summaryrefslogtreecommitdiff
path: root/src/mesa/tnl/t_imm_fixup.c
blob: cf3aa2ed9636ba22dbb7b9b4fd3652ec6b170cbd (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
785
786
/*
 * Mesa 3-D graphics library
 * Version:  5.1
 *
 * Copyright (C) 1999-2003  Brian Paul   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
 * BRIAN PAUL 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.
 */

/*
 * Authors:
 *    Keith Whitwell <keith@tungstengraphics.com>
 */


#include "glheader.h"
#include "context.h"
#include "enums.h"
#include "dlist.h"
#include "colormac.h"
#include "light.h"
#include "macros.h"
#include "imports.h"
#include "state.h"
#include "mtypes.h"

#include "math/m_matrix.h"
#include "math/m_xform.h"

#include "t_context.h"
#include "t_imm_alloc.h"
#include "t_imm_debug.h"
#include "t_imm_elt.h"
#include "t_imm_fixup.h"
#include "t_imm_exec.h"
#include "t_pipeline.h"


/*
 * Indexed by primitive type (GL_POINTS=0, GL_LINES=1, GL_LINE_LOOP=2, etc)
 */
static const GLuint increment[GL_POLYGON+2] = { 1,2,1,1,3,1,1,4,2,1,1 };
static const GLuint intro[GL_POLYGON+2]     = { 0,0,2,2,0,2,2,0,2,2,0 };


void
_tnl_fixup_4f( GLfloat data[][4], const GLuint flag[],
               GLuint start, GLuint match )
{
   GLuint i = start;

   for (;;) {
      if ((flag[++i] & match) == 0) {
	 COPY_4FV(data[i], data[i-1]);
	 if (flag[i] & VERT_BIT_END_VB) break;
      }
   }
}


void
_tnl_fixup_1ui( GLuint *data, GLuint flag[], GLuint start, GLuint match )
{
   GLuint i = start;

   for (;;) {
      if ((flag[++i] & match) == 0) {
	 data[i] = data[i-1];
	 if (flag[i] & VERT_BIT_END_VB) break;
      }
   }
   flag[i] |= match;
}


void
_tnl_fixup_1ub( GLubyte *data, GLuint flag[], GLuint start, GLuint match)
{
   GLuint i = start;

   for (;;) {
      if ((flag[++i] & match) == 0) {
	 data[i] = data[i-1];
	 if (flag[i] & VERT_BIT_END_VB) break;
      }
   }
   flag[i] |= match;
}


static void
fixup_first_4f( GLfloat data[][4], const GLuint flag[], GLuint match,
		GLuint start, const GLfloat *dflt )
{
   GLuint i = start-1;
   match |= VERT_BIT_END_VB;

   while ((flag[++i]&match) == 0)
      COPY_4FV(data[i], dflt);
}


static void
fixup_first_1ui( GLuint data[], const GLuint flag[], GLuint match,
		 GLuint start, const GLuint dflt )
{
   GLuint i = start-1;
   match |= VERT_BIT_END_VB;

   while ((flag[++i]&match) == 0)
      data[i] = dflt;
}


static void
fixup_first_1ub( GLubyte data[], const GLuint flag[], GLuint match,
		 GLuint start, GLubyte dflt )
{
   GLuint i = start-1;
   match |= VERT_BIT_END_VB;

   while ((flag[++i]&match) == 0)
      data[i] = dflt;
}


/**
 * Copy vertex attributes from the ctx->Current group into the immediate
 * struct at the given position according to copyMask.
 */
static void
copy_from_current( GLcontext *ctx, struct immediate *IM, 
                   GLuint pos, GLuint copyMask )
{
   GLuint attrib, attribBit;

   if (MESA_VERBOSE&VERBOSE_IMMEDIATE)
      _tnl_print_vert_flags("copy from current", copyMask); 

