var_shared = (1 << 8),
nir_var_all = ~0,
} nir_variable_mode;
/**
* Rounding modes.
*/
typedef enum {
nir_rounding_mode_undef = 0,
nir_rounding_mode_rtne = 1, /* round to nearest even */
nir_rounding_mode_ru = 2, /* round up */
nir_rounding_mode_rd = 3, /* round down */
nir_rounding_mode_rtz = 4, /* round towards zero */
} nir_rounding_mode;
typedef union {
float f32[NIR_MAX_VEC_COMPONENTS];
double f64[NIR_MAX_VEC_COMPONENTS];
int8_t i8[NIR_MAX_VEC_COMPONENTS];
uint8_t u8[NIR_MAX_VEC_COMPONENTS];
int16_t i16[NIR_MAX_VEC_COMPONENTS];
uint16_t u16[NIR_MAX_VEC_COMPONENTS];
int32_t i32[NIR_MAX_VEC_COMPONENTS];
uint32_t u32[NIR_MAX_VEC_COMPONENTS];
int64_t i64[NIR_MAX_VEC_COMPONENTS];
uint64_t u64[NIR_MAX_VEC_COMPONENTS];
} nir_const_value;
typedef struct nir_constant {
/**
* Value of the constant.
*
* The field used to back the values supplied by the constant is determined
* by the type associated with the \c nir_variable. Constants may be
* scalars, vectors, or matrices.
*/
nir_const_value values[NIR_MAX_VEC_COMPONENTS];
/* we could get this from the var->type but makes clone *much* easier to
* not have to care about the type.
*/
unsigned num_elements;
/* Array elements / Structure Fields */
struct nir_constant **elements;
} nir_constant;
/**
* \brief Layout qualifiers for gl_FragDepth.
*
* The AMD/ARB_conservative_depth extensions allow gl_FragDepth to be redeclared
* with a layout qualifier.
*/
typedef enum {
nir_depth_layout_none, /**< No depth layout is specified. */
nir_depth_layout_any,
nir_depth_layout_greater,
nir_depth_layout_less,
nir_depth_layout_unchanged
} nir_depth_layout;
/**
* Enum keeping track of how a variable was declared.
*/
typedef enum {
/**
* Normal declaration.
*/
nir_var_declared_normally = 0,
/**
* Variable is implicitly generated by the compiler and should not be
* visible via the API.
*/
nir_var_hidden,
} nir_var_declaration_type;
/**
* Either a uniform, global variable, shader input, or shader output. Based on
* ir_variable - it should be easy to translate between the two.
*/
typedef struct nir_variable {
struct exec_node node;
/**
* Declared type of the variable
*/
const struct glsl_type *type;
/**
* Declared name of the variable
*/
char *name;
struct nir_variable_data {
/**
* Storage class of the variable.
*
* \sa nir_variable_mode
*/
nir_variable_mode mode;
/**
* Is the variable read-only?
*
* This is set for variables declared as \c const, shader inputs,
* and uniforms.
*/
unsigned read_only:1;
unsigned centroid:1;
unsigned sample:1;
unsigned patch:1;
unsigned invariant:1;
/**
* When separate shader programs are enabled, only input/outputs between
* the stages of a multi-stage separate program can be safely removed
* from the shader interface. Other input/outputs must remains active.
*
* This is also used to make sure xfb varyings that are unused by the
* fragment shader are not removed.
*/
unsigned always_active_io:1;
/**
* Interpolation mode for shader inputs / outputs
*
* \sa glsl_interp_mode
*/
unsigned interpolation:2;
/**
* \name ARB_fragment_coord_conventions
* @{
*/
unsigned origin_upper_left:1;
unsigned pixel_center_integer:1;
/*@}*/
/**
* If non-zero, then this variable may be packed along with other variables
* into a single varying slot, so this offset should be applied when
* accessing components. For example, an offset of 1 means that the x
* component of this variable is actually stored in component y of the
* location specified by \c location.
*/
unsigned location_frac:2;
/**
* If true, this variable represents an array of scalars that should
* be tightly packed. In other words, consecutive array elements
* should be stored one component apart, rather than one slot apart.
*/
unsigned compact:1;
/**
* Whether this is a fragment shader output implicitly initialized with
* the previous contents of the specified render target at the
* framebuffer location corresponding to this shader invocation.
*/
unsigned fb_fetch_output:1;
/**
* Non-zero if this variable is considered bindless as defined by
* ARB_bindless_texture.
*/
unsigned bindless:1;
/**
* Was an explicit binding set in the shader?
*/
unsigned explicit_binding:1;
/**
* Was a transfer feedback buffer set in the shader?
*/
unsigned explicit_xfb_buffer:1;
/**
* Was a transfer feedback stride set in the shader?
*/
unsigned explicit_xfb_stride:1;
/**
* Was an explicit offset set in the shader?
*/
unsigned explicit_offset:1;
/**
* \brief Layout qualifier for gl_FragDepth.
*
* This is not equal to \c ir_depth_layout_none if and only if this
* variable is \c gl_FragDepth and a layout qualifier is specified.
*/
nir_depth_layout depth_layout;
/**
* Storage location of the base of this variable
*
* The precise meaning of this field depends on the nature of the variable.
*
* - Vertex shader input: one of the values from \c gl_vert_attrib.
* - Vertex shader output: one of the values from \c gl_varying_slot.
* - Geometry shader input: one of the values from \c gl_varying_slot.
* - Geometry shader output: one of the values from \c gl_varying_slot.
* - Fragment shader input: one of the values from \c gl_varying_slot.
* - Fragment shader output: one of the values from \c gl_frag_result.
* - Uniforms: Per-stage uniform slot number for default uniform block.
* - Uniforms: Index within the uniform block definition for UBO members.
* - Non-UBO Uniforms: uniform slot number.
* - Other: This field is not currently used.
*
* If the variable is a uniform, shader input, or shader output, and the
* slot has not been assigned, the value will be -1.
*/
int location;
/**
* The actual location of the variable in the IR. Only valid for inputs
* and outputs.
*/
unsigned int driver_location;
/**
* Vertex stream output identifier.
*
* For packed outputs, bit 31 is set and bits [2*i+1,2*i] indicate the
* stream of the i-th component.
*/
unsigned stream;
/**
* output index for dual source blending.
*/
int index;
/**
* Descriptor set binding for sampler or UBO.
*/
int descriptor_set;
/**
* Initial binding point for a sampler or UBO.
*
* For array types, this represents the binding point for the first element.
*/
int binding;
/**
* Location an atomic counter or transform feedback is stored at.
*/
unsigned offset;
/**
* Transform feedback buffer.
*/
unsigned xfb_buffer;
/**
* Transform feedback stride.
*/
unsigned xfb_stride;
/**
* How the variable was declared. See nir_var_declaration_type.
*
* This is used to detect variables generated by the compiler, so should
* not be visible via the API.
*/
unsigned how_declared:2;
/**
* ARB_shader_image_load_store qualifiers.
*/
struct {
enum gl_access_qualifier access;
/** Image internal format if specified explicitly, otherwise GL_NONE. */
GLenum format;
} image;
} data;
/**
* Built-in state that backs this uniform
*
* Once set at variable creation, \c state_slots must remain invariant.
* This is because, ideally, this array would be shared by all clones of
* this variable in the IR tree. In other words, we'd really like for it
* to be a fly-weight.
*
* If the variable is not a uniform, \c num_state_slots will be zero and
* \c state_slots will be \c NULL.
*/
/*@{*/
unsigned num_state_slots; /**< Number of state slots used */
nir_state_slot *state_slots; /**< State descriptors. */
/*@}*/
/**
* Constant expression assigned in the initializer of the variable
*
* This field should only be used temporarily by creators of NIR shaders
* and then lower_constant_initializers can be used to get rid of them.
* Most of the rest of NIR ignores this field or asserts that it's NULL.
*/
nir_constant *constant_initializer;
/**
* For variables that are in an interface block or are an instance of an
* interface block, this is the \c GLSL_TYPE_INTERFACE type for that block.
*
* \sa ir_variable::location
*/
const struct glsl_type *interface_type;
/**
* Description of per-member data for per-member struct variables
*
* This is used for variables which are actually an amalgamation of
* multiple entities such as a struct of built-in values or a struct of
* inputs each with their own layout specifier. This is only allowed on
* variables with a struct or array of array of struct type.
*/
unsigned num_members;
struct nir_variable_data *members;
} nir_variable;
#define nir_foreach_variable(var, var_list) \
foreach_list_typed(nir_variable, var, node, var_list)
#define nir_foreach_variable_safe(var, var_list) \
foreach_list_typed_safe(nir_variable, var, node, var_list)
static inline bool
nir_variable_is_global(const nir_variable *var)
{
return var->data.mode != nir_var_local;
}
typedef struct nir_register {
struct exec_node node;
unsigned num_components; /** < number of vector components */
unsigned num_array_elems; /** < size of array (0 for no array) */
/* The bit-size of each channel; must be one of 8, 16, 32, or 64 */
uint8_t bit_size;
/** generic register index. */
unsigned index;
/** only for debug purposes, can be NULL */
const char *name;
/** whether this register is local (per-function) or global (per-shader) */
bool is_global;
/**
* If this flag is set to true, then accessing channels >= num_components
* is well-defined, and simply spills over to the next array element. This
* is useful for backends that can do per-component accessing, in
* particular scalar backends. By setting this flag and making
* num_components equal to 1, structures can be packed tightly into
* registers and then registers can be accessed per-component to get to
* each structure member, even if it crosses vec4 boundaries.
*/
bool is_packed;
/** set of nir_srcs where this register is used (read from) */
struct list_head uses;
/** set of nir_dests where this register is defined (written to) */
struct list_head defs;
/** set of nir_ifs where this register is used as a condition */
struct list_head if_uses;
} nir_register;
#define nir_foreach_register(reg, reg_list) \
foreach_list_typed(nir_register, reg, node, reg_list)
#define nir_foreach_register_safe(reg, reg_list) \
foreach_list_typed_safe(nir_register, reg, node, reg_list)
typedef enum {
nir_instr_type_alu,
nir_instr_type_deref,
nir_instr_type_call,
nir_instr_type_tex,
nir_instr_type_intrinsic,
nir_instr_type_load_const,
nir_instr_type_jump,
nir_instr_type_ssa_undef,
nir_instr_type_phi,
nir_instr_type_parallel_copy,
} nir_instr_type;
typedef struct nir_instr {
struct exec_node node;
nir_instr_type type;
struct nir_block *block;
/** generic instruction index. */
unsigned index;
/* A temporary for optimization and analysis passes to use for storing
* flags. For instance, DCE uses this to store the "dead/live" info.
