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
|
#ifndef _BLENDING_H_
#define _BLENDING_H_
//#include <bfc/common.h>
#include <bfc/platform/types.h>
#include <tataki/export.h>
#ifdef _WIN64
#define NO_MMX
#endif
class TATAKIAPI Blenders
{
public:
static void init();
static unsigned int inline BLEND_ADJ1(unsigned int a, unsigned int b, int alpha);
static unsigned int inline BLEND_ADJ2(unsigned int a, unsigned int b);
static unsigned int inline BLEND_ADJ3(unsigned int a, unsigned int b, int alpha);
static unsigned int inline BLEND_MUL(unsigned int a, int v);
static unsigned int inline BLEND_AVG(unsigned int a, unsigned int b);
static unsigned int inline BLEND4(unsigned int *p1, unsigned int w, int xp, int yp);
#ifndef NO_MMX
static int inline MMX_AVAILABLE() { return mmx_available; }
static unsigned int inline BLEND_ADJ1_MMX(unsigned int a, unsigned int b, int alpha);
static unsigned int inline BLEND_ADJ2_MMX(unsigned int a, unsigned int b);
static unsigned int inline BLEND_ADJ3_MMX(unsigned int a, unsigned int b, int alpha);
static unsigned int inline BLEND_MUL_MMX(unsigned int a, int v);
static unsigned int inline BLEND_AVG_MMX(unsigned int a, unsigned int b);
static unsigned int inline BLEND4_MMX(unsigned int *p1, unsigned int w, int xp, int yp);
static void inline BLEND_MMX_END()
{
#if defined(WIN32) && !defined(_WIN64)
if (mmx_available) __asm emms;
#endif
#ifdef LINUX
if (mmx_available) __asm__ volatile ( "emms" : : );
#endif
}
#endif
//private:
static uint8_t alphatable[256][256];
#ifndef NO_MMX
static int mmx_available;
#endif
};
// NON MMX
// average blend of a and b.
unsigned int inline Blenders::BLEND_AVG(unsigned int a, unsigned int b)
{
return ((a >> 1)&~((1 << 7) | (1 << 15) | (1 << 23))) + ((b >> 1)&~((1 << 7) | (1 << 15) | (1 << 23)));
}
// multiplies 32 bit color A by scalar V (0-255)
unsigned int inline Blenders::BLEND_MUL(unsigned int a, int v)
{
register int t;
t = Blenders::alphatable[a & 0xFF][v];
t |= Blenders::alphatable[(a & 0xFF00) >> 8][v] << 8;
t |= Blenders::alphatable[(a & 0xFF0000) >> 16][v] << 16;
t |= Blenders::alphatable[(a & 0xFF000000) >> 24][v] << 24;
return t;
}
// V is scalar (0-255), (1.0-V)*b + V*a
unsigned int inline Blenders::BLEND_ADJ1(unsigned int a, unsigned int b, int v)
{
register int t;
t = Blenders::alphatable[b & 0xFF][0xFF - v] + Blenders::alphatable[a & 0xFF][v];
t |= (Blenders::alphatable[(b & 0xFF00) >> 8][0xFF - v] + Blenders::alphatable[(a & 0xFF00) >> 8][v]) << 8;
t |= (Blenders::alphatable[(b & 0xFF0000) >> 16][0xFF - v] + Blenders::alphatable[(a & 0xFF0000) >> 16][v]) << 16;
t |= (Blenders::alphatable[(b & 0xFF000000) >> 24][0xFF - v] + Blenders::alphatable[(a & 0xFF000000) >> 24][v]) << 24;
return t;
}
// returns a*(1.0-Alpha(b)) + b
unsigned int inline Blenders::BLEND_ADJ2(unsigned int a, unsigned int b)
{
register int t, z;
int v = 0xff - ((b >> 24) & 0xff);
t = Blenders::alphatable[a & 0xFF][v] + (b & 0xFF);
if (t > 0xFF) t = 0xff;
z = (Blenders::alphatable[(a & 0xFF00) >> 8][v] << 8) + (b & 0xFF00);
if (z > 0xFF00) z = 0xff00;
t |= z;
z = (Blenders::alphatable[(a & 0xFF0000) >> 16][v] << 16) + ((b & 0xFF0000));
if (z > 0xFF0000) z = 0xff0000;
t |= z;
z = (Blenders::alphatable[(a & 0xFF000000) >> 24][v]) + ((b & 0xFF000000) >> 24);
if (z > 0xFF) z = 0xff;
return t | (z << 24);
}
// returns a*(1-Alpha(b)*W) + b*W, clamped (W is scalar 0-0xff).
