/
SDL_blit_A.c
2268 lines (2044 loc) · 73.5 KB
/
SDL_blit_A.c
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
/*
SDL - Simple DirectMedia Layer
Copyright (C) 1997-2006 Sam Lantinga
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
Sam Lantinga
slouken@libsdl.org
*/
#include "SDL_config.h"
#include "SDL_video.h"
#include "SDL_blit.h"
#if defined(__GNUC__) && (defined(__i386__) || defined(__x86_64__)) && SDL_ASSEMBLY_ROUTINES
#define MMX_ASMBLIT 1
#endif
/* Function to check the CPU flags */
#include "SDL_cpuinfo.h"
#if MMX_ASMBLIT
#include "mmx.h"
#endif
/* Functions to perform alpha blended blitting */
/* N->1 blending with per-surface alpha */
static void BlitNto1SurfaceAlpha(SDL_BlitInfo *info)
{
int width = info->d_width;
int height = info->d_height;
Uint8 *src = info->s_pixels;
int srcskip = info->s_skip;
Uint8 *dst = info->d_pixels;
int dstskip = info->d_skip;
Uint8 *palmap = info->table;
SDL_PixelFormat *srcfmt = info->src;
SDL_PixelFormat *dstfmt = info->dst;
int srcbpp = srcfmt->BytesPerPixel;
const unsigned A = srcfmt->alpha;
while ( height-- ) {
DUFFS_LOOP4(
{
Uint32 Pixel;
unsigned sR;
unsigned sG;
unsigned sB;
unsigned dR;
unsigned dG;
unsigned dB;
DISEMBLE_RGB(src, srcbpp, srcfmt, Pixel, sR, sG, sB);
dR = dstfmt->palette->colors[*dst].r;
dG = dstfmt->palette->colors[*dst].g;
dB = dstfmt->palette->colors[*dst].b;
ALPHA_BLEND(sR, sG, sB, A, dR, dG, dB);
dR &= 0xff;
dG &= 0xff;
dB &= 0xff;
/* Pack RGB into 8bit pixel */
if ( palmap == NULL ) {
*dst =((dR>>5)<<(3+2))|
((dG>>5)<<(2))|
((dB>>6)<<(0));
} else {
*dst = palmap[((dR>>5)<<(3+2))|
((dG>>5)<<(2)) |
((dB>>6)<<(0))];
}
dst++;
src += srcbpp;
},
width);
src += srcskip;
dst += dstskip;
}
}
/* N->1 blending with pixel alpha */
static void BlitNto1PixelAlpha(SDL_BlitInfo *info)
{
int width = info->d_width;
int height = info->d_height;
Uint8 *src = info->s_pixels;
int srcskip = info->s_skip;
Uint8 *dst = info->d_pixels;
int dstskip = info->d_skip;
Uint8 *palmap = info->table;
SDL_PixelFormat *srcfmt = info->src;
SDL_PixelFormat *dstfmt = info->dst;
int srcbpp = srcfmt->BytesPerPixel;
/* FIXME: fix alpha bit field expansion here too? */
while ( height-- ) {
DUFFS_LOOP4(
{
Uint32 Pixel;
unsigned sR;
unsigned sG;
unsigned sB;
unsigned sA;
unsigned dR;
unsigned dG;
unsigned dB;
DISEMBLE_RGBA(src,srcbpp,srcfmt,Pixel,sR,sG,sB,sA);
dR = dstfmt->palette->colors[*dst].r;
dG = dstfmt->palette->colors[*dst].g;
dB = dstfmt->palette->colors[*dst].b;
ALPHA_BLEND(sR, sG, sB, sA, dR, dG, dB);
dR &= 0xff;
dG &= 0xff;
dB &= 0xff;
/* Pack RGB into 8bit pixel */
if ( palmap == NULL ) {
*dst =((dR>>5)<<(3+2))|
((dG>>5)<<(2))|
((dB>>6)<<(0));
} else {
*dst = palmap[((dR>>5)<<(3+2))|
((dG>>5)<<(2)) |
((dB>>6)<<(0)) ];
}
dst++;
src += srcbpp;
},
width);
src += srcskip;
dst += dstskip;
}
}
/* colorkeyed N->1 blending with per-surface alpha */
static void BlitNto1SurfaceAlphaKey(SDL_BlitInfo *info)
{
int width = info->d_width;
int height = info->d_height;
Uint8 *src = info->s_pixels;
int srcskip = info->s_skip;
Uint8 *dst = info->d_pixels;
int dstskip = info->d_skip;
Uint8 *palmap = info->table;
SDL_PixelFormat *srcfmt = info->src;
SDL_PixelFormat *dstfmt = info->dst;
