aboutsummaryrefslogtreecommitdiff
path: root/mesalib/src/mesa/swrast/s_texcombine.c
blob: df3a74a06adee0c0563fbaa2d3593660bf279bb4 (plain)
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
/*
 * Mesa 3-D graphics library
 * Version:  7.5
 *
 * Copyright (C) 1999-2008  Brian Paul   All Rights Reserved.
 * Copyright (C) 2009  VMware, Inc.   All Rights Reserved.
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included
 * in all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
 * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR
 * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
 * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
 * OTHER DEALINGS IN THE SOFTWARE.
 */


#include "main/glheader.h"
#include "main/context.h"
#include "main/colormac.h"
#include "main/imports.h"
#include "main/pixeltransfer.h"
#include "main/samplerobj.h"
#include "program/prog_instruction.h"

#include "s_context.h"
#include "s_texcombine.h"


/**
 * Pointer to array of float[4]
 * This type makes the code below more concise and avoids a lot of casting.
 */
typedef float (*float4_array)[4];


/**
 * Return array of texels for given unit.
 */
static inline float4_array
get_texel_array(SWcontext *swrast, GLuint unit)
{
#ifdef _OPENMP
   return (float4_array) (swrast->TexelBuffer + unit * SWRAST_MAX_WIDTH * 4 * omp_get_num_threads() + (SWRAST_MAX_WIDTH * 4 * omp_get_thread_num()));
#else
   return (float4_array) (swrast->TexelBuffer + unit * SWRAST_MAX_WIDTH * 4);
#endif
}



/**
 * Do texture application for:
 *  GL_EXT_texture_env_combine
 *  GL_ARB_texture_env_combine
 *  GL_EXT_texture_env_dot3
 *  GL_ARB_texture_env_dot3
 *  GL_ATI_texture_env_combine3
 *  GL_NV_texture_env_combine4
 *  conventional GL texture env modes
 *
 * \param ctx          rendering context
 * \param unit         the texture combiner unit
 * \param primary_rgba incoming fragment color array
 * \param texelBuffer  pointer to texel colors for all texture units
 * 
 * \param span         two fields are used in this function:
 *                       span->end: number of fragments to process
 *                       span->array->rgba: incoming/result fragment colors
 */
static void
texture_combine( struct gl_context *ctx, GLuint unit,
                 const float4_array primary_rgba,
                 const GLfloat *texelBuffer,
                 SWspan *span )
{
   SWcontext *swrast = SWRAST_CONTEXT(ctx);
   const struct gl_texture_unit *textureUnit = &(ctx->Texture.Unit[unit]);
   const struct gl_tex_env_combine_state *combine = textureUnit->_CurrentCombine;
   float4_array argRGB[MAX_COMBINER_TERMS];
   float4_array argA[MAX_COMBINER_TERMS];
   const GLfloat scaleRGB = (GLfloat) (1 << combine->ScaleShiftRGB);
   const GLfloat scaleA = (GLfloat) (1 << combine->ScaleShiftA);
   const GLuint numArgsRGB = combine->_NumArgsRGB;
   const GLuint numArgsA = combine->_NumArgsA;
   float4_array ccolor[4], rgba;
   GLuint i, term;
   GLuint n = span->end;
   GLchan (*rgbaChan)[4] = span->array->rgba;

   /* alloc temp pixel buffers */
   rgba = malloc(4 * n * sizeof(GLfloat));
   if (!rgba) {
      _mesa_error(ctx, GL_OUT_OF_MEMORY, "texture_combine");
      return;
   }

   for (i = 0; i < numArgsRGB || i < numArgsA; i++) {
      ccolor[i] = malloc(4 * n * sizeof(GLfloat));
      if (!ccolor[i]) {
         while (i) {
            free(ccolor[i]);
            i--;
         }
         _mesa_error(ctx, GL_OUT_OF_MEMORY, "texture_combine");
         free(rgba);
         return;
      }
   }

   for (i = 0; i < n; i++) {
      rgba[i][RCOMP] = CHAN_TO_FLOAT(rgbaChan[i][RCOMP]);
      rgba[i][GCOMP] = CHAN_TO_FLOAT(rgbaChan[i][GCOMP]);
      rgba[i][BCOMP] = CHAN_TO_FLOAT(rgbaChan[i][BCOMP]);
      rgba[i][ACOMP] = CHAN_TO_FLOAT(rgbaChan[i][ACOMP]);
   }

