/
GLES_GPU.cpp
1913 lines (1629 loc) · 62.7 KB
/
GLES_GPU.cpp
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
// Copyright (c) 2012- PPSSPP Project.
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, version 2.0 or later versions.
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License 2.0 for more details.
// A copy of the GPL 2.0 should have been included with the program.
// If not, see http://www.gnu.org/licenses/
// Official git repository and contact information can be found at
// https://github.com/hrydgard/ppsspp and http://www.ppsspp.org/.
#include "base/logging.h"
#include "gfx_es2/gl_state.h"
#include "Common/ChunkFile.h"
#include "Core/Debugger/Breakpoints.h"
#include "Core/MemMap.h"
#include "Core/Host.h"
#include "Core/Config.h"
#include "Core/Reporting.h"
#include "Core/System.h"
#include "GPU/GPUState.h"
#include "GPU/ge_constants.h"
#include "GPU/GeDisasm.h"
#include "GPU/GLES/ShaderManager.h"
#include "GPU/GLES/GLES_GPU.h"
#include "GPU/GLES/Framebuffer.h"
#include "GPU/GLES/TransformPipeline.h"
#include "GPU/GLES/TextureCache.h"
#include "Core/MIPS/MIPS.h"
#include "Core/HLE/sceKernelThread.h"
#include "Core/HLE/sceKernelInterrupt.h"
#include "Core/HLE/sceGe.h"
enum {
FLAG_FLUSHBEFORE = 1,
FLAG_FLUSHBEFOREONCHANGE = 2,
FLAG_EXECUTE = 4, // needs to actually be executed. unused for now.
FLAG_EXECUTEONCHANGE = 8,
FLAG_ANY_EXECUTE = 4 | 8,
FLAG_READS_PC = 16,
FLAG_WRITES_PC = 32,
};
struct CommandTableEntry {
u8 cmd;
u8 flags;
GLES_GPU::CmdFunc func;
};
static const CommandTableEntry commandTable[] = {
// Changes that dirty the framebuffer
{GE_CMD_FRAMEBUFPTR, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_FramebufType},
{GE_CMD_FRAMEBUFWIDTH, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_FramebufType},
{GE_CMD_FRAMEBUFPIXFORMAT, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_FramebufType},
{GE_CMD_ZBUFPTR, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_ZBUFWIDTH, FLAG_FLUSHBEFOREONCHANGE},
// Changes that dirty uniforms
{GE_CMD_FOGCOLOR, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_FOG1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_FOG2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
// Should these maybe flush?
{GE_CMD_MINZ, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_MAXZ, FLAG_FLUSHBEFOREONCHANGE},
// Changes that dirty texture scaling.
{GE_CMD_TEXMAPMODE, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXSCALEU, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXSCALEV, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXOFFSETU, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXOFFSETV, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
// Changes that dirty the current texture. Really should be possible to avoid executing these if we compile
// by adding some more flags.
{GE_CMD_TEXSIZE0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTE, &GLES_GPU::Execute_TexSize0},
{GE_CMD_TEXSIZE1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXSIZE2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXSIZE3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXSIZE4, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXSIZE5, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXSIZE6, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXSIZE7, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXFORMAT, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXADDR0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_TexAddr0},
{GE_CMD_TEXADDR1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXADDR2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXADDR3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXADDR4, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXADDR5, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXADDR6, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXADDR7, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXBUFWIDTH0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXBUFWIDTH1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXBUFWIDTH2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXBUFWIDTH3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXBUFWIDTH4, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXBUFWIDTH5, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXBUFWIDTH6, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXBUFWIDTH7, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
// These must flush on change, so that LoadClut doesn't have to always flush.
{GE_CMD_CLUTADDR, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_CLUTADDRUPPER, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_CLUTFORMAT, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
// These affect the fragment shader so need flushing.
{GE_CMD_CLEARMODE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_TEXTUREMAPENABLE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_FOGENABLE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_TEXMODE, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXSHADELS, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_SHADEMODE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_TEXFUNC, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_COLORTEST, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_ALPHATESTENABLE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_COLORTESTENABLE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_COLORTESTMASK, FLAG_FLUSHBEFOREONCHANGE},
// These change the vertex shader so need flushing.
{GE_CMD_REVERSENORMAL, FLAG_FLUSHBEFOREONCHANGE}, // TODO: This one is actually processed during vertex decoding which is wrong.
{GE_CMD_LIGHTINGENABLE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_LIGHTENABLE0, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_LIGHTENABLE1, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_LIGHTENABLE2, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_LIGHTENABLE3, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_LIGHTTYPE0, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_LIGHTTYPE1, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_LIGHTTYPE2, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_LIGHTTYPE3, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_MATERIALUPDATE, FLAG_FLUSHBEFOREONCHANGE},
// This changes both shaders so need flushing.
{GE_CMD_LIGHTMODE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_TEXFILTER, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXWRAP, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
// Uniform changes
{GE_CMD_ALPHATEST, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_COLORREF, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_TEXENVCOLOR, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
// Simple render state changes. Handled in StateMapping.cpp.