unsigned int inline Blenders::BLEND_ADJ3(unsigned int a, unsigned int b, int w)
{
register int t, z;
int v = 0xff - Blenders::alphatable[(b >> 24) & 0xff][w];
t = Blenders::alphatable[a & 0xFF][v] + Blenders::alphatable[b & 0xFF][w];
if (t > 0xFF) t = 0xFF;
z = Blenders::alphatable[(a & 0xFF00) >> 8][v] + Blenders::alphatable[(b & 0xFF00) >> 8][w];
if (z > 0xFF) z = 0xFF;
t |= z << 8;
z = Blenders::alphatable[(a & 0xFF0000) >> 16][v] + Blenders::alphatable[(b & 0xFF0000) >> 16][w];
if (z > 0xFF) z = 0xFF;
t |= z << 16;
z = Blenders::alphatable[(a & 0xFF000000) >> 24][v] + Blenders::alphatable[(b & 0xFF000000) >> 24][w];
if (z > 0xFF) z = 0xFF;
return t | (z << 24);
}
unsigned int __inline Blenders::BLEND4(unsigned int *p1, unsigned int w, int xp, int yp)
{
register int t;
uint8_t a1, a2, a3, a4;
xp = (xp >> 8) & 0xff;
yp = (yp >> 8) & 0xff;
a1 = alphatable[255 - xp][255 - yp];
a2 = alphatable[xp][255 - yp];
a3 = alphatable[255 - xp][yp];
a4 = alphatable[xp][yp];
t = alphatable[p1[0] & 0xff][a1] + alphatable[p1[1] & 0xff][a2] + alphatable[p1[w] & 0xff][a3] + alphatable[p1[w + 1] & 0xff][a4];
t |= (alphatable[(p1[0] >> 8) & 0xff][a1] + alphatable[(p1[1] >> 8) & 0xff][a2] + alphatable[(p1[w] >> 8) & 0xff][a3] + alphatable[(p1[w + 1] >> 8) & 0xff][a4]) << 8;
t |= (alphatable[(p1[0] >> 16) & 0xff][a1] + alphatable[(p1[1] >> 16) & 0xff][a2] + alphatable[(p1[w] >> 16) & 0xff][a3] + alphatable[(p1[w + 1] >> 16) & 0xff][a4]) << 16;
t |= (alphatable[(p1[0] >> 24) & 0xff][a1] + alphatable[(p1[1] >> 24) & 0xff][a2] + alphatable[(p1[w] >> 24) & 0xff][a3] + alphatable[(p1[w + 1] >> 24) & 0xff][a4]) << 24;
return t;
}
#ifndef NO_MMX
#ifdef WIN32
#pragma warning( push, 1 )
#pragma warning(disable: 4799)
#endif
#ifdef WIN32
#define MMX_CONST const
#else
#define MMX_CONST
#endif
static unsigned int MMX_CONST Blenders__mmx_revn2[2] = {0x01000100, 0x01000100};
static unsigned int MMX_CONST Blenders__mmx_zero[2];
static unsigned int MMX_CONST Blenders__mmx_one[2] = {1, 0};
#undef MMX_CONST
/// MMX
// average blend of a and b.