*/
uint8_t pass_flags;
} nir_instr;
static inline nir_instr *
nir_instr_next(nir_instr *instr)
{
struct exec_node *next = exec_node_get_next(&instr->node);
if (exec_node_is_tail_sentinel(next))
return NULL;
else
return exec_node_data(nir_instr, next, node);
}
static inline nir_instr *
nir_instr_prev(nir_instr *instr)
{
struct exec_node *prev = exec_node_get_prev(&instr->node);
if (exec_node_is_head_sentinel(prev))
return NULL;
else
return exec_node_data(nir_instr, prev, node);
}
static inline bool
nir_instr_is_first(const nir_instr *instr)
{
return exec_node_is_head_sentinel(exec_node_get_prev_const(&instr->node));
}
static inline bool
nir_instr_is_last(const nir_instr *instr)
{
return exec_node_is_tail_sentinel(exec_node_get_next_const(&instr->node));
}
typedef struct nir_ssa_def {
/** for debugging only, can be NULL */
const char* name;
/** generic SSA definition index. */
unsigned index;
/** Index into the live_in and live_out bitfields */
unsigned live_index;
/** Instruction which produces this SSA value. */
nir_instr *parent_instr;
/** set of nir_instrs where this register is used (read from) */
struct list_head uses;
/** set of nir_ifs where this register is used as a condition */
struct list_head if_uses;
uint8_t num_components;
/* The bit-size of each channel; must be one of 8, 16, 32, or 64 */
uint8_t bit_size;
} nir_ssa_def;
struct nir_src;
typedef struct {
nir_register *reg;
struct nir_src *indirect; /** < NULL for no indirect offset */
unsigned base_offset;
/* TODO use-def chain goes here */
} nir_reg_src;
typedef struct {
nir_instr *parent_instr;
struct list_head def_link;
nir_register *reg;
struct nir_src *indirect; /** < NULL for no indirect offset */
unsigned base_offset;
/* TODO def-use chain goes here */
} nir_reg_dest;
struct nir_if;
typedef struct nir_src {
union {
/** Instruction that consumes this value as a source. */
nir_instr *parent_instr;
struct nir_if *parent_if;
};
struct list_head use_link;
union {
nir_reg_src reg;
nir_ssa_def *ssa;
};
bool is_ssa;
} nir_src;
static inline nir_src
nir_src_init(void)
{
nir_src src = { { NULL } };
return src;
}
#define NIR_SRC_INIT nir_src_init()
#define nir_foreach_use(src, reg_or_ssa_def) \
list_for_each_entry(nir_src, src, &(reg_or_ssa_def)->uses, use_link)
#define nir_foreach_use_safe(src, reg_or_ssa_def) \
list_for_each_entry_safe(nir_src, src, &(reg_or_ssa_def)->uses, use_link)
#define nir_foreach_if_use(src, reg_or_ssa_def) \
list_for_each_entry(nir_src, src, &(reg_or_ssa_def)->if_uses, use_link)
#define nir_foreach_if_use_safe(src, reg_or_ssa_def) \
list_for_each_entry_safe(nir_src, src, &(reg_or_ssa_def)->if_uses, use_link)
typedef struct {
union {
nir_reg_dest reg;
nir_ssa_def ssa;
};
bool is_ssa;
} nir_dest;
static inline nir_dest
nir_dest_init(void)
{
nir_dest dest = { { { NULL } } };
return dest;
}
#define NIR_DEST_INIT nir_dest_init()
#define nir_foreach_def(dest, reg) \
list_for_each_entry(nir_dest, dest, &(reg)->defs, reg.def_link)
#define nir_foreach_def_safe(dest, reg) \
list_for_each_entry_safe(nir_dest, dest, &(reg)->defs, reg.def_link)
static inline nir_src
nir_src_for_ssa(nir_ssa_def *def)
{
nir_src src = NIR_SRC_INIT;
src.is_ssa = true;
src.ssa = def;
return src;
}
static inline nir_src
nir_src_for_reg(nir_register *reg)
{
nir_src src = NIR_SRC_INIT;
src.is_ssa = false;
src.reg.reg = reg;
src.reg.indirect = NULL;
src.reg.base_offset = 0;
return src;
}
static inline nir_dest
nir_dest_for_reg(nir_register *reg)
{
nir_dest dest = NIR_DEST_INIT;
dest.reg.reg = reg;
return dest;
}
static inline unsigned
nir_src_bit_size(nir_src src)
{
return src.is_ssa ? src.ssa->bit_size : src.reg.reg->bit_size;
}
static inline unsigned
nir_src_num_components(nir_src src)
{
return src.is_ssa ? src.ssa->num_components : src.reg.reg->num_components;
}
static inline bool
nir_src_is_const(nir_src src)
{
return src.is_ssa &&
src.ssa->parent_instr->type == nir_instr_type_load_const;
}
int64_t nir_src_as_int(nir_src src);
uint64_t nir_src_as_uint(nir_src src);
bool nir_src_as_bool(nir_src src);
double nir_src_as_float(nir_src src);
int64_t nir_src_comp_as_int(nir_src src, unsigned component);
uint64_t nir_src_comp_as_uint(nir_src src, unsigned component);
bool nir_src_comp_as_bool(nir_src src, unsigned component);
double nir_src_comp_as_float(nir_src src, unsigned component);
static inline unsigned
nir_dest_bit_size(nir_dest dest)
{
return dest.is_ssa ? dest.ssa.bit_size : dest.reg.reg->bit_size;
}
static inline unsigned
nir_dest_num_components(nir_dest dest)
{
return dest.is_ssa ? dest.ssa.num_components : dest.reg.reg->num_components;
}
void nir_src_copy(nir_src *dest, const nir_src *src, void *instr_or_if);
void nir_dest_copy(nir_dest *dest, const nir_dest *src, nir_instr *instr);
typedef struct {
nir_src src;
/**
* \name input modifiers
*/
/*@{*/
/**
* For inputs interpreted as floating point, flips the sign bit. For
* inputs interpreted as integers, performs the two's complement negation.
*/
bool negate;
/**
* Clears the sign bit for floating point values, and computes the integer
* absolute value for integers. Note that the negate modifier acts after
* the absolute value modifier, therefore if both are set then all inputs
* will become negative.
*/
bool abs;
/*@}*/
/**
* For each input component, says which component of the register it is
* chosen from. Note that which elements of the swizzle are used and which
* are ignored are based on the write mask for most opcodes - for example,
* a statement like "foo.xzw = bar.zyx" would have a writemask of 1101b and
* a swizzle of {2, x, 1, 0} where x means "don't care."
*/
uint8_t swizzle[NIR_MAX_VEC_COMPONENTS];
} nir_alu_src;
typedef struct {
nir_dest dest;
/**
* \name saturate output modifier
*
* Only valid for opcodes that output floating-point numbers. Clamps the
* output to between 0.0 and 1.0 inclusive.
*/
bool saturate;
unsigned write_mask : NIR_MAX_VEC_COMPONENTS; /* ignored if dest.is_ssa is true */
} nir_alu_dest;
typedef enum {
nir_type_invalid = 0, /* Not a valid type */
nir_type_float,
nir_type_int,
nir_type_uint,
nir_type_bool,
nir_type_bool32 = 32 | nir_type_bool,
nir_type_int8 = 8 | nir_type_int,
nir_type_int16 = 16 | nir_type_int,
nir_type_int32 = 32 | nir_type_int,
nir_type_int64 = 64 | nir_type_int,
nir_type_uint8 = 8 | nir_type_uint,
nir_type_uint16 = 16 | nir_type_uint,
nir_type_uint32 = 32 | nir_type_uint,
nir_type_uint64 = 64 | nir_type_uint,
nir_type_float16 = 16 | nir_type_float,
nir_type_float32 = 32 | nir_type_float,
nir_type_float64 = 64 | nir_type_float,
} nir_alu_type;
#define NIR_ALU_TYPE_SIZE_MASK 0xfffffff8
#define NIR_ALU_TYPE_BASE_TYPE_MASK 0x00000007
static inline unsigned
nir_alu_type_get_type_size(nir_alu_type type)
{
return type & NIR_ALU_TYPE_SIZE_MASK;
}
static inline unsigned
nir_alu_type_get_base_type(nir_alu_type type)
{
return type & NIR_ALU_TYPE_BASE_TYPE_MASK;
}
static inline nir_alu_type
nir_get_nir_type_for_glsl_base_type(enum glsl_base_type base_type)
{
switch (base_type) {
case GLSL_TYPE_BOOL:
return nir_type_bool32;
break;
case GLSL_TYPE_UINT:
return nir_type_uint32;
break;
case GLSL_TYPE_INT:
return nir_type_int32;
break;
case GLSL_TYPE_UINT16:
return nir_type_uint16;
break;
case GLSL_TYPE_INT16:
return nir_type_int16;
break;
case GLSL_TYPE_UINT8:
return nir_type_uint8;
case GLSL_TYPE_INT8:
return nir_type_int8;
case GLSL_TYPE_UINT64:
return nir_type_uint64;
break;
case GLSL_TYPE_INT64:
return nir_type_int64;
break;
case GLSL_TYPE_FLOAT:
return nir_type_float32;
break;
case GLSL_TYPE_FLOAT16:
return nir_type_float16;
break;
case GLSL_TYPE_DOUBLE:
return nir_type_float64;
break;
default:
unreachable("unknown type");
}
}
static inline nir_alu_type
nir_get_nir_type_for_glsl_type(const struct glsl_type *type)
{
return nir_get_nir_type_for_glsl_base_type(glsl_get_base_type(type));
}
nir_op nir_type_conversion_op(nir_alu_type src, nir_alu_type dst,
nir_rounding_mode rnd);
typedef enum {
NIR_OP_IS_COMMUTATIVE = (1 << 0),
NIR_OP_IS_ASSOCIATIVE = (1 << 1),
} nir_op_algebraic_property;
typedef struct {
const char *name;
unsigned num_inputs;
/**
* The number of components in the output
*
* If non-zero, this is the size of the output and input sizes are
* explicitly given; swizzle and writemask are still in effect, but if
* the output component is masked out, then the input component may
* still be in use.
*
* If zero, the opcode acts in the standard, per-component manner; the
* operation is performed on each component (except the ones that are
* masked out) with the input being taken from the input swizzle for
* that component.
*
* The size of some of the inputs may be given (i.e. non-zero) even
* though output_size is zero; in that case, the inputs with a zero
* size act per-component, while the inputs with non-zero size don't.
*/
unsigned output_size;
/**
* The type of vector that the instruction outputs. Note that the
* staurate modifier is only allowed on outputs with the float type.
*/
nir_alu_type output_type;
/**
* The number of components in each input
*/
unsigned input_sizes[NIR_MAX_VEC_COMPONENTS];
/**
* The type of vector that each input takes. Note that negate and
* absolute value are only allowed on inputs with int or float type and
* behave differently on the two.
*/
nir_alu_type input_types[NIR_MAX_VEC_COMPONENTS];
nir_op_algebraic_property algebraic_properties;
} nir_op_info;
extern const nir_op_info nir_op_infos[nir_num_opcodes];
typedef struct nir_alu_instr {
nir_instr instr;
nir_op op;
/** Indicates that this ALU instruction generates an exact value
*
* This is kind of a mixture of GLSL "precise" and "invariant" and not
* really equivalent to either. This indicates that the value generated by
* this operation is high-precision and any code transformations that touch
* it must ensure that the resulting value is bit-for-bit identical to the
* original.