int srcbpp = srcfmt->BytesPerPixel;
Uint32 ckey = srcfmt->colorkey;
const int A = srcfmt->alpha;
while ( height-- ) {
DUFFS_LOOP(
{
Uint32 Pixel;
unsigned sR;
unsigned sG;
unsigned sB;
unsigned dR;
unsigned dG;
unsigned dB;
DISEMBLE_RGB(src, srcbpp, srcfmt, Pixel, sR, sG, sB);
if ( Pixel != ckey ) {
dR = dstfmt->palette->colors[*dst].r;
dG = dstfmt->palette->colors[*dst].g;
dB = dstfmt->palette->colors[*dst].b;
ALPHA_BLEND(sR, sG, sB, A, dR, dG, dB);
dR &= 0xff;
dG &= 0xff;
dB &= 0xff;
/* Pack RGB into 8bit pixel */
if ( palmap == NULL ) {
*dst =((dR>>5)<<(3+2))|
((dG>>5)<<(2)) |
((dB>>6)<<(0));
} else {
*dst = palmap[((dR>>5)<<(3+2))|
((dG>>5)<<(2)) |
((dB>>6)<<(0)) ];
}
}
dst++;
src += srcbpp;
},
width);
src += srcskip;
dst += dstskip;
}
}
#if MMX_ASMBLIT
/* fast RGB888->(A)RGB888 blending with surface alpha=128 special case */
static void BlitRGBtoRGBSurfaceAlpha128MMX(SDL_BlitInfo *info)
{
int width = info->d_width;
int height = info->d_height;
Uint32 *srcp = (Uint32 *)info->s_pixels;
int srcskip = info->s_skip >> 2;
Uint32 *dstp = (Uint32 *)info->d_pixels;
int dstskip = info->d_skip >> 2;
Uint8 load[8];
*(Uint64 *)load = 0x00fefefe00fefefeULL;/* alpha128 mask */
movq_m2r(*load, mm4); /* alpha128 mask -> mm4 */
*(Uint64 *)load = 0x0001010100010101ULL;/* !alpha128 mask */
movq_m2r(*load, mm3); /* !alpha128 mask -> mm3 */
*(Uint64 *)load = 0xFF000000FF000000ULL;/* dst alpha mask */
movq_m2r(*load, mm7); /* dst alpha mask -> mm7 */
while(height--) {
DUFFS_LOOP_DOUBLE2(
{
Uint32 s = *srcp++;
Uint32 d = *dstp;
*dstp++ = ((((s & 0x00fefefe) + (d & 0x00fefefe)) >> 1)
+ (s & d & 0x00010101)) | 0xff000000;
},{
movq_m2r((*dstp), mm2);/* 2 x dst -> mm2(ARGBARGB) */
movq_r2r(mm2, mm6); /* 2 x dst -> mm6(ARGBARGB) */
movq_m2r((*srcp), mm1);/* 2 x src -> mm1(ARGBARGB) */
movq_r2r(mm1, mm5); /* 2 x src -> mm5(ARGBARGB) */
pand_r2r(mm4, mm6); /* dst & mask -> mm6 */
pand_r2r(mm4, mm5); /* src & mask -> mm5 */
paddd_r2r(mm6, mm5); /* mm6 + mm5 -> mm5 */
psrld_i2r(1, mm5); /* mm5 >> 1 -> mm5 */
pand_r2r(mm1, mm2); /* src & dst -> mm2 */
pand_r2r(mm3, mm2); /* mm2 & !mask -> mm2 */
paddd_r2r(mm5, mm2); /* mm5 + mm2 -> mm2 */
por_r2r(mm7, mm2); /* mm7(full alpha) | mm2 -> mm2 */
movq_r2m(mm2, (*dstp));/* mm2 -> 2 x dst pixels */
dstp += 2;
srcp += 2;
}, width);
srcp += srcskip;
dstp += dstskip;
}
emms();
}
/* fast RGB888->(A)RGB888 blending with surface alpha */
static void BlitRGBtoRGBSurfaceAlphaMMX(SDL_BlitInfo *info)
{
unsigned alpha = info->src->alpha;
if(alpha == 128) {
BlitRGBtoRGBSurfaceAlpha128MMX(info);
} else {
int width = info->d_width;
int height = info->d_height;
Uint32 *srcp = (Uint32 *)info->s_pixels;
int srcskip = info->s_skip >> 2;
Uint32 *dstp = (Uint32 *)info->d_pixels;
int dstskip = info->d_skip >> 2;
Uint8 load[8] = {alpha, alpha, alpha, alpha,
alpha, alpha, alpha, alpha};
movq_m2r(*load, mm4); /* alpha -> mm4 */
*(Uint64 *)load = 0x00FF00FF00FF00FFULL;
movq_m2r(*load, mm3); /* mask -> mm3 */
pand_r2r(mm3, mm4); /* mm4 & mask -> 0A0A0A0A -> mm4 */
*(Uint64 *)load = 0xFF000000FF000000ULL;/* dst alpha mask */
movq_m2r(*load, mm7); /* dst alpha mask -> mm7 */
while(height--) {
DUFFS_LOOP_DOUBLE2({
/* One Pixel Blend */
