/
SDL_blit_A.c
1470 lines (1328 loc) · 44.6 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) 1997-2004 Sam Lantinga
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Library General Public
License as published by the Free Software Foundation; either
version 2 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
Library General Public License for more details.
You should have received a copy of the GNU Library General Public
License along with this library; if not, write to the Free
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Sam Lantinga
slouken@libsdl.org
*/
#ifdef SAVE_RCSID
static char rcsid =
"@(#) $Id$";
#endif
#include <stdio.h>
#include "SDL_types.h"
#include "SDL_video.h"
#include "SDL_blit.h"
#if (defined(i386) || defined(__x86_64__)) && defined(__GNUC__) && defined(USE_ASMBLIT)
#define MMX_ASMBLIT
#endif
#ifdef MMX_ASMBLIT
/* Function to check the CPU flags */
#include "SDL_cpuinfo.h"
#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;
}
}
#ifdef 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
/* fast RGB888->(A)RGB888 blending with surface alpha=128 special case */
static void BlitRGBtoRGBSurfaceAlpha128(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;
while(height--) {
DUFFS_LOOP4({
Uint32 s = *srcp++;
Uint32 d = *dstp;
*dstp++ = ((((s & 0x00fefefe) + (d & 0x00fefefe)) >> 1)
+ (s & d & 0x00010101)) | 0xff000000;
}, width);
srcp += srcskip;
dstp += dstskip;
}
}
/* fast RGB888->(A)RGB888 blending with surface alpha */
static void BlitRGBtoRGBSurfaceAlpha(SDL_BlitInfo *info)
{
unsigned alpha = info->src->alpha;
if(alpha == 128) {
BlitRGBtoRGBSurfaceAlpha128(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;
Uint32 s;
Uint32 d;
Uint32 s1;
Uint32 d1;
while(height--) {
DUFFS_LOOP_DOUBLE2({
/* One Pixel Blend */
s = *srcp;
d = *dstp;
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 | 0xff000000;
++srcp;
++dstp;
},{
/* Two Pixels Blend */
s = *srcp;
d = *dstp;
s1 = s & 0xff00ff;
d1 = d & 0xff00ff;
d1 += (s1 - d1) * alpha >> 8;
d1 &= 0xff00ff;
s = ((s & 0xff00) >> 8) |
((srcp[1] & 0xff00) << 8);
d = ((d & 0xff00) >> 8) |
((dstp[1] & 0xff00) << 8);
d += (s - d) * alpha >> 8;
d &= 0x00ff00ff;
*dstp++ = d1 | ((d << 8) & 0xff00) | 0xff000000;
++srcp;
s1 = *srcp;
d1 = *dstp;
s1 &= 0xff00ff;
d1 &= 0xff00ff;
d1 += (s1 - d1) * alpha >> 8;
d1 &= 0xff00ff;
*dstp = d1 | ((d >> 8) & 0xff00) | 0xff000000;
++srcp;
++dstp;
}, width);
srcp += srcskip;
dstp += dstskip;
}
}
}
/* fast ARGB888->(A)RGB888 blending with pixel alpha */
static void BlitRGBtoRGBPixelAlpha(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;
while(height--) {
DUFFS_LOOP4({
Uint32 dalpha;
Uint32 d;
Uint32 s1;
Uint32 d1;
Uint32 s = *srcp;
Uint32 alpha = s >> 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) {
*dstp = (s & 0x00ffffff) | (*dstp & 0xff000000);
} else {
/*
* take out the middle component (green), and process
* the other two in parallel. One multiply less.
