summaryrefslogtreecommitdiffstatshomepage
path: root/src/devices/video/voodoo.cpp
blob: 472f5c5fd47715753b5263a5de6005bacee5a3b3 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
// license:BSD-3-Clause
// copyright-holders:Aaron Giles
/***************************************************************************

    voodoo.c

    3dfx Voodoo Graphics SST-1/2 emulator.

****************************************************************************

    Specs:

    Voodoo 1 (SST1):
        2,4MB frame buffer RAM
        1,2,4MB texture RAM
        50MHz clock frequency
        clears @ 2 pixels/clock (RGB and depth simultaneously)
        renders @ 1 pixel/clock
        64 entry PCI FIFO
        memory FIFO up to 65536 entries

    --------------------------

    still to be implemented:
        * trilinear textures

    things to verify:
        * floating Z buffer


iterated RGBA = 12.12 [24 bits]
iterated Z    = 20.12 [32 bits]
iterated W    = 18.32 [48 bits]

>mamepm blitz
Stall PCI for HWM: 1
PCI FIFO Empty Entries LWM: D
LFB -> FIFO: 1
Texture -> FIFO: 1
Memory FIFO: 1
Memory FIFO HWM: 2000
Memory FIFO Write Burst HWM: 36
Memory FIFO LWM for PCI: 5
Memory FIFO row start: 120
Memory FIFO row rollover: 3FF
Video dither subtract: 0
DRAM banking: 1
Triple buffer: 0
Video buffer offset: 60
DRAM banking: 1

>mamepm wg3dh
Stall PCI for HWM: 1
PCI FIFO Empty Entries LWM: D
LFB -> FIFO: 1
Texture -> FIFO: 1
Memory FIFO: 1
Memory FIFO HWM: 2000
Memory FIFO Write Burst HWM: 36
Memory FIFO LWM for PCI: 5
Memory FIFO row start: C0
Memory FIFO row rollover: 3FF
Video dither subtract: 0
DRAM banking: 1
Triple buffer: 0
Video buffer offset: 40
DRAM banking: 1


As a point of reference, the 3D engine uses the following algorithm to calculate the linear memory address as a
function of the video buffer offset (fbiInit2 bits(19:11)), the number of 32x32 tiles in the X dimension (fbiInit1
bits(7:4) and bit(24)), X, and Y:

    tilesInX[4:0] = {fbiInit1[24], fbiInit1[7:4], fbiInit6[30]}
    rowBase = fbiInit2[19:11]
    rowStart = ((Y>>5) * tilesInX) >> 1

    if (!(tilesInX & 1))
    {
        rowOffset = (X>>6);
        row[9:0] = rowStart + rowOffset (for color buffer 0)
        row[9:0] = rowBase + rowStart + rowOffset (for color buffer 1)
        row[9:0] = (rowBase<<1) + rowStart + rowOffset (for depth/alpha buffer when double color buffering[fbiInit5[10:9]=0])
        row[9:0] = (rowBase<<1) + rowStart + rowOffset (for color buffer 2 when triple color buffering[fbiInit5[10:9]=1 or 2])
        row[9:0] = (rowBase<<1) + rowBase + rowStart + rowOffset (for depth/alpha buffer when triple color buffering[fbiInit5[10:9]=2])
        column[8:0] = ((Y % 32) <<4) + ((X % 32)>>1)
        ramSelect[1] = ((X&0x20) ? 1 : 0) (for color buffers)
        ramSelect[1] = ((X&0x20) ? 0 : 1) (for depth/alpha buffers)
    }
    else
    {
        rowOffset = (!(Y&0x20)) ? (X>>6) : ((X>31) ? (((X-32)>>6)+1) : 0)
        row[9:0] = rowStart + rowOffset (for color buffer 0)
        row[9:0] = rowBase + rowStart + rowOffset (for color buffer 1)
        row[9:0] = (rowBase<<1) + rowStart + rowOffset (for depth/alpha buffer when double color buffering[fbiInit5[10:9]=0])
        row[9:0] = (rowBase<<1) + rowStart + rowOffset (for color buffer 2 when triple color buffering[fbiInit5[10:9]=1 or 2])
        row[9:0] = (rowBase<<1) + rowBase + rowStart + rowOffset (for depth/alpha buffer when triple color buffering[fbiInit5[10:9]=2])
        column[8:0] = ((Y % 32) <<4) + ((X % 32)>>1)
        ramSelect[1] = (((X&0x20)^(Y&0x20)) ? 1 : 0) (for color buffers)
        ramSelect[1] = (((X&0x20)^(Y&0x20)) ? 0 : 1) (for depth/alpha buffers)
    }
    ramSelect[0] = X % 2
    pixelMemoryAddress[21:0] = (row[9:0]<<12) + (column[8:0]<<3) + (ramSelect[1:0]<<1)
    bankSelect = pixelMemoryAddress[21]

**************************************************************************/

/*

TODO:
 - look at speed on Konami games (nbapbp, racingj, etc)
 - look at timing issues on IT games
 - bad textures in some Voodoo 3 games (mocapb for example)
 - update callers to use maps

*/

#include "emu.h"
#include "voodoo.h"

using namespace voodoo;


//**************************************************************************
//  GLOBAL HELPERS
//**************************************************************************

//-------------------------------------------------
//  float_to_int32 - convert a floating-point
//  value in raw IEEE format into an integer with
//  the given number of fractional bits
//-------------------------------------------------

inline s32 float_to_int32(u32 data, int fixedbits)
{
	// compute the effective exponent
	int exponent = ((data >> 23) & 0xff) - 127 - 23 + fixedbits;

	// extract the mantissa and return the implied leading 1 bit
	s32 result = (data & 0x7fffff) | 0x800000;

	// shift by the exponent, handling minimum/maximum
	if (exponent < 0)
	{
		if (exponent > -32)
			result >>= -exponent;
		else
			result = 0;
	}
	else
	{
		if (exponent < 32)
			result <<= exponent;
		else
			result = 0x7fffffff;
	}

	// negate based on the sign
	return (data & 0x80000000) ? -result : result;
}


//-------------------------------------------------
//  float_to_int64 - convert a floating-point
//  value in raw IEEE format into an integer with
//  the given number of fractional bits
//-------------------------------------------------

inline s64 float_to_int64(u32 data, int fixedbits)
{
	// compute the effective exponent
	int exponent = ((data >> 23) & 0xff) - 127 - 23 + fixedbits;

	// extract the mantissa and return the implied leading 1 bit
	s64 result = (data & 0x7fffff) | 0x800000;

	// shift by the exponent, handling minimum/maximum
	if (exponent < 0)
	{
		if (exponent > -64)
			result >>= -exponent;
		else
			result = 0;
	}
	else
	{
		if (exponent < 64)
			result <<= exponent;
		else
			result = 0x7fffffffffffffffull;
	}

	// negate based on the sign
	return (data & 0x80000000) ? -result : result;
}


//**************************************************************************
//  VOODOO REGISTERS
//**************************************************************************

//-------------------------------------------------
//  register_save - save live state
//-------------------------------------------------

void voodoo_regs::register_save(save_proxy &save)
{
	save.save_item(NAME(m_regs));
	save.save_item(NAME(m_starts));
	save.save_item(NAME(m_startt));
	save.save_item(NAME(m_startw));
	save.save_item(NAME(m_dsdx));
	save.save_item(NAME(m_dtdx));
	save.save_item(NAME(m_dwdx));
	save.save_item(NAME(m_dsdy));
	save.save_item(NAME(m_dtdy));
	save.save_item(NAME(m_dwdy));
}


//-------------------------------------------------
//  s_alias_map - remap of first 64 registers
//-------------------------------------------------

u8 const voodoo_regs::s_alias_map[0x40] =
{
	voodoo_regs::reg_vdstatus,   0x004/4,                     voodoo_regs::reg_vertexAx,   voodoo_regs::reg_vertexAy,
	voodoo_regs::reg_vertexBx,   voodoo_regs::reg_vertexBy,   voodoo_regs::reg_vertexCx,   voodoo_regs::reg_vertexCy,
	voodoo_regs::reg_startR,     voodoo_regs::reg_dRdX,       voodoo_regs::reg_dRdY,       voodoo_regs::reg_startG,
	voodoo_regs::reg_dGdX,       voodoo_regs::reg_dGdY,       voodoo_regs::reg_startB,     voodoo_regs::reg_dBdX,
	voodoo_regs::reg_dBdY,       voodoo_regs::reg_startZ,     voodoo_regs::reg_dZdX,       voodoo_regs::reg_dZdY,
	voodoo_regs::reg_startA,     voodoo_regs::reg_dAdX,       voodoo_regs::reg_dAdY,       voodoo_regs::reg_startS,
	voodoo_regs::reg_dSdX,       voodoo_regs::reg_dSdY,       voodoo_regs::reg_startT,     voodoo_regs::reg_dTdX,
	voodoo_regs::reg_dTdY,       voodoo_regs::reg_startW,     voodoo_regs::reg_dWdX,       voodoo_regs::reg_dWdY,

	voodoo_regs::reg_triangleCMD,0x084/4,                     voodoo_regs::reg_fvertexAx,  voodoo_regs::reg_fvertexAy,
	voodoo_regs::reg_fvertexBx,  voodoo_regs::reg_fvertexBy,  voodoo_regs::reg_fvertexCx,  voodoo_regs::reg_fvertexCy,
	voodoo_regs::reg_fstartR,    voodoo_regs::reg_fdRdX,      voodoo_regs::reg_fdRdY,      voodoo_regs::reg_fstartG,
	voodoo_regs::reg_fdGdX,      voodoo_regs::reg_fdGdY,      voodoo_regs::reg_fstartB,    voodoo_regs::reg_fdBdX,
	voodoo_regs::reg_fdBdY,      voodoo_regs::reg_fstartZ,    voodoo_regs::reg_fdZdX,      voodoo_regs::reg_fdZdY,
	voodoo_regs::reg_fstartA,    voodoo_regs::reg_fdAdX,      voodoo_regs::reg_fdAdY,      voodoo_regs::reg_fstartS,
	voodoo_regs::reg_fdSdX,      voodoo_regs::reg_fdSdY,      voodoo_regs::reg_fstartT,    voodoo_regs::reg_fdTdX,
	voodoo_regs::reg_fdTdY,      voodoo_regs::reg_fstartW,    voodoo_regs::reg_fdWdX,      voodoo_regs::reg_fdWdY
};


//**************************************************************************
//  SHARED TABLES
//**************************************************************************

//-------------------------------------------------
//  shared_tables - constructor
//-------------------------------------------------

shared_tables::shared_tables()
{
	// configure the array of texel formats
	texel[0] = rgb332;
	texel[1] = nullptr;
	texel[2] = alpha8;
	texel[3] = int8;
	texel[4] = ai44;
	texel[5] = nullptr;
	texel[6] = nullptr;
	texel[7] = nullptr;
	texel[8] = rgb332;
	texel[9] = nullptr;
	texel[10] = rgb565;
	texel[11] = argb1555;
	texel[12] = argb4444;
	texel[13] = int8;
	texel[14] = nullptr;
	texel[15] = nullptr;

	// build static 8-bit texel tables
	for (int val = 0; val < 256; val++)
	{
		// 8-bit RGB (3-3-2)
		rgb332[val] = rgbexpand<3,3,2>(val, 5, 2, 0).set_a(0xff);

		// 8-bit alpha
		alpha8[val] = rgb_t(val, val, val, val);

		// 8-bit intensity
		int8[val] = rgb_t(0xff, val, val, val);

		// 8-bit alpha, intensity
		ai44[val] = argbexpand<4,4,4,4>(val, 4, 0, 0, 0);
	}

	// build static 16-bit texel tables
	for (int val = 0; val < 65536; val++)
	{
		// table 10 = 16-bit RGB (5-6-5)
		rgb565[val] = rgbexpand<5,6,5>(val, 11, 5, 0).set_a(0xff);

		// table 11 = 16 ARGB (1-5-5-5)
		argb1555[val] = argbexpand<1,5,5,5>(val, 15, 10, 5, 0);

		// table 12 = 16-bit ARGB (4-4-4-4)
		argb4444[val] = argbexpand<4,4,4,4>(val, 12, 8, 4, 0);
	}
}


//**************************************************************************
//  TMU STATE
//**************************************************************************

//-------------------------------------------------
//  tmu_state - constructor
//-------------------------------------------------

tmu_state::tmu_state() :
	m_index(0),
	m_ram(nullptr),
	m_mask(0),
	m_basemask(0xfffff),
	m_baseshift(3)
{
}


//-------------------------------------------------
//  init - configure local state
//-------------------------------------------------

void tmu_state::init(int index, shared_tables const &share, u8 *ram, u32 size)
{
	// configure texture RAM
	m_index = index;
	m_ram = ram;
	m_mask = size - 1;
	m_regdirty = true;
	m_palette_dirty[0] = m_palette_dirty[1] = m_palette_dirty[2] = m_palette_dirty[3] = true;
	m_texel_lookup = &share.texel[0];
}


//-------------------------------------------------
//  register_save - register for save states
//-------------------------------------------------

void tmu_state::register_save(save_proxy &save)
{
	// register state
	save.save_class(NAME(m_reg));
	save.save_item(NAME(m_palette));
}


//-------------------------------------------------
//  post_load - mark everything dirty following a
//  state load
//-------------------------------------------------

void tmu_state::post_load()
{
	m_regdirty = true;
	m_palette_dirty[0] = m_palette_dirty[1] = m_palette_dirty[2] = m_palette_dirty[3] = true;
}


//-------------------------------------------------
//  ncc_w - handle a write to the NCC/palette
//  registers
//-------------------------------------------------

void tmu_state::ncc_w(offs_t regnum, u32 data)
{
	u32 regindex = regnum - voodoo_regs::reg_nccTable;

	// I/Q entries in NCC 0 reference the palette if the high bit is set
	if (BIT(data, 31) && regindex >= 4 && regindex < 12)
	{
		// extract the palette index
		int const index = (BIT(data, 24, 7) << 1) | BIT(regindex, 0);

		// compute RGB and ARGB values
		rgb_t rgb = 0xff000000 | data;
		rgb_t argb = argbexpand<6,6,6,6>(data, 18, 12, 6, 0);

		// set and mark dirty
		if (m_palette[0][index] != rgb)
		{
			m_palette[0][index] = rgb;
			m_palette_dirty[0] = true;
		}
		if (m_palette[1][index] != argb)
		{
			m_palette[1][index] = argb;
			m_palette_dirty[1] = true;
		}
		return;
	}

	// if no delta, don't mark dirty
	if (m_reg.read(regnum) == data)
		return;

	// write the updated data and mark dirty
	m_reg.write(regnum, data);
	m_palette_dirty[2 + regindex / 12] = true;
}


//-------------------------------------------------
//  prepare_texture - handle updating the texture
//  state if the texture configuration is dirty
//-------------------------------------------------

inline rasterizer_texture &tmu_state::prepare_texture(voodoo_renderer &renderer)
{
	// if the texture parameters are dirty, update them
	if (m_regdirty)
	{
		// determine the lookup
		auto const texmode = m_reg.texture_mode();
		u32 const texformat = texmode.format();
		rgb_t const *lookup = m_texel_lookup[texformat];

		// if null lookup, then we need something dynamic
		if (lookup == nullptr)
		{
			// could be either straight palette or NCC table
			int palindex;
			if ((texformat & 7) == 1)
			{
				// NCC case: palindex = 2 or 3 based on table select
				palindex = 2 + texmode.ncc_table_select();
				if (m_palette_dirty[palindex])
				{
					u32 const *regs = m_reg.subset(voodoo_regs::reg_nccTable + 12 * (palindex & 1));
					renderer.alloc_palette(m_index * 4 + palindex).compute_ncc(regs);
				}
			}
			else
			{
				// palette case: palindex = 0 or 1 based on RGB vs RGBA
				palindex = (texformat == 6) ? 1 : 0;
				if (m_palette_dirty[palindex])
					renderer.alloc_palette(m_index * 4 + palindex).copy(&m_palette[palindex & 1][0]);
			}

			// clear the dirty flag and fetch the texels
			m_palette_dirty[palindex] = false;
			lookup = renderer.last_palette(m_index * 4 + palindex).texels();
		}