   for (attrib = 0, attribBit = 1; attrib < 16; attrib++, attribBit <<= 1) {
      if (copyMask & attribBit) {
         if (!IM->Attrib[attrib]) {
            IM->Attrib[attrib] = _mesa_malloc(IMM_SIZE * 4 * sizeof(GLfloat));
            if (!IM->Attrib[attrib]) {
               _mesa_error(ctx, GL_OUT_OF_MEMORY, "vertex processing3");
               return;
            }
         }
         COPY_4FV( IM->Attrib[attrib][pos], ctx->Current.Attrib[attrib]);
      }
   }

   if (copyMask & VERT_BIT_INDEX)
      IM->Index[pos] = ctx->Current.Index;

   if (copyMask & VERT_BIT_EDGEFLAG)
      IM->EdgeFlag[pos] = ctx->Current.EdgeFlag;
}


/**
 * Fill in missing vertex attributes in the incoming immediate structure.
 * For example, suppose the following calls are made:
 * glBegin()
 * glColor(c1)
 * glVertex(v1)
 * glVertex(v2)
 * glEnd()
 * The v2 vertex should get color c1 since glColor wasn't called for v2.
 * This function will make c2 be c1.  Same story for all vertex attribs.
 */
void
_tnl_fixup_input( GLcontext *ctx, struct immediate *IM )
{
   TNLcontext *tnl = TNL_CONTEXT(ctx);
   GLuint start = IM->CopyStart;
   GLuint andflag = IM->CopyAndFlag;
   GLuint orflag = IM->CopyOrFlag | IM->Evaluated;
   GLuint fixup;

   IM->CopyTexSize = IM->TexSize;

/*     _mesa_debug(ctx, "Fixup input, Start: %u Count: %u LastData: %u\n", */
/*  	   IM->Start, IM->Count, IM->LastData); */
/*     _tnl_print_vert_flags("Orflag", orflag); */
/*     _tnl_print_vert_flags("Andflag", andflag); */


   fixup = ~andflag & VERT_BITS_FIXUP;

   if (!ctx->CompileFlag)
      fixup &= tnl->pipeline.inputs;

   if (!ctx->ExecuteFlag)
      fixup &= orflag;

   if ((orflag & (VERT_BIT_POS|VERT_BITS_EVAL_ANY)) == 0)
      fixup = 0;

   if (fixup) {
      const GLuint copy = fixup & ~IM->Flag[start];
      GLuint attr;

      /* Equivalent to a lazy copy-from-current when setting up the
       * immediate.
       */
      if (ctx->ExecuteFlag && copy) 
	 copy_from_current( ctx, IM, start, copy );

      if (MESA_VERBOSE&VERBOSE_IMMEDIATE)
  	 _tnl_print_vert_flags("fixup", fixup); 

      for (attr = 1; attr < VERT_ATTRIB_MAX; attr++) { /* skip 0 (POS) */
         const GLuint attrBit = 1 << attr;
         if (fixup & attrBit) {
            if (!IM->Attrib[attr]) {
               IM->Attrib[attr] = _mesa_malloc(IMM_SIZE * 4 * sizeof(GLfloat));
               if (!IM->Attrib[attr]) {
                  _mesa_error(ctx, GL_OUT_OF_MEMORY, "vertex processing");
                  return;
               }
            }

            if (attr == VERT_BIT_COLOR0) {
               /* special case, darn - try to remove someday */
               if (orflag & VERT_BIT_COLOR0) {
                  _tnl_fixup_4f( IM->Attrib[VERT_ATTRIB_COLOR0], IM->Flag,
                                 start, VERT_BIT_COLOR0 );
               }
               /* No need for else case as the drivers understand stride
                * zero here.  (TODO - propogate this)
                */
            }
            else {
               /* general case */
               if (orflag & attrBit)
                  _tnl_fixup_4f( IM->Attrib[attr], IM->Flag, start, attrBit );
               else
                  fixup_first_4f( IM->Attrib[attr], IM->Flag, VERT_BIT_END_VB,
                                  start, IM->Attrib[attr][start] );
            }
         }
      }