movd_m2r((*srcp), mm1);/* src(ARGB) -> mm1 (0000ARGB)*/
punpcklbw_r2r(mm1, mm1); /* AARRGGBB -> mm1 */
pand_r2r(mm3, mm1); /* 0A0R0G0B -> mm1 */
movd_m2r((*dstp), mm2);/* dst(ARGB) -> mm2 (0000ARGB)*/
movq_r2r(mm2, mm6);/* dst(ARGB) -> mm6 (0000ARGB)*/
punpcklbw_r2r(mm2, mm2); /* AARRGGBB -> mm2 */
pand_r2r(mm3, mm2); /* 0A0R0G0B -> mm2 */
psubw_r2r(mm2, mm1);/* src - dst -> mm1 */
pmullw_r2r(mm4, mm1); /* mm1 * alpha -> mm1 */
psrlw_i2r(8, mm1); /* mm1 >> 8 -> mm1 */
paddw_r2r(mm1, mm2); /* mm1 + mm2(dst) -> mm2 */
pand_r2r(mm3, mm2); /* 0A0R0G0B -> mm2 */
packuswb_r2r(mm2, mm2); /* ARGBARGB -> mm2 */
por_r2r(mm7, mm2); /* mm7(full alpha) | mm2 -> mm2 */
movd_r2m(mm2, *dstp);/* mm2 -> Pixel */
++srcp;
++dstp;
},{
/* Two Pixels Blend */
movq_m2r((*srcp), mm0);/* 2 x src -> mm0(ARGBARGB)*/
movq_r2r(mm0, mm1); /* 2 x src -> mm1(ARGBARGB) */
punpcklbw_r2r(mm0, mm0); /* low - AARRGGBB -> mm0 */
pand_r2r(mm3, mm0); /* 0A0R0G0B -> mm0(src1) */
punpckhbw_r2r(mm1, mm1); /* high - AARRGGBB -> mm1 */
pand_r2r(mm3, mm1); /* 0A0R0G0B -> mm1(src2) */
movq_m2r((*dstp), mm2);/* 2 x dst -> mm2(ARGBARGB) */
movq_r2r(mm2, mm5); /* 2 x dst -> mm5(ARGBARGB) */
movq_r2r(mm2, mm6); /* 2 x dst -> mm6(ARGBARGB) */
punpcklbw_r2r(mm2, mm2); /* low - AARRGGBB -> mm2 */
punpckhbw_r2r(mm6, mm6); /* high - AARRGGBB -> mm6 */
pand_r2r(mm3, mm2); /* 0A0R0G0B -> mm2(dst1) */
psubw_r2r(mm2, mm0);/* src1 - dst1 -> mm0 */
pmullw_r2r(mm4, mm0); /* mm0 * alpha -> mm0 */
pand_r2r(mm3, mm6); /* 0A0R0G0B -> mm6(dst2) */
psrlw_i2r(8, mm0); /* mm0 >> 8 -> mm1 */
psubw_r2r(mm6, mm1);/* src2 - dst2 -> mm1 */
pmullw_r2r(mm4, mm1); /* mm1 * alpha -> mm1 */
paddw_r2r(mm0, mm2); /* mm0 + mm2(dst1) -> mm2 */
psrlw_i2r(8, mm1); /* mm1 >> 8 -> mm0 */
pand_r2r(mm3, mm2); /* 0A0R0G0B -> mm2 */
paddw_r2r(mm1, mm6); /* mm1 + mm6(dst2) -> mm6 */
pand_r2r(mm3, mm6); /* 0A0R0G0B -> mm6 */
packuswb_r2r(mm2, mm2); /* ARGBARGB -> mm2 */
packuswb_r2r(mm6, mm6); /* ARGBARGB -> mm6 */
psrlq_i2r(32, mm2); /* mm2 >> 32 -> mm2 */
psllq_i2r(32, mm6); /* mm6 << 32 -> mm6 */
por_r2r(mm6, mm2); /* mm6 | mm2 -> mm2 */
por_r2r(mm7, mm2); /* mm7(full alpha) | mm2 -> mm2 */
movq_r2m(mm2, *dstp);/* mm2 -> 2 x Pixel */
srcp += 2;
dstp += 2;
}, width);
srcp += srcskip;
dstp += dstskip;
}
emms();
}
}
/* fast ARGB888->(A)RGB888 blending with pixel alpha */
static void BlitRGBtoRGBPixelAlphaMMX(SDL_BlitInfo *info)
{
int width = info->d_width;
int height = info->d_height;
Uint32 *srcp = (Uint32 *)info->s_pixels;
int srcskip = info->s_skip >> 2;
Uint32 *dstp = (Uint32 *)info->d_pixels;
int dstskip = info->d_skip >> 2;
Uint32 alpha = 0;
Uint8 load[8];
*(Uint64 *)load = 0x00FF00FF00FF00FFULL;
movq_m2r(*load, mm3); /* mask -> mm2 */
*(Uint64 *)load = 0x00FF000000000000ULL;
movq_m2r(*load, mm7); /* dst alpha mask -> mm2 */
*(Uint64 *)load = 0x00FFFFFF00FFFFFFULL;
movq_m2r(*load, mm0); /* alpha 255 mask -> mm0 */
*(Uint64 *)load = 0xFF000000FF000000ULL;
movq_m2r(*load, mm6); /* alpha 255 !mask -> mm6 */
while(height--) {
DUFFS_LOOP4({
alpha = *srcp;
alpha >>= 24;
/* FIXME: Here we special-case opaque alpha since the
compositioning used (>>8 instead of /255) doesn't handle
it correctly. Also special-case alpha=0 for speed?
