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 * \file brw_vec4_tes.cpp 26 * 27 * Tessellaton evaluation shader specific code derived from the vec4_visitor class. 28 */ 29 30#include "brw_vec4_tes.h" 31#include "brw_cfg.h" 32#include "dev/intel_debug.h" 33 34namespace brw { 35 36vec4_tes_visitor::vec4_tes_visitor(const struct brw_compiler *compiler, 37 void *log_data, 38 const struct brw_tes_prog_key *key, 39 struct brw_tes_prog_data *prog_data, 40 const nir_shader *shader, 41 void *mem_ctx, 42 bool debug_enabled) 43 : vec4_visitor(compiler, log_data, &key->base.tex, &prog_data->base, 44 shader, mem_ctx, false, debug_enabled) 45{ 46} 47 48void 49vec4_tes_visitor::setup_payload() 50{ 51 int reg = 0; 52 53 /* The payload always contains important data in r0 and r1, which contains 54 * the URB handles that are passed on to the URB write at the end 55 * of the thread. 56 */ 57 reg += 2; 58 59 reg = setup_uniforms(reg); 60 61 foreach_block_and_inst(block, vec4_instruction, inst, cfg) { 62 for (int i = 0; i < 3; i++) { 63 if (inst->src[i].file != ATTR) 64 continue; 65 66 unsigned slot = inst->src[i].nr + inst->src[i].offset / 16; 67 struct brw_reg grf = brw_vec4_grf(reg + slot / 2, 4 * (slot % 2)); 68 grf = stride(grf, 0, 4, 1); 69 grf.swizzle = inst->src[i].swizzle; 70 grf.type = inst->src[i].type; 71 grf.abs = inst->src[i].abs; 72 grf.negate = inst->src[i].negate; 73 inst->src[i] = grf; 74 } 75 } 76 77 reg += 8 * prog_data->urb_read_length; 78 79 this->first_non_payload_grf = reg; 80} 81 82 83void 84vec4_tes_visitor::emit_prolog() 85{ 86 input_read_header = src_reg(this, glsl_type::uvec4_type); 87 emit(TES_OPCODE_CREATE_INPUT_READ_HEADER, dst_reg(input_read_header)); 88 89 this->current_annotation = NULL; 90} 91 92 93void 94vec4_tes_visitor::emit_urb_write_header(int mrf) 95{ 96 /* No need to do anything for DS; an implied write to this MRF will be 97 * performed by VEC4_VS_OPCODE_URB_WRITE. 98 */ 99 (void) mrf; 100} 101 102 103vec4_instruction * 104vec4_tes_visitor::emit_urb_write_opcode(bool complete) 105{ 106 vec4_instruction *inst = emit(VEC4_VS_OPCODE_URB_WRITE); 107 inst->urb_write_flags = complete ? 108 BRW_URB_WRITE_EOT_COMPLETE : BRW_URB_WRITE_NO_FLAGS; 109 110 return inst; 111} 112 113void 114vec4_tes_visitor::nir_emit_intrinsic(nir_intrinsic_instr *instr) 115{ 116 const struct brw_tes_prog_data *tes_prog_data = 117 (const struct brw_tes_prog_data *) prog_data; 118 119 switch (instr->intrinsic) { 120 case nir_intrinsic_load_tess_coord: 121 /* gl_TessCoord is part of the payload in g1 channels 0-2 and 4-6. */ 122 emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_F), 123 src_reg(brw_vec8_grf(1, 0)))); 124 break; 125 case nir_intrinsic_load_tess_level_outer: 126 if (tes_prog_data->domain == BRW_TESS_DOMAIN_ISOLINE) { 127 emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_F), 128 swizzle(src_reg(ATTR, 1, glsl_type::vec4_type), 129 BRW_SWIZZLE_ZWZW))); 130 } else { 131 emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_F), 132 swizzle(src_reg(ATTR, 1, glsl_type::vec4_type), 133 