{GE_CMD_OFFSETX, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_OFFSETY, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_CULL, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_CULLFACEENABLE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_DITHERENABLE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_STENCILOP, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_STENCILTEST, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_STENCILTESTENABLE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_ALPHABLENDENABLE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_BLENDMODE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_BLENDFIXEDA, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_BLENDFIXEDB, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_MASKRGB, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_MASKALPHA, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_ZTEST, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_ZTESTENABLE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_ZWRITEDISABLE, FLAG_FLUSHBEFOREONCHANGE},
#ifndef USING_GLES2
{GE_CMD_LOGICOP, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_LOGICOPENABLE, FLAG_FLUSHBEFOREONCHANGE},
#else
{GE_CMD_LOGICOP, 0},
{GE_CMD_LOGICOPENABLE, 0},
#endif
// Can probably ignore this one as we don't support AA lines.
{GE_CMD_ANTIALIASENABLE, FLAG_FLUSHBEFOREONCHANGE},
// Morph weights. TODO: Remove precomputation?
{GE_CMD_MORPHWEIGHT0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_MORPHWEIGHT1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_MORPHWEIGHT2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_MORPHWEIGHT3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_MORPHWEIGHT4, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_MORPHWEIGHT5, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_MORPHWEIGHT6, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_MORPHWEIGHT7, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
// Control spline/bezier patches. Don't really require flushing as such, but meh.
{GE_CMD_PATCHDIVISION, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_PATCHPRIMITIVE, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_PATCHFACING, FLAG_FLUSHBEFOREONCHANGE},
{GE_CMD_PATCHCULLENABLE, FLAG_FLUSHBEFOREONCHANGE},
// Viewport.
{GE_CMD_VIEWPORTX1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_VIEWPORTY1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_VIEWPORTX2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_VIEWPORTY2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_VIEWPORTZ1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_VIEWPORTZ2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
// Region
{GE_CMD_REGION1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_Region},
{GE_CMD_REGION2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_Region},
// Scissor
{GE_CMD_SCISSOR1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_SCISSOR2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
// These dirty various vertex shader uniforms. Could embed information about that in this table and call dirtyuniform directly, hm...
{GE_CMD_AMBIENTCOLOR, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_Ambient},
{GE_CMD_AMBIENTALPHA, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_Ambient},
{GE_CMD_MATERIALDIFFUSE, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_MaterialDiffuse},
{GE_CMD_MATERIALEMISSIVE, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_MaterialEmissive},
{GE_CMD_MATERIALAMBIENT, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_MaterialAmbient},
{GE_CMD_MATERIALALPHA, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_MaterialAmbient},
{GE_CMD_MATERIALSPECULAR, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_MaterialSpecular},
{GE_CMD_MATERIALSPECULARCOEF, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_MaterialSpecular},
// These precompute a value. not sure if worth it. Also dirty uniforms, which could be table-ized to avoid execute.
{GE_CMD_LX0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LY0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LZ0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LX1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LY1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LZ1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LX2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LY2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LZ2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LX3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LY3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LZ3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDX0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDY0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDZ0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDX1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDY1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDZ1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDX2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDY2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDZ2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDX3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDY3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDZ3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKA0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKB0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKC0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKA1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKB1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKC1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKA2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKB2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKC2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKA3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKB3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKC3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKS0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKS1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKS2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKS3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKO0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKO1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKO2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LKO3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LAC0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDC0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LSC0, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LAC1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDC1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LSC1, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LAC2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDC2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LSC2, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LAC3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LDC3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
{GE_CMD_LSC3, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE},
// Ignored commands
{GE_CMD_CLIPENABLE, 0},
{GE_CMD_TEXFLUSH, 0},
{GE_CMD_TEXLODSLOPE, 0},
{GE_CMD_TEXLEVEL, FLAG_EXECUTEONCHANGE}, // we don't support this anyway, no need to flush.
{GE_CMD_TEXSYNC, 0},
// These are just nop or part of other later commands.
{GE_CMD_NOP, 0},
{GE_CMD_BASE, 0},
{GE_CMD_TRANSFERSRC, 0},
{GE_CMD_TRANSFERSRCW, 0},
{GE_CMD_TRANSFERDST, 0},
{GE_CMD_TRANSFERDSTW, 0},
{GE_CMD_TRANSFERSRCPOS, 0},
{GE_CMD_TRANSFERDSTPOS, 0},
{GE_CMD_TRANSFERSIZE, 0},
// From Common. No flushing but definitely need execute.
{GE_CMD_OFFSETADDR, FLAG_EXECUTE, &GPUCommon::Execute_OffsetAddr},
{GE_CMD_ORIGIN, FLAG_EXECUTE | FLAG_READS_PC, &GPUCommon::Execute_Origin}, // Really?
