summaryrefslogtreecommitdiff
path: root/src/mesa/drivers/dri/r300/r300_swtcl.c
blob: 4a6762ff8303f2deded4e35d81d375df3ae60bd9 (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
/**************************************************************************

Copyright (C) 2007 Dave Airlie

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
on 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
THE COPYRIGHT OWNER(S) 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.

**************************************************************************/

/*
 * Authors:
 *   Dave Airlie <airlied@linux.ie>
 *   Maciej Cencora <m.cencora@gmail.com>
 */

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

#include "r300_state.h"
#include "r300_swtcl.h"
#include "r300_emit.h"
#include "r300_tex.h"
#include "r300_render.h"
#include "main/simple_list.h"

#define EMIT_ATTR( ATTR, STYLE )					\
do {									\
	rmesa->radeon.swtcl.vertex_attrs[rmesa->radeon.swtcl.vertex_attr_count].attrib = (ATTR);	\
	rmesa->radeon.swtcl.vertex_attrs[rmesa->radeon.swtcl.vertex_attr_count].format = (STYLE);	\
	rmesa->radeon.swtcl.vertex_attr_count++;					\
} while (0)

#define EMIT_PAD( N )							\
do {									\
   rmesa->radeon.swtcl.vertex_attrs[rmesa->radeon.swtcl.vertex_attr_count].attrib = 0;		\
   rmesa->radeon.swtcl.vertex_attrs[rmesa->radeon.swtcl.vertex_attr_count].format = EMIT_PAD;	\
   rmesa->radeon.swtcl.vertex_attrs[rmesa->radeon.swtcl.vertex_attr_count].offset = (N);		\
   rmesa->radeon.swtcl.vertex_attr_count++;					\
} while (0)

#define ADD_ATTR(_attr, _format, _dst_loc, _swizzle, _write_mask, _normalize) \
do { \
	attrs[num_attrs].element = (_attr); \
	attrs[num_attrs].data_type = (_format); \
	attrs[num_attrs].dst_loc = (_dst_loc); \
	attrs[num_attrs].swizzle = (_swizzle); \
	attrs[num_attrs].write_mask = (_write_mask); \
	attrs[num_attrs]._signed = 0; \
	attrs[num_attrs].normalize = (_normalize); \
	++num_attrs; \
} while (0)

void r300ChooseSwtclVertexFormat(struct gl_context *ctx, GLuint *_InputsRead,  GLuint *_OutputsWritten)
{
	r300ContextPtr rmesa = R300_CONTEXT( ctx );
	TNLcontext *tnl = TNL_CONTEXT(ctx);
	struct vertex_buffer *VB = &tnl->vb;
	int first_free_tex = 0;
	GLuint InputsRead = 0;
	GLuint OutputsWritten = 0;
	int num_attrs = 0;
	GLuint fp_reads = rmesa->selected_fp->InputsRead;
	struct vertex_attribute *attrs = rmesa->vbuf.attribs;

	radeon_print(RADEON_SWRENDER, RADEON_VERBOSE, "%s\n", __func__);
	rmesa->swtcl.coloroffset = rmesa->swtcl.specoffset = 0;
	rmesa->radeon.swtcl.vertex_attr_count = 0;

	if (RADEON_DEBUG & RADEON_VERTS)
		fprintf(stderr, "%s\n", __func__);

	/* We always want non Ndc coords format */
	VB->AttribPtr[VERT_ATTRIB_POS] = VB->ClipPtr;