   /*
   printf("modeRGB 0x%x  modeA 0x%x  srcRGB1 0x%x  srcA1 0x%x  srcRGB2 0x%x  srcA2 0x%x\n",
          combine->ModeRGB,
          combine->ModeA,
          combine->SourceRGB[0],
          combine->SourceA[0],
          combine->SourceRGB[1],
          combine->SourceA[1]);
   */

   /*
    * Do operand setup for up to 4 operands.  Loop over the terms.
    */
   for (term = 0; term < numArgsRGB; term++) {
      const GLenum srcRGB = combine->SourceRGB[term];
      const GLenum operandRGB = combine->OperandRGB[term];

      switch (srcRGB) {
         case GL_TEXTURE:
            argRGB[term] = get_texel_array(swrast, unit);
            break;
         case GL_PRIMARY_COLOR:
            argRGB[term] = primary_rgba;
            break;
         case GL_PREVIOUS:
            argRGB[term] = rgba;
            break;
         case GL_CONSTANT:
            {
               float4_array c = ccolor[term];
               GLfloat red   = textureUnit->EnvColor[0];
               GLfloat green = textureUnit->EnvColor[1];
               GLfloat blue  = textureUnit->EnvColor[2];
               GLfloat alpha = textureUnit->EnvColor[3];
               for (i = 0; i < n; i++) {
                  ASSIGN_4V(c[i], red, green, blue, alpha);
               }
               argRGB[term] = ccolor[term];
            }
            break;
	 /* GL_ATI_texture_env_combine3 allows GL_ZERO & GL_ONE as sources.
	  */
	 case GL_ZERO:
            {
               float4_array c = ccolor[term];
               for (i = 0; i < n; i++) {
                  ASSIGN_4V(c[i], 0.0F, 0.0F, 0.0F, 0.0F);
               }
               argRGB[term] = ccolor[term];
            }
            break;
	 case GL_ONE:
            {
               float4_array c = ccolor[term];
               for (i = 0; i < n; i++) {
                  ASSIGN_4V(c[i], 1.0F, 1.0F, 1.0F, 1.0F);
               }
               argRGB[term] = ccolor[term];
            }
            break;
         default:
            /* ARB_texture_env_crossbar source */
            {
               const GLuint srcUnit = srcRGB - GL_TEXTURE0;
               ASSERT(srcUnit < ctx->Const.MaxTextureUnits);
               if (!ctx->Texture.Unit[srcUnit]._ReallyEnabled)
                  goto end;
               argRGB[term] = get_texel_array(swrast, srcUnit);
            }
      }

      if (operandRGB != GL_SRC_COLOR) {
         float4_array src = argRGB[term];
         float4_array dst = ccolor[term];

         /* point to new arg[term] storage */
         argRGB[term] = ccolor[term];

         switch (operandRGB) {
         case GL_ONE_MINUS_SRC_COLOR:
            for (i = 0; i < n; i++) {
               dst[i][RCOMP] = 1.0F - src[i][RCOMP];
               dst[i][GCOMP] = 1.0F - src[i][GCOMP];
               dst[i][BCOMP] = 1.0F - src[i][BCOMP];
            }
            break;
         case GL_SRC_ALPHA:
            for (i = 0; i < n; i++) {
               dst[i][RCOMP] =
               dst[i][GCOMP] =
               dst[i][BCOMP] = src[i][ACOMP];
            }
            break;
         case GL_ONE_MINUS_SRC_ALPHA:
            for (i = 0; i < n; i++) {
               dst[i][RCOMP] =
               dst[i][GCOMP] =
               dst[i][BCOMP] = 1.0F - src[i][ACOMP];
            }
            break;
         default:
            _mesa_problem(ctx, "Bad operandRGB");
         }
      }
   }