*/
bool exact;
nir_alu_dest dest;
nir_alu_src src[];
} nir_alu_instr;
void nir_alu_src_copy(nir_alu_src *dest, const nir_alu_src *src,
nir_alu_instr *instr);
void nir_alu_dest_copy(nir_alu_dest *dest, const nir_alu_dest *src,
nir_alu_instr *instr);
/* is this source channel used? */
static inline bool
nir_alu_instr_channel_used(const nir_alu_instr *instr, unsigned src,
unsigned channel)
{
if (nir_op_infos[instr->op].input_sizes[src] > 0)
return channel < nir_op_infos[instr->op].input_sizes[src];
return (instr->dest.write_mask >> channel) & 1;
}
static inline nir_component_mask_t
nir_alu_instr_src_read_mask(const nir_alu_instr *instr, unsigned src)
{
nir_component_mask_t read_mask = 0;
for (unsigned c = 0; c < NIR_MAX_VEC_COMPONENTS; c++) {
if (!nir_alu_instr_channel_used(instr, src, c))
continue;
read_mask |= (1 << instr->src[src].swizzle[c]);
}
return read_mask;
}
/*
* For instructions whose destinations are SSA, get the number of channels
* used for a source
*/
static inline unsigned
nir_ssa_alu_instr_src_components(const nir_alu_instr *instr, unsigned src)
{
assert(instr->dest.dest.is_ssa);
if (nir_op_infos[instr->op].input_sizes[src] > 0)
return nir_op_infos[instr->op].input_sizes[src];
return instr->dest.dest.ssa.num_components;
}
bool nir_alu_srcs_equal(const nir_alu_instr *alu1, const nir_alu_instr *alu2,
unsigned src1, unsigned src2);
typedef enum {
nir_deref_type_var,
nir_deref_type_array,
nir_deref_type_array_wildcard,
nir_deref_type_struct,
nir_deref_type_cast,
} nir_deref_type;
typedef struct {
nir_instr instr;
/** The type of this deref instruction */
nir_deref_type deref_type;
/** The mode of the underlying variable */
nir_variable_mode mode;
/** The dereferenced type of the resulting pointer value */
const struct glsl_type *type;
union {
/** Variable being dereferenced if deref_type is a deref_var */
nir_variable *var;
/** Parent deref if deref_type is not deref_var */
nir_src parent;
};
/** Additional deref parameters */
union {
struct {
nir_src index;
} arr;
struct {
unsigned index;
} strct;
};
/** Destination to store the resulting "pointer" */
nir_dest dest;
} nir_deref_instr;
NIR_DEFINE_CAST(nir_instr_as_deref, nir_instr, nir_deref_instr, instr,
type, nir_instr_type_deref)
static inline nir_deref_instr *
nir_src_as_deref(nir_src src)
{
if (!src.is_ssa)
return NULL;
if (src.ssa->parent_instr->type != nir_instr_type_deref)
return NULL;
return nir_instr_as_deref(src.ssa->parent_instr);
}
static inline nir_deref_instr *
nir_deref_instr_parent(const nir_deref_instr *instr)
{
if (instr->deref_type == nir_deref_type_var)
return NULL;
else
return nir_src_as_deref(instr->parent);
}
static inline nir_variable *
nir_deref_instr_get_variable(const nir_deref_instr *instr)
{
while (instr->deref_type != nir_deref_type_var) {
if (instr->deref_type == nir_deref_type_cast)
return NULL;
instr = nir_deref_instr_parent(instr);
}
return instr->var;
}
bool nir_deref_instr_has_indirect(nir_deref_instr *instr);
bool nir_deref_instr_remove_if_unused(nir_deref_instr *instr);
typedef struct {
nir_instr instr;
struct nir_function *callee;
unsigned num_params;
nir_src params[];
} nir_call_instr;
#include "nir_intrinsics.h"
#define NIR_INTRINSIC_MAX_CONST_INDEX 4
/** Represents an intrinsic
*
* An intrinsic is an instruction type for handling things that are
* more-or-less regular operations but don't just consume and produce SSA
* values like ALU operations do. Intrinsics are not for things that have
* special semantic meaning such as phi nodes and parallel copies.
* Examples of intrinsics include variable load/store operations, system
* value loads, and the like. Even though texturing more-or-less falls
* under this category, texturing is its own instruction type because
* trying to represent texturing with intrinsics would lead to a
* combinatorial explosion of intrinsic opcodes.
*
* By having a single instruction type for handling a lot of different
* cases, optimization passes can look for intrinsics and, for the most
* part, completely ignore them. Each intrinsic type also has a few
* possible flags that govern whether or not they can be reordered or
* eliminated. That way passes like dead code elimination can still work
* on intrisics without understanding the meaning of each.
*
* Each intrinsic has some number of constant indices, some number of
* variables, and some number of sources. What these sources, variables,
* and indices mean depends on the intrinsic and is documented with the
* intrinsic declaration in nir_intrinsics.h. Intrinsics and texture
* instructions are the only types of instruction that can operate on
* variables.
*/
typedef struct {
nir_instr instr;
nir_intrinsic_op intrinsic;
nir_dest dest;
/** number of components if this is a vectorized intrinsic
*
* Similarly to ALU operations, some intrinsics are vectorized.
* An intrinsic is vectorized if nir_intrinsic_infos.dest_components == 0.
* For vectorized intrinsics, the num_components field specifies the
* number of destination components and the number of source components
* for all sources with nir_intrinsic_infos.src_components[i] == 0.
*/
uint8_t num_components;
int const_index[NIR_INTRINSIC_MAX_CONST_INDEX];
nir_src src[];
} nir_intrinsic_instr;
static inline nir_variable *
nir_intrinsic_get_var(nir_intrinsic_instr *intrin, unsigned i)
{
return nir_deref_instr_get_variable(nir_src_as_deref(intrin->src[i]));
}
/**
* \name NIR intrinsics semantic flags
*
* information about what the compiler can do with the intrinsics.
*
* \sa nir_intrinsic_info::flags
*/
typedef enum {
/**
* whether the intrinsic can be safely eliminated if none of its output
* value is not being used.
*/
NIR_INTRINSIC_CAN_ELIMINATE = (1 << 0),
/**
* Whether the intrinsic can be reordered with respect to any other
* intrinsic, i.e. whether the only reordering dependencies of the
* intrinsic are due to the register reads/writes.
*/
NIR_INTRINSIC_CAN_REORDER = (1 << 1),
} nir_intrinsic_semantic_flag;
/**
* \name NIR intrinsics const-index flag
*
* Indicates the usage of a const_index slot.
*
* \sa nir_intrinsic_info::index_map
*/
typedef enum {
/**
* Generally instructions that take a offset src argument, can encode
* a constant 'base' value which is added to the offset.
*/
NIR_INTRINSIC_BASE = 1,
/**
* For store instructions, a writemask for the store.
*/
NIR_INTRINSIC_WRMASK = 2,
/**
* The stream-id for GS emit_vertex/end_primitive intrinsics.
*/
NIR_INTRINSIC_STREAM_ID = 3,
/**
* The clip-plane id for load_user_clip_plane intrinsic.
*/
NIR_INTRINSIC_UCP_ID = 4,
/**
* The amount of data, starting from BASE, that this instruction may
* access. This is used to provide bounds if the offset is not constant.
*/
NIR_INTRINSIC_RANGE = 5,
/**
* The Vulkan descriptor set for vulkan_resource_index intrinsic.
*/
NIR_INTRINSIC_DESC_SET = 6,
/**
* The Vulkan descriptor set binding for vulkan_resource_index intrinsic.
*/
NIR_INTRINSIC_BINDING = 7,
/**
* Component offset.
*/
NIR_INTRINSIC_COMPONENT = 8,
/**
* Interpolation mode (only meaningful for FS inputs).
*/
NIR_INTRINSIC_INTERP_MODE = 9,
/**
* A binary nir_op to use when performing a reduction or scan operation
*/
NIR_INTRINSIC_REDUCTION_OP = 10,
/**
* Cluster size for reduction operations
*/
NIR_INTRINSIC_CLUSTER_SIZE = 11,
/**
* Parameter index for a load_param intrinsic
*/
NIR_INTRINSIC_PARAM_IDX = 12,
/**
* Image dimensionality for image intrinsics
*
* One of GLSL_SAMPLER_DIM_*
*/
NIR_INTRINSIC_IMAGE_DIM = 13,
/**
* Non-zero if we are accessing an array image
*/
NIR_INTRINSIC_IMAGE_ARRAY = 14,
/**
* Image format for image intrinsics
*/
NIR_INTRINSIC_FORMAT = 15,
/**
* Access qualifiers for image intrinsics
*/
NIR_INTRINSIC_ACCESS = 16,
/**
* Alignment for offsets and addresses
*
* These two parameters, specify an alignment in terms of a multiplier and
* an offset. The offset or address parameter X of the intrinsic is
* guaranteed to satisfy the following:
*
* (X - align_offset) % align_mul == 0
*/
NIR_INTRINSIC_ALIGN_MUL = 17,
NIR_INTRINSIC_ALIGN_OFFSET = 18,
NIR_INTRINSIC_NUM_INDEX_FLAGS,
} nir_intrinsic_index_flag;
#define NIR_INTRINSIC_MAX_INPUTS 5
typedef struct {
const char *name;
unsigned num_srcs; /** < number of register/SSA inputs */
/** number of components of each input register
*
* If this value is 0, the number of components is given by the
* num_components field of nir_intrinsic_instr.
*/
unsigned src_components[NIR_INTRINSIC_MAX_INPUTS];
bool has_dest;
/** number of components of the output register
*
* If this value is 0, the number of components is given by the
* num_components field of nir_intrinsic_instr.
*/
unsigned dest_components;
/** the number of constant indices used by the intrinsic */
unsigned num_indices;
/** indicates the usage of intr->const_index[n] */
unsigned index_map[NIR_INTRINSIC_NUM_INDEX_FLAGS];
/** semantic flags for calls to this intrinsic */
nir_intrinsic_semantic_flag flags;
} nir_intrinsic_info;
extern const nir_intrinsic_info nir_intrinsic_infos[nir_num_intrinsics];
static inline unsigned
nir_intrinsic_src_components(nir_intrinsic_instr *intr, unsigned srcn)
{
const nir_intrinsic_info *info = &nir_intrinsic_infos[intr->intrinsic];
assert(srcn < info->num_srcs);
if (info->src_components[srcn])
return info->src_components[srcn];
else
return intr->num_components;
}
static inline unsigned
nir_intrinsic_dest_components(nir_intrinsic_instr *intr)
{
const nir_intrinsic_info *info = &nir_intrinsic_infos[intr->intrinsic];
if (!info->has_dest)
return 0;
else if (info->dest_components)
return info->dest_components;
else
return intr->num_components;
}
#define INTRINSIC_IDX_ACCESSORS(name, flag, type) \
static inline type \
nir_intrinsic_##name(const nir_intrinsic_instr *instr) \
{ \
const nir_intrinsic_info *info = &nir_intrinsic_infos[instr->intrinsic]; \
assert(info->index_map[NIR_INTRINSIC_##flag] > 0); \
return (type)instr->const_index[info->index_map[NIR_INTRINSIC_##flag] - 1]; \
} \
static inline void \
nir_intrinsic_set_##name(nir_intrinsic_instr *instr, type val) \
{ \
const nir_intrinsic_info *info = &nir_intrinsic_infos[instr->intrinsic]; \
assert(info->index_map[NIR_INTRINSIC_##flag] > 0); \
instr->const_index[info->index_map[NIR_INTRINSIC_##flag] - 1] = val; \
}
INTRINSIC_IDX_ACCESSORS(write_mask, WRMASK, unsigned)
INTRINSIC_IDX_ACCESSORS(base, BASE, int)
INTRINSIC_IDX_ACCESSORS(stream_id, STREAM_ID, unsigned)
INTRINSIC_IDX_ACCESSORS(ucp_id, UCP_ID, unsigned)
INTRINSIC_IDX_ACCESSORS(range, RANGE, unsigned)
INTRINSIC_IDX_ACCESSORS(desc_set, DESC_SET, unsigned)
INTRINSIC_IDX_ACCESSORS(binding, BINDING, unsigned)
INTRINSIC_IDX_ACCESSORS(component, COMPONENT, unsigned)
INTRINSIC_IDX_ACCESSORS(interp_mode, INTERP_MODE, unsigned)
INTRINSIC_IDX_ACCESSORS(reduction_op, REDUCTION_OP, unsigned)
INTRINSIC_IDX_ACCESSORS(cluster_size, CLUSTER_SIZE, unsigned)
INTRINSIC_IDX_ACCESSORS(param_idx, PARAM_IDX, unsigned)
INTRINSIC_IDX_ACCESSORS(image_dim, IMAGE_DIM, enum glsl_sampler_dim)
INTRINSIC_IDX_ACCESSORS(image_array, IMAGE_ARRAY, bool)
INTRINSIC_IDX_ACCESSORS(access, ACCESS, enum gl_access_qualifier)
INTRINSIC_IDX_ACCESSORS(format, FORMAT, unsigned)
INTRINSIC_IDX_ACCESSORS(align_mul, ALIGN_MUL, unsigned)
INTRINSIC_IDX_ACCESSORS(align_offset, ALIGN_OFFSET, unsigned)
static inline void
nir_intrinsic_set_align(nir_intrinsic_instr *intrin,
unsigned align_mul, unsigned align_offset)
{
assert(util_is_power_of_two_nonzero(align_mul));
assert(align_offset < align_mul);
nir_intrinsic_set_align_mul(intrin, align_mul);
nir_intrinsic_set_align_offset(intrin, align_offset);
}
/** Returns a simple alignment for a load/store intrinsic offset
*
* Instead of the full mul+offset alignment scheme provided by the ALIGN_MUL
* and ALIGN_OFFSET parameters, this helper takes both into account and
* provides a single simple alignment parameter. The offset X is guaranteed
* to satisfy X % align == 0.
