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
|
/*
* Mesa 3-D graphics library
* Version: 6.5
*
* Copyright (C) 1999-2005 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.
*/
/* Vertices are just an array of floats, with all the attributes
* packed. We currently assume a layout like:
*
* attr[0][0..3] - window position
* attr[1..n][0..3] - remaining attributes.
*
* Attributes are assumed to be 4 floats wide but are packed so that
* all the enabled attributes run contiguously.
*/
#include "glheader.h"
#include "imports.h"
#include "sp_context.h"
#include "sp_headers.h"
#include "sp_quad.h"
#include "core/tgsi_core.h"
struct exec_machine {
const struct setup_coefficient *coef; /**< will point to quad->coef */
GLfloat attr[FRAG_ATTRIB_MAX][4][QUAD_SIZE];
};
/**
* Compute quad's attributes values, as constants (GL_FLAT shading).
*/
static INLINE void cinterp( struct exec_machine *exec,
GLuint attrib,
GLuint i )
{
GLuint j;
for (j = 0; j < QUAD_SIZE; j++) {
exec->attr[attrib][i][j] = exec->coef[attrib].a0[i];
}
}
/**
* Compute quad's attribute values by linear interpolation.
*
* Push into the fp:
*
* INPUT[attr] = MAD COEF_A0[attr], COEF_DADX[attr], INPUT_WPOS.xxxx
* INPUT[attr] = MAD INPUT[attr], COEF_DADY[attr], INPUT_WPOS.yyyy
*/
static INLINE void linterp( struct exec_machine *exec,
GLuint attrib,
GLuint i )
{
GLuint j;
for (j = 0; j < QUAD_SIZE; j++) {
const GLfloat x = exec->attr[FRAG_ATTRIB_WPOS][0][j];
const GLfloat y = exec->attr[FRAG_ATTRIB_WPOS][1][j];
exec->attr[attrib][i][j] = (exec->coef[attrib].a0[i] +
exec->coef[attrib].dadx[i] * x +
exec->coef[attrib].dady[i] * y);
}
}
/**
* Compute quad's attribute values by linear interpolation with
* perspective correction.
*
* Push into the fp:
*
* INPUT[attr] = MAD COEF_A0[attr], COEF_DADX[attr], INPUT_WPOS.xxxx
* INPUT[attr] = MAD INPUT[attr], COEF_DADY[attr], INPUT_WPOS.yyyy
* INPUT[attr] = MUL INPUT[attr], INPUT_WPOS.wwww
*
* (Or should that be 1/w ???)
*/
static INLINE void pinterp( struct exec_machine *exec,
GLuint attrib,
GLuint i )
{
GLuint j;
for (j = 0; j < QUAD_SIZE; j++) {
const GLfloat x = exec->attr[FRAG_ATTRIB_WPOS][0][j];
const GLfloat y = exec->attr[FRAG_ATTRIB_WPOS][1][j];
const GLfloat invW = exec->attr[FRAG_ATTRIB_WPOS][3][j];
exec->attr[attrib][i][j] = ((exec->coef[attrib].a0[i] +
exec->coef[attrib].dadx[i] * x +
exec->coef[attrib].dady[i] * y) * invW);
}
}
/* This should be done by the fragment shader execution unit (code
* generated from the decl instructions). Do it here for now.
*/
static void
shade_quad( struct quad_stage *qs, struct quad_header *quad )
{
const struct softpipe_context *softpipe = qs->softpipe;
struct exec_machine exec;
const GLfloat fx = quad->x0;
const GLfloat fy = quad->y0;
GLuint i, j;
exec.coef = quad->coef;
/* Position:
*/
exec.attr[FRAG_ATTRIB_WPOS][0][0] = fx;
exec.attr[FRAG_ATTRIB_WPOS][0][1] = fx + 1.0;
exec.attr[FRAG_ATTRIB_WPOS][0][2] = fx;
exec.attr[FRAG_ATTRIB_WPOS][0][3] = fx + 1.0;
exec.attr[FRAG_ATTRIB_WPOS][1][0] = fy;
exec.attr[FRAG_ATTRIB_WPOS][1][1] = fy;
exec.attr[FRAG_ATTRIB_WPOS][1][2] = fy + 1.0;
exec.attr[FRAG_ATTRIB_WPOS][1][3] = fy + 1.0;
/* Z and W are done by linear interpolation */
if (softpipe->need_z) {
linterp(&exec, 0, 2); /* attr[0].z */
}
if (softpipe->need_w) {
linterp(&exec, 0, 3); /* attr[0].w */
/*invert(&exec, 0, 3);*/
}
/* Interpolate all the remaining attributes. This will get pushed
* into the fragment program's responsibilities at some point.