*/
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;
}, width);
srcp += srcskip;
dstp += dstskip;
}
}
#ifdef MMX_ASMBLIT
/* fast (as in MMX with prefetch) ARGB888->(A)RGB888 blending with pixel alpha */
inline static void BlitRGBtoRGBPixelAlphaMMX3DNOW(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 s;
Uint32 alpha;
__asm__ (
/* make mm6 all zeros. */
"pxor %%mm6, %%mm6\n"
/* Make a mask to preserve the alpha. */
"pcmpeqb %%mm7, %%mm7\n\t" /* mm7(s) = FF FF FF FF | FF FF FF FF */
"psrlq $16, %%mm7\n\t" /* mm7(s) = 00 00 FF FF | FF FF FF FF */
: );
while(height--) {
DUFFS_LOOP4({
__asm__ (
"prefetch 64(%0)\n"
"prefetch 64(%1)\n"
: : "r" (srcp), "r" (dstp) );
s = *srcp;
alpha = s >> 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 == SDL_ALPHA_OPAQUE) {
*dstp = (s & 0x00ffffff) | (*dstp & 0xff000000);
}
else {
__asm__ (
/* load in the source, and dst. */
"movd (%0), %%mm0\n" /* mm0(s) = 0 0 0 0 | As Rs Gs Bs */
"movd (%1), %%mm1\n" /* mm1(d) = 0 0 0 0 | Ad Rd Gd Bd */
/* Move the src alpha into mm2 */
/* if supporting pshufw */
/*"pshufw $0x55, %%mm0, %%mm2\n" */ /* mm2 = 0 As 0 As | 0 As 0 As */
/*"psrlw $8, %%mm2\n" */
/* else: */
"movq %%mm0, %%mm2\n"
"psrld $24, %%mm2\n" /* mm2 = 0 0 0 0 | 0 0 0 As */
"punpcklwd %%mm2, %%mm2\n" /* mm2 = 0 0 0 0 | 0 As 0 As */
"punpckldq %%mm2, %%mm2\n" /* mm2 = 0 As 0 As | 0 As 0 As */
/* move the colors into words. */
"punpcklbw %%mm6, %%mm0\n" /* mm0 = 0 As 0 Rs | 0 Gs 0 Bs */
"punpcklbw %%mm6, %%mm1\n" /* mm0 = 0 Ad 0 Rd | 0 Gd 0 Bd */
/* src - dst */
"psubw %%mm1, %%mm0\n" /* mm0 = As-Ad Rs-Rd | Gs-Gd Bs-Bd */
/* A * (src-dst) */
"pmullw %%mm2, %%mm0\n" /* mm0 = As*As-d As*Rs-d | As*Gs-d As*Bs-d */
"pand %%mm7, %%mm0\n" /* to preserve dest alpha */
"psrlw $8, %%mm0\n" /* mm0 = Ac>>8 Rc>>8 | Gc>>8 Bc>>8 */
"paddb %%mm1, %%mm0\n" /* mm0 = Ac+Ad Rc+Rd | Gc+Gd Bc+Bd */
"packuswb %%mm0, %%mm0\n" /* mm0 = | Ac Rc Gc Bc */
"movd %%mm0, (%1)\n" /* result in mm0 */
: : "r" (srcp), "r" (dstp) );
}
++srcp;
++dstp;
}, width);
srcp += srcskip;
dstp += dstskip;
}
__asm__ (
"emms\n"
: );
}
#endif
/* 16bpp special case for per-surface alpha=50%: blend 2 pixels in parallel */
/* blend a single 16 bit pixel at 50% */
#define BLEND16_50(d, s, mask) \
((((s & mask) + (d & mask)) >> 1) + (s & d & (~mask & 0xffff)))
/* blend two 16 bit pixels at 50% */
#define BLEND2x16_50(d, s, mask) \
(((s & (mask | mask << 16)) >> 1) + ((d & (mask | mask << 16)) >> 1) \
+ (s & d & (~(mask | mask << 16))))
static void Blit16to16SurfaceAlpha128(SDL_BlitInfo *info, Uint16 mask)
{
int width = info->d_width;
int height = info->d_height;
Uint16 *srcp = (Uint16 *)info->s_pixels;
int srcskip = info->s_skip >> 1;
Uint16 *dstp = (Uint16 *)info->d_pixels;
int dstskip = info->d_skip >> 1;
while(height--) {
if(((unsigned long)srcp ^ (unsigned long)dstp) & 2) {
/*
* Source and destination not aligned, pipeline it.