*/
static inline unsigned
nir_intrinsic_align(nir_intrinsic_instr *intrin)
{
const unsigned align_mul = nir_intrinsic_align_mul(intrin);
const unsigned align_offset = nir_intrinsic_align_offset(intrin);
assert(align_offset < align_mul);
return align_offset ? 1 << (ffs(align_offset) - 1) : align_mul;
}
/**
* \group texture information
*
* This gives semantic information about textures which is useful to the
* frontend, the backend, and lowering passes, but not the optimizer.
*/
typedef enum {
nir_tex_src_coord,
nir_tex_src_projector,
nir_tex_src_comparator, /* shadow comparator */
nir_tex_src_offset,
nir_tex_src_bias,
nir_tex_src_lod,
nir_tex_src_min_lod,
nir_tex_src_ms_index, /* MSAA sample index */
nir_tex_src_ms_mcs, /* MSAA compression value */
nir_tex_src_ddx,
nir_tex_src_ddy,
nir_tex_src_texture_deref, /* < deref pointing to the texture */
nir_tex_src_sampler_deref, /* < deref pointing to the sampler */
nir_tex_src_texture_offset, /* < dynamically uniform indirect offset */
nir_tex_src_sampler_offset, /* < dynamically uniform indirect offset */
nir_tex_src_plane, /* < selects plane for planar textures */
nir_num_tex_src_types
} nir_tex_src_type;
typedef struct {
nir_src src;
nir_tex_src_type src_type;
} nir_tex_src;
typedef enum {
nir_texop_tex, /**< Regular texture look-up */
nir_texop_txb, /**< Texture look-up with LOD bias */
nir_texop_txl, /**< Texture look-up with explicit LOD */
nir_texop_txd, /**< Texture look-up with partial derivatives */
nir_texop_txf, /**< Texel fetch with explicit LOD */
nir_texop_txf_ms, /**< Multisample texture fetch */
nir_texop_txf_ms_mcs, /**< Multisample compression value fetch */
nir_texop_txs, /**< Texture size */
nir_texop_lod, /**< Texture lod query */
nir_texop_tg4, /**< Texture gather */
nir_texop_query_levels, /**< Texture levels query */
nir_texop_texture_samples, /**< Texture samples query */
nir_texop_samples_identical, /**< Query whether all samples are definitely
* identical.
*/
} nir_texop;
typedef struct {
nir_instr instr;
enum glsl_sampler_dim sampler_dim;
nir_alu_type dest_type;
nir_texop op;
nir_dest dest;
nir_tex_src *src;
unsigned num_srcs, coord_components;
bool is_array, is_shadow;
/**
* If is_shadow is true, whether this is the old-style shadow that outputs 4
* components or the new-style shadow that outputs 1 component.
*/
bool is_new_style_shadow;
/* gather component selector */
unsigned component : 2;
/** The texture index
*
* If this texture instruction has a nir_tex_src_texture_offset source,
* then the texture index is given by texture_index + texture_offset.
*/
unsigned texture_index;
/** The size of the texture array or 0 if it's not an array */
unsigned texture_array_size;
/** The sampler index
*
* The following operations do not require a sampler and, as such, this
* field should be ignored:
* - nir_texop_txf
* - nir_texop_txf_ms
* - nir_texop_txs
* - nir_texop_lod
* - nir_texop_query_levels
* - nir_texop_texture_samples
* - nir_texop_samples_identical
*
* If this texture instruction has a nir_tex_src_sampler_offset source,
* then the sampler index is given by sampler_index + sampler_offset.
*/
unsigned sampler_index;
} nir_tex_instr;
static inline unsigned
nir_tex_instr_dest_size(const nir_tex_instr *instr)
{
switch (instr->op) {
case nir_texop_txs: {
unsigned ret;
switch (instr->sampler_dim) {
case GLSL_SAMPLER_DIM_1D:
case GLSL_SAMPLER_DIM_BUF:
ret = 1;
break;
case GLSL_SAMPLER_DIM_2D:
case GLSL_SAMPLER_DIM_CUBE:
case GLSL_SAMPLER_DIM_MS:
case GLSL_SAMPLER_DIM_RECT:
case GLSL_SAMPLER_DIM_EXTERNAL:
case GLSL_SAMPLER_DIM_SUBPASS:
ret = 2;
break;
case GLSL_SAMPLER_DIM_3D:
ret = 3;
break;
default:
unreachable("not reached");
}
if (instr->is_array)
ret++;
return ret;
}
case nir_texop_lod:
return 2;
case nir_texop_texture_samples:
case nir_texop_query_levels:
case nir_texop_samples_identical:
return 1;
default:
if (instr->is_shadow && instr->is_new_style_shadow)
return 1;
return 4;
}
}
/* Returns true if this texture operation queries something about the texture
* rather than actually sampling it.
*/
static inline bool
nir_tex_instr_is_query(const nir_tex_instr *instr)
{
switch (instr->op) {
case nir_texop_txs:
case nir_texop_lod:
case nir_texop_texture_samples:
case nir_texop_query_levels:
case nir_texop_txf_ms_mcs:
return true;
case nir_texop_tex:
case nir_texop_txb:
case nir_texop_txl:
case nir_texop_txd:
case nir_texop_txf:
case nir_texop_txf_ms:
case nir_texop_tg4:
return false;
default:
unreachable("Invalid texture opcode");
}
}
static inline bool
nir_alu_instr_is_comparison(const nir_alu_instr *instr)
{
switch (instr->op) {
case nir_op_flt:
case nir_op_fge:
case nir_op_feq:
case nir_op_fne:
case nir_op_ilt:
case nir_op_ult:
case nir_op_ige:
case nir_op_uge:
case nir_op_ieq:
case nir_op_ine:
case nir_op_i2b32:
case nir_op_f2b32:
case nir_op_inot:
case nir_op_fnot:
return true;
default:
return false;
}
}
static inline nir_alu_type
nir_tex_instr_src_type(const nir_tex_instr *instr, unsigned src)
{
switch (instr->src[src].src_type) {
case nir_tex_src_coord:
switch (instr->op) {
case nir_texop_txf:
case nir_texop_txf_ms:
case nir_texop_txf_ms_mcs:
case nir_texop_samples_identical:
return nir_type_int;
default:
return nir_type_float;
}
case nir_tex_src_lod:
switch (instr->op) {
case nir_texop_txs:
case nir_texop_txf:
return nir_type_int;
default:
return nir_type_float;
}
case nir_tex_src_projector:
case nir_tex_src_comparator:
case nir_tex_src_bias:
case nir_tex_src_ddx:
case nir_tex_src_ddy:
return nir_type_float;
case nir_tex_src_offset:
case nir_tex_src_ms_index:
case nir_tex_src_texture_offset:
case nir_tex_src_sampler_offset:
return nir_type_int;
default:
unreachable("Invalid texture source type");
}
}
static inline unsigned
nir_tex_instr_src_size(const nir_tex_instr *instr, unsigned src)
{
if (instr->src[src].src_type == nir_tex_src_coord)
return instr->coord_components;
/* The MCS value is expected to be a vec4 returned by a txf_ms_mcs */
if (instr->src[src].src_type == nir_tex_src_ms_mcs)
return 4;
if (instr->src[src].src_type == nir_tex_src_ddx ||
instr->src[src].src_type == nir_tex_src_ddy) {
if (instr->is_array)
return instr->coord_components - 1;
else
return instr->coord_components;
}
/* Usual APIs don't allow cube + offset, but we allow it, with 2 coords for
* the offset, since a cube maps to a single face.
*/
if (instr->src[src].src_type == nir_tex_src_offset) {
if (instr->sampler_dim == GLSL_SAMPLER_DIM_CUBE)
return 2;
else if (instr->is_array)
return instr->coord_components - 1;
else
return instr->coord_components;
}
return 1;
}
static inline int
nir_tex_instr_src_index(const nir_tex_instr *instr, nir_tex_src_type type)
{
for (unsigned i = 0; i < instr->num_srcs; i++)
if (instr->src[i].src_type == type)
return (int) i;
return -1;
}
void nir_tex_instr_add_src(nir_tex_instr *tex,
nir_tex_src_type src_type,
nir_src src);
void nir_tex_instr_remove_src(nir_tex_instr *tex, unsigned src_idx);
typedef struct {
nir_instr instr;
nir_const_value value;
nir_ssa_def def;
} nir_load_const_instr;
typedef enum {
nir_jump_return,
nir_jump_break,
nir_jump_continue,
} nir_jump_type;
typedef struct {
nir_instr instr;
nir_jump_type type;
} nir_jump_instr;
/* creates a new SSA variable in an undefined state */
typedef struct {
nir_instr instr;
nir_ssa_def def;
} nir_ssa_undef_instr;
typedef struct {
struct exec_node node;
/* The predecessor block corresponding to this source */
struct nir_block *pred;
nir_src src;
} nir_phi_src;
#define nir_foreach_phi_src(phi_src, phi) \
foreach_list_typed(nir_phi_src, phi_src, node, &(phi)->srcs)
#define nir_foreach_phi_src_safe(phi_src, phi) \
foreach_list_typed_safe(nir_phi_src, phi_src, node, &(phi)->srcs)
typedef struct {
nir_instr instr;
struct exec_list srcs; /** < list of nir_phi_src */
nir_dest dest;
} nir_phi_instr;
typedef struct {
struct exec_node node;
nir_src src;
nir_dest dest;
} nir_parallel_copy_entry;
#define nir_foreach_parallel_copy_entry(entry, pcopy) \
foreach_list_typed(nir_parallel_copy_entry, entry, node, &(pcopy)->entries)
typedef struct {
nir_instr instr;
/* A list of nir_parallel_copy_entrys. The sources of all of the
* entries are copied to the corresponding destinations "in parallel".
* In other words, if we have two entries: a -> b and b -> a, the values
* get swapped.
*/
struct exec_list entries;
} nir_parallel_copy_instr;
NIR_DEFINE_CAST(nir_instr_as_alu, nir_instr, nir_alu_instr, instr,
type, nir_instr_type_alu)
NIR_DEFINE_CAST(nir_instr_as_call, nir_instr, nir_call_instr, instr,
type, nir_instr_type_call)
NIR_DEFINE_CAST(nir_instr_as_jump, nir_instr, nir_jump_instr, instr,
type, nir_instr_type_jump)
NIR_DEFINE_CAST(nir_instr_as_tex, nir_instr, nir_tex_instr, instr,
type, nir_instr_type_tex)
NIR_DEFINE_CAST(nir_instr_as_intrinsic, nir_instr, nir_intrinsic_instr, instr,
type, nir_instr_type_intrinsic)
NIR_DEFINE_CAST(nir_instr_as_load_const, nir_instr, nir_load_const_instr, instr,
type, nir_instr_type_load_const)
NIR_DEFINE_CAST(nir_instr_as_ssa_undef, nir_instr, nir_ssa_undef_instr, instr,
type, nir_instr_type_ssa_undef)
NIR_DEFINE_CAST(nir_instr_as_phi, nir_instr, nir_phi_instr, instr,
type, nir_instr_type_phi)
NIR_DEFINE_CAST(nir_instr_as_parallel_copy, nir_instr,
nir_parallel_copy_instr, instr,
type, nir_instr_type_parallel_copy)
/*
* Control flow
*
* Control flow consists of a tree of control flow nodes, which include
* if-statements and loops. The leaves of the tree are basic blocks, lists of
* instructions that always run start-to-finish. Each basic block also keeps
* track of its successors (blocks which may run immediately after the current
* block) and predecessors (blocks which could have run immediately before the
* current block). Each function also has a start block and an end block which
* all return statements point to (which is always empty). Together, all the
* blocks with their predecessors and successors make up the control flow
* graph (CFG) of the function. There are helpers that modify the tree of
* control flow nodes while modifying the CFG appropriately; these should be
* used instead of modifying the tree directly.
