summaryrefslogtreecommitdiffstats
path: root/src/mesa/state_tracker/st_draw.c
blob: ff3008a5fc0d01577bc9ca3f599d916e2ff7c059 (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
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
/**************************************************************************
 * 
 * Copyright 2007 Tungsten Graphics, Inc., Cedar Park, Texas.
 * 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, sub license, 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 (including the
 * next paragraph) 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 NON-INFRINGEMENT.
 * IN NO EVENT SHALL TUNGSTEN GRAPHICS AND/OR ITS SUPPLIERS 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.
 * 
 **************************************************************************/

/*
 * This file implements the st_draw_vbo() function which is called from
 * Mesa's VBO module.  All point/line/triangle rendering is done through
 * this function whether the user called glBegin/End, glDrawArrays,
 * glDrawElements, glEvalMesh, or glCalList, etc.
 *
 * We basically convert the VBO's vertex attribute/array information into
 * Gallium vertex state, bind the vertex buffer objects and call
 * pipe->draw_vbo().
 *
 * Authors:
 *   Keith Whitwell <keith@tungstengraphics.com>
 */


#include "main/imports.h"
#include "main/image.h"
#include "main/bufferobj.h"
#include "main/macros.h"
#include "main/mfeatures.h"
#include "program/prog_uniform.h"

#include "vbo/vbo.h"

#include "st_context.h"
#include "st_atom.h"
#include "st_cb_bufferobjects.h"
#include "st_draw.h"
#include "st_program.h"

#include "pipe/p_context.h"
#include "pipe/p_defines.h"
#include "util/u_inlines.h"
#include "util/u_format.h"
#include "util/u_prim.h"
#include "util/u_draw_quad.h"
#include "draw/draw_context.h"
#include "cso_cache/cso_context.h"


static GLuint double_types[4] = {
   PIPE_FORMAT_R64_FLOAT,
   PIPE_FORMAT_R64G64_FLOAT,
   PIPE_FORMAT_R64G64B64_FLOAT,
   PIPE_FORMAT_R64G64B64A64_FLOAT
};

static GLuint float_types[4] = {
   PIPE_FORMAT_R32_FLOAT,
   PIPE_FORMAT_R32G32_FLOAT,
   PIPE_FORMAT_R32G32B32_FLOAT,
   PIPE_FORMAT_R32G32B32A32_FLOAT
};

static GLuint half_float_types[4] = {
   PIPE_FORMAT_R16_FLOAT,
   PIPE_FORMAT_R16G16_FLOAT,
   PIPE_FORMAT_R16G16B16_FLOAT,
   PIPE_FORMAT_R16G16B16A16_FLOAT
};

static GLuint uint_types_norm[4] = {
   PIPE_FORMAT_R32_UNORM,
   PIPE_FORMAT_R32G32_UNORM,
   PIPE_FORMAT_R32G32B32_UNORM,
   PIPE_FORMAT_R32G32B32A32_UNORM
};

static GLuint uint_types_scale[4] = {
   PIPE_FORMAT_R32_USCALED,
   PIPE_FORMAT_R32G32_USCALED,
   PIPE_FORMAT_R32G32B32_USCALED,
   PIPE_FORMAT_R32G32B32A32_USCALED
};

static GLuint int_types_norm[4] = {
   PIPE_FORMAT_R32_SNORM,
   PIPE_FORMAT_R32G32_SNORM,
   PIPE_FORMAT_R32G32B32_SNORM,
   PIPE_FORMAT_R32G32B32A32_SNORM
};

static GLuint int_types_scale[4] = {
   PIPE_FORMAT_R32_SSCALED,
   PIPE_FORMAT_R32G32_SSCALED,
   PIPE_FORMAT_R32G32B32_SSCALED,
   PIPE_FORMAT_R32G32B32A32_SSCALED
};

static GLuint ushort_types_norm[4] = {
   PIPE_FORMAT_R16_UNORM,
   PIPE_FORMAT_R16G16_UNORM,
   PIPE_FORMAT_R16G16B16_UNORM,
   PIPE_FORMAT_R16G16B16A16_UNORM
};

static GLuint ushort_types_scale[4] = {
   PIPE_FORMAT_R16_USCALED,
   PIPE_FORMAT_R16G16_USCALED,
   PIPE_FORMAT_R16G16B16_USCALED,
   PIPE_FORMAT_R16G16B16A16_USCALED
};

