aboutsummaryrefslogtreecommitdiff
path: root/mesalib/src/mesa/tnl/t_vb_program.c
blob: 5ade54343dddede80cf018cf158570ca6efd8e38 (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
/*
 * Mesa 3-D graphics library
 * Version:  7.6
 *
 * 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.
 */


/**
 * \file tnl/t_vb_program.c
 * \brief Pipeline stage for executing vertex programs.
 * \author Brian Paul,  Keith Whitwell
 */


#include "main/glheader.h"
#include "main/colormac.h"
#include "main/macros.h"
#include "main/imports.h"
#include "main/samplerobj.h"
#include "math/m_xform.h"
#include "program/prog_instruction.h"
#include "program/prog_statevars.h"
#include "program/prog_execute.h"
#include "swrast/s_context.h"

#include "tnl/tnl.h"
#include "tnl/t_context.h"
#include "tnl/t_pipeline.h"


#ifdef NAN_CHECK
/** Check for NaNs and very large values */
static inline void
check_float(float x)
{
   assert(!IS_INF_OR_NAN(x));
   assert(1.0e-15 <= x && x <= 1.0e15);
}
#endif


/*!
 * Private storage for the vertex program pipeline stage.
 */
struct vp_stage_data {
   /** The results of running the vertex program go into these arrays. */
   GLvector4f results[VARYING_SLOT_MAX];

   GLvector4f ndcCoords;              /**< normalized device coords */
   GLubyte *clipmask;                 /**< clip flags */
   GLubyte ormask, andmask;           /**< for clipping */

   GLboolean vertex_textures;

   struct gl_program_machine machine;
};


#define VP_STAGE_DATA(stage) ((struct vp_stage_data *)(stage->privatePtr))


static void
userclip( struct gl_context *ctx,
          GLvector4f *clip,
          GLubyte *clipmask,
          GLubyte *clipormask,
          GLubyte *clipandmask )
{
   GLuint p;

   for (p = 0; p < ctx->Const.MaxClipPlanes; p++) {
      if (ctx->Transform.ClipPlanesEnabled & (1 << p)) {
	 GLuint nr, i;
	 const GLfloat a = ctx->Transform._ClipUserPlane[p][0];
	 const GLfloat b = ctx->Transform._ClipUserPlane[p][1];
	 const GLfloat c = ctx->Transform._ClipUserPlane[p][2];
	 const GLfloat d = ctx->Transform._ClipUserPlane[p][3];
         GLfloat *coord = (GLfloat *)clip->data;
         GLuint stride = clip->stride;
         GLuint count = clip->count;

	 for (nr = 0, i = 0 ; i < count ; i++) {
	    GLfloat dp = (coord[0] * a + 
			  coord[1] * b +
			  coord[2] * c +
			  coord[3] * d);

	    if (dp < 0) {
	       nr++;
	       clipmask[i] |= CLIP_USER_BIT;
	    }

	    STRIDE_F(coord, stride);
	 }

	 if (nr > 0) {
	    *clipormask |= CLIP_USER_BIT;
	    if (nr == count) {
	       *clipandmask |= CLIP_USER_BIT;
	       return;
	    }
	 }
      }
   }
}


static GLboolean
do_ndc_cliptest(struct gl_context *ctx, struct vp_stage_data *store)
{
   TNLcontext *tnl = TNL_CONTEXT(ctx);
   struct vertex_buffer *VB = &tnl->vb;
   /* Cliptest and perspective divide.  Clip functions must clear
    * the clipmask.
    */
   store->ormask = 0;
   store->andmask = CLIP_FRUSTUM_BITS;

   tnl_clip_prepare(ctx);

   if (tnl->NeedNdcCoords) {
      VB->NdcPtr =
         _mesa_clip_tab[VB->ClipPtr->size]( VB->ClipPtr,
                                            &store->ndcCoords,
                                            store->clipmask,
                                            &store->ormask,
                                            &store->andmask,
					    !ctx->Transform.DepthClamp );
   }
   else {
      VB->NdcPtr = NULL;
      _mesa_clip_np_tab[VB->ClipPtr->size]( VB->ClipPtr,
                                            NULL,
                                            store->clipmask,
                                            &store->ormask,
                                            &store->andmask,
					    !ctx->Transform.DepthClamp );
   }

   if (store->andmask) {
      /* All vertices are outside the frustum */
      return GL_FALSE;
   }

