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
|
/*
* Copyright © 2009 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.
*
* Authors:
* Eric Anholt <eric@anholt.net>
*
*/
#include "brw_context.h"
#include "brw_state.h"
#include "brw_defines.h"
#include "intel_batchbuffer.h"
#include "main/shaderapi.h"
static void
gen6_upload_gs_push_constants(struct brw_context *brw)
{
struct brw_stage_state *stage_state = &brw->gs.base;
/* BRW_NEW_GEOMETRY_PROGRAM */
const struct brw_geometry_program *gp =
(struct brw_geometry_program *) brw->geometry_program;
if (gp) {
/* BRW_NEW_GS_PROG_DATA */
struct brw_stage_prog_data *prog_data = &brw->gs.prog_data->base.base;
_mesa_shader_write_subroutine_indices(&brw->ctx, MESA_SHADER_GEOMETRY);
gen6_upload_push_constants(brw, &gp->program.Base, prog_data,
stage_state, AUB_TRACE_VS_CONSTANTS);
}
if (brw->gen >= 7)
gen7_upload_constant_state(brw, stage_state, gp, _3DSTATE_CONSTANT_GS);
}
const struct brw_tracked_state gen6_gs_push_constants = {
.dirty = {
.mesa = _NEW_PROGRAM_CONSTANTS |
_NEW_TRANSFORM,
.brw = BRW_NEW_BATCH |
BRW_NEW_BLORP |
BRW_NEW_GEOMETRY_PROGRAM |
BRW_NEW_GS_PROG_DATA |
BRW_NEW_PUSH_CONSTANT_ALLOCATION,
},
.emit = gen6_upload_gs_push_constants,
};
static void
upload_gs_state_for_tf(struct brw_context *brw)
{
const struct gen_device_info *devinfo = &brw->screen->devinfo;
BEGIN_BATCH(7);
OUT_BATCH(_3DSTATE_GS << 16 | (7 - 2));
OUT_BATCH(brw->ff_gs.prog_offset);
OUT_BATCH(GEN6_GS_SPF_MODE | GEN6_GS_VECTOR_MASK_ENABLE);
OUT_BATCH(0); /* no scratch space */
OUT_BATCH((2 << GEN6_GS_DISPATCH_START_GRF_SHIFT) |
(brw->ff_gs.prog_data->urb_read_length << GEN6_GS_URB_READ_LENGTH_SHIFT));
OUT_BATCH(((devinfo->max_gs_threads - 1) << GEN6_GS_MAX_THREADS_SHIFT) |
GEN6_GS_STATISTICS_ENABLE |
GEN6_GS_SO_STATISTICS_ENABLE |
GEN6_GS_RENDERING_ENABLE);
OUT_BATCH(GEN6_GS_SVBI_PAYLOAD_ENABLE |
GEN6_GS_SVBI_POSTINCREMENT_ENABLE |
(brw->ff_gs.prog_data->svbi_postincrement_value <<
GEN6_GS_SVBI_POSTINCREMENT_VALUE_SHIFT) |
GEN6_GS_ENABLE);
ADVANCE_BATCH();
}
static void
upload_gs_state(struct brw_context *brw)
{
const struct gen_device_info *devinfo = &brw->screen->devinfo;
/* BRW_NEW_GEOMETRY_PROGRAM */
bool active = brw->geometry_program;
/* BRW_NEW_GS_PROG_DATA */
const struct brw_vue_prog_data *prog_data = &brw->gs.prog_data->base;
const struct brw_stage_state *stage_state = &brw->gs.base;
if (!active || stage_state->push_const_size == 0) {
/* Disable the push constant buffers. */
BEGIN_BATCH(5);
OUT_BATCH(_3DSTATE_CONSTANT_GS << 16 | (5 - 2));
OUT_BATCH(0);
OUT_BATCH(0);
OUT_BATCH(0);
OUT_BATCH(0);
ADVANCE_BATCH();
} else {
BEGIN_BATCH(5);
OUT_BATCH(_3DSTATE_CONSTANT_GS << 16 |
GEN6_CONSTANT_BUFFER_0_ENABLE |
(5 - 2));
/* Pointer to the GS constant buffer. Covered by the set of
* state flags from gen6_upload_vs_constants
*/
OUT_BATCH(stage_state->push_const_offset +
stage_state->push_const_size - 1);
OUT_BATCH(0);
OUT_BATCH(0);
OUT_BATCH(0);
ADVANCE_BATCH();
}
if (active) {
BEGIN_BATCH(7);
OUT_BATCH(_3DSTATE_GS << 16 | (7 - 2));
OUT_BATCH(stage_state->prog_offset);
/* GEN6_GS_SPF_MODE and GEN6_GS_VECTOR_MASK_ENABLE are enabled as it
* was previously done for gen6.
