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
|
/*
Copyright (C) Intel Corp. 2006. All Rights Reserved.
Intel funded Tungsten Graphics (http://www.tungstengraphics.com) to
develop this 3D driver.
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 (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 NONINFRINGEMENT.
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:
* Keith Whitwell <keith@tungstengraphics.com>
*/
#include "brw_context.h"
#include "brw_wm.h"
/***********************************************************************
*/
static struct brw_wm_ref *get_ref( struct brw_wm_compile *c )
{
assert(c->nr_refs < BRW_WM_MAX_REF);
return &c->refs[c->nr_refs++];
}
static struct brw_wm_value *get_value( struct brw_wm_compile *c)
{
assert(c->nr_refs < BRW_WM_MAX_VREG);
return &c->vreg[c->nr_vreg++];
}
/** return pointer to a newly allocated instruction */
static struct brw_wm_instruction *get_instruction( struct brw_wm_compile *c )
{
assert(c->nr_insns < BRW_WM_MAX_INSN);
return &c->instruction[c->nr_insns++];
}
/***********************************************************************
*/
/** Init the "undef" register */
static void pass0_init_undef( struct brw_wm_compile *c)
{
struct brw_wm_ref *ref = &c->undef_ref;
ref->value = &c->undef_value;
ref->hw_reg = brw_vec8_grf(0, 0);
ref->insn = 0;
ref->prevuse = NULL;
}
/** Set a FP register to a value */
static void pass0_set_fpreg_value( struct brw_wm_compile *c,
GLuint file,
GLuint idx,
GLuint component,
struct brw_wm_value *value )
{
struct brw_wm_ref *ref = get_ref(c);
ref->value = value;
ref->hw_reg = brw_vec8_grf(0, 0);
ref->insn = 0;
ref->prevuse = NULL;
c->pass0_fp_reg[file][idx][component] = ref;
}
/** Set a FP register to a ref */
static void pass0_set_fpreg_ref( struct brw_wm_compile *c,
GLuint file,
GLuint idx,
GLuint component,
const struct brw_wm_ref *src_ref )
{
c->pass0_fp_reg[file][idx][component] = src_ref;
}
static const struct brw_wm_ref *get_param_ref( struct brw_wm_compile *c,
const GLfloat *param_ptr )
{
GLuint i = c->prog_data.nr_params++;
if (i >= BRW_WM_MAX_PARAM) {
_mesa_printf("%s: out of params\n", __FUNCTION__);
c->prog_data.error = 1;
return NULL;
}
else {
struct brw_wm_ref *ref = get_ref(c);
c->prog_data.param[i] = param_ptr;
c->nr_creg = (i+16)/16;
/* Push the offsets into hw_reg. These will be added to the
* real register numbers once one is allocated in pass2.