unsigned int inline Blenders::BLEND_AVG_MMX(unsigned int a, unsigned int b)
{
return ((a >> 1)&~((1 << 7) | (1 << 15) | (1 << 23))) + ((b >> 1)&~((1 << 7) | (1 << 15) | (1 << 23)));
}
// multiplies 32 bit color A by scalar V (0-255)
unsigned int inline Blenders::BLEND_MUL_MMX(unsigned int a, int v)
{
#ifdef WIN32
__asm
{
movd mm3, [v] // VVVVVVVV
movd mm0, [a]
packuswb mm3, mm3 // 0000HHVV
punpcklbw mm0, [Blenders__mmx_zero]
punpcklwd mm3, mm3 // HHVVHHVV
punpckldq mm3, mm3 // HHVVHHVV HHVVHHVV
pmullw mm0, mm3
psrlw mm0, 8
packuswb mm0, mm0
movd eax, mm0
}
#else
__asm__ volatile (
"movd %0, %%mm3\n"
"movd %1, %%mm0\n"
"packuswb %%mm3, %%mm3\n"
"punpcklbw (Blenders__mmx_zero), %%mm0\n"
"punpcklwd %%mm3, %%mm3\n"
"punpckldq %%mm3, %%mm3\n"
"pmullw %%mm3, %%mm0\n"
"psrlw $8, %%mm0\n"
"packuswb %%mm0, %%mm0\n"
"movd %%mm0, %%eax\n"
:
: "m" (v), "m" (a)
: "%mm0", "%mm3" );
#endif
}
// V is scalar (0-255), (1.0-V)*b + V*a
unsigned int inline Blenders::BLEND_ADJ1_MMX(unsigned int a, unsigned int b, int v)
{
#ifdef WIN32
__asm
{
movd mm3, [v] // VVVVVVVV
movd mm0, [a]
packuswb mm3, mm3 // 0000HHVV
movd mm1, [b]
paddusw mm3, [Blenders__mmx_one]
movq mm4, [Blenders__mmx_revn2]
punpcklwd mm3, mm3 // HHVVHHVV
punpcklbw mm0, [Blenders__mmx_zero]
punpckldq mm3, mm3 // HHVVHHVV HHVVHHVV
punpcklbw mm1, [Blenders__mmx_zero]
psubw mm4, mm3
pmullw mm0, mm3
pmullw mm1, mm4
paddw mm0, mm1
psrlw mm0, 8
packuswb mm0, mm0
movd eax, mm0
}
#else
__asm__ volatile (
"movd %0, %%mm3\n"
"movd %1, %%mm0\n"
"packuswb %%mm3, %%mm3\n"
"movd %2, %%mm1\n"
"paddusw (Blenders__mmx_one), %%mm3\n"
"movq (Blenders__mmx_revn2), %%mm4\n"
"punpcklwd %%mm3, %%mm3\n"
"punpcklbw (Blenders__mmx_zero), %%mm0\n"
"punpckldq %%mm3, %%mm3\n"
"punpcklbw (Blenders__mmx_zero), %%mm1\n"
"psubw %%mm3, %%mm4\n"
"pmullw %%mm3, %%mm0\n"
"pmullw %%mm4, %%mm1\n"
"paddw %%mm1, %%mm0\n"
"psrlw $8, %%mm0\n"
"packuswb %%mm0, %%mm0\n"
"movd %%mm0, %%eax\n"
:
: "m" (v), "m" (a), "m" (b)
: "%mm0", "%mm1", "%mm3", "%mm4" );
#endif
}
// returns a*(1.0-Alpha(b)) + b
unsigned int inline Blenders::BLEND_ADJ2_MMX(unsigned int a, unsigned int b)
{
#ifdef WIN32
__asm
{
movd mm3, [b] // VVVVVVVV
movq mm4, [Blenders__mmx_revn2]
movd mm0, [a]
psrld mm3, 24
movd mm1, [b]
paddusw mm3, [Blenders__mmx_one]
punpcklwd mm3, mm3 // HHVVHHVV
punpcklbw mm0, [Blenders__mmx_zero]