	/* Always write position vector */
	InputsRead |= 1 << VERT_ATTRIB_POS;
	OutputsWritten |= 1 << VERT_RESULT_HPOS;
	EMIT_ATTR( _TNL_ATTRIB_POS, EMIT_4F );
	ADD_ATTR(VERT_ATTRIB_POS, R300_DATA_TYPE_FLOAT_4, SWTCL_OVM_POS, SWIZZLE_XYZW, MASK_XYZW, 0);
	rmesa->swtcl.coloroffset = 4;

	if (fp_reads & FRAG_BIT_COL0) {
		InputsRead |= 1 << VERT_ATTRIB_COLOR0;
		OutputsWritten |= 1 << VERT_RESULT_COL0;
#if MESA_LITTLE_ENDIAN
		EMIT_ATTR( _TNL_ATTRIB_COLOR0, EMIT_4UB_4F_RGBA );
		ADD_ATTR(VERT_ATTRIB_COLOR0, R300_DATA_TYPE_BYTE, SWTCL_OVM_COLOR0, SWIZZLE_XYZW, MASK_XYZW, 1);
#else
		EMIT_ATTR( _TNL_ATTRIB_COLOR0, EMIT_4UB_4F_ABGR );
		ADD_ATTR(VERT_ATTRIB_COLOR0, R300_DATA_TYPE_BYTE, SWTCL_OVM_COLOR0, SWIZZLE_XYZW, MASK_XYZW, 1);
#endif
	}

	if (fp_reads & FRAG_BIT_COL1) {
		GLuint swiz = MAKE_SWIZZLE4(SWIZZLE_X, SWIZZLE_Y, SWIZZLE_Z, SWIZZLE_ONE);
		InputsRead |= 1 << VERT_ATTRIB_COLOR1;
		OutputsWritten |= 1 << VERT_RESULT_COL1;
#if MESA_LITTLE_ENDIAN
		EMIT_ATTR( _TNL_ATTRIB_COLOR1, EMIT_4UB_4F_RGBA );
		ADD_ATTR(VERT_ATTRIB_COLOR1, R300_DATA_TYPE_BYTE, SWTCL_OVM_COLOR1, swiz, MASK_XYZW, 1);
#else
		EMIT_ATTR( _TNL_ATTRIB_COLOR1, EMIT_4UB_4F_ABGR );
		ADD_ATTR(VERT_ATTRIB_COLOR1, R300_DATA_TYPE_BYTE, SWTCL_OVM_COLOR1, swiz, MASK_XYZW, 1);
#endif
		rmesa->swtcl.specoffset = rmesa->swtcl.coloroffset + 1;
	}

	if (ctx->Light.Enabled && ctx->Light.Model.TwoSide) {
		VB->AttribPtr[VERT_ATTRIB_GENERIC0] = VB->BackfaceColorPtr;
		OutputsWritten |= 1 << VERT_RESULT_BFC0;
#if MESA_LITTLE_ENDIAN
		EMIT_ATTR( _TNL_ATTRIB_GENERIC0, EMIT_4UB_4F_RGBA );
		ADD_ATTR(VERT_ATTRIB_GENERIC0, R300_DATA_TYPE_BYTE, SWTCL_OVM_COLOR2, SWIZZLE_XYZW, MASK_XYZW, 1);
#else
		EMIT_ATTR( _TNL_ATTRIB_GENERIC0, EMIT_4UB_4F_ABGR );
		ADD_ATTR(VERT_ATTRIB_GENERIC0, R300_DATA_TYPE_BYTE, SWTCL_OVM_COLOR2, SWIZZLE_XYZW, MASK_XYZW, 1);
#endif
		if (fp_reads & FRAG_BIT_COL1) {
			VB->AttribPtr[VERT_ATTRIB_GENERIC1] = VB->BackfaceSecondaryColorPtr;
			GLuint swiz = MAKE_SWIZZLE4(SWIZZLE_X, SWIZZLE_Y, SWIZZLE_Z, SWIZZLE_ONE);
			OutputsWritten |= 1 << VERT_RESULT_BFC1;
#if MESA_LITTLE_ENDIAN
			EMIT_ATTR( _TNL_ATTRIB_GENERIC1, EMIT_4UB_4F_RGBA );
			ADD_ATTR(VERT_ATTRIB_GENERIC1, R300_DATA_TYPE_BYTE, SWTCL_OVM_COLOR3, swiz, MASK_XYZW, 1);
#else
			EMIT_ATTR( _TNL_ATTRIB_GENERIC1, EMIT_4UB_4F_ABGR );
			ADD_ATTR(VERT_ATTRIB_GENERIC1, R300_DATA_TYPE_BYTE, SWTCL_OVM_COLOR3, swiz, MASK_XYZW, 1);
#endif
		}
	}

