This repository has been archived by the owner on Feb 11, 2021. It is now read-only.
/
SDL_blit.h
533 lines (497 loc) · 14.9 KB
1
2
/*
SDL - Simple DirectMedia Layer
3
Copyright (C) 1997-2011 Sam Lantinga
4
5
This library is free software; you can redistribute it and/or
6
modify it under the terms of the GNU Lesser General Public
7
License as published by the Free Software Foundation; either
8
version 2.1 of the License, or (at your option) any later version.
9
10
11
12
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
13
Lesser General Public License for more details.
14
15
16
17
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
18
19
Sam Lantinga
20
slouken@libsdl.org
21
*/
22
#include "SDL_config.h"
23
24
25
26
#ifndef _SDL_blit_h
#define _SDL_blit_h
27
#include "SDL_cpuinfo.h"
28
#include "SDL_endian.h"
29
#include "SDL_surface.h"
30
31
/* SDL blit copy flags */
32
33
#define SDL_COPY_MODULATE_COLOR 0x00000001
#define SDL_COPY_MODULATE_ALPHA 0x00000002
34
35
#define SDL_COPY_BLEND 0x00000010
#define SDL_COPY_ADD 0x00000020
36
#define SDL_COPY_MOD 0x00000040
37
38
39
40
41
#define SDL_COPY_COLORKEY 0x00000100
#define SDL_COPY_NEAREST 0x00000200
#define SDL_COPY_RLE_DESIRED 0x00001000
#define SDL_COPY_RLE_COLORKEY 0x00002000
#define SDL_COPY_RLE_ALPHAKEY 0x00004000
42
#define SDL_COPY_RLE_MASK (SDL_COPY_RLE_DESIRED|SDL_COPY_RLE_COLORKEY|SDL_COPY_RLE_ALPHAKEY)
43
44
/* SDL blit CPU flags */
45
46
#define SDL_CPU_ANY 0x00000000
#define SDL_CPU_MMX 0x00000001
47
#define SDL_CPU_3DNOW 0x00000002
48
49
#define SDL_CPU_SSE 0x00000004
#define SDL_CPU_SSE2 0x00000008
50
51
#define SDL_CPU_ALTIVEC_PREFETCH 0x00000010
#define SDL_CPU_ALTIVEC_NOPREFETCH 0x00000020
52
53
54
typedef struct
{
55
56
57
Uint8 *src;
int src_w, src_h;
int src_pitch;
58
int src_skip;
59
60
61
Uint8 *dst;
int dst_w, dst_h;
int dst_pitch;
62
int dst_skip;
63
64
SDL_PixelFormat *src_fmt;
SDL_PixelFormat *dst_fmt;
65
Uint8 *table;
66
67
68
int flags;
Uint32 colorkey;
Uint8 r, g, b, a;
69
70
} SDL_BlitInfo;
71
typedef void (SDLCALL * SDL_BlitFunc) (SDL_BlitInfo * info);
72
73
74
typedef struct
{
75
76
77
78
79
80
Uint32 src_format;
Uint32 dst_format;
int flags;
int cpu;
SDL_BlitFunc func;
} SDL_BlitFuncEntry;
81
82
/* Blit mapping definition */
83
84
85
86
typedef struct SDL_BlitMap
{
SDL_Surface *dst;
int identity;
87
88
SDL_blit blit;
void *data;
89
SDL_BlitInfo info;
90
91
92
/* the version count matches the destination; mismatch indicates
an invalid mapping */
93
Uint32 palette_version;
94
95
96
} SDL_BlitMap;
/* Functions found in SDL_blit.c */
97
extern int SDL_CalculateBlit(SDL_Surface * surface);
98
99
100
101
102
103
/* Functions found in SDL_blit_*.c */
extern SDL_BlitFunc SDL_CalculateBlit0(SDL_Surface * surface);
extern SDL_BlitFunc SDL_CalculateBlit1(SDL_Surface * surface);
extern SDL_BlitFunc SDL_CalculateBlitN(SDL_Surface * surface);