static GLuint short_types_norm[4] = {
   PIPE_FORMAT_R16_SNORM,
   PIPE_FORMAT_R16G16_SNORM,
   PIPE_FORMAT_R16G16B16_SNORM,
   PIPE_FORMAT_R16G16B16A16_SNORM
};

static GLuint short_types_scale[4] = {
   PIPE_FORMAT_R16_SSCALED,
   PIPE_FORMAT_R16G16_SSCALED,
   PIPE_FORMAT_R16G16B16_SSCALED,
   PIPE_FORMAT_R16G16B16A16_SSCALED
};

static GLuint ubyte_types_norm[4] = {
   PIPE_FORMAT_R8_UNORM,
   PIPE_FORMAT_R8G8_UNORM,
   PIPE_FORMAT_R8G8B8_UNORM,
   PIPE_FORMAT_R8G8B8A8_UNORM
};

static GLuint ubyte_types_scale[4] = {
   PIPE_FORMAT_R8_USCALED,
   PIPE_FORMAT_R8G8_USCALED,
   PIPE_FORMAT_R8G8B8_USCALED,
   PIPE_FORMAT_R8G8B8A8_USCALED
};

static GLuint byte_types_norm[4] = {
   PIPE_FORMAT_R8_SNORM,
   PIPE_FORMAT_R8G8_SNORM,
   PIPE_FORMAT_R8G8B8_SNORM,
   PIPE_FORMAT_R8G8B8A8_SNORM
};

static GLuint byte_types_scale[4] = {
   PIPE_FORMAT_R8_SSCALED,
   PIPE_FORMAT_R8G8_SSCALED,
   PIPE_FORMAT_R8G8B8_SSCALED,
   PIPE_FORMAT_R8G8B8A8_SSCALED
};

static GLuint fixed_types[4] = {
   PIPE_FORMAT_R32_FIXED,
   PIPE_FORMAT_R32G32_FIXED,
   PIPE_FORMAT_R32G32B32_FIXED,
   PIPE_FORMAT_R32G32B32A32_FIXED
};



/**
 * Return a PIPE_FORMAT_x for the given GL datatype and size.
 */
enum pipe_format
st_pipe_vertex_format(GLenum type, GLuint size, GLenum format,
                      GLboolean normalized)
{
   assert((type >= GL_BYTE && type <= GL_DOUBLE) ||
          type == GL_FIXED || type == GL_HALF_FLOAT ||
          type == GL_INT_2_10_10_10_REV ||
          type == GL_UNSIGNED_INT_2_10_10_10_REV);
   assert(size >= 1);
   assert(size <= 4);
   assert(format == GL_RGBA || format == GL_BGRA);

   if (type == GL_INT_2_10_10_10_REV ||
       type == GL_UNSIGNED_INT_2_10_10_10_REV) {
      assert(size == 4);

      if (format == GL_BGRA) {
         if (type == GL_INT_2_10_10_10_REV) {
            if (normalized)
               return PIPE_FORMAT_B10G10R10A2_SNORM;
            else
               return PIPE_FORMAT_B10G10R10A2_SSCALED;
         } else {
            if (normalized)
               return PIPE_FORMAT_B10G10R10A2_UNORM;
            else
               return PIPE_FORMAT_B10G10R10A2_USCALED;
         }
      } else {
         if (type == GL_INT_2_10_10_10_REV) {
            if (normalized)
               return PIPE_FORMAT_R10G10B10A2_SNORM;
            else
               return PIPE_FORMAT_R10G10B10A2_SSCALED;
         } else {
            if (normalized)
               return PIPE_FORMAT_R10G10B10A2_UNORM;
            else
               return PIPE_FORMAT_R10G10B10A2_USCALED;
         }
      }
   }

   if (format == GL_BGRA) {
      /* this is an odd-ball case */
      assert(type == GL_UNSIGNED_BYTE);
      assert(normalized);
      return PIPE_FORMAT_B8G8R8A8_UNORM;
   }