	if (RENDERINPUTS_TEST(tnl->render_inputs_bitset, _TNL_ATTRIB_POINTSIZE )) {
		GLuint swiz = MAKE_SWIZZLE4(SWIZZLE_X, SWIZZLE_ZERO, SWIZZLE_ZERO, SWIZZLE_ZERO);
		InputsRead |= 1 << VERT_ATTRIB_POINT_SIZE;
		OutputsWritten |= 1 << VERT_RESULT_PSIZ;
		EMIT_ATTR( _TNL_ATTRIB_POINTSIZE, EMIT_1F );
		ADD_ATTR(VERT_ATTRIB_POINT_SIZE, R300_DATA_TYPE_FLOAT_1, SWTCL_OVM_POINT_SIZE, swiz, MASK_X, 0);
	}

	if (rmesa->selected_fp->wpos_attr != FRAG_ATTRIB_MAX) {
		int tex_id = rmesa->selected_fp->wpos_attr - FRAG_ATTRIB_TEX0;

		VB->AttribPtr[VERT_ATTRIB_TEX0 + tex_id] = VB->AttribPtr[VERT_ATTRIB_POS];
		VB->AttribPtr[_TNL_ATTRIB_TEX0 + tex_id] = VB->AttribPtr[VERT_ATTRIB_POS];
		RENDERINPUTS_SET(tnl->render_inputs_bitset, _TNL_ATTRIB_TEX0 + tex_id);
	}

	if (rmesa->selected_fp->fog_attr != FRAG_ATTRIB_MAX) {
		int tex_id = rmesa->selected_fp->fog_attr - FRAG_ATTRIB_TEX0;

		VB->AttribPtr[VERT_ATTRIB_TEX0 + tex_id] = VB->AttribPtr[VERT_ATTRIB_FOG];
		VB->AttribPtr[_TNL_ATTRIB_TEX0 + tex_id] = VB->AttribPtr[VERT_ATTRIB_FOG];
		RENDERINPUTS_SET(tnl->render_inputs_bitset, _TNL_ATTRIB_TEX0 + tex_id);
	}

	/**
	 *  Sending only one texcoord component may lead to lock up,
	 *  so for all textures always output 4 texcoord components to RS.
	 */
	{
		int i;
		GLuint swiz, format, hw_format;
		for (i = 0; i < ctx->Const.MaxTextureUnits; i++) {
			if (fp_reads & FRAG_BIT_TEX(i)) {
				switch (VB->AttribPtr[_TNL_ATTRIB_TEX0 + i]->size) {
					case 1:
						format = EMIT_1F;
						hw_format = R300_DATA_TYPE_FLOAT_1;
						swiz = MAKE_SWIZZLE4(SWIZZLE_X, SWIZZLE_ZERO, SWIZZLE_ZERO, SWIZZLE_ONE);
						break;
					case 2:
						format = EMIT_2F;
						hw_format = R300_DATA_TYPE_FLOAT_2;
						swiz = MAKE_SWIZZLE4(SWIZZLE_X, SWIZZLE_Y, SWIZZLE_ZERO, SWIZZLE_ONE);
						break;
					case 3:
						format = EMIT_3F;
						hw_format = R300_DATA_TYPE_FLOAT_3;
						swiz = MAKE_SWIZZLE4(SWIZZLE_X, SWIZZLE_Y, SWIZZLE_Z, SWIZZLE_ONE);
						break;
					case 4:
						format = EMIT_4F;
						hw_format = R300_DATA_TYPE_FLOAT_4;
						swiz = SWIZZLE_XYZW;
						break;
					default:
						continue;
				}
				InputsRead |= 1 << (VERT_ATTRIB_TEX0 + i);
				OutputsWritten |= 1 << (VERT_RESULT_TEX0 + i);
				EMIT_ATTR(_TNL_ATTRIB_TEX(i), format);
				ADD_ATTR(VERT_ATTRIB_TEX0 + i, hw_format, SWTCL_OVM_TEX(first_free_tex), swiz, MASK_XYZW, 0);
				++first_free_tex;
			}
		}
	}

	if (first_free_tex >= ctx->Const.MaxTextureUnits) {
		fprintf(stderr, "\tout of free texcoords to write fog coordinate\n");
		exit(-1);
	}