{GE_CMD_PRIM, FLAG_EXECUTE, &GLES_GPU::Execute_Prim},
{GE_CMD_JUMP, FLAG_EXECUTE | FLAG_READS_PC | FLAG_WRITES_PC, &GPUCommon::Execute_Jump},
{GE_CMD_CALL, FLAG_EXECUTE | FLAG_READS_PC | FLAG_WRITES_PC, &GPUCommon::Execute_Call},
{GE_CMD_RET, FLAG_EXECUTE | FLAG_READS_PC | FLAG_WRITES_PC, &GPUCommon::Execute_Ret},
{GE_CMD_END, FLAG_FLUSHBEFORE | FLAG_EXECUTE | FLAG_READS_PC | FLAG_WRITES_PC, &GPUCommon::Execute_End}, // Flush?
{GE_CMD_VADDR, FLAG_EXECUTE, &GLES_GPU::Execute_Vaddr},
{GE_CMD_IADDR, FLAG_EXECUTE, &GLES_GPU::Execute_Iaddr},
{GE_CMD_BJUMP, FLAG_EXECUTE | FLAG_READS_PC | FLAG_WRITES_PC, &GPUCommon::Execute_BJump}, // EXECUTE
{GE_CMD_BOUNDINGBOX, FLAG_EXECUTE}, // + FLUSHBEFORE when we implement... or not, do we need to?
// Changing the vertex type requires us to flush.
{GE_CMD_VERTEXTYPE, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTEONCHANGE, &GLES_GPU::Execute_VertexType},
{GE_CMD_BEZIER, FLAG_FLUSHBEFORE | FLAG_EXECUTE},
{GE_CMD_SPLINE, FLAG_FLUSHBEFORE | FLAG_EXECUTE},
// These two are actually processed in CMD_END.
{GE_CMD_SIGNAL, FLAG_FLUSHBEFORE},
{GE_CMD_FINISH, FLAG_FLUSHBEFORE},
// Changes that trigger data copies. Only flushing on change for LOADCLUT must be a bit of a hack...
{GE_CMD_LOADCLUT, FLAG_FLUSHBEFOREONCHANGE | FLAG_EXECUTE, &GLES_GPU::Execute_LoadClut},
{GE_CMD_TRANSFERSTART, FLAG_FLUSHBEFORE | FLAG_EXECUTE | FLAG_READS_PC},
// We don't use the dither table.
{GE_CMD_DITH0},
{GE_CMD_DITH1},
{GE_CMD_DITH2},
{GE_CMD_DITH3},
// These handle their own flushing.
{GE_CMD_WORLDMATRIXNUMBER, FLAG_EXECUTE | FLAG_READS_PC | FLAG_WRITES_PC},
{GE_CMD_WORLDMATRIXDATA, FLAG_EXECUTE},
{GE_CMD_VIEWMATRIXNUMBER, FLAG_EXECUTE | FLAG_READS_PC | FLAG_WRITES_PC},
{GE_CMD_VIEWMATRIXDATA, FLAG_EXECUTE},
{GE_CMD_PROJMATRIXNUMBER, FLAG_EXECUTE | FLAG_READS_PC | FLAG_WRITES_PC},
{GE_CMD_PROJMATRIXDATA, FLAG_EXECUTE},
{GE_CMD_TGENMATRIXNUMBER, FLAG_EXECUTE | FLAG_READS_PC | FLAG_WRITES_PC},
{GE_CMD_TGENMATRIXDATA, FLAG_EXECUTE},
{GE_CMD_BONEMATRIXNUMBER, FLAG_EXECUTE | FLAG_READS_PC | FLAG_WRITES_PC},
{GE_CMD_BONEMATRIXDATA, FLAG_EXECUTE},
// "Missing" commands (gaps in the sequence)
{GE_CMD_UNKNOWN_03, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_0D, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_11, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_29, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_34, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_35, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_39, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_4E, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_4F, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_52, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_59, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_5A, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_B6, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_B7, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_D1, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_ED, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_EF, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_F0, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_F1, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_F2, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_F3, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_F4, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_F5, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_F6, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_F7, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_F8, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_F9, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_FA, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_FB, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_FC, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_FD, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_FE, FLAG_EXECUTE},
{GE_CMD_UNKNOWN_FF, FLAG_EXECUTE},
};
GLES_GPU::CommandInfo GLES_GPU::cmdInfo_[256];
GLES_GPU::GLES_GPU()
: resized_(false) {
#ifdef _WIN32
lastVsync_ = g_Config.bVSync ? 1 : 0;
// Disabled EXT_swap_control_tear for now, it never seems to settle at the correct timing
// so it just keeps tearing. Not what I hoped for...