   if (normalized) {
      switch (type) {
      case GL_DOUBLE: return double_types[size-1];
      case GL_FLOAT: return float_types[size-1];
      case GL_HALF_FLOAT: return half_float_types[size-1];
      case GL_INT: return int_types_norm[size-1];
      case GL_SHORT: return short_types_norm[size-1];
      case GL_BYTE: return byte_types_norm[size-1];
      case GL_UNSIGNED_INT: return uint_types_norm[size-1];
      case GL_UNSIGNED_SHORT: return ushort_types_norm[size-1];
      case GL_UNSIGNED_BYTE: return ubyte_types_norm[size-1];
      case GL_FIXED: return fixed_types[size-1];
      default: assert(0); return 0;
      }
   }
   else {
      switch (type) {
      case GL_DOUBLE: return double_types[size-1];
      case GL_FLOAT: return float_types[size-1];
      case GL_HALF_FLOAT: return half_float_types[size-1];
      case GL_INT: return int_types_scale[size-1];
      case GL_SHORT: return short_types_scale[size-1];
      case GL_BYTE: return byte_types_scale[size-1];
      case GL_UNSIGNED_INT: return uint_types_scale[size-1];
      case GL_UNSIGNED_SHORT: return ushort_types_scale[size-1];
      case GL_UNSIGNED_BYTE: return ubyte_types_scale[size-1];
      case GL_FIXED: return fixed_types[size-1];
      default: assert(0); return 0;
      }
   }
   return PIPE_FORMAT_NONE; /* silence compiler warning */
}


/**
 * This is very similar to vbo_all_varyings_in_vbos() but we are
 * only interested in per-vertex data.  See bug 38626.
 */
static GLboolean
all_varyings_in_vbos(const struct gl_client_array *arrays[])
{
   GLuint i;
   
   for (i = 0; i < VERT_ATTRIB_MAX; i++)
      if (arrays[i]->StrideB &&
          !arrays[i]->InstanceDivisor &&
          !_mesa_is_bufferobj(arrays[i]->BufferObj))
	 return GL_FALSE;

   return GL_TRUE;
}


/**
 * Examine the active arrays to determine if we have interleaved
 * vertex arrays all living in one VBO, or all living in user space.
 */
static GLboolean
is_interleaved_arrays(const struct st_vertex_program *vp,
                      const struct st_vp_variant *vpv,
                      const struct gl_client_array **arrays)
{
   GLuint attr;
   const struct gl_buffer_object *firstBufObj = NULL;
   GLint firstStride = -1;
   const GLubyte *firstPtr = NULL;
   GLboolean userSpaceBuffer = GL_FALSE;

   for (attr = 0; attr < vpv->num_inputs; attr++) {
      const GLuint mesaAttr = vp->index_to_input[attr];
      const struct gl_client_array *array = arrays[mesaAttr];
      const struct gl_buffer_object *bufObj = array->BufferObj;
      const GLsizei stride = array->StrideB; /* in bytes */

      if (attr == 0) {
         /* save info about the first array */
         firstStride = stride;
         firstPtr = array->Ptr;         
         firstBufObj = bufObj;
         userSpaceBuffer = !bufObj || !bufObj->Name;
      }
      else {
         /* check if other arrays interleave with the first, in same buffer */
         if (stride != firstStride)
            return GL_FALSE; /* strides don't match */

         if (bufObj != firstBufObj)
            return GL_FALSE; /* arrays in different VBOs */

         if (abs(array->Ptr - firstPtr) > firstStride)
            return GL_FALSE; /* arrays start too far apart */

         if ((!bufObj || !_mesa_is_bufferobj(bufObj)) != userSpaceBuffer)
            return GL_FALSE; /* mix of VBO and user-space arrays */
      }
   }

   return GL_TRUE;
}


/**
 * Set up for drawing interleaved arrays that all live in one VBO
 * or all live in user space.
 * \param vbuffer  returns vertex buffer info
 * \param velements  returns vertex element info
 */
static void
setup_interleaved_attribs(struct gl_context *ctx,
                          const struct st_vertex_program *vp,
                          const struct st_vp_variant *vpv,
                          const struct gl_client_array **arrays,
                          struct pipe_vertex_buffer *vbuffer,
                          struct pipe_vertex_element velements[],
                          unsigned max_index,
                          unsigned num_instances)
{
   struct st_context *st = st_context(ctx);
   struct pipe_context *pipe = st->pipe;
   GLuint attr;
   const GLubyte *low_addr = NULL;
   GLboolean usingVBO;      /* all arrays in a VBO? */
   struct gl_buffer_object *bufobj;
   GLuint user_buffer_size = 0;
   GLuint vertex_size = 0;  /* bytes per vertex, in bytes */
   GLsizei stride;

