This repository has been archived by the owner on Feb 11, 2021. It is now read-only.
/
SDL_blit.h
495 lines (459 loc) · 14.1 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
32
33
/* Table to do pixel byte expansion */
extern Uint8* SDL_expand_byte[9];
34
/* SDL blit copy flags */
35
36
#define SDL_COPY_MODULATE_COLOR 0x00000001
#define SDL_COPY_MODULATE_ALPHA 0x00000002
37
38
#define SDL_COPY_BLEND 0x00000010
#define SDL_COPY_ADD 0x00000020
39
#define SDL_COPY_MOD 0x00000040
40
41
42
43
44
#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
45
#define SDL_COPY_RLE_MASK (SDL_COPY_RLE_DESIRED|SDL_COPY_RLE_COLORKEY|SDL_COPY_RLE_ALPHAKEY)
46
47
/* SDL blit CPU flags */
48
49
#define SDL_CPU_ANY 0x00000000
#define SDL_CPU_MMX 0x00000001
50
#define SDL_CPU_3DNOW 0x00000002
51
52
#define SDL_CPU_SSE 0x00000004
#define SDL_CPU_SSE2 0x00000008
53
54
#define SDL_CPU_ALTIVEC_PREFETCH 0x00000010
#define SDL_CPU_ALTIVEC_NOPREFETCH 0x00000020
55
56
57
typedef struct
{
58
59
60
Uint8 *src;
int src_w, src_h;
int src_pitch;
61
int src_skip;
62
63
64
Uint8 *dst;
int dst_w, dst_h;
int dst_pitch;
65
int dst_skip;
66
67
SDL_PixelFormat *src_fmt;
SDL_PixelFormat *dst_fmt;
68
Uint8 *table;
69
70
71
int flags;
Uint32 colorkey;
Uint8 r, g, b, a;
72
73
} SDL_BlitInfo;
74
typedef void (SDLCALL * SDL_BlitFunc) (SDL_BlitInfo * info);
75
76
77
typedef struct
{
78
79
80
81
82
83
Uint32 src_format;
Uint32 dst_format;
int flags;
int cpu;
SDL_BlitFunc func;
} SDL_BlitFuncEntry;
84
85
/* Blit mapping definition */
86
87
88
89
typedef struct SDL_BlitMap
{
SDL_Surface *dst;
int identity;
90
91
SDL_blit blit;
void *data;
92
SDL_BlitInfo info;
93
94
95
/* the version count matches the destination; mismatch indicates
an invalid mapping */
96
Uint32 palette_version;
97
98
99
} SDL_BlitMap;
/* Functions found in SDL_blit.c */
100
extern int SDL_CalculateBlit(SDL_Surface * surface);
101
102
103
104
105
106
/* 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);
107
108
109
110
111
/*
* Useful macros for blitting routines
*/
112
113
114
#if defined(__GNUC__)
#define DECLARE_ALIGNED(t,v,a) t __attribute__((aligned(a))) v
#elif defined(_MSC_VER)
115
#define DECLARE_ALIGNED(t,v,a) __declspec(align(a)) t v
116
117
118
119
#else
#define DECLARE_ALIGNED(t,v,a) t v
#endif
120
/* Load pixel of the specified format from a buffer and get its R-G-B values */
121
#define RGB_FROM_PIXEL(Pixel, fmt, r, g, b) \
122
{ \
123
124
125
r = SDL_expand_byte[fmt->Rloss][((Pixel&fmt->Rmask)>>fmt->Rshift)]; \
g = SDL_expand_byte[fmt->Gloss][((Pixel&fmt->Gmask)>>fmt->Gshift)]; \
b = SDL_expand_byte[fmt->Bloss][((Pixel&fmt->Bmask)>>fmt->Bshift)]; \
126
}
127
#define RGB_FROM_RGB565(Pixel, r, g, b) \
128
{ \
129
130
131