extern SDL_BlitFunc SDL_CalculateBlitA(SDL_Surface * surface);
104
105
106
107
108
/*
* Useful macros for blitting routines
*/
109
110
111
#if defined(__GNUC__)
#define DECLARE_ALIGNED(t,v,a) t __attribute__((aligned(a))) v
#elif defined(_MSC_VER)
112
#define DECLARE_ALIGNED(t,v,a) __declspec(align(a)) t v
113
114
115
116
#else
#define DECLARE_ALIGNED(t,v,a) t v
#endif
117
118
119
120
121
122
123
124
125
126
/* The Nintendo surfaces are special. Bit 15 is the transparency
* bit. It must be set for the pixel to be displayed. By setting that
* value to 0 for other platforms, their compiler should optimize it
* out. */
#ifdef __NDS__
#define NDS_BIT15 0x8000
#else
#define NDS_BIT15 0
#endif
127
128
/* Load pixel of the specified format from a buffer and get its R-G-B values */
/* FIXME: rescale values to 0..255 here? */
129
#define RGB_FROM_PIXEL(Pixel, fmt, r, g, b) \
130
{ \
131
132
133
r = (((Pixel&fmt->Rmask)>>fmt->Rshift)<<fmt->Rloss); \
g = (((Pixel&fmt->Gmask)>>fmt->Gshift)<<fmt->Gloss); \
b = (((Pixel&fmt->Bmask)>>fmt->Bshift)<<fmt->Bloss); \
134
}
135
#define RGB_FROM_RGB565(Pixel, r, g, b) \
136
{ \
137
138
139
r = (((Pixel&0xF800)>>11)<<3); \
g = (((Pixel&0x07E0)>>5)<<2); \
b = ((Pixel&0x001F)<<3); \
140
}
141
#define RGB_FROM_RGB555(Pixel, r, g, b) \
142
{ \
143
144
145
r = (((Pixel&0x7C00)>>10)<<3); \
g = (((Pixel&0x03E0)>>5)<<3); \
b = ((Pixel&0x001F)<<3); \
146
}
147
#define RGB_FROM_RGB888(Pixel, r, g, b) \
148
{ \
149
150
151
r = ((Pixel&0xFF0000)>>16); \
g = ((Pixel&0xFF00)>>8); \
b = (Pixel&0xFF); \
152
}
153
#define RETRIEVE_RGB_PIXEL(buf, bpp, Pixel) \
154
155
156
do { \
switch (bpp) { \
case 2: \
157
Pixel = *((Uint16 *)(buf)); \
158
159
160
161
break; \
\
case 3: { \
Uint8 *B = (Uint8 *)(buf); \
162
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
163
Pixel = B[0] + (B[1] << 8) + (B[2] << 16); \
164
} else { \
165
Pixel = (B[0] << 16) + (B[1] << 8) + B[2]; \
166
167
168
169
170
} \
} \
break; \
\
case 4: \
171
Pixel = *((Uint32 *)(buf)); \
172
173
174
break; \
\
default: \
175
Pixel; /* stop gcc complaints */ \
176
177
break; \
} \
178
} while (0)
179
180
#define DISEMBLE_RGB(buf, bpp, fmt, Pixel, r, g, b) \
181
182
183
do { \
switch (bpp) { \
case 2: \
184
Pixel = *((Uint16 *)(buf)); \
185
RGB_FROM_PIXEL(Pixel, fmt, r, g, b); \
186
187
break; \
\
188
189
190
191
192
case 3: { \
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
r = *((buf)+fmt->Rshift/8); \
g = *((buf)+fmt->Gshift/8); \
b = *((buf)+fmt->Bshift/8); \
193
} else { \
194
195
196
r = *((buf)+2-fmt->Rshift/8); \
g = *((buf)+2-fmt->Gshift/8); \
b = *((buf)+2-fmt->Bshift/8); \
197
198
199
200
201
} \
} \
break; \
\
case 4: \
202
Pixel = *((Uint32 *)(buf)); \
203
RGB_FROM_PIXEL(Pixel, fmt, r, g, b); \
204
205
break; \
\
206
207
default: \
Pixel; /* stop gcc complaints */ \
208
209
break; \
} \
210