   /* Find the lowest address of the arrays we're drawing,
    * Init bufobj and stride.
    */
   if (vpv->num_inputs) {
      const GLuint mesaAttr0 = vp->index_to_input[0];
      const struct gl_client_array *array = arrays[mesaAttr0];

      /* Since we're doing interleaved arrays, we know there'll be at most
       * one buffer object and the stride will be the same for all arrays.
       * Grab them now.
       */
      bufobj = array->BufferObj;
      stride = array->StrideB;

      low_addr = arrays[vp->index_to_input[0]]->Ptr;

      for (attr = 1; attr < vpv->num_inputs; attr++) {
         const GLubyte *start = arrays[vp->index_to_input[attr]]->Ptr;
         low_addr = MIN2(low_addr, start);
      }
   }
   else {
      /* not sure we'll ever have zero inputs, but play it safe */
      bufobj = NULL;
      stride = 0;
      low_addr = 0;
   }

   /* are the arrays in user space? */
   usingVBO = bufobj && _mesa_is_bufferobj(bufobj);

   for (attr = 0; attr < vpv->num_inputs; attr++) {
      const GLuint mesaAttr = vp->index_to_input[attr];
      const struct gl_client_array *array = arrays[mesaAttr];
      unsigned src_offset = (unsigned) (array->Ptr - low_addr);
      GLuint element_size = array->_ElementSize;

      assert(element_size == array->Size * _mesa_sizeof_type(array->Type));

      velements[attr].src_offset = src_offset;
      velements[attr].instance_divisor = array->InstanceDivisor;
      velements[attr].vertex_buffer_index = 0;
      velements[attr].src_format = st_pipe_vertex_format(array->Type,
                                                         array->Size,
                                                         array->Format,
                                                         array->Normalized);
      assert(velements[attr].src_format);

      if (!usingVBO) {
         /* how many bytes referenced by this attribute array? */
         uint divisor = array->InstanceDivisor;
         uint last_index = divisor ? num_instances / divisor : max_index;
         uint bytes = src_offset + stride * last_index + element_size;

         user_buffer_size = MAX2(user_buffer_size, bytes);

         /* update vertex size */
         vertex_size = MAX2(vertex_size, src_offset + element_size);
      }
   }

   /*
    * Return the vbuffer info and setup user-space attrib info, if needed.
    */
   if (vpv->num_inputs == 0) {
      /* just defensive coding here */
      vbuffer->buffer = NULL;
      vbuffer->buffer_offset = 0;
      vbuffer->stride = 0;
      st->num_user_attribs = 0;
   }
   else if (usingVBO) {
      /* all interleaved arrays in a VBO */
      struct st_buffer_object *stobj = st_buffer_object(bufobj);

      vbuffer->buffer = NULL;
      pipe_resource_reference(&vbuffer->buffer, stobj->buffer);
      vbuffer->buffer_offset = pointer_to_offset(low_addr);
      vbuffer->stride = stride;
      st->num_user_attribs = 0;
   }
   else {
      /* all interleaved arrays in user memory */
      vbuffer->buffer = pipe_user_buffer_create(pipe->screen,
                                                (void*) low_addr,
                                                user_buffer_size,
                                                PIPE_BIND_VERTEX_BUFFER);
      vbuffer->buffer_offset = 0;
      vbuffer->stride = stride;

      /* Track user vertex buffers. */
      pipe_resource_reference(&st->user_attrib[0].buffer, vbuffer->buffer);
      st->user_attrib[0].element_size = vertex_size;
      st->user_attrib[0].stride = stride;
      st->num_user_attribs = 1;
   }
}


/**
 * Set up a separate pipe_vertex_buffer and pipe_vertex_element for each
 * vertex attribute.
 * \param vbuffer  returns vertex buffer info
 * \param velements  returns vertex element info
 */
static void
setup_non_interleaved_attribs(struct gl_context *ctx,
                              const struct st_vertex_program *vp,
                              const struct st_vp_variant *vpv,
                              const struct gl_client_array **arrays,
                              struct pipe_vertex_buffer vbuffer[],
                              struct pipe_vertex_element velements[],
                              unsigned max_index,
                              unsigned num_instances)
{
   struct st_context *st = st_context(ctx);
   struct pipe_context *pipe = st->pipe;
   GLuint attr;

   for (attr = 0; attr < vpv->num_inputs; attr++) {
      const GLuint mesaAttr = vp->index_to_input[attr];
      const struct gl_client_array *array = arrays[mesaAttr];
      struct gl_buffer_object *bufobj = array->BufferObj;
      GLuint element_size = array->_ElementSize;
      GLsizei stride = array->StrideB;