r = SDL_expand_byte[3][((Pixel&0xF800)>>11)]; \
g = SDL_expand_byte[2][((Pixel&0x07E0)>>5)]; \
b = SDL_expand_byte[3][(Pixel&0x001F)]; \
132
}
133
#define RGB_FROM_RGB555(Pixel, r, g, b) \
134
{ \
135
136
137
r = SDL_expand_byte[3][((Pixel&0x7C00)>>10)]; \
g = SDL_expand_byte[3][((Pixel&0x03E0)>>5)]; \
b = SDL_expand_byte[3][(Pixel&0x001F)]; \
138
}
139
#define RGB_FROM_RGB888(Pixel, r, g, b) \
140
{ \
141
142
143
r = ((Pixel&0xFF0000)>>16); \
g = ((Pixel&0xFF00)>>8); \
b = (Pixel&0xFF); \
144
}
145
#define RETRIEVE_RGB_PIXEL(buf, bpp, Pixel) \
146
147
148
do { \
switch (bpp) { \
case 2: \
149
Pixel = *((Uint16 *)(buf)); \
150
151
152
153
break; \
\
case 3: { \
Uint8 *B = (Uint8 *)(buf); \
154
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
155
Pixel = B[0] + (B[1] << 8) + (B[2] << 16); \
156
} else { \
157
Pixel = (B[0] << 16) + (B[1] << 8) + B[2]; \
158
159
160
161
162
} \
} \
break; \
\
case 4: \
163
Pixel = *((Uint32 *)(buf)); \
164
165
166
break; \
\
default: \
167
Pixel = 0; /* stop gcc complaints */ \
168
169
break; \
} \
170
} while (0)
171
172
#define DISEMBLE_RGB(buf, bpp, fmt, Pixel, r, g, b) \
173
174
175
do { \
switch (bpp) { \
case 2: \
176
Pixel = *((Uint16 *)(buf)); \
177
RGB_FROM_PIXEL(Pixel, fmt, r, g, b); \
178
179
break; \
\
180
case 3: { \
181
Pixel = 0; \
182
183
184
185
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
r = *((buf)+fmt->Rshift/8); \
g = *((buf)+fmt->Gshift/8); \
b = *((buf)+fmt->Bshift/8); \
186
} else { \
187
188
189
r = *((buf)+2-fmt->Rshift/8); \
g = *((buf)+2-fmt->Gshift/8); \
b = *((buf)+2-fmt->Bshift/8); \
190
191
192
193
194
} \
} \
break; \
\
case 4: \
195
Pixel = *((Uint32 *)(buf)); \
196
RGB_FROM_PIXEL(Pixel, fmt, r, g, b); \
197
198
break; \
\
199
default: \
200
201
202
/* stop gcc complaints */ \
Pixel = 0; \
r = g = b = 0; \
203
204
break; \
} \
205
} while (0)
206
207
/* Assemble R-G-B values into a specified pixel format and store them */
208
#define PIXEL_FROM_RGB(Pixel, fmt, r, g, b) \
209
{ \
210
Pixel = ((r>>fmt->Rloss)<<fmt->Rshift)| \
211
((g>>fmt->Gloss)<<fmt->Gshift)| \
212
213
((b>>fmt->Bloss)<<fmt->Bshift)|
fmt->Amask; \
214
}
215
#define RGB565_FROM_RGB(Pixel, r, g, b) \
216
{ \
217
Pixel = ((r>>3)<<11)|((g>>2)<<5)|(b>>3); \
218
}
219
#define RGB555_FROM_RGB(Pixel, r, g, b) \
220
{ \
221
Pixel = ((r>>3)<<10)|((g>>3)<<5)|(b>>3); \
222
}
223
#define RGB888_FROM_RGB(Pixel, r, g, b) \
224
{ \
225
Pixel = (r<<16)|(g<<8)|b; \
226
}
227
228
229
230
#define ARGB8888_FROM_RGBA(Pixel, r, g, b, a) \
{ \
Pixel = (a<<24)|(r<<16)|(g<<8)|b; \
}
231
232
233
234
235
236
237
238
239
240
241
242
#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; \
}
243
244
245
246
#define ASSEMBLE_RGB(buf, bpp, fmt, r, g, b) \