		// recompute the rasterization parameters
		renderer.alloc_texture(m_index).recompute(m_reg, m_ram, m_mask, lookup, m_basemask, m_baseshift);
		m_regdirty = false;
	}
	return renderer.last_texture(m_index);
}



//**************************************************************************
//  MEMORY FIFO
//**************************************************************************

//-------------------------------------------------
//  memory_fifo - constructor
//-------------------------------------------------

memory_fifo::memory_fifo() :
	m_base(nullptr),
	m_size(0),
	m_in(0),
	m_out(0)
{
}


//-------------------------------------------------
//  configure - set the base/size and reset
//-------------------------------------------------

void memory_fifo::configure(u32 *base, u32 size)
{
	m_base = base;
	m_size = size;
	reset();
}


//-------------------------------------------------
//  register_save - register for save states
//-------------------------------------------------

void memory_fifo::register_save(save_proxy &save)
{
	save.save_item(NAME(m_size));
	save.save_item(NAME(m_in));
	save.save_item(NAME(m_out));
}


//-------------------------------------------------
//  add - append an item to the fifo
//-------------------------------------------------

inline void memory_fifo::add(u32 data)
{
	// compute the value of 'in' after we add this item
	s32 next_in = m_in + 1;
	if (next_in >= m_size)
		next_in = 0;

	// as long as it's not equal to the output pointer, we can do it
	if (next_in != m_out)
	{
		m_base[m_in] = data;
		m_in = next_in;
	}
}


//-------------------------------------------------
//  remove - remove the next item from the fifo
//-------------------------------------------------

inline u32 memory_fifo::remove()
{
	// return invalid data if empty
	if (m_out == m_in)
		return 0xffffffff;

	// determine next output
	s32 next_out = m_out + 1;
	if (next_out >= m_size)
		next_out = 0;

	// fetch current and advance
	u32 data = m_base[m_out];
	m_out = next_out;
	return data;
}


//**************************************************************************
//  DEBUG STATS
//**************************************************************************

//-------------------------------------------------
//  debug_stats - constructor
//-------------------------------------------------

debug_stats::debug_stats() :
	m_lastkey(false),
	m_display(false)
{
	reset();
}


//-------------------------------------------------
//  add_emulation_stats - add in statistics from
//  the emulation stats
//-------------------------------------------------

void debug_stats::add_emulation_stats(thread_stats_block const &block)
{
	m_pixels_in += block.pixels_in;
	m_pixels_out += block.pixels_out;
	m_chroma_fail += block.chroma_fail;
	m_zfunc_fail += block.zfunc_fail;
	m_afunc_fail += block.afunc_fail;
	m_clipped += block.clip_fail;
	m_stippled += block.stipple_count;
}


//-------------------------------------------------
//  reset - reset per-swap statistics
//-------------------------------------------------

void debug_stats::reset()
{
	m_stalls = 0;
	m_triangles = 0;
	m_pixels_in = 0;
	m_pixels_out = 0;
	m_chroma_fail = 0;
	m_zfunc_fail = 0;
	m_afunc_fail = 0;
	m_clipped = 0;
	m_stippled = 0;
	m_reg_writes = 0;
	m_reg_reads = 0;
	m_lfb_writes = 0;
	m_lfb_reads = 0;
	m_tex_writes = 0;
	std::fill_n(&m_texture_mode[0], std::size(m_texture_mode), 0);
}


//-------------------------------------------------
//  update_string - compute the string to display
//  all the statistics
//-------------------------------------------------

void debug_stats::update_string(rectangle const &visarea, u32 swap_history)
{
	// create a string of texture modes used
	char texmodes[17] = { 0 };
	char *texptr = &texmodes[0];
	for (int mode = 0; mode < 16; mode++)
		if (m_texture_mode[mode])
			*texptr++ = "0123456789ABCDEF"[mode];
	*texptr = 0;

	// build the string
	m_string = string_format("Swap:%6d\n"
							 "Hist:%08X\n"
							 "Stal:%6d\n"
							 "Rend:%6d%%\n"
							 "Poly:%6d\n"
							 "PxIn:%6d\n"
							 "POut:%6d\n"
							 "Clip:%6d\n"
							 "Stip:%6d\n"
							 "Chro:%6d\n"
							 "ZFun:%6d\n"
							 "AFun:%6d\n"
							 "RegW:%6d\n"
							 "RegR:%6d\n"
							 "LFBW:%6d\n"
							 "LFBR:%6d\n"
							 "TexW:%6d\n"
							 "TexM:%s",
							 m_swaps, swap_history, m_stalls, m_pixels_out * 100 / (visarea.width() * visarea.height()),
							 m_triangles, m_pixels_in, m_pixels_out, m_clipped, m_stippled,
							 m_chroma_fail, m_zfunc_fail, m_afunc_fail,
							 m_reg_writes, m_reg_reads, m_lfb_writes, m_lfb_reads, m_tex_writes, texmodes);
}


//-------------------------------------------------
//  update_display_state - based on the current key
//  state, update and return whether stats should
//  be shown
//-------------------------------------------------

bool debug_stats::update_display_state(bool key_pressed)
{
	if (key_pressed && key_pressed != m_lastkey)
		m_display = !m_display;
	m_lastkey = key_pressed;
	return m_display;
}


//**************************************************************************
//  GENERIC VOODOO DEVICE
//**************************************************************************

//-------------------------------------------------
//  generic_voodoo_device - constructor
//-------------------------------------------------

generic_voodoo_device::generic_voodoo_device(const machine_config &mconfig, device_type type, const char *tag, device_t *owner, u32 clock, voodoo_model model) :
	device_t(mconfig, type, tag, owner, clock),
	device_video_interface(mconfig, *this),
	m_model(model),
	m_fbmem_in_mb(0),
	m_tmumem0_in_mb(0),
	m_tmumem1_in_mb(0),
	m_status_cycles(0),
	m_cpu(*this, finder_base::DUMMY_TAG),
	m_vblank_cb(*this),
	m_stall_cb(*this),
	m_pciint_cb(*this)
{
}


//-------------------------------------------------
//  device_start - device startup
//-------------------------------------------------

void generic_voodoo_device::device_start()
{
	// resolve callbacks
	m_vblank_cb.resolve();
	m_stall_cb.resolve();
	m_pciint_cb.resolve();
}


//**************************************************************************
//  VOODOO 1 DEVICE
//**************************************************************************

//-------------------------------------------------
//  voodoo_1_device - constructor
//-------------------------------------------------

DEFINE_DEVICE_TYPE(VOODOO_1, voodoo_1_device, "voodoo_1", "3dfx Voodoo Graphics")

voodoo_1_device::voodoo_1_device(const machine_config &mconfig, device_type type, const char *tag, device_t *owner, u32 clock, voodoo_model model) :
	generic_voodoo_device(mconfig, type, tag, owner, clock, model),
	m_chipmask(1),
	m_init_enable(0),
	m_stall_state(NOT_STALLED),
	m_stall_trigger(0),
	m_operation_end(attotime::zero),
	m_flush_flag(false),
	m_fbram(nullptr),
	m_fbmask(0),
	m_rgboffs{ u32(~0), u32(~0), u32(~0) },
	m_auxoffs(~0),
	m_frontbuf(0),
	m_backbuf(1),
	m_video_changed(true),
	m_lfb_stride(0),
	m_width(512),
	m_height(384),
	m_xoffs(0),
	m_yoffs(0),
	m_vsyncstart(0),
	m_vsyncstop(0),
	m_swaps_pending(0),
	m_vblank(0),
	m_vblank_count(0),
	m_vblank_swap_pending(0),
	m_vblank_swap(0),
	m_vblank_dont_swap(0),
	m_vsync_start_timer(nullptr),
	m_vsync_stop_timer(nullptr),
	m_stall_resume_timer(nullptr),
	m_last_status_pc(0),
	m_last_status_value(0),
	m_clut_dirty(true),
	m_clut(33),
	m_pen(65536)
{
	for (int index = 0; index < std::size(m_regtable); index++)
		m_regtable[index].unpack(s_register_table[index], *this);
}


//-------------------------------------------------
//  ~voodoo_1_device - destructor
//-------------------------------------------------

voodoo_1_device::~voodoo_1_device()
{
}


//-------------------------------------------------
//  core_map - device map for core memory access
//-------------------------------------------------

void voodoo_1_device::core_map(address_map &map)
{
	// Voodoo-1 memory map:
	//
	//   00ab----`--ccccrr`rrrrrr-- Register access
	//                                a = alternate register map if fbi_init3().tri_register_remap()
	//                                b = byte swizzle data if fbi_init0().swizzle_reg_writes()
	//                                c = chip mask select
	//                                r = register index ($00-$FF)
	//   01-yyyyy`yyyyyxxx`xxxxxxx- Linear frame buffer access (16-bit)
	//   01yyyyyy`yyyyxxxx`xxxxxx-- Linear frame buffer access (32-bit)
	//   1-ccllll`tttttttt`sssssss- Texture memory access, where:
	//                                c = chip mask select
	//                                l = LOD
	//                                t = Y index
	//                                s = X index
	//
	map(0x000000, 0x3fffff).rw(FUNC(voodoo_1_device::map_register_r), FUNC(voodoo_1_device::map_register_w));
	map(0x400000, 0x7fffff).rw(FUNC(voodoo_1_device::map_lfb_r), FUNC(voodoo_1_device::map_lfb_w));
	map(0x800000, 0xffffff).w(FUNC(voodoo_1_device::map_texture_w));
}


//-------------------------------------------------
//  read - generic read handler until everyone is
//  using the memory map
//-------------------------------------------------

u32 voodoo_1_device::read(offs_t offset, u32 mem_mask)
{
	switch (offset >> (22-2))
	{
		case 0x000000 >> 22:
			return map_register_r(offset);

		case 0x400000 >> 22:
			return map_lfb_r(offset - 0x400000/4);

		default:
			return 0xffffffff;
	}
}


//-------------------------------------------------
//  write - generic write handler until everyone is
//  using the memory map
//-------------------------------------------------

void voodoo_1_device::write(offs_t offset, u32 data, u32 mem_mask)
{
	switch (offset >> (22-2))
	{
		case 0x000000 >> 22:
			map_register_w(offset, data, mem_mask);
			break;

		case 0x400000 >> 22:
			map_lfb_w(offset - 0x400000/4, data, mem_mask);
			break;

		case 0x800000 >> 22:
		case 0xc00000 >> 22:
			map_texture_w(offset - 0x800000/4, data, mem_mask);
			break;
	}
}


//-------------------------------------------------
//  set_init_enable - set the externally-controlled
//  init_en register
//-------------------------------------------------

void voodoo_1_device::set_init_enable(u32 newval)
{
	m_init_enable = reg_init_en(newval);
	if (LOG_REGISTERS)
		logerror("VOODOO.REG:initEnable write = %08X\n", newval);
}


//-------------------------------------------------
//  update - update the screen bitmap
//-------------------------------------------------

int voodoo_1_device::update(bitmap_rgb32 &bitmap, const rectangle &cliprect)
{
	// if we are blank, just fill with black
	if (m_reg.fbi_init1().software_blank())
	{
		bitmap.fill(0, cliprect);
		int changed = m_video_changed;
		m_video_changed = false;
		return changed;
	}

	// if the CLUT is dirty, recompute the pens array
	if (m_clut_dirty)
	{
		rgb_t const *clutbase = &m_clut[0];

		// kludge: some of the Midway games write 0 to the last entry when they obviously mean FF
		if ((m_clut[32] & 0xffffff) == 0 && (m_clut[31] & 0xffffff) != 0)
			m_clut[32] = 0x20ffffff;

		// compute the R/B pens first
		u8 rtable[32], gtable[64], btable[32];
		for (u32 rawcolor = 0; rawcolor < 32; rawcolor++)
		{
			// treat rawcolor as a 5-bit value, scale up to 8 bits, and linear interpolate for red/blue
			u32 color = pal5bit(rawcolor);
			rtable[rawcolor] = (clutbase[color >> 3].r() * (8 - (color & 7)) + clutbase[(color >> 3) + 1].r() * (color & 7)) >> 3;
			btable[rawcolor] = (clutbase[color >> 3].b() * (8 - (color & 7)) + clutbase[(color >> 3) + 1].b() * (color & 7)) >> 3;
		}

		// then the G pens
		for (u32 rawcolor = 0; rawcolor < 64; rawcolor++)
		{
			// treat rawcolor as a 6-bit value, scale up to 8 bits, and linear interpolate
			u32 color = pal6bit(rawcolor);
			gtable[rawcolor] = (clutbase[color >> 3].g() * (8 - (color & 7)) + clutbase[(color >> 3) + 1].g() * (color & 7)) >> 3;
		}

		// now assemble the values into their final form
		for (u32 pen = 0; pen < 65536; pen++)
			m_pen[pen] = rgb_t(rtable[BIT(pen, 11, 5)], gtable[BIT(pen, 5, 6)], btable[BIT(pen, 0, 5)]);

		// no longer dirty
		m_clut_dirty = false;
		m_video_changed = true;
	}
	return update_common(bitmap, cliprect, &m_pen[0]);
}


//-------------------------------------------------
//  device_start - device startup
//-------------------------------------------------

void voodoo_1_device::device_start()
{
	// resolve configuration-related items
	generic_voodoo_device::device_start();

	// validate configuration
	if (m_fbmem_in_mb == 0)
		fatalerror("%s: Invalid Voodoo memory configuration", tag());
	if (!BIT(m_chipmask, 1) && m_tmumem0_in_mb == 0)
		fatalerror("%s: Invalid Voodoo memory configuration", tag());

	// create shared tables
	m_shared = std::make_unique<shared_tables>();
	voodoo::dither_helper::init_static();

	// determine our index within the system, then set our trigger
	u32 index = 0;
	for (device_t &scan : device_enumerator(machine().root_device()))
		if (scan.type() == this->type())
		{
			if (&scan == this)
				break;
			index++;
		}
	m_stall_trigger = 51324 + index;

	// allocate timers for VBLANK
	m_vsync_stop_timer = machine().scheduler().timer_alloc(timer_expired_delegate(FUNC(voodoo_1_device::vblank_stop), this), this);
	m_vsync_start_timer = machine().scheduler().timer_alloc(timer_expired_delegate(FUNC(voodoo_1_device::vblank_start),this), this);

	// add TMUs to the chipmask if memory is specified (later chips leave
	// the tmumem values at 0 and set the chipmask directly to indicate
	// that RAM is shared)
	if (m_tmumem0_in_mb != 0)
	{
		m_chipmask |= 2;
		if (m_tmumem1_in_mb != 0)
			m_chipmask |= 4;
	}

	// allocate memory
	u32 total_allocation = m_fbmem_in_mb + m_tmumem0_in_mb + m_tmumem1_in_mb;
	m_memory = std::make_unique<u8[]>(total_allocation * 1024 * 1024 + 4096);

	// configure frame buffer memory, aligning the base to a 4k boundary
	m_fbram = (u8 *)(((uintptr_t(m_memory.get()) + 4095) >> 12) << 12);
	m_fbmask = m_fbmem_in_mb * 1024 * 1024 - 1;

	// configure texture memory
	u8 *tmumem[2] = { nullptr, nullptr };
	u8 tmusize[2] = { m_tmumem0_in_mb, m_tmumem1_in_mb };
	if (tmusize[0] != 0)
	{
		// separate framebuffer and texture RAM (Voodoo 1/2)
		tmumem[0] = m_fbram + m_fbmem_in_mb * 1024 * 1024;
		tmumem[1] = tmumem[0] + tmusize[0] * 1024 * 1024;
	}
	else
	{
		// shared framebuffer and texture RAM (Voodoo Banshee/3)
		tmumem[0] = tmumem[1] = m_fbram;
		tmusize[0] = tmusize[1] = m_fbmem_in_mb;
	}

	// initialize the frame buffer
	m_rgboffs[0] = m_rgboffs[1] = m_rgboffs[2] = 0;
	m_auxoffs = ~0;

	m_frontbuf = 0;
	m_backbuf = 1;
	m_swaps_pending = 0;
	m_video_changed = true;

	m_lfb_stride = 10;

	m_width = 512;
	m_height = 384;
	m_xoffs = 0;
	m_yoffs = 0;
	m_vsyncstart = 0;
	m_vsyncstop = 0;

	m_vblank = 0;
	m_vblank_count = 0;
	m_vblank_swap_pending = 0;
	m_vblank_swap = 0;
	m_vblank_dont_swap = 0;

	m_lfb_stats.reset();