*/
ref->hw_reg = brw_vec1_grf((i&8)?1:0, i%8);
ref->value = &c->creg[i/16];
ref->insn = 0;
ref->prevuse = NULL;
return ref;
}
}
/** Return a ref to an immediate value */
static const struct brw_wm_ref *get_imm_ref( struct brw_wm_compile *c,
const GLfloat *imm1f )
{
GLuint i;
/* Search for an existing const value matching the request:
*/
for (i = 0; i < c->nr_imm_refs; i++) {
if (c->imm_ref[i].imm_val == *imm1f)
return c->imm_ref[i].ref;
}
/* Else try to add a new one:
*/
if (c->nr_imm_refs < BRW_WM_MAX_IMM) {
GLuint i = c->nr_imm_refs++;
/* An immediate is a special type of parameter:
*/
c->imm_ref[i].imm_val = *imm_val;
c->imm_ref[i].ref = get_param_ref(c, imm_val);
return c->imm_ref[i].ref;
}
else {
_mesa_printf("%s: out of imm_refs\n", __FUNCTION__);
c->prog_data.error = 1;
return NULL;
}
}
/* Lookup our internal registers
*/
static const struct brw_wm_ref *pass0_get_reg( struct brw_wm_compile *c,
GLuint file,
GLuint idx,
GLuint component )
{
const struct brw_wm_ref *ref = c->pass0_fp_reg[file][idx][component];
if (!ref) {
switch (file) {
case PROGRAM_INPUT:
case PROGRAM_PAYLOAD:
case PROGRAM_TEMPORARY:
case PROGRAM_OUTPUT:
case PROGRAM_VARYING:
break;
case PROGRAM_LOCAL_PARAM:
ref = get_param_ref(c, &c->fp->program.Base.LocalParams[idx][component]);
break;
case PROGRAM_ENV_PARAM:
ref = get_param_ref(c, &c->env_param[idx][component]);
break;
case PROGRAM_STATE_VAR:
case PROGRAM_UNIFORM:
case PROGRAM_CONSTANT:
case PROGRAM_NAMED_PARAM: {
struct gl_program_parameter_list *plist = c->fp->program.Base.Parameters;
/* There's something really hokey about parameters parsed in
* arb programs - they all end up in here, whether they be
* state values, parameters or constants. This duplicates the
* structure above & also seems to subvert the limits set for
* each type of constant/param.
*/
switch (plist->Parameters[idx].Type) {
case PROGRAM_NAMED_PARAM:
case PROGRAM_CONSTANT:
/* These are invarient:
*/
ref = get_imm_ref(c, &plist->ParameterValues[idx][component]);
break;
case PROGRAM_STATE_VAR:
case PROGRAM_UNIFORM:
/* These may change from run to run:
*/
ref = get_param_ref(c, &plist->ParameterValues[idx][component] );
break;
default:
assert(0);
break;
}
break;
}
default:
assert(0);
break;
}
c->pass0_fp_reg[file][idx][component] = ref;
}
if (!ref)
ref = &c->undef_ref;
return ref;
}
/***********************************************************************
* Straight translation to internal instruction format
*/
static void pass0_set_dst( struct brw_wm_compile *c,
struct brw_wm_instruction *out,
const struct prog_instruction *inst,
GLuint writemask )
{
const struct prog_dst_register *dst = &inst->DstReg;
GLuint i;
for (i = 0; i < 4; i++) {
if (writemask & (1<<i)) {
out->dst[i] = get_value(c);
pass0_set_fpreg_value(c, dst->File, dst->Index, i, out->dst[i]);
}
}
out->writemask = writemask;
}
static const struct brw_wm_ref *get_fp_src_reg_ref( struct brw_wm_compile *c,
struct prog_src_register src,
GLuint i )
{
GLuint component = GET_SWZ(src.Swizzle,i);
const struct brw_wm_ref *src_ref;
static const GLfloat const_zero = 0.0;
static const GLfloat const_one = 1.0;
if (component == SWIZZLE_ZERO)
src_ref = get_imm_ref(c, &const_zero);
else if (component == SWIZZLE_ONE)
src_ref = get_imm_ref(c, &const_one);
else
src_ref = pass0_get_reg(c, src.File, src.Index, component);
return src_ref;
}
static struct brw_wm_ref *get_new_ref( struct brw_wm_compile *c,
struct prog_src_register src,
GLuint i,
struct brw_wm_instruction *insn)
{
