63612e7544000c72239eeb20d01d71b79f4c0049
[mesa.git] / src / mesa / drivers / dri / i965 / gen8_ps_state.c
1 /*
2 * Copyright © 2012 Intel Corporation
3 *
4 * Permission is hereby granted, free of charge, to any person obtaining a
5 * copy of this software and associated documentation files (the "Software"),
6 * to deal in the Software without restriction, including without limitation
7 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
8 * and/or sell copies of the Software, and to permit persons to whom the
9 * Software is furnished to do so, subject to the following conditions:
10 *
11 * The above copyright notice and this permission notice (including the next
12 * paragraph) shall be included in all copies or substantial portions of the
13 * Software.
14 *
15 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
18 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
20 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
21 * IN THE SOFTWARE.
22 */
23
24 #include <stdbool.h>
25 #include "brw_state.h"
26 #include "brw_defines.h"
27 #include "intel_batchbuffer.h"
28
29 static void
30 upload_ps_extra(struct brw_context *brw)
31 {
32 /* BRW_NEW_FRAGMENT_PROGRAM */
33 const struct brw_fragment_program *fp =
34 brw_fragment_program_const(brw->fragment_program);
35 uint32_t dw1 = 0;
36
37 dw1 |= GEN8_PSX_PIXEL_SHADER_VALID;
38
39 if (fp->program.UsesKill)
40 dw1 |= GEN8_PSX_KILL_ENABLE;
41
42 /* BRW_NEW_FRAGMENT_PROGRAM */
43 if (brw->wm.prog_data->num_varying_inputs != 0)
44 dw1 |= GEN8_PSX_ATTRIBUTE_ENABLE;
45
46 if (fp->program.Base.OutputsWritten & BITFIELD64_BIT(FRAG_RESULT_DEPTH)) {
47 switch (fp->program.FragDepthLayout) {
48 case FRAG_DEPTH_LAYOUT_NONE:
49 case FRAG_DEPTH_LAYOUT_ANY:
50 dw1 |= GEN8_PSX_PSCDEPTH_ON;
51 break;
52 case FRAG_DEPTH_LAYOUT_GREATER:
53 dw1 |= GEN8_PSX_PSCDEPTH_ON_GE;
54 break;
55 case FRAG_DEPTH_LAYOUT_LESS:
56 dw1 |= GEN8_PSX_PSCDEPTH_ON_LE;
57 break;
58 case FRAG_DEPTH_LAYOUT_UNCHANGED:
59 break;
60 }
61 }
62
63 if (fp->program.Base.InputsRead & VARYING_BIT_POS)
64 dw1 |= GEN8_PSX_USES_SOURCE_DEPTH | GEN8_PSX_USES_SOURCE_W;
65
66 BEGIN_BATCH(2);
67 OUT_BATCH(_3DSTATE_PS_EXTRA << 16 | (2 - 2));
68 OUT_BATCH(dw1);
69 ADVANCE_BATCH();
70 }
71
72 const struct brw_tracked_state gen8_ps_extra = {
73 .dirty = {
74 .mesa = 0,
75 .brw = BRW_NEW_CONTEXT | BRW_NEW_FRAGMENT_PROGRAM,
76 .cache = 0,
77 },
78 .emit = upload_ps_extra,
79 };
80
81 static void
82 upload_wm_state(struct brw_context *brw)
83 {
84 struct gl_context *ctx = &brw->ctx;
85 uint32_t dw1 = 0;
86
87 dw1 |= GEN7_WM_STATISTICS_ENABLE;
88 dw1 |= GEN7_WM_LINE_AA_WIDTH_1_0;
89 dw1 |= GEN7_WM_LINE_END_CAP_AA_WIDTH_0_5;
90 dw1 |= GEN7_WM_POINT_RASTRULE_UPPER_RIGHT;
91
92 /* _NEW_LINE */
93 if (ctx->Line.StippleFlag)
94 dw1 |= GEN7_WM_LINE_STIPPLE_ENABLE;
95
96 /* _NEW_POLYGON */
97 if (ctx->Polygon.StippleFlag)
98 dw1 |= GEN7_WM_POLYGON_STIPPLE_ENABLE;
99
100 /* CACHE_NEW_WM_PROG */
101 dw1 |= brw->wm.prog_data->barycentric_interp_modes <<
102 GEN7_WM_BARYCENTRIC_INTERPOLATION_MODE_SHIFT;
103
104 BEGIN_BATCH(2);
105 OUT_BATCH(_3DSTATE_WM << 16 | (2 - 2));
106 OUT_BATCH(dw1);
107 ADVANCE_BATCH();
108 }
109
110 const struct brw_tracked_state gen8_wm_state = {
111 .dirty = {
112 .mesa = _NEW_LINE | _NEW_POLYGON,
113 .brw = BRW_NEW_CONTEXT,
114 .cache = CACHE_NEW_WM_PROG,
115 },
116 .emit = upload_wm_state,
117 };
118
119 static void
120 upload_ps_state(struct brw_context *brw)
121 {
122 struct gl_context *ctx = &brw->ctx;
123 uint32_t dw3 = 0, dw6 = 0, dw7 = 0;
124
125 /* BRW_NEW_PS_BINDING_TABLE */
126 BEGIN_BATCH(2);
127 OUT_BATCH(_3DSTATE_BINDING_TABLE_POINTERS_PS << 16 | (2 - 2));
128 OUT_BATCH(brw->wm.base.bind_bo_offset);
129 ADVANCE_BATCH();
130
131 /* CACHE_NEW_SAMPLER */
132 BEGIN_BATCH(2);
133 OUT_BATCH(_3DSTATE_SAMPLER_STATE_POINTERS_PS << 16 | (2 - 2));
134 OUT_BATCH(brw->wm.base.sampler_offset);
135 ADVANCE_BATCH();
136
137 /* CACHE_NEW_WM_PROG */
138 gen8_upload_constant_state(brw, &brw->wm.base, true, _3DSTATE_CONSTANT_PS);
139
140 /* Initialize the execution mask with VMask. Otherwise, derivatives are
141 * incorrect for subspans where some of the pixels are unlit. We believe
142 * the bit just didn't take effect in previous generations.
