1/* 2 * Copyright © 2010 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 ir_optimization.h 26 * 27 * Prototypes for optimization passes to be called by the compiler and drivers. 28 */ 29 30#ifndef GLSL_IR_OPTIMIZATION_H 31#define GLSL_IR_OPTIMIZATION_H 32 33struct gl_linked_shader; 34struct gl_shader_program; 35 36/* Operations for lower_instructions() */ 37#define SUB_TO_ADD_NEG 0x01 38#define LDEXP_TO_ARITH 0x80 39#define CARRY_TO_ARITH 0x100 40#define BORROW_TO_ARITH 0x200 41#define DOPS_TO_DFRAC 0x800 42#define DFREXP_DLDEXP_TO_ARITH 0x1000 43#define BIT_COUNT_TO_MATH 0x02000 44#define EXTRACT_TO_SHIFTS 0x04000 45#define INSERT_TO_SHIFTS 0x08000 46#define REVERSE_TO_SHIFTS 0x10000 47#define FIND_LSB_TO_FLOAT_CAST 0x20000 48#define FIND_MSB_TO_FLOAT_CAST 0x40000 49#define IMUL_HIGH_TO_MUL 0x80000 50#define SQRT_TO_ABS_SQRT 0x200000 51 52/* Operations for lower_64bit_integer_instructions() */ 53#define DIV64 (1U << 0) 54#define MOD64 (1U << 1) 55 56/** 57 * \see class lower_packing_builtins_visitor 58 */ 59enum lower_packing_builtins_op { 60 LOWER_PACK_UNPACK_NONE = 0x0000, 61 62 LOWER_PACK_SNORM_2x16 = 0x0001, 63 LOWER_UNPACK_SNORM_2x16 = 0x0002, 64 65 LOWER_PACK_UNORM_2x16 = 0x0004, 66 LOWER_UNPACK_UNORM_2x16 = 0x0008, 67 68 LOWER_PACK_HALF_2x16 = 0x0010, 69 LOWER_UNPACK_HALF_2x16 = 0x0020, 70 71 LOWER_PACK_SNORM_4x8 = 0x0040, 72 LOWER_UNPACK_SNORM_4x8 = 0x0080, 73 74 LOWER_PACK_UNORM_4x8 = 0x0100, 75 LOWER_UNPACK_UNORM_4x8 = 0x0200, 76 77 LOWER_PACK_USE_BFI = 0x0400, 78 LOWER_PACK_USE_BFE = 0x0800, 79}; 80 81bool do_common_optimization(exec_list *ir, bool linked, 82 const struct gl_shader_compiler_options *options, 83 bool native_integers); 84 85bool ir_constant_fold(ir_rvalue **rvalue); 86 87bool do_rebalance_tree(exec_list *instructions); 88bool do_algebraic(exec_list *instructions, bool native_integers, 89 const struct gl_shader_compiler_options *options); 90bool do_constant_folding(exec_list *instructions); 91bool do_constant_variable(exec_list *instructions); 92bool do_constant_variable_unlinked(exec_list *instructions); 93bool do_copy_propagation_elements(exec_list *instructions); 94bool do_constant_propagation(exec_list *instructions); 95bool do_dead_code(exec_list *instructions); 96bool do_dead_code_local(exec_list *instructions); 97bool do_dead_code_unlinked(exec_list *instructions); 98bool do_dead_functions(exec_list *instructions); 99bool opt_flip_matrices(exec_list *instructions); 100bool do_function_inlining(exec_list *instructions); 101bool do_lower_jumps(exec_list *instructions, bool pull_out_jumps = true, bool lower_sub_return = true, bool lower_main_return = false, bool lower_continue = false); 102bool do_if_simplification(exec_list *instructions); 103bool opt_flatten_nested_if_blocks(exec_list *instructions); 104bool do_mat_op_to_vec(exec_list *instructions); 105bool do_minmax_prune(exec_list *instructions); 106bool do_structure_splitting(exec_list *instructions); 107bool do_tree_grafting(exec_list *instructions); 108bool do_vec_index_to_cond_assign(exec_list *instructions); 109bool do_vec_index_to_swizzle(exec_list *instructions); 110bool lower_discard(exec_list *instructions); 111void lower_discard_flow(exec_list *instructions); 112bool lower_instructions(exec_list *instructions, unsigned what_to_lower); 113bool lower_clip_cull_distance(struct gl_shader_program *prog, 114 gl_linked_shader *shader); 115void lower_output_reads(unsigned stage, exec_list *instructions); 116bool lower_packing_builtins(exec_list *instructions, int op_mask); 117bool lower_vector_insert(exec_list *instructions, bool lower_nonconstant_index); 118bool lower_vector_derefs(gl_linked_shader *shader); 119void lower_named_interface_blocks(void *mem_ctx, gl_linked_shader *shader); 120bool optimize_split_arrays(exec_list *instructions, bool linked); 121void optimize_dead_builtin_variables(exec_list *instructions, 122 enum ir_variable_mode other); 123bool lower_tess_level(gl_linked_shader *shader); 124 125bool lower_vertex_id(gl_linked_shader *shader); 126bool lower_cs_derived(gl_linked_shader *shader); 127bool lower_blend_equation_advanced(gl_linked_shader *shader, bool coherent); 128 129bool lower_builtins(exec_list *instructions); 130bool lower_subroutine(exec_list *instructions, struct _mesa_glsl_parse_state *state); 131bool propagate_invariance(exec_list *instructions); 132 133namespace ir_builder { class ir_factory; }; 134 135ir_variable *compare_index_block(ir_builder::ir_factory &body, 136 ir_variable *index, 137 unsigned base, unsigned components); 138 139bool lower_64bit_integer_instructions(exec_list *instructions, 140 unsigned what_to_lower); 141 142void lower_precision(const struct gl_shader_compiler_options *options, 143 exec_list *instructions); 144 145#endif /* GLSL_IR_OPTIMIZATION_H */ 146