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
|
/*
* Copyright © 2010 Intel Corporation
*
* 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 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.
*/
extern "C" {
#include "main/macros.h"
#include "brw_context.h"
#include "brw_vs.h"
}
#include "brw_fs.h"
#include "glsl/ir_optimization.h"
#include "glsl/ir_print_visitor.h"
struct gl_shader *
brw_new_shader(struct gl_context *ctx, GLuint name, GLuint type)
{
struct brw_shader *shader;
shader = rzalloc(NULL, struct brw_shader);
if (shader) {
shader->base.Type = type;
shader->base.Name = name;
_mesa_init_shader(ctx, &shader->base);
}
return &shader->base;
}
struct gl_shader_program *
brw_new_shader_program(struct gl_context *ctx, GLuint name)
{
struct brw_shader_program *prog;
prog = rzalloc(NULL, struct brw_shader_program);
if (prog) {
prog->base.Name = name;
_mesa_init_shader_program(ctx, &prog->base);
}
return &prog->base;
}
/**
* Performs a compile of the shader stages even when we don't know
* what non-orthogonal state will be set, in the hope that it reflects
* the eventual NOS used, and thus allows us to produce link failures.
*/
bool
brw_shader_precompile(struct gl_context *ctx, struct gl_shader_program *prog)
{
if (!brw_fs_precompile(ctx, prog))
return false;
if (!brw_vs_precompile(ctx, prog))
return false;
return true;
}
GLboolean
brw_link_shader(struct gl_context *ctx, struct gl_shader_program *prog)
{
struct brw_context *brw = brw_context(ctx);
struct intel_context *intel = &brw->intel;
unsigned int stage;
for (stage = 0; stage < ARRAY_SIZE(prog->_LinkedShaders); stage++) {
struct brw_shader *shader =
(struct brw_shader *)prog->_LinkedShaders[stage];
if (!shader)
continue;
void *mem_ctx = ralloc_context(NULL);
bool progress;
if (shader->ir)
ralloc_free(shader->ir);
shader->ir = new(shader) exec_list;
clone_ir_list(mem_ctx, shader->ir, shader->base.ir);
do_mat_op_to_vec(shader->ir);
lower_instructions(shader->ir,
MOD_TO_FRACT |
DIV_TO_MUL_RCP |
INT_DIV_TO_MUL_RCP |
SUB_TO_ADD_NEG |
EXP_TO_EXP2 |
LOG_TO_LOG2);
/* Pre-gen6 HW can only nest if-statements 16 deep. Beyond this,
* if-statements need to be flattened.
*/
if (intel->gen < 6)
lower_if_to_cond_assign(shader->ir, 16);
do_lower_texture_projection(shader->ir);
do_vec_index_to_cond_assign(shader->ir);
brw_do_cubemap_normalize(shader->ir);
lower_noise(shader->ir);
lower_quadop_vector(shader->ir, false);
bool input = true;
bool output = stage == MESA_SHADER_FRAGMENT;
bool temp = stage == MESA_SHADER_FRAGMENT;
bool uniform = stage == MESA_SHADER_FRAGMENT;
lower_variable_index_to_cond_assign(shader->ir,
input, output, temp, uniform);
do {
progress = false;
if (stage == MESA_SHADER_FRAGMENT) {
brw_do_channel_expressions(shader->ir);
brw_do_vector_splitting(shader->ir);
}
progress = do_lower_jumps(shader->ir, true, true,
true, /* main return */
false, /* continue */
false /* loops */
) || progress;
progress = do_common_optimization(shader->ir, true, 32) || progress;
} while (progress);
validate_ir_tree(shader->ir);
reparent_ir(shader->ir, shader->ir);
ralloc_free(mem_ctx);
}
if (!_mesa_ir_link_shader(ctx, prog))
return GL_FALSE;
if (!brw_shader_precompile(ctx, prog))
return GL_FALSE;
return GL_TRUE;
}
int
brw_type_for_base_type(const struct glsl_type *type)
{
switch (type->base_type) {
case GLSL_TYPE_FLOAT:
return BRW_REGISTER_TYPE_F;
case GLSL_TYPE_INT:
case GLSL_TYPE_BOOL:
return BRW_REGISTER_TYPE_D;
case GLSL_TYPE_UINT:
return BRW_REGISTER_TYPE_UD;
case GLSL_TYPE_ARRAY:
case GLSL_TYPE_STRUCT:
case GLSL_TYPE_SAMPLER:
/* These should be overridden with the type of the member when
* dereferenced into. BRW_REGISTER_TYPE_UD seems like a likely
* way to trip up if we don't.
*/
return BRW_REGISTER_TYPE_UD;
default:
assert(!"not reached");
return BRW_REGISTER_TYPE_F;
}
}
uint32_t
brw_conditional_for_comparison(unsigned int op)
{
switch (op) {
case ir_binop_less:
return BRW_CONDITIONAL_L;
case ir_binop_greater:
return BRW_CONDITIONAL_G;
case ir_binop_lequal:
return BRW_CONDITIONAL_LE;
case ir_binop_gequal:
return BRW_CONDITIONAL_GE;
case ir_binop_equal:
case ir_binop_all_equal: /* same as equal for scalars */
return BRW_CONDITIONAL_Z;
case ir_binop_nequal:
case ir_binop_any_nequal: /* same as nequal for scalars */
return BRW_CONDITIONAL_NZ;
default:
assert(!"not reached: bad operation for comparison");
return BRW_CONDITIONAL_NZ;
}
}
uint32_t
brw_math_function(enum opcode op)
{
switch (op) {
case SHADER_OPCODE_RCP:
return BRW_MATH_FUNCTION_INV;
case SHADER_OPCODE_RSQ:
return BRW_MATH_FUNCTION_RSQ;
case SHADER_OPCODE_SQRT:
return BRW_MATH_FUNCTION_SQRT;
case SHADER_OPCODE_EXP2:
return BRW_MATH_FUNCTION_EXP;
case SHADER_OPCODE_LOG2:
return BRW_MATH_FUNCTION_LOG;
case SHADER_OPCODE_POW:
return BRW_MATH_FUNCTION_POW;
case SHADER_OPCODE_SIN:
return BRW_MATH_FUNCTION_SIN;
case SHADER_OPCODE_COS:
return BRW_MATH_FUNCTION_COS;
default:
assert(!"not reached: unknown math function");
return 0;
}
}
|