Benchmark this! */
if(alpha) {
if(alpha == SDL_ALPHA_OPAQUE) {
movd_m2r((*srcp), mm1);/* src(ARGB) -> mm1 (0000ARGB)*/
movd_m2r((*dstp), mm2);/* dst(ARGB) -> mm2 (0000ARGB)*/
pand_r2r(mm0, mm1);
pand_r2r(mm6, mm2);
por_r2r(mm1, mm2);
movd_r2m(mm2, (*dstp));
} else {
movd_m2r((*srcp), mm1);/* src(ARGB) -> mm1 (0000ARGB)*/
punpcklbw_r2r(mm1, mm1); /* AARRGGBB -> mm1 */
pand_r2r(mm3, mm1); /* 0A0R0G0B -> mm1 */
movd_m2r((*dstp), mm2);/* dst(ARGB) -> mm2 (0000ARGB)*/
punpcklbw_r2r(mm2, mm2); /* AARRGGBB -> mm2 */
pand_r2r(mm3, mm2); /* 0A0R0G0B -> mm2 */
movq_r2r(mm2, mm5);/* mm2(0A0R0G0B) -> mm5 */
pand_r2r(mm7, mm5); /* mm5 & dst alpha mask -> mm5(0A000000) */
psrlq_i2r(24, mm5); /* mm5 >> 24 -> mm5 (0000A000)*/
movq_r2r(mm1, mm4);/* mm1(0A0R0G0B) -> mm4 */
psrlq_i2r(48, mm4); /* mm4 >> 48 -> mm4(0000000A) */
punpcklwd_r2r(mm4, mm4); /* 00000A0A -> mm4 */
punpcklwd_r2r(mm4, mm4); /* 0A0A0A0A -> mm4 */
/* blend */
psubw_r2r(mm2, mm1);/* src - dst -> mm1 */
pmullw_r2r(mm4, mm1); /* mm1 * alpha -> mm1 */
psrlw_i2r(8, mm1); /* mm1 >> 8 -> mm1 */
paddw_r2r(mm1, mm2); /* mm1 + mm2(dst) -> mm2 */
pand_r2r(mm3, mm2); /* 0A0R0G0B -> mm2 */
packuswb_r2r(mm2, mm2); /* ARGBARGB -> mm2 */
pand_r2r(mm0, mm2); /* 0RGB0RGB -> mm2 */
por_r2r(mm5, mm2); /* dst alpha | mm2 -> mm2 */
movd_r2m(mm2, *dstp);/* mm2 -> dst */
}
}
++srcp;
++dstp;
}, width);
srcp += srcskip;
dstp += dstskip;
}
emms();
}
#endif
#if SDL_ALTIVEC_BLITTERS
#if HAVE_ALTIVEC_H
#include <altivec.h>
#endif
#include <assert.h>
#if (defined(__MACOSX__) && (__GNUC__ < 4))
#define VECUINT8_LITERAL(a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p) \
(vector unsigned char) ( a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p )
#define VECUINT16_LITERAL(a,b,c,d,e,f,g,h) \
(vector unsigned short) ( a,b,c,d,e,f,g,h )
#else
#define VECUINT8_LITERAL(a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p) \
(vector unsigned char) { a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p }
#define VECUINT16_LITERAL(a,b,c,d,e,f,g,h) \
(vector unsigned short) { a,b,c,d,e,f,g,h }
#endif
#define UNALIGNED_PTR(x) (((size_t) x) & 0x0000000F)
#define VECPRINT(msg, v) do { \
vector unsigned int tmpvec = (vector unsigned int)(v); \
unsigned int *vp = (unsigned int *)&tmpvec; \
printf("%s = %08X %08X %08X %08X\n", msg, vp[0], vp[1], vp[2], vp[3]); \
} while (0)
/* the permuation vector that takes the high bytes out of all the appropriate shorts
(vector unsigned char)(
0x00, 0x10, 0x02, 0x12,
0x04, 0x14, 0x06, 0x16,
0x08, 0x18, 0x0A, 0x1A,
0x0C, 0x1C, 0x0E, 0x1E );
*/
#define VEC_MERGE_PERMUTE() (vec_add(vec_lvsl(0, (int*)NULL), (vector unsigned char)vec_splat_u16(0x0F)))
#define VEC_U32_24() (vec_add(vec_splat_u32(12), vec_splat_u32(12)))
#define VEC_ALPHA_MASK() ((vector unsigned char)vec_sl((vector unsigned int)vec_splat_s8(-1), VEC_U32_24()))
#define VEC_ALIGNER(src) ((UNALIGNED_PTR(src)) \
? vec_lvsl(0, src) \
: vec_add(vec_lvsl(8, src), vec_splat_u8(8)))
#define VEC_MULTIPLY_ALPHA(vs, vd, valpha, mergePermute, v1_16, v8_16) do { \
/* vtemp1 contains source AAGGAAGGAAGGAAGG */ \
vector unsigned short vtemp1 = vec_mule(vs, valpha); \
/* vtemp2 contains source RRBBRRBBRRBBRRBB */ \
vector unsigned short vtemp2 = vec_mulo(vs, valpha); \
/* valpha2 is 255-alpha */ \
vector unsigned char valpha2 = vec_nor(valpha, valpha); \
/* vtemp3 contains dest AAGGAAGGAAGGAAGG */ \
vector unsigned short vtemp3 = vec_mule(vd, valpha2); \
/* vtemp4 contains dest RRBBRRBBRRBBRRBB */ \
vector unsigned short vtemp4 = vec_mulo(vd, valpha2); \
/* add source and dest */ \
vtemp1 = vec_add(vtemp1, vtemp3); \
vtemp2 = vec_add(vtemp2, vtemp4); \
/* vtemp1 = (vtemp1 + 1) + ((vtemp1 + 1) >> 8) */ \
vtemp1 = vec_add(vtemp1, v1_16); \
vtemp3 = vec_sr(vtemp1, v8_16); \
vtemp1 = vec_add(vtemp1, vtemp3); \
/* vtemp2 = (vtemp2 + 1) + ((vtemp2 + 1) >> 8) */ \
vtemp2 = vec_add(vtemp2, v1_16); \
vtemp4 = vec_sr(vtemp2, v8_16); \
vtemp2 = vec_add(vtemp2, vtemp4); \
/* (>>8) and get ARGBARGBARGBARGB */ \
vd = (vector unsigned char)vec_perm(vtemp1, vtemp2, mergePermute); \
} while (0)
/* Calculate the permute vector used for 32->32 swizzling */
static vector unsigned char calc_swizzle32(const SDL_PixelFormat *srcfmt,
const SDL_PixelFormat *dstfmt)
{
/*
* We have to assume that the bits that aren't used by other
* colors is alpha, and it's one complete byte, since some formats
* leave alpha with a zero mask, but we should still swizzle the bits.