   /*
    * Set up the argA[term] pointers
    */
   for (term = 0; term < numArgsA; term++) {
      const GLenum srcA = combine->SourceA[term];
      const GLenum operandA = combine->OperandA[term];

      switch (srcA) {
         case GL_TEXTURE:
            argA[term] = get_texel_array(swrast, unit);
            break;
         case GL_PRIMARY_COLOR:
            argA[term] = primary_rgba;
            break;
         case GL_PREVIOUS:
            argA[term] = rgba;
            break;
         case GL_CONSTANT:
            {
               float4_array c = ccolor[term];
               GLfloat alpha = textureUnit->EnvColor[3];
               for (i = 0; i < n; i++)
                  c[i][ACOMP] = alpha;
               argA[term] = ccolor[term];
            }
            break;
	 /* GL_ATI_texture_env_combine3 allows GL_ZERO & GL_ONE as sources.
	  */
	 case GL_ZERO:
            {
               float4_array c = ccolor[term];
               for (i = 0; i < n; i++)
                  c[i][ACOMP] = 0.0F;
               argA[term] = ccolor[term];
            }
            break;
	 case GL_ONE:
            {
               float4_array c = ccolor[term];
               for (i = 0; i < n; i++)
                  c[i][ACOMP] = 1.0F;
               argA[term] = ccolor[term];
            }
            break;
         default:
            /* ARB_texture_env_crossbar source */
            {
               const GLuint srcUnit = srcA - GL_TEXTURE0;
               ASSERT(srcUnit < ctx->Const.MaxTextureUnits);
               if (!ctx->Texture.Unit[srcUnit]._ReallyEnabled)
                  goto end;
               argA[term] = get_texel_array(swrast, srcUnit);
            }
      }

      if (operandA == GL_ONE_MINUS_SRC_ALPHA) {
         float4_array src = argA[term];
         float4_array dst = ccolor[term];
         argA[term] = ccolor[term];
         for (i = 0; i < n; i++) {
            dst[i][ACOMP] = 1.0F - src[i][ACOMP];
         }
      }
   }

   /* RGB channel combine */
   {
      float4_array arg0 = argRGB[0];
      float4_array arg1 = argRGB[1];
      float4_array arg2 = argRGB[2];
      float4_array arg3 = argRGB[3];