      if (fixup & VERT_BIT_EDGEFLAG) {
	 if (orflag & VERT_BIT_EDGEFLAG)
	    _tnl_fixup_1ub( IM->EdgeFlag, IM->Flag, start, VERT_BIT_EDGEFLAG );
	 else
	    fixup_first_1ub( IM->EdgeFlag, IM->Flag, VERT_BIT_END_VB, start,
			     IM->EdgeFlag[start] );
      }

      if (fixup & VERT_BIT_INDEX) {
	 if (orflag & VERT_BIT_INDEX)
	    _tnl_fixup_1ui( IM->Index, IM->Flag, start, VERT_BIT_INDEX );
	 else
	    fixup_first_1ui( IM->Index, IM->Flag, VERT_BIT_END_VB, start, 
			     IM->Index[start] );
      }

   }
      
   /* Prune possible half-filled slot.
    */
   IM->Flag[IM->LastData+1] &= ~VERT_BIT_END_VB;
   IM->Flag[IM->Count] |= VERT_BIT_END_VB;


   /* Materials:
    */
   if (IM->MaterialOrMask & ~IM->MaterialAndMask) {
      GLuint vulnerable = IM->MaterialOrMask;
      GLuint i = IM->Start;

      do {
	 while (!(IM->Flag[i] & VERT_BIT_MATERIAL))
	    i++;

	 vulnerable &= ~IM->MaterialMask[i];
	 _mesa_copy_material_pairs( IM->Material[i],
				    ctx->Light.Material,
				    vulnerable );


	++i;
      } while (vulnerable);
   }
}


static void
copy_material( struct immediate *next,
               struct immediate *prev,
               GLuint dst, GLuint src )
{
/*     _mesa_debug(NULL, "%s\n", __FUNCTION__); */

   if (next->Material == 0) {
      next->Material = (struct gl_material (*)[2])
         MALLOC( sizeof(struct gl_material) * IMM_SIZE * 2 );
      next->MaterialMask = (GLuint *) MALLOC( sizeof(GLuint) * IMM_SIZE );
   }

   next->MaterialMask[dst] = prev->MaterialOrMask;
   MEMCPY(next->Material[dst], prev->Material[src],
          2 * sizeof(struct gl_material));
}



static GLboolean is_fan_like[GL_POLYGON+1] = {
   GL_FALSE,
   GL_FALSE,
   GL_TRUE,			/* line loop */
   GL_FALSE,
   GL_FALSE,
   GL_FALSE,
   GL_TRUE,			/* tri fan */
   GL_FALSE,
   GL_FALSE,
   GL_TRUE			/* polygon */
};


/**
 * Copy the untransformed data from the shared vertices of a primitive
 * that wraps over two immediate structs.  This is done prior to
 * set_immediate so that prev and next may point to the same
 * structure.  In general it's difficult to avoid this copy on long
 * primitives.
 *
 * Have to be careful with the transitions between display list
 * replay, compile and normal execute modes.
 */
void
_tnl_copy_immediate_vertices( GLcontext *ctx, struct immediate *next )
{
   TNLcontext *tnl = TNL_CONTEXT(ctx);
   struct immediate *prev = tnl->ExecCopySource;
   struct vertex_arrays *inputs = &tnl->imm_inputs;
   GLuint count = tnl->ExecCopyCount;
   GLuint *elts = tnl->ExecCopyElts;
   GLuint offset = IMM_MAX_COPIED_VERTS - count;
   GLuint i;

   if (!prev) {
      ASSERT(tnl->ExecCopyCount == 0);
      return;
   }

   next->CopyStart = next->Start - count;