{ \
switch (bpp) { \
case 2: { \
247
Uint16 Pixel; \
248
\
249
PIXEL_FROM_RGB(Pixel, fmt, r, g, b); \
250
*((Uint16 *)(buf)) = Pixel; \
251
252
253
254
} \
break; \
\
case 3: { \
255
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
256
257
258
259
260
261
262
263
264
265
266
267
*((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: { \
268
Uint32 Pixel; \
269
\
270
271
PIXEL_FROM_RGB(Pixel, fmt, r, g, b); \
*((Uint32 *)(buf)) = Pixel; \
272
273
274
275
276
277
} \
break; \
} \
}
/* FIXME: Should we rescale alpha into 0..255 here? */
278
#define RGBA_FROM_PIXEL(Pixel, fmt, r, g, b, a) \
279
{ \
280
281
282
283
r = SDL_expand_byte[fmt->Rloss][((Pixel&fmt->Rmask)>>fmt->Rshift)]; \
g = SDL_expand_byte[fmt->Gloss][((Pixel&fmt->Gmask)>>fmt->Gshift)]; \
b = SDL_expand_byte[fmt->Bloss][((Pixel&fmt->Bmask)>>fmt->Bshift)]; \
a = SDL_expand_byte[fmt->Aloss][((Pixel&fmt->Amask)>>fmt->Ashift)]; \
284
}
285
#define RGBA_FROM_8888(Pixel, fmt, r, g, b, a) \
286
{ \
287
288
289
290
r = (Pixel&fmt->Rmask)>>fmt->Rshift; \
g = (Pixel&fmt->Gmask)>>fmt->Gshift; \
b = (Pixel&fmt->Bmask)>>fmt->Bshift; \
a = (Pixel&fmt->Amask)>>fmt->Ashift; \
291
}
292
#define RGBA_FROM_RGBA8888(Pixel, r, g, b, a) \
293
{ \
294
295
296
297
r = (Pixel>>24); \
g = ((Pixel>>16)&0xFF); \
b = ((Pixel>>8)&0xFF); \
a = (Pixel&0xFF); \
298
}
299
#define RGBA_FROM_ARGB8888(Pixel, r, g, b, a) \
300
{ \
301
302
303
304
r = ((Pixel>>16)&0xFF); \
g = ((Pixel>>8)&0xFF); \
b = (Pixel&0xFF); \
a = (Pixel>>24); \
305
}
306
#define RGBA_FROM_ABGR8888(Pixel, r, g, b, a) \
307
{ \
308
309
310
311
r = (Pixel&0xFF); \
g = ((Pixel>>8)&0xFF); \
b = ((Pixel>>16)&0xFF); \
a = (Pixel>>24); \
312
}
313
314
315
316
317
318
319
#define RGBA_FROM_BGRA8888(Pixel, r, g, b, a) \
{ \
r = ((Pixel>>8)&0xFF); \
g = ((Pixel>>16)&0xFF); \
b = (Pixel>>24); \
a = (Pixel&0xFF); \
}
320
#define DISEMBLE_RGBA(buf, bpp, fmt, Pixel, r, g, b, a) \
321
322
323
do { \
switch (bpp) { \
case 2: \
324
Pixel = *((Uint16 *)(buf)); \
325
RGBA_FROM_PIXEL(Pixel, fmt, r, g, b, a); \
326
327
break; \
\
328
case 3: { \
329
Pixel = 0; \
330
331
332
333
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
r = *((buf)+fmt->Rshift/8); \
g = *((buf)+fmt->Gshift/8); \
b = *((buf)+fmt->Bshift/8); \
334
} else { \
335
336
337
r = *((buf)+2-fmt->Rshift/8); \
g = *((buf)+2-fmt->Gshift/8); \
b = *((buf)+2-fmt->Bshift/8); \
338
} \
339
a = 0xFF; \
340
341
342
343
} \
break; \
\
case 4: \
344
Pixel = *((Uint32 *)(buf)); \
345
RGBA_FROM_PIXEL(Pixel, fmt, r, g, b, a); \
346
347
348
break; \
\
default: \
349
350
351
/* stop gcc complaints */ \
Pixel = 0; \
r = g = b = a = 0; \
352
353
break; \
} \
354
} while (0)
355
356
/* FIXME: this isn't correct, especially for Alpha (maximum != 255) */