      switch (combine->ModeRGB) {
      case GL_REPLACE:
         for (i = 0; i < n; i++) {
            rgba[i][RCOMP] = arg0[i][RCOMP] * scaleRGB;
            rgba[i][GCOMP] = arg0[i][GCOMP] * scaleRGB;
            rgba[i][BCOMP] = arg0[i][BCOMP] * scaleRGB;
         }
         break;
      case GL_MODULATE:
         for (i = 0; i < n; i++) {
            rgba[i][RCOMP] = arg0[i][RCOMP] * arg1[i][RCOMP] * scaleRGB;
            rgba[i][GCOMP] = arg0[i][GCOMP] * arg1[i][GCOMP] * scaleRGB;
            rgba[i][BCOMP] = arg0[i][BCOMP] * arg1[i][BCOMP] * scaleRGB;
         }
         break;
      case GL_ADD:
         if (textureUnit->EnvMode == GL_COMBINE4_NV) {
            /* (a * b) + (c * d) */
            for (i = 0; i < n; i++) {
               rgba[i][RCOMP] = (arg0[i][RCOMP] * arg1[i][RCOMP] +
                                 arg2[i][RCOMP] * arg3[i][RCOMP]) * scaleRGB;
               rgba[i][GCOMP] = (arg0[i][GCOMP] * arg1[i][GCOMP] +
                                 arg2[i][GCOMP] * arg3[i][GCOMP]) * scaleRGB;
               rgba[i][BCOMP] = (arg0[i][BCOMP] * arg1[i][BCOMP] +
                                 arg2[i][BCOMP] * arg3[i][BCOMP]) * scaleRGB;
            }
         }
         else {
            /* 2-term addition */
            for (i = 0; i < n; i++) {
               rgba[i][RCOMP] = (arg0[i][RCOMP] + arg1[i][RCOMP]) * scaleRGB;
               rgba[i][GCOMP] = (arg0[i][GCOMP] + arg1[i][GCOMP]) * scaleRGB;
               rgba[i][BCOMP] = (arg0[i][BCOMP] + arg1[i][BCOMP]) * scaleRGB;
            }
         }
         break;
      case GL_ADD_SIGNED:
         if (textureUnit->EnvMode == GL_COMBINE4_NV) {
            /* (a * b) + (c * d) - 0.5 */
            for (i = 0; i < n; i++) {
               rgba[i][RCOMP] = (arg0[i][RCOMP] * arg1[i][RCOMP] +
                                 arg2[i][RCOMP] * arg3[i][RCOMP] - 0.5F) * scaleRGB;
               rgba[i][GCOMP] = (arg0[i][GCOMP] * arg1[i][GCOMP] +
                                 arg2[i][GCOMP] * arg3[i][GCOMP] - 0.5F) * scaleRGB;
               rgba[i][BCOMP] = (arg0[i][BCOMP] * arg1[i][BCOMP] +
                                 arg2[i][BCOMP] * arg3[i][BCOMP] - 0.5F) * scaleRGB;
            }
         }
         else {
            for (i = 0; i < n; i++) {
               rgba[i][RCOMP] = (arg0[i][RCOMP] + arg1[i][RCOMP] - 0.5F) * scaleRGB;
               rgba[i][GCOMP] = (arg0[i][GCOMP] + arg1[i][GCOMP] - 0.5F) * scaleRGB;
               rgba[i][BCOMP] = (arg0[i][BCOMP] + arg1[i][BCOMP] - 0.5F) * scaleRGB;
            }
         }
         break;
      case GL_INTERPOLATE:
         for (i = 0; i < n; i++) {
            rgba[i][RCOMP] = (arg0[i][RCOMP] * arg2[i][RCOMP] +
                          arg1[i][RCOMP] * (1.0F - arg2[i][RCOMP])) * scaleRGB;
            rgba[i][GCOMP] = (arg0[i][GCOMP] * arg2[i][GCOMP] +
                          arg1[i][GCOMP] * (1.0F - arg2[i][GCOMP])) * scaleRGB;
            rgba[i][BCOMP] = (arg0[i][BCOMP] * arg2[i][BCOMP] +
                          arg1[i][BCOMP] * (1.0F - arg2[i][BCOMP])) * scaleRGB;
         }
         break;
      case GL_SUBTRACT:
         for (i = 0; i < n; i++) {