   if ((prev->CopyOrFlag & VERT_BITS_DATA) == VERT_BIT_ELT &&
       ctx->Array.LockCount &&
       (ctx->Array.Vertex.Enabled ||
        (ctx->VertexProgram.Enabled && ctx->Array.VertexAttrib[0].Enabled)))
   {
      /* Copy Elt values only
       */
      for (i = 0 ; i < count ; i++)
      {
	 GLuint src = elts[i+offset];
	 GLuint dst = next->CopyStart+i;
	 next->Elt[dst] = prev->Elt[src];
	 next->Flag[dst] = VERT_BIT_ELT;
	 elts[i+offset] = dst;
      }
/*        _mesa_debug(ctx, "ADDING VERT_BIT_ELT!\n"); */
      next->CopyOrFlag |= VERT_BIT_ELT;
      next->CopyAndFlag &= VERT_BIT_ELT;
   }
   else {
      GLuint copy = tnl->pipeline.inputs & (prev->CopyOrFlag|prev->Evaluated);
      GLuint flag, attr;

      if (is_fan_like[ctx->Driver.CurrentExecPrimitive]) {
	 flag = ((prev->CopyOrFlag|prev->Evaluated) & VERT_BITS_FIXUP);
	 next->CopyOrFlag |= flag;
      } 
      else {
	 /* Don't let an early 'glColor', etc. poison the elt path.
	  */
	 flag = ((prev->OrFlag|prev->Evaluated) & VERT_BITS_FIXUP);
      }

      next->TexSize |= tnl->ExecCopyTexSize;
      next->CopyAndFlag &= flag;
	 

/*        _tnl_print_vert_flags("copy vertex components", copy); */
/*        _tnl_print_vert_flags("prev copyorflag", prev->CopyOrFlag); */
/*        _tnl_print_vert_flags("flag", flag); */

      /* Allocate attribute arrays in the destination immediate struct */
      for (attr = 0; attr < VERT_ATTRIB_MAX; attr++) {
         if ((copy & (1 << attr)) && !next->Attrib[attr]) {
            next->Attrib[attr] = _mesa_malloc(IMM_SIZE * 4 * sizeof(GLfloat));
            if (!next->Attrib[attr]) {
               _mesa_error(ctx, GL_OUT_OF_MEMORY, "vertex processing");
               return;
            }
         }
      }

      /* Copy whole vertices
       */
      for (i = 0 ; i < count ; i++)
      {
	 GLuint src = elts[i+offset];
	 GLuint isrc = src - prev->CopyStart;
	 GLuint dst = next->CopyStart+i;

	 /* Values subject to eval must be copied out of the 'inputs'
	  * struct.  (Copied rows should not be evaluated twice).
	  *
	  * Note these pointers are null when inactive.
          *
          * XXX try to use a loop over vertex attribs here.
	  */
	 COPY_4FV( next->Attrib[VERT_ATTRIB_POS][dst],
                   inputs->Obj.data[isrc] );

	 if (copy & VERT_BIT_NORMAL) {
	    COPY_3FV( next->Attrib[VERT_ATTRIB_NORMAL][dst],
                      inputs->Normal.data[isrc] );
	 }

	 if (copy & VERT_BIT_COLOR0)
	    COPY_4FV( next->Attrib[VERT_ATTRIB_COLOR0][dst], 
		      ((GLfloat (*)[4])inputs->Color.Ptr)[isrc] );

	 if (copy & VERT_BITS_TEX_ANY) {
	    GLuint i;
	    for (i = 0 ; i < prev->MaxTextureUnits ; i++) {
	       if (copy & VERT_BIT_TEX(i))
		  COPY_4FV( next->Attrib[VERT_ATTRIB_TEX0 + i][dst], 
			    inputs->TexCoord[i].data[isrc] );
	    }
	 }