143 */
144 dw3 |= GEN7_PS_VECTOR_MASK_ENABLE;
145
146 /* CACHE_NEW_SAMPLER */
147 dw3 |=
148 (ALIGN(brw->wm.base.sampler_count, 4) / 4) << GEN7_PS_SAMPLER_COUNT_SHIFT;
149
150 /* CACHE_NEW_WM_PROG */
151 dw3 |=
152 ((brw->wm.prog_data->base.binding_table.size_bytes / 4) <<
153 GEN7_PS_BINDING_TABLE_ENTRY_COUNT_SHIFT);
154
155 /* Use ALT floating point mode for ARB fragment programs, because they
156 * require 0^0 == 1. Even though _CurrentFragmentProgram is used for
157 * rendering, CurrentFragmentProgram is used for this check to
158 * differentiate between the GLSL and non-GLSL cases.
159 */
160 if (ctx->Shader.CurrentProgram[MESA_SHADER_FRAGMENT] == NULL)
161 dw3 |= GEN7_PS_FLOATING_POINT_MODE_ALT;
162
163 dw6 |= (brw->max_wm_threads - 2) << HSW_PS_MAX_THREADS_SHIFT;
164
165 /* CACHE_NEW_WM_PROG */
166 if (brw->wm.prog_data->base.nr_params > 0)
167 dw6 |= GEN7_PS_PUSH_CONSTANT_ENABLE;
168
169 dw6 |= GEN7_PS_8_DISPATCH_ENABLE;
170 if (brw->wm.prog_data->prog_offset_16)
171 dw6 |= GEN7_PS_16_DISPATCH_ENABLE;
172
173 dw7 |=
174 brw->wm.prog_data->first_curbe_grf << GEN7_PS_DISPATCH_START_GRF_SHIFT_0 |
175 brw->wm.prog_data->first_curbe_grf_16<< GEN7_PS_DISPATCH_START_GRF_SHIFT_2;
176
177 BEGIN_BATCH(12);
178 OUT_BATCH(_3DSTATE_PS << 16 | (12 - 2));
179 OUT_BATCH(brw->wm.base.prog_offset);
180 OUT_BATCH(0);
181 OUT_BATCH(dw3);
182 if (brw->wm.prog_data->total_scratch) {
183 OUT_RELOC64(brw->wm.base.scratch_bo,
184 I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
185 ffs(brw->wm.prog_data->total_scratch) - 11);
186 } else {
187 OUT_BATCH(0);
188 OUT_BATCH(0);
189 }
190 OUT_BATCH(dw6);
191 OUT_BATCH(dw7);
192 OUT_BATCH(0); /* kernel 1 pointer */
193 OUT_BATCH(0);
194 OUT_BATCH(brw->wm.base.prog_offset + brw->wm.prog_data->prog_offset_16);
195 OUT_BATCH(0);
196 ADVANCE_BATCH();
197 }
198
199 const struct brw_tracked_state gen8_ps_state = {
200 .dirty = {
201 .mesa = _NEW_PROGRAM_CONSTANTS,
202 .brw = BRW_NEW_FRAGMENT_PROGRAM |
203 BRW_NEW_PS_BINDING_TABLE |
204 BRW_NEW_BATCH |
205 BRW_NEW_PUSH_CONSTANT_ALLOCATION,
206 .cache = CACHE_NEW_SAMPLER | CACHE_NEW_WM_PROG
207 },
208 .emit = upload_ps_state,
209 };