      assert(element_size == array->Size * _mesa_sizeof_type(array->Type));

      if (bufobj && _mesa_is_bufferobj(bufobj)) {
         /* Attribute data is in a VBO.
          * Recall that for VBOs, the gl_client_array->Ptr field is
          * really an offset from the start of the VBO, not a pointer.
          */
         struct st_buffer_object *stobj = st_buffer_object(bufobj);
         assert(stobj->buffer);

         vbuffer[attr].buffer = NULL;
         pipe_resource_reference(&vbuffer[attr].buffer, stobj->buffer);
         vbuffer[attr].buffer_offset = pointer_to_offset(array->Ptr);
      }
      else {
         /* wrap user data */
         uint bytes;
         void *ptr;

         if (array->Ptr) {
            uint divisor = array->InstanceDivisor;
            uint last_index = divisor ? num_instances / divisor : max_index;

            bytes = stride * last_index + element_size;

            ptr = (void *) array->Ptr;
         }
         else {
            /* no array, use ctx->Current.Attrib[] value */
            bytes = element_size = sizeof(ctx->Current.Attrib[0]);
            ptr = (void *) ctx->Current.Attrib[mesaAttr];
            stride = 0;
         }

         assert(ptr);
         assert(bytes);

         vbuffer[attr].buffer =
            pipe_user_buffer_create(pipe->screen, ptr, bytes,
                                    PIPE_BIND_VERTEX_BUFFER);

         vbuffer[attr].buffer_offset = 0;

         /* Track user vertex buffers. */
         pipe_resource_reference(&st->user_attrib[attr].buffer, vbuffer[attr].buffer);
         st->user_attrib[attr].element_size = element_size;
         st->user_attrib[attr].stride = stride;
         st->num_user_attribs = MAX2(st->num_user_attribs, attr + 1);
      }

      /* common-case setup */
      vbuffer[attr].stride = stride; /* in bytes */

      velements[attr].src_offset = 0;
      velements[attr].instance_divisor = array->InstanceDivisor;
      velements[attr].vertex_buffer_index = attr;
      velements[attr].src_format = st_pipe_vertex_format(array->Type,
                                                         array->Size,
                                                         array->Format,
                                                         array->Normalized);
      assert(velements[attr].src_format);
   }
}


static void
setup_index_buffer(struct gl_context *ctx,
                   const struct _mesa_index_buffer *ib,
                   struct pipe_index_buffer *ibuffer)
{
   struct st_context *st = st_context(ctx);
   struct pipe_context *pipe = st->pipe;

   memset(ibuffer, 0, sizeof(*ibuffer));
   if (ib) {
      struct gl_buffer_object *bufobj = ib->obj;

      switch (ib->type) {
      case GL_UNSIGNED_INT:
         ibuffer->index_size = 4;
         break;
      case GL_UNSIGNED_SHORT:
         ibuffer->index_size = 2;
         break;
      case GL_UNSIGNED_BYTE:
         ibuffer->index_size = 1;
         break;
      default:
         assert(0);
	 return;
      }

      /* get/create the index buffer object */
      if (bufobj && _mesa_is_bufferobj(bufobj)) {
         /* elements/indexes are in a real VBO */
         struct st_buffer_object *stobj = st_buffer_object(bufobj);
         pipe_resource_reference(&ibuffer->buffer, stobj->buffer);
         ibuffer->offset = pointer_to_offset(ib->ptr);
      }
      else {
         /* element/indicies are in user space memory */
         ibuffer->buffer =
            pipe_user_buffer_create(pipe->screen, (void *) ib->ptr,
                                    ib->count * ibuffer->index_size,
                                    PIPE_BIND_INDEX_BUFFER);
      }
   }
}


/**
 * Prior to drawing, check that any uniforms referenced by the
 * current shader have been set.  If a uniform has not been set,
 * issue a warning.
 */
static void
check_uniforms(struct gl_context *ctx)
{
   struct gl_shader_program *shProg[3] = {
      ctx->Shader.CurrentVertexProgram,
      ctx->Shader.CurrentGeometryProgram,
      ctx->Shader.CurrentFragmentProgram,
   };
   unsigned j;

   for (j = 0; j < 3; j++) {
      unsigned i;

      if (shProg[j] == NULL || !shProg[j]->LinkStatus)
	 continue;

      for (i = 0; i < shProg[j]->Uniforms->NumUniforms; i++) {
         const struct gl_uniform *u = &shProg[j]->Uniforms->Uniforms[i];
         if (!u->Initialized) {
            _mesa_warning(ctx,
                          "Using shader with uninitialized uniform: %s",
                          u->Name);
         }
      }
   }
}