         /* the rest aren't used for evaluators */
         if (copy & VERT_BIT_COLOR1)
	    COPY_4FV( next->Attrib[VERT_ATTRIB_COLOR1][dst],
                      prev->Attrib[VERT_ATTRIB_COLOR1][src] );

         if (copy & VERT_BIT_FOG)
	    COPY_4FV( next->Attrib[VERT_ATTRIB_FOG][dst],
                      prev->Attrib[VERT_ATTRIB_FOG][src] );


	 if (copy & VERT_BIT_INDEX)
	    next->Index[dst] = inputs->Index.data[isrc];

         if (copy & VERT_BIT_EDGEFLAG)
	    next->EdgeFlag[dst] = prev->EdgeFlag[src];

	 if (copy & VERT_BIT_ELT)
	    next->Elt[dst] = prev->Elt[src];

         if (copy & VERT_BIT_MATERIAL)
	    if (prev->Flag[src] & VERT_BIT_MATERIAL)
	       copy_material(next, prev, dst, src);

	 next->Flag[dst] = flag;
	 next->CopyOrFlag |= prev->Flag[src] & (VERT_BITS_FIXUP|
						VERT_BIT_MATERIAL|
						VERT_BIT_POS);
 	 elts[i+offset] = dst;
      }
   }

   if (--tnl->ExecCopySource->ref_count == 0) 
      _tnl_free_immediate( ctx, tnl->ExecCopySource );
  
   tnl->ExecCopySource = next; next->ref_count++;
}



/**
 * Revive a compiled immediate struct - propogate new 'Current'
 * values.  Often this is redundant because the current values were
 * known and fixed up at compile time (or in the first execution of
 * the cassette).
 */
void
_tnl_fixup_compiled_cassette( GLcontext *ctx, struct immediate *IM )
{
   TNLcontext *tnl = TNL_CONTEXT(ctx);
   GLuint fixup;
   GLuint start = IM->Start;

/*     _mesa_debug(ctx, "%s\n", __FUNCTION__); */

   IM->Evaluated = 0;
   IM->CopyOrFlag = IM->OrFlag;	  
   IM->CopyAndFlag = IM->AndFlag; 
   IM->CopyTexSize = IM->TexSize | tnl->ExecCopyTexSize;

   _tnl_copy_immediate_vertices( ctx, IM );

   if (ctx->Driver.CurrentExecPrimitive == GL_POLYGON+1) {
      ASSERT(IM->CopyStart == IM->Start);
   }

   /* Naked array elements can be copied into the first cassette in a
    * display list.  Need to translate them away:
    */
   if (IM->CopyOrFlag & VERT_BIT_ELT) {
      GLuint copy = tnl->pipeline.inputs & ~ctx->Array._Enabled;
      GLuint i;

      ASSERT(IM->CopyStart < IM->Start);

      _tnl_translate_array_elts( ctx, IM, IM->CopyStart, IM->Start );

      for (i = IM->CopyStart ; i < IM->Start ; i++)
	 copy_from_current( ctx, IM, i, copy ); 

      _tnl_copy_to_current( ctx, IM, ctx->Array._Enabled, IM->Start );
   }

   fixup = tnl->pipeline.inputs & ~IM->Flag[start] & VERT_BITS_FIXUP;

/*     _tnl_print_vert_flags("fixup compiled", fixup); */

   if (fixup) {
      GLuint attr;

      for (attr = 1; attr < VERT_ATTRIB_MAX; attr++) { /* skip 0 (POS) */
         const GLuint attrBit = 1 << attr;
         if (fixup & attrBit) {
            if (!IM->Attrib[attr]) {
               IM->Attrib[attr] = _mesa_malloc(IMM_SIZE * 4 * sizeof(GLfloat));
               if (!IM->Attrib[attr]) {
                  _mesa_error(ctx, GL_OUT_OF_MEMORY, "vertex processing");
               }
            }
            if (attr == VERT_ATTRIB_COLOR0) {
               /* special case, darn */
               if (IM->CopyOrFlag & VERT_BIT_COLOR0)
                  fixup_first_4f(IM->Attrib[VERT_ATTRIB_COLOR0], IM->Flag,
                                 VERT_BIT_COLOR0, start,
                                 ctx->Current.Attrib[VERT_ATTRIB_COLOR0] );
               else
                  fixup &= ~VERT_BIT_COLOR0;
            }
            else {
               fixup_first_4f(IM->Attrib[attr], IM->Flag,
                              attrBit, start, ctx->Current.Attrib[attr] );
            }
         }
      }

      if (fixup & VERT_BIT_EDGEFLAG)
	 fixup_first_1ub(IM->EdgeFlag, IM->Flag, VERT_BIT_EDGEFLAG, start,
			 ctx->Current.EdgeFlag );

      if (fixup & VERT_BIT_INDEX)
	 fixup_first_1ui(IM->Index, IM->Flag, VERT_BIT_INDEX, start,
			 ctx->Current.Index );