*/
typedef enum {
nir_cf_node_block,
nir_cf_node_if,
nir_cf_node_loop,
nir_cf_node_function
} nir_cf_node_type;
typedef struct nir_cf_node {
struct exec_node node;
nir_cf_node_type type;
struct nir_cf_node *parent;
} nir_cf_node;
typedef struct nir_block {
nir_cf_node cf_node;
struct exec_list instr_list; /** < list of nir_instr */
/** generic block index; generated by nir_index_blocks */
unsigned index;
/*
* Each block can only have up to 2 successors, so we put them in a simple
* array - no need for anything more complicated.
*/
struct nir_block *successors[2];
/* Set of nir_block predecessors in the CFG */
struct set *predecessors;
/*
* this node's immediate dominator in the dominance tree - set to NULL for
* the start block.
*/
struct nir_block *imm_dom;
/* This node's children in the dominance tree */
unsigned num_dom_children;
struct nir_block **dom_children;
/* Set of nir_blocks on the dominance frontier of this block */
struct set *dom_frontier;
/*
* These two indices have the property that dom_{pre,post}_index for each
* child of this block in the dominance tree will always be between
* dom_pre_index and dom_post_index for this block, which makes testing if
* a given block is dominated by another block an O(1) operation.
*/
unsigned dom_pre_index, dom_post_index;
/* live in and out for this block; used for liveness analysis */
BITSET_WORD *live_in;
BITSET_WORD *live_out;
} nir_block;
static inline nir_instr *
nir_block_first_instr(nir_block *block)
{
struct exec_node *head = exec_list_get_head(&block->instr_list);
return exec_node_data(nir_instr, head, node);
}
static inline nir_instr *
nir_block_last_instr(nir_block *block)
{
struct exec_node *tail = exec_list_get_tail(&block->instr_list);
return exec_node_data(nir_instr, tail, node);
}
static inline bool
nir_block_ends_in_jump(nir_block *block)
{
return !exec_list_is_empty(&block->instr_list) &&
nir_block_last_instr(block)->type == nir_instr_type_jump;
}
#define nir_foreach_instr(instr, block) \
foreach_list_typed(nir_instr, instr, node, &(block)->instr_list)
#define nir_foreach_instr_reverse(instr, block) \
foreach_list_typed_reverse(nir_instr, instr, node, &(block)->instr_list)
#define nir_foreach_instr_safe(instr, block) \
foreach_list_typed_safe(nir_instr, instr, node, &(block)->instr_list)
#define nir_foreach_instr_reverse_safe(instr, block) \
foreach_list_typed_reverse_safe(nir_instr, instr, node, &(block)->instr_list)
typedef struct nir_if {
nir_cf_node cf_node;
nir_src condition;
struct exec_list then_list; /** < list of nir_cf_node */
struct exec_list else_list; /** < list of nir_cf_node */
} nir_if;
typedef struct {
nir_if *nif;
nir_instr *conditional_instr;
nir_block *break_block;
nir_block *continue_from_block;
bool continue_from_then;
struct list_head loop_terminator_link;
} nir_loop_terminator;
typedef struct {
/* Number of instructions in the loop */
unsigned num_instructions;
/* Maximum number of times the loop is run (if known) */
unsigned max_trip_count;
/* Do we know the exact number of times the loop will be run */
bool exact_trip_count_known;
/* Unroll the loop regardless of its size */
bool force_unroll;
/* Does the loop contain complex loop terminators, continues or other
* complex behaviours? If this is true we can't rely on
* loop_terminator_list to be complete or accurate.
*/
bool complex_loop;
nir_loop_terminator *limiting_terminator;
/* A list of loop_terminators terminating this loop. */
struct list_head loop_terminator_list;
} nir_loop_info;
typedef struct {
nir_cf_node cf_node;
struct exec_list body; /** < list of nir_cf_node */
nir_loop_info *info;
} nir_loop;
/**
* Various bits of metadata that can may be created or required by
* optimization and analysis passes
*/
typedef enum {
nir_metadata_none = 0x0,
nir_metadata_block_index = 0x1,
nir_metadata_dominance = 0x2,
nir_metadata_live_ssa_defs = 0x4,
nir_metadata_not_properly_reset = 0x8,
nir_metadata_loop_analysis = 0x10,
} nir_metadata;
typedef struct {
nir_cf_node cf_node;
/** pointer to the function of which this is an implementation */
struct nir_function *function;
struct exec_list body; /** < list of nir_cf_node */
nir_block *end_block;
/** list for all local variables in the function */
struct exec_list locals;
/** list of local registers in the function */
struct exec_list registers;
/** next available local register index */
unsigned reg_alloc;
/** next available SSA value index */
unsigned ssa_alloc;
/* total number of basic blocks, only valid when block_index_dirty = false */
unsigned num_blocks;
nir_metadata valid_metadata;
} nir_function_impl;
ATTRIBUTE_RETURNS_NONNULL static inline nir_block *
nir_start_block(nir_function_impl *impl)
{
return (nir_block *) impl->body.head_sentinel.next;
}
ATTRIBUTE_RETURNS_NONNULL static inline nir_block *
nir_impl_last_block(nir_function_impl *impl)
{
return (nir_block *) impl->body.tail_sentinel.prev;
}
static inline nir_cf_node *
nir_cf_node_next(nir_cf_node *node)
{
struct exec_node *next = exec_node_get_next(&node->node);
if (exec_node_is_tail_sentinel(next))
return NULL;
else
return exec_node_data(nir_cf_node, next, node);
}
static inline nir_cf_node *
nir_cf_node_prev(nir_cf_node *node)
{
struct exec_node *prev = exec_node_get_prev(&node->node);
if (exec_node_is_head_sentinel(prev))
return NULL;
else
return exec_node_data(nir_cf_node, prev, node);
}
static inline bool
nir_cf_node_is_first(const nir_cf_node *node)
{
return exec_node_is_head_sentinel(node->node.prev);
}
static inline bool
nir_cf_node_is_last(const nir_cf_node *node)
{
return exec_node_is_tail_sentinel(node->node.next);
}
NIR_DEFINE_CAST(nir_cf_node_as_block, nir_cf_node, nir_block, cf_node,
type, nir_cf_node_block)
NIR_DEFINE_CAST(nir_cf_node_as_if, nir_cf_node, nir_if, cf_node,
type, nir_cf_node_if)
NIR_DEFINE_CAST(nir_cf_node_as_loop, nir_cf_node, nir_loop, cf_node,
type, nir_cf_node_loop)
NIR_DEFINE_CAST(nir_cf_node_as_function, nir_cf_node,
nir_function_impl, cf_node, type, nir_cf_node_function)
static inline nir_block *
nir_if_first_then_block(nir_if *if_stmt)
{
struct exec_node *head = exec_list_get_head(&if_stmt->then_list);
return nir_cf_node_as_block(exec_node_data(nir_cf_node, head, node));
}
static inline nir_block *
nir_if_last_then_block(nir_if *if_stmt)
{
struct exec_node *tail = exec_list_get_tail(&if_stmt->then_list);
return nir_cf_node_as_block(exec_node_data(nir_cf_node, tail, node));
}
static inline nir_block *
nir_if_first_else_block(nir_if *if_stmt)
{
struct exec_node *head = exec_list_get_head(&if_stmt->else_list);
return nir_cf_node_as_block(exec_node_data(nir_cf_node, head, node));
}
static inline nir_block *
nir_if_last_else_block(nir_if *if_stmt)
{
struct exec_node *tail = exec_list_get_tail(&if_stmt->else_list);
return nir_cf_node_as_block(exec_node_data(nir_cf_node, tail, node));
}
static inline nir_block *
nir_loop_first_block(nir_loop *loop)
{
struct exec_node *head = exec_list_get_head(&loop->body);
return nir_cf_node_as_block(exec_node_data(nir_cf_node, head, node));
}
static inline nir_block *
nir_loop_last_block(nir_loop *loop)
{
struct exec_node *tail = exec_list_get_tail(&loop->body);
return nir_cf_node_as_block(exec_node_data(nir_cf_node, tail, node));
}
typedef struct {
uint8_t num_components;
uint8_t bit_size;
} nir_parameter;
typedef struct nir_function {
struct exec_node node;
const char *name;
struct nir_shader *shader;
unsigned num_params;
nir_parameter *params;
/** The implementation of this function.
*
* If the function is only declared and not implemented, this is NULL.
*/
nir_function_impl *impl;
} nir_function;
typedef struct nir_shader_compiler_options {
bool lower_fdiv;
bool lower_ffma;
bool fuse_ffma;
bool lower_flrp32;
/** Lowers flrp when it does not support doubles */
bool lower_flrp64;
bool lower_fpow;
bool lower_fsat;
bool lower_fsqrt;
bool lower_fmod32;
bool lower_fmod64;
/** Lowers ibitfield_extract/ubitfield_extract to ibfe/ubfe. */
bool lower_bitfield_extract;
/** Lowers ibitfield_extract/ubitfield_extract to bfm, compares, shifts. */
bool lower_bitfield_extract_to_shifts;
/** Lowers bitfield_insert to bfi/bfm */
bool lower_bitfield_insert;
/** Lowers bitfield_insert to bfm, compares, and shifts. */
bool lower_bitfield_insert_to_shifts;
/** Lowers bitfield_reverse to shifts. */
bool lower_bitfield_reverse;
/** Lowers bit_count to shifts. */
bool lower_bit_count;
/** Lowers bfm to shifts and subtracts. */
bool lower_bfm;
/** Lowers ifind_msb to compare and ufind_msb */
bool lower_ifind_msb;
/** Lowers find_lsb to ufind_msb and logic ops */
bool lower_find_lsb;
bool lower_uadd_carry;
bool lower_usub_borrow;
/** Lowers imul_high/umul_high to 16-bit multiplies and carry operations. */
bool lower_mul_high;
/** lowers fneg and ineg to fsub and isub. */
bool lower_negate;
/** lowers fsub and isub to fadd+fneg and iadd+ineg. */
bool lower_sub;
/* lower {slt,sge,seq,sne} to {flt,fge,feq,fne} + b2f: */
bool lower_scmp;
/** enables rules to lower idiv by power-of-two: */
bool lower_idiv;
/* lower b2f to iand */
bool lower_b2f;
/* Does the native fdot instruction replicate its result for four
* components? If so, then opt_algebraic_late will turn all fdotN
* instructions into fdot_replicatedN instructions.
*/
bool fdot_replicates;
/** lowers ffloor to fsub+ffract: */
bool lower_ffloor;
/** lowers ffract to fsub+ffloor: */
bool lower_ffract;
/** lowers fceil to fneg+ffloor+fneg: */
bool lower_fceil;
bool lower_ldexp;
bool lower_pack_half_2x16;
bool lower_pack_unorm_2x16;
bool lower_pack_snorm_2x16;
bool lower_pack_unorm_4x8;
bool lower_pack_snorm_4x8;
bool lower_unpack_half_2x16;
bool lower_unpack_unorm_2x16;
bool lower_unpack_snorm_2x16;
bool lower_unpack_unorm_4x8;
bool lower_unpack_snorm_4x8;
bool lower_extract_byte;
bool lower_extract_word;
bool lower_all_io_to_temps;
/**
* Does the driver support real 32-bit integers? (Otherwise, integers
* are simulated by floats.)
