1/*
2 * Copyright 2015-2021 Arm Limited
3 * SPDX-License-Identifier: Apache-2.0 OR MIT
4 *
5 * Licensed under the Apache License, Version 2.0 (the "License");
6 * you may not use this file except in compliance with the License.
7 * You may obtain a copy of the License at
8 *
9 * http://www.apache.org/licenses/LICENSE-2.0
10 *
11 * Unless required by applicable law or agreed to in writing, software
12 * distributed under the License is distributed on an "AS IS" BASIS,
13 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14 * See the License for the specific language governing permissions and
15 * limitations under the License.
16 */
17
18/*
19 * At your option, you may choose to accept this material under either:
20 * 1. The Apache License, Version 2.0, found at <http://www.apache.org/licenses/LICENSE-2.0>, or
21 * 2. The MIT License, found at <http://opensource.org/licenses/MIT>.
22 */
23
24#ifndef SPIRV_CROSS_GLSL_HPP
25#define SPIRV_CROSS_GLSL_HPP
26
27#include "GLSL.std.450.h"
28#include "spirv_cross.hpp"
29#include <unordered_map>
30#include <unordered_set>
31#include <utility>
32
33namespace SPIRV_CROSS_NAMESPACE
34{
35enum PlsFormat
36{
37 PlsNone = 0,
38
39 PlsR11FG11FB10F,
40 PlsR32F,
41 PlsRG16F,
42 PlsRGB10A2,
43 PlsRGBA8,
44 PlsRG16,
45
46 PlsRGBA8I,
47 PlsRG16I,
48
49 PlsRGB10A2UI,
50 PlsRGBA8UI,
51 PlsRG16UI,
52 PlsR32UI
53};
54
55struct PlsRemap
56{
57 uint32_t id;
58 PlsFormat format;
59};
60
61enum AccessChainFlagBits
62{
63 ACCESS_CHAIN_INDEX_IS_LITERAL_BIT = 1 << 0,
64 ACCESS_CHAIN_CHAIN_ONLY_BIT = 1 << 1,
65 ACCESS_CHAIN_PTR_CHAIN_BIT = 1 << 2,
66 ACCESS_CHAIN_SKIP_REGISTER_EXPRESSION_READ_BIT = 1 << 3,
67 ACCESS_CHAIN_LITERAL_MSB_FORCE_ID = 1 << 4,
68 ACCESS_CHAIN_FLATTEN_ALL_MEMBERS_BIT = 1 << 5,
69 ACCESS_CHAIN_FORCE_COMPOSITE_BIT = 1 << 6
70};
71typedef uint32_t AccessChainFlags;
72
73class CompilerGLSL : public Compiler
74{
75public:
76 struct Options
77 {
78 // The shading language version. Corresponds to #version $VALUE.
79 uint32_t version = 450;
80
81 // Emit the OpenGL ES shading language instead of desktop OpenGL.
82 bool es = false;
83
84 // Debug option to always emit temporary variables for all expressions.
85 bool force_temporary = false;
86 // Debug option, can be increased in an attempt to workaround SPIRV-Cross bugs temporarily.
87 // If this limit has to be increased, it points to an implementation bug.
88 // In certain scenarios, the maximum number of debug iterations may increase beyond this limit
89 // as long as we can prove we're making certain kinds of forward progress.
90 uint32_t force_recompile_max_debug_iterations = 3;
91
92 // If true, Vulkan GLSL features are used instead of GL-compatible features.
93 // Mostly useful for debugging SPIR-V files.
94 bool vulkan_semantics = false;
95
96 // If true, gl_PerVertex is explicitly redeclared in vertex, geometry and tessellation shaders.
97 // The members of gl_PerVertex is determined by which built-ins are declared by the shader.
98 // This option is ignored in ES versions, as redeclaration in ES is not required, and it depends on a different extension
99 // (EXT_shader_io_blocks) which makes things a bit more fuzzy.
100 bool separate_shader_objects = false;
101
102 // Flattens multidimensional arrays, e.g. float foo[a][b][c] into single-dimensional arrays,
103 // e.g. float foo[a * b * c].
104 // This function does not change the actual SPIRType of any object.
105 // Only the generated code, including declarations of interface variables are changed to be single array dimension.
106 bool flatten_multidimensional_arrays = false;
107
108 // For older desktop GLSL targets than version 420, the
109 // GL_ARB_shading_language_420pack extensions is used to be able to support
110 // layout(binding) on UBOs and samplers.
111 // If disabled on older targets, binding decorations will be stripped.
112 bool enable_420pack_extension = true;
113
114 // In non-Vulkan GLSL, emit push constant blocks as UBOs rather than plain uniforms.
115 bool emit_push_constant_as_uniform_buffer = false;
116
117 // Always emit uniform blocks as plain uniforms, regardless of the GLSL version, even when UBOs are supported.
118 // Does not apply to shader storage or push constant blocks.
119 bool emit_uniform_buffer_as_plain_uniforms = false;
120
121 // Emit OpLine directives if present in the module.
122 // May not correspond exactly to original source, but should be a good approximation.
123 bool emit_line_directives = false;
124
125 // In cases where readonly/writeonly decoration are not used at all,
126 // we try to deduce which qualifier(s) we should actually used, since actually emitting
127 // read-write decoration is very rare, and older glslang/HLSL compilers tend to just emit readwrite as a matter of fact.
128 // The default (true) is to enable automatic deduction for these cases, but if you trust the decorations set
129 // by the SPIR-V, it's recommended to set this to false.
130 bool enable_storage_image_qualifier_deduction = true;
131
132 // On some targets (WebGPU), uninitialized variables are banned.
133 // If this is enabled, all variables (temporaries, Private, Function)
134 // which would otherwise be uninitialized will now be initialized to 0 instead.
135 bool force_zero_initialized_variables = false;
136
137 // In GLSL, force use of I/O block flattening, similar to
138 // what happens on legacy GLSL targets for blocks and structs.
139 bool force_flattened_io_blocks = false;
140
141 // For opcodes where we have to perform explicit additional nan checks, very ugly code is generated.
142 // If we opt-in, ignore these requirements.
143 // In opcodes like NClamp/NMin/NMax and FP compare, ignore NaN behavior.
144 // Use FClamp/FMin/FMax semantics for clamps and lets implementation choose ordered or unordered
145 // compares.
146 bool relax_nan_checks = false;
147
148 // Loading row-major matrices from UBOs on older AMD Windows OpenGL drivers is problematic.
149 // To load these types correctly, we must generate a wrapper. them in a dummy function which only purpose is to
150 // ensure row_major decoration is actually respected.
151 // This workaround may cause significant performance degeneration on some Android devices.