//if (false && gl_extensions.EXT_swap_control_tear) {
// See http://developer.download.nvidia.com/opengl/specs/WGL_EXT_swap_control_tear.txt
// glstate.SetVSyncInterval(g_Config.bVSync ? -1 :0);
//} else {
glstate.SetVSyncInterval(g_Config.bVSync ? 1 : 0);
//}
#endif
shaderManager_ = new ShaderManager();
transformDraw_.SetShaderManager(shaderManager_);
transformDraw_.SetTextureCache(&textureCache_);
transformDraw_.SetFramebufferManager(&framebufferManager_);
framebufferManager_.SetTextureCache(&textureCache_);
framebufferManager_.SetShaderManager(shaderManager_);
textureCache_.SetFramebufferManager(&framebufferManager_);
// Sanity check gstate
if ((int *)&gstate.transferstart - (int *)&gstate != 0xEA) {
ERROR_LOG(G3D, "gstate has drifted out of sync!");
}
// Sanity check cmdInfo_ table - no dupes please
std::set<u8> dupeCheck;
memset(cmdInfo_, 0, sizeof(cmdInfo_));
for (size_t i = 0; i < ARRAY_SIZE(commandTable); i++) {
const u8 cmd = commandTable[i].cmd;
if (dupeCheck.find(cmd) != dupeCheck.end()) {
ERROR_LOG(G3D, "Command table Dupe: %02x (%i)", (int)cmd, (int)cmd);
} else {
dupeCheck.insert(cmd);
}
cmdInfo_[cmd].flags |= commandTable[i].flags;
cmdInfo_[cmd].func = commandTable[i].func;
if (!cmdInfo_[cmd].func) {
cmdInfo_[cmd].func = &GLES_GPU::ExecuteOpInternal;
}
}
// Find commands missing from the table.
for (int i = 0; i < 0xEF; i++) {
if (dupeCheck.find((u8)i) == dupeCheck.end()) {
ERROR_LOG(G3D, "Command missing from table: %02x (%i)", i, i);
}
}
// No need to flush before the tex scale/offset commands if we are baking
// the tex scale/offset into the vertices anyway.
if (g_Config.bPrescaleUV) {
cmdInfo_[GE_CMD_TEXSCALEU].flags &= ~FLAG_FLUSHBEFOREONCHANGE;
cmdInfo_[GE_CMD_TEXSCALEV].flags &= ~FLAG_FLUSHBEFOREONCHANGE;
cmdInfo_[GE_CMD_TEXOFFSETU].flags &= ~FLAG_FLUSHBEFOREONCHANGE;
cmdInfo_[GE_CMD_TEXOFFSETV].flags &= ~FLAG_FLUSHBEFOREONCHANGE;
}
if (g_Config.bSoftwareSkinning) {
cmdInfo_[GE_CMD_VERTEXTYPE].flags &= ~FLAG_FLUSHBEFOREONCHANGE;
}
BuildReportingInfo();
}
GLES_GPU::~GLES_GPU() {
framebufferManager_.DestroyAllFBOs();
shaderManager_->ClearCache(true);
delete shaderManager_;
}
// Let's avoid passing nulls into snprintf().
static const char *GetGLStringAlways(GLenum name) {
const GLubyte *value = glGetString(name);
if (!value)
return "?";
return (const char *)value;
}
// Needs to be called on GPU thread, not reporting thread.
void GLES_GPU::BuildReportingInfo() {
const char *glVendor = GetGLStringAlways(GL_VENDOR);
const char *glRenderer = GetGLStringAlways(GL_RENDERER);
const char *glVersion = GetGLStringAlways(GL_VERSION);
const char *glSlVersion = GetGLStringAlways(GL_SHADING_LANGUAGE_VERSION);
const char *glExtensions = GetGLStringAlways(GL_EXTENSIONS);
char temp[16384];
snprintf(temp, sizeof(temp), "%s (%s %s), %s (extensions: %s)", glVersion, glVendor, glRenderer, glSlVersion, glExtensions);
reportingPrimaryInfo_ = glVendor;
reportingFullInfo_ = temp;
Reporting::UpdateConfig();
}
void GLES_GPU::DeviceLost() {
ILOG("GLES_GPU: DeviceLost");
// Should only be executed on the GL thread.
// Simply drop all caches and textures.
// FBOs appear to survive? Or no?
// TransformDraw has registered as a GfxResourceHolder.
shaderManager_->ClearCache(false);
textureCache_.Clear(false);
framebufferManager_.DeviceLost();
}
void GLES_GPU::InitClear() {
ScheduleEvent(GPU_EVENT_INIT_CLEAR);
}
void GLES_GPU::InitClearInternal() {
bool useNonBufferedRendering = g_Config.iRenderingMode == FB_NON_BUFFERED_MODE;
if (useNonBufferedRendering) {
glstate.depthWrite.set(GL_TRUE);
glstate.colorMask.set(GL_TRUE, GL_TRUE, GL_TRUE, GL_TRUE);
glClearColor(0,0,0,1);
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT | GL_STENCIL_BUFFER_BIT);
}
glstate.viewport.set(0, 0, PSP_CoreParameter().pixelWidth, PSP_CoreParameter().pixelHeight);
}
void GLES_GPU::DumpNextFrame() {
dumpNextFrame_ = true;
}
void GLES_GPU::BeginFrame() {
ScheduleEvent(GPU_EVENT_BEGIN_FRAME);
}
void GLES_GPU::BeginFrameInternal() {
#ifdef _WIN32
// Turn off vsync when unthrottled
int desiredVSyncInterval = g_Config.bVSync ? 1 : 0;
if ((PSP_CoreParameter().unthrottle) || (PSP_CoreParameter().fpsLimit == 1))
desiredVSyncInterval = 0;
if (desiredVSyncInterval != lastVsync_) {
// Disabled EXT_swap_control_tear for now, it never seems to settle at the correct timing
// so it just keeps tearing. Not what I hoped for...