*/
/* ARGB */
const static struct SDL_PixelFormat default_pixel_format = {
NULL, 0, 0,
0, 0, 0, 0,
16, 8, 0, 24,
0x00FF0000, 0x0000FF00, 0x000000FF, 0xFF000000,
0, 0};
if (!srcfmt) {
srcfmt = &default_pixel_format;
}
if (!dstfmt) {
dstfmt = &default_pixel_format;
}
vector unsigned char plus = VECUINT8_LITERAL
( 0x00, 0x00, 0x00, 0x00,
0x04, 0x04, 0x04, 0x04,
0x08, 0x08, 0x08, 0x08,
0x0C, 0x0C, 0x0C, 0x0C );
vector unsigned char vswiz;
vector unsigned int srcvec;
#define RESHIFT(X) (3 - ((X) >> 3))
Uint32 rmask = RESHIFT(srcfmt->Rshift) << (dstfmt->Rshift);
Uint32 gmask = RESHIFT(srcfmt->Gshift) << (dstfmt->Gshift);
Uint32 bmask = RESHIFT(srcfmt->Bshift) << (dstfmt->Bshift);
Uint32 amask;
/* Use zero for alpha if either surface doesn't have alpha */
if (dstfmt->Amask) {
amask = ((srcfmt->Amask) ? RESHIFT(srcfmt->Ashift) : 0x10) << (dstfmt->Ashift);
} else {
amask = 0x10101010 & ((dstfmt->Rmask | dstfmt->Gmask | dstfmt->Bmask) ^ 0xFFFFFFFF);
}
#undef RESHIFT
((unsigned int *)(char*)&srcvec)[0] = (rmask | gmask | bmask | amask);
vswiz = vec_add(plus, (vector unsigned char)vec_splat(srcvec, 0));
return(vswiz);
}
static void Blit32to565PixelAlphaAltivec(SDL_BlitInfo *info)
{
int height = info->d_height;
Uint8 *src = (Uint8 *)info->s_pixels;
int srcskip = info->s_skip;
Uint8 *dst = (Uint8 *)info->d_pixels;
int dstskip = info->d_skip;
SDL_PixelFormat *srcfmt = info->src;
vector unsigned char v0 = vec_splat_u8(0);
vector unsigned short v8_16 = vec_splat_u16(8);
vector unsigned short v1_16 = vec_splat_u16(1);
vector unsigned short v2_16 = vec_splat_u16(2);
vector unsigned short v3_16 = vec_splat_u16(3);
vector unsigned int v8_32 = vec_splat_u32(8);
vector unsigned int v16_32 = vec_add(v8_32, v8_32);
vector unsigned short v3f = VECUINT16_LITERAL(
0x003f, 0x003f, 0x003f, 0x003f,
0x003f, 0x003f, 0x003f, 0x003f);
vector unsigned short vfc = VECUINT16_LITERAL(
0x00fc, 0x00fc, 0x00fc, 0x00fc,
0x00fc, 0x00fc, 0x00fc, 0x00fc);
/*
0x10 - 0x1f is the alpha
0x00 - 0x0e evens are the red
0x01 - 0x0f odds are zero
*/
vector unsigned char vredalpha1 = VECUINT8_LITERAL(
0x10, 0x00, 0x01, 0x01,
0x10, 0x02, 0x01, 0x01,
0x10, 0x04, 0x01, 0x01,
0x10, 0x06, 0x01, 0x01
);
vector unsigned char vredalpha2 = (vector unsigned char)(
vec_add((vector unsigned int)vredalpha1, vec_sl(v8_32, v16_32))
);
/*
0x00 - 0x0f is ARxx ARxx ARxx ARxx
0x11 - 0x0f odds are blue
*/
vector unsigned char vblue1 = VECUINT8_LITERAL(
0x00, 0x01, 0x02, 0x11,
0x04, 0x05, 0x06, 0x13,
0x08, 0x09, 0x0a, 0x15,
0x0c, 0x0d, 0x0e, 0x17
);
vector unsigned char vblue2 = (vector unsigned char)(
vec_add((vector unsigned int)vblue1, v8_32)
);
/*
0x00 - 0x0f is ARxB ARxB ARxB ARxB
0x10 - 0x0e evens are green
*/
vector unsigned char vgreen1 = VECUINT8_LITERAL(
0x00, 0x01, 0x10, 0x03,
0x04, 0x05, 0x12, 0x07,
0x08, 0x09, 0x14, 0x0b,
0x0c, 0x0d, 0x16, 0x0f
);
vector unsigned char vgreen2 = (vector unsigned char)(
vec_add((vector unsigned int)vgreen1, vec_sl(v8_32, v8_32))
);
vector unsigned char vgmerge = VECUINT8_LITERAL(
0x00, 0x02, 0x00, 0x06,
0x00, 0x0a, 0x00, 0x0e,
0x00, 0x12, 0x00, 0x16,
0x00, 0x1a, 0x00, 0x1e);
vector unsigned char mergePermute = VEC_MERGE_PERMUTE();
vector unsigned char vpermute = calc_swizzle32(srcfmt, NULL);
vector unsigned char valphaPermute = vec_and(vec_lvsl(0, (int *)NULL), vec_splat_u8(0xC));
vector unsigned short vf800 = (vector unsigned short)vec_splat_u8(-7);
vf800 = vec_sl(vf800, vec_splat_u16(8));
while(height--) {
int extrawidth;
vector unsigned char valigner;
vector unsigned char vsrc;
vector unsigned char voverflow;
int width = info->d_width;
#define ONE_PIXEL_BLEND(condition, widthvar) \
while (condition) { \
Uint32 Pixel; \