   /* Test userclip planes.  This contributes to VB->ClipMask.
    */
   /** XXX NEW_SLANG _Enabled ??? */
   if (ctx->Transform.ClipPlanesEnabled && (!ctx->VertexProgram._Enabled ||
      ctx->VertexProgram.Current->IsPositionInvariant)) {
      userclip( ctx,
		VB->ClipPtr,
		store->clipmask,
		&store->ormask,
		&store->andmask );

      if (store->andmask) {
	 return GL_FALSE;
      }
   }

   VB->ClipAndMask = store->andmask;
   VB->ClipOrMask = store->ormask;
   VB->ClipMask = store->clipmask;

   return GL_TRUE;
}


/**
 * XXX the texture sampling code in this module is a bit of a hack.
 * The texture sampling code is in swrast, though it doesn't have any
 * real dependencies on the rest of swrast.  It should probably be
 * moved into main/ someday.
 */
static void
vp_fetch_texel(struct gl_context *ctx, const GLfloat texcoord[4], GLfloat lambda,
               GLuint unit, GLfloat color[4])
{
   SWcontext *swrast = SWRAST_CONTEXT(ctx);

   /* XXX use a float-valued TextureSample routine here!!! */
   swrast->TextureSample[unit](ctx, _mesa_get_samplerobj(ctx, unit),
                               ctx->Texture.Unit[unit]._Current,
                               1, (const GLfloat (*)[4]) texcoord,
                               &lambda,  (GLfloat (*)[4]) color);
}


/**
 * Called via ctx->Driver.ProgramStringNotify() after a new vertex program
 * string has been parsed.
 */
GLboolean
_tnl_program_string(struct gl_context *ctx, GLenum target, struct gl_program *program)
{
   /* No-op.
    * If we had derived anything from the program that was private to this
    * stage we'd recompute/validate it here.
    */
   return GL_TRUE;
}


/**
 * Initialize virtual machine state prior to executing vertex program.
 */
static void
init_machine(struct gl_context *ctx, struct gl_program_machine *machine,
             GLuint instID)
{
   /* Input registers get initialized from the current vertex attribs */
   memcpy(machine->VertAttribs, ctx->Current.Attrib,
          MAX_VERTEX_GENERIC_ATTRIBS * 4 * sizeof(GLfloat));

   machine->NumDeriv = 0;

   /* init condition codes */
   machine->CondCodes[0] = COND_EQ;
   machine->CondCodes[1] = COND_EQ;
   machine->CondCodes[2] = COND_EQ;
   machine->CondCodes[3] = COND_EQ;

   /* init call stack */
   machine->StackDepth = 0;

   machine->FetchTexelLod = vp_fetch_texel;
   machine->FetchTexelDeriv = NULL; /* not used by vertex programs */

   machine->Samplers = ctx->VertexProgram._Current->Base.SamplerUnits;

   machine->SystemValues[SYSTEM_VALUE_INSTANCE_ID][0] = (GLfloat) instID;
}


/**
 * Map the texture images which the vertex program will access (if any).
 */
static void
map_textures(struct gl_context *ctx, const struct gl_vertex_program *vp)
{
   GLuint u;

   for (u = 0; u < ctx->Const.VertexProgram.MaxTextureImageUnits; u++) {
      if (vp->Base.TexturesUsed[u]) {
         /* Note: _Current *should* correspond to the target indicated
          * in TexturesUsed[u].
          */
         _swrast_map_texture(ctx, ctx->Texture.Unit[u]._Current);
      }
   }
}


/**
 * Unmap the texture images which were used by the vertex program (if any).
 */
static void
unmap_textures(struct gl_context *ctx, const struct gl_vertex_program *vp)
{
   GLuint u;

   for (u = 0; u < ctx->Const.VertexProgram.MaxTextureImageUnits; u++) {
      if (vp->Base.TexturesUsed[u]) {
         /* Note: _Current *should* correspond to the target indicated
          * in TexturesUsed[u].
          */
         _swrast_unmap_texture(ctx, ctx->Texture.Unit[u]._Current);
      }
   }
}


/**
 * This function executes vertex programs
 */
static GLboolean
run_vp( struct gl_context *ctx, struct tnl_pipeline_stage *stage )
{
   TNLcontext *tnl = TNL_CONTEXT(ctx);
   struct vp_stage_data *store = VP_STAGE_DATA(stage);
   struct vertex_buffer *VB = &tnl->vb;
   struct gl_vertex_program *program = ctx->VertexProgram._Current;
   struct gl_program_machine *machine = &store->machine;
   GLuint outputs[VARYING_SLOT_MAX], numOutputs;
   GLuint i, j;

   if (!program)
      return GL_TRUE;