152 bool enable_row_major_load_workaround = true;
153
154 // If non-zero, controls layout(num_views = N) in; in GL_OVR_multiview2.
155 uint32_t ovr_multiview_view_count = 0;
156
157 enum Precision
158 {
159 DontCare,
160 Lowp,
161 Mediump,
162 Highp
163 };
164
165 struct VertexOptions
166 {
167 // "Vertex-like shader" here is any shader stage that can write BuiltInPosition.
168
169 // GLSL: In vertex-like shaders, rewrite [0, w] depth (Vulkan/D3D style) to [-w, w] depth (GL style).
170 // MSL: In vertex-like shaders, rewrite [-w, w] depth (GL style) to [0, w] depth.
171 // HLSL: In vertex-like shaders, rewrite [-w, w] depth (GL style) to [0, w] depth.
172 bool fixup_clipspace = false;
173
174 // In vertex-like shaders, inverts gl_Position.y or equivalent.
175 bool flip_vert_y = false;
176
177 // GLSL only, for HLSL version of this option, see CompilerHLSL.
178 // If true, the backend will assume that InstanceIndex will need to apply
179 // a base instance offset. Set to false if you know you will never use base instance
180 // functionality as it might remove some internal uniforms.
181 bool support_nonzero_base_instance = true;
182 } vertex;
183
184 struct FragmentOptions
185 {
186 // Add precision mediump float in ES targets when emitting GLES source.
187 // Add precision highp int in ES targets when emitting GLES source.
188 Precision default_float_precision = Mediump;
189 Precision default_int_precision = Highp;
190 } fragment;
191 };
192
193 void remap_pixel_local_storage(std::vector<PlsRemap> inputs, std::vector<PlsRemap> outputs)
194 {
195 pls_inputs = std::move(inputs);
196 pls_outputs = std::move(outputs);
197 remap_pls_variables();
198 }
199
200 // Redirect a subpassInput reading from input_attachment_index to instead load its value from
201 // the color attachment at location = color_location. Requires ESSL.
202 // If coherent, uses GL_EXT_shader_framebuffer_fetch, if not, uses noncoherent variant.
203 void remap_ext_framebuffer_fetch(uint32_t input_attachment_index, uint32_t color_location, bool coherent);
204
205 explicit CompilerGLSL(std::vector<uint32_t> spirv_)
206 : Compiler(std::move(spirv_))
207 {
208 init();
209 }
210
211 CompilerGLSL(const uint32_t *ir_, size_t word_count)
212 : Compiler(ir_, word_count)
213 {
214 init();
215 }
216
217 explicit CompilerGLSL(const ParsedIR &ir_)
218 : Compiler(ir_)
219 {
220 init();
221 }
222
223 explicit CompilerGLSL(ParsedIR &&ir_)
224 : Compiler(std::move(ir_))
225 {
226 init();
227 }
228
229 const Options &get_common_options() const
230 {
231 return options;
232 }
233
234 void set_common_options(const Options &opts)
235 {
236 options = opts;
237 }
238
239 std::string compile() override;
240
241 // Returns the current string held in the conversion buffer. Useful for
242 // capturing what has been converted so far when compile() throws an error.
243 std::string get_partial_source();
244
245 // Adds a line to be added right after #version in GLSL backend.
246 // This is useful for enabling custom extensions which are outside the scope of SPIRV-Cross.
247 // This can be combined with variable remapping.
248 // A new-line will be added.
249 //
250 // While add_header_line() is a more generic way of adding arbitrary text to the header
251 // of a GLSL file, require_extension() should be used when adding extensions since it will
252 // avoid creating collisions with SPIRV-Cross generated extensions.
253 //
254 // Code added via add_header_line() is typically backend-specific.
255 void add_header_line(const std::string &str);
256
257 // Adds an extension which is required to run this shader, e.g.
258 // require_extension("GL_KHR_my_extension");
259 void require_extension(const std::string &ext);
260
261 // Returns the list of required extensions. After compilation this will contains any other
262 // extensions that the compiler used automatically, in addition to the user specified ones.
263 const SmallVector<std::string> &get_required_extensions() const;
264
265 // Legacy GLSL compatibility method.
266 // Takes a uniform or push constant variable and flattens it into a (i|u)vec4 array[N]; array instead.
267 // For this to work, all types in the block must be the same basic type, e.g. mixing vec2 and vec4 is fine, but
268 // mixing int and float is not.
269 // The name of the uniform array will be the same as the interface block name.
270 void flatten_buffer_block(VariableID id);
271
272 // After compilation, query if a variable ID was used as a depth resource.
273 // This is meaningful for MSL since descriptor types depend on this knowledge.
274 // Cases which return true:
275 // - Images which are declared with depth = 1 image type.
276 // - Samplers which are statically used at least once with Dref opcodes.
277 // - Images which are statically used at least once with Dref opcodes.
278 bool variable_is_depth_or_compare(VariableID id) const;
279
280 // If a shader output is active in this stage, but inactive in a subsequent stage,
281 // this can be signalled here. This can be used to work around certain cross-stage matching problems
282 // which plagues MSL and HLSL in certain scenarios.
283 // An output which matches one of these will not be emitted in stage output interfaces, but rather treated as a private
284 // variable.
285 // This option is only meaningful for MSL and HLSL, since GLSL matches by location directly.
286 // Masking builtins only takes effect if the builtin in question is part of the stage output interface.
287 void mask_stage_output_by_location(uint32_t location, uint32_t component);
288 void mask_stage_output_by_builtin(spv::BuiltIn builtin);
289
290 // Allow to control how to format float literals in the output.
291 // Set to "nullptr" to use the default "convert_to_string" function.
292 // This handle is not owned by SPIRV-Cross and must remain valid until compile() has been called.