	// initialize the memory FIFO
	m_fbmem_fifo.configure(nullptr, 0);

	// initialize the CLUT
	for (int pen = 0; pen < 32; pen++)
		m_clut[pen] = rgb_t(pen, pal5bit(pen), pal5bit(pen), pal5bit(pen));
	m_clut[32] = rgb_t(32,0xff,0xff,0xff);
	m_clut_dirty = true;

	// initialize the TMUs
	u16 tmu_config = 0x11;
	m_tmu[0].init(0, *m_shared.get(), tmumem[0], tmusize[0] * 1024 * 1024);
	if (BIT(m_chipmask, 2))
	{
		m_tmu[1].init(1, *m_shared.get(), tmumem[1], tmusize[1] * 1024 * 1024);
		tmu_config |= 0xc0;
	}

	// create the renderer
	m_renderer = std::make_unique<voodoo_renderer>(machine(), tmu_config, m_shared->rgb565, m_reg, &m_tmu[0].regs(), BIT(m_chipmask, 2) ? &m_tmu[1].regs() : nullptr);

	// set up the PCI FIFO
	m_pci_fifo.configure(m_pci_fifo_mem, 64*2);
	m_stall_state = NOT_STALLED;
	m_stall_resume_timer = machine().scheduler().timer_alloc(timer_expired_delegate(FUNC(voodoo_1_device::stall_resume_callback), this));

	// initialize registers
	m_init_enable = 0;
	m_reg.write(voodoo_regs::reg_fbiInit0, (1 << 4) | (0x10 << 6));
	m_reg.write(voodoo_regs::reg_fbiInit1, (1 << 1) | (1 << 8) | (1 << 12) | (2 << 20));
	m_reg.write(voodoo_regs::reg_fbiInit2, (1 << 6) | (0x100 << 23));
	m_reg.write(voodoo_regs::reg_fbiInit3, (2 << 13) | (0xf << 17));
	m_reg.write(voodoo_regs::reg_fbiInit4, (1 << 0));

	// do a soft reset to reset everything else
	soft_reset();

	// register for save states
	save_proxy save(*this);
	register_save(save, total_allocation);
}


//-------------------------------------------------
//  device_stop - device-specific stop
//-------------------------------------------------

void voodoo_1_device::device_stop()
{
	m_renderer->wait("device_stop");
}


//-------------------------------------------------
//  device_reset - device-specific reset
//-------------------------------------------------

void voodoo_1_device::device_reset()
{
	soft_reset();
}


//-------------------------------------------------
//  device_post_load - update after loading save
//  state
//-------------------------------------------------

void voodoo_1_device::device_post_load()
{
	// dirty everything so it gets recomputed
	m_clut_dirty = true;
	for (tmu_state &tm : m_tmu)
		tm.post_load();

	// recompute FBI memory FIFO to get the base pointer set
	if (m_fbmem_fifo.configured())
		recompute_fbmem_fifo();
}


//-------------------------------------------------
//  soft_reset - handle reset when initiated by
//  a register write
//-------------------------------------------------

void voodoo_1_device::soft_reset()
{
	reset_counters();
	m_reg.write(voodoo_regs::reg_fbiTrianglesOut, 0);
	m_pci_fifo.reset();
	m_fbmem_fifo.reset();
}


//-------------------------------------------------
//  register_save - register items for save states
//-------------------------------------------------

ALLOW_SAVE_TYPE(reg_init_en);
ALLOW_SAVE_TYPE(voodoo_regs::register_data);
ALLOW_SAVE_TYPE(voodoo_1_device::stall_state);

void voodoo_1_device::register_save(save_proxy &save, u32 total_allocation)
{
	// PCI state/FIFOs
	save.save_item(NAME(m_init_enable));
	save.save_item(NAME(m_stall_state));
	save.save_item(NAME(m_operation_end));
	save.save_class(NAME(m_pci_fifo));
	save.save_class(NAME(m_fbmem_fifo));

	// allocated memory
	save.save_pointer(NAME(m_fbram), 1024 * 1024 * total_allocation);
	save.save_class(NAME(*m_renderer.get()));

	// video buffer configuration
	save.save_item(NAME(m_rgboffs));
	save.save_item(NAME(m_auxoffs));
	save.save_item(NAME(m_frontbuf));
	save.save_item(NAME(m_backbuf));

	// linear frame buffer access configuration
	save.save_item(NAME(m_lfb_stride));

	// video configuration
	save.save_item(NAME(m_width));
	save.save_item(NAME(m_height));
	save.save_item(NAME(m_xoffs));
	save.save_item(NAME(m_yoffs));
	save.save_item(NAME(m_vsyncstart));
	save.save_item(NAME(m_vsyncstop));

	// VBLANK/swapping state
	save.save_item(NAME(m_swaps_pending));
	save.save_item(NAME(m_vblank));
	save.save_item(NAME(m_vblank_count));
	save.save_item(NAME(m_vblank_swap_pending));
	save.save_item(NAME(m_vblank_swap));
	save.save_item(NAME(m_vblank_dont_swap));

	// register state
	save.save_class(NAME(m_reg));
	save.save_class(NAME(m_tmu[0]));
	save.save_class(NAME(m_tmu[1]));
	save.save_item(NAME(m_dac_reg));
	save.save_item(NAME(m_dac_read_result));

	// memory for PCI FIFO
	save.save_item(NAME(m_pci_fifo_mem));

	// pens and CLUT
	save.save_item(NAME(m_clut));
}


//-------------------------------------------------
//  draw_buffer_indirect - given a 2-bit index,
//  return the front/back buffer for drawing
//-------------------------------------------------

u16 *voodoo_1_device::draw_buffer_indirect(int index)
{
	switch (index)
	{
		case 0: m_video_changed = true; return front_buffer();
		case 1: return back_buffer();
		default: return nullptr;
	}
}


//-------------------------------------------------
//  lfb_buffer_indirect - given a 2-bit index,
//  return the front/back/depth buffer for LFB
//  access
//-------------------------------------------------

u16 *voodoo_1_device::lfb_buffer_indirect(int index)
{
	switch (index)
	{
		case 0: m_video_changed = true; return front_buffer();
		case 1: return back_buffer();
		case 2: return aux_buffer();
		default: return nullptr;
	}
}


//-------------------------------------------------
//  prepare_for_read - handle housekeeping before
//  processing a direct PCI read
//-------------------------------------------------

void voodoo_1_device::prepare_for_read()
{
	// if we have something pending, flush the FIFOs up to the current time
	if (operation_pending())
		flush_fifos(machine().time());
}


//-------------------------------------------------
//  prepare_for_write - handle housekeeping before
//  processing a direct PCI write
//-------------------------------------------------

bool voodoo_1_device::prepare_for_write()
{
	// should not be getting accesses while stalled (but we do)
	if (m_stall_state != NOT_STALLED)
		logerror("voodoo_1_device::write while stalled!\n");

	// if we have something pending, flush the FIFOs up to the current time
	bool pending = operation_pending();
	if (pending)
	{
		flush_fifos(machine().time());
		pending = operation_pending();
	}
	return pending;
}


//-------------------------------------------------
//  recompute_fbmem_fifo - recompute and configure
//  the framebuffer RAM-based FIFO based on the
//  fbiInit registers
//-------------------------------------------------

void voodoo_1_device::recompute_fbmem_fifo()
{
	// compute the memory FIFO location and size
	u32 fifo_last_page = m_reg.fbi_init4().memory_fifo_stop_row();
	if (fifo_last_page > m_fbmask / 0x1000)
		fifo_last_page = m_fbmask / 0x1000;

	// is it valid and enabled?
	u32 const fifo_start_page = m_reg.fbi_init4().memory_fifo_start_row();
	if (fifo_start_page <= fifo_last_page && m_reg.fbi_init0().enable_memory_fifo())
	{
		u32 size = std::min<u32>((fifo_last_page + 1 - fifo_start_page) * 0x1000 / 4, 65536*2);
		m_fbmem_fifo.configure((u32 *)(m_fbram + fifo_start_page * 0x1000), size);
	}

	// if not, disable the FIFO
	else
		m_fbmem_fifo.configure(nullptr, 0);
}


//-------------------------------------------------
//  add_to_fifo - add a write to the PCI FIFO,
//  spilling to the memory FIFO as configured
//-------------------------------------------------

void voodoo_1_device::add_to_fifo(u32 offset, u32 data, u32 mem_mask)
{
	// add flags to the offset based on the mem_mask
	if (!ACCESSING_BITS_16_31)
		offset |= memory_fifo::NO_16_31;
	if (!ACCESSING_BITS_0_15)
		offset |= memory_fifo::NO_0_15;

	// if there's room in the PCI FIFO, add there
	if (LOG_FIFO_VERBOSE)
		logerror("VOODOO.%d.FIFO:adding to PCI FIFO @ %08X=%08X\n", this, offset, data);
	assert(!m_pci_fifo.full());

	// add as offset/data pair
	m_pci_fifo.add(offset);
	m_pci_fifo.add(data);

	// handle flushing to the memory FIFO
	if (m_reg.fbi_init0().enable_memory_fifo() && m_pci_fifo.space() <= 2 * m_reg.fbi_init4().memory_fifo_lwm())
	{
		u8 valid[4];

		// determine which types of data can go to the memory FIFO
		valid[0] = true;
		valid[1] = m_reg.fbi_init0().lfb_to_memory_fifo();
		valid[2] = valid[3] = m_reg.fbi_init0().texmem_to_memory_fifo();

		// flush everything we can
		if (LOG_FIFO_VERBOSE) logerror("VOODOO.FIFO:moving PCI FIFO to memory FIFO\n");
		while (!m_pci_fifo.empty() && valid[(m_pci_fifo.peek() >> 22) & 3])
		{
			m_fbmem_fifo.add(m_pci_fifo.remove());
			m_fbmem_fifo.add(m_pci_fifo.remove());
		}

		// if we're above the HWM as a result, stall
		if (m_reg.fbi_init0().stall_pcie_for_hwm() && m_fbmem_fifo.items() >= 2 * 32 * m_reg.fbi_init0().memory_fifo_hwm())
		{
			if (LOG_FIFO) logerror("VOODOO.FIFO:hit memory FIFO HWM -- stalling\n");
			stall_cpu(STALLED_UNTIL_FIFO_LWM);
		}
	}

	// if we're at the LWM for the PCI FIFO, stall
	if (m_reg.fbi_init0().stall_pcie_for_hwm() && m_pci_fifo.space() <= 2 * m_reg.fbi_init0().pci_fifo_lwm())
	{
		if (LOG_FIFO) logerror("VOODOO.FIFO:hit PCI FIFO free LWM -- stalling\n");
		stall_cpu(STALLED_UNTIL_FIFO_LWM);
	}
}


//-------------------------------------------------
//  flush_fifos - flush data out of FIFOs up to
//  the current time
//-------------------------------------------------

void voodoo_1_device::flush_fifos(attotime current_time)
{
	// check for recursive calls
	if (m_flush_flag)
		return;
	m_flush_flag = true;

	// should only be called if something is pending
	assert(operation_pending());

	if (LOG_FIFO_VERBOSE)
		logerror("VOODOO.FIFO:flush_fifos start -- pending=%s cur=%s\n", m_operation_end.as_string(18), current_time.as_string(18));

	// loop while we still have cycles to burn
	while (m_operation_end <= current_time)
	{
		// execute from the FIFOs until we get something that's non-zero
		u32 cycles = execute_fifos();

		// if nothing remains, we're done; clear the flags
		if (cycles == 0)
		{
			clear_pending_operation();
			if (LOG_FIFO_VERBOSE)
				logerror("VOODOO.FIFO:flush_fifos end -- FIFOs empty\n");
			m_flush_flag = false;
			return;
		}

		// account for those cycles
		m_operation_end += clocks_to_attotime(cycles);

		if (LOG_FIFO_VERBOSE)
			logerror("VOODOO.FIFO:update -- pending=%s cur=%s\n", m_operation_end.as_string(18), current_time.as_string(18));
	}

	if (LOG_FIFO_VERBOSE)
		logerror("VOODOO.FIFO:flush_fifos end -- pending command complete at %s\n", m_operation_end.as_string(18));

	m_flush_flag = false;
}


//-------------------------------------------------
//  execute_fifos - execute commands from the FIFOs
//  until a non-zero cycle count operation is run
//-------------------------------------------------

u32 voodoo_1_device::execute_fifos()
{
	// loop until FIFOs are empty or until we get a non-zero cycle count
	while (1)
	{
		// prioritize framebuffer FIFO over PCI FIFO
		voodoo::memory_fifo &memfifo = !m_fbmem_fifo.empty() ? m_fbmem_fifo : m_pci_fifo;

		// if empty, return 0
		if (memfifo.empty())
			return 0;

		// extract address and data
		u32 offset = memfifo.remove();
		u32 data = memfifo.remove();

		// target the appropriate location
		switch (offset & memory_fifo::TYPE_MASK)
		{
			case memory_fifo::TYPE_REGISTER:
			{
				// just use the chipmask raw since it was adjusted prior to being added to the FIFO
				u32 regnum = BIT(offset, 0, 8);
				u32 chipmask = BIT(offset, 8, 4);

				// if we got a non-zero number of cycles back, return
				u32 cycles = m_regtable[regnum].write(*this, chipmask, regnum, data);
				if (cycles > 0)
					return cycles;
				break;
			}

			case memory_fifo::TYPE_TEXTURE:
				internal_texture_w(offset & ~memory_fifo::FLAGS_MASK, data);
				break;

			case memory_fifo::TYPE_LFB:
			{
				u32 mem_mask = 0xffffffff;
				if (offset & memory_fifo::NO_16_31)
					mem_mask &= 0x0000ffff;
				if (offset & memory_fifo::NO_0_15)
					mem_mask &= 0xffff0000;
				internal_lfb_w(offset & ~memory_fifo::FLAGS_MASK, data, mem_mask);
				break;
			}
		}
	}
}


//-------------------------------------------------
//  map_register_r - handle a mapped read from
//  regular register space
//-------------------------------------------------

u32 voodoo_1_device::map_register_r(offs_t offset)
{
	prepare_for_read();

	// extract chipmask and register
	u32 chipmask = chipmask_from_offset(offset);
	u32 regnum = BIT(offset, 0, 8);
	return m_regtable[regnum].read(*this, chipmask, regnum);
}


//-------------------------------------------------
//  map_lfb_r - handle a mapped read from LFB space
//-------------------------------------------------

u32 voodoo_1_device::map_lfb_r(offs_t offset)
{
	prepare_for_read();
	return internal_lfb_r(offset);
}


//-------------------------------------------------
//  map_register_w - handle a mapped write to
//  regular register space
//-------------------------------------------------

void voodoo_1_device::map_register_w(offs_t offset, u32 data, u32 mem_mask)
{
	bool pending = prepare_for_write();

	// extract chipmask and register
	u32 chipmask = chipmask_from_offset(offset);
	u32 regnum = BIT(offset, 0, 8);

	// handle register swizzling -- manual says bit 21; voodoo2 manual says bit 20
	// guessing it does not overlap with the alternate register mapping bit
	if (BIT(offset, 20-2) && m_reg.fbi_init0().swizzle_reg_writes())
		data = swapendian_int32(data);

	// handle aliasing
	if (BIT(offset, 21-2) && m_reg.fbi_init3().tri_register_remap())
		regnum = voodoo_regs::alias(regnum);

	// look up the register
	auto const &regentry = m_regtable[regnum];

	// if this is non-FIFO command, execute immediately
	if (!regentry.is_fifo())
		return void(regentry.write(*this, chipmask, regnum, data));

	// track swap buffer commands seen
	if (regnum == voodoo_regs::reg_swapbufferCMD)
		m_swaps_pending++;

	// if we're busy add to the FIFO
	if (pending && m_init_enable.enable_pci_fifo())
		return add_to_fifo(memory_fifo::TYPE_REGISTER | (chipmask << 8) | regnum, data, mem_mask);

	// if we get a non-zero number of cycles back, mark things pending
	u32 cycles = regentry.write(*this, chipmask, regnum, data);
	if (cycles > 0)
	{
		m_operation_end = machine().time() + clocks_to_attotime(cycles);
		if (LOG_FIFO_VERBOSE)
			logerror("VOODOO.FIFO:direct write start at %s end at %s\n", machine().time().as_string(18), m_operation_end.as_string(18));
	}
}