357
#define PIXEL_FROM_RGBA(Pixel, fmt, r, g, b, a) \
358
{ \
359
Pixel = ((r>>fmt->Rloss)<<fmt->Rshift)| \
360
361
((g>>fmt->Gloss)<<fmt->Gshift)| \
((b>>fmt->Bloss)<<fmt->Bshift)| \
362
((a>>fmt->Aloss)<<fmt->Ashift); \
363
364
365
366
367
}
#define ASSEMBLE_RGBA(buf, bpp, fmt, r, g, b, a) \
{ \
switch (bpp) { \
case 2: { \
368
Uint16 Pixel; \
369
\
370
PIXEL_FROM_RGBA(Pixel, fmt, r, g, b, a); \
371
*((Uint16 *)(buf)) = Pixel; \
372
373
374
} \
break; \
\
375
376
case 3: { \
if (SDL_BYTEORDER == SDL_LIL_ENDIAN) { \
377
378
379
380
381
382
383
384
385
386
387
388
*((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: { \
389
Uint32 Pixel; \
390
\
391
392
PIXEL_FROM_RGBA(Pixel, fmt, r, g, b, a); \
*((Uint32 *)(buf)) = Pixel; \
393
394
395
396
397
} \
break; \
} \
}
398
/* Blend the RGB values of two Pixels based on a source alpha value */
399
400
#define ALPHA_BLEND(sR, sG, sB, A, dR, dG, dB) \
do { \
401
402
403
dR = ((((int)(sR-dR)*(int)A)/255)+dR); \
dG = ((((int)(sG-dG)*(int)A)/255)+dG); \
dB = ((((int)(sB-dB)*(int)A)/255)+dB); \
404
405
406
} while(0)
407
/* This is a very useful loop for optimizing blitters */
408
409
410
#if defined(_MSC_VER) && (_MSC_VER == 1300)
/* There's a bug in the Visual C++ 7 optimizer when compiling this code */
#else
411
#define USE_DUFFS_LOOP
412
#endif
413
414
415
416
417
#ifdef USE_DUFFS_LOOP
/* 8-times unrolled loop */
#define DUFFS_LOOP8(pixel_copy_increment, width) \
{ int n = (width+7)/8; \
418
switch (width & 7) { \
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
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; \
434
switch (width & 3) { \
435
436
437
438
case 0: do { pixel_copy_increment; \
case 3: pixel_copy_increment; \
case 2: pixel_copy_increment; \
case 1: pixel_copy_increment; \
439
} while (--n > 0); \
440
441
442
} \
}
443
444
445
446
/* Use the 8-times version of the loop by default */
#define DUFFS_LOOP(pixel_copy_increment, width) \
DUFFS_LOOP8(pixel_copy_increment, width)
447
448
449
450
451
452
453
454
455
456
/* 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; \
457
} \
458
459
460
461
462
463
464
if (n) { \
n = (n+7)/ 8; \
switch (n & 4) { \
case 0: do { pixel_copy_increment4; \
case 4: pixel_copy_increment4; \
} while (--n > 0); \
} \
465
466
467
} \
}
468
469
#else
470
471
472
473
474
475
476
477
478
479
480
/* 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)
481
482
483
484
#define DUFFS_LOOP_124(pixel_copy_increment1, \
pixel_copy_increment2, \
pixel_copy_increment4, width) \
DUFFS_LOOP(pixel_copy_increment1, width)
485
486
487
488
489
490
491
492
493
#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 */
494
495
/* vi: set ts=4 sw=4 expandtab: */