293 void set_float_formatter(FloatFormatter *formatter)
294 {
295 float_formatter = formatter;
296 }
297
298protected:
299 struct ShaderSubgroupSupportHelper
300 {
301 // lower enum value = greater priority
302 enum Candidate
303 {
304 KHR_shader_subgroup_ballot,
305 KHR_shader_subgroup_basic,
306 KHR_shader_subgroup_vote,
307 KHR_shader_subgroup_arithmetic,
308 NV_gpu_shader_5,
309 NV_shader_thread_group,
310 NV_shader_thread_shuffle,
311 ARB_shader_ballot,
312 ARB_shader_group_vote,
313 AMD_gcn_shader,
314
315 CandidateCount
316 };
317
318 static const char *get_extension_name(Candidate c);
319 static SmallVector<std::string> get_extra_required_extension_names(Candidate c);
320 static const char *get_extra_required_extension_predicate(Candidate c);
321
322 enum Feature
323 {
324 SubgroupMask = 0,
325 SubgroupSize = 1,
326 SubgroupInvocationID = 2,
327 SubgroupID = 3,
328 NumSubgroups = 4,
329 SubgroupBroadcast_First = 5,
330 SubgroupBallotFindLSB_MSB = 6,
331 SubgroupAll_Any_AllEqualBool = 7,
332 SubgroupAllEqualT = 8,
333 SubgroupElect = 9,
334 SubgroupBarrier = 10,
335 SubgroupMemBarrier = 11,
336 SubgroupBallot = 12,
337 SubgroupInverseBallot_InclBitCount_ExclBitCout = 13,
338 SubgroupBallotBitExtract = 14,
339 SubgroupBallotBitCount = 15,
340 SubgroupArithmeticIAddReduce = 16,
341 SubgroupArithmeticIAddExclusiveScan = 17,
342 SubgroupArithmeticIAddInclusiveScan = 18,
343 SubgroupArithmeticFAddReduce = 19,
344 SubgroupArithmeticFAddExclusiveScan = 20,
345 SubgroupArithmeticFAddInclusiveScan = 21,
346 SubgroupArithmeticIMulReduce = 22,
347 SubgroupArithmeticIMulExclusiveScan = 23,
348 SubgroupArithmeticIMulInclusiveScan = 24,
349 SubgroupArithmeticFMulReduce = 25,
350 SubgroupArithmeticFMulExclusiveScan = 26,
351 SubgroupArithmeticFMulInclusiveScan = 27,
352 FeatureCount
353 };
354
355 using FeatureMask = uint32_t;
356 static_assert(sizeof(FeatureMask) * 8u >= FeatureCount, "Mask type needs more bits.");
357
358 using CandidateVector = SmallVector<Candidate, CandidateCount>;
359 using FeatureVector = SmallVector<Feature>;
360
361 static FeatureVector get_feature_dependencies(Feature feature);
362 static FeatureMask get_feature_dependency_mask(Feature feature);
363 static bool can_feature_be_implemented_without_extensions(Feature feature);
364 static Candidate get_KHR_extension_for_feature(Feature feature);
365
366 struct Result
367 {
368 Result();
369 uint32_t weights[CandidateCount];
370 };
371
372 void request_feature(Feature feature);
373 bool is_feature_requested(Feature feature) const;
374 Result resolve() const;
375
376 static CandidateVector get_candidates_for_feature(Feature ft, const Result &r);
377
378 private:
379 static CandidateVector get_candidates_for_feature(Feature ft);
380 static FeatureMask build_mask(const SmallVector<Feature> &features);
381 FeatureMask feature_mask = 0;
382 };
383
384 // TODO remove this function when all subgroup ops are supported (or make it always return true)
385 static bool is_supported_subgroup_op_in_opengl(spv::Op op, const uint32_t *ops);
386
387 void reset(uint32_t iteration_count);
388 void emit_function(SPIRFunction &func, const Bitset &return_flags);
389
390 bool has_extension(const std::string &ext) const;
391 void require_extension_internal(const std::string &ext);
392
393 // Virtualize methods which need to be overridden by subclass targets like C++ and such.
394 virtual void emit_function_prototype(SPIRFunction &func, const Bitset &return_flags);
395
396 SPIRBlock *current_emitting_block = nullptr;
397 SmallVector<SPIRBlock *> current_emitting_switch_stack;
398 bool current_emitting_switch_fallthrough = false;
399
400 virtual void emit_instruction(const Instruction &instr);
401 struct TemporaryCopy
402 {
403 uint32_t dst_id;
404 uint32_t src_id;
405 };
406 TemporaryCopy handle_instruction_precision(const Instruction &instr);
407 void emit_block_instructions(SPIRBlock &block);
408 void emit_block_instructions_with_masked_debug(SPIRBlock &block);
409
410 // For relax_nan_checks.
411 GLSLstd450 get_remapped_glsl_op(GLSLstd450 std450_op) const;
412 spv::Op get_remapped_spirv_op(spv::Op op) const;
413
414 virtual void emit_glsl_op(uint32_t result_type, uint32_t result_id, uint32_t op, const uint32_t *args,
415 uint32_t count);
416 virtual void emit_spv_amd_shader_ballot_op(uint32_t result_type, uint32_t result_id, uint32_t op,
417 const uint32_t *args, uint32_t count);
418 virtual void emit_spv_amd_shader_explicit_vertex_parameter_op(uint32_t result_type, uint32_t result_id, uint32_t op,
419 const uint32_t *args, uint32_t count);
420 virtual void emit_spv_amd_shader_trinary_minmax_op(uint32_t result_type, uint32_t result_id, uint32_t op,
421 const uint32_t *args, uint32_t count);
422 virtual void emit_spv_amd_gcn_shader_op(uint32_t result_type, uint32_t result_id, uint32_t op, const uint32_t *args,
423 uint32_t count);
424 virtual void emit_header();
425 void emit_line_directive(uint32_t file_id, uint32_t line_literal);
426 void build_workgroup_size(SmallVector<std::string> &arguments, const SpecializationConstant &x,
427 const SpecializationConstant &y, const SpecializationConstant &z);
428
429 void request_subgroup_feature(ShaderSubgroupSupportHelper::Feature feature);
430
431 virtual void emit_sampled_image_op(uint32_t result_type, uint32_t result_id, uint32_t image_id, uint32_t samp_id);
432 virtual void emit_texture_op(const Instruction &i, bool sparse);
433 virtual std::string to_texture_op(const Instruction &i, bool sparse, bool *forward,
434 SmallVector<uint32_t> &inherited_expressions);
435 virtual void emit_subgroup_op(const Instruction &i);
436 virtual std::string type_to_glsl(const SPIRType &type, uint32_t id = 0);
437 virtual std::string builtin_to_glsl(spv::BuiltIn builtin, spv::StorageClass storage);
438 virtual void emit_struct_member(const SPIRType &type, uint32_t member_type_id, uint32_t index,
439 const std::string &qualifier = "", uint32_t base_offset = 0);
440 virtual void emit_struct_padding_target(const SPIRType &type);
441 virtual std::string image_type_glsl(const SPIRType &type, uint32_t id = 0, bool member = false);
442 std::string constant_expression(const SPIRConstant &c,
443 bool inside_block_like_struct_scope = false,
444 bool inside_struct_scope = false);
445 virtual std::string constant_op_expression(const SPIRConstantOp &cop);
446 virtual std::string constant_expression_vector(const SPIRConstant &c, uint32_t vector);
447 virtual void emit_fixup();
448 virtual std::string variable_decl(const SPIRType &type, const std::string &name, uint32_t id = 0);
449 virtual bool variable_decl_is_remapped_storage(const SPIRVariable &var, spv::StorageClass storage) const;
450 virtual std::string to_func_call_arg(const SPIRFunction::Parameter &arg, uint32_t id);
451
452 struct TextureFunctionBaseArguments
453 {
454 // GCC 4.8 workarounds, it doesn't understand '{}' constructor here, use explicit default constructor.