      IM->CopyOrFlag |= fixup;
   }
   

   /* Materials:
    */
   if (IM->MaterialOrMask & ~IM->MaterialAndMask) {
      GLuint vulnerable = IM->MaterialOrMask;
      GLuint i = IM->Start;

      do {
	 while (!(IM->Flag[i] & VERT_BIT_MATERIAL))
	    i++;

	 vulnerable &= ~IM->MaterialMask[i];
	 _mesa_copy_material_pairs( IM->Material[i],
				    ctx->Light.Material,
				    vulnerable );


	 ++i;
      } while (vulnerable);
   }
}



static void
copy_none( TNLcontext *tnl, GLuint start, GLuint count, GLuint ovf)
{
   (void) (start && ovf && tnl && count);
}

static void
copy_last( TNLcontext *tnl, GLuint start, GLuint count, GLuint ovf)
{
   (void) start; (void) ovf;
   tnl->ExecCopyCount = 1;
   tnl->ExecCopyElts[2] = count-1;
}

static void
copy_first_and_last( TNLcontext *tnl, GLuint start, GLuint count, GLuint ovf)
{
   (void) ovf;
   tnl->ExecCopyCount = 2;
   tnl->ExecCopyElts[1] = start;
   tnl->ExecCopyElts[2] = count-1;
}

static void
copy_last_three( TNLcontext *tnl, GLuint start, GLuint count, GLuint ovf )
{
   (void) start;
   tnl->ExecCopyCount = 2+ovf;
   tnl->ExecCopyElts[0] = count-3;
   tnl->ExecCopyElts[1] = count-2;
   tnl->ExecCopyElts[2] = count-1;
}

static void
copy_overflow( TNLcontext *tnl, GLuint start, GLuint count, GLuint ovf )
{
   (void) start;
   tnl->ExecCopyCount = ovf;
   tnl->ExecCopyElts[0] = count-3;
   tnl->ExecCopyElts[1] = count-2;
   tnl->ExecCopyElts[2] = count-1;
}


typedef void (*copy_func)( TNLcontext *tnl, GLuint start, GLuint count,
			   GLuint ovf );

static copy_func copy_tab[GL_POLYGON+2] =
{
   copy_none,            /* GL_POINTS */
   copy_overflow,        /* GL_LINES */
   copy_first_and_last,  /* GL_LINE_LOOP */
   copy_last,            /* GL_LINE_STRIP */
   copy_overflow,        /* GL_TRIANGLES */
   copy_last_three,      /* GL_TRIANGLE_STRIP */
   copy_first_and_last,  /* GL_TRIANGLE_FAN */
   copy_overflow,        /* GL_QUADS */
   copy_last_three,      /* GL_QUAD_STRIP */
   copy_first_and_last,  /* GL_POLYGON */
   copy_none
};



/**
 * Figure out what vertices need to be copied next time.
 */
void
_tnl_get_exec_copy_verts( GLcontext *ctx, struct immediate *IM )
{

   TNLcontext *tnl = TNL_CONTEXT(ctx);
   GLuint last = IM->LastPrimitive;
   GLuint prim = ctx->Driver.CurrentExecPrimitive;
   GLuint pincr = increment[prim];
   GLuint pintro = intro[prim];
   GLuint ovf = 0;