punpckldq mm3, mm3 // HHVVHHVV HHVVHHVV
punpcklbw mm1, [Blenders__mmx_zero]
psubw mm4, mm3
pmullw mm0, mm4
// stall
// stall
// stall
psrlw mm0, 8
// stall
paddw mm0, mm1
// stall
packuswb mm0, mm0
// stall
movd eax, mm0
}
#else
__asm__ volatile (
"movd %1, %%mm3\n"
"movq (Blenders__mmx_revn2), %%mm4\n"
"movd %0, %%mm0\n"
"psrld $24, %%mm3\n"
"movd %1, %%mm1\n"
"paddusw (Blenders__mmx_one), %%mm3\n"
"punpcklwd %%mm3, %%mm3\n"
"punpcklbw (Blenders__mmx_zero), %%mm0\n"
"punpckldq %%mm3, %%mm3\n"
"punpcklbw (Blenders__mmx_zero), %%mm1\n"
"psubw %%mm3, %%mm4\n"
"pmullw %%mm4, %%mm0\n"
"psrlw $8, %%mm0\n"
"paddw %%mm1, %%mm0\n"
"packuswb %%mm0, %%mm0\n"
"movd %%mm0, %%eax\n"
:
: "m" (a), "m" (b)
: "%esi", "%mm0", "%mm1", "%mm3", "%mm4" );
#endif
}
// returns a*(1-Alpha(b)*W) + b*W, clamped (W is scalar 0-0xff).
unsigned int inline Blenders::BLEND_ADJ3_MMX(unsigned int a, unsigned int b, int w)
{
#ifdef WIN32
__asm
{
movd mm3, [b] // VVVVVVVV
movd mm5, [w]
movd mm0, [a]
psrld mm3, 24
movd mm1, [b]
paddusw mm3, [Blenders__mmx_one]
movq mm4, [Blenders__mmx_revn2]
pmullw mm3, mm5
packuswb mm5, mm5
punpcklbw mm0, [Blenders__mmx_zero]
punpcklwd mm5, mm5
punpcklbw mm1, [Blenders__mmx_zero]
psrlw mm3, 8
punpckldq mm5, mm5
paddusw mm3, [Blenders__mmx_one]
punpcklwd mm3, mm3 // HHVVHHVV
punpckldq mm3, mm3 // HHVVHHVV HHVVHHVV
psubw mm4, mm3
pmullw mm0, mm4
pmullw mm1, mm5
paddusw mm0, mm1
psrlw mm0, 8
packuswb mm0, mm0
movd eax, mm0
}
#else
__asm__ volatile (
"movd %2, %%mm3\n"
"movd %0, %%mm5\n"
"movd %1, %%mm0\n"
"psrld $24, %%mm3\n"
"movd %2, %%mm1\n"
"paddusw (Blenders__mmx_one), %%mm3\n"
"movq (Blenders__mmx_revn2), %%mm4\n"
"pmullw %%mm5, %%mm3\n"
"packuswb %%mm5, %%mm5 \n"
"punpcklbw (Blenders__mmx_zero), %%mm0\n"
"punpcklwd %%mm5, %%mm5\n"
"punpcklbw (Blenders__mmx_zero), %%mm1\n"
"psrlw $8, %%mm3\n"
"punpckldq %%mm5, %%mm5\n"
"paddusw (Blenders__mmx_one), %%mm3\n"
"punpcklwd %%mm3, %%mm3\n"
"punpckldq %%mm3, %%mm3\n"
"psubw %%mm3, %%mm4\n"
"pmullw %%mm4, %%mm0\n"
"pmullw %%mm5, %%mm1\n"
"paddusw %%mm1, %%mm0\n"
"psrlw $8, %%mm0\n"
"packuswb %%mm0, %%mm0\n"
"movd %%mm0, %%eax\n"
:
: "m" (w), "m" (a), "m" (b)
: "%mm0", "%mm1", "%mm4", "%mm3", "%mm5" );
#endif
}
// does bilinear filtering. p1 is upper left pixel, w is width of framebuffer
// xp and yp's low 16 bits are used for the subpixel positioning.