455 TextureFunctionBaseArguments() = default;
456 VariableID img = 0;
457 const SPIRType *imgtype = nullptr;
458 bool is_fetch = false, is_gather = false, is_proj = false;
459 };
460
461 struct TextureFunctionNameArguments
462 {
463 // GCC 4.8 workarounds, it doesn't understand '{}' constructor here, use explicit default constructor.
464 TextureFunctionNameArguments() = default;
465 TextureFunctionBaseArguments base;
466 bool has_array_offsets = false, has_offset = false, has_grad = false;
467 bool has_dref = false, is_sparse_feedback = false, has_min_lod = false;
468 uint32_t lod = 0;
469 };
470 virtual std::string to_function_name(const TextureFunctionNameArguments &args);
471
472 struct TextureFunctionArguments
473 {
474 // GCC 4.8 workarounds, it doesn't understand '{}' constructor here, use explicit default constructor.
475 TextureFunctionArguments() = default;
476 TextureFunctionBaseArguments base;
477 uint32_t coord = 0, coord_components = 0, dref = 0;
478 uint32_t grad_x = 0, grad_y = 0, lod = 0, offset = 0;
479 uint32_t bias = 0, component = 0, sample = 0, sparse_texel = 0, min_lod = 0;
480 bool nonuniform_expression = false, has_array_offsets = false;
481 };
482 virtual std::string to_function_args(const TextureFunctionArguments &args, bool *p_forward);
483
484 void emit_sparse_feedback_temporaries(uint32_t result_type_id, uint32_t id, uint32_t &feedback_id,
485 uint32_t &texel_id);
486 uint32_t get_sparse_feedback_texel_id(uint32_t id) const;
487 virtual void emit_buffer_block(const SPIRVariable &type);
488 virtual void emit_push_constant_block(const SPIRVariable &var);
489 virtual void emit_uniform(const SPIRVariable &var);
490 virtual std::string unpack_expression_type(std::string expr_str, const SPIRType &type, uint32_t physical_type_id,
491 bool packed_type, bool row_major);
492
493 virtual bool builtin_translates_to_nonarray(spv::BuiltIn builtin) const;
494
495 virtual bool is_user_type_structured(uint32_t id) const;
496
497 void emit_copy_logical_type(uint32_t lhs_id, uint32_t lhs_type_id, uint32_t rhs_id, uint32_t rhs_type_id,
498 SmallVector<uint32_t> chain);
499
500 StringStream<> buffer;
501
502 template <typename T>
503 inline void statement_inner(T &&t)
504 {
505 buffer << std::forward<T>(t);
506 statement_count++;
507 }
508
509 template <typename T, typename... Ts>
510 inline void statement_inner(T &&t, Ts &&... ts)
511 {
512 buffer << std::forward<T>(t);
513 statement_count++;
514 statement_inner(std::forward<Ts>(ts)...);
515 }
516
517 template <typename... Ts>
518 inline void statement(Ts &&... ts)
519 {
520 if (is_forcing_recompilation())
521 {
522 // Do not bother emitting code while force_recompile is active.
523 // We will compile again.
524 statement_count++;
525 return;
526 }
527
528 if (redirect_statement)
529 {
530 redirect_statement->push_back(join(std::forward<Ts>(ts)...));
531 statement_count++;
532 }
533 else
534 {
535 for (uint32_t i = 0; i < indent; i++)
536 buffer << " ";
537 statement_inner(std::forward<Ts>(ts)...);
538 buffer << '\n';
539 }
540 }
541
542 template <typename... Ts>
543 inline void statement_no_indent(Ts &&... ts)
544 {
545 auto old_indent = indent;
546 indent = 0;
547 statement(std::forward<Ts>(ts)...);
548 indent = old_indent;
549 }
550
551 // Used for implementing continue blocks where
552 // we want to obtain a list of statements we can merge
553 // on a single line separated by comma.
554 SmallVector<std::string> *redirect_statement = nullptr;
555 const SPIRBlock *current_continue_block = nullptr;
556 bool block_temporary_hoisting = false;
557 bool block_debug_directives = false;
558
559 void begin_scope();
560 void end_scope();
561 void end_scope(const std::string &trailer);
562 void end_scope_decl();
563 void end_scope_decl(const std::string &decl);
564
565 Options options;
566
567 // Allow Metal to use the array<T> template to make arrays a value type
568 virtual std::string type_to_array_glsl(const SPIRType &type, uint32_t variable_id);
569 std::string to_array_size(const SPIRType &type, uint32_t index);
570 uint32_t to_array_size_literal(const SPIRType &type, uint32_t index) const;
571 uint32_t to_array_size_literal(const SPIRType &type) const;
572 virtual std::string variable_decl(const SPIRVariable &variable); // Threadgroup arrays can't have a wrapper type
573 std::string variable_decl_function_local(SPIRVariable &variable);
574
575 void add_local_variable_name(uint32_t id);
576 void add_resource_name(uint32_t id);
577 void add_member_name(SPIRType &type, uint32_t name);
578 void add_function_overload(const SPIRFunction &func);
579
580 virtual bool is_non_native_row_major_matrix(uint32_t id);
581 virtual bool member_is_non_native_row_major_matrix(const SPIRType &type, uint32_t index);
582 bool member_is_remapped_physical_type(const SPIRType &type, uint32_t index) const;
583 bool member_is_packed_physical_type(const SPIRType &type, uint32_t index) const;
584 virtual std::string convert_row_major_matrix(std::string exp_str, const SPIRType &exp_type,
585 uint32_t physical_type_id, bool is_packed,
586 bool relaxed = false);
587
588 std::unordered_set<std::string> local_variable_names;
589 std::unordered_set<std::string> resource_names;
590 std::unordered_set<std::string> block_input_names;
591 std::unordered_set<std::string> block_output_names;
592 std::unordered_set<std::string> block_ubo_names;
593 std::unordered_set<std::string> block_ssbo_names;
594 std::unordered_set<std::string> block_names; // A union of all block_*_names.
595 std::unordered_map<std::string, std::unordered_set<uint64_t>> function_overloads;
596 std::unordered_map<uint32_t, std::string> preserved_aliases;
597 void preserve_alias_on_reset(uint32_t id);
598 void reset_name_caches();
599
600 bool processing_entry_point = false;
601
602 // Can be overriden by subclass backends for trivial things which
603 // shouldn't need polymorphism.