            rgba[i][RCOMP] = (arg0[i][RCOMP] - arg1[i][RCOMP]) * scaleRGB;
            rgba[i][GCOMP] = (arg0[i][GCOMP] - arg1[i][GCOMP]) * scaleRGB;
            rgba[i][BCOMP] = (arg0[i][BCOMP] - arg1[i][BCOMP]) * scaleRGB;
         }
         break;
      case GL_DOT3_RGB_EXT:
      case GL_DOT3_RGBA_EXT:
         /* Do not scale the result by 1 2 or 4 */
         for (i = 0; i < n; i++) {
            GLfloat dot = ((arg0[i][RCOMP] - 0.5F) * (arg1[i][RCOMP] - 0.5F) +
                           (arg0[i][GCOMP] - 0.5F) * (arg1[i][GCOMP] - 0.5F) +
                           (arg0[i][BCOMP] - 0.5F) * (arg1[i][BCOMP] - 0.5F))
               * 4.0F;
            dot = CLAMP(dot, 0.0F, 1.0F);
            rgba[i][RCOMP] = rgba[i][GCOMP] = rgba[i][BCOMP] = dot;
         }
         break;
      case GL_DOT3_RGB:
      case GL_DOT3_RGBA:
         /* DO scale the result by 1 2 or 4 */
         for (i = 0; i < n; i++) {
            GLfloat dot = ((arg0[i][RCOMP] - 0.5F) * (arg1[i][RCOMP] - 0.5F) +
                           (arg0[i][GCOMP] - 0.5F) * (arg1[i][GCOMP] - 0.5F) +
                           (arg0[i][BCOMP] - 0.5F) * (arg1[i][BCOMP] - 0.5F))
               * 4.0F * scaleRGB;
            dot = CLAMP(dot, 0.0F, 1.0F);
            rgba[i][RCOMP] = rgba[i][GCOMP] = rgba[i][BCOMP] = dot;
         }
         break;
      case GL_MODULATE_ADD_ATI:
         for (i = 0; i < n; i++) {
            rgba[i][RCOMP] = ((arg0[i][RCOMP] * arg2[i][RCOMP]) +
                              arg1[i][RCOMP]) * scaleRGB;
            rgba[i][GCOMP] = ((arg0[i][GCOMP] * arg2[i][GCOMP]) +
                              arg1[i][GCOMP]) * scaleRGB;
            rgba[i][BCOMP] = ((arg0[i][BCOMP] * arg2[i][BCOMP]) +
                              arg1[i][BCOMP]) * scaleRGB;
	 }
         break;
      case GL_MODULATE_SIGNED_ADD_ATI:
         for (i = 0; i < n; i++) {
            rgba[i][RCOMP] = ((arg0[i][RCOMP] * arg2[i][RCOMP]) +
                              arg1[i][RCOMP] - 0.5F) * scaleRGB;
            rgba[i][GCOMP] = ((arg0[i][GCOMP] * arg2[i][GCOMP]) +
                              arg1[i][GCOMP] - 0.5F) * scaleRGB;
            rgba[i][BCOMP] = ((arg0[i][BCOMP] * arg2[i][BCOMP]) +
                              arg1[i][BCOMP] - 0.5F) * scaleRGB;
	 }
         break;
      case GL_MODULATE_SUBTRACT_ATI:
         for (i = 0; i < n; i++) {
            rgba[i][RCOMP] = ((arg0[i][RCOMP] * arg2[i][RCOMP]) -
                              arg1[i][RCOMP]) * scaleRGB;
            rgba[i][GCOMP] = ((arg0[i][GCOMP] * arg2[i][GCOMP]) -
                              arg1[i][GCOMP]) * scaleRGB;
            rgba[i][BCOMP] = ((arg0[i][BCOMP] * arg2[i][BCOMP]) -
                              arg1[i][BCOMP]) * scaleRGB;
	 }
         break;
      case GL_BUMP_ENVMAP_ATI:
         /* this produces a fixed rgba color, and the coord calc is done elsewhere */
         for (i = 0; i < n; i++) {
            /* rgba result is 0,0,0,1 */
            rgba[i][RCOMP] = 0.0;
            rgba[i][GCOMP] = 0.0;
            rgba[i][BCOMP] = 0.0;
            rgba[i][ACOMP] = 1.0;
	 }
         goto end; /* no alpha processing */
      default:
         _mesa_problem(ctx, "invalid combine mode");
      }
   }