	R300_NEWPRIM(rmesa);
	rmesa->vbuf.num_attribs = num_attrs;
	*_InputsRead = InputsRead;
	*_OutputsWritten = OutputsWritten;

	RENDERINPUTS_COPY(rmesa->render_inputs_bitset, tnl->render_inputs_bitset);
}

static void r300PrepareVertices(struct gl_context *ctx)
{
	r300ContextPtr rmesa = R300_CONTEXT(ctx);
	GLuint InputsRead, OutputsWritten;
	radeon_print(RADEON_SWRENDER, RADEON_TRACE, "%s\n", __func__);

	r300ChooseSwtclVertexFormat(ctx, &InputsRead, &OutputsWritten);
	r300SetupVAP(ctx, InputsRead, OutputsWritten);

	rmesa->radeon.swtcl.vertex_size =
		_tnl_install_attrs( ctx,
				    rmesa->radeon.swtcl.vertex_attrs,
				    rmesa->radeon.swtcl.vertex_attr_count,
				    NULL, 0 );

	rmesa->radeon.swtcl.vertex_size /= 4;
}

static void r300_predict_emit_size( r300ContextPtr rmesa )
{
	if (!rmesa->radeon.swtcl.emit_prediction) {
		const int vertex_size = 7;
		const int prim_size = 3;
		const int cache_flush_size = 4;
		const int pre_emit_state = 4;
		const int scissor_size = 3;
		const int state_size = radeonCountStateEmitSize(&rmesa->radeon);

		if (rcommonEnsureCmdBufSpace(&rmesa->radeon,
					state_size + pre_emit_state + scissor_size
					+ vertex_size + prim_size + cache_flush_size * 2,
					__FUNCTION__))
			rmesa->radeon.swtcl.emit_prediction = radeonCountStateEmitSize(&rmesa->radeon);
		else
			rmesa->radeon.swtcl.emit_prediction = state_size;

		rmesa->radeon.swtcl.emit_prediction += rmesa->radeon.cmdbuf.cs->cdw
			+ vertex_size + scissor_size + prim_size + cache_flush_size * 2 + pre_emit_state;
		radeon_print(RADEON_SWRENDER, RADEON_VERBOSE,
				"%s, size %d\n",
				__func__, rmesa->radeon.cmdbuf.cs->cdw
				+ vertex_size + scissor_size + prim_size + cache_flush_size * 2 + pre_emit_state);
	}
}


static GLuint reduced_prim[] = {
	GL_POINTS,
	GL_LINES,
	GL_LINES,
	GL_LINES,
	GL_TRIANGLES,
	GL_TRIANGLES,
	GL_TRIANGLES,
	GL_TRIANGLES,
	GL_TRIANGLES,
	GL_TRIANGLES,
};

static void r300RasterPrimitive( struct gl_context *ctx, GLuint prim );

/***********************************************************************
 *                    Emit primitives as inline vertices               *
 ***********************************************************************/


#define HAVE_POINTS      1
#define HAVE_LINES       1
#define HAVE_LINE_STRIPS 1
#define HAVE_TRIANGLES   1
#define HAVE_TRI_STRIPS  1
#define HAVE_TRI_STRIP_1 0
#define HAVE_TRI_FANS    1
#define HAVE_QUADS       0
#define HAVE_QUAD_STRIPS 0
#define HAVE_POLYGONS    1
#define HAVE_ELTS        1

static void* r300_alloc_verts(r300ContextPtr rmesa, GLuint n, GLuint size)
{
	void *rv;
	do {
		r300_predict_emit_size( rmesa );
		rv = rcommonAllocDmaLowVerts( &rmesa->radeon, n, size * 4 );
	} while (!rv);
	return rv;
}