*
* TODO: test with both disabled to see if the HW is behaving
* as expected, like in gen7.
*/
OUT_BATCH(GEN6_GS_SPF_MODE | GEN6_GS_VECTOR_MASK_ENABLE |
((ALIGN(stage_state->sampler_count, 4)/4) <<
GEN6_GS_SAMPLER_COUNT_SHIFT) |
((prog_data->base.binding_table.size_bytes / 4) <<
GEN6_GS_BINDING_TABLE_ENTRY_COUNT_SHIFT));
if (prog_data->base.total_scratch) {
OUT_RELOC(stage_state->scratch_bo,
I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
ffs(stage_state->per_thread_scratch) - 11);
} else {
OUT_BATCH(0); /* no scratch space */
}
OUT_BATCH((prog_data->urb_read_length <<
GEN6_GS_URB_READ_LENGTH_SHIFT) |
(0 << GEN6_GS_URB_ENTRY_READ_OFFSET_SHIFT) |
(prog_data->base.dispatch_grf_start_reg <<
GEN6_GS_DISPATCH_START_GRF_SHIFT));
OUT_BATCH(((devinfo->max_gs_threads - 1) << GEN6_GS_MAX_THREADS_SHIFT) |
GEN6_GS_STATISTICS_ENABLE |
GEN6_GS_SO_STATISTICS_ENABLE |
GEN6_GS_RENDERING_ENABLE);
if (brw->gs.prog_data->gen6_xfb_enabled) {
/* GEN6_GS_REORDER is equivalent to GEN7_GS_REORDER_TRAILING
* in gen7. SNB and IVB specs are the same regarding the reordering of
* TRISTRIP/TRISTRIP_REV vertices and triangle orientation, so we do
* the same thing in both generations. For more details, see the
* comment in gen7_gs_state.c
*/
OUT_BATCH(GEN6_GS_REORDER |
GEN6_GS_SVBI_PAYLOAD_ENABLE |
GEN6_GS_ENABLE);
} else {
OUT_BATCH(GEN6_GS_REORDER | GEN6_GS_ENABLE);
}
ADVANCE_BATCH();
} else if (brw->ff_gs.prog_active) {
/* In gen6, transform feedback for the VS stage is done with an ad-hoc GS
* program. This function provides the needed 3DSTATE_GS for this.
*/
upload_gs_state_for_tf(brw);
} else {
/* No GS function required */
BEGIN_BATCH(7);
OUT_BATCH(_3DSTATE_GS << 16 | (7 - 2));
OUT_BATCH(0); /* prog_bo */
OUT_BATCH((0 << GEN6_GS_SAMPLER_COUNT_SHIFT) |
(0 << GEN6_GS_BINDING_TABLE_ENTRY_COUNT_SHIFT));
OUT_BATCH(0); /* scratch space base offset */
OUT_BATCH((1 << GEN6_GS_DISPATCH_START_GRF_SHIFT) |
(0 << GEN6_GS_URB_READ_LENGTH_SHIFT) |
(0 << GEN6_GS_URB_ENTRY_READ_OFFSET_SHIFT));
OUT_BATCH((0 << GEN6_GS_MAX_THREADS_SHIFT) |
GEN6_GS_STATISTICS_ENABLE |
GEN6_GS_RENDERING_ENABLE);
OUT_BATCH(0);
ADVANCE_BATCH();
}
brw->gs.enabled = active;
}
const struct brw_tracked_state gen6_gs_state = {
.dirty = {
.mesa = _NEW_PROGRAM_CONSTANTS |
_NEW_TRANSFORM,
.brw = BRW_NEW_BATCH |
BRW_NEW_BLORP |
BRW_NEW_CONTEXT |
BRW_NEW_FF_GS_PROG_DATA |
BRW_NEW_GEOMETRY_PROGRAM |
BRW_NEW_GS_PROG_DATA |
BRW_NEW_PUSH_CONSTANT_ALLOCATION,
},
.emit = upload_gs_state,
};
|