//if (false && gl_extensions.EXT_swap_control_tear) {
// See http://developer.download.nvidia.com/opengl/specs/WGL_EXT_swap_control_tear.txt
// glstate.SetVSyncInterval(-desiredVSyncInterval);
//} else {
glstate.SetVSyncInterval(desiredVSyncInterval);
//}
lastVsync_ = desiredVSyncInterval;
}
#endif
textureCache_.StartFrame();
transformDraw_.DecimateTrackedVertexArrays();
if (dumpNextFrame_) {
NOTICE_LOG(G3D, "DUMPING THIS FRAME");
dumpThisFrame_ = true;
dumpNextFrame_ = false;
} else if (dumpThisFrame_) {
dumpThisFrame_ = false;
}
shaderManager_->DirtyShader();
// Not sure if this is really needed.
shaderManager_->DirtyUniform(DIRTY_ALL);
framebufferManager_.BeginFrame();
}
void GLES_GPU::SetDisplayFramebuffer(u32 framebuf, u32 stride, GEBufferFormat format) {
host->GPUNotifyDisplay(framebuf, stride, format);
framebufferManager_.SetDisplayFramebuffer(framebuf, stride, format);
}
bool GLES_GPU::FramebufferDirty() {
if (g_Config.bSeparateCPUThread) {
// Allow it to process fully before deciding if it's dirty.
SyncThread();
}
VirtualFramebuffer *vfb = framebufferManager_.GetDisplayVFB();
if (vfb) {
bool dirty = vfb->dirtyAfterDisplay;
vfb->dirtyAfterDisplay = false;
return dirty;
}
return true;
}
bool GLES_GPU::FramebufferReallyDirty() {
if (g_Config.bSeparateCPUThread) {
// Allow it to process fully before deciding if it's dirty.
SyncThread();
}
VirtualFramebuffer *vfb = framebufferManager_.GetDisplayVFB();
if (vfb) {
bool dirty = vfb->reallyDirtyAfterDisplay;
vfb->reallyDirtyAfterDisplay = false;
return dirty;
}
return true;
}
void GLES_GPU::CopyDisplayToOutput() {
ScheduleEvent(GPU_EVENT_COPY_DISPLAY_TO_OUTPUT);
}
void GLES_GPU::CopyDisplayToOutputInternal() {
glstate.depthWrite.set(GL_TRUE);
glstate.colorMask.set(GL_TRUE, GL_TRUE, GL_TRUE, GL_TRUE);
transformDraw_.Flush();
framebufferManager_.CopyDisplayToOutput();
framebufferManager_.EndFrame();
shaderManager_->DirtyLastShader();
// If buffered, discard the depth buffer of the backbuffer. Don't even know if we need one.
#if 0
#ifdef USING_GLES2
if (gl_extensions.EXT_discard_framebuffer && g_Config.iRenderingMode != 0) {
GLenum attachments[] = {GL_DEPTH_EXT, GL_STENCIL_EXT};
glDiscardFramebufferEXT(GL_FRAMEBUFFER, 2, attachments);
}
#endif
#endif
gstate_c.textureChanged = TEXCHANGE_UPDATED;
}
// Maybe should write this in ASM...
void GLES_GPU::FastRunLoop(DisplayList &list) {
const CommandInfo *cmdInfo = cmdInfo_;
for (; downcount > 0; --downcount) {
// We know that display list PCs have the upper nibble == 0 - no need to mask the pointer
const u32 op = *(const u32 *)(Memory::base + list.pc);
const u32 cmd = op >> 24;
const CommandInfo info = cmdInfo[cmd];
const u8 cmdFlags = info.flags; // If we stashed the cmdFlags in the top bits of the cmdmem, we could get away with one table lookup instead of two
const u32 diff = op ^ gstate.cmdmem[cmd];
// Inlined CheckFlushOp here to get rid of the dumpThisFrame_ check.
if ((cmdFlags & FLAG_FLUSHBEFORE) || (diff && (cmdFlags & FLAG_FLUSHBEFOREONCHANGE))) {
transformDraw_.Flush();
}
gstate.cmdmem[cmd] = op; // TODO: no need to write if diff==0...