* This is mostly a win for big blits but no loss for
* small ones
*/
Uint32 prev_sw;
int w = width;
/* handle odd destination */
if((unsigned long)dstp & 2) {
Uint16 d = *dstp, s = *srcp;
*dstp = BLEND16_50(d, s, mask);
dstp++;
srcp++;
w--;
}
srcp++; /* srcp is now 32-bit aligned */
/* bootstrap pipeline with first halfword */
prev_sw = ((Uint32 *)srcp)[-1];
while(w > 1) {
Uint32 sw, dw, s;
sw = *(Uint32 *)srcp;
dw = *(Uint32 *)dstp;
if(SDL_BYTEORDER == SDL_BIG_ENDIAN)
s = (prev_sw << 16) + (sw >> 16);
else
s = (prev_sw >> 16) + (sw << 16);
prev_sw = sw;
*(Uint32 *)dstp = BLEND2x16_50(dw, s, mask);
dstp += 2;
srcp += 2;
w -= 2;
}
/* final pixel if any */
if(w) {
Uint16 d = *dstp, s;
if(SDL_BYTEORDER == SDL_BIG_ENDIAN)
s = prev_sw;
else
s = prev_sw >> 16;
*dstp = BLEND16_50(d, s, mask);
srcp++;
dstp++;
}
srcp += srcskip - 1;
dstp += dstskip;
} else {
/* source and destination are aligned */
int w = width;
/* first odd pixel? */
if((unsigned long)srcp & 2) {
Uint16 d = *dstp, s = *srcp;
*dstp = BLEND16_50(d, s, mask);
srcp++;
dstp++;
w--;
}
/* srcp and dstp are now 32-bit aligned */
while(w > 1) {
Uint32 sw = *(Uint32 *)srcp;
Uint32 dw = *(Uint32 *)dstp;
*(Uint32 *)dstp = BLEND2x16_50(dw, sw, mask);
srcp += 2;
dstp += 2;
w -= 2;
}
/* last odd pixel? */
if(w) {
Uint16 d = *dstp, s = *srcp;
*dstp = BLEND16_50(d, s, mask);
srcp++;
dstp++;
}
srcp += srcskip;
dstp += dstskip;
}
}
}
#ifdef MMX_ASMBLIT
/* fast RGB565->RGB565 blending with surface alpha */
static void Blit565to565SurfaceAlphaMMX(SDL_BlitInfo *info)
{
unsigned alpha = info->src->alpha; /* downscale alpha to 5 bits */
if(alpha == 128) {
Blit16to16SurfaceAlpha128(info, 0xf7de);
} else {
int width = info->d_width;
int height = info->d_height;
Uint16 *srcp = (Uint16 *)info->s_pixels;
int srcskip = info->s_skip >> 1;
Uint16 *dstp = (Uint16 *)info->d_pixels;
int dstskip = info->d_skip >> 1;
Uint32 s, d;
Uint8 load[8];
alpha &= ~(1+2+4); /* cut alpha to get the exact same behaviour */
*(Uint64 *)load = alpha;
alpha >>= 3; /* downscale alpha to 5 bits */
movq_m2r(*load, mm0); /* alpha(0000000A) -> mm0 */
punpcklwd_r2r(mm0, mm0); /* 00000A0A -> mm0 */
punpcklwd_r2r(mm0, mm0); /* 0A0A0A0A -> mm0 */
/* Setup the 565 color channel masks */
*(Uint64 *)load = 0xF800F800F800F800ULL;
movq_m2r(*load, mm1); /* MASKRED -> mm1 */
*(Uint64 *)load = 0x07E007E007E007E0ULL;
movq_m2r(*load, mm4); /* MASKGREEN -> mm4 */
*(Uint64 *)load = 0x001F001F001F001FULL;
movq_m2r(*load, mm7); /* MASKBLUE -> mm7 */
while(height--) {
DUFFS_LOOP_QUATRO2(
{
s = *srcp++;
d = *dstp;
/*
* shift out the middle component (green) to
* the high 16 bits, and process all three RGB
* components at the same time.
*/
s = (s | s << 16) & 0x07e0f81f;
d = (d | d << 16) & 0x07e0f81f;
d += (s - d) * alpha >> 5;
d &= 0x07e0f81f;
*dstp++ = d | d >> 16;
},{
s = *srcp++;
d = *dstp;
/*
* shift out the middle component (green) to
* the high 16 bits, and process all three RGB
* components at the same time.