} while (0)
211
212
/* Assemble R-G-B values into a specified pixel format and store them */
213
#define PIXEL_FROM_RGB(Pixel, fmt, r, g, b) \
214
{ \
215
Pixel = ((r>>fmt->Rloss)<<fmt->Rshift)| \
216
217
218
((g>>fmt->Gloss)<<fmt->Gshift)| \
((b>>fmt->Bloss)<<fmt->Bshift); \
}
219
#define RGB565_FROM_RGB(Pixel, r, g, b) \
220
{ \
221
Pixel = ((r>>3)<<11)|((g>>2)<<5)|(b>>3); \
222
}
223
#define RGB555_FROM_RGB(Pixel, r, g, b) \
224
{ \
225
Pixel = ((r>>3)<<10)|((g>>3)<<5)|(b>>3); \
226
}
227
#define RGB888_FROM_RGB(Pixel, r, g, b) \
228
{ \
229
Pixel = (r<<16)|(g<<8)|b; \
230
}
231
232
233
234
#define ARGB8888_FROM_RGBA(Pixel, r, g, b, a) \
{ \
Pixel = (a<<24)|(r<<16)|(g<<8)|b; \
}
235
236
237
238
239
240
241
242
243
244
245
246
#define RGBA8888_FROM_RGBA(Pixel, r, g, b, a) \
{ \
Pixel = (r<<24)|(g<<16)|(b<<8)|a; \
}
#define ABGR8888_FROM_RGBA(Pixel, r, g, b, a) \
{ \
Pixel = (a<<24)|(b<<16)|(g<<8)|r; \
}
#define BGRA8888_FROM_RGBA(Pixel, r, g, b, a) \
{ \
Pixel = (b<<24)|(g<<16)|(r<<8)|a; \
}
247
248
249
250
#define ASSEMBLE_RGB(buf, bpp, fmt, r, g, b) \
{ \
switch (bpp) { \
case 2: { \
251
Uint16 Pixel; \
252
\
253
PIXEL_FROM_RGB(Pixel, fmt, r, g, b); \
254
*((Uint16 *)(buf)) = Pixel | NDS_BIT15; \
255
256
257
258
} \
break; \
\
case 3: { \
259
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
260
261
262
263
264
265
266
267
268
269
270
271
*((buf)+fmt->Rshift/8) = r; \
*((buf)+fmt->Gshift/8) = g; \
*((buf)+fmt->Bshift/8) = b; \
} else { \
*((buf)+2-fmt->Rshift/8) = r; \
*((buf)+2-fmt->Gshift/8) = g; \
*((buf)+2-fmt->Bshift/8) = b; \
} \
} \
break; \
\
case 4: { \
272
Uint32 Pixel; \
273
\
274
275
PIXEL_FROM_RGB(Pixel, fmt, r, g, b); \
*((Uint32 *)(buf)) = Pixel; \
276
277
278
279
280
281
282
283
284
} \
break; \
} \
}
#define ASSEMBLE_RGB_AMASK(buf, bpp, fmt, r, g, b, Amask) \
{ \
switch (bpp) { \
case 2: { \
Uint16 *bufp; \
285
Uint16 Pixel; \
286
287
\
bufp = (Uint16 *)buf; \
288
289
PIXEL_FROM_RGB(Pixel, fmt, r, g, b); \
*bufp = Pixel | (*bufp & Amask); \
290
291
292
293
} \
break; \
\
case 3: { \
294
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
295
296
297
298
299
300
301
302
303
304
305
306
307
*((buf)+fmt->Rshift/8) = r; \
*((buf)+fmt->Gshift/8) = g; \
*((buf)+fmt->Bshift/8) = b; \
} else { \
*((buf)+2-fmt->Rshift/8) = r; \
*((buf)+2-fmt->Gshift/8) = g; \
*((buf)+2-fmt->Bshift/8) = b; \
} \
} \
break; \
\
case 4: { \
Uint32 *bufp; \
308
Uint32 Pixel; \
309
310
\
bufp = (Uint32 *)buf; \
311
312
PIXEL_FROM_RGB(Pixel, fmt, r, g, b); \
*bufp = Pixel | (*bufp & Amask); \
313
314
315
316
317
318
} \
break; \
} \
}
/* FIXME: Should we rescale alpha into 0..255 here? */
319
#define RGBA_FROM_PIXEL(Pixel, fmt, r, g, b, a) \
320
{ \
321
322
323
324
r = ((Pixel&fmt->Rmask)>>fmt->Rshift)<<fmt->Rloss; \
g = ((Pixel&fmt->Gmask)>>fmt->Gshift)<<fmt->Gloss; \
b = ((Pixel&fmt->Bmask)>>fmt->Bshift)<<fmt->Bloss; \