if ((cmdFlags & FLAG_EXECUTE) || (diff && (cmdFlags & FLAG_EXECUTEONCHANGE))) {
(this->*info.func)(op, diff);
}
list.pc += 4;
}
}
void GLES_GPU::ProcessEvent(GPUEvent ev) {
switch (ev.type) {
case GPU_EVENT_INIT_CLEAR:
InitClearInternal();
break;
case GPU_EVENT_BEGIN_FRAME:
BeginFrameInternal();
break;
case GPU_EVENT_COPY_DISPLAY_TO_OUTPUT:
CopyDisplayToOutputInternal();
break;
case GPU_EVENT_INVALIDATE_CACHE:
InvalidateCacheInternal(ev.invalidate_cache.addr, ev.invalidate_cache.size, ev.invalidate_cache.type);
break;
default:
GPUCommon::ProcessEvent(ev);
}
}
inline void GLES_GPU::CheckFlushOp(int cmd, u32 diff) {
const u8 cmdFlags = cmdInfo_[cmd].flags;
if ((cmdFlags & FLAG_FLUSHBEFORE) || (diff && (cmdFlags & FLAG_FLUSHBEFOREONCHANGE))) {
if (dumpThisFrame_) {
NOTICE_LOG(G3D, "================ FLUSH ================");
}
transformDraw_.Flush();
}
}
void GLES_GPU::PreExecuteOp(u32 op, u32 diff) {
CheckFlushOp(op >> 24, diff);
}
void GLES_GPU::ExecuteOp(u32 op, u32 diff) {
const u8 cmd = op >> 24;
const CommandInfo info = cmdInfo_[cmd];
const u8 cmdFlags = info.flags;
if ((cmdFlags & FLAG_EXECUTE) || (diff && (cmdFlags & FLAG_EXECUTEONCHANGE))) {
(this->*info.func)(op, diff);
}
}
void GLES_GPU::Execute_Vaddr(u32 op, u32 diff) {
gstate_c.vertexAddr = gstate_c.getRelativeAddress(op & 0x00FFFFFF);
}
void GLES_GPU::Execute_Iaddr(u32 op, u32 diff) {
gstate_c.indexAddr = gstate_c.getRelativeAddress(op & 0x00FFFFFF);
}
void GLES_GPU::Execute_Prim(u32 op, u32 diff) {
// This drives all drawing. All other state we just buffer up, then we apply it only
// when it's time to draw. As most PSP games set state redundantly ALL THE TIME, this is a huge optimization.
u32 data = op & 0xFFFFFF;
u32 count = data & 0xFFFF;
GEPrimitiveType prim = static_cast<GEPrimitiveType>(data >> 16);
if (count == 0)
return;
// Discard AA lines as we can't do anything that makes sense with these anyway. The SW plugin might, though.
if (gstate.isAntiAliasEnabled()) {
// Discard AA lines in DOA
if (prim == GE_PRIM_LINE_STRIP)
return;
// Discard AA lines in Summon Night 5
if ((prim == GE_PRIM_LINES) && gstate.isSkinningEnabled())
return;
}
// This also make skipping drawing very effective.
framebufferManager_.SetRenderFrameBuffer();
if (gstate_c.skipDrawReason & (SKIPDRAW_SKIPFRAME | SKIPDRAW_NON_DISPLAYED_FB)) {
transformDraw_.SetupVertexDecoder(gstate.vertType);
// Rough estimate, not sure what's correct.
int vertexCost = transformDraw_.EstimatePerVertexCost();
cyclesExecuted += vertexCost * count;
return;
}
if (!Memory::IsValidAddress(gstate_c.vertexAddr)) {
ERROR_LOG_REPORT(G3D, "Bad vertex address %08x!", gstate_c.vertexAddr);
return;
}
// TODO: Split this so that we can collect sequences of primitives, can greatly speed things up
// on platforms where draw calls are expensive like mobile and D3D
void *verts = Memory::GetPointerUnchecked(gstate_c.vertexAddr);
void *inds = 0;
if ((gstate.vertType & GE_VTYPE_IDX_MASK) != GE_VTYPE_IDX_NONE) {
if (!Memory::IsValidAddress(gstate_c.indexAddr)) {
ERROR_LOG_REPORT(G3D, "Bad index address %08x!", gstate_c.indexAddr);
return;
}
inds = Memory::GetPointerUnchecked(gstate_c.indexAddr);
}
#ifndef MOBILE_DEVICE
if (prim > GE_PRIM_RECTANGLES) {
ERROR_LOG_REPORT_ONCE(reportPrim, G3D, "Unexpected prim type: %d", prim);
}
#endif
int bytesRead;
transformDraw_.SubmitPrim(verts, inds, prim, count, gstate.vertType, &bytesRead);
int vertexCost = transformDraw_.EstimatePerVertexCost();
gpuStats.vertexGPUCycles += vertexCost * count;
cyclesExecuted += vertexCost * count;
// After drawing, we advance the vertexAddr (when non indexed) or indexAddr (when indexed).
// Some games rely on this, they don't bother reloading VADDR and IADDR.