/** Helper code for primitive restart fallback */
#define DO_DRAW(pipe, cur_start, cur_count) \
   do { \
      info.start = cur_start; \
      info.count = cur_count; \
      if (u_trim_pipe_prim(info.mode, &info.count)) { \
         if (transfer) \
            pipe_buffer_unmap(pipe, transfer); \
         pipe->draw_vbo(pipe, &info); \
         if (transfer) { \
            ptr = pipe_buffer_map(pipe, ibuffer->buffer, PIPE_TRANSFER_READ, &transfer); \
            assert(ptr != NULL); \
            ptr = ADD_POINTERS(ptr, ibuffer->offset); \
         } \
      } \
   } while(0)
      
/** More helper code for primitive restart fallback */
#define PRIM_RESTART_LOOP(elements) \
   do { \
      for (i = start; i < end; i++) { \
         if (elements[i] == info.restart_index) { \
            if (cur_count > 0) { \
               /* draw elts up to prev pos */ \
               DO_DRAW(pipe, cur_start, cur_count); \
            } \
            /* begin new prim at next elt */ \
            cur_start = i + 1; \
            cur_count = 0; \
         } \
         else { \
            cur_count++; \
         } \
      } \
      if (cur_count > 0) { \
         DO_DRAW(pipe, cur_start, cur_count); \
      } \
   } while (0)

static void
handle_fallback_primitive_restart(struct pipe_context *pipe,
                                  const struct _mesa_index_buffer *ib,
                                  struct pipe_index_buffer *ibuffer,
                                  struct pipe_draw_info *orig_info)
{
   const unsigned start = orig_info->start;
   const unsigned count = orig_info->count;
   const unsigned end = start + count;
   struct pipe_draw_info info = *orig_info;
   struct pipe_transfer *transfer = NULL;
   unsigned instance, i, cur_start, cur_count;
   const void *ptr;

   info.primitive_restart = FALSE;

   if (!info.indexed) {
      /* Splitting the draw arrays call is handled by the VBO module */
      if (u_trim_pipe_prim(info.mode, &info.count))
         pipe->draw_vbo(pipe, &info);

      return;
   }

   /* info.indexed == TRUE */
   assert(ibuffer);
   assert(ibuffer->buffer);

   if (ib) {
      struct gl_buffer_object *bufobj = ib->obj;
      if (bufobj && bufobj->Name) {
         ptr = NULL;
      }
      else {
         ptr = ib->ptr;
      }
   } else {
      ptr = NULL;
   }

   if (!ptr)
      ptr = pipe_buffer_map(pipe, ibuffer->buffer, PIPE_TRANSFER_READ, &transfer);

   if (!ptr)
     return;
   ptr = ADD_POINTERS(ptr, ibuffer->offset);

   /* Need to loop over instances as well to preserve draw order */
   for (instance = 0; instance < orig_info->instance_count; instance++) {
      info.start_instance = instance + orig_info->start_instance;
      info.instance_count = 1;
      cur_start = start;
      cur_count = 0;

      switch (ibuffer->index_size) {
      case 1:
         {
            const ubyte *elt_ub = (const ubyte *)ptr; 
            PRIM_RESTART_LOOP(elt_ub);
         }
         break;
      case 2:
         {
            const ushort *elt_us = (const ushort *)ptr;
            PRIM_RESTART_LOOP(elt_us);
         }
         break;
      case 4:
         {
            const uint *elt_ui = (const uint *)ptr;
            PRIM_RESTART_LOOP(elt_ui);
         }
         break;
      default:
         assert(0 && "bad index_size in handle_fallback_primitive_restart()");
      }
   }

   if (transfer)
      pipe_buffer_unmap(pipe, transfer);
}


/**
 * Translate OpenGL primtive type (GL_POINTS, GL_TRIANGLE_STRIP, etc) to
 * the corresponding Gallium type.
 */
static unsigned
translate_prim(const struct gl_context *ctx, unsigned prim)
{
   /* GL prims should match Gallium prims, spot-check a few */
   assert(GL_POINTS == PIPE_PRIM_POINTS);
   assert(GL_QUADS == PIPE_PRIM_QUADS);
   assert(GL_TRIANGLE_STRIP_ADJACENCY == PIPE_PRIM_TRIANGLE_STRIP_ADJACENCY);