604 struct BackendVariations
605 {
606 std::string discard_literal = "discard";
607 std::string demote_literal = "demote";
608 std::string null_pointer_literal = "";
609 bool float_literal_suffix = false;
610 bool double_literal_suffix = true;
611 bool uint32_t_literal_suffix = true;
612 bool long_long_literal_suffix = false;
613 const char *basic_int_type = "int";
614 const char *basic_uint_type = "uint";
615 const char *basic_int8_type = "int8_t";
616 const char *basic_uint8_type = "uint8_t";
617 const char *basic_int16_type = "int16_t";
618 const char *basic_uint16_type = "uint16_t";
619 const char *int16_t_literal_suffix = "s";
620 const char *uint16_t_literal_suffix = "us";
621 const char *nonuniform_qualifier = "nonuniformEXT";
622 const char *boolean_mix_function = "mix";
623 SPIRType::BaseType boolean_in_struct_remapped_type = SPIRType::Boolean;
624 bool swizzle_is_function = false;
625 bool shared_is_implied = false;
626 bool unsized_array_supported = true;
627 bool explicit_struct_type = false;
628 bool use_initializer_list = false;
629 bool use_typed_initializer_list = false;
630 bool can_declare_struct_inline = true;
631 bool can_declare_arrays_inline = true;
632 bool native_row_major_matrix = true;
633 bool use_constructor_splatting = true;
634 bool allow_precision_qualifiers = false;
635 bool can_swizzle_scalar = false;
636 bool force_gl_in_out_block = false;
637 bool force_merged_mesh_block = false;
638 bool can_return_array = true;
639 bool allow_truncated_access_chain = false;
640 bool supports_extensions = false;
641 bool supports_empty_struct = false;
642 bool array_is_value_type = true;
643 bool array_is_value_type_in_buffer_blocks = true;
644 bool comparison_image_samples_scalar = false;
645 bool native_pointers = false;
646 bool support_small_type_sampling_result = false;
647 bool support_case_fallthrough = true;
648 bool use_array_constructor = false;
649 bool needs_row_major_load_workaround = false;
650 bool support_pointer_to_pointer = false;
651 bool support_precise_qualifier = false;
652 bool support_64bit_switch = false;
653 bool workgroup_size_is_hidden = false;
654 bool requires_relaxed_precision_analysis = false;
655 bool implicit_c_integer_promotion_rules = false;
656 } backend;
657
658 void emit_struct(SPIRType &type);
659 void emit_resources();
660 void emit_extension_workarounds(spv::ExecutionModel model);
661 void emit_subgroup_arithmetic_workaround(const std::string &func, spv::Op op, spv::GroupOperation group_op);
662 void emit_polyfills(uint32_t polyfills, bool relaxed);
663 void emit_buffer_block_native(const SPIRVariable &var);
664 void emit_buffer_reference_block(uint32_t type_id, bool forward_declaration);
665 void emit_buffer_block_legacy(const SPIRVariable &var);
666 void emit_buffer_block_flattened(const SPIRVariable &type);
667 void fixup_implicit_builtin_block_names(spv::ExecutionModel model);
668 void emit_declared_builtin_block(spv::StorageClass storage, spv::ExecutionModel model);
669 bool should_force_emit_builtin_block(spv::StorageClass storage);
670 void emit_push_constant_block_vulkan(const SPIRVariable &var);
671 void emit_push_constant_block_glsl(const SPIRVariable &var);
672 void emit_interface_block(const SPIRVariable &type);
673 void emit_flattened_io_block(const SPIRVariable &var, const char *qual);
674 void emit_flattened_io_block_struct(const std::string &basename, const SPIRType &type, const char *qual,
675 const SmallVector<uint32_t> &indices);
676 void emit_flattened_io_block_member(const std::string &basename, const SPIRType &type, const char *qual,
677 const SmallVector<uint32_t> &indices);
678 void emit_block_chain(SPIRBlock &block);
679 void emit_hoisted_temporaries(SmallVector<std::pair<TypeID, ID>> &temporaries);
680 std::string constant_value_macro_name(uint32_t id);
681 int get_constant_mapping_to_workgroup_component(const SPIRConstant &constant) const;
682 void emit_constant(const SPIRConstant &constant);
683 void emit_specialization_constant_op(const SPIRConstantOp &constant);
684 std::string emit_continue_block(uint32_t continue_block, bool follow_true_block, bool follow_false_block);
685 bool attempt_emit_loop_header(SPIRBlock &block, SPIRBlock::Method method);
686
687 void branch(BlockID from, BlockID to);
688 void branch_to_continue(BlockID from, BlockID to);
689 void branch(BlockID from, uint32_t cond, BlockID true_block, BlockID false_block);
690 void flush_phi(BlockID from, BlockID to);
691 void flush_variable_declaration(uint32_t id);
692 void flush_undeclared_variables(SPIRBlock &block);
693 void emit_variable_temporary_copies(const SPIRVariable &var);
694
695 bool should_dereference(uint32_t id);
696 bool should_forward(uint32_t id) const;
697 bool should_suppress_usage_tracking(uint32_t id) const;
698 void emit_mix_op(uint32_t result_type, uint32_t id, uint32_t left, uint32_t right, uint32_t lerp);
699 void emit_nminmax_op(uint32_t result_type, uint32_t id, uint32_t op0, uint32_t op1, GLSLstd450 op);
700 void emit_emulated_ahyper_op(uint32_t result_type, uint32_t result_id, uint32_t op0, GLSLstd450 op);
701 bool to_trivial_mix_op(const SPIRType &type, std::string &op, uint32_t left, uint32_t right, uint32_t lerp);
702 void emit_quaternary_func_op(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, uint32_t op2,
703 uint32_t op3, const char *op);
704 void emit_trinary_func_op(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, uint32_t op2,
705 const char *op);
706 void emit_binary_func_op(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, const char *op);
707 void emit_atomic_func_op(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, const char *op);
708 void emit_atomic_func_op(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, uint32_t op2, const char *op);
709
710 void emit_unary_func_op_cast(uint32_t result_type, uint32_t result_id, uint32_t op0, const char *op,
711 SPIRType::BaseType input_type, SPIRType::BaseType expected_result_type);
712 void emit_binary_func_op_cast(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, const char *op,
713 SPIRType::BaseType input_type, bool skip_cast_if_equal_type);
714 void emit_binary_func_op_cast_clustered(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1,
715 const char *op, SPIRType::BaseType input_type);
716 void emit_trinary_func_op_cast(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, uint32_t op2,
717 const char *op, SPIRType::BaseType input_type);
718 void emit_trinary_func_op_bitextract(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1,
719 uint32_t op2, const char *op, SPIRType::BaseType expected_result_type,
720 SPIRType::BaseType input_type0, SPIRType::BaseType input_type1,