// Q: Are these changed reflected in the real registers? Needs testing.
if (inds) {
int indexSize = 1;
if ((gstate.vertType & GE_VTYPE_IDX_MASK) == GE_VTYPE_IDX_16BIT)
indexSize = 2;
gstate_c.indexAddr += count * indexSize;
} else {
gstate_c.vertexAddr += bytesRead;
}
}
void GLES_GPU::Execute_VertexType(u32 op, u32 diff) {
if (!g_Config.bSoftwareSkinning) {
if (diff & (GE_VTYPE_TC_MASK | GE_VTYPE_THROUGH_MASK))
shaderManager_->DirtyUniform(DIRTY_UVSCALEOFFSET);
} else {
// Don't flush when weight count changes, unless morph is enabled.
if ((diff & ~GE_VTYPE_WEIGHTCOUNT_MASK) || (op & GE_VTYPE_MORPHCOUNT_MASK) != 0) {
// Restore and flush
gstate.vertType ^= diff;
Flush();
gstate.vertType ^= diff;
if (diff & (GE_VTYPE_TC_MASK | GE_VTYPE_THROUGH_MASK))
shaderManager_->DirtyUniform(DIRTY_UVSCALEOFFSET);
}
}
}
void GLES_GPU::Execute_Region(u32 op, u32 diff) {
gstate_c.framebufChanged = true;
gstate_c.textureChanged |= TEXCHANGE_PARAMSONLY;
}
void GLES_GPU::Execute_FramebufType(u32 op, u32 diff) {
gstate_c.framebufChanged = true;
gstate_c.textureChanged |= TEXCHANGE_PARAMSONLY;
}
void GLES_GPU::Execute_TexAddr0(u32 op, u32 diff) {
gstate_c.textureChanged = TEXCHANGE_UPDATED;
shaderManager_->DirtyUniform(DIRTY_UVSCALEOFFSET);
}
void GLES_GPU::Execute_LoadClut(u32 op, u32 diff) {
gstate_c.textureChanged |= TEXCHANGE_PARAMSONLY;
textureCache_.LoadClut();
// This could be used to "dirty" textures with clut.
}
void GLES_GPU::Execute_TexSize0(u32 op, u32 diff) {
// Render to texture may have overridden the width/height.
// Don't reset it unless the size is different / the texture has changed.
if (diff || gstate_c.textureChanged != TEXCHANGE_UNCHANGED) {
gstate_c.curTextureWidth = gstate.getTextureWidth(0);
gstate_c.curTextureHeight = gstate.getTextureHeight(0);
shaderManager_->DirtyUniform(DIRTY_UVSCALEOFFSET);
// We will need to reset the texture now.
gstate_c.textureChanged |= TEXCHANGE_PARAMSONLY;
}
}
void GLES_GPU::Execute_Ambient(u32 op, u32 diff) {
shaderManager_->DirtyUniform(DIRTY_AMBIENT);
}
void GLES_GPU::Execute_MaterialDiffuse(u32 op, u32 diff) {
shaderManager_->DirtyUniform(DIRTY_MATDIFFUSE);
}
void GLES_GPU::Execute_MaterialEmissive(u32 op, u32 diff) {
shaderManager_->DirtyUniform(DIRTY_MATEMISSIVE);
}
void GLES_GPU::Execute_MaterialAmbient(u32 op, u32 diff) {
shaderManager_->DirtyUniform(DIRTY_MATAMBIENTALPHA);
}
void GLES_GPU::Execute_MaterialSpecular(u32 op, u32 diff) {
shaderManager_->DirtyUniform(DIRTY_MATSPECULAR);
}
void GLES_GPU::ExecuteOpInternal(u32 op, u32 diff) {
u32 cmd = op >> 24;
u32 data = op & 0xFFFFFF;
// Handle control and drawing commands here directly. The others we delegate.
switch (cmd) {
case GE_CMD_BASE:
break;
case GE_CMD_VADDR:
Execute_Vaddr(op, diff);
break;
case GE_CMD_IADDR:
Execute_Iaddr(op, diff);
break;
case GE_CMD_PRIM:
Execute_Prim(op, diff);
break;
// The arrow and other rotary items in Puzbob are bezier patches, strangely enough.
case GE_CMD_BEZIER:
{
// This also make skipping drawing very effective.
framebufferManager_.SetRenderFrameBuffer();
if (gstate_c.skipDrawReason & (SKIPDRAW_SKIPFRAME | SKIPDRAW_NON_DISPLAYED_FB)) {
// TODO: Should this eat some cycles? Probably yes. Not sure if important.