*/
bool native_integers;
/* Indicates that the driver only has zero-based vertex id */
bool vertex_id_zero_based;
/**
* If enabled, gl_BaseVertex will be lowered as:
* is_indexed_draw (~0/0) & firstvertex
*/
bool lower_base_vertex;
/**
* If enabled, gl_HelperInvocation will be lowered as:
*
* !((1 << sample_id) & sample_mask_in))
*
* This depends on some possibly hw implementation details, which may
* not be true for all hw. In particular that the FS is only executed
* for covered samples or for helper invocations. So, do not blindly
* enable this option.
*
* Note: See also issue #22 in ARB_shader_image_load_store
*/
bool lower_helper_invocation;
bool lower_cs_local_index_from_id;
bool lower_cs_local_id_from_index;
bool lower_device_index_to_zero;
/* Set if nir_lower_wpos_ytransform() should also invert gl_PointCoord. */
bool lower_wpos_pntc;
/**
* Should nir_lower_io() create load_interpolated_input intrinsics?
*
* If not, it generates regular load_input intrinsics and interpolation
* information must be inferred from the list of input nir_variables.
*/
bool use_interpolated_input_intrinsics;
unsigned max_unroll_iterations;
} nir_shader_compiler_options;
typedef struct nir_shader {
/** list of uniforms (nir_variable) */
struct exec_list uniforms;
/** list of inputs (nir_variable) */
struct exec_list inputs;
/** list of outputs (nir_variable) */
struct exec_list outputs;
/** list of shared compute variables (nir_variable) */
struct exec_list shared;
/** Set of driver-specific options for the shader.
*
* The memory for the options is expected to be kept in a single static
* copy by the driver.
*/
const struct nir_shader_compiler_options *options;
/** Various bits of compile-time information about a given shader */
struct shader_info info;
/** list of global variables in the shader (nir_variable) */
struct exec_list globals;
/** list of system value variables in the shader (nir_variable) */
struct exec_list system_values;
struct exec_list functions; /** < list of nir_function */
/** list of global register in the shader */
struct exec_list registers;
/** next available global register index */
unsigned reg_alloc;
/**
* the highest index a load_input_*, load_uniform_*, etc. intrinsic can
* access plus one
*/
unsigned num_inputs, num_uniforms, num_outputs, num_shared;
/** Constant data associated with this shader.
*
* Constant data is loaded through load_constant intrinsics. See also
* nir_opt_large_constants.
*/
void *constant_data;
unsigned constant_data_size;
} nir_shader;
static inline nir_function_impl *
nir_shader_get_entrypoint(nir_shader *shader)
{
assert(exec_list_length(&shader->functions) == 1);
struct exec_node *func_node = exec_list_get_head(&shader->functions);
nir_function *func = exec_node_data(nir_function, func_node, node);
assert(func->num_params == 0);
assert(func->impl);
return func->impl;
}
#define nir_foreach_function(func, shader) \
foreach_list_typed(nir_function, func, node, &(shader)->functions)
nir_shader *nir_shader_create(void *mem_ctx,
gl_shader_stage stage,
const nir_shader_compiler_options *options,
shader_info *si);
/** creates a register, including assigning it an index and adding it to the list */
nir_register *nir_global_reg_create(nir_shader *shader);
nir_register *nir_local_reg_create(nir_function_impl *impl);
void nir_reg_remove(nir_register *reg);
/** Adds a variable to the appropriate list in nir_shader */
void nir_shader_add_variable(nir_shader *shader, nir_variable *var);
static inline void
nir_function_impl_add_variable(nir_function_impl *impl, nir_variable *var)
{
assert(var->data.mode == nir_var_local);
exec_list_push_tail(&impl->locals, &var->node);
}
/** creates a variable, sets a few defaults, and adds it to the list */
nir_variable *nir_variable_create(nir_shader *shader,
nir_variable_mode mode,
const struct glsl_type *type,
const char *name);
/** creates a local variable and adds it to the list */
nir_variable *nir_local_variable_create(nir_function_impl *impl,
const struct glsl_type *type,
const char *name);
/** creates a function and adds it to the shader's list of functions */
nir_function *nir_function_create(nir_shader *shader, const char *name);
nir_function_impl *nir_function_impl_create(nir_function *func);
/** creates a function_impl that isn't tied to any particular function */
nir_function_impl *nir_function_impl_create_bare(nir_shader *shader);
nir_block *nir_block_create(nir_shader *shader);
nir_if *nir_if_create(nir_shader *shader);
nir_loop *nir_loop_create(nir_shader *shader);
nir_function_impl *nir_cf_node_get_function(nir_cf_node *node);
/** requests that the given pieces of metadata be generated */
void nir_metadata_require(nir_function_impl *impl, nir_metadata required, ...);
/** dirties all but the preserved metadata */
void nir_metadata_preserve(nir_function_impl *impl, nir_metadata preserved);
/** creates an instruction with default swizzle/writemask/etc. with NULL registers */
nir_alu_instr *nir_alu_instr_create(nir_shader *shader, nir_op op);
nir_deref_instr *nir_deref_instr_create(nir_shader *shader,
nir_deref_type deref_type);
nir_jump_instr *nir_jump_instr_create(nir_shader *shader, nir_jump_type type);
nir_load_const_instr *nir_load_const_instr_create(nir_shader *shader,
unsigned num_components,
unsigned bit_size);
nir_intrinsic_instr *nir_intrinsic_instr_create(nir_shader *shader,
nir_intrinsic_op op);
nir_call_instr *nir_call_instr_create(nir_shader *shader,
nir_function *callee);
nir_tex_instr *nir_tex_instr_create(nir_shader *shader, unsigned num_srcs);
nir_phi_instr *nir_phi_instr_create(nir_shader *shader);
nir_parallel_copy_instr *nir_parallel_copy_instr_create(nir_shader *shader);
nir_ssa_undef_instr *nir_ssa_undef_instr_create(nir_shader *shader,
unsigned num_components,
unsigned bit_size);
nir_const_value nir_alu_binop_identity(nir_op binop, unsigned bit_size);
/**
* NIR Cursors and Instruction Insertion API
* @{
*
* A tiny struct representing a point to insert/extract instructions or
* control flow nodes. Helps reduce the combinatorial explosion of possible
* points to insert/extract.
*
* \sa nir_control_flow.h
*/
typedef enum {
nir_cursor_before_block,
nir_cursor_after_block,
nir_cursor_before_instr,
nir_cursor_after_instr,
} nir_cursor_option;
typedef struct {
nir_cursor_option option;
union {
nir_block *block;
nir_instr *instr;
};
} nir_cursor;
static inline nir_block *
nir_cursor_current_block(nir_cursor cursor)
{
if (cursor.option == nir_cursor_before_instr ||
cursor.option == nir_cursor_after_instr) {
return cursor.instr->block;
} else {
return cursor.block;
}
}
bool nir_cursors_equal(nir_cursor a, nir_cursor b);
static inline nir_cursor
nir_before_block(nir_block *block)
{
nir_cursor cursor;
cursor.option = nir_cursor_before_block;
cursor.block = block;
return cursor;
}
static inline nir_cursor
nir_after_block(nir_block *block)
{
nir_cursor cursor;
cursor.option = nir_cursor_after_block;
cursor.block = block;
return cursor;
}
static inline nir_cursor
nir_before_instr(nir_instr *instr)
{
nir_cursor cursor;
cursor.option = nir_cursor_before_instr;
cursor.instr = instr;
return cursor;
}
static inline nir_cursor
nir_after_instr(nir_instr *instr)
{
nir_cursor cursor;
cursor.option = nir_cursor_after_instr;
cursor.instr = instr;
return cursor;
}
static inline nir_cursor
nir_after_block_before_jump(nir_block *block)
{
nir_instr *last_instr = nir_block_last_instr(block);
if (last_instr && last_instr->type == nir_instr_type_jump) {
return nir_before_instr(last_instr);
} else {
return nir_after_block(block);
}
}
static inline nir_cursor
nir_before_src(nir_src *src, bool is_if_condition)
{
if (is_if_condition) {
nir_block *prev_block =
nir_cf_node_as_block(nir_cf_node_prev(&src->parent_if->cf_node));
assert(!nir_block_ends_in_jump(prev_block));
return nir_after_block(prev_block);
} else if (src->parent_instr->type == nir_instr_type_phi) {
#ifndef NDEBUG
nir_phi_instr *cond_phi = nir_instr_as_phi(src->parent_instr);
bool found = false;
nir_foreach_phi_src(phi_src, cond_phi) {
if (phi_src->src.ssa == src->ssa) {
found = true;
break;
}
}
assert(found);
#endif
/* The LIST_ENTRY macro is a generic container-of macro, it just happens
* to have a more specific name.
*/
nir_phi_src *phi_src = LIST_ENTRY(nir_phi_src, src, src);
return nir_after_block_before_jump(phi_src->pred);
} else {
return nir_before_instr(src->parent_instr);
}
}
static inline nir_cursor
nir_before_cf_node(nir_cf_node *node)
{
if (node->type == nir_cf_node_block)
return nir_before_block(nir_cf_node_as_block(node));
return nir_after_block(nir_cf_node_as_block(nir_cf_node_prev(node)));
}
static inline nir_cursor
nir_after_cf_node(nir_cf_node *node)
{
if (node->type == nir_cf_node_block)
return nir_after_block(nir_cf_node_as_block(node));
return nir_before_block(nir_cf_node_as_block(nir_cf_node_next(node)));
}
static inline nir_cursor
nir_after_phis(nir_block *block)
{
nir_foreach_instr(instr, block) {
if (instr->type != nir_instr_type_phi)
return nir_before_instr(instr);
}
return nir_after_block(block);
}
static inline nir_cursor
nir_after_cf_node_and_phis(nir_cf_node *node)
{
if (node->type == nir_cf_node_block)
return nir_after_block(nir_cf_node_as_block(node));
nir_block *block = nir_cf_node_as_block(nir_cf_node_next(node));
return nir_after_phis(block);
}
static inline nir_cursor
nir_before_cf_list(struct exec_list *cf_list)
{
nir_cf_node *first_node = exec_node_data(nir_cf_node,
exec_list_get_head(cf_list), node);
return nir_before_cf_node(first_node);
}
static inline nir_cursor
nir_after_cf_list(struct exec_list *cf_list)
{
nir_cf_node *last_node = exec_node_data(nir_cf_node,
exec_list_get_tail(cf_list), node);
return nir_after_cf_node(last_node);
}
/**
* Insert a NIR instruction at the given cursor.
*
* Note: This does not update the cursor.