a = ((Pixel&fmt->Amask)>>fmt->Ashift)<<fmt->Aloss; \
325
}
326
#define RGBA_FROM_8888(Pixel, fmt, r, g, b, a) \
327
{ \
328
329
330
331
r = (Pixel&fmt->Rmask)>>fmt->Rshift; \
g = (Pixel&fmt->Gmask)>>fmt->Gshift; \
b = (Pixel&fmt->Bmask)>>fmt->Bshift; \
a = (Pixel&fmt->Amask)>>fmt->Ashift; \
332
}
333
#define RGBA_FROM_RGBA8888(Pixel, r, g, b, a) \
334
{ \
335
336
337
338
r = (Pixel>>24); \
g = ((Pixel>>16)&0xFF); \
b = ((Pixel>>8)&0xFF); \
a = (Pixel&0xFF); \
339
}
340
#define RGBA_FROM_ARGB8888(Pixel, r, g, b, a) \
341
{ \
342
343
344
345
r = ((Pixel>>16)&0xFF); \
g = ((Pixel>>8)&0xFF); \
b = (Pixel&0xFF); \
a = (Pixel>>24); \
346
}
347
#define RGBA_FROM_ABGR8888(Pixel, r, g, b, a) \
348
{ \
349
350
351
352
r = (Pixel&0xFF); \
g = ((Pixel>>8)&0xFF); \
b = ((Pixel>>16)&0xFF); \
a = (Pixel>>24); \
353
}
354
355
356
357
358
359
360
#define RGBA_FROM_BGRA8888(Pixel, r, g, b, a) \
{ \
r = ((Pixel>>8)&0xFF); \
g = ((Pixel>>16)&0xFF); \
b = (Pixel>>24); \
a = (Pixel&0xFF); \
}
361
#define DISEMBLE_RGBA(buf, bpp, fmt, Pixel, r, g, b, a) \
362
363
364
do { \
switch (bpp) { \
case 2: \
365
Pixel = *((Uint16 *)(buf)); \
366
RGBA_FROM_PIXEL(Pixel, fmt, r, g, b, a); \
367
368
break; \
\
369
370
371
372
373
case 3: { \
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
r = *((buf)+fmt->Rshift/8); \
g = *((buf)+fmt->Gshift/8); \
b = *((buf)+fmt->Bshift/8); \
374
} else { \
375
376
377
r = *((buf)+2-fmt->Rshift/8); \
g = *((buf)+2-fmt->Gshift/8); \
b = *((buf)+2-fmt->Bshift/8); \
378
} \
379
a = 0xFF; \
380
381
382
383
} \
break; \
\
case 4: \
384
Pixel = *((Uint32 *)(buf)); \
385
RGBA_FROM_PIXEL(Pixel, fmt, r, g, b, a); \
386
387
388
break; \
\
default: \
389
Pixel; /* stop gcc complaints */ \
390
391
break; \
} \
392
} while (0)
393
394
/* FIXME: this isn't correct, especially for Alpha (maximum != 255) */
395
#define PIXEL_FROM_RGBA(Pixel, fmt, r, g, b, a) \
396
{ \
397
Pixel = ((r>>fmt->Rloss)<<fmt->Rshift)| \
398
399
((g>>fmt->Gloss)<<fmt->Gshift)| \
((b>>fmt->Bloss)<<fmt->Bshift)| \
400
((a>>fmt->Aloss)<<fmt->Ashift); \
401
402
403
404
405
}
#define ASSEMBLE_RGBA(buf, bpp, fmt, r, g, b, a) \
{ \
switch (bpp) { \
case 2: { \
406
Uint16 Pixel; \
407
\
408
PIXEL_FROM_RGBA(Pixel, fmt, r, g, b, a); \
409
*((Uint16 *)(buf)) = Pixel | NDS_BIT15; \
410
411
412
} \
break; \
\
413
414
case 3: { \
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
415
416
417
418
419
420
421
422
423
424
425
426
*((buf)+fmt->Rshift/8) = r; \
*((buf)+fmt->Gshift/8) = g; \
*((buf)+fmt->Bshift/8) = b; \
} else { \
*((buf)+2-fmt->Rshift/8) = r; \
*((buf)+2-fmt->Gshift/8) = g; \
*((buf)+2-fmt->Bshift/8) = b; \
} \
} \
break; \
\
case 4: { \
427
Uint32 Pixel; \
428
\
429
430
PIXEL_FROM_RGBA(Pixel, fmt, r, g, b, a); \
*((Uint32 *)(buf)) = Pixel; \
431
432
433
434
435
} \
break; \
} \
}
436
/* Blend the RGB values of two Pixels based on a source alpha value */