unsigned int inline Blenders::BLEND4_MMX(unsigned int *p1, unsigned int w, int xp, int yp)
{
#ifdef WIN32
__asm
{
movd mm6, xp
mov eax, p1
movd mm7, yp
mov esi, w
movq mm4, Blenders__mmx_revn2
psrlw mm6, 8
movq mm5, Blenders__mmx_revn2
psrlw mm7, 8
movd mm0, [eax]
punpcklwd mm6, mm6
movd mm1, [eax + 4]
punpcklwd mm7, mm7
movd mm2, [eax + esi*4]
punpckldq mm6, mm6
movd mm3, [eax + esi*4 + 4]
punpckldq mm7, mm7
punpcklbw mm0, [Blenders__mmx_zero]
psubw mm4, mm6
punpcklbw mm1, [Blenders__mmx_zero]
pmullw mm0, mm4
punpcklbw mm2, [Blenders__mmx_zero]
pmullw mm1, mm6
punpcklbw mm3, [Blenders__mmx_zero]
psubw mm5, mm7
pmullw mm2, mm4
pmullw mm3, mm6
paddw mm0, mm1
// stall (mm0)
psrlw mm0, 8
// stall (waiting for mm3/mm2)
paddw mm2, mm3
pmullw mm0, mm5
psrlw mm2, 8
// stall (mm2)
pmullw mm2, mm7
// stall
// stall (mm2)
paddw mm0, mm2
// stall
psrlw mm0, 8
// stall
packuswb mm0, mm0
// stall
movd eax, mm0
}
#else
__asm__ volatile (
"movd %2, %%mm6\n"
"mov %0, %%eax\n"
"movd %3, %%mm7\n"
"mov %1, %%esi\n"
"movq (Blenders__mmx_revn2), %%mm4\n"
"psrlw $8, %%mm6\n"
"movq (Blenders__mmx_revn2), %%mm5\n"
"psrlw $8, %%mm7\n"
"movd (%%eax), %%mm0\n"
"punpcklwd %%mm6,%%mm6\n"
"movd 4(%%eax), %%mm1\n"
"punpcklwd %%mm7,%%mm7\n"
"movd (%%eax,%%esi,4), %%mm2\n"
"punpckldq %%mm6,%%mm6\n"
"movd 4(%%eax,%%esi,4), %%mm3\n"
"punpckldq %%mm7,%%mm7\n"
"punpcklbw (Blenders__mmx_zero), %%mm0\n"
"psubw %%mm6, %%mm4\n"
"punpcklbw (Blenders__mmx_zero), %%mm1\n"
"pmullw %%mm4, %%mm0\n"
"punpcklbw (Blenders__mmx_zero), %%mm2\n"
"pmullw %%mm6, %%mm1\n"
"punpcklbw (Blenders__mmx_zero), %%mm3\n"
"psubw %%mm7, %%mm5\n"
"pmullw %%mm4, %%mm2\n"
"pmullw %%mm6, %%mm3\n"
"paddw %%mm1, %%mm0\n"
"psrlw $8, %%mm0\n"
"paddw %%mm3, %%mm2\n"
"pmullw %%mm5, %%mm0\n"
"psrlw $8, %%mm2\n"
"pmullw %%mm7, %%mm2\n"
"paddw %%mm2, %%mm0\n"
"psrlw $8, %%mm0\n"
"packuswb %%mm0, %%mm0\n"
"movd %%mm0, %%eax\n"
:
: "m" (p1), "m" (w), "m" (xp), "m" (yp)
: "%mm0", "%mm1", "%mm4", "%mm3", "%mm5" );
#endif
}
#ifdef WIN32
#pragma warning( pop )
#endif
#endif // ndef NO_MMX
#endif
|