BRW_SWIZZLE_WZYX))); 134 } 135 break; 136 case nir_intrinsic_load_tess_level_inner: 137 if (tes_prog_data->domain == BRW_TESS_DOMAIN_QUAD) { 138 emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_F), 139 swizzle(src_reg(ATTR, 0, glsl_type::vec4_type), 140 BRW_SWIZZLE_WZYX))); 141 } else { 142 emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_F), 143 src_reg(ATTR, 1, glsl_type::float_type))); 144 } 145 break; 146 case nir_intrinsic_load_primitive_id: 147 emit(TES_OPCODE_GET_PRIMITIVE_ID, 148 get_nir_dest(instr->dest, BRW_REGISTER_TYPE_UD)); 149 break; 150 151 case nir_intrinsic_load_input: 152 case nir_intrinsic_load_per_vertex_input: { 153 assert(nir_dest_bit_size(instr->dest) == 32); 154 src_reg indirect_offset = get_indirect_offset(instr); 155 unsigned imm_offset = instr->const_index[0]; 156 src_reg header = input_read_header; 157 unsigned first_component = nir_intrinsic_component(instr); 158 159 if (indirect_offset.file != BAD_FILE) { 160 src_reg clamped_indirect_offset = src_reg(this, glsl_type::uvec4_type); 161 162 /* Page 190 of "Volume 7: 3D Media GPGPU Engine (Haswell)" says the 163 * valid range of the offset is [0, 0FFFFFFFh]. 164 */ 165 emit_minmax(BRW_CONDITIONAL_L, 166 dst_reg(clamped_indirect_offset), 167 retype(indirect_offset, BRW_REGISTER_TYPE_UD), 168 brw_imm_ud(0x0fffffffu)); 169 170 header = src_reg(this, glsl_type::uvec4_type); 171 emit(TES_OPCODE_ADD_INDIRECT_URB_OFFSET, dst_reg(header), 172 input_read_header, clamped_indirect_offset); 173 } else { 174 /* Arbitrarily only push up to 24 vec4 slots worth of data, 175 * which is 12 registers (since each holds 2 vec4 slots). 176 */ 177 const unsigned max_push_slots = 24; 178 if (imm_offset < max_push_slots) { 179 src_reg src = src_reg(ATTR, imm_offset, glsl_type::ivec4_type); 180 src.swizzle = BRW_SWZ_COMP_INPUT(first_component); 181 182 emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_D), src)); 183 184 prog_data->urb_read_length = 185 MAX2(prog_data->urb_read_length, 186 DIV_ROUND_UP(imm_offset + 1, 2)); 187 break; 188 } 189 } 190 191 dst_reg temp(this, glsl_type::ivec4_type); 192 vec4_instruction *read = 193 emit(VEC4_OPCODE_URB_READ, temp, src_reg(header)); 194 read->offset = imm_offset; 195 read->urb_write_flags = BRW_URB_WRITE_PER_SLOT_OFFSET; 196 197 src_reg src = src_reg(temp); 198 src.swizzle = BRW_SWZ_COMP_INPUT(first_component); 199 200 /* Copy to target. We might end up with some funky writemasks landing 201 * in here, but we really don't want them in the above pseudo-ops. 202 */ 203 dst_reg dst = get_nir_dest(instr->dest, BRW_REGISTER_TYPE_D); 204 dst.writemask = brw_writemask_for_size(instr->num_components); 205 emit(MOV(dst, src)); 206 break; 207 } 208 default: 209 vec4_visitor::nir_emit_intrinsic(instr); 210 } 211} 212 213 214void 215vec4_tes_visitor::emit_thread_end() 216{ 217 /* For DS, we always end the thread by emitting a single vertex. 218 * emit_urb_write_opcode() will take care of setting the eot flag on the 219 * SEND instruction. 220 */ 221 emit_vertex(); 222} 223 224} /* namespace brw */ 225