//-------------------------------------------------
//  map_lfb_w - handle a mapped write to LFB space
//-------------------------------------------------

void voodoo_1_device::map_lfb_w(offs_t offset, u32 data, u32 mem_mask)
{
	// if we're busy add to the FIFO, else just execute immediately
	if (prepare_for_write() && m_init_enable.enable_pci_fifo())
		add_to_fifo(memory_fifo::TYPE_LFB | offset, data, mem_mask);
	else
		internal_lfb_w(offset, data, mem_mask);
}


//-------------------------------------------------
//  map_texture_w - handle a mapped write to
//  texture space
//-------------------------------------------------

void voodoo_1_device::map_texture_w(offs_t offset, u32 data, u32 mem_mask)
{
	// if we're busy add to the FIFO, else just execute immediately
	if (prepare_for_write() && m_init_enable.enable_pci_fifo())
		add_to_fifo(memory_fifo::TYPE_TEXTURE | offset, data, mem_mask);
	else
		internal_texture_w(offset, data);
}


//-------------------------------------------------
//  internal_lfb_r - handle a read from the linear
//  frame buffer
//-------------------------------------------------

u32 voodoo_1_device::internal_lfb_r(offs_t offset)
{
	// statistics
	if (DEBUG_STATS)
		m_stats.m_lfb_reads++;

	// linear frame buffer reads are inherently 16-bit; convert offset to an pixel index
	offset <<= 1;

	// convert offset into X/Y coordinates
	s32 x = offset & ((1 << m_lfb_stride) - 1);
	s32 y = offset >> m_lfb_stride;
	s32 scry = y;

	// effective Y is determined by the Y origin bit
	scry &= 0x3ff;
	auto const lfbmode = m_reg.lfb_mode();
	if (lfbmode.y_origin())
		scry = m_renderer->yorigin() - scry;

	// select the target buffer
	u16 *buffer = lfb_buffer_indirect(lfbmode.read_buffer_select());
	if (buffer == nullptr)
		return 0xffffffff;

	// advance pointers to the proper row
	buffer += scry * m_renderer->rowpixels() + x;
	if (buffer + 1 >= ram_end())
	{
		logerror("internal_lfb_r: Buffer offset out of bounds x=%i y=%i offset=%08X bufoffs=%08X\n", x, y, offset, u32(buffer - lfb_buffer_indirect(lfbmode.read_buffer_select())));
		return 0xffffffff;
	}

	// wait for any outstanding work to finish before reading
	m_renderer->wait("internal_lfb_r");

	// read and assemble two pixels
	u32 data = buffer[0] | (buffer[1] << 16);

	// word swapping
	if (lfbmode.word_swap_reads())
		data = (data << 16) | (data >> 16);

	// byte swizzling
	if (lfbmode.byte_swizzle_reads())
		data = swapendian_int32(data);

	if (LOG_LFB)
		logerror("VOODOO.LFB:read (%d,%d) = %08X\n", x, y, data);
	return data;
}


//-------------------------------------------------
//  internal_lfb_w - handle a write to the linear
//  frame buffer
//-------------------------------------------------

void voodoo_1_device::internal_lfb_w(offs_t offset, u32 data, u32 mem_mask)
{
	// statistics
	if (DEBUG_STATS)
		m_stats.m_lfb_writes++;

	// byte swizzling
	auto const lfbmode = m_reg.lfb_mode();
	if (lfbmode.byte_swizzle_writes())
	{
		data = swapendian_int32(data);
		mem_mask = swapendian_int32(mem_mask);
	}

	// word swapping
	if (lfbmode.word_swap_writes())
	{
		data = (data << 16) | (data >> 16);
		mem_mask = (mem_mask << 16) | (mem_mask >> 16);
	}

	// convert the incoming data
	rgb_t src_color[2];
	u16 src_depth[2];
	u32 mask = expand_lfb_data(lfbmode, data, src_color, src_depth);

	// if there are two pixels, then the offset is *2
	if ((mask & LFB_PIXEL1_MASK) != 0)
		offset <<= 1;

	// compute X,Y
	s32 x = offset & ((1 << m_lfb_stride) - 1);
	s32 y = (offset >> m_lfb_stride) & 0x3ff;

	// adjust the mask based on which half of the data is written
	if (!ACCESSING_BITS_0_15)
		mask &= ~(LFB_PIXEL0_MASK - LFB_DEPTH_PRESENT_MSW_0);
	if (!ACCESSING_BITS_16_31)
		mask &= ~(LFB_PIXEL1_MASK + LFB_DEPTH_PRESENT_MSW_0);

	// select the target buffers
	u16 *dest = draw_buffer_indirect(lfbmode.write_buffer_select());
	if (dest == nullptr)
		return;
	u16 *depth = aux_buffer();
	u16 *end = ram_end();

	// simple case: no pipeline
	auto const fbzmode = m_reg.fbz_mode();
	if (!lfbmode.enable_pixel_pipeline())
	{
		if (LOG_LFB)
			logerror("VOODOO.LFB:write raw mode %X (%d,%d) = %08X & %08X\n", lfbmode.write_format(), x, y, data, mem_mask);

		// determine the screen Y
		s32 scry = y;
		if (lfbmode.y_origin())
			scry = m_renderer->yorigin() - y;

		// advance pointers to the proper row
		dest += scry * m_renderer->rowpixels() + x;
		if (depth != nullptr)
			depth += scry * m_renderer->rowpixels() + x;

		// wait for any outstanding work to finish
		m_renderer->wait("internal_lfb_w(raw)");

		// loop over up to two pixels
		voodoo::dither_helper dither(scry, fbzmode);
		for (int pix = 0; mask != 0; pix++)
		{
			// make sure we care about this pixel
			if ((mask & LFB_PIXEL0_MASK) != 0)
			{
				// write to the RGB buffer
				rgb_t pixel = src_color[pix];
				if ((mask & LFB_RGB_PRESENT_0) != 0 && dest + pix < end)
					dest[pix] = dither.pixel(x, pixel.r(), pixel.g(), pixel.b());

				// make sure we have an aux buffer to write to
				if (depth != nullptr && depth + pix < end)
				{
					if (fbzmode.enable_alpha_planes())
					{
						// write to the alpha buffer
						if ((mask & LFB_ALPHA_PRESENT_0) != 0)
							depth[pix] = pixel.a();
					}
					else
					{
						// write to the depth buffer
						if ((mask & (LFB_DEPTH_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0)) != 0)
							depth[pix] = src_depth[pix];
					}
				}

				// track pixel writes to the frame buffer regardless of mask
				m_reg.add(voodoo_regs::reg_fbiPixelsOut, 1);
			}

			// advance our pointers
			x++;
			mask >>= 4;
		}
	}

	// tricky case: run the full pixel pipeline on the pixel
	else
	{
		if (LOG_LFB)
			logerror("VOODOO.LFB:write pipelined mode %X (%d,%d) = %08X & %08X\n", lfbmode.write_format(), x, y, data, mem_mask);

		// determine the screen Y
		s32 scry = y;
		if (fbzmode.y_origin())
			scry = m_renderer->yorigin() - y;

		// advance pointers to the proper row
		dest += scry * m_renderer->rowpixels();
		if (depth != nullptr)
			depth += scry * m_renderer->rowpixels();

		// make a dummy poly_extra_data structure with some cached values
		if (m_reg.fbz_mode().enable_stipple() && !m_reg.fbz_mode().stipple_pattern())
			logerror("Warning: rotated stipple pattern used in LFB write\n");

		// loop over up to two pixels
		thread_stats_block &threadstats = m_lfb_stats;
		rgbaint_t iterargb(0);
		for (int pix = 0; mask != 0; pix++)
		{
			// make sure we care about this pixel
			if ((mask & LFB_PIXEL0_MASK) != 0)
				m_renderer->pixel_pipeline(threadstats, dest, depth, x, y, src_color[pix], src_depth[pix]);

			// advance our pointers
			x++;
			mask >>= 4;
		}
	}
}


//-------------------------------------------------
//  expand_lfb_data - expand a 32-bit raw data
//  value into 1 or 2 expanded RGBA and depth
//  values
//-------------------------------------------------

u32 voodoo_1_device::expand_lfb_data(reg_lfb_mode const lfbmode, u32 data, rgb_t src_color[2], u16 src_depth[2])
{
	// extract default depth value from low bits of zaColor
	src_depth[0] = src_depth[1] = m_reg.za_color() & 0xffff;

	// if not otherwise specified, alpha defaults to the upper bits of zaColor
	u32 src_alpha = m_reg.za_color() >> 24;

	// extract color information from the data
	switch (16 * lfbmode.rgba_lanes() + lfbmode.write_format())
	{
		case 16*0 + 0:      // ARGB, format 0: 16-bit RGB 5-6-5
		case 16*2 + 0:      // RGBA, format 0: 16-bit RGB 5-6-5
			src_color[0] = rgbexpand<5,6,5>(data, 11,  5,  0).set_a(src_alpha);
			src_color[1] = rgbexpand<5,6,5>(data, 27, 21, 16).set_a(src_alpha);
			return LFB_RGB_PRESENT_0 | LFB_RGB_PRESENT_1;

		case 16*1 + 0:      // ABGR, format 0: 16-bit RGB 5-6-5
		case 16*3 + 0:      // BGRA, format 0: 16-bit RGB 5-6-5
			src_color[0] = rgbexpand<5,6,5>(data,  0,  5, 11).set_a(src_alpha);
			src_color[1] = rgbexpand<5,6,5>(data, 16, 21, 27).set_a(src_alpha);
			return LFB_RGB_PRESENT_0 | LFB_RGB_PRESENT_1;

		case 16*0 + 1:      // ARGB, format 1: 16-bit RGB x-5-5-5
			src_color[0] = rgbexpand<5,5,5>(data, 10,  5,  0).set_a(src_alpha);
			src_color[1] = rgbexpand<5,5,5>(data, 26, 21, 16).set_a(src_alpha);
			return LFB_RGB_PRESENT_0 | LFB_RGB_PRESENT_1;

		case 16*1 + 1:      // ABGR, format 1: 16-bit RGB x-5-5-5
			src_color[0] = rgbexpand<5,5,5>(data,  0,  5, 10).set_a(src_alpha);
			src_color[1] = rgbexpand<5,5,5>(data, 16, 21, 26).set_a(src_alpha);
			return LFB_RGB_PRESENT_0 | LFB_RGB_PRESENT_1;

		case 16*2 + 1:      // RGBA, format 1: 16-bit RGB x-5-5-5
			src_color[0] = rgbexpand<5,5,5>(data, 11,  6,  1).set_a(src_alpha);
			src_color[1] = rgbexpand<5,5,5>(data, 27, 22, 17).set_a(src_alpha);
			return LFB_RGB_PRESENT_0 | LFB_RGB_PRESENT_1;

		case 16*3 + 1:      // BGRA, format 1: 16-bit RGB x-5-5-5
			src_color[0] = rgbexpand<5,5,5>(data,  1,  6, 11).set_a(src_alpha);
			src_color[1] = rgbexpand<5,5,5>(data, 17, 22, 27).set_a(src_alpha);
			return LFB_RGB_PRESENT_0 | LFB_RGB_PRESENT_1;

		case 16*0 + 2:      // ARGB, format 2: 16-bit ARGB 1-5-5-5
			src_color[0] = argbexpand<1,5,5,5>(data, 15, 10,  5,  0);
			src_color[1] = argbexpand<1,5,5,5>(data, 31, 26, 21, 16);
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0 | LFB_RGB_PRESENT_1 | LFB_ALPHA_PRESENT_1;

		case 16*1 + 2:      // ABGR, format 2: 16-bit ARGB 1-5-5-5
			src_color[0] = argbexpand<1,5,5,5>(data, 15,  0,  5, 10);
			src_color[1] = argbexpand<1,5,5,5>(data, 31, 16, 21, 26);
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0 | LFB_RGB_PRESENT_1 | LFB_ALPHA_PRESENT_1;

		case 16*2 + 2:      // RGBA, format 2: 16-bit ARGB 1-5-5-5
			src_color[0] = argbexpand<1,5,5,5>(data,  0, 11,  6,  1);
			src_color[1] = argbexpand<1,5,5,5>(data, 16, 27, 22, 17);
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0 | LFB_RGB_PRESENT_1 | LFB_ALPHA_PRESENT_1;

		case 16*3 + 2:      // BGRA, format 2: 16-bit ARGB 1-5-5-5
			src_color[0] = argbexpand<1,5,5,5>(data,  0,  1,  6, 11);
			src_color[1] = argbexpand<1,5,5,5>(data, 16, 17, 22, 27);
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0 | LFB_RGB_PRESENT_1 | LFB_ALPHA_PRESENT_1;

		case 16*0 + 4:      // ARGB, format 4: 32-bit RGB x-8-8-8
			src_color[0] = rgbexpand<8,8,8>(data, 16,  8,  0).set_a(src_alpha);
			return LFB_RGB_PRESENT_0;

		case 16*1 + 4:      // ABGR, format 4: 32-bit RGB x-8-8-8
			src_color[0] = rgbexpand<8,8,8>(data,  0,  8, 16).set_a(src_alpha);
			return LFB_RGB_PRESENT_0;

		case 16*2 + 4:      // RGBA, format 4: 32-bit RGB x-8-8-8
			src_color[0] = rgbexpand<8,8,8>(data, 24, 16,  8).set_a(src_alpha);
			return LFB_RGB_PRESENT_0;

		case 16*3 + 4:      // BGRA, format 4: 32-bit RGB x-8-8-8
			src_color[0] = rgbexpand<8,8,8>(data,  8, 16, 24).set_a(src_alpha);
			return LFB_RGB_PRESENT_0;

		case 16*0 + 5:      // ARGB, format 5: 32-bit ARGB 8-8-8-8
			src_color[0] = argbexpand<8,8,8,8>(data, 24, 16,  8,  0);
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0;

		case 16*1 + 5:      // ABGR, format 5: 32-bit ARGB 8-8-8-8
			src_color[0] = argbexpand<8,8,8,8>(data, 24,  0,  8, 16);
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0;

		case 16*2 + 5:      // RGBA, format 5: 32-bit ARGB 8-8-8-8
			src_color[0] = argbexpand<8,8,8,8>(data,  0, 24, 16,  8);
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0;

		case 16*3 + 5:      // BGRA, format 5: 32-bit ARGB 8-8-8-8
			src_color[0] = argbexpand<8,8,8,8>(data,  0,  8, 16, 24);
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0;

		case 16*0 + 12:     // ARGB, format 12: 32-bit depth+RGB 5-6-5
		case 16*2 + 12:     // RGBA, format 12: 32-bit depth+RGB 5-6-5
			src_color[0] = rgbexpand<5,6,5>(data, 11,  5,  0).set_a(src_alpha);
			src_depth[0] = data >> 16;
			return LFB_RGB_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0;

		case 16*1 + 12:     // ABGR, format 12: 32-bit depth+RGB 5-6-5
		case 16*3 + 12:     // BGRA, format 12: 32-bit depth+RGB 5-6-5
			src_color[0] = rgbexpand<5,6,5>(data,  0,  5, 11).set_a(src_alpha);
			src_depth[0] = data >> 16;
			return LFB_RGB_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0;

		case 16*0 + 13:     // ARGB, format 13: 32-bit depth+RGB x-5-5-5
			src_color[0] = rgbexpand<5,5,5>(data, 10,  5,  0).set_a(src_alpha);
			src_depth[0] = data >> 16;
			return LFB_RGB_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0;

		case 16*1 + 13:     // ABGR, format 13: 32-bit depth+RGB x-5-5-5
			src_color[0] = rgbexpand<5,5,5>(data,  0,  5, 10).set_a(src_alpha);
			src_depth[0] = data >> 16;
			return LFB_RGB_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0;

		case 16*2 + 13:     // RGBA, format 13: 32-bit depth+RGB x-5-5-5
			src_color[0] = rgbexpand<5,5,5>(data, 11,  6,  1).set_a(src_alpha);
			src_depth[0] = data >> 16;
			return LFB_RGB_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0;

		case 16*3 + 13:     // BGRA, format 13: 32-bit depth+RGB x-5-5-5
			src_color[0] = rgbexpand<5,5,5>(data,  1,  6, 11).set_a(src_alpha);
			src_depth[0] = data >> 16;
			return LFB_RGB_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0;

		case 16*0 + 14:     // ARGB, format 14: 32-bit depth+ARGB 1-5-5-5
			src_color[0] = argbexpand<1,5,5,5>(data, 15, 10,  5,  0);
			src_depth[0] = data >> 16;
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0;

		case 16*1 + 14:     // ABGR, format 14: 32-bit depth+ARGB 1-5-5-5
			src_color[0] = argbexpand<1,5,5,5>(data, 15,  0,  5, 10);
			src_depth[0] = data >> 16;
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0;

		case 16*2 + 14:     // RGBA, format 14: 32-bit depth+ARGB 1-5-5-5
			src_color[0] = argbexpand<1,5,5,5>(data,  0, 11,  6,  1);
			src_depth[0] = data >> 16;
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0;

		case 16*3 + 14:     // BGRA, format 14: 32-bit depth+ARGB 1-5-5-5
			src_color[0] = argbexpand<1,5,5,5>(data,  0,  1,  6, 11);
			src_depth[0] = data >> 16;
			return LFB_RGB_PRESENT_0 | LFB_ALPHA_PRESENT_0 | LFB_DEPTH_PRESENT_MSW_0;

		case 16*0 + 15:     // ARGB, format 15: 16-bit depth
		case 16*1 + 15:     // ARGB, format 15: 16-bit depth
		case 16*2 + 15:     // ARGB, format 15: 16-bit depth
		case 16*3 + 15:     // ARGB, format 15: 16-bit depth
			src_depth[0] = data & 0xffff;
			src_depth[1] = data >> 16;
			return LFB_DEPTH_PRESENT_0 | LFB_DEPTH_PRESENT_1;

		default:            // reserved
			logerror("internal_lfb_w: Unknown format\n");
			return 0;
	}
}