#undef LOCAL_VARS
#undef ALLOC_VERTS
#define CTX_ARG r300ContextPtr rmesa
#define GET_VERTEX_DWORDS() rmesa->radeon.swtcl.vertex_size
#define ALLOC_VERTS( n, size ) r300_alloc_verts(rmesa, n, size);
#define LOCAL_VARS						\
   r300ContextPtr rmesa = R300_CONTEXT(ctx);		\
   const char *r300verts = (char *)rmesa->radeon.swtcl.verts;
#define VERT(x) (r300Vertex *)(r300verts + ((x) * vertsize * sizeof(int)))
#define VERTEX r300Vertex
#undef TAG
#define TAG(x) r300_##x
#include "tnl_dd/t_dd_triemit.h"



/***********************************************************************
 *          Macros for t_dd_tritmp.h to draw basic primitives          *
 ***********************************************************************/

#define QUAD( a, b, c, d ) r300_quad( rmesa, a, b, c, d )
#define TRI( a, b, c )     r300_triangle( rmesa, a, b, c )
#define LINE( a, b )       r300_line( rmesa, a, b )
#define POINT( a )         r300_point( rmesa, a )

/***********************************************************************
 *              Build render functions from dd templates               *
 ***********************************************************************/

#define R300_UNFILLED_BIT	0x01
#define R300_MAX_TRIFUNC	0x02

static struct {
   tnl_points_func	        points;
   tnl_line_func		line;
   tnl_triangle_func	triangle;
   tnl_quad_func		quad;
} rast_tab[R300_MAX_TRIFUNC];

#define DO_FALLBACK  0
#define DO_UNFILLED (IND & R300_UNFILLED_BIT)
#define DO_TWOSIDE   0
#define DO_FLAT      0
#define DO_OFFSET    0
#define DO_TRI       1
#define DO_QUAD      1
#define DO_LINE      1
#define DO_POINTS    1
#define DO_FULL_QUAD 1

#define HAVE_SPEC   1
#define HAVE_BACK_COLORS  0
#define HAVE_HW_FLATSHADE 1
#define TAB rast_tab

#define DEPTH_SCALE 1.0
#define UNFILLED_TRI unfilled_tri
#define UNFILLED_QUAD unfilled_quad
#define VERT_X(_v) _v->v.x
#define VERT_Y(_v) _v->v.y
#define VERT_Z(_v) _v->v.z
#define AREA_IS_CCW( a ) (a < 0)
#define GET_VERTEX(e) (rmesa->radeon.swtcl.verts + (e*rmesa->radeon.swtcl.vertex_size*sizeof(int)))

#define VERT_SET_RGBA( v, c ) \
do { \
   r300_color_t *color = (r300_color_t *)&((v)->ui[coloroffset]); \
   UNCLAMPED_FLOAT_TO_UBYTE(color->red, (c)[0]); \
   UNCLAMPED_FLOAT_TO_UBYTE(color->green, (c)[1]); \
   UNCLAMPED_FLOAT_TO_UBYTE(color->blue, (c)[2]); \
   UNCLAMPED_FLOAT_TO_UBYTE(color->alpha, (c)[3]); \
} while (0)

#define VERT_COPY_RGBA( v0, v1 ) v0->ui[coloroffset] = v1->ui[coloroffset]

#define VERT_SET_SPEC( v0, c ) \
do { \
   if (specoffset) { \
   UNCLAMPED_FLOAT_TO_UBYTE(v0->v.specular.red, (c)[0]); \
   UNCLAMPED_FLOAT_TO_UBYTE(v0->v.specular.green, (c)[1]); \
   UNCLAMPED_FLOAT_TO_UBYTE(v0->v.specular.blue, (c)[2]); \
   } \
} while (0)