   /* ARB program or vertex shader */
   _mesa_load_state_parameters(ctx, program->Base.Parameters);

   /* make list of outputs to save some time below */
   numOutputs = 0;
   for (i = 0; i < VARYING_SLOT_MAX; i++) {
      if (program->Base.OutputsWritten & BITFIELD64_BIT(i)) {
         outputs[numOutputs++] = i;
      }
   }

   /* Allocate result vectors.  We delay this until now to avoid allocating
    * memory that would never be used if we don't run the software tnl pipeline.
    */
   if (!store->results[0].storage) {
      for (i = 0; i < VARYING_SLOT_MAX; i++) {
         assert(!store->results[i].storage);
         _mesa_vector4f_alloc( &store->results[i], 0, VB->Size, 32 );
         store->results[i].size = 4;
      }
   }

   map_textures(ctx, program);

   for (i = 0; i < VB->Count; i++) {
      GLuint attr;

      init_machine(ctx, machine, tnl->CurInstance);

#if 0
      printf("Input  %d: %f, %f, %f, %f\n", i,
             VB->AttribPtr[0]->data[i][0],
             VB->AttribPtr[0]->data[i][1],
             VB->AttribPtr[0]->data[i][2],
             VB->AttribPtr[0]->data[i][3]);
      printf("   color: %f, %f, %f, %f\n",
             VB->AttribPtr[3]->data[i][0],
             VB->AttribPtr[3]->data[i][1],
             VB->AttribPtr[3]->data[i][2],
             VB->AttribPtr[3]->data[i][3]);
      printf("  normal: %f, %f, %f, %f\n",
             VB->AttribPtr[2]->data[i][0],
             VB->AttribPtr[2]->data[i][1],
             VB->AttribPtr[2]->data[i][2],
             VB->AttribPtr[2]->data[i][3]);
#endif

      /* the vertex array case */
      for (attr = 0; attr < VERT_ATTRIB_MAX; attr++) {
	 if (program->Base.InputsRead & BITFIELD64_BIT(attr)) {
	    const GLubyte *ptr = (const GLubyte*) VB->AttribPtr[attr]->data;
	    const GLuint size = VB->AttribPtr[attr]->size;
	    const GLuint stride = VB->AttribPtr[attr]->stride;
	    const GLfloat *data = (GLfloat *) (ptr + stride * i);
#ifdef NAN_CHECK
            check_float(data[0]);
            check_float(data[1]);
            check_float(data[2]);
            check_float(data[3]);
#endif
	    COPY_CLEAN_4V(machine->VertAttribs[attr], size, data);
	 }
      }

      /* execute the program */
      _mesa_execute_program(ctx, &program->Base, machine);

      /* copy the output registers into the VB->attribs arrays */
      for (j = 0; j < numOutputs; j++) {
         const GLuint attr = outputs[j];
#ifdef NAN_CHECK
         check_float(machine->Outputs[attr][0]);
         check_float(machine->Outputs[attr][1]);
         check_float(machine->Outputs[attr][2]);
         check_float(machine->Outputs[attr][3]);
#endif
         COPY_4V(store->results[attr].data[i], machine->Outputs[attr]);
      }

      /* FOGC is a special case.  Fragment shader expects (f,0,0,1) */
      if (program->Base.OutputsWritten & BITFIELD64_BIT(VARYING_SLOT_FOGC)) {
         store->results[VARYING_SLOT_FOGC].data[i][1] = 0.0;
         store->results[VARYING_SLOT_FOGC].data[i][2] = 0.0;
         store->results[VARYING_SLOT_FOGC].data[i][3] = 1.0;
      }
#ifdef NAN_CHECK
      ASSERT(machine->Outputs[0][3] != 0.0F);
#endif
#if 0
      printf("HPOS: %f %f %f %f\n",
             machine->Outputs[0][0], 
             machine->Outputs[0][1], 
             machine->Outputs[0][2], 
             machine->Outputs[0][3]);
#endif
   }

   unmap_textures(ctx, program);

   if (program->IsPositionInvariant) {
      /* We need the exact same transform as in the fixed function path here
       * to guarantee invariance, depending on compiler optimization flags
       * results could be different otherwise.
       */
      VB->ClipPtr = TransformRaw( &store->results[0],
				  &ctx->_ModelProjectMatrix,
				  VB->AttribPtr[0] );