721 SPIRType::BaseType input_type2);
722 void emit_bitfield_insert_op(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, uint32_t op2,
723 uint32_t op3, const char *op, SPIRType::BaseType offset_count_type);
724
725 void emit_unary_func_op(uint32_t result_type, uint32_t result_id, uint32_t op0, const char *op);
726 void emit_unrolled_unary_op(uint32_t result_type, uint32_t result_id, uint32_t operand, const char *op);
727 void emit_binary_op(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, const char *op);
728 void emit_unrolled_binary_op(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, const char *op,
729 bool negate, SPIRType::BaseType expected_type);
730 void emit_binary_op_cast(uint32_t result_type, uint32_t result_id, uint32_t op0, uint32_t op1, const char *op,
731 SPIRType::BaseType input_type, bool skip_cast_if_equal_type, bool implicit_integer_promotion);
732
733 SPIRType binary_op_bitcast_helper(std::string &cast_op0, std::string &cast_op1, SPIRType::BaseType &input_type,
734 uint32_t op0, uint32_t op1, bool skip_cast_if_equal_type);
735
736 virtual bool emit_complex_bitcast(uint32_t result_type, uint32_t id, uint32_t op0);
737
738 std::string to_ternary_expression(const SPIRType &result_type, uint32_t select, uint32_t true_value,
739 uint32_t false_value);
740
741 void emit_unary_op(uint32_t result_type, uint32_t result_id, uint32_t op0, const char *op);
742 void emit_unary_op_cast(uint32_t result_type, uint32_t result_id, uint32_t op0, const char *op);
743 virtual void emit_mesh_tasks(SPIRBlock &block);
744 bool expression_is_forwarded(uint32_t id) const;
745 bool expression_suppresses_usage_tracking(uint32_t id) const;
746 bool expression_read_implies_multiple_reads(uint32_t id) const;
747 SPIRExpression &emit_op(uint32_t result_type, uint32_t result_id, const std::string &rhs, bool forward_rhs,
748 bool suppress_usage_tracking = false);
749
750 void access_chain_internal_append_index(std::string &expr, uint32_t base, const SPIRType *type,
751 AccessChainFlags flags, bool &access_chain_is_arrayed, uint32_t index);
752
753 std::string access_chain_internal(uint32_t base, const uint32_t *indices, uint32_t count, AccessChainFlags flags,
754 AccessChainMeta *meta);
755
756 spv::StorageClass get_expression_effective_storage_class(uint32_t ptr);
757 virtual bool access_chain_needs_stage_io_builtin_translation(uint32_t base);
758
759 virtual void check_physical_type_cast(std::string &expr, const SPIRType *type, uint32_t physical_type);
760 virtual bool prepare_access_chain_for_scalar_access(std::string &expr, const SPIRType &type,
761 spv::StorageClass storage, bool &is_packed);
762
763 std::string access_chain(uint32_t base, const uint32_t *indices, uint32_t count, const SPIRType &target_type,
764 AccessChainMeta *meta = nullptr, bool ptr_chain = false);
765
766 std::string flattened_access_chain(uint32_t base, const uint32_t *indices, uint32_t count,
767 const SPIRType &target_type, uint32_t offset, uint32_t matrix_stride,
768 uint32_t array_stride, bool need_transpose);
769 std::string flattened_access_chain_struct(uint32_t base, const uint32_t *indices, uint32_t count,
770 const SPIRType &target_type, uint32_t offset);
771 std::string flattened_access_chain_matrix(uint32_t base, const uint32_t *indices, uint32_t count,
772 const SPIRType &target_type, uint32_t offset, uint32_t matrix_stride,
773 bool need_transpose);
774 std::string flattened_access_chain_vector(uint32_t base, const uint32_t *indices, uint32_t count,
775 const SPIRType &target_type, uint32_t offset, uint32_t matrix_stride,
776 bool need_transpose);
777 std::pair<std::string, uint32_t> flattened_access_chain_offset(const SPIRType &basetype, const uint32_t *indices,
778 uint32_t count, uint32_t offset,
779 uint32_t word_stride, bool *need_transpose = nullptr,
780 uint32_t *matrix_stride = nullptr,
781 uint32_t *array_stride = nullptr,
782 bool ptr_chain = false);
783
784 const char *index_to_swizzle(uint32_t index);
785 std::string remap_swizzle(const SPIRType &result_type, uint32_t input_components, const std::string &expr);
786 std::string declare_temporary(uint32_t type, uint32_t id);
787 void emit_uninitialized_temporary(uint32_t type, uint32_t id);
788 SPIRExpression &emit_uninitialized_temporary_expression(uint32_t type, uint32_t id);
789 void append_global_func_args(const SPIRFunction &func, uint32_t index, SmallVector<std::string> &arglist);
790 std::string to_non_uniform_aware_expression(uint32_t id);
791 std::string to_expression(uint32_t id, bool register_expression_read = true);
792 std::string to_composite_constructor_expression(const SPIRType &parent_type, uint32_t id, bool block_like_type);
793 std::string to_rerolled_array_expression(const SPIRType &parent_type, const std::string &expr, const SPIRType &type);
794 std::string to_enclosed_expression(uint32_t id, bool register_expression_read = true);
795 std::string to_unpacked_expression(uint32_t id, bool register_expression_read = true);
796 std::string to_unpacked_row_major_matrix_expression(uint32_t id);
797 std::string to_enclosed_unpacked_expression(uint32_t id, bool register_expression_read = true);
798 std::string to_dereferenced_expression(uint32_t id, bool register_expression_read = true);
799 std::string to_pointer_expression(uint32_t id, bool register_expression_read = true);
800 std::string to_enclosed_pointer_expression(uint32_t id, bool register_expression_read = true);
801 std::string to_extract_component_expression(uint32_t id, uint32_t index);
802 std::string to_extract_constant_composite_expression(uint32_t result_type, const SPIRConstant &c,
803 const uint32_t *chain, uint32_t length);
804 static bool needs_enclose_expression(const std::string &expr);
805 std::string enclose_expression(const std::string &expr);
806 std::string dereference_expression(const SPIRType &expression_type, const std::string &expr);
807 std::string address_of_expression(const std::string &expr);
808 void strip_enclosed_expression(std::string &expr);
809 std::string to_member_name(const SPIRType &type, uint32_t index);
810 virtual std::string to_member_reference(uint32_t base, const SPIRType &type, uint32_t index, bool ptr_chain_is_resolved);
811 std::string to_multi_member_reference(const SPIRType &type, const SmallVector<uint32_t> &indices);
812 std::string type_to_glsl_constructor(const SPIRType &type);
813 std::string argument_decl(const SPIRFunction::Parameter &arg);
814 virtual std::string to_qualifiers_glsl(uint32_t id);
815 void fixup_io_block_patch_primitive_qualifiers(const SPIRVariable &var);
816 void emit_output_variable_initializer(const SPIRVariable &var);
817 std::string to_precision_qualifiers_glsl(uint32_t id);