unsigned sR, sG, sB, dR, dG, dB, sA; \
DISEMBLE_RGBA(src, 4, srcfmt, Pixel, sR, sG, sB, sA); \
if(sA) { \
unsigned short dstpixel = *((unsigned short *)dst); \
dR = (dstpixel >> 8) & 0xf8; \
dG = (dstpixel >> 3) & 0xfc; \
dB = (dstpixel << 3) & 0xf8; \
ACCURATE_ALPHA_BLEND(sR, sG, sB, sA, dR, dG, dB); \
*((unsigned short *)dst) = ( \
((dR & 0xf8) << 8) | ((dG & 0xfc) << 3) | (dB >> 3) \
); \
} \
src += 4; \
dst += 2; \
widthvar--; \
}
ONE_PIXEL_BLEND((UNALIGNED_PTR(dst)) && (width), width);
extrawidth = (width % 8);
valigner = VEC_ALIGNER(src);
vsrc = (vector unsigned char)vec_ld(0, src);
width -= extrawidth;
while (width) {
vector unsigned char valpha;
vector unsigned char vsrc1, vsrc2;
vector unsigned char vdst1, vdst2;
vector unsigned short vR, vG, vB;
vector unsigned short vpixel, vrpixel, vgpixel, vbpixel;
/* Load 8 pixels from src as ARGB */
voverflow = (vector unsigned char)vec_ld(15, src);
vsrc = vec_perm(vsrc, voverflow, valigner);
vsrc1 = vec_perm(vsrc, vsrc, vpermute);
src += 16;
vsrc = (vector unsigned char)vec_ld(15, src);
voverflow = vec_perm(voverflow, vsrc, valigner);
vsrc2 = vec_perm(voverflow, voverflow, vpermute);
src += 16;
/* Load 8 pixels from dst as XRGB */
voverflow = vec_ld(0, dst);
vR = vec_and((vector unsigned short)voverflow, vf800);
vB = vec_sl((vector unsigned short)voverflow, v3_16);
vG = vec_sl(vB, v2_16);
vdst1 = (vector unsigned char)vec_perm((vector unsigned char)vR, (vector unsigned char)vR, vredalpha1);
vdst1 = vec_perm(vdst1, (vector unsigned char)vB, vblue1);
vdst1 = vec_perm(vdst1, (vector unsigned char)vG, vgreen1);
vdst2 = (vector unsigned char)vec_perm((vector unsigned char)vR, (vector unsigned char)vR, vredalpha2);
vdst2 = vec_perm(vdst2, (vector unsigned char)vB, vblue2);
vdst2 = vec_perm(vdst2, (vector unsigned char)vG, vgreen2);
/* Alpha blend 8 pixels as ARGB */
valpha = vec_perm(vsrc1, v0, valphaPermute);
VEC_MULTIPLY_ALPHA(vsrc1, vdst1, valpha, mergePermute, v1_16, v8_16);
valpha = vec_perm(vsrc2, v0, valphaPermute);
VEC_MULTIPLY_ALPHA(vsrc2, vdst2, valpha, mergePermute, v1_16, v8_16);
/* Convert 8 pixels to 565 */
vpixel = (vector unsigned short)vec_packpx((vector unsigned int)vdst1, (vector unsigned int)vdst2);
vgpixel = (vector unsigned short)vec_perm(vdst1, vdst2, vgmerge);
vgpixel = vec_and(vgpixel, vfc);
vgpixel = vec_sl(vgpixel, v3_16);
vrpixel = vec_sl(vpixel, v1_16);
vrpixel = vec_and(vrpixel, vf800);
vbpixel = vec_and(vpixel, v3f);
vdst1 = vec_or((vector unsigned char)vrpixel, (vector unsigned char)vgpixel);
vdst1 = vec_or(vdst1, (vector unsigned char)vbpixel);
/* Store 8 pixels */
vec_st(vdst1, 0, dst);
width -= 8;
dst += 16;
}
ONE_PIXEL_BLEND((extrawidth), extrawidth);
#undef ONE_PIXEL_BLEND
src += srcskip;
dst += dstskip;
}
}
static void Blit32to32SurfaceAlphaKeyAltivec(SDL_BlitInfo *info)
{
unsigned alpha = info->src->alpha;
int height = info->d_height;
Uint32 *srcp = (Uint32 *)info->s_pixels;
int srcskip = info->s_skip >> 2;
Uint32 *dstp = (Uint32 *)info->d_pixels;
int dstskip = info->d_skip >> 2;
SDL_PixelFormat *srcfmt = info->src;
SDL_PixelFormat *dstfmt = info->dst;
unsigned sA = srcfmt->alpha;
unsigned dA = dstfmt->Amask ? SDL_ALPHA_OPAQUE : 0;
Uint32 rgbmask = srcfmt->Rmask | srcfmt->Gmask | srcfmt->Bmask;
Uint32 ckey = info->src->colorkey;
vector unsigned char mergePermute;
vector unsigned char vsrcPermute;
vector unsigned char vdstPermute;
vector unsigned char vsdstPermute;
vector unsigned char valpha;
vector unsigned char valphamask;
vector unsigned char vbits;
vector unsigned char v0;
vector unsigned short v1;
vector unsigned short v8;
vector unsigned int vckey;
vector unsigned int vrgbmask;
mergePermute = VEC_MERGE_PERMUTE();