/*     _mesa_debug(ctx, "_tnl_get_exec_copy_verts %s\n",  */
/*  	   _mesa_lookup_enum_by_nr(prim)); */

   if (tnl->ExecCopySource)
      if (--tnl->ExecCopySource->ref_count == 0) 
	 _tnl_free_immediate( ctx, tnl->ExecCopySource );

   if (prim == GL_POLYGON+1) {
      tnl->ExecCopySource = 0;
      tnl->ExecCopyCount = 0;
      tnl->ExecCopyTexSize = 0;
      tnl->ExecParity = 0;
   }
   else {
      /* Remember this immediate as the one to copy from.
       */
      tnl->ExecCopySource = IM; IM->ref_count++;
      tnl->ExecCopyCount = 0;
      tnl->ExecCopyTexSize = IM->CopyTexSize;

      if (IM->LastPrimitive != IM->CopyStart)
	 tnl->ExecParity = 0;
	 
      tnl->ExecParity ^= IM->PrimitiveLength[IM->LastPrimitive] & 1;


      if (pincr != 1 && (IM->Count - last - pintro))
	 ovf = (IM->Count - last - pintro) % pincr;

      if (last < IM->Count)
	 copy_tab[prim]( tnl, last, IM->Count, ovf );
   }
}


/**
 * Recalculate ExecCopyElts, ExecParity, etc.  
 */
void 
_tnl_get_purged_copy_verts( GLcontext *ctx, struct immediate *IM ) 
{
   TNLcontext *tnl = TNL_CONTEXT(ctx);

   if (ctx->Driver.CurrentExecPrimitive != GL_POLYGON+1) {
      GLuint last = IM->LastPrimitive;
      GLenum prim = IM->Primitive[last];
      GLuint pincr = increment[prim];
      GLuint pintro = intro[prim];
      GLuint ovf = 0, i;

      tnl->ExecCopyCount = 0;
      if (IM->LastPrimitive != IM->CopyStart)
	 tnl->ExecParity = 0;
	 
      tnl->ExecParity ^= IM->PrimitiveLength[IM->LastPrimitive] & 1;

      if (pincr != 1 && (IM->Count - last - pintro))
	 ovf = (IM->Count - last - pintro) % pincr;

      if (last < IM->Count)
	 copy_tab[prim]( tnl, last, IM->Count, ovf );

      for (i = 0 ; i < tnl->ExecCopyCount ; i++)
	 tnl->ExecCopyElts[i] = IM->Elt[tnl->ExecCopyElts[i]];
   }
}


/**
 * Called via the VB->import_data function pointer.
 */
void
_tnl_upgrade_current_data( GLcontext *ctx, GLuint required, GLuint flags )
{
   TNLcontext *tnl = TNL_CONTEXT(ctx);
   struct vertex_buffer *VB = &tnl->vb;
   struct immediate *IM = (struct immediate *)VB->import_source;

   ASSERT(IM);

/*     _tnl_print_vert_flags("_tnl_upgrade_client_data", required); */

   if ((required & VERT_BIT_COLOR0) && (VB->ColorPtr[0]->Flags & CA_CLIENT_DATA)) {
      struct gl_client_array *tmp = &tnl->imm_inputs.Color;
      GLuint start = IM->CopyStart;

      tmp->Ptr = IM->Attrib[VERT_ATTRIB_COLOR0] + start;
      tmp->StrideB = 4 * sizeof(GLfloat);
      tmp->Flags = 0;

      COPY_4FV( IM->Attrib[VERT_ATTRIB_COLOR0][start],
                ctx->Current.Attrib[VERT_ATTRIB_COLOR0]);   

      /*
      ASSERT(IM->Flag[IM->LastData+1] & VERT_BIT_END_VB);
      */

      fixup_first_4f( IM->Attrib[VERT_ATTRIB_COLOR0], IM->Flag,
                      VERT_BIT_END_VB,
                      start, IM->Attrib[VERT_ATTRIB_COLOR0][start] );

      VB->importable_data &= ~VERT_BIT_COLOR0;
   }
}