   /* Alpha channel combine */
   {
      float4_array arg0 = argA[0];
      float4_array arg1 = argA[1];
      float4_array arg2 = argA[2];
      float4_array arg3 = argA[3];

      switch (combine->ModeA) {
      case GL_REPLACE:
         for (i = 0; i < n; i++) {
            rgba[i][ACOMP] = arg0[i][ACOMP] * scaleA;
         }
         break;
      case GL_MODULATE:
         for (i = 0; i < n; i++) {
            rgba[i][ACOMP] = arg0[i][ACOMP] * arg1[i][ACOMP] * scaleA;
         }
         break;
      case GL_ADD:
         if (textureUnit->EnvMode == GL_COMBINE4_NV) {
            /* (a * b) + (c * d) */
            for (i = 0; i < n; i++) {
               rgba[i][ACOMP] = (arg0[i][ACOMP] * arg1[i][ACOMP] +
                                 arg2[i][ACOMP] * arg3[i][ACOMP]) * scaleA;
            }
         }
         else {
            /* two-term add */
            for (i = 0; i < n; i++) {
               rgba[i][ACOMP] = (arg0[i][ACOMP] + arg1[i][ACOMP]) * scaleA;
            }
         }
         break;
      case GL_ADD_SIGNED:
         if (textureUnit->EnvMode == GL_COMBINE4_NV) {
            /* (a * b) + (c * d) - 0.5 */
            for (i = 0; i < n; i++) {
               rgba[i][ACOMP] = (arg0[i][ACOMP] * arg1[i][ACOMP] +
                                 arg2[i][ACOMP] * arg3[i][ACOMP] -
                                 0.5F) * scaleA;
            }
         }
         else {
            /* a + b - 0.5 */
            for (i = 0; i < n; i++) {
               rgba[i][ACOMP] = (arg0[i][ACOMP] + arg1[i][ACOMP] - 0.5F) * scaleA;
            }
         }
         break;
      case GL_INTERPOLATE:
         for (i = 0; i < n; i++) {
            rgba[i][ACOMP] = (arg0[i][ACOMP] * arg2[i][ACOMP] +
                              arg1[i][ACOMP] * (1.0F - arg2[i][ACOMP]))
               * scaleA;
         }
         break;
      case GL_SUBTRACT:
         for (i = 0; i < n; i++) {
            rgba[i][ACOMP] = (arg0[i][ACOMP] - arg1[i][ACOMP]) * scaleA;
         }
         break;
      case GL_MODULATE_ADD_ATI:
         for (i = 0; i < n; i++) {
            rgba[i][ACOMP] = ((arg0[i][ACOMP] * arg2[i][ACOMP])
                              + arg1[i][ACOMP]) * scaleA;
         }
         break;
      case GL_MODULATE_SIGNED_ADD_ATI:
         for (i = 0; i < n; i++) {
            rgba[i][ACOMP] = ((arg0[i][ACOMP] * arg2[i][ACOMP]) +
                              arg1[i][ACOMP] - 0.5F) * scaleA;
         }
         break;
      case GL_MODULATE_SUBTRACT_ATI:
         for (i = 0; i < n; i++) {
            rgba[i][ACOMP] = ((arg0[i][ACOMP] * arg2[i][ACOMP])
                              - arg1[i][ACOMP]) * scaleA;
         }
         break;
      default:
         _mesa_problem(ctx, "invalid combine mode");
      }
   }

   /* Fix the alpha component for GL_DOT3_RGBA_EXT/ARB combining.
    * This is kind of a kludge.  It would have been better if the spec
    * were written such that the GL_COMBINE_ALPHA value could be set to
    * GL_DOT3.
    */
   if (combine->ModeRGB == GL_DOT3_RGBA_EXT ||
       combine->ModeRGB == GL_DOT3_RGBA) {
      for (i = 0; i < n; i++) {
	 rgba[i][ACOMP] = rgba[i][RCOMP];
      }
   }

   for (i = 0; i < n; i++) {
      UNCLAMPED_FLOAT_TO_CHAN(rgbaChan[i][RCOMP], rgba[i][RCOMP]);
      UNCLAMPED_FLOAT_TO_CHAN(rgbaChan[i][GCOMP], rgba[i][GCOMP]);
      UNCLAMPED_FLOAT_TO_CHAN(rgbaChan[i][BCOMP], rgba[i][BCOMP]);
      UNCLAMPED_FLOAT_TO_CHAN(rgbaChan[i][ACOMP], rgba[i][ACOMP]);
   }
   /* The span->array->rgba values are of CHAN type so set
    * span->array->ChanType field accordingly.
    */
   span->array->ChanType = CHAN_TYPE;

end:
   for (i = 0; i < numArgsRGB || i < numArgsA; i++) {
      free(ccolor[i]);
   }
   free(rgba);
}


/**
 * Apply X/Y/Z/W/0/1 swizzle to an array of colors/texels.
 * See GL_EXT_texture_swizzle.
 */
static void
swizzle_texels(GLuint swizzle, GLuint count, float4_array texels)
{
   const GLuint swzR = GET_SWZ(swizzle, 0);
   const GLuint swzG = GET_SWZ(swizzle, 1);
   const GLuint swzB = GET_SWZ(swizzle, 2);
   const GLuint swzA = GET_SWZ(swizzle, 3);
   GLfloat vector[6];
   GLuint i;

   vector[SWIZZLE_ZERO] = 0;
   vector[SWIZZLE_ONE] = 1.0F;

   for (i = 0; i < count; i++) {
      vector[SWIZZLE_X] = texels[i][0];
      vector[SWIZZLE_Y] = texels[i][1];
      vector[SWIZZLE_Z] = texels[i][2];
      vector[SWIZZLE_W] = texels[i][3];
      texels[i][RCOMP] = vector[swzR];
      texels[i][GCOMP] = vector[swzG];
      texels[i][BCOMP] = vector[swzB];
      texels[i][ACOMP] = vector[swzA];
   }
}