437
438
#define ALPHA_BLEND(sR, sG, sB, A, dR, dG, dB) \
do { \
439
440
441
dR = ((((int)(sR-dR)*(int)A)/255)+dR); \
dG = ((((int)(sG-dG)*(int)A)/255)+dG); \
dB = ((((int)(sB-dB)*(int)A)/255)+dB); \
442
443
444
} while(0)
445
/* This is a very useful loop for optimizing blitters */
446
447
448
#if defined(_MSC_VER) && (_MSC_VER == 1300)
/* There's a bug in the Visual C++ 7 optimizer when compiling this code */
#else
449
#define USE_DUFFS_LOOP
450
#endif
451
452
453
454
455
#ifdef USE_DUFFS_LOOP
/* 8-times unrolled loop */
#define DUFFS_LOOP8(pixel_copy_increment, width) \
{ int n = (width+7)/8; \
456
switch (width & 7) { \
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
case 0: do { pixel_copy_increment; \
case 7: pixel_copy_increment; \
case 6: pixel_copy_increment; \
case 5: pixel_copy_increment; \
case 4: pixel_copy_increment; \
case 3: pixel_copy_increment; \
case 2: pixel_copy_increment; \
case 1: pixel_copy_increment; \
} while ( --n > 0 ); \
} \
}
/* 4-times unrolled loop */
#define DUFFS_LOOP4(pixel_copy_increment, width) \
{ int n = (width+3)/4; \
472
switch (width & 3) { \
473
474
475
476
case 0: do { pixel_copy_increment; \
case 3: pixel_copy_increment; \
case 2: pixel_copy_increment; \
case 1: pixel_copy_increment; \
477
} while (--n > 0); \
478
479
480
} \
}
481
482
483
484
/* Use the 8-times version of the loop by default */
#define DUFFS_LOOP(pixel_copy_increment, width) \
DUFFS_LOOP8(pixel_copy_increment, width)
485
486
487
488
489
490
491
492
493
494
/* Special version of Duff's device for even more optimization */
#define DUFFS_LOOP_124(pixel_copy_increment1, \
pixel_copy_increment2, \
pixel_copy_increment4, width) \
{ int n = width; \
if (n & 1) { \
pixel_copy_increment1; n -= 1; \
} \
if (n & 2) { \
pixel_copy_increment2; n -= 2; \
495
} \
496
497
498
499
500
501
502
if (n) { \
n = (n+7)/ 8; \
switch (n & 4) { \
case 0: do { pixel_copy_increment4; \
case 4: pixel_copy_increment4; \
} while (--n > 0); \
} \
503
504
505
} \
}
506
507
#else
508
509
510
511
512
513
514
515
516
517
518
/* Don't use Duff's device to unroll loops */
#define DUFFS_LOOP(pixel_copy_increment, width) \
{ int n; \
for ( n=width; n > 0; --n ) { \
pixel_copy_increment; \
} \
}
#define DUFFS_LOOP8(pixel_copy_increment, width) \
DUFFS_LOOP(pixel_copy_increment, width)
#define DUFFS_LOOP4(pixel_copy_increment, width) \
DUFFS_LOOP(pixel_copy_increment, width)
519
520
521
522
#define DUFFS_LOOP_124(pixel_copy_increment1, \
pixel_copy_increment2, \
pixel_copy_increment4, width) \
DUFFS_LOOP(pixel_copy_increment1, width)
523
524
525
526
527
528
529
530
531
#endif /* USE_DUFFS_LOOP */
/* Prevent Visual C++ 6.0 from printing out stupid warnings */
#if defined(_MSC_VER) && (_MSC_VER >= 600)
#pragma warning(disable: 4550)
#endif
#endif /* _SDL_blit_h */
532
533
/* vi: set ts=4 sw=4 expandtab: */