*/
void nir_instr_insert(nir_cursor cursor, nir_instr *instr);
static inline void
nir_instr_insert_before(nir_instr *instr, nir_instr *before)
{
nir_instr_insert(nir_before_instr(instr), before);
}
static inline void
nir_instr_insert_after(nir_instr *instr, nir_instr *after)
{
nir_instr_insert(nir_after_instr(instr), after);
}
static inline void
nir_instr_insert_before_block(nir_block *block, nir_instr *before)
{
nir_instr_insert(nir_before_block(block), before);
}
static inline void
nir_instr_insert_after_block(nir_block *block, nir_instr *after)
{
nir_instr_insert(nir_after_block(block), after);
}
static inline void
nir_instr_insert_before_cf(nir_cf_node *node, nir_instr *before)
{
nir_instr_insert(nir_before_cf_node(node), before);
}
static inline void
nir_instr_insert_after_cf(nir_cf_node *node, nir_instr *after)
{
nir_instr_insert(nir_after_cf_node(node), after);
}
static inline void
nir_instr_insert_before_cf_list(struct exec_list *list, nir_instr *before)
{
nir_instr_insert(nir_before_cf_list(list), before);
}
static inline void
nir_instr_insert_after_cf_list(struct exec_list *list, nir_instr *after)
{
nir_instr_insert(nir_after_cf_list(list), after);
}
void nir_instr_remove_v(nir_instr *instr);
static inline nir_cursor
nir_instr_remove(nir_instr *instr)
{
nir_cursor cursor;
nir_instr *prev = nir_instr_prev(instr);
if (prev) {
cursor = nir_after_instr(prev);
} else {
cursor = nir_before_block(instr->block);
}
nir_instr_remove_v(instr);
return cursor;
}
/** @} */
typedef bool (*nir_foreach_ssa_def_cb)(nir_ssa_def *def, void *state);
typedef bool (*nir_foreach_dest_cb)(nir_dest *dest, void *state);
typedef bool (*nir_foreach_src_cb)(nir_src *src, void *state);
bool nir_foreach_ssa_def(nir_instr *instr, nir_foreach_ssa_def_cb cb,
void *state);
bool nir_foreach_dest(nir_instr *instr, nir_foreach_dest_cb cb, void *state);
bool nir_foreach_src(nir_instr *instr, nir_foreach_src_cb cb, void *state);
nir_const_value *nir_src_as_const_value(nir_src src);
static inline struct nir_instr *
nir_src_instr(const struct nir_src *src)
{
return src->is_ssa ? src->ssa->parent_instr : NULL;
}
#define NIR_SRC_AS_(name, c_type, type_enum, cast_macro) \
static inline c_type * \
nir_src_as_ ## name (struct nir_src *src) \
{ \
return src->is_ssa && src->ssa->parent_instr->type == type_enum \
? cast_macro(src->ssa->parent_instr) : NULL; \
} \
static inline const c_type * \
nir_src_as_ ## name ## _const(const struct nir_src *src) \
{ \
return src->is_ssa && src->ssa->parent_instr->type == type_enum \
? cast_macro(src->ssa->parent_instr) : NULL; \
}
NIR_SRC_AS_(alu_instr, nir_alu_instr, nir_instr_type_alu, nir_instr_as_alu)
bool nir_src_is_dynamically_uniform(nir_src src);
bool nir_srcs_equal(nir_src src1, nir_src src2);
void nir_instr_rewrite_src(nir_instr *instr, nir_src *src, nir_src new_src);
void nir_instr_move_src(nir_instr *dest_instr, nir_src *dest, nir_src *src);
void nir_if_rewrite_condition(nir_if *if_stmt, nir_src new_src);
void nir_instr_rewrite_dest(nir_instr *instr, nir_dest *dest,
nir_dest new_dest);
void nir_ssa_dest_init(nir_instr *instr, nir_dest *dest,
unsigned num_components, unsigned bit_size,
const char *name);
void nir_ssa_def_init(nir_instr *instr, nir_ssa_def *def,
unsigned num_components, unsigned bit_size,
const char *name);
static inline void
nir_ssa_dest_init_for_type(nir_instr *instr, nir_dest *dest,
const struct glsl_type *type,
const char *name)
{
assert(glsl_type_is_vector_or_scalar(type));
nir_ssa_dest_init(instr, dest, glsl_get_components(type),
glsl_get_bit_size(type), name);
}
void nir_ssa_def_rewrite_uses(nir_ssa_def *def, nir_src new_src);
void nir_ssa_def_rewrite_uses_after(nir_ssa_def *def, nir_src new_src,
nir_instr *after_me);
nir_component_mask_t nir_ssa_def_components_read(const nir_ssa_def *def);
/*
* finds the next basic block in source-code order, returns NULL if there is
* none
*/
nir_block *nir_block_cf_tree_next(nir_block *block);
/* Performs the opposite of nir_block_cf_tree_next() */
nir_block *nir_block_cf_tree_prev(nir_block *block);
/* Gets the first block in a CF node in source-code order */
nir_block *nir_cf_node_cf_tree_first(nir_cf_node *node);
/* Gets the last block in a CF node in source-code order */
nir_block *nir_cf_node_cf_tree_last(nir_cf_node *node);
/* Gets the next block after a CF node in source-code order */
nir_block *nir_cf_node_cf_tree_next(nir_cf_node *node);
/* Macros for loops that visit blocks in source-code order */
#define nir_foreach_block(block, impl) \
for (nir_block *block = nir_start_block(impl); block != NULL; \
block = nir_block_cf_tree_next(block))
#define nir_foreach_block_safe(block, impl) \
for (nir_block *block = nir_start_block(impl), \
*next = nir_block_cf_tree_next(block); \
block != NULL; \
block = next, next = nir_block_cf_tree_next(block))
#define nir_foreach_block_reverse(block, impl) \
for (nir_block *block = nir_impl_last_block(impl); block != NULL; \
block = nir_block_cf_tree_prev(block))
#define nir_foreach_block_reverse_safe(block, impl) \
for (nir_block *block = nir_impl_last_block(impl), \
*prev = nir_block_cf_tree_prev(block); \
block != NULL; \
block = prev, prev = nir_block_cf_tree_prev(block))
#define nir_foreach_block_in_cf_node(block, node) \
for (nir_block *block = nir_cf_node_cf_tree_first(node); \
block != nir_cf_node_cf_tree_next(node); \
block = nir_block_cf_tree_next(block))
/* If the following CF node is an if, this function returns that if.
* Otherwise, it returns NULL.
*/
nir_if *nir_block_get_following_if(nir_block *block);
nir_loop *nir_block_get_following_loop(nir_block *block);
void nir_index_local_regs(nir_function_impl *impl);
void nir_index_global_regs(nir_shader *shader);
void nir_index_ssa_defs(nir_function_impl *impl);
unsigned nir_index_instrs(nir_function_impl *impl);
void nir_index_blocks(nir_function_impl *impl);
void nir_print_shader(nir_shader *shader, FILE *fp);
void nir_print_shader_annotated(nir_shader *shader, FILE *fp, struct hash_table *errors);
void nir_print_instr(const nir_instr *instr, FILE *fp);
nir_shader *nir_shader_clone(void *mem_ctx, const nir_shader *s);
nir_function_impl *nir_function_impl_clone(const nir_function_impl *fi);
nir_constant *nir_constant_clone(const nir_constant *c, nir_variable *var);
nir_variable *nir_variable_clone(const nir_variable *c, nir_shader *shader);
nir_shader *nir_shader_serialize_deserialize(void *mem_ctx, nir_shader *s);
#ifndef NDEBUG
void nir_validate_shader(nir_shader *shader, const char *when);
void nir_metadata_set_validation_flag(nir_shader *shader);
void nir_metadata_check_validation_flag(nir_shader *shader);
static inline bool
should_clone_nir(void)
{
static int should_clone = -1;
if (should_clone < 0)
should_clone = env_var_as_boolean("NIR_TEST_CLONE", false);
return should_clone;
}
static inline bool
should_serialize_deserialize_nir(void)
{
static int test_serialize = -1;
if (test_serialize < 0)
test_serialize = env_var_as_boolean("NIR_TEST_SERIALIZE", false);
return test_serialize;
}
static inline bool
should_print_nir(void)
{
static int should_print = -1;
if (should_print < 0)
should_print = env_var_as_boolean("NIR_PRINT", false);
return should_print;
}
#else
static inline void nir_validate_shader(nir_shader *shader, const char *when) { (void) shader; (void)when; }
static inline void nir_metadata_set_validation_flag(nir_shader *shader) { (void) shader; }
static inline void nir_metadata_check_validation_flag(nir_shader *shader) { (void) shader; }
static inline bool should_clone_nir(void) { return false; }
static inline bool should_serialize_deserialize_nir(void) { return false; }
static inline bool should_print_nir(void) { return false; }
#endif /* NDEBUG */
#define _PASS(pass, nir, do_pass) do { \
do_pass \
nir_validate_shader(nir, "after " #pass); \
if (should_clone_nir()) { \
nir_shader *clone = nir_shader_clone(ralloc_parent(nir), nir); \
ralloc_free(nir); \
nir = clone; \
} \
if (should_serialize_deserialize_nir()) { \
void *mem_ctx = ralloc_parent(nir); \
nir = nir_shader_serialize_deserialize(mem_ctx, nir); \
} \
} while (0)
#define NIR_PASS(progress, nir, pass, ...) _PASS(pass, nir, \
nir_metadata_set_validation_flag(nir); \
if (should_print_nir()) \
printf("%s\n", #pass); \
if (pass(nir, ##__VA_ARGS__)) { \
progress = true; \
if (should_print_nir()) \
nir_print_shader(nir, stdout); \
nir_metadata_check_validation_flag(nir); \
} \
)
#define NIR_PASS_V(nir, pass, ...) _PASS(pass, nir, \
if (should_print_nir()) \
printf("%s\n", #pass); \
pass(nir, ##__VA_ARGS__); \
if (should_print_nir()) \
nir_print_shader(nir, stdout); \
)
void nir_calc_dominance_impl(nir_function_impl *impl);
void nir_calc_dominance(nir_shader *shader);
nir_block *nir_dominance_lca(nir_block *b1, nir_block *b2);
bool nir_block_dominates(nir_block *parent, nir_block *child);
void nir_dump_dom_tree_impl(nir_function_impl *impl, FILE *fp);
void nir_dump_dom_tree(nir_shader *shader, FILE *fp);
void nir_dump_dom_frontier_impl(nir_function_impl *impl, FILE *fp);
void nir_dump_dom_frontier(nir_shader *shader, FILE *fp);
void nir_dump_cfg_impl(nir_function_impl *impl, FILE *fp);
void nir_dump_cfg(nir_shader *shader, FILE *fp);
int nir_gs_count_vertices(const nir_shader *shader);
bool nir_shrink_vec_array_vars(nir_shader *shader, nir_variable_mode modes);
bool nir_split_array_vars(nir_shader *shader, nir_variable_mode modes);
bool nir_split_var_copies(nir_shader *shader);
bool nir_split_per_member_structs(nir_shader *shader);
bool nir_split_struct_vars(nir_shader *shader, nir_variable_mode modes);
bool nir_lower_returns_impl(nir_function_impl *impl);
bool nir_lower_returns(nir_shader *shader);
bool nir_inline_functions(nir_shader *shader);
bool nir_propagate_invariant(nir_shader *shader);
void nir_lower_var_copy_instr(nir_intrinsic_instr *copy, nir_shader *shader);
void nir_lower_deref_copy_instr(struct nir_builder *b,
nir_intrinsic_instr *copy);
bool nir_lower_var_copies(nir_shader *shader);
void nir_fixup_deref_modes(nir_shader *shader);
bool nir_lower_global_vars_to_local(nir_shader *shader);
bool nir_lower_indirect_derefs(nir_shader *shader, nir_variable_mode modes);
bool nir_lower_locals_to_regs(nir_shader *shader);
void nir_lower_io_to_temporaries(nir_shader *shader,
nir_function_impl *entrypoint,
bool outputs, bool inputs);
void nir_shader_gather_info(nir_shader *shader, nir_function_impl *entrypoint);
void nir_assign_var_locations(struct exec_list *var_list, unsigned *size,
int (*type_size)(const struct glsl_type *));
/* Some helpers to do very simple linking */
bool nir_remove_unused_varyings(nir_shader *producer, nir_shader *consumer);
bool nir_remove_unused_io_vars(nir_shader *shader, struct exec_list *var_list,
uint64_t *used_by_other_stage,
uint64_t *used_by_other_stage_patches);
void nir_compact_varyings(nir_shader *producer, nir_shader *consumer,
bool default_to_smooth_interp);
void nir_link_xfb_varyings(nir_shader *producer, nir_shader *consumer);
bool nir_link_constant_varyings(nir_shader *producer, nir_shader *consumer);
typedef enum {
/* If set, this forces all non-flat fragment shader inputs to be
* interpolated as if with the "sample" qualifier. This requires
* nir_shader_compiler_options::use_interpolated_input_intrinsics.