818 virtual const char *to_storage_qualifiers_glsl(const SPIRVariable &var);
819 std::string flags_to_qualifiers_glsl(const SPIRType &type, const Bitset &flags);
820 const char *format_to_glsl(spv::ImageFormat format);
821 virtual std::string layout_for_member(const SPIRType &type, uint32_t index);
822 virtual std::string to_interpolation_qualifiers(const Bitset &flags);
823 std::string layout_for_variable(const SPIRVariable &variable);
824 std::string to_combined_image_sampler(VariableID image_id, VariableID samp_id);
825 virtual bool skip_argument(uint32_t id) const;
826 virtual bool emit_array_copy(const char *expr, uint32_t lhs_id, uint32_t rhs_id,
827 spv::StorageClass lhs_storage, spv::StorageClass rhs_storage);
828 virtual void emit_block_hints(const SPIRBlock &block);
829 virtual std::string to_initializer_expression(const SPIRVariable &var);
830 virtual std::string to_zero_initialized_expression(uint32_t type_id);
831 bool type_can_zero_initialize(const SPIRType &type) const;
832
833 bool buffer_is_packing_standard(const SPIRType &type, BufferPackingStandard packing,
834 uint32_t *failed_index = nullptr, uint32_t start_offset = 0,
835 uint32_t end_offset = ~(0u));
836 std::string buffer_to_packing_standard(const SPIRType &type,
837 bool support_std430_without_scalar_layout,
838 bool support_enhanced_layouts);
839
840 uint32_t type_to_packed_base_size(const SPIRType &type, BufferPackingStandard packing);
841 uint32_t type_to_packed_alignment(const SPIRType &type, const Bitset &flags, BufferPackingStandard packing);
842 uint32_t type_to_packed_array_stride(const SPIRType &type, const Bitset &flags, BufferPackingStandard packing);
843 uint32_t type_to_packed_size(const SPIRType &type, const Bitset &flags, BufferPackingStandard packing);
844 uint32_t type_to_location_count(const SPIRType &type) const;
845
846 std::string bitcast_glsl(const SPIRType &result_type, uint32_t arg);
847 virtual std::string bitcast_glsl_op(const SPIRType &result_type, const SPIRType &argument_type);
848
849 std::string bitcast_expression(SPIRType::BaseType target_type, uint32_t arg);
850 std::string bitcast_expression(const SPIRType &target_type, SPIRType::BaseType expr_type, const std::string &expr);
851
852 std::string build_composite_combiner(uint32_t result_type, const uint32_t *elems, uint32_t length);
853 bool remove_duplicate_swizzle(std::string &op);
854 bool remove_unity_swizzle(uint32_t base, std::string &op);
855
856 // Can modify flags to remote readonly/writeonly if image type
857 // and force recompile.
858 bool check_atomic_image(uint32_t id);
859
860 virtual void replace_illegal_names();
861 void replace_illegal_names(const std::unordered_set<std::string> &keywords);
862 virtual void emit_entry_point_declarations();
863
864 void replace_fragment_output(SPIRVariable &var);
865 void replace_fragment_outputs();
866 std::string legacy_tex_op(const std::string &op, const SPIRType &imgtype, uint32_t id);
867
868 void forward_relaxed_precision(uint32_t dst_id, const uint32_t *args, uint32_t length);
869 void analyze_precision_requirements(uint32_t type_id, uint32_t dst_id, uint32_t *args, uint32_t length);
870 Options::Precision analyze_expression_precision(const uint32_t *args, uint32_t length) const;
871
872 uint32_t indent = 0;
873
874 std::unordered_set<uint32_t> emitted_functions;
875
876 // Ensure that we declare phi-variable copies even if the original declaration isn't deferred
877 std::unordered_set<uint32_t> flushed_phi_variables;
878
879 std::unordered_set<uint32_t> flattened_buffer_blocks;
880 std::unordered_map<uint32_t, bool> flattened_structs;
881
882 ShaderSubgroupSupportHelper shader_subgroup_supporter;
883
884 std::string load_flattened_struct(const std::string &basename, const SPIRType &type);
885 std::string to_flattened_struct_member(const std::string &basename, const SPIRType &type, uint32_t index);
886 void store_flattened_struct(uint32_t lhs_id, uint32_t value);
887 void store_flattened_struct(const std::string &basename, uint32_t rhs, const SPIRType &type,
888 const SmallVector<uint32_t> &indices);
889 std::string to_flattened_access_chain_expression(uint32_t id);
890
891 // Usage tracking. If a temporary is used more than once, use the temporary instead to
892 // avoid AST explosion when SPIRV is generated with pure SSA and doesn't write stuff to variables.
893 std::unordered_map<uint32_t, uint32_t> expression_usage_counts;
894 void track_expression_read(uint32_t id);
895
896 SmallVector<std::string> forced_extensions;
897 SmallVector<std::string> header_lines;
898
899 // Used when expressions emit extra opcodes with their own unique IDs,
900 // and we need to reuse the IDs across recompilation loops.
901 // Currently used by NMin/Max/Clamp implementations.
902 std::unordered_map<uint32_t, uint32_t> extra_sub_expressions;
903
904 SmallVector<TypeID> workaround_ubo_load_overload_types;
905 void request_workaround_wrapper_overload(TypeID id);
906 void rewrite_load_for_wrapped_row_major(std::string &expr, TypeID loaded_type, ID ptr);
907
908 uint32_t statement_count = 0;
909
910 inline bool is_legacy() const
911 {
912 return (options.es && options.version < 300) || (!options.es && options.version < 130);
913 }
914
915 inline bool is_legacy_es() const
916 {
917 return options.es && options.version < 300;
918 }
919
920 inline bool is_legacy_desktop() const
921 {
922 return !options.es && options.version < 130;
923 }
924
925 enum Polyfill : uint32_t
926 {
927 PolyfillTranspose2x2 = 1 << 0,
928 PolyfillTranspose3x3 = 1 << 1,
929 PolyfillTranspose4x4 = 1 << 2,
930 PolyfillDeterminant2x2 = 1 << 3,
931 PolyfillDeterminant3x3 = 1 << 4,
932 PolyfillDeterminant4x4 = 1 << 5,
933 PolyfillMatrixInverse2x2 = 1 << 6,
934 PolyfillMatrixInverse3x3 = 1 << 7,
935 PolyfillMatrixInverse4x4 = 1 << 8,
936 PolyfillNMin16 = 1 << 9,
937 PolyfillNMin32 = 1 << 10,
938 PolyfillNMin64 = 1 << 11,
939 PolyfillNMax16 = 1 << 12,
940 PolyfillNMax32 = 1 << 13,
941 PolyfillNMax64 = 1 << 14,
942 PolyfillNClamp16 = 1 << 15,
943 PolyfillNClamp32 = 1 << 16,
944 PolyfillNClamp64 = 1 << 17,
945 };
946
947 uint32_t required_polyfills = 0;
948 uint32_t required_polyfills_relaxed = 0;
949 void require_polyfill(Polyfill polyfill, bool relaxed);
950
951 bool ray_tracing_is_khr = false;
952 bool barycentric_is_nv = false;
953 void ray_tracing_khr_fixup_locations();
954
955 bool args_will_forward(uint32_t id, const uint32_t *args, uint32_t num_args, bool pure);
956 void register_call_out_argument(uint32_t id);
957 void register_impure_function_call();
958 void register_control_dependent_expression(uint32_t expr);
959
960 // GL_EXT_shader_pixel_local_storage support.