/**
 * Apply texture mapping to a span of fragments.
 */
void
_swrast_texture_span( struct gl_context *ctx, SWspan *span )
{
   SWcontext *swrast = SWRAST_CONTEXT(ctx);
   float4_array primary_rgba;
   GLuint unit;

   if (!swrast->TexelBuffer) {
#ifdef _OPENMP
      const GLint maxThreads = omp_get_max_threads();
#else
      const GLint maxThreads = 1;
#endif

      /* TexelBuffer is also global and normally shared by all SWspan
       * instances; when running with multiple threads, create one per
       * thread.
       */
      swrast->TexelBuffer =
	 malloc(ctx->Const.FragmentProgram.MaxTextureImageUnits * maxThreads *
			    SWRAST_MAX_WIDTH * 4 * sizeof(GLfloat));
      if (!swrast->TexelBuffer) {
	 _mesa_error(ctx, GL_OUT_OF_MEMORY, "texture_combine");
	 return;
      }
   }

   primary_rgba = malloc(span->end * 4 * sizeof(GLfloat));

   if (!primary_rgba) {
      _mesa_error(ctx, GL_OUT_OF_MEMORY, "texture_span");
      return;
   }

   ASSERT(span->end <= SWRAST_MAX_WIDTH);

   /*
    * Save copy of the incoming fragment colors (the GL_PRIMARY_COLOR)
    */
   if (swrast->_TextureCombinePrimary) {
      GLuint i;
      for (i = 0; i < span->end; i++) {
         primary_rgba[i][RCOMP] = CHAN_TO_FLOAT(span->array->rgba[i][RCOMP]);
         primary_rgba[i][GCOMP] = CHAN_TO_FLOAT(span->array->rgba[i][GCOMP]);
         primary_rgba[i][BCOMP] = CHAN_TO_FLOAT(span->array->rgba[i][BCOMP]);
         primary_rgba[i][ACOMP] = CHAN_TO_FLOAT(span->array->rgba[i][ACOMP]);
      }
   }

   /* First must sample all bump maps */
   for (unit = 0; unit < ctx->Const.MaxTextureUnits; unit++) {
      const struct gl_texture_unit *texUnit = &ctx->Texture.Unit[unit];

      if (texUnit->_ReallyEnabled &&
         texUnit->_CurrentCombine->ModeRGB == GL_BUMP_ENVMAP_ATI) {
         const GLfloat (*texcoords)[4] = (const GLfloat (*)[4])
            span->array->attribs[VARYING_SLOT_TEX0 + unit];
         float4_array targetcoords =
            span->array->attribs[VARYING_SLOT_TEX0 +
               ctx->Texture.Unit[unit].BumpTarget - GL_TEXTURE0];

         const struct gl_sampler_object *samp = _mesa_get_samplerobj(ctx, unit);
         GLfloat *lambda = span->array->lambda[unit];
         float4_array texels = get_texel_array(swrast, unit);
         GLuint i;
         GLfloat rotMatrix00 = ctx->Texture.Unit[unit].RotMatrix[0];
         GLfloat rotMatrix01 = ctx->Texture.Unit[unit].RotMatrix[1];
         GLfloat rotMatrix10 = ctx->Texture.Unit[unit].RotMatrix[2];
         GLfloat rotMatrix11 = ctx->Texture.Unit[unit].RotMatrix[3];

         /* adjust texture lod (lambda) */
         if (span->arrayMask & SPAN_LAMBDA) {
            if (texUnit->LodBias + samp->LodBias != 0.0F) {
               /* apply LOD bias, but don't clamp yet */
               const GLfloat bias = CLAMP(texUnit->LodBias + samp->LodBias,
                                          -ctx->Const.MaxTextureLodBias,
                                          ctx->Const.MaxTextureLodBias);
               GLuint i;
               for (i = 0; i < span->end; i++) {
                  lambda[i] += bias;
               }
            }

            if (samp->MinLod != -1000.0 ||
                samp->MaxLod != 1000.0) {
               /* apply LOD clamping to lambda */
               const GLfloat min = samp->MinLod;
               const GLfloat max = samp->MaxLod;
               GLuint i;
               for (i = 0; i < span->end; i++) {
                  GLfloat l = lambda[i];
                  lambda[i] = CLAMP(l, min, max);
               }
            }
         }

