1 /*
2 * Copyright © 2013 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
21 * DEALINGS IN THE SOFTWARE.
22 */
23
24
25 #include "brw_vec4_vs.h"
26 #include "common/gen_debug.h"
27
28 namespace brw {
29
30 void
emit_prolog()31 vec4_vs_visitor::emit_prolog()
32 {
33 }
34
35
36 void
emit_urb_write_header(int mrf)37 vec4_vs_visitor::emit_urb_write_header(int mrf)
38 {
39 /* No need to do anything for VS; an implied write to this MRF will be
40 * performed by VS_OPCODE_URB_WRITE.
41 */
42 (void) mrf;
43 }
44
45
46 vec4_instruction *
emit_urb_write_opcode(bool complete)47 vec4_vs_visitor::emit_urb_write_opcode(bool complete)
48 {
49 /* For VS, the URB writes end the thread. */
50 if (complete) {
51 if (INTEL_DEBUG & DEBUG_SHADER_TIME)
52 emit_shader_time_end();
53 }
54
55 vec4_instruction *inst = emit(VS_OPCODE_URB_WRITE);
56 inst->urb_write_flags = complete ?
57 BRW_URB_WRITE_EOT_COMPLETE : BRW_URB_WRITE_NO_FLAGS;
58
59 return inst;
60 }
61
62
63 void
emit_urb_slot(dst_reg reg,int varying)64 vec4_vs_visitor::emit_urb_slot(dst_reg reg, int varying)
65 {
66 reg.type = BRW_REGISTER_TYPE_F;
67 output_reg[varying][0].type = reg.type;
68
69 switch (varying) {
70 case VARYING_SLOT_COL0:
71 case VARYING_SLOT_COL1:
72 case VARYING_SLOT_BFC0:
73 case VARYING_SLOT_BFC1: {
74 /* These built-in varyings are only supported in compatibility mode,
75 * and we only support GS in core profile. So, this must be a vertex
76 * shader.
77 */
78 vec4_instruction *inst = emit_generic_urb_slot(reg, varying, 0);
79 if (inst && key->clamp_vertex_color)
80 inst->saturate = true;
81 break;
82 }
83 default:
84 return vec4_visitor::emit_urb_slot(reg, varying);
85 }
86 }
87
88
89 void
emit_clip_distances(dst_reg reg,int offset)90 vec4_vs_visitor::emit_clip_distances(dst_reg reg, int offset)
91 {
92 /* From the GLSL 1.30 spec, section 7.1 (Vertex Shader Special Variables):
93 *
94 * "If a linked set of shaders forming the vertex stage contains no
95 * static write to gl_ClipVertex or gl_ClipDistance, but the
96 * application has requested clipping against user clip planes through
97 * the API, then the coordinate written to gl_Position is used for
98 * comparison against the user clip planes."
99 *
100 * This function is only called if the shader didn't write to
101 * gl_ClipDistance. Accordingly, we use gl_ClipVertex to perform clipping
102 * if the user wrote to it; otherwise we use gl_Position.
103 */
104 gl_varying_slot clip_vertex = VARYING_SLOT_CLIP_VERTEX;
105 if (!(prog_data->vue_map.slots_valid & VARYING_BIT_CLIP_VERTEX)) {
106 clip_vertex = VARYING_SLOT_POS;
107 }
108
109 for (int i = 0; i + offset < key->nr_userclip_plane_consts && i < 4;
110 ++i) {
111 reg.writemask = 1 << i;
112 emit(DP4(reg,
113 src_reg(output_reg[clip_vertex][0]),
114 src_reg(this->userplane[i + offset])));
115 }
116 }
117
118
119 void
setup_uniform_clipplane_values()120 vec4_vs_visitor::setup_uniform_clipplane_values()
121 {
122 if (key->nr_userclip_plane_consts == 0)
123 return;
124
125 assert(stage_prog_data->nr_params == (unsigned)this->uniforms * 4);
126 brw_stage_prog_data_add_params(stage_prog_data,
127 key->nr_userclip_plane_consts * 4);
128
129 for (int i = 0; i < key->nr_userclip_plane_consts; ++i) {
130 this->userplane[i] = dst_reg(UNIFORM, this->uniforms);
131 this->userplane[i].type = BRW_REGISTER_TYPE_F;
132 for (int j = 0; j < 4; ++j) {
133 stage_prog_data->param[this->uniforms * 4 + j] =
134 BRW_PARAM_BUILTIN_CLIP_PLANE(i, j);
135 }
136 ++this->uniforms;
137 }
138 }
139
140
141 void
emit_thread_end()142 vec4_vs_visitor::emit_thread_end()
143 {
144 setup_uniform_clipplane_values();
145
146 /* Lower legacy ff and ClipVertex clipping to clip distances */
147 if (key->nr_userclip_plane_consts > 0) {
148 current_annotation = "user clip distances";
149
150 output_reg[VARYING_SLOT_CLIP_DIST0][0] =
151 dst_reg(this, glsl_type::vec4_type);
152 output_reg[VARYING_SLOT_CLIP_DIST1][0] =
153 dst_reg(this, glsl_type::vec4_type);
154 output_num_components[VARYING_SLOT_CLIP_DIST0][0] = 4;
155 output_num_components[VARYING_SLOT_CLIP_DIST1][0] = 4;
156
157 emit_clip_distances(output_reg[VARYING_SLOT_CLIP_DIST0][0], 0);
158 emit_clip_distances(output_reg[VARYING_SLOT_CLIP_DIST1][0], 4);
159 }
160
161 /* For VS, we always end the thread by emitting a single vertex.
162 * emit_urb_write_opcode() will take care of setting the eot flag on the
163 * SEND instruction.
164 */
165 emit_vertex();
166 }
167
168
vec4_vs_visitor(const struct brw_compiler * compiler,void * log_data,const struct brw_vs_prog_key * key,struct brw_vs_prog_data * vs_prog_data,const nir_shader * shader,void * mem_ctx,int shader_time_index)169 vec4_vs_visitor::vec4_vs_visitor(const struct brw_compiler *compiler,
170 void *log_data,
171 const struct brw_vs_prog_key *key,
172 struct brw_vs_prog_data *vs_prog_data,
173 const nir_shader *shader,
174 void *mem_ctx,
175 int shader_time_index)
176 : vec4_visitor(compiler, log_data, &key->tex, &vs_prog_data->base, shader,
177 mem_ctx, false /* no_spills */, shader_time_index),
178 key(key),
179 vs_prog_data(vs_prog_data)
180 {
181 }
182
183
184 } /* namespace brw */
185