v0 = vec_splat_u8(0);
v1 = vec_splat_u16(1);
v8 = vec_splat_u16(8);
/* set the alpha to 255 on the destination surf */
valphamask = VEC_ALPHA_MASK();
vsrcPermute = calc_swizzle32(srcfmt, NULL);
vdstPermute = calc_swizzle32(NULL, dstfmt);
vsdstPermute = calc_swizzle32(dstfmt, NULL);
/* set a vector full of alpha and 255-alpha */
((unsigned char *)&valpha)[0] = alpha;
valpha = vec_splat(valpha, 0);
vbits = (vector unsigned char)vec_splat_s8(-1);
ckey &= rgbmask;
((unsigned int *)(char*)&vckey)[0] = ckey;
vckey = vec_splat(vckey, 0);
((unsigned int *)(char*)&vrgbmask)[0] = rgbmask;
vrgbmask = vec_splat(vrgbmask, 0);
while(height--) {
int width = info->d_width;
#define ONE_PIXEL_BLEND(condition, widthvar) \
while (condition) { \
Uint32 Pixel; \
unsigned sR, sG, sB, dR, dG, dB; \
RETRIEVE_RGB_PIXEL(((Uint8 *)srcp), 4, Pixel); \
if(sA && Pixel != ckey) { \
RGB_FROM_PIXEL(Pixel, srcfmt, sR, sG, sB); \
DISEMBLE_RGB(((Uint8 *)dstp), 4, dstfmt, Pixel, dR, dG, dB); \
ACCURATE_ALPHA_BLEND(sR, sG, sB, sA, dR, dG, dB); \
ASSEMBLE_RGBA(((Uint8 *)dstp), 4, dstfmt, dR, dG, dB, dA); \
} \
dstp++; \
srcp++; \
widthvar--; \
}
ONE_PIXEL_BLEND((UNALIGNED_PTR(dstp)) && (width), width);
if (width > 0) {
int extrawidth = (width % 4);
vector unsigned char valigner = VEC_ALIGNER(srcp);
vector unsigned char vs = (vector unsigned char)vec_ld(0, srcp);
width -= extrawidth;
while (width) {
vector unsigned char vsel;
vector unsigned char voverflow;
vector unsigned char vd;
vector unsigned char vd_orig;
/* s = *srcp */
voverflow = (vector unsigned char)vec_ld(15, srcp);
vs = vec_perm(vs, voverflow, valigner);
/* vsel is set for items that match the key */
vsel = (vector unsigned char)vec_and((vector unsigned int)vs, vrgbmask);
vsel = (vector unsigned char)vec_cmpeq((vector unsigned int)vsel, vckey);
/* permute to source format */
vs = vec_perm(vs, valpha, vsrcPermute);
/* d = *dstp */
vd = (vector unsigned char)vec_ld(0, dstp);
vd_orig = vd = vec_perm(vd, v0, vsdstPermute);
VEC_MULTIPLY_ALPHA(vs, vd, valpha, mergePermute, v1, v8);
/* set the alpha channel to full on */
vd = vec_or(vd, valphamask);
/* mask out color key */
vd = vec_sel(vd, vd_orig, vsel);
/* permute to dest format */
vd = vec_perm(vd, vbits, vdstPermute);
/* *dstp = res */
vec_st((vector unsigned int)vd, 0, dstp);
srcp += 4;
dstp += 4;
width -= 4;
vs = voverflow;
}
ONE_PIXEL_BLEND((extrawidth), extrawidth);
}
#undef ONE_PIXEL_BLEND
srcp += srcskip;
dstp += dstskip;
}
}
static void Blit32to32PixelAlphaAltivec(SDL_BlitInfo *info)
{
int width = info->d_width;
int height = info->d_height;
Uint32 *srcp = (Uint32 *)info->s_pixels;
int srcskip = info->s_skip >> 2;
Uint32 *dstp = (Uint32 *)info->d_pixels;
int dstskip = info->d_skip >> 2;
SDL_PixelFormat *srcfmt = info->src;
SDL_PixelFormat *dstfmt = info->dst;
vector unsigned char mergePermute;
vector unsigned char valphaPermute;
vector unsigned char vsrcPermute;
vector unsigned char vdstPermute;
vector unsigned char vsdstPermute;
vector unsigned char valphamask;
vector unsigned char vpixelmask;
vector unsigned char v0;
vector unsigned short v1;
vector unsigned short v8;
v0 = vec_splat_u8(0);
v1 = vec_splat_u16(1);
v8 = vec_splat_u16(8);
mergePermute = VEC_MERGE_PERMUTE();
valphamask = VEC_ALPHA_MASK();
valphaPermute = vec_and(vec_lvsl(0, (int *)NULL), vec_splat_u8(0xC));
vpixelmask = vec_nor(valphamask, v0);
vsrcPermute = calc_swizzle32(srcfmt, NULL);
vdstPermute = calc_swizzle32(NULL, dstfmt);
vsdstPermute = calc_swizzle32(dstfmt, NULL);
while ( height-- ) {
width = info->d_width;
#define ONE_PIXEL_BLEND(condition, widthvar) while ((condition)) { \
Uint32 Pixel; \
unsigned sR, sG, sB, dR, dG, dB, sA, dA; \