//-------------------------------------------------
//  internal_texture_w - handle writes to texture
//  RAM
//-------------------------------------------------

void voodoo_1_device::internal_texture_w(offs_t offset, u32 data)
{
	// statistics
	if (DEBUG_STATS)
		m_stats.m_tex_writes++;

	// point to the right TMU
	int tmunum = BIT(offset, 19, 2);
	if (!BIT(m_chipmask, 1 + tmunum))
		return;

	// the seq_8_downld flag seems to always come from TMU #0
	bool seq_8_downld = m_tmu[0].regs().texture_mode().seq_8_downld();

	// pull out modes from the TMU and update state
	auto &regs = m_tmu[tmunum].regs();
	auto const texlod = regs.texture_lod();
	auto const texmode = regs.texture_mode();
	auto &texture = m_tmu[tmunum].prepare_texture(*m_renderer.get());

	// texture direct not handled (but never seen so far)
	if (texlod.tdirect_write())
		fatalerror("%s: Unsupported texture direct write", tag());

	// swizzle the data
	if (texlod.tdata_swizzle())
		data = swapendian_int32(data);
	if (texlod.tdata_swap())
		data = (data >> 16) | (data << 16);

	// determine destination pointer
	u32 bytes_per_texel = (texmode.format() < 8) ? 1 : 2;
	u32 lod = BIT(offset, 15, 4);
	u32 tt = BIT(offset, 7, 8);
	u32 ts = (offset << ((seq_8_downld && bytes_per_texel == 1) ? 2 : 1)) & 0xff;

	// validate parameters
	if (lod > 8)
		return;
	u8 *dest = texture.write_ptr(lod, ts, tt, bytes_per_texel);

	// wait for any outstanding work to finish
	m_renderer->wait("internal_texture_w");

	// write the four bytes in little-endian order
	if (bytes_per_texel == 1)
	{
		dest[BYTE4_XOR_LE(0)] = (data >> 0) & 0xff;
		dest[BYTE4_XOR_LE(1)] = (data >> 8) & 0xff;
		dest[BYTE4_XOR_LE(2)] = (data >> 16) & 0xff;
		dest[BYTE4_XOR_LE(3)] = (data >> 24) & 0xff;
	}
	else
	{
		u16 *dest16 = reinterpret_cast<u16 *>(dest);
		dest16[BYTE_XOR_LE(0)] = (data >> 0) & 0xffff;
		dest16[BYTE_XOR_LE(1)] = (data >> 16) & 0xffff;
	}
}


//-------------------------------------------------
//  reg_invalid_r - generic invalid register read
//-------------------------------------------------

u32 voodoo_1_device::reg_invalid_r(u32 chipmask, u32 regnum)
{
	// funkball does invalid reads of textureMode and will leave
	// improper bits set if this returns 0xffffffff
	logerror("%s: Unexpected read from register %s[%X.%02X]\n", machine().describe_context(), m_regtable[regnum].name(), chipmask, regnum);
	return 0x00000000;
}


//-------------------------------------------------
//  reg_passive_r - generic passive register read
//-------------------------------------------------

u32 voodoo_1_device::reg_passive_r(u32 chipmask, u32 regnum)
{
	return m_reg.read(regnum);
}


//-------------------------------------------------
//  reg_status_r - status register read
//-------------------------------------------------

u32 voodoo_1_device::reg_status_r(u32 chipmask, u32 regnum)
{
	u32 result = 0;

	// bits 5:0 are the PCI FIFO free space
	result |= std::min(m_pci_fifo.space() / 2, 0x3f) << 0;

	// bit 6 is the vertical retrace
	result |= m_vblank << 6;

	// bit 7 is FBI graphics engine busy
	// bit 8 is TREX busy
	// bit 9 is overall busy
	if (operation_pending())
		result |= (1 << 7) | (1 << 8) | (1 << 9);

	// bits 10-11 is displayed buffer
	result |= m_frontbuf << 10;

	// bits 12-27 is memory FIFO free space
	if (m_reg.fbi_init0().enable_memory_fifo() == 0)
		result |= 0xffff << 12;
	else
		result |= std::min(m_fbmem_fifo.space() / 2, 0xffff) << 12;

	// bits 30:28 are the number of pending swaps
	result |= std::min<s32>(m_swaps_pending, 7) << 28;

	// eat some cycles since people like polling here
	if (m_status_cycles != 0)
		m_cpu->eat_cycles(m_status_cycles);

	// bit 31 is PCI interrupt pending (not implemented)
	return result;
}


//-------------------------------------------------
//  reg_fbiinit2_r - fbiInit2 register read
//-------------------------------------------------

u32 voodoo_1_device::reg_fbiinit2_r(u32 chipmask, u32 regnum)
{
	// bit 2 of the initEnable register maps this to dacRead
	return m_init_enable.remap_init_to_dac() ? m_dac_read_result : m_reg.read(regnum);
}


//-------------------------------------------------
//  reg_vretrace_r - vRetrace register read
//-------------------------------------------------

u32 voodoo_1_device::reg_vretrace_r(u32 chipmask, u32 regnum)
{
	// sfrush needs this to be at least 1 extra cycle slower or else it won't boot
	// mace needs this to be at least 2 extra cycles
	m_cpu->eat_cycles(2);

	// return 0 if vblank is active
	return m_vblank ? 0 : screen().vpos();
}


//-------------------------------------------------
//  reg_stats_r - statistics register reads
//-------------------------------------------------

u32 voodoo_1_device::reg_stats_r(u32 chipmask, u32 regnum)
{
	update_statistics(true);
	return m_reg.read(regnum);
}


//-------------------------------------------------
//  reg_invalid_w - generic invalid register write
//-------------------------------------------------

u32 voodoo_1_device::reg_invalid_w(u32 chipmask, u32 regnum, u32 data)
{
	logerror("%s: Unexpected write to register %s[%X.%02X] = %08X\n", machine().describe_context(), m_regtable[regnum].name(), chipmask, regnum, data);
	return 0;
}


//-------------------------------------------------
//  reg_status_w - status register write (Voodoo 1)
//-------------------------------------------------

u32 voodoo_1_device::reg_unimplemented_w(u32 chipmask, u32 regnum, u32 data)
{
	logerror("%s: Unimplemented write to register %s[%X.%02X] = %08X\n", machine().describe_context(), m_regtable[regnum].name(), chipmask, regnum, data);
	return 0;
}

//-------------------------------------------------
//  reg_passive_w - generic passive register write
//-------------------------------------------------

u32 voodoo_1_device::reg_passive_w(u32 chipmask, u32 regnum, u32 data)
{
	if (BIT(chipmask, 0)) m_reg.write(regnum, data);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write(regnum, data);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write(regnum, data);
	return 0;
}


//-------------------------------------------------
//  reg_fpassive_4_w -- passive write with floating
//  point to x.4 fixed point conversion
//-------------------------------------------------

u32 voodoo_1_device::reg_fpassive_4_w(u32 chipmask, u32 regnum, u32 data)
{
	return reg_passive_w(chipmask, regnum - 0x80/4, float_to_int32(data, 4));
}


//-------------------------------------------------
//  reg_fpassive_12_w -- passive write with
//  floating point to x.12 fixed point conversion
//-------------------------------------------------

u32 voodoo_1_device::reg_fpassive_12_w(u32 chipmask, u32 regnum, u32 data)
{
	return reg_passive_w(chipmask, regnum - 0x80/4, float_to_int32(data, 12));
}


//-------------------------------------------------
//  reg_starts_w -- write to startS (14.18)
//  reg_starts_w -- write to startT (14.18)
//  reg_dsdx_w -- write to dSdX (14.18)
//  reg_dtdx_w -- write to dTdX (14.18)
//  reg_dsdy_w -- write to dSdY (14.18)
//  reg_dtdy_w -- write to dTdY (14.18)
//-------------------------------------------------

u32 voodoo_1_device::reg_starts_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = s64(s32(data)) << 14;
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_start_s(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_start_s(data64);
	return 0;
}
u32 voodoo_1_device::reg_startt_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = s64(s32(data)) << 14;
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_start_t(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_start_t(data64);
	return 0;
}
u32 voodoo_1_device::reg_dsdx_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = s64(s32(data)) << 14;
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_ds_dx(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_ds_dx(data64);
	return 0;
}
u32 voodoo_1_device::reg_dtdx_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = s64(s32(data)) << 14;
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_dt_dx(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_dt_dx(data64);
	return 0;
}
u32 voodoo_1_device::reg_dsdy_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = s64(s32(data)) << 14;
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_ds_dy(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_ds_dy(data64);
	return 0;
}
u32 voodoo_1_device::reg_dtdy_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = s64(s32(data)) << 14;
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_dt_dy(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_dt_dy(data64);
	return 0;
}


//-------------------------------------------------
//  reg_fstarts_w -- write to fstartS
//  reg_fstartt_w -- write to fstartT
//  reg_fdsdx_w -- write to fdSdX
//  reg_fdtdx_w -- write to fdTdX
//  reg_fdsdy_w -- write to fdSdY
//  reg_fdtdy_w -- write to fdTdY
//-------------------------------------------------

u32 voodoo_1_device::reg_fstarts_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = float_to_int64(data, 32);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_start_s(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_start_s(data64);
	return 0;
}
u32 voodoo_1_device::reg_fstartt_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = float_to_int64(data, 32);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_start_t(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_start_t(data64);
	return 0;
}
u32 voodoo_1_device::reg_fdsdx_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = float_to_int64(data, 32);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_ds_dx(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_ds_dx(data64);
	return 0;
}
u32 voodoo_1_device::reg_fdtdx_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = float_to_int64(data, 32);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_dt_dx(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_dt_dx(data64);
	return 0;
}
u32 voodoo_1_device::reg_fdsdy_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = float_to_int64(data, 32);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_ds_dy(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_ds_dy(data64);
	return 0;
}
u32 voodoo_1_device::reg_fdtdy_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = float_to_int64(data, 32);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_dt_dy(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_dt_dy(data64);
	return 0;
}


//-------------------------------------------------
//  reg_startw_w -- write to startW (2.30 -> 16.32)
//  reg_dwdx_w -- write to dWdX (2.30 -> 16.32)
//  reg_dwdy_w -- write to dWdY (2.30 -> 16.32)
//-------------------------------------------------

u32 voodoo_1_device::reg_startw_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = s64(s32(data)) << 2;
	if (BIT(chipmask, 0))          m_reg.write_start_w(data64);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_start_w(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_start_w(data64);
	return 0;
}
u32 voodoo_1_device::reg_dwdx_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = s64(s32(data)) << 2;
	if (BIT(chipmask, 0))          m_reg.write_dw_dx(data64);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_dw_dx(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_dw_dx(data64);
	return 0;
}
u32 voodoo_1_device::reg_dwdy_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = s64(s32(data)) << 2;
	if (BIT(chipmask, 0))          m_reg.write_dw_dy(data64);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_dw_dy(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_dw_dy(data64);
	return 0;
}


//-------------------------------------------------
//  reg_fstartw_w -- write to fstartW
//  reg_fdwdx_w -- write to fdWdX
//  reg_fdwdy_w -- write to fdWdY
//-------------------------------------------------

u32 voodoo_1_device::reg_fstartw_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = float_to_int64(data, 32);
	if (BIT(chipmask, 0))          m_reg.write_start_w(data64);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_start_w(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_start_w(data64);
	return 0;
}
u32 voodoo_1_device::reg_fdwdx_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = float_to_int64(data, 32);
	if (BIT(chipmask, 0))          m_reg.write_dw_dx(data64);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_dw_dx(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_dw_dx(data64);
	return 0;
}
u32 voodoo_1_device::reg_fdwdy_w(u32 chipmask, u32 regnum, u32 data)
{
	s64 data64 = float_to_int64(data, 32);
	if (BIT(chipmask, 0))          m_reg.write_dw_dy(data64);
	if (BIT(chipmask, 1)) m_tmu[0].regs().write_dw_dy(data64);
	if (BIT(chipmask, 2)) m_tmu[1].regs().write_dw_dy(data64);
	return 0;
}


//-------------------------------------------------
//  reg_triangle_w -- write to triangleCMD/
//  ftriangleCMD
//-------------------------------------------------

u32 voodoo_1_device::reg_triangle_w(u32 chipmask, u32 regnum, u32 data)
{
	return triangle();
}


//-------------------------------------------------
//  reg_nop_w -- write to nopCMD
//-------------------------------------------------

u32 voodoo_1_device::reg_nop_w(u32 chipmask, u32 regnum, u32 data)
{
	// NOP should synchronize the pipeline; in theory we can mostly get away without
	// it, but gtfore06 shows flicker on some golfers if we don't respect it; some
	// games (notably gradius4) take a noticeable hit when this is present, so it
	// may be worth adding an option to not block here
	m_renderer->wait("reg_nop_w");

	if (BIT(data, 0))
		reset_counters();
	if (BIT(data, 1))
		m_reg.write(voodoo_regs::reg_fbiTrianglesOut, 0);
	return 0;
}


//-------------------------------------------------
//  reg_fastfill_w -- write to fastfillCMD
//-------------------------------------------------

u32 voodoo_1_device::reg_fastfill_w(u32 chipmask, u32 regnum, u32 data)
{
	auto &poly = m_renderer->alloc_poly();

	// determine the draw buffer (Banshee and later are hard-coded to the back buffer)
	poly.destbase = draw_buffer_indirect(m_reg.fbz_mode().draw_buffer());
	if (poly.destbase == nullptr)
		return 0;
	poly.depthbase = aux_buffer();
	poly.clipleft = m_reg.clip_left();
	poly.clipright = m_reg.clip_right();
	poly.cliptop = m_reg.clip_top();
	poly.clipbottom = m_reg.clip_bottom();
	poly.color1 = m_reg.color1().argb();
	poly.zacolor = m_reg.za_color();

	// 2 pixels per clock
	return m_renderer->enqueue_fastfill(poly) / 2;
}


//-------------------------------------------------
//  reg_swapbuffer_w -- write to swapbufferCMD
//-------------------------------------------------

u32 voodoo_1_device::reg_swapbuffer_w(u32 chipmask, u32 regnum, u32 data)
{
	// the don't swap value is Voodoo 2-only, masked off by the register engine
	m_vblank_swap_pending = true;
	m_vblank_swap = BIT(data, 1, 8);
	m_vblank_dont_swap = BIT(data, 9);

	// if we're not syncing to the retrace, process the command immediately
	if (!BIT(data, 0))
	{
		swap_buffers();
		return 0;
	}