      /* Drivers expect this to be clean to element 4...
       */
      switch (VB->ClipPtr->size) {
      case 1:
	 /* impossible */
      case 2:
	 _mesa_vector4f_clean_elem( VB->ClipPtr, VB->Count, 2 );
	 /* fall-through */
      case 3:
	 _mesa_vector4f_clean_elem( VB->ClipPtr, VB->Count, 3 );
	 /* fall-through */
      case 4:
	 break;
      }
   }
   else {
      /* Setup the VB pointers so that the next pipeline stages get
       * their data from the right place (the program output arrays).
       */
      VB->ClipPtr = &store->results[VARYING_SLOT_POS];
      VB->ClipPtr->size = 4;
      VB->ClipPtr->count = VB->Count;
   }

   VB->AttribPtr[VERT_ATTRIB_COLOR0] = &store->results[VARYING_SLOT_COL0];
   VB->AttribPtr[VERT_ATTRIB_COLOR1] = &store->results[VARYING_SLOT_COL1];
   VB->AttribPtr[VERT_ATTRIB_FOG] = &store->results[VARYING_SLOT_FOGC];
   VB->AttribPtr[_TNL_ATTRIB_POINTSIZE] = &store->results[VARYING_SLOT_PSIZ];
   VB->BackfaceColorPtr = &store->results[VARYING_SLOT_BFC0];
   VB->BackfaceSecondaryColorPtr = &store->results[VARYING_SLOT_BFC1];

   for (i = 0; i < ctx->Const.MaxTextureCoordUnits; i++) {
      VB->AttribPtr[_TNL_ATTRIB_TEX0 + i]
         = &store->results[VARYING_SLOT_TEX0 + i];
   }

   for (i = 0; i < ctx->Const.MaxVarying; i++) {
      if (program->Base.OutputsWritten & BITFIELD64_BIT(VARYING_SLOT_VAR0 + i)) {
         /* Note: varying results get put into the generic attributes */
	 VB->AttribPtr[VERT_ATTRIB_GENERIC0+i]
            = &store->results[VARYING_SLOT_VAR0 + i];
      }
   }


   /* Perform NDC and cliptest operations:
    */
   return do_ndc_cliptest(ctx, store);
}


/**
 * Called the first time stage->run is called.  In effect, don't
 * allocate data until the first time the stage is run.
 */
static GLboolean
init_vp(struct gl_context *ctx, struct tnl_pipeline_stage *stage)
{
   TNLcontext *tnl = TNL_CONTEXT(ctx);
   struct vertex_buffer *VB = &(tnl->vb);
   struct vp_stage_data *store;
   const GLuint size = VB->Size;

   stage->privatePtr = calloc(1, sizeof(*store));
   store = VP_STAGE_DATA(stage);
   if (!store)
      return GL_FALSE;

   /* a few other misc allocations */
   _mesa_vector4f_alloc( &store->ndcCoords, 0, size, 32 );
   store->clipmask = _mesa_align_malloc(sizeof(GLubyte)*size, 32 );

   return GL_TRUE;
}


/**
 * Destructor for this pipeline stage.
 */
static void
dtr(struct tnl_pipeline_stage *stage)
{
   struct vp_stage_data *store = VP_STAGE_DATA(stage);

   if (store) {
      GLuint i;

      /* free the vertex program result arrays */
      for (i = 0; i < VARYING_SLOT_MAX; i++)
         _mesa_vector4f_free( &store->results[i] );

      /* free misc arrays */
      _mesa_vector4f_free( &store->ndcCoords );
      _mesa_align_free( store->clipmask );

      free( store );
      stage->privatePtr = NULL;
   }
}


static void
validate_vp_stage(struct gl_context *ctx, struct tnl_pipeline_stage *stage)
{
   if (ctx->VertexProgram._Current) {
      _swrast_update_texture_samplers(ctx);
   }
}



/**
 * Public description of this pipeline stage.
 */
const struct tnl_pipeline_stage _tnl_vertex_program_stage =
{
   "vertex-program",
   NULL,			/* private_data */
   init_vp,			/* create */
   dtr,				/* destroy */
   validate_vp_stage, 		/* validate */
   run_vp			/* run -- initially set to ctr */
};