*/
s = (s | s << 16) & 0x07e0f81f;
d = (d | d << 16) & 0x07e0f81f;
d += (s - d) * alpha >> 5;
d &= 0x07e0f81f;
*dstp++ = d | d >> 16;
s = *srcp++;
d = *dstp;
/*
* shift out the middle component (green) to
* the high 16 bits, and process all three RGB
* components at the same time.
*/
s = (s | s << 16) & 0x07e0f81f;
d = (d | d << 16) & 0x07e0f81f;
d += (s - d) * alpha >> 5;
d &= 0x07e0f81f;
*dstp++ = d | d >> 16;
},{
movq_m2r((*dstp), mm3);/* 4 dst pixels -> mm3 */
movq_m2r((*srcp), mm2);/* 4 src pixels -> mm2 */
/* RED */
movq_r2r(mm2, mm5); /* src -> mm5 */
pand_r2r(mm1 , mm5); /* src & MASKRED -> mm5 */
psrlq_i2r(11, mm5); /* mm5 >> 11 -> mm5 [000r 000r 000r 000r] */
movq_r2r(mm3, mm6); /* dst -> mm6 */
pand_r2r(mm1 , mm6); /* dst & MASKRED -> mm6 */
psrlq_i2r(11, mm6); /* mm6 >> 11 -> mm6 [000r 000r 000r 000r] */
/* blend */
psubw_r2r(mm6, mm5);/* src - dst -> mm5 */
pmullw_r2r(mm0, mm5); /* mm5 * alpha -> mm5 */
psrlw_i2r(8, mm5); /* mm5 >> 8 -> mm5 */
paddw_r2r(mm5, mm6); /* mm5 + mm6(dst) -> mm6 */
psllq_i2r(11, mm6); /* mm6 << 11 -> mm6 */
pand_r2r(mm1, mm6); /* mm6 & MASKRED -> mm6 */
movq_r2r(mm4, mm5); /* MASKGREEN -> mm5 */
por_r2r(mm7, mm5); /* MASKBLUE | mm5 -> mm5 */
pand_r2r(mm5, mm3); /* mm3 & mm5(!MASKRED) -> mm3 */
por_r2r(mm6, mm3); /* save new reds in dsts */
/* green */
movq_r2r(mm2, mm5); /* src -> mm5 */
pand_r2r(mm4 , mm5); /* src & MASKGREEN -> mm5 */
psrlq_i2r(5, mm5); /* mm5 >> 5 -> mm5 [000g 000g 000g 000g] */
movq_r2r(mm3, mm6); /* dst -> mm6 */
pand_r2r(mm4 , mm6); /* dst & MASKGREEN -> mm6 */
psrlq_i2r(5, mm6); /* mm6 >> 5 -> mm6 [000g 000g 000g 000g] */
/* blend */
psubw_r2r(mm6, mm5);/* src - dst -> mm5 */
pmullw_r2r(mm0, mm5); /* mm5 * alpha -> mm5 */
psrlw_i2r(8, mm5); /* mm5 >> 8 -> mm5 */
paddw_r2r(mm5, mm6); /* mm5 + mm6(dst) -> mm6 */
psllq_i2r(5, mm6); /* mm6 << 5 -> mm6 */
pand_r2r(mm4, mm6); /* mm6 & MASKGREEN -> mm6 */
movq_r2r(mm1, mm5); /* MASKRED -> mm5 */
por_r2r(mm7, mm5); /* MASKBLUE | mm5 -> mm5 */
pand_r2r(mm5, mm3); /* mm3 & mm5(!MASKGREEN) -> mm3 */
por_r2r(mm6, mm3); /* save new greens in dsts */
/* blue */
movq_r2r(mm2, mm5); /* src -> mm5 */
pand_r2r(mm7 , mm5); /* src & MASKRED -> mm5[000b 000b 000b 000b] */
movq_r2r(mm3, mm6); /* dst -> mm6 */
pand_r2r(mm7 , mm6); /* dst & MASKBLUE -> mm6[000b 000b 000b 000b] */
/* blend */
psubw_r2r(mm6, mm5);/* src - dst -> mm5 */
pmullw_r2r(mm0, mm5); /* mm5 * alpha -> mm5 */
psrlw_i2r(8, mm5); /* mm5 >> 8 -> mm5 */
paddw_r2r(mm5, mm6); /* mm5 + mm6(dst) -> mm6 */
pand_r2r(mm7, mm6); /* mm6 & MASKBLUE -> mm6 */
movq_r2r(mm1, mm5); /* MASKRED -> mm5 */
por_r2r(mm4, mm5); /* MASKGREEN | mm5 -> mm5 */