	// determine how many cycles to wait; we deliberately overshoot here because
	// the final count gets updated on the VBLANK
	return (m_vblank_swap + 1) * clock() / 10;
}


//-------------------------------------------------
//  reg_fogtable_w -- write to fogTable
//-------------------------------------------------

u32 voodoo_1_device::reg_fogtable_w(u32 chipmask, u32 regnum, u32 data)
{
	if (BIT(chipmask, 0)) m_renderer->write_fog(2 * (regnum - voodoo_regs::reg_fogTable), data);
	return 0;
}


//-------------------------------------------------
//  reg_fbiinit_w -- write to an fbiinit register
//-------------------------------------------------

u32 voodoo_1_device::reg_fbiinit_w(u32 chipmask, u32 regnum, u32 data)
{
	if (BIT(chipmask, 0) && m_init_enable.enable_hw_init())
	{
		m_renderer->wait("reg_fbiinit_w");
		m_reg.write(regnum, data);

		// handle resets written to fbiInit0
		if (regnum == voodoo_regs::reg_fbiInit0 && m_reg.fbi_init0().graphics_reset())
			soft_reset();
		if (regnum == voodoo_regs::reg_fbiInit0 && m_reg.fbi_init0().fifo_reset())
			m_pci_fifo.reset();

		// compute FIFO layout when fbiInit0 or fbiInit4 change
		if (regnum == voodoo_regs::reg_fbiInit0 || regnum == voodoo_regs::reg_fbiInit4)
			recompute_fbmem_fifo();

		// recompute video memory when fbiInit1 or fbiInit2 change
		if (regnum == voodoo_regs::reg_fbiInit1 || regnum == voodoo_regs::reg_fbiInit2)
			recompute_video_memory();

		// update Y origina when fbiInit3 changes
		if (regnum == voodoo_regs::reg_fbiInit3)
			m_renderer->set_yorigin(m_reg.fbi_init3().yorigin_subtract());
	}
	return 0;
}


//-------------------------------------------------
//  reg_video_w -- write to a video configuration
//  register; synchronize then recompute everything
//-------------------------------------------------

u32 voodoo_1_device::reg_video_w(u32 chipmask, u32 regnum, u32 data)
{
	if (BIT(chipmask, 0))
	{
		m_renderer->wait("reg_video_w");
		m_reg.write(regnum, data);

		auto const hsync = m_reg.hsync<true>();
		auto const vsync = m_reg.vsync<true>();
		auto const back_porch = m_reg.back_porch<true>();
		auto const video_dimensions = m_reg.video_dimensions<true>();
		if (hsync.raw() != 0 && vsync.raw() != 0 && video_dimensions.raw() != 0 && back_porch.raw() != 0)
		{
			recompute_video_timing(
					hsync.hsync_on(), hsync.hsync_off(),
					video_dimensions.xwidth(), back_porch.horizontal() + 2,
					vsync.vsync_on(), vsync.vsync_off(),
					video_dimensions.yheight(), back_porch.vertical());
		}
	}
	return 0;
}


//-------------------------------------------------
//  reg_clut_w -- write to clutData; mark dirty if
//  changed
//-------------------------------------------------

u32 voodoo_1_device::reg_clut_w(u32 chipmask, u32 regnum, u32 data)
{
	if (BIT(chipmask, 0))
	{
		if (m_reg.fbi_init1().video_timing_reset() == 0)
		{
			int index = BIT(data, 24, 8);
			if (index <= 32 && m_clut[index] != data)
			{
				screen().update_partial(screen().vpos());
				m_clut[index] = data;
				m_clut_dirty = true;
			}
		}
		else
			logerror("clutData ignored because video timing reset = 1\n");
	}
	return 0;
}


//-------------------------------------------------
//  reg_dac_w -- write to dacData
//-------------------------------------------------

u32 voodoo_1_device::reg_dac_w(u32 chipmask, u32 regnum, u32 data)
{
	if (BIT(chipmask, 0))
	{
		// upper 2 address bits are only on Voodoo2+ but are masked by the
		// register entry for Voodoo 1 so safe to just use them as presented
		u32 regnum = BIT(data, 8, 3) + 8 * BIT(data, 12, 2);
		if (!BIT(data, 11))
			m_dac_reg[regnum] = BIT(data, 0, 8);
		else
		{
			// this is just to make startup happy
			m_dac_read_result = m_dac_reg[regnum];
			switch (m_dac_reg[7])
			{
				case 0x01:  m_dac_read_result = 0x55; break;
				case 0x07:  m_dac_read_result = 0x71; break;
				case 0x0b:  m_dac_read_result = 0x79; break;
			}
		}
	}
	return 0;
}


//-------------------------------------------------
//  reg_texture_w -- passive write to a TMU; mark
//  dirty if changed
//-------------------------------------------------

u32 voodoo_1_device::reg_texture_w(u32 chipmask, u32 regnum, u32 data)
{
	if (BIT(chipmask, 1))
	{
		if (data != m_tmu[0].regs().read(regnum))
		{
			m_tmu[0].regs().write(regnum, data);
			m_tmu[0].mark_dirty();
		}
	}
	if (BIT(chipmask, 2))
	{
		if (data != m_tmu[1].regs().read(regnum))
		{
			m_tmu[1].regs().write(regnum, data);
			m_tmu[1].mark_dirty();
		}
	}
	return 0;
}


//-------------------------------------------------
//  reg_palette_w -- passive write to a palette or
//  NCC table; mark dirty if changed
//-------------------------------------------------

u32 voodoo_1_device::reg_palette_w(u32 chipmask, u32 regnum, u32 data)
{
	if (BIT(chipmask, 1)) m_tmu[0].ncc_w(regnum, data);
	if (BIT(chipmask, 2)) m_tmu[1].ncc_w(regnum, data);
	return 0;
}


//-------------------------------------------------
//  adjust_vblank_start_timer -- adjust the VBLANK
//  start timer based on latest information
//-------------------------------------------------

void voodoo_1_device::adjust_vblank_start_timer()
{
	attotime time_until_blank = screen().time_until_pos(m_vsyncstart);
	if (LOG_VBLANK_SWAP)
		logerror("adjust_vblank_start_timer: period: %s\n", time_until_blank.as_string());

	// if zero, adjust to next frame, otherwise we may get stuck in an infinite loop
	if (time_until_blank == attotime::zero)
		time_until_blank = screen().frame_period();
	m_vsync_start_timer->adjust(time_until_blank);
}


//-------------------------------------------------
//  vblank_start -- timer callback for the start
//  of VBLANK
//-------------------------------------------------

void voodoo_1_device::vblank_start(void *ptr, s32 param)
{
	if (LOG_VBLANK_SWAP)
		logerror("--- vblank start\n");

	// flush the pipes
	if (operation_pending())
	{
		if (LOG_VBLANK_SWAP)
			logerror("---- vblank flush begin\n");
		flush_fifos(machine().time());
		if (LOG_VBLANK_SWAP)
			logerror("---- vblank flush end\n");
	}

	// increment the count
	m_vblank_count = std::min(m_vblank_count + 1, 250);

	// logging
	if (LOG_VBLANK_SWAP)
		logerror("---- vblank count = %u swap = %u pending = %u", m_vblank_count, m_vblank_swap, m_vblank_swap_pending);
	if (LOG_VBLANK_SWAP && m_vblank_swap_pending)
		logerror(" (target=%d)", m_vblank_swap);
	if (LOG_VBLANK_SWAP)
		logerror("\n");

	// if we're past the swap count, do the swap
	if (m_vblank_swap_pending && m_vblank_count >= m_vblank_swap)
		swap_buffers();

	// set a timer for the next off state
	m_vsync_stop_timer->adjust(screen().time_until_pos(m_vsyncstop));

	// set internal state and call the client
	m_vblank = true;

	// notify external VBLANK handler on all models
	if (!m_vblank_cb.isnull())
		m_vblank_cb(true);
}


//-------------------------------------------------
//  vblank_stop -- timer callback for the end of
//  VBLANK
//-------------------------------------------------

void voodoo_1_device::vblank_stop(void *ptr, s32 param)
{
	if (LOG_VBLANK_SWAP)
		logerror("--- vblank end\n");

	// set internal state and call the client
	m_vblank = false;

	// notify external VBLANK handler on all models
	if (!m_vblank_cb.isnull())
		m_vblank_cb(false);

	// go to the end of the next frame
	adjust_vblank_start_timer();
}


//-------------------------------------------------
//  swap_buffers -- perform a buffer swap; in most
//  cases this comes at VBLANK time
//-------------------------------------------------

void voodoo_1_device::swap_buffers()
{
	if (LOG_VBLANK_SWAP)
		logerror("--- swap_buffers @ %d\n", screen().vpos());

	// force a partial update
	m_renderer->wait("swap_buffers");
	screen().update_partial(screen().vpos());
	m_video_changed = true;

	// keep a history of swap intervals
	m_reg.update_swap_history(std::min<u8>(m_vblank_count, 15));

	// rotate the buffers; implementation differs between models
	rotate_buffers();

	// decrement the pending count and reset our state
	if (m_swaps_pending != 0)
		m_swaps_pending--;
	m_vblank_count = 0;
	m_vblank_swap_pending = false;

	// reset the last_op_time to now and start processing the next command
	if (operation_pending())
	{
		if (LOG_VBLANK_SWAP)
			logerror("---- swap_buffers flush begin\n");
		flush_fifos(m_operation_end = machine().time());
		if (LOG_VBLANK_SWAP)
			logerror("---- swap_buffers flush end\n");
	}

	// we may be able to unstall now
	if (m_stall_state != NOT_STALLED)
		check_stalled_cpu(machine().time());

	// periodically log rasterizer info
	m_stats.m_swaps++;
	if (m_stats.m_swaps % 1000 == 0)
		m_renderer->dump_rasterizer_stats();

	// update the statistics (debug)
	if (DEBUG_STATS)
	{
		if (m_stats.displayed())
		{
			update_statistics(true);
			m_stats.update_string(screen().visible_area(), m_reg.swap_history());
		}
		m_stats.reset();
	}
}


//-------------------------------------------------
//  rotate_buffers -- rotate the buffers according
//  to the current buffer config; this is split
//  out so later devices can override
//-------------------------------------------------

void voodoo_1_device::rotate_buffers()
{
	if (!m_vblank_dont_swap)
	{
		u32 buffers = (m_rgboffs[2] == ~0) ? 2 : 3;
		m_frontbuf = (m_frontbuf + 1) % buffers;
		m_backbuf = (m_frontbuf + 1) % buffers;
	}
}


//-------------------------------------------------
//  update_common -- shared update function
//-------------------------------------------------

int voodoo_1_device::update_common(bitmap_rgb32 &bitmap, const rectangle &cliprect, rgb_t const *pens)
{
	// flush the pipes
	if (operation_pending())
	{
		if (LOG_VBLANK_SWAP)
			logerror("---- update flush begin\n");
		flush_fifos(machine().time());
		if (LOG_VBLANK_SWAP)
			logerror("---- update flush end\n");
	}

	// reset the video changed flag
	bool changed = m_video_changed;
	m_video_changed = false;

	// select the buffer to draw
	int drawbuf = m_frontbuf;
	if (DEBUG_BACKBUF && machine().input().code_pressed(KEYCODE_L))
		drawbuf = m_backbuf;

	// copy from the current front buffer
	u32 rowpixels = m_renderer->rowpixels();
	u16 *buffer_base = draw_buffer(drawbuf);
	if (LOG_VBLANK_SWAP) logerror("--- update_common %d-%d @ %d from %08X\n", cliprect.min_y, cliprect.max_y, screen().vpos(), u32((u8 *)buffer_base - m_fbram));
	for (s32 y = cliprect.min_y; y <= cliprect.max_y; y++)
	{
		if (y < m_yoffs)
			continue;
		u16 const *const src = buffer_base + (y - m_yoffs) * rowpixels - m_xoffs;
		u32 *dst = &bitmap.pix(y);
		for (s32 x = cliprect.min_x; x <= cliprect.max_x; x++)
			dst[x] = pens[src[x]];
	}

	// update stats display
	if (DEBUG_STATS && m_stats.update_display_state(machine().input().code_pressed(KEYCODE_BACKSLASH)))
		popmessage(m_stats.string(), 0, 0);

	// overwrite with the depth buffer if debugging and the ENTER key is pressed
	if (DEBUG_DEPTH && machine().input().code_pressed(KEYCODE_ENTER))
		for (s32 y = cliprect.min_y; y <= cliprect.max_y; y++)
		{
			u16 const *const src = aux_buffer() + (y - m_yoffs) * rowpixels - m_xoffs;
			u32 *const dst = &bitmap.pix(y);
			for (s32 x = cliprect.min_x; x <= cliprect.max_x; x++)
				dst[x] = ((src[x] << 8) & 0xff0000) | ((src[x] >> 0) & 0xff00) | ((src[x] >> 8) & 0xff);
		}

	return changed;
}


//-------------------------------------------------
//  recompute_video_timing -- given hsync and
//  vsync parameter, find the best match for known
//  monitor types and select the best fit
//-------------------------------------------------

void voodoo_1_device::recompute_video_timing(u32 hsyncon, u32 hsyncoff, u32 hvis, u32 hbp, u32 vsyncon, u32 vsyncoff, u32 vvis, u32 vbp)
{
	u32 htotal = hsyncoff + 1 + hsyncon + 1;
	u32 vtotal = vsyncoff + vsyncon;

	// create a new visarea from the backporch and visible values
	rectangle visarea(hbp, hbp + std::max(s32(hvis) - 1, 0), vbp, vbp + std::max(s32(vvis) - 1, 0));

	// keep within bounds
	visarea.max_x = std::min<s32>(visarea.max_x, htotal - 1);
	visarea.max_y = std::min<s32>(visarea.max_y, vtotal - 1);

	// compute the new period for standard res, medium res, and VGA res
	attoseconds_t stdperiod = HZ_TO_ATTOSECONDS(15750) * vtotal;
	attoseconds_t medperiod = HZ_TO_ATTOSECONDS(25000) * vtotal;
	attoseconds_t vgaperiod = HZ_TO_ATTOSECONDS(31500) * vtotal;

	// compute a diff against the current refresh period
	attoseconds_t refresh = screen().frame_period().attoseconds();
	attoseconds_t stddiff = std::abs(stdperiod - refresh);
	attoseconds_t meddiff = std::abs(medperiod - refresh);
	attoseconds_t vgadiff = std::abs(vgaperiod - refresh);

	logerror("hSync=%d-%d, bp=%d, vis=%d  vSync=%d-%d, bp=%d, vis=%d\n", hsyncon, hsyncoff, hbp, hvis, vsyncon, vsyncoff, vbp, vvis);
	logerror("Horiz: %d-%d (%d total)  Vert: %d-%d (%d total) -- ", visarea.min_x, visarea.max_x, htotal, visarea.min_y, visarea.max_y, vtotal);

	// configure the screen based on which one matches the closest
	if (stddiff < meddiff && stddiff < vgadiff)
	{
		screen().configure(htotal, vtotal, visarea, stdperiod);
		logerror("Standard resolution, %f Hz\n", ATTOSECONDS_TO_HZ(stdperiod));
	}
	else if (meddiff < vgadiff)
	{
		screen().configure(htotal, vtotal, visarea, medperiod);
		logerror("Medium resolution, %f Hz\n", ATTOSECONDS_TO_HZ(medperiod));
	}
	else
	{
		screen().configure(htotal, vtotal, visarea, vgaperiod);
		logerror("VGA resolution, %f Hz\n", ATTOSECONDS_TO_HZ(vgaperiod));
	}

	// configure the new framebuffer info
	m_width = hvis;
	m_height = vvis;
	m_xoffs = hbp;
	m_yoffs = vbp;
	m_vsyncstart = vsyncoff;
	m_vsyncstop = vsyncon;
	logerror("yoffs: %d vsyncstart: %d vsyncstop: %d\n", vbp, m_vsyncstart, m_vsyncstop);

	adjust_vblank_start_timer();
}


//-------------------------------------------------
//  recompute_video_memory -- compute the layout
//  of video memory
//-------------------------------------------------

void voodoo_1_device::recompute_video_memory()
{
	// configuration is either double-buffered (0) or triple-buffered (1)
	u32 config = m_reg.fbi_init2().enable_triple_buf();

	// 4-bit tile count; tiles are 64x16
	u32 xtiles = m_reg.fbi_init1().x_video_tiles();
	recompute_video_memory_common(config, xtiles * 64);
}