   /* Avoid quadstrips if it's easy to do so:
    * Note: it's important to do the correct trimming if we change the
    * prim type!  We do that wherever this function is called.
    */
   if (prim == GL_QUAD_STRIP &&
       ctx->Light.ShadeModel != GL_FLAT &&
       ctx->Polygon.FrontMode == GL_FILL &&
       ctx->Polygon.BackMode == GL_FILL)
      prim = GL_TRIANGLE_STRIP;

   return prim;
}


static void
st_validate_varrays(struct gl_context *ctx,
                    const struct gl_client_array **arrays,
                    unsigned max_index,
                    unsigned num_instances)
{
   struct st_context *st = st_context(ctx);
   const struct st_vertex_program *vp;
   const struct st_vp_variant *vpv;
   struct pipe_vertex_buffer vbuffer[PIPE_MAX_SHADER_INPUTS];
   struct pipe_vertex_element velements[PIPE_MAX_ATTRIBS];
   unsigned num_vbuffers, num_velements;
   GLuint attr;
   unsigned i;

   /* must get these after state validation! */
   vp = st->vp;
   vpv = st->vp_variant;

   memset(velements, 0, sizeof(struct pipe_vertex_element) * vpv->num_inputs);

   /* Unreference any user vertex buffers. */
   for (i = 0; i < st->num_user_attribs; i++) {
      pipe_resource_reference(&st->user_attrib[i].buffer, NULL);
   }
   st->num_user_attribs = 0;

   /*
    * Setup the vbuffer[] and velements[] arrays.
    */
   if (is_interleaved_arrays(vp, vpv, arrays)) {
      setup_interleaved_attribs(ctx, vp, vpv, arrays, vbuffer, velements,
                                max_index, num_instances);

      num_vbuffers = 1;
      num_velements = vpv->num_inputs;
      if (num_velements == 0)
         num_vbuffers = 0;
   }
   else {
      setup_non_interleaved_attribs(ctx, vp, vpv, arrays,
                                    vbuffer, velements, max_index,
                                    num_instances);
      num_vbuffers = vpv->num_inputs;
      num_velements = vpv->num_inputs;
   }

   cso_set_vertex_buffers(st->cso_context, num_vbuffers, vbuffer);
   cso_set_vertex_elements(st->cso_context, num_velements, velements);

   /* unreference buffers (frees wrapped user-space buffer objects)
    * This is OK, because the pipe driver should reference buffers by itself
    * in set_vertex_buffers. */
   for (attr = 0; attr < num_vbuffers; attr++) {
      pipe_resource_reference(&vbuffer[attr].buffer, NULL);
      assert(!vbuffer[attr].buffer);
   }
}


/**
 * This function gets plugged into the VBO module and is called when
 * we have something to render.
 * Basically, translate the information into the format expected by gallium.
 */
void
st_draw_vbo(struct gl_context *ctx,
            const struct gl_client_array **arrays,
            const struct _mesa_prim *prims,
            GLuint nr_prims,
            const struct _mesa_index_buffer *ib,
	    GLboolean index_bounds_valid,
            GLuint min_index,
            GLuint max_index)
{
   struct st_context *st = st_context(ctx);
   struct pipe_context *pipe = st->pipe;
   struct pipe_index_buffer ibuffer;
   struct pipe_draw_info info;
   unsigned i, num_instances = 1;
   GLboolean new_array =
      st->dirty.st &&
      (st->dirty.mesa & (_NEW_ARRAY | _NEW_PROGRAM | _NEW_BUFFER_OBJECT)) != 0;

   /* Mesa core state should have been validated already */
   assert(ctx->NewState == 0x0);

   if (ib) {
      /* Gallium probably doesn't want this in some cases. */
      if (!index_bounds_valid)
         if (!all_varyings_in_vbos(arrays))
            vbo_get_minmax_index(ctx, prims, ib, &min_index, &max_index);

      for (i = 0; i < nr_prims; i++) {
         num_instances = MAX2(num_instances, prims[i].num_instances);
      }
   }
   else {
      /* Get min/max index for non-indexed drawing. */
      min_index = ~0;
      max_index = 0;

      for (i = 0; i < nr_prims; i++) {
         min_index = MIN2(min_index, prims[i].start);
         max_index = MAX2(max_index, prims[i].start + prims[i].count - 1);
         num_instances = MAX2(num_instances, prims[i].num_instances);
      }
   }