return;
}
if (!Memory::IsValidAddress(gstate_c.vertexAddr)) {
ERROR_LOG_REPORT(G3D, "Bad vertex address %08x!", gstate_c.vertexAddr);
break;
}
void *control_points = Memory::GetPointerUnchecked(gstate_c.vertexAddr);
void *indices = NULL;
if ((gstate.vertType & GE_VTYPE_IDX_MASK) != GE_VTYPE_IDX_NONE) {
if (!Memory::IsValidAddress(gstate_c.indexAddr)) {
ERROR_LOG_REPORT(G3D, "Bad index address %08x!", gstate_c.indexAddr);
break;
}
indices = Memory::GetPointerUnchecked(gstate_c.indexAddr);
}
if (gstate.getPatchPrimitiveType() != GE_PATCHPRIM_TRIANGLES) {
ERROR_LOG_REPORT(G3D, "Unsupported patch primitive %x", gstate.getPatchPrimitiveType());
break;
}
if (gstate.vertType & GE_VTYPE_MORPHCOUNT_MASK) {
DEBUG_LOG_REPORT(G3D, "Bezier + morph: %i", (gstate.vertType & GE_VTYPE_MORPHCOUNT_MASK) >> GE_VTYPE_MORPHCOUNT_SHIFT);
}
if (vertTypeIsSkinningEnabled(gstate.vertType)) {
DEBUG_LOG_REPORT(G3D, "Bezier + skinning: %i", vertTypeGetNumBoneWeights(gstate.vertType));
}
GEPatchPrimType patchPrim = gstate.getPatchPrimitiveType();
int bz_ucount = data & 0xFF;
int bz_vcount = (data >> 8) & 0xFF;
transformDraw_.SubmitBezier(control_points, indices, bz_ucount, bz_vcount, patchPrim, gstate.vertType);
}
break;
case GE_CMD_SPLINE:
{
// This also make skipping drawing very effective.
framebufferManager_.SetRenderFrameBuffer();
if (gstate_c.skipDrawReason & (SKIPDRAW_SKIPFRAME | SKIPDRAW_NON_DISPLAYED_FB)) {
// TODO: Should this eat some cycles? Probably yes. Not sure if important.
return;
}
if (!Memory::IsValidAddress(gstate_c.vertexAddr)) {
ERROR_LOG_REPORT(G3D, "Bad vertex address %08x!", gstate_c.vertexAddr);
break;
}
void *control_points = Memory::GetPointerUnchecked(gstate_c.vertexAddr);
void *indices = NULL;
if ((gstate.vertType & GE_VTYPE_IDX_MASK) != GE_VTYPE_IDX_NONE) {
if (!Memory::IsValidAddress(gstate_c.indexAddr)) {
ERROR_LOG_REPORT(G3D, "Bad index address %08x!", gstate_c.indexAddr);
break;
}
indices = Memory::GetPointerUnchecked(gstate_c.indexAddr);
}
if (gstate.getPatchPrimitiveType() != GE_PATCHPRIM_TRIANGLES) {
ERROR_LOG_REPORT(G3D, "Unsupported patch primitive %x", gstate.getPatchPrimitiveType());
break;
}
if (gstate.vertType & GE_VTYPE_MORPHCOUNT_MASK) {
DEBUG_LOG_REPORT(G3D, "Spline + morph: %i", (gstate.vertType & GE_VTYPE_MORPHCOUNT_MASK) >> GE_VTYPE_MORPHCOUNT_SHIFT);
}
if (vertTypeIsSkinningEnabled(gstate.vertType)) {
DEBUG_LOG_REPORT(G3D, "Spline + skinning: %i", vertTypeGetNumBoneWeights(gstate.vertType));
}
int sp_ucount = data & 0xFF;
int sp_vcount = (data >> 8) & 0xFF;
int sp_utype = (data >> 16) & 0x3;
int sp_vtype = (data >> 18) & 0x3;
GEPatchPrimType patchPrim = gstate.getPatchPrimitiveType();
transformDraw_.SubmitSpline(control_points, indices, sp_ucount, sp_vcount, sp_utype, sp_vtype, patchPrim, gstate.vertType);
}
break;
case GE_CMD_BOUNDINGBOX:
// Just resetting, nothing to bound.
if (data == 0) {
// TODO: Should this set the bboxResult? Let's set it true for now.
currentList->bboxResult = true;
break;
}
if (((data & 7) == 0) && data <= 64) { // Sanity check
void *control_points = Memory::GetPointer(gstate_c.vertexAddr);
if (gstate.vertType & GE_VTYPE_IDX_MASK) {
ERROR_LOG_REPORT_ONCE(boundingbox, G3D, "Indexed bounding box data not supported.");
// Data seems invalid. Let's assume the box test passed.
currentList->bboxResult = true;
break;
}
// Test if the bounding box is within the drawing region.
currentList->bboxResult = transformDraw_.TestBoundingBox(control_points, data, gstate.vertType);
} else {
ERROR_LOG_REPORT_ONCE(boundingbox, G3D, "Bad bounding box data: %06x", data);
// Data seems invalid. Let's assume the box test passed.
currentList->bboxResult = true;
}
break;
case GE_CMD_VERTEXTYPE:
Execute_VertexType(op, diff);
break;
case GE_CMD_REGION1:
case GE_CMD_REGION2:
Execute_Region(op, diff);
break;
case GE_CMD_CLIPENABLE:
//we always clip, this is opengl
break;
case GE_CMD_CULLFACEENABLE:
case GE_CMD_CULL:
break;
case GE_CMD_TEXTUREMAPENABLE:
// Don't need to dirty the texture here, already dirtied at list start/etc.
break;
case GE_CMD_LIGHTINGENABLE:
break;