//-------------------------------------------------
//  recompute_video_memory_common -- core logic
//  for video memory layout based on 2-bit config
//  and the computed rowpixels
//-------------------------------------------------

void voodoo_1_device::recompute_video_memory_common(u32 config, u32 rowpixels)
{
	// remember the front buffer configuration to check for changes
	u16 *starting_front = front_buffer();
	u32 starting_rowpix = m_renderer->rowpixels();

	// first RGB buffer always starts at 0
	m_rgboffs[0] = 0;

	// second RGB buffer starts immediately afterwards
	u32 const buffer_pages = m_reg.fbi_init2().video_buffer_offset();
	m_rgboffs[1] = buffer_pages * 0x1000;

	// remaining buffers are based on the config
	switch (config)
	{
		case 3: // reserved
//          logerror("VOODOO.ERROR:Unexpected memory configuration in recompute_video_memory!\n");
			[[fallthrough]];
		case 0: // 2 color buffers, 1 aux buffer
			m_rgboffs[2] = ~0;
			m_auxoffs = 2 * buffer_pages * 0x1000;
			break;

		case 1: // 3 color buffers, 0 aux buffers
			m_rgboffs[2] = 2 * buffer_pages * 0x1000;
			m_auxoffs = ~0;
			break;

		case 2: // 3 color buffers, 1 aux buffers
			m_rgboffs[2] = 2 * buffer_pages * 0x1000;
			m_auxoffs = 3 * buffer_pages * 0x1000;
			break;
	}

	// clamp the RGB buffers to video memory
	for (int buf = 0; buf < 3; buf++)
		if (m_rgboffs[buf] != ~0 && m_rgboffs[buf] > m_fbmask)
			m_rgboffs[buf] = m_fbmask;

	// clamp the aux buffer to video memory
	if (m_auxoffs != ~0 && m_auxoffs > m_fbmask)
		m_auxoffs = m_fbmask;

	// reset our front/back buffers if they are out of range
	if (m_rgboffs[2] == ~0)
	{
		if (m_frontbuf == 2)
			m_frontbuf = 0;
		if (m_backbuf == 2)
			m_backbuf = 0;
	}

	// mark video changed if the front buffer configuration is different
	if (front_buffer() != starting_front || rowpixels != starting_rowpix)
		m_video_changed = true;
	m_renderer->set_rowpixels(rowpixels);
}


//-------------------------------------------------
//  triangle - execute the 'triangle' command
//-------------------------------------------------

s32 voodoo_1_device::triangle()
{
	g_profiler.start(PROFILER_USER2);

	// allocate polygon information now
	auto &poly = m_renderer->alloc_poly();

	// determine the draw buffer
	poly.destbase = draw_buffer_indirect(m_reg.fbz_mode().draw_buffer());
	if (poly.destbase == nullptr)
		return TRIANGLE_SETUP_CLOCKS;
	poly.depthbase = aux_buffer();
	poly.clipleft = m_reg.clip_left();
	poly.clipright = m_reg.clip_right();
	poly.cliptop = m_reg.clip_top();
	poly.clipbottom = m_reg.clip_bottom();

	// fill in triangle parameters
	poly.ax = m_reg.ax();
	poly.ay = m_reg.ay();
	poly.startr = m_reg.start_r();
	poly.startg = m_reg.start_g();
	poly.startb = m_reg.start_b();
	poly.starta = m_reg.start_a();
	poly.startz = m_reg.start_z();
	poly.startw = m_reg.start_w();
	poly.drdx = m_reg.dr_dx();
	poly.dgdx = m_reg.dg_dx();
	poly.dbdx = m_reg.db_dx();
	poly.dadx = m_reg.da_dx();
	poly.dzdx = m_reg.dz_dx();
	poly.dwdx = m_reg.dw_dx();
	poly.drdy = m_reg.dr_dy();
	poly.dgdy = m_reg.dg_dy();
	poly.dbdy = m_reg.db_dy();
	poly.dady = m_reg.da_dy();
	poly.dzdy = m_reg.dz_dy();
	poly.dwdy = m_reg.dw_dy();

	// perform subpixel adjustments -- note that the documentation indicates this
	// is done in the internal registers, so do it there
	if (m_reg.fbz_colorpath().cca_subpixel_adjust())
	{
		s32 dx = 8 - (poly.ax & 15);
		s32 dy = 8 - (poly.ay & 15);

		// adjust iterated R,G,B,A and W/Z
		m_reg.write(voodoo_regs::reg_startR, poly.startr += (dy * poly.drdy + dx * poly.drdx) >> 4);
		m_reg.write(voodoo_regs::reg_startG, poly.startg += (dy * poly.dgdy + dx * poly.dgdx) >> 4);
		m_reg.write(voodoo_regs::reg_startB, poly.startb += (dy * poly.dbdy + dx * poly.dbdx) >> 4);
		m_reg.write(voodoo_regs::reg_startA, poly.starta += (dy * poly.dady + dx * poly.dadx) >> 4);
		m_reg.write(voodoo_regs::reg_startZ, poly.startz += (dy * poly.dzdy + dx * poly.dzdx) >> 4);
		m_reg.write_start_w(poly.startw += (dy * poly.dwdy + dx * poly.dwdx) >> 4);

		// adjust iterated W/S/T for TMU 0
		auto &tmu0regs = m_tmu[0].regs();
		tmu0regs.write_start_w(tmu0regs.start_w() + ((dy * tmu0regs.dw_dy() + dx * tmu0regs.dw_dx()) >> 4));
		tmu0regs.write_start_s(tmu0regs.start_s() + ((dy * tmu0regs.ds_dy() + dx * tmu0regs.ds_dx()) >> 4));
		tmu0regs.write_start_t(tmu0regs.start_t() + ((dy * tmu0regs.dt_dy() + dx * tmu0regs.dt_dx()) >> 4));

		// adjust iterated W/S/T for TMU 1
		if (BIT(m_chipmask, 2))
		{
			auto &tmu1regs = m_tmu[1].regs();
			tmu1regs.write_start_w(tmu1regs.start_w() + ((dy * tmu1regs.dw_dy() + dx * tmu1regs.dw_dx()) >> 4));
			tmu1regs.write_start_s(tmu1regs.start_s() + ((dy * tmu1regs.ds_dy() + dx * tmu1regs.ds_dx()) >> 4));
			tmu1regs.write_start_t(tmu1regs.start_t() + ((dy * tmu1regs.dt_dy() + dx * tmu1regs.dt_dx()) >> 4));
		}
	}

	// fill in texture 0 parameters
	poly.tex0 = nullptr;
	if (poly.raster.texmode0().raw() != 0xffffffff)
	{
		auto &tmu0regs = m_tmu[0].regs();
		poly.starts0 = tmu0regs.start_s();
		poly.startt0 = tmu0regs.start_t();
		poly.startw0 = tmu0regs.start_w();
		poly.ds0dx = tmu0regs.ds_dx();
		poly.dt0dx = tmu0regs.dt_dx();
		poly.dw0dx = tmu0regs.dw_dx();
		poly.ds0dy = tmu0regs.ds_dy();
		poly.dt0dy = tmu0regs.dt_dy();
		poly.dw0dy = tmu0regs.dw_dy();
		poly.tex0 = &m_tmu[0].prepare_texture(*m_renderer.get());
		if (DEBUG_STATS)
			m_stats.m_texture_mode[tmu0regs.texture_mode().format()]++;
	}

	// fill in texture 1 parameters
	poly.tex1 = nullptr;
	if (poly.raster.texmode1().raw() != 0xffffffff)
	{
		auto &tmu1regs = m_tmu[1].regs();
		poly.starts1 = tmu1regs.start_s();
		poly.startt1 = tmu1regs.start_t();
		poly.startw1 = tmu1regs.start_w();
		poly.ds1dx = tmu1regs.ds_dx();
		poly.dt1dx = tmu1regs.dt_dx();
		poly.dw1dx = tmu1regs.dw_dx();
		poly.ds1dy = tmu1regs.ds_dy();
		poly.dt1dy = tmu1regs.dt_dy();
		poly.dw1dy = tmu1regs.dw_dy();
		poly.tex1 = &m_tmu[1].prepare_texture(*m_renderer.get());
		if (DEBUG_STATS)
			m_stats.m_texture_mode[tmu1regs.texture_mode().format()]++;
	}

	// fill in color parameters
	poly.color0 = m_reg.color0().argb();
	poly.color1 = m_reg.color1().argb();
	poly.chromakey = m_reg.chroma_key().argb();
	poly.fogcolor = m_reg.fog_color().argb();
	poly.zacolor = m_reg.za_color();
	poly.stipple = m_reg.stipple();
	poly.alpharef = m_reg.alpha_mode().alpharef();

	// fill in the vertex data
	voodoo_renderer::vertex_t vert[3];
	vert[0].x = float(m_reg.ax()) * (1.0f / 16.0f);
	vert[0].y = float(m_reg.ay()) * (1.0f / 16.0f);
	vert[1].x = float(m_reg.bx()) * (1.0f / 16.0f);
	vert[1].y = float(m_reg.by()) * (1.0f / 16.0f);
	vert[2].x = float(m_reg.cx()) * (1.0f / 16.0f);
	vert[2].y = float(m_reg.cy()) * (1.0f / 16.0f);

	// enqueue a triangle
	s32 pixels = m_renderer->enqueue_triangle(poly, vert);

	// update stats
	m_reg.add(voodoo_regs::reg_fbiTrianglesOut, 1);
	if (DEBUG_STATS)
		m_stats.m_triangles++;

	g_profiler.stop();

	if (LOG_REGISTERS)
		logerror("cycles = %d\n", TRIANGLE_SETUP_CLOCKS + pixels);

	// 1 pixel per clock, plus some setup time
	return TRIANGLE_SETUP_CLOCKS + pixels;
}


//-------------------------------------------------
//  accumulate_statistics - add the statistics
//  from the given thread block to the shared
//  statistics
//-------------------------------------------------

void voodoo_1_device::accumulate_statistics(thread_stats_block const &block)
{
	// update live voodoo statistics
	m_reg.add(voodoo_regs::reg_fbiPixelsIn, block.pixels_in);
	m_reg.add(voodoo_regs::reg_fbiPixelsOut, block.pixels_out);
	m_reg.add(voodoo_regs::reg_fbiChromaFail, block.chroma_fail);
	m_reg.add(voodoo_regs::reg_fbiZfuncFail, block.zfunc_fail);
	m_reg.add(voodoo_regs::reg_fbiAfuncFail, block.afunc_fail);

	// update emulation statistics
	if (DEBUG_STATS)
		m_stats.add_emulation_stats(block);
}


//-------------------------------------------------
//  update_statistics - gather statistics from
//  all threads and then reset the thread-local
//  information
//-------------------------------------------------

void voodoo_1_device::update_statistics(bool accumulate)
{
	// accumulate/reset statistics from all units
	for (auto &stats : m_renderer->thread_stats())
	{
		if (accumulate)
			accumulate_statistics(stats);
		stats.reset();
	}

	// accumulate/reset statistics from the LFB
	if (accumulate)
		accumulate_statistics(m_lfb_stats);
	m_lfb_stats.reset();
}


//-------------------------------------------------
//  reset_counters - reset the exposed statistics
//  counters to 0
//-------------------------------------------------

void voodoo_1_device::reset_counters()
{
	update_statistics(false);
	m_reg.write(voodoo_regs::reg_fbiPixelsIn, 0);
	m_reg.write(voodoo_regs::reg_fbiChromaFail, 0);
	m_reg.write(voodoo_regs::reg_fbiZfuncFail, 0);
	m_reg.write(voodoo_regs::reg_fbiAfuncFail, 0);
	m_reg.write(voodoo_regs::reg_fbiPixelsOut, 0);
}


//-------------------------------------------------
//  check_stalled_cpu - determine if it's time to
//  un-stall a CPU given pending operations
//-------------------------------------------------

void voodoo_1_device::check_stalled_cpu(attotime current_time)
{
	bool resume = false;

	// flush anything we can
	if (operation_pending())
		flush_fifos(current_time);

	// if we're just stalled until the LWM is passed, see if we're ok now
	if (m_stall_state == STALLED_UNTIL_FIFO_LWM)
	{
		// if there's room in the memory FIFO now, we can proceed
		if (m_reg.fbi_init0().enable_memory_fifo())
		{
			if (m_fbmem_fifo.items() < 2 * 32 * m_reg.fbi_init0().memory_fifo_hwm())
				resume = true;
		}
		else if (m_pci_fifo.space() > 2 * m_reg.fbi_init0().pci_fifo_lwm())
			resume = true;
	}

	// if we're stalled until the FIFOs are empty, check now
	else if (m_stall_state == STALLED_UNTIL_FIFO_EMPTY)
	{
		if (m_reg.fbi_init0().enable_memory_fifo())
		{
			if (m_fbmem_fifo.empty() && m_pci_fifo.empty())
				resume = true;
		}
		else if (m_pci_fifo.empty())
			resume = true;
	}

	// resume if necessary
	if (resume || !operation_pending())
	{
		if (LOG_FIFO)
			logerror("VOODOO.FIFO:Stall condition cleared; resuming\n");
		m_stall_state = NOT_STALLED;

		// either call the callback, or trigger the trigger
		if (!m_stall_cb.isnull())
			m_stall_cb(false);
		else
			machine().scheduler().trigger(m_stall_trigger);
	}

	// if not, set a timer for the next one
	else
		m_stall_resume_timer->adjust(m_operation_end - current_time);
}


//-------------------------------------------------
//  stall_cpu - stall our associated CPU until
//  operations are complete
//-------------------------------------------------

void voodoo_1_device::stall_cpu(stall_state state)
{
	// sanity check
	assert(operation_pending());

	// set the state and update statistics
	m_stall_state = state;
	if (DEBUG_STATS)
		m_stats.m_stalls++;

	// either call the callback, or spin the CPU
	if (!m_stall_cb.isnull())
		m_stall_cb(true);
	else
		m_cpu->spin_until_trigger(m_stall_trigger);

	// set a timer to clear the stall
	m_stall_resume_timer->adjust(m_operation_end - machine().time());
}


//-------------------------------------------------
//  stall_resume_callback - timer callback to
//  check the stall state for our CPU
//-------------------------------------------------

void voodoo_1_device::stall_resume_callback(void *ptr, s32 param)
{
	check_stalled_cpu(machine().time());
}


//**************************************************************************
//  VOODOO 1 REGISTER MAP
//**************************************************************************

#define REGISTER_ENTRY(name, reader, writer, bits, chips, sync, fifo) \
	{ static_register_table_entry<voodoo_1_device>::make_mask(bits), register_table_entry::CHIPMASK_##chips | register_table_entry::SYNC_##sync | register_table_entry::FIFO_##fifo, #name, &voodoo_1_device::reg_##writer##_w, &voodoo_1_device::reg_##reader##_r },

#define RESERVED_ENTRY REGISTER_ENTRY(reserved, invalid, invalid, 32, FBI, NOSYNC, FIFO)