   /* Validate state. */
   if (st->dirty.st) {
      GLboolean vertDataEdgeFlags;

      /* sanity check for pointer arithmetic below */
      assert(sizeof(arrays[0]->Ptr[0]) == 1);

      vertDataEdgeFlags = arrays[VERT_ATTRIB_EDGEFLAG]->BufferObj &&
                          arrays[VERT_ATTRIB_EDGEFLAG]->BufferObj->Name;
      if (vertDataEdgeFlags != st->vertdata_edgeflags) {
         st->vertdata_edgeflags = vertDataEdgeFlags;
         st->dirty.st |= ST_NEW_EDGEFLAGS_DATA;
      }

      st_validate_state(st);

      if (new_array) {
         st_validate_varrays(ctx, arrays, max_index, num_instances);
      }

#if 0
      if (MESA_VERBOSE & VERBOSE_GLSL) {
         check_uniforms(ctx);
      }
#else
      (void) check_uniforms;
#endif
   }

   /* Notify the driver that the content of user buffers may have been
    * changed. */
   assert(max_index >= min_index);
   if (!new_array && st->num_user_attribs) {
      for (i = 0; i < st->num_user_attribs; i++) {
         if (st->user_attrib[i].buffer) {
            unsigned element_size = st->user_attrib[i].element_size;
            unsigned stride = st->user_attrib[i].stride;
            unsigned min_offset = min_index * stride;
            unsigned max_offset = max_index * stride + element_size;

            assert(max_offset > min_offset);

            pipe->redefine_user_buffer(pipe, st->user_attrib[i].buffer,
                                       min_offset,
                                       max_offset - min_offset);
         }
      }
   }

   setup_index_buffer(ctx, ib, &ibuffer);
   pipe->set_index_buffer(pipe, &ibuffer);

   util_draw_init_info(&info);
   if (ib) {
      info.indexed = TRUE;
      if (min_index != ~0 && max_index != ~0) {
         info.min_index = min_index;
         info.max_index = max_index;
      }
   }

   info.primitive_restart = ctx->Array.PrimitiveRestart;
   info.restart_index = ctx->Array.RestartIndex;

   /* do actual drawing */
   for (i = 0; i < nr_prims; i++) {
      info.mode = translate_prim( ctx, prims[i].mode );
      info.start = prims[i].start;
      info.count = prims[i].count;
      info.instance_count = prims[i].num_instances;
      info.index_bias = prims[i].basevertex;
      if (!ib) {
         info.min_index = info.start;
         info.max_index = info.start + info.count - 1;
      }

      if (info.primitive_restart) {
         /*
          * Handle primitive restart for drivers that doesn't support it.
          *
          * The VBO module handles restart inside of draw_arrays for us,
          * but we should still remove the primitive_restart flag on the
          * info struct, the fallback function does this for us. Just
          * remove the flag for all drivers in this case as well.
          */
         if (st->sw_primitive_restart || !info.indexed)
            handle_fallback_primitive_restart(pipe, ib, &ibuffer, &info);
         else
            /* don't trim, restarts might be inside index list */
            pipe->draw_vbo(pipe, &info);
      }
      else if (u_trim_pipe_prim(info.mode, &info.count))
         pipe->draw_vbo(pipe, &info);
   }

   pipe_resource_reference(&ibuffer.buffer, NULL);
}


void
st_init_draw(struct st_context *st)
{
   struct gl_context *ctx = st->ctx;

   vbo_set_draw_func(ctx, st_draw_vbo);

#if FEATURE_feedback || FEATURE_rastpos
   st->draw = draw_create(st->pipe); /* for selection/feedback */

   /* Disable draw options that might convert points/lines to tris, etc.
    * as that would foul-up feedback/selection mode.
    */
   draw_wide_line_threshold(st->draw, 1000.0f);
   draw_wide_point_threshold(st->draw, 1000.0f);
   draw_enable_line_stipple(st->draw, FALSE);
   draw_enable_point_sprites(st->draw, FALSE);
#endif
}


void
st_destroy_draw(struct st_context *st)
{
#if FEATURE_feedback || FEATURE_rastpos
   draw_destroy(st->draw);
#endif
}