* Start at 1 to skip fragment position attribute (computed above).
*/
for (i = 1; i < quad->nr_attrs; i++) {
switch (softpipe->interp[i]) {
case INTERP_CONSTANT:
for (j = 0; j < NUM_CHANNELS; j++)
cinterp(&exec, i, j);
break;
case INTERP_LINEAR:
for (j = 0; j < NUM_CHANNELS; j++)
linterp(&exec, i, j);
break;
case INTERP_PERSPECTIVE:
for (j = 0; j < NUM_CHANNELS; j++)
pinterp(&exec, i, j);
break;
}
}
#if 1
/*softpipe->run_fs( tri->fp, quad, &tri->outputs );*/
{
struct tgsi_exec_machine machine;
struct tgsi_exec_vector inputs[FRAG_ATTRIB_MAX + 1];
struct tgsi_exec_vector outputs[FRAG_ATTRIB_MAX + 1];
struct tgsi_exec_vector *ainputs;
struct tgsi_exec_vector *aoutputs;
GLuint i /*, total*/;
#ifdef DEBUG
memset(&machine, 0, sizeof(machine));
#endif
ainputs = (struct tgsi_exec_vector *) tgsi_align_128bit( inputs );
aoutputs = (struct tgsi_exec_vector *) tgsi_align_128bit( outputs );
#if 0
for( i = total = 0; i < PIPE_ATTRIB_MAX; i++ ) {
GLuint attr;
attr = softpipe->fp_attr_to_slot[i];
if( attr || total == 0) {
assert( total < FRAG_ATTRIB_MAX );
assert( attr < FRAG_ATTRIB_MAX );
assert( sizeof( ainputs[0] ) == sizeof( exec.attr[0] ) );
memcpy(
&ainputs[total],
exec.attr[attr],
sizeof( ainputs[0] ) );
total++;
}
}
#else
/* load input registers */
/* XXX simpler than above, but might not be right... */
for (i = 0; i < softpipe->nr_attrs; i++) {
memcpy(
&ainputs[i],
exec.attr[i],
sizeof( ainputs[0] ) );
}
#endif
/* init machine state */
tgsi_exec_machine_init(
&machine,
softpipe->fs.tokens );
machine.Inputs = ainputs;
machine.Outputs = aoutputs;
machine.Consts = softpipe->fs.constants->constant; /* XXX alignment? */
/* run shader */
tgsi_exec_machine_run( &machine );
/* store result color */
memcpy(quad->outputs.color,
&aoutputs[FRAG_ATTRIB_COL0].xyzw[0].f[0],
sizeof(quad->outputs.color));
if (softpipe->need_z) {
/* XXX temporary */
quad->outputs.depth[0] = exec.attr[0][2][0];
quad->outputs.depth[1] = exec.attr[0][2][1];
quad->outputs.depth[2] = exec.attr[0][2][2];
quad->outputs.depth[3] = exec.attr[0][2][3];
}
}
#else
{
GLuint attr = softpipe->fp_attr_to_slot[FRAG_ATTRIB_COL0];
assert(attr);
memcpy(quad->outputs.color,
exec.attr[attr],
sizeof(quad->outputs.color));
if (softpipe->need_z) {
quad->outputs.depth[0] = exec.attr[0][2][0];
quad->outputs.depth[1] = exec.attr[0][2][1];
quad->outputs.depth[2] = exec.attr[0][2][2];
quad->outputs.depth[3] = exec.attr[0][2][3];
}
}
#endif
/* shader may cull fragments */
if (quad->mask)
qs->next->run(qs->next, quad);
}
struct quad_stage *sp_quad_shade_stage( struct softpipe_context *softpipe )
{
struct quad_stage *stage = CALLOC_STRUCT(quad_stage);
stage->softpipe = softpipe;
stage->run = shade_quad;
return stage;
}
|