#define VERT_COPY_SPEC( v0, v1 ) \
do { \
   if (specoffset) { \
       v0->v.specular.red = v1->v.specular.red; \
       v0->v.specular.green = v1->v.specular.green; \
       v0->v.specular.blue = v1->v.specular.blue; \
   } \
} while (0)

#define VERT_SAVE_RGBA( idx )    color[idx] = v[idx]->ui[coloroffset]
#define VERT_RESTORE_RGBA( idx ) v[idx]->ui[coloroffset] = color[idx]
#define VERT_SAVE_SPEC( idx )    if (specoffset) spec[idx] = v[idx]->ui[specoffset]
#define VERT_RESTORE_SPEC( idx ) if (specoffset) v[idx]->ui[specoffset] = spec[idx]

#undef LOCAL_VARS
#undef TAG
#undef INIT

#define LOCAL_VARS(n)							\
   r300ContextPtr rmesa = R300_CONTEXT(ctx);			\
   GLuint color[n] = { 0, }, spec[n] = { 0, };				\
   GLuint coloroffset = rmesa->swtcl.coloroffset;	\
   GLuint specoffset = rmesa->swtcl.specoffset;			\
   (void) color; (void) spec; (void) coloroffset; (void) specoffset;

/***********************************************************************
 *                Helpers for rendering unfilled primitives            *
 ***********************************************************************/

#define RASTERIZE(x) r300RasterPrimitive( ctx, reduced_prim[x] )
#define RENDER_PRIMITIVE rmesa->radeon.swtcl.render_primitive
#undef TAG
#define TAG(x) x
#include "tnl_dd/t_dd_unfilled.h"
#undef IND


/***********************************************************************
 *                      Generate GL render functions                   *
 ***********************************************************************/


#define IND (0)
#define TAG(x) x
#include "tnl_dd/t_dd_tritmp.h"

#define IND (R300_UNFILLED_BIT)
#define TAG(x) x##_unfilled
#include "tnl_dd/t_dd_tritmp.h"


static void init_rast_tab( void )
{
   init();
   init_unfilled();
}

/**********************************************************************/
/*               Render unclipped begin/end objects                   */
/**********************************************************************/

#define RENDER_POINTS( start, count )		\
   for ( ; start < count ; start++)		\
      r300_point( rmesa, VERT(start) )
#define RENDER_LINE( v0, v1 ) \
   r300_line( rmesa, VERT(v0), VERT(v1) )
#define RENDER_TRI( v0, v1, v2 )  \
   r300_triangle( rmesa, VERT(v0), VERT(v1), VERT(v2) )
#define RENDER_QUAD( v0, v1, v2, v3 ) \
   r300_quad( rmesa, VERT(v0), VERT(v1), VERT(v2), VERT(v3) )
#define INIT(x) do {					\
   r300RenderPrimitive( ctx, x );			\
} while (0)
#undef LOCAL_VARS
#define LOCAL_VARS						\
   r300ContextPtr rmesa = R300_CONTEXT(ctx);		\
   const GLuint vertsize = rmesa->radeon.swtcl.vertex_size;		\
   const char *r300verts = (char *)rmesa->radeon.swtcl.verts;		\
   const GLuint * const elt = TNL_CONTEXT(ctx)->vb.Elts;	\
   const GLboolean stipple = ctx->Line.StippleFlag;		\
   (void) elt; (void) stipple;
#define RESET_STIPPLE	//if ( stipple ) r200ResetLineStipple( ctx );
#define RESET_OCCLUSION
#define PRESERVE_VB_DEFS
#define ELT(x) (x)
#define TAG(x) r300_##x##_verts
#include "tnl/t_vb_rendertmp.h"
#undef ELT
#undef TAG
#define TAG(x) r300_##x##_elts
#define ELT(x) elt[x]
#include "tnl/t_vb_rendertmp.h"