#define RESERVED_ENTRY_x8 RESERVED_ENTRY RESERVED_ENTRY RESERVED_ENTRY RESERVED_ENTRY RESERVED_ENTRY RESERVED_ENTRY RESERVED_ENTRY RESERVED_ENTRY

static_register_table_entry<voodoo_1_device> const voodoo_1_device::s_register_table[256] =
{
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(status,          status,      unimplemented,32,FBI,      NOSYNC,   FIFO)    // 000
	RESERVED_ENTRY                                                                             // 004
	REGISTER_ENTRY(vertexAx,        invalid,     passive,     16, FBI_TREX, NOSYNC,   FIFO)    // 008
	REGISTER_ENTRY(vertexAy,        invalid,     passive,     16, FBI_TREX, NOSYNC,   FIFO)    // 00c
	REGISTER_ENTRY(vertexBx,        invalid,     passive,     16, FBI_TREX, NOSYNC,   FIFO)    // 010
	REGISTER_ENTRY(vertexBy,        invalid,     passive,     16, FBI_TREX, NOSYNC,   FIFO)    // 014
	REGISTER_ENTRY(vertexCx,        invalid,     passive,     16, FBI_TREX, NOSYNC,   FIFO)    // 018
	REGISTER_ENTRY(vertexCy,        invalid,     passive,     16, FBI_TREX, NOSYNC,   FIFO)    // 01c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(startR,          invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 020
	REGISTER_ENTRY(startG,          invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 024
	REGISTER_ENTRY(startB,          invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 028
	REGISTER_ENTRY(startZ,          invalid,     passive,     32, FBI,      NOSYNC,   FIFO)    // 02c
	REGISTER_ENTRY(startA,          invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 030
	REGISTER_ENTRY(startS,          invalid,     starts,      32, TREX,     NOSYNC,   FIFO)    // 034
	REGISTER_ENTRY(startT,          invalid,     startt,      32, TREX,     NOSYNC,   FIFO)    // 038
	REGISTER_ENTRY(startW,          invalid,     startw,      32, FBI_TREX, NOSYNC,   FIFO)    // 03c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(dRdX,            invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 040
	REGISTER_ENTRY(dGdX,            invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 044
	REGISTER_ENTRY(dBdX,            invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 048
	REGISTER_ENTRY(dZdX,            invalid,     passive,     32, FBI,      NOSYNC,   FIFO)    // 04c
	REGISTER_ENTRY(dAdX,            invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 050
	REGISTER_ENTRY(dSdX,            invalid,     dsdx,        32, TREX,     NOSYNC,   FIFO)    // 054
	REGISTER_ENTRY(dTdX,            invalid,     dtdx,        32, TREX,     NOSYNC,   FIFO)    // 058
	REGISTER_ENTRY(dWdX,            invalid,     dwdx,        32, FBI_TREX, NOSYNC,   FIFO)    // 05c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(dRdY,            invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 060
	REGISTER_ENTRY(dGdY,            invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 064
	REGISTER_ENTRY(dBdY,            invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 068
	REGISTER_ENTRY(dZdY,            invalid,     passive,     32, FBI,      NOSYNC,   FIFO)    // 06c
	REGISTER_ENTRY(dAdY,            invalid,     passive,     24, FBI,      NOSYNC,   FIFO)    // 070
	REGISTER_ENTRY(dSdY,            invalid,     dsdy,        32, TREX,     NOSYNC,   FIFO)    // 074
	REGISTER_ENTRY(dTdY,            invalid,     dtdy,        32, TREX,     NOSYNC,   FIFO)    // 078
	REGISTER_ENTRY(dWdY,            invalid,     dwdy,        32, FBI_TREX, NOSYNC,   FIFO)    // 07c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(triangleCMD,     invalid,     triangle,    32, FBI_TREX, NOSYNC,   FIFO)    // 080
	RESERVED_ENTRY                                                                             // 084
	REGISTER_ENTRY(fvertexAx,       invalid,     fpassive_4,  32, FBI_TREX, NOSYNC,   FIFO)    // 088
	REGISTER_ENTRY(fvertexAy,       invalid,     fpassive_4,  32, FBI_TREX, NOSYNC,   FIFO)    // 08c
	REGISTER_ENTRY(fvertexBx,       invalid,     fpassive_4,  32, FBI_TREX, NOSYNC,   FIFO)    // 090
	REGISTER_ENTRY(fvertexBy,       invalid,     fpassive_4,  32, FBI_TREX, NOSYNC,   FIFO)    // 094
	REGISTER_ENTRY(fvertexCx,       invalid,     fpassive_4,  32, FBI_TREX, NOSYNC,   FIFO)    // 098
	REGISTER_ENTRY(fvertexCy,       invalid,     fpassive_4,  32, FBI_TREX, NOSYNC,   FIFO)    // 09c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(fstartR,         invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0a0
	REGISTER_ENTRY(fstartG,         invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0a4
	REGISTER_ENTRY(fstartB,         invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0a8
	REGISTER_ENTRY(fstartZ,         invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0ac
	REGISTER_ENTRY(fstartA,         invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0b0
	REGISTER_ENTRY(fstartS,         invalid,     fstarts,     32, TREX,     NOSYNC,   FIFO)    // 0b4
	REGISTER_ENTRY(fstartT,         invalid,     fstartt,     32, TREX,     NOSYNC,   FIFO)    // 0b8
	REGISTER_ENTRY(fstartW,         invalid,     fstartw,     32, FBI_TREX, NOSYNC,   FIFO)    // 0bc
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(fdRdX,           invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0c0
	REGISTER_ENTRY(fdGdX,           invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0c4
	REGISTER_ENTRY(fdBdX,           invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0c8
	REGISTER_ENTRY(fdZdX,           invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0cc
	REGISTER_ENTRY(fdAdX,           invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0d0
	REGISTER_ENTRY(fdSdX,           invalid,     fdsdx,       32, TREX,     NOSYNC,   FIFO)    // 0d4
	REGISTER_ENTRY(fdTdX,           invalid,     fdtdx,       32, TREX,     NOSYNC,   FIFO)    // 0d8
	REGISTER_ENTRY(fdWdX,           invalid,     fdwdx,       32, FBI_TREX, NOSYNC,   FIFO)    // 0dc
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(fdRdY,           invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0e0
	REGISTER_ENTRY(fdGdY,           invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0e4
	REGISTER_ENTRY(fdBdY,           invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0e8
	REGISTER_ENTRY(fdZdY,           invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0ec
	REGISTER_ENTRY(fdAdY,           invalid,     fpassive_12, 32, FBI,      NOSYNC,   FIFO)    // 0f0
	REGISTER_ENTRY(fdSdY,           invalid,     fdsdy,       32, TREX,     NOSYNC,   FIFO)    // 0f4
	REGISTER_ENTRY(fdTdY,           invalid,     fdtdy,       32, TREX,     NOSYNC,   FIFO)    // 0f8
	REGISTER_ENTRY(fdWdY,           invalid,     fdwdy,       32, FBI_TREX, NOSYNC,   FIFO)    // 0fc
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(ftriangleCMD,    invalid,     triangle,    32, FBI_TREX, NOSYNC,   FIFO)    // 100
	REGISTER_ENTRY(fbzColorPath,    passive,     passive,     28, FBI_TREX, NOSYNC,   FIFO)    // 104
	REGISTER_ENTRY(fogMode,         passive,     passive,      6, FBI_TREX, NOSYNC,   FIFO)    // 108
	REGISTER_ENTRY(alphaMode,       passive,     passive,     32, FBI_TREX, NOSYNC,   FIFO)    // 10c
	REGISTER_ENTRY(fbzMode,         passive,     passive,     21, FBI_TREX,   SYNC,   FIFO)    // 110
	REGISTER_ENTRY(lfbMode,         passive,     passive,     17, FBI_TREX,   SYNC,   FIFO)    // 114
	REGISTER_ENTRY(clipLeftRight,   passive,     passive,     26, FBI_TREX,   SYNC,   FIFO)    // 118
	REGISTER_ENTRY(clipLowYHighY,   passive,     passive,     26, FBI_TREX,   SYNC,   FIFO)    // 11c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(nopCMD,          invalid,     nop,          1, FBI_TREX,   SYNC,   FIFO)    // 120
	REGISTER_ENTRY(fastfillCMD,     invalid,     fastfill,     0, FBI,        SYNC,   FIFO)    // 124
	REGISTER_ENTRY(swapbufferCMD,   invalid,     swapbuffer,   9, FBI,        SYNC,   FIFO)    // 128
	REGISTER_ENTRY(fogColor,        invalid,     passive,     24, FBI,        SYNC,   FIFO)    // 12c
	REGISTER_ENTRY(zaColor,         invalid,     passive,     32, FBI,        SYNC,   FIFO)    // 130
	REGISTER_ENTRY(chromaKey,       invalid,     passive,     24, FBI,        SYNC,   FIFO)    // 134
	RESERVED_ENTRY                                                                             // 138
	RESERVED_ENTRY                                                                             // 13c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(stipple,         passive,     passive,     32, FBI,        SYNC,   FIFO)    // 140
	REGISTER_ENTRY(color0,          passive,     passive,     32, FBI,        SYNC,   FIFO)    // 144
	REGISTER_ENTRY(color1,          passive,     passive,     32, FBI,        SYNC,   FIFO)    // 148
	REGISTER_ENTRY(fbiPixelsIn,     stats,       invalid,     24, FBI,          NA,     NA)    // 14c
	REGISTER_ENTRY(fbiChromaFail,   stats,       invalid,     24, FBI,          NA,     NA)    // 150
	REGISTER_ENTRY(fbiZfuncFail,    stats,       invalid,     24, FBI,          NA,     NA)    // 154
	REGISTER_ENTRY(fbiAfuncFail,    stats,       invalid,     24, FBI,          NA,     NA)    // 158
	REGISTER_ENTRY(fbiPixelsOut,    stats,       invalid,     24, FBI,          NA,     NA)    // 15c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(fogTable[0],     invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 160
	REGISTER_ENTRY(fogTable[1],     invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 164
	REGISTER_ENTRY(fogTable[2],     invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 168
	REGISTER_ENTRY(fogTable[3],     invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 16c
	REGISTER_ENTRY(fogTable[4],     invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 170
	REGISTER_ENTRY(fogTable[5],     invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 174
	REGISTER_ENTRY(fogTable[6],     invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 178
	REGISTER_ENTRY(fogTable[7],     invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 17c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(fogTable[8],     invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 180
	REGISTER_ENTRY(fogTable[9],     invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 184
	REGISTER_ENTRY(fogTable[10],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 188
	REGISTER_ENTRY(fogTable[11],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 18c
	REGISTER_ENTRY(fogTable[12],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 190
	REGISTER_ENTRY(fogTable[13],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 194
	REGISTER_ENTRY(fogTable[14],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 198
	REGISTER_ENTRY(fogTable[15],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 19c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(fogTable[16],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1a0
	REGISTER_ENTRY(fogTable[17],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1a4
	REGISTER_ENTRY(fogTable[18],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1a8
	REGISTER_ENTRY(fogTable[19],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1ac
	REGISTER_ENTRY(fogTable[20],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1b0
	REGISTER_ENTRY(fogTable[21],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1b4
	REGISTER_ENTRY(fogTable[22],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1b8
	REGISTER_ENTRY(fogTable[23],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1bc
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(fogTable[24],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1c0
	REGISTER_ENTRY(fogTable[25],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1c4
	REGISTER_ENTRY(fogTable[26],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1c8
	REGISTER_ENTRY(fogTable[27],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1cc
	REGISTER_ENTRY(fogTable[28],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1d0
	REGISTER_ENTRY(fogTable[29],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1d4
	REGISTER_ENTRY(fogTable[30],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1d8
	REGISTER_ENTRY(fogTable[31],    invalid,     fogtable,    32, FBI,        SYNC,   FIFO)    // 1dc
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	RESERVED_ENTRY_x8                                                                // 1e0-1fc
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(fbiInit4,        passive,     fbiinit,     28, FBI,      NOSYNC, NOFIFO)    // 200
	REGISTER_ENTRY(vRetrace,        vretrace,    invalid,     12, FBI,          NA,     NA)    // 204
	REGISTER_ENTRY(backPorch,       passive,     video,       24, FBI,      NOSYNC, NOFIFO)    // 208
	REGISTER_ENTRY(videoDimensions, passive,     video,       26, FBI,      NOSYNC, NOFIFO)    // 20c
	REGISTER_ENTRY(fbiInit0,        passive,     fbiinit,     31, FBI,      NOSYNC, NOFIFO)    // 210
	REGISTER_ENTRY(fbiInit1,        passive,     fbiinit,     32, FBI,      NOSYNC, NOFIFO)    // 214
	REGISTER_ENTRY(fbiInit2,        fbiinit2,    fbiinit,     32, FBI,      NOSYNC, NOFIFO)    // 218
	REGISTER_ENTRY(fbiInit3,        passive,     fbiinit,     32, FBI,      NOSYNC, NOFIFO)    // 21c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(hSync,           invalid,     video,       26, FBI,      NOSYNC, NOFIFO)    // 220
	REGISTER_ENTRY(vSync,           invalid,     video,       28, FBI,      NOSYNC, NOFIFO)    // 224
	REGISTER_ENTRY(clutData,        invalid,     clut,        30, FBI,      NOSYNC, NOFIFO)    // 228
	REGISTER_ENTRY(dacData,         invalid,     dac,         12, FBI,      NOSYNC, NOFIFO)    // 22c
	REGISTER_ENTRY(maxRgbDelta,     invalid,     unimplemented,24,FBI,      NOSYNC, NOFIFO)    // 230
	RESERVED_ENTRY                                                                             // 234
	RESERVED_ENTRY                                                                             // 238
	RESERVED_ENTRY                                                                             // 23c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	RESERVED_ENTRY_x8                                                                          // 240-25c
	RESERVED_ENTRY_x8                                                                          // 260-27c
	RESERVED_ENTRY_x8                                                                          // 280-29c
	RESERVED_ENTRY_x8                                                                          // 2a0-2bc
	RESERVED_ENTRY_x8                                                                          // 2c0-2dc
	RESERVED_ENTRY_x8                                                                          // 2e0-2fc
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(textureMode,     invalid,     texture,     32, TREX,     NOSYNC,   FIFO)    // 300
	REGISTER_ENTRY(tLOD,            invalid,     texture,     32, TREX,     NOSYNC,   FIFO)    // 304
	REGISTER_ENTRY(tDetail,         invalid,     texture,     17, TREX,     NOSYNC,   FIFO)    // 308
	REGISTER_ENTRY(texBaseAddr,     invalid,     texture,     19, TREX,     NOSYNC,   FIFO)    // 30c
	REGISTER_ENTRY(texBaseAddr_1,   invalid,     texture,     19, TREX,     NOSYNC,   FIFO)    // 310
	REGISTER_ENTRY(texBaseAddr_2,   invalid,     texture,     19, TREX,     NOSYNC,   FIFO)    // 314
	REGISTER_ENTRY(texBaseAddr_3_8, invalid,     texture,     19, TREX,     NOSYNC,   FIFO)    // 318
	REGISTER_ENTRY(trexInit0,       invalid,     passive,     32, TREX,       SYNC,   FIFO)    // 31c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(trexInit1,       invalid,     passive,     32, TREX,       SYNC,   FIFO)    // 320
	REGISTER_ENTRY(nccTable0[0],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 324
	REGISTER_ENTRY(nccTable0[1],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 328
	REGISTER_ENTRY(nccTable0[2],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 32c
	REGISTER_ENTRY(nccTable0[3],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 330
	REGISTER_ENTRY(nccTable0[4],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 334
	REGISTER_ENTRY(nccTable0[5],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 338
	REGISTER_ENTRY(nccTable0[6],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 33c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(nccTable0[7],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 340
	REGISTER_ENTRY(nccTable0[8],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 344
	REGISTER_ENTRY(nccTable0[9],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 348
	REGISTER_ENTRY(nccTable0[10],   invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 34c
	REGISTER_ENTRY(nccTable0[11],   invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 350
	REGISTER_ENTRY(nccTable1[0],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 354
	REGISTER_ENTRY(nccTable1[1],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 358
	REGISTER_ENTRY(nccTable1[2],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 35c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(nccTable1[3],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 360
	REGISTER_ENTRY(nccTable1[4],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 364
	REGISTER_ENTRY(nccTable1[5],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 368
	REGISTER_ENTRY(nccTable1[6],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 36c
	REGISTER_ENTRY(nccTable1[7],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 370
	REGISTER_ENTRY(nccTable1[8],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 374
	REGISTER_ENTRY(nccTable1[9],    invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 378
	REGISTER_ENTRY(nccTable1[10],   invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 37c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	REGISTER_ENTRY(nccTable1[11],   invalid,     palette,     32, TREX,       SYNC,   FIFO)    // 380
	RESERVED_ENTRY                                                                             // 384
	RESERVED_ENTRY                                                                             // 388
	RESERVED_ENTRY                                                                             // 38c
	RESERVED_ENTRY                                                                             // 390
	RESERVED_ENTRY                                                                             // 394
	RESERVED_ENTRY                                                                             // 398
	RESERVED_ENTRY                                                                             // 39c
	//             name             rd handler   wr handler  bits chips     sync?     fifo?
	RESERVED_ENTRY_x8                                                                          // 3a0-3bc
	RESERVED_ENTRY_x8                                                                          // 3c0-3dc
	RESERVED_ENTRY_x8                                                                          // 3e0-3fc
};