*/
nir_lower_io_force_sample_interpolation = (1 << 1),
} nir_lower_io_options;
bool nir_lower_io(nir_shader *shader,
nir_variable_mode modes,
int (*type_size)(const struct glsl_type *),
nir_lower_io_options);
nir_src *nir_get_io_offset_src(nir_intrinsic_instr *instr);
nir_src *nir_get_io_vertex_index_src(nir_intrinsic_instr *instr);
bool nir_is_per_vertex_io(const nir_variable *var, gl_shader_stage stage);
bool nir_lower_regs_to_ssa_impl(nir_function_impl *impl);
bool nir_lower_regs_to_ssa(nir_shader *shader);
bool nir_lower_vars_to_ssa(nir_shader *shader);
bool nir_remove_dead_derefs(nir_shader *shader);
bool nir_remove_dead_derefs_impl(nir_function_impl *impl);
bool nir_remove_dead_variables(nir_shader *shader, nir_variable_mode modes);
bool nir_lower_constant_initializers(nir_shader *shader,
nir_variable_mode modes);
bool nir_move_load_const(nir_shader *shader);
bool nir_move_vec_src_uses_to_dest(nir_shader *shader);
bool nir_lower_vec_to_movs(nir_shader *shader);
void nir_lower_alpha_test(nir_shader *shader, enum compare_func func,
bool alpha_to_one);
bool nir_lower_alu(nir_shader *shader);
bool nir_lower_alu_to_scalar(nir_shader *shader);
bool nir_lower_load_const_to_scalar(nir_shader *shader);
bool nir_lower_read_invocation_to_scalar(nir_shader *shader);
bool nir_lower_phis_to_scalar(nir_shader *shader);
void nir_lower_io_arrays_to_elements(nir_shader *producer, nir_shader *consumer);
void nir_lower_io_arrays_to_elements_no_indirects(nir_shader *shader,
bool outputs_only);
void nir_lower_io_to_scalar(nir_shader *shader, nir_variable_mode mask);
void nir_lower_io_to_scalar_early(nir_shader *shader, nir_variable_mode mask);
typedef struct nir_lower_subgroups_options {
uint8_t subgroup_size;
uint8_t ballot_bit_size;
bool lower_to_scalar:1;
bool lower_vote_trivial:1;
bool lower_vote_eq_to_ballot:1;
bool lower_subgroup_masks:1;
bool lower_shuffle:1;
bool lower_shuffle_to_32bit:1;
bool lower_quad:1;
} nir_lower_subgroups_options;
bool nir_lower_subgroups(nir_shader *shader,
const nir_lower_subgroups_options *options);
bool nir_lower_system_values(nir_shader *shader);
typedef struct nir_lower_tex_options {
/**
* bitmask of (1 << GLSL_SAMPLER_DIM_x) to control for which
* sampler types a texture projector is lowered.
*/
unsigned lower_txp;
/**
* If true, lower away nir_tex_src_offset for all texelfetch instructions.
*/
bool lower_txf_offset;
/**
* If true, lower away nir_tex_src_offset for all rect textures.
*/
bool lower_rect_offset;
/**
* If true, lower rect textures to 2D, using txs to fetch the
* texture dimensions and dividing the texture coords by the
* texture dims to normalize.
*/
bool lower_rect;
/**
* If true, convert yuv to rgb.
*/
unsigned lower_y_uv_external;
unsigned lower_y_u_v_external;
unsigned lower_yx_xuxv_external;
unsigned lower_xy_uxvx_external;
unsigned lower_ayuv_external;
/**
* To emulate certain texture wrap modes, this can be used
* to saturate the specified tex coord to [0.0, 1.0]. The
* bits are according to sampler #, ie. if, for example:
*
* (conf->saturate_s & (1 << n))
*
* is true, then the s coord for sampler n is saturated.
*
* Note that clamping must happen *after* projector lowering
* so any projected texture sample instruction with a clamped
* coordinate gets automatically lowered, regardless of the
* 'lower_txp' setting.
*/
unsigned saturate_s;
unsigned saturate_t;
unsigned saturate_r;
/* Bitmask of textures that need swizzling.
*
* If (swizzle_result & (1 << texture_index)), then the swizzle in
* swizzles[texture_index] is applied to the result of the texturing
* operation.
*/
unsigned swizzle_result;
/* A swizzle for each texture. Values 0-3 represent x, y, z, or w swizzles
* while 4 and 5 represent 0 and 1 respectively.
*/
uint8_t swizzles[32][4];
/**
* Bitmap of textures that need srgb to linear conversion. If
* (lower_srgb & (1 << texture_index)) then the rgb (xyz) components
* of the texture are lowered to linear.
*/
unsigned lower_srgb;
/**
* If true, lower nir_texop_txd on cube maps with nir_texop_txl.
*/
bool lower_txd_cube_map;
/**
* If true, lower nir_texop_txd on shadow samplers (except cube maps)
* with nir_texop_txl. Notice that cube map shadow samplers are lowered
* with lower_txd_cube_map.
*/
bool lower_txd_shadow;
/**
* If true, lower nir_texop_txd on all samplers to a nir_texop_txl.
* Implies lower_txd_cube_map and lower_txd_shadow.
*/
bool lower_txd;
} nir_lower_tex_options;
bool nir_lower_tex(nir_shader *shader,
const nir_lower_tex_options *options);
bool nir_lower_idiv(nir_shader *shader);
bool nir_lower_clip_vs(nir_shader *shader, unsigned ucp_enables, bool use_vars);
bool nir_lower_clip_fs(nir_shader *shader, unsigned ucp_enables);
bool nir_lower_clip_cull_distance_arrays(nir_shader *nir);
void nir_lower_two_sided_color(nir_shader *shader);
bool nir_lower_clamp_color_outputs(nir_shader *shader);
void nir_lower_passthrough_edgeflags(nir_shader *shader);
bool nir_lower_patch_vertices(nir_shader *nir, unsigned static_count,
const gl_state_index16 *uniform_state_tokens);
typedef struct nir_lower_wpos_ytransform_options {
gl_state_index16 state_tokens[STATE_LENGTH];
bool fs_coord_origin_upper_left :1;
bool fs_coord_origin_lower_left :1;
bool fs_coord_pixel_center_integer :1;
bool fs_coord_pixel_center_half_integer :1;
} nir_lower_wpos_ytransform_options;
bool nir_lower_wpos_ytransform(nir_shader *shader,
const nir_lower_wpos_ytransform_options *options);
bool nir_lower_wpos_center(nir_shader *shader, const bool for_sample_shading);
typedef struct nir_lower_drawpixels_options {
gl_state_index16 texcoord_state_tokens[STATE_LENGTH];
gl_state_index16 scale_state_tokens[STATE_LENGTH];
gl_state_index16 bias_state_tokens[STATE_LENGTH];
unsigned drawpix_sampler;
unsigned pixelmap_sampler;
bool pixel_maps :1;
bool scale_and_bias :1;
} nir_lower_drawpixels_options;
void nir_lower_drawpixels(nir_shader *shader,
const nir_lower_drawpixels_options *options);
typedef struct nir_lower_bitmap_options {
unsigned sampler;
bool swizzle_xxxx;
} nir_lower_bitmap_options;
void nir_lower_bitmap(nir_shader *shader, const nir_lower_bitmap_options *options);
bool nir_lower_atomics_to_ssbo(nir_shader *shader, unsigned ssbo_offset);
typedef enum {
nir_lower_int_source_mods = 1 << 0,
nir_lower_float_source_mods = 1 << 1,
nir_lower_all_source_mods = (1 << 2) - 1
} nir_lower_to_source_mods_flags;
bool nir_lower_to_source_mods(nir_shader *shader, nir_lower_to_source_mods_flags options);
bool nir_lower_gs_intrinsics(nir_shader *shader);
typedef unsigned (*nir_lower_bit_size_callback)(const nir_alu_instr *, void *);
bool nir_lower_bit_size(nir_shader *shader,
nir_lower_bit_size_callback callback,
void *callback_data);
typedef enum {
nir_lower_imul64 = (1 << 0),
nir_lower_isign64 = (1 << 1),
/** Lower all int64 modulus and division opcodes */
nir_lower_divmod64 = (1 << 2),
} nir_lower_int64_options;
bool nir_lower_int64(nir_shader *shader, nir_lower_int64_options options);
typedef enum {
nir_lower_drcp = (1 << 0),
nir_lower_dsqrt = (1 << 1),
nir_lower_drsq = (1 << 2),
nir_lower_dtrunc = (1 << 3),
nir_lower_dfloor = (1 << 4),
nir_lower_dceil = (1 << 5),
nir_lower_dfract = (1 << 6),
nir_lower_dround_even = (1 << 7),
nir_lower_dmod = (1 << 8)
} nir_lower_doubles_options;
bool nir_lower_doubles(nir_shader *shader, nir_lower_doubles_options options);
bool nir_lower_pack(nir_shader *shader);
bool nir_normalize_cubemap_coords(nir_shader *shader);
void nir_live_ssa_defs_impl(nir_function_impl *impl);
void nir_loop_analyze_impl(nir_function_impl *impl,
nir_variable_mode indirect_mask);
bool nir_ssa_defs_interfere(nir_ssa_def *a, nir_ssa_def *b);
bool nir_repair_ssa_impl(nir_function_impl *impl);
bool nir_repair_ssa(nir_shader *shader);
void nir_convert_loop_to_lcssa(nir_loop *loop);
/* If phi_webs_only is true, only convert SSA values involved in phi nodes to
* registers. If false, convert all values (even those not involved in a phi
* node) to registers.
*/
bool nir_convert_from_ssa(nir_shader *shader, bool phi_webs_only);
bool nir_lower_phis_to_regs_block(nir_block *block);
bool nir_lower_ssa_defs_to_regs_block(nir_block *block);
bool nir_rematerialize_derefs_in_use_blocks_impl(nir_function_impl *impl);
bool nir_opt_algebraic(nir_shader *shader);
bool nir_opt_algebraic_before_ffma(nir_shader *shader);
bool nir_opt_algebraic_late(nir_shader *shader);
bool nir_opt_constant_folding(nir_shader *shader);
bool nir_opt_global_to_local(nir_shader *shader);
bool nir_copy_prop(nir_shader *shader);
bool nir_opt_copy_prop_vars(nir_shader *shader);
bool nir_opt_cse(nir_shader *shader);
bool nir_opt_dce(nir_shader *shader);
bool nir_opt_dead_cf(nir_shader *shader);
bool nir_opt_dead_write_vars(nir_shader *shader);
bool nir_opt_find_array_copies(nir_shader *shader);
bool nir_opt_gcm(nir_shader *shader, bool value_number);
bool nir_opt_if(nir_shader *shader);
bool nir_opt_intrinsics(nir_shader *shader);
bool nir_opt_large_constants(nir_shader *shader,
glsl_type_size_align_func size_align,
unsigned threshold);
bool nir_opt_loop_unroll(nir_shader *shader, nir_variable_mode indirect_mask);
bool nir_opt_move_comparisons(nir_shader *shader);
bool nir_opt_move_load_ubo(nir_shader *shader);
bool nir_opt_peephole_select(nir_shader *shader, unsigned limit);
bool nir_opt_remove_phis_impl(nir_function_impl *impl);
bool nir_opt_remove_phis(nir_shader *shader);
bool nir_opt_shrink_load(nir_shader *shader);
bool nir_opt_trivial_continues(nir_shader *shader);
bool nir_opt_undef(nir_shader *shader);
bool nir_opt_conditional_discard(nir_shader *shader);
void nir_sweep(nir_shader *shader);
void nir_remap_dual_slot_attributes(nir_shader *shader,
uint64_t *dual_slot_inputs);
uint64_t nir_get_single_slot_attribs_mask(uint64_t attribs, uint64_t dual_slot);
nir_intrinsic_op nir_intrinsic_from_system_value(gl_system_value val);
gl_system_value nir_system_value_from_intrinsic(nir_intrinsic_op intrin);
#ifdef __cplusplus
} /* extern "C" */
#endif
#endif /* NIR_H */
|