DISEMBLE_RGBA((Uint8 *)srcp, 4, srcfmt, Pixel, sR, sG, sB, sA); \
if(sA) { \
DISEMBLE_RGBA((Uint8 *)dstp, 4, dstfmt, Pixel, dR, dG, dB, dA); \
ACCURATE_ALPHA_BLEND(sR, sG, sB, sA, dR, dG, dB); \
ASSEMBLE_RGBA((Uint8 *)dstp, 4, dstfmt, dR, dG, dB, dA); \
} \
++srcp; \
++dstp; \
widthvar--; \
}
ONE_PIXEL_BLEND((UNALIGNED_PTR(dstp)) && (width), width);
if (width > 0) {
// vsrcPermute
// vdstPermute
int extrawidth = (width % 4);
vector unsigned char valigner = VEC_ALIGNER(srcp);
vector unsigned char vs = (vector unsigned char)vec_ld(0, srcp);
width -= extrawidth;
while (width) {
vector unsigned char voverflow;
vector unsigned char vd;
vector unsigned char valpha;
vector unsigned char vdstalpha;
/* s = *srcp */
voverflow = (vector unsigned char)vec_ld(15, srcp);
vs = vec_perm(vs, voverflow, valigner);
vs = vec_perm(vs, v0, vsrcPermute);
valpha = vec_perm(vs, v0, valphaPermute);
/* d = *dstp */
vd = (vector unsigned char)vec_ld(0, dstp);
vd = vec_perm(vd, v0, vsdstPermute);
vdstalpha = vec_and(vd, valphamask);
VEC_MULTIPLY_ALPHA(vs, vd, valpha, mergePermute, v1, v8);
/* set the alpha to the dest alpha */
vd = vec_and(vd, vpixelmask);
vd = vec_or(vd, vdstalpha);
vd = vec_perm(vd, v0, vdstPermute);
/* *dstp = res */
vec_st((vector unsigned int)vd, 0, dstp);
srcp += 4;
dstp += 4;
width -= 4;
vs = voverflow;
}
ONE_PIXEL_BLEND((extrawidth), extrawidth);
}
srcp += srcskip;
dstp += dstskip;
#undef ONE_PIXEL_BLEND
}
}
/* fast ARGB888->(A)RGB888 blending with pixel alpha */
static void BlitRGBtoRGBPixelAlphaAltivec(SDL_BlitInfo *info)
{
int width = info->d_width;
int height = info->d_height;
Uint32 *srcp = (Uint32 *)info->s_pixels;
int srcskip = info->s_skip >> 2;
Uint32 *dstp = (Uint32 *)info->d_pixels;
int dstskip = info->d_skip >> 2;
vector unsigned char mergePermute;
vector unsigned char valphaPermute;
vector unsigned char valphamask;
vector unsigned char vpixelmask;
vector unsigned char v0;
vector unsigned short v1;
vector unsigned short v8;
v0 = vec_splat_u8(0);
v1 = vec_splat_u16(1);
v8 = vec_splat_u16(8);
mergePermute = VEC_MERGE_PERMUTE();
valphamask = VEC_ALPHA_MASK();
valphaPermute = vec_and(vec_lvsl(0, (int *)NULL), vec_splat_u8(0xC));
vpixelmask = vec_nor(valphamask, v0);
while(height--) {
width = info->d_width;
#define ONE_PIXEL_BLEND(condition, widthvar) \
while ((condition)) { \
Uint32 dalpha; \
Uint32 d; \
Uint32 s1; \
Uint32 d1; \
Uint32 s = *srcp; \
Uint32 alpha = s >> 24; \
if(alpha) { \
if(alpha == SDL_ALPHA_OPAQUE) { \
*dstp = (s & 0x00ffffff) | (*dstp & 0xff000000); \
} else { \
d = *dstp; \
dalpha = d & 0xff000000; \
s1 = s & 0xff00ff; \
d1 = d & 0xff00ff; \
d1 = (d1 + ((s1 - d1) * alpha >> 8)) & 0xff00ff; \
s &= 0xff00; \
d &= 0xff00; \
d = (d + ((s - d) * alpha >> 8)) & 0xff00; \
*dstp = d1 | d | dalpha; \
} \
} \
++srcp; \
++dstp; \
widthvar--; \
}
ONE_PIXEL_BLEND((UNALIGNED_PTR(dstp)) && (width), width);
if (width > 0) {
int extrawidth = (width % 4);
vector unsigned char valigner = VEC_ALIGNER(srcp);
vector unsigned char vs = (vector unsigned char)vec_ld(0, srcp);
width -= extrawidth;
while (width) {
vector unsigned char voverflow;
vector unsigned char vd;
vector unsigned char valpha;
vector unsigned char vdstalpha;
/* s = *srcp */
voverflow = (vector unsigned char)vec_ld(15, srcp);
vs = vec_perm(vs, voverflow, valigner);
valpha = vec_perm(vs, v0, valphaPermute);
/* d = *dstp */
vd = (vector unsigned char)vec_ld(0, dstp);
vdstalpha = vec_and(vd, valphamask);
VEC_MULTIPLY_ALPHA(vs, vd, valpha, mergePermute, v1, v8);
/* set the alpha to the dest alpha */
vd = vec_and(vd, vpixelmask);
vd = vec_or(vd, vdstalpha);
/* *dstp = res */
vec_st((vector unsigned int)vd, 0, dstp);
srcp += 4;
dstp += 4;
width -= 4;
vs = voverflow;