/**********************************************************************/
/*                    Choose render functions                         */
/**********************************************************************/
static void r300ChooseRenderState( struct gl_context *ctx )
{
	TNLcontext *tnl = TNL_CONTEXT(ctx);
	r300ContextPtr rmesa = R300_CONTEXT(ctx);
	GLuint index = 0;
	GLuint flags = ctx->_TriangleCaps;
	radeon_print(RADEON_SWRENDER, RADEON_VERBOSE, "%s\n", __func__);

	if (flags & DD_TRI_UNFILLED)      index |= R300_UNFILLED_BIT;

	if (index != rmesa->radeon.swtcl.RenderIndex) {
		tnl->Driver.Render.Points = rast_tab[index].points;
		tnl->Driver.Render.Line = rast_tab[index].line;
		tnl->Driver.Render.ClippedLine = rast_tab[index].line;
		tnl->Driver.Render.Triangle = rast_tab[index].triangle;
		tnl->Driver.Render.Quad = rast_tab[index].quad;

		if (index == 0) {
			tnl->Driver.Render.PrimTabVerts = r300_render_tab_verts;
			tnl->Driver.Render.PrimTabElts = r300_render_tab_elts;
			tnl->Driver.Render.ClippedPolygon = r300_fast_clipped_poly;
		} else {
			tnl->Driver.Render.PrimTabVerts = _tnl_render_tab_verts;
			tnl->Driver.Render.PrimTabElts = _tnl_render_tab_elts;
			tnl->Driver.Render.ClippedPolygon = _tnl_RenderClippedPolygon;
		}

		rmesa->radeon.swtcl.RenderIndex = index;
	}
}

void r300RenderStart(struct gl_context *ctx)
{
	radeon_print(RADEON_SWRENDER, RADEON_VERBOSE, "%s\n", __func__);
	r300ContextPtr rmesa = R300_CONTEXT( ctx );

	r300ChooseRenderState(ctx);

	r300UpdateShaders(rmesa);

	r300PrepareVertices(ctx);

	r300ValidateBuffers(ctx);

	r300UpdateShaderStates(rmesa);


	/* investigate if we can put back flush optimisation if needed */
	if (rmesa->radeon.dma.flush != NULL) {
		rmesa->radeon.dma.flush(ctx);
	}
}

void r300RenderFinish(struct gl_context *ctx)
{
}

static void r300RasterPrimitive( struct gl_context *ctx, GLuint hwprim )
{
	r300ContextPtr rmesa = R300_CONTEXT(ctx);
	radeon_print(RADEON_SWRENDER, RADEON_TRACE, "%s\n", __func__);

	if (rmesa->radeon.swtcl.hw_primitive != hwprim) {
		R300_NEWPRIM( rmesa );
		rmesa->radeon.swtcl.hw_primitive = hwprim;
	}
}

void r300RenderPrimitive(struct gl_context *ctx, GLenum prim)
{

	r300ContextPtr rmesa = R300_CONTEXT(ctx);
	rmesa->radeon.swtcl.render_primitive = prim;
	radeon_print(RADEON_SWRENDER, RADEON_TRACE, "%s\n", __func__);

	if ((prim == GL_TRIANGLES) && (ctx->_TriangleCaps & DD_TRI_UNFILLED))
		return;

	r300RasterPrimitive( ctx, reduced_prim[prim] );
}

void r300ResetLineStipple(struct gl_context *ctx)
{
	if (RADEON_DEBUG & RADEON_VERTS)
		fprintf(stderr, "%s\n", __func__);
}

void r300InitSwtcl(struct gl_context *ctx)
{
	TNLcontext *tnl = TNL_CONTEXT(ctx);
	r300ContextPtr rmesa = R300_CONTEXT(ctx);
	static int firsttime = 1;
	radeon_print(RADEON_SWRENDER, RADEON_NORMAL, "%s\n", __func__);

	if (firsttime) {
		init_rast_tab();
		firsttime = 0;
	}
	rmesa->radeon.swtcl.emit_prediction = 0;

	tnl->Driver.Render.Start = r300RenderStart;
	tnl->Driver.Render.Finish = r300RenderFinish;
	tnl->Driver.Render.PrimitiveNotify = r300RenderPrimitive;
	tnl->Driver.Render.ResetLineStipple = r300ResetLineStipple;
	tnl->Driver.Render.BuildVertices = _tnl_build_vertices;
	tnl->Driver.Render.CopyPV = _tnl_copy_pv;
	tnl->Driver.Render.Interp = _tnl_interp;