pand_r2r(mm5, mm3); /* mm3 & mm5(!MASKBLUE) -> mm3 */
por_r2r(mm6, mm3); /* save new blues in dsts */
movq_r2m(mm3, *dstp);/* mm3 -> 4 dst pixels */
srcp += 4;
dstp += 4;
}, width);
srcp += srcskip;
dstp += dstskip;
}
emms();
}
}
/* fast RGB555->RGB555 blending with surface alpha */
static void Blit555to555SurfaceAlphaMMX(SDL_BlitInfo *info)
{
unsigned alpha = info->src->alpha; /* downscale alpha to 5 bits */
if(alpha == 128) {
Blit16to16SurfaceAlpha128(info, 0xfbde);
} else {
int width = info->d_width;
int height = info->d_height;
Uint16 *srcp = (Uint16 *)info->s_pixels;
int srcskip = info->s_skip >> 1;
Uint16 *dstp = (Uint16 *)info->d_pixels;
int dstskip = info->d_skip >> 1;
Uint32 s, d;
Uint8 load[8];
alpha &= ~(1+2+4); /* cut alpha to get the exact same behaviour */
*(Uint64 *)load = alpha;
alpha >>= 3; /* downscale alpha to 5 bits */
movq_m2r(*load, mm0); /* alpha(0000000A) -> mm0 */
punpcklwd_r2r(mm0, mm0); /* 00000A0A -> mm0 */
punpcklwd_r2r(mm0, mm0); /* 0A0A0A0A -> mm0 */
/* Setup the 555 color channel masks */
*(Uint64 *)load = 0x7C007C007C007C00ULL;
movq_m2r(*load, mm1); /* MASKRED -> mm1 */
*(Uint64 *)load = 0x03E003E003E003E0ULL;
movq_m2r(*load, mm4); /* MASKGREEN -> mm4 */
*(Uint64 *)load = 0x001F001F001F001FULL;
movq_m2r(*load, mm7); /* MASKBLUE -> mm7 */
while(height--) {
DUFFS_LOOP_QUATRO2(
{
s = *srcp++;
d = *dstp;
/*
* shift out the middle component (green) to
* the high 16 bits, and process all three RGB
* components at the same time.
*/
s = (s | s << 16) & 0x03e07c1f;
d = (d | d << 16) & 0x03e07c1f;
d += (s - d) * alpha >> 5;
d &= 0x03e07c1f;
*dstp++ = d | d >> 16;
},{
s = *srcp++;
d = *dstp;
/*
* shift out the middle component (green) to
* the high 16 bits, and process all three RGB
* components at the same time.
*/
s = (s | s << 16) & 0x03e07c1f;
d = (d | d << 16) & 0x03e07c1f;
d += (s - d) * alpha >> 5;
d &= 0x03e07c1f;
*dstp++ = d | d >> 16;
s = *srcp++;
d = *dstp;
/*
* shift out the middle component (green) to
* the high 16 bits, and process all three RGB
* components at the same time.
*/
s = (s | s << 16) & 0x03e07c1f;
d = (d | d << 16) & 0x03e07c1f;
d += (s - d) * alpha >> 5;
d &= 0x03e07c1f;
*dstp++ = d | d >> 16;
},{
movq_m2r((*dstp), mm3);/* 4 dst pixels -> mm3 */
movq_m2r((*srcp), mm2);/* 4 src pixels -> mm2 */
/* RED */
movq_r2r(mm2, mm5); /* src -> mm5 */
pand_r2r(mm1 , mm5); /* src & MASKRED -> mm5 */
psrlq_i2r(10, mm5); /* mm5 >> 10 -> mm5 [000r 000r 000r 000r] */
movq_r2r(mm3, mm6); /* dst -> mm6 */
pand_r2r(mm1 , mm6); /* dst & MASKRED -> mm6 */
psrlq_i2r(10, mm6); /* mm6 >> 10 -> mm6 [000r 000r 000r 000r] */