const struct brw_wm_ref *ref = get_fp_src_reg_ref(c, src, i);
struct brw_wm_ref *newref = get_ref(c);
newref->value = ref->value;
newref->hw_reg = ref->hw_reg;
if (insn) {
newref->insn = insn - c->instruction;
newref->prevuse = newref->value->lastuse;
newref->value->lastuse = newref;
}
if (src.Negate & (1 << i))
newref->hw_reg.negate ^= 1;
if (src.Abs) {
newref->hw_reg.negate = 0;
newref->hw_reg.abs = 1;
}
return newref;
}
static void
translate_insn(struct brw_wm_compile *c,
const struct prog_instruction *inst)
{
struct brw_wm_instruction *out = get_instruction(c);
GLuint writemask = inst->DstReg.WriteMask;
GLuint nr_args = brw_wm_nr_args(inst->Opcode);
GLuint i, j;
/* Copy some data out of the instruction
*/
out->opcode = inst->Opcode;
out->saturate = (inst->SaturateMode != SATURATE_OFF);
out->tex_unit = inst->TexSrcUnit;
out->tex_idx = inst->TexSrcTarget;
out->tex_shadow = inst->TexShadow;
out->eot = inst->Aux & 1;
out->target = inst->Aux >> 1;
/* Args:
*/
for (i = 0; i < nr_args; i++) {
for (j = 0; j < 4; j++) {
out->src[i][j] = get_new_ref(c, inst->SrcReg[i], j, out);
}
}
/* Dst:
*/
pass0_set_dst(c, out, inst, writemask);
}
/***********************************************************************
* Optimize moves and swizzles away:
*/
static void pass0_precalc_mov( struct brw_wm_compile *c,
const struct prog_instruction *inst )
{
const struct prog_dst_register *dst = &inst->DstReg;
GLuint writemask = inst->DstReg.WriteMask;
struct brw_wm_ref *refs[4];
GLuint i;
/* Get the effect of a MOV by manipulating our register table:
* First get all refs, then assign refs. This ensures that "in-place"
* swizzles such as:
* MOV t, t.xxyx
* are handled correctly. Previously, these two steps were done in
* one loop and the above case was incorrectly handled.
*/
for (i = 0; i < 4; i++) {
refs[i] = get_new_ref(c, inst->SrcReg[0], i, NULL);
}
for (i = 0; i < 4; i++) {
if (writemask & (1 << i)) {
pass0_set_fpreg_ref( c, dst->File, dst->Index, i, refs[i]);
}
}
}
/* Initialize payload "registers".
*/
static void pass0_init_payload( struct brw_wm_compile *c )
{
GLuint i;
for (i = 0; i < 4; i++) {
GLuint j = i >= c->key.nr_depth_regs ? 0 : i;
pass0_set_fpreg_value( c, PROGRAM_PAYLOAD, PAYLOAD_DEPTH, i,
&c->payload.depth[j] );
}
#if 0
/* This seems to be an alternative to the INTERP_WPOS stuff I do
* elsewhere:
*/
if (c->key.source_depth_reg)
pass0_set_fpreg_value(c, PROGRAM_INPUT, FRAG_ATTRIB_WPOS, 2,
&c->payload.depth[c->key.source_depth_reg/2]);
#endif
for (i = 0; i < FRAG_ATTRIB_MAX; i++)
pass0_set_fpreg_value( c, PROGRAM_PAYLOAD, i, 0,
&c->payload.input_interp[i] );
}
/***********************************************************************
* PASS 0
*
* Work forwards to give each calculated value a unique number. Where
* an instruction produces duplicate values (eg DP3), all are given
* the same number.
*
* Translate away swizzling and eliminate non-saturating moves.
*/
void brw_wm_pass0( struct brw_wm_compile *c )
{
GLuint insn;
c->nr_vreg = 0;
c->nr_insns = 0;
pass0_init_undef(c);
pass0_init_payload(c);
for (insn = 0; insn < c->nr_fp_insns; insn++) {
const struct prog_instruction *inst = &c->prog_instructions[insn];
/* Optimize away moves, otherwise emit translated instruction:
*/
switch (inst->Opcode) {
case OPCODE_MOV:
if (!inst->SaturateMode) {
pass0_precalc_mov(c, inst);
}
else {
translate_insn(c, inst);
}
break;
default:
translate_insn(c, inst);
break;
}
}
if (INTEL_DEBUG & DEBUG_WM) {
brw_wm_print_program(c, "pass0");
}
}
|