         /* Sample the texture (span->end = number of fragments) */
         swrast->TextureSample[unit]( ctx, samp,
                                      ctx->Texture.Unit[unit]._Current,
                                      span->end, texcoords, lambda, texels );

         /* manipulate the span values of the bump target
            not sure this can work correctly even ignoring
            the problem that channel is unsigned */
         for (i = 0; i < span->end; i++) {
            targetcoords[i][0] += (texels[i][0] * rotMatrix00 + texels[i][1] *
                                  rotMatrix01) / targetcoords[i][3];
            targetcoords[i][1] += (texels[i][0] * rotMatrix10 + texels[i][1] *
                                  rotMatrix11) / targetcoords[i][3];
         }
      }
   }

   /*
    * Must do all texture sampling before combining in order to
    * accomodate GL_ARB_texture_env_crossbar.
    */
   for (unit = 0; unit < ctx->Const.MaxTextureUnits; unit++) {
      const struct gl_texture_unit *texUnit = &ctx->Texture.Unit[unit];
      if (texUnit->_ReallyEnabled &&
          texUnit->_CurrentCombine->ModeRGB != GL_BUMP_ENVMAP_ATI) {
         const GLfloat (*texcoords)[4] = (const GLfloat (*)[4])
            span->array->attribs[VARYING_SLOT_TEX0 + unit];
         const struct gl_texture_object *curObj = texUnit->_Current;
         const struct gl_sampler_object *samp = _mesa_get_samplerobj(ctx, unit);
         GLfloat *lambda = span->array->lambda[unit];
         float4_array texels = get_texel_array(swrast, unit);

         /* adjust texture lod (lambda) */
         if (span->arrayMask & SPAN_LAMBDA) {
            if (texUnit->LodBias + samp->LodBias != 0.0F) {
               /* apply LOD bias, but don't clamp yet */
               const GLfloat bias = CLAMP(texUnit->LodBias + samp->LodBias,
                                          -ctx->Const.MaxTextureLodBias,
                                          ctx->Const.MaxTextureLodBias);
               GLuint i;
               for (i = 0; i < span->end; i++) {
                  lambda[i] += bias;
               }
            }

            if (samp->MinLod != -1000.0 ||
                samp->MaxLod != 1000.0) {
               /* apply LOD clamping to lambda */
               const GLfloat min = samp->MinLod;
               const GLfloat max = samp->MaxLod;
               GLuint i;
               for (i = 0; i < span->end; i++) {
                  GLfloat l = lambda[i];
                  lambda[i] = CLAMP(l, min, max);
               }
            }
         }
         else if (samp->MaxAnisotropy > 1.0 &&
                  samp->MinFilter == GL_LINEAR_MIPMAP_LINEAR) {
            /* sample_lambda_2d_aniso is beeing used as texture_sample_func,
             * it requires the current SWspan *span as an additional parameter.
             * In order to keep the same function signature, the unused lambda
             * parameter will be modified to actually contain the SWspan pointer.
             * This is a Hack. To make it right, the texture_sample_func
             * signature and all implementing functions need to be modified.
             */
            /* "hide" SWspan struct; cast to (GLfloat *) to suppress warning */
            lambda = (GLfloat *)span;
         }

         /* Sample the texture (span->end = number of fragments) */
         swrast->TextureSample[unit]( ctx, samp,
                                      ctx->Texture.Unit[unit]._Current,
                                      span->end, texcoords, lambda, texels );

         /* GL_EXT_texture_swizzle */
         if (curObj->_Swizzle != SWIZZLE_NOOP) {
            swizzle_texels(curObj->_Swizzle, span->end, texels);
         }
      }
   }

   /*
    * OK, now apply the texture (aka texture combine/blend).
    * We modify the span->color.rgba values.
    */
   for (unit = 0; unit < ctx->Const.MaxTextureUnits; unit++) {
      if (ctx->Texture.Unit[unit]._ReallyEnabled)
         texture_combine(ctx, unit, primary_rgba, swrast->TexelBuffer, span);
   }

   free(primary_rgba);
}