	/* FIXME: what are these numbers? */
	_tnl_init_vertices( ctx, ctx->Const.MaxArrayLockSize + 12,
			    48 * sizeof(GLfloat) );

	rmesa->radeon.swtcl.verts = (GLubyte *)tnl->clipspace.vertex_buf;
	rmesa->radeon.swtcl.RenderIndex = ~0;
	rmesa->radeon.swtcl.render_primitive = GL_TRIANGLES;
	rmesa->radeon.swtcl.hw_primitive = 0;

	_tnl_invalidate_vertex_state( ctx, ~0 );
	_tnl_invalidate_vertices( ctx, ~0 );

	_tnl_need_projected_coords( ctx, GL_FALSE );
}

void r300DestroySwtcl(struct gl_context *ctx)
{
}

static void r300EmitVertexAOS(r300ContextPtr rmesa, GLuint vertex_size, struct radeon_bo *bo, GLuint offset)
{
	BATCH_LOCALS(&rmesa->radeon);

	radeon_print(RADEON_SWRENDER, RADEON_TRACE,
		"%s:  vertex_size %d, offset 0x%x \n",
			__FUNCTION__, vertex_size, offset);

	BEGIN_BATCH(7);
	OUT_BATCH_PACKET3(R300_PACKET3_3D_LOAD_VBPNTR, 2);
	OUT_BATCH(1);
	OUT_BATCH(vertex_size | (vertex_size << 8));
	OUT_BATCH_RELOC(offset, bo, offset, RADEON_GEM_DOMAIN_GTT, 0, 0);
	END_BATCH();
}

static void r300EmitVbufPrim(r300ContextPtr rmesa, GLuint primitive, GLuint vertex_nr)
{
	BATCH_LOCALS(&rmesa->radeon);
	int type, num_verts;
	if (RADEON_DEBUG & RADEON_VERTS)
		fprintf(stderr, "%s\n", __func__);

	type = r300PrimitiveType(rmesa, primitive);
	num_verts = r300NumVerts(rmesa, vertex_nr, primitive);

	BEGIN_BATCH(3);
	OUT_BATCH_PACKET3(R300_PACKET3_3D_DRAW_VBUF_2, 0);
	OUT_BATCH(R300_VAP_VF_CNTL__PRIM_WALK_VERTEX_LIST | (num_verts << 16) | type);
	END_BATCH();
}

void r300_swtcl_flush(struct gl_context *ctx, uint32_t current_offset)
{
	radeon_print(RADEON_SWRENDER, RADEON_TRACE, "%s\n", __func__);
	r300ContextPtr rmesa = R300_CONTEXT(ctx);

	r300EmitCacheFlush(rmesa);

	radeonEmitState(&rmesa->radeon);
	r300_emit_scissor(ctx);
	r300EmitVertexAOS(rmesa,
			  rmesa->radeon.swtcl.vertex_size,
			  rmesa->radeon.swtcl.bo,
			  current_offset);

	r300EmitVbufPrim(rmesa,
		   rmesa->radeon.swtcl.hw_primitive,
		   rmesa->radeon.swtcl.numverts);
	r300EmitCacheFlush(rmesa);
	if ( rmesa->radeon.swtcl.emit_prediction < rmesa->radeon.cmdbuf.cs->cdw )
		WARN_ONCE("Rendering was %d commands larger than predicted size."
			" We might overflow  command buffer.\n",
			rmesa->radeon.cmdbuf.cs->cdw - rmesa->radeon.swtcl.emit_prediction );
	rmesa->radeon.swtcl.emit_prediction = 0;
	COMMIT_BATCH();
}