961 std::vector<PlsRemap> pls_inputs;
962 std::vector<PlsRemap> pls_outputs;
963 std::string pls_decl(const PlsRemap &variable);
964 const char *to_pls_qualifiers_glsl(const SPIRVariable &variable);
965 void emit_pls();
966 void remap_pls_variables();
967
968 // GL_EXT_shader_framebuffer_fetch support.
969 std::vector<std::pair<uint32_t, uint32_t>> subpass_to_framebuffer_fetch_attachment;
970 std::vector<std::pair<uint32_t, bool>> inout_color_attachments;
971 bool location_is_framebuffer_fetch(uint32_t location) const;
972 bool location_is_non_coherent_framebuffer_fetch(uint32_t location) const;
973 bool subpass_input_is_framebuffer_fetch(uint32_t id) const;
974 void emit_inout_fragment_outputs_copy_to_subpass_inputs();
975 const SPIRVariable *find_subpass_input_by_attachment_index(uint32_t index) const;
976 const SPIRVariable *find_color_output_by_location(uint32_t location) const;
977
978 // A variant which takes two sets of name. The secondary is only used to verify there are no collisions,
979 // but the set is not updated when we have found a new name.
980 // Used primarily when adding block interface names.
981 void add_variable(std::unordered_set<std::string> &variables_primary,
982 const std::unordered_set<std::string> &variables_secondary, std::string &name);
983
984 void check_function_call_constraints(const uint32_t *args, uint32_t length);
985 void handle_invalid_expression(uint32_t id);
986 void force_temporary_and_recompile(uint32_t id);
987 void find_static_extensions();
988
989 uint32_t consume_temporary_in_precision_context(uint32_t type_id, uint32_t id, Options::Precision precision);
990 std::unordered_map<uint32_t, uint32_t> temporary_to_mirror_precision_alias;
991 std::unordered_set<uint32_t> composite_insert_overwritten;
992 std::unordered_set<uint32_t> block_composite_insert_overwrite;
993
994 std::string emit_for_loop_initializers(const SPIRBlock &block);
995 void emit_while_loop_initializers(const SPIRBlock &block);
996 bool for_loop_initializers_are_same_type(const SPIRBlock &block);
997 bool optimize_read_modify_write(const SPIRType &type, const std::string &lhs, const std::string &rhs);
998 void fixup_image_load_store_access();
999
1000 bool type_is_empty(const SPIRType &type);
1001
1002 bool can_use_io_location(spv::StorageClass storage, bool block);
1003 const Instruction *get_next_instruction_in_block(const Instruction &instr);
1004 static uint32_t mask_relevant_memory_semantics(uint32_t semantics);
1005
1006 std::string convert_half_to_string(const SPIRConstant &value, uint32_t col, uint32_t row);
1007 std::string convert_float_to_string(const SPIRConstant &value, uint32_t col, uint32_t row);
1008 std::string convert_double_to_string(const SPIRConstant &value, uint32_t col, uint32_t row);
1009
1010 std::string convert_separate_image_to_expression(uint32_t id);
1011
1012 // Builtins in GLSL are always specific signedness, but the SPIR-V can declare them
1013 // as either unsigned or signed.
1014 // Sometimes we will need to automatically perform casts on load and store to make this work.
1015 virtual SPIRType::BaseType get_builtin_basetype(spv::BuiltIn builtin, SPIRType::BaseType default_type);
1016 virtual void cast_to_variable_store(uint32_t target_id, std::string &expr, const SPIRType &expr_type);
1017 virtual void cast_from_variable_load(uint32_t source_id, std::string &expr, const SPIRType &expr_type);
1018 void unroll_array_from_complex_load(uint32_t target_id, uint32_t source_id, std::string &expr);
1019 bool unroll_array_to_complex_store(uint32_t target_id, uint32_t source_id);
1020 void convert_non_uniform_expression(std::string &expr, uint32_t ptr_id);
1021
1022 void handle_store_to_invariant_variable(uint32_t store_id, uint32_t value_id);
1023 void disallow_forwarding_in_expression_chain(const SPIRExpression &expr);
1024
1025 bool expression_is_constant_null(uint32_t id) const;
1026 bool expression_is_non_value_type_array(uint32_t ptr);
1027 virtual void emit_store_statement(uint32_t lhs_expression, uint32_t rhs_expression);
1028
1029 uint32_t get_integer_width_for_instruction(const Instruction &instr) const;
1030 uint32_t get_integer_width_for_glsl_instruction(GLSLstd450 op, const uint32_t *arguments, uint32_t length) const;
1031
1032 bool variable_is_lut(const SPIRVariable &var) const;
1033
1034 char current_locale_radix_character = '.';
1035
1036 void fixup_type_alias();
1037 void reorder_type_alias();
1038 void fixup_anonymous_struct_names();
1039 void fixup_anonymous_struct_names(std::unordered_set<uint32_t> &visited, const SPIRType &type);
1040
1041 static const char *vector_swizzle(int vecsize, int index);
1042
1043 bool is_stage_output_location_masked(uint32_t location, uint32_t component) const;
1044 bool is_stage_output_builtin_masked(spv::BuiltIn builtin) const;
1045 bool is_stage_output_variable_masked(const SPIRVariable &var) const;
1046 bool is_stage_output_block_member_masked(const SPIRVariable &var, uint32_t index, bool strip_array) const;
1047 bool is_per_primitive_variable(const SPIRVariable &var) const;
1048 uint32_t get_accumulated_member_location(const SPIRVariable &var, uint32_t mbr_idx, bool strip_array) const;
1049 uint32_t get_declared_member_location(const SPIRVariable &var, uint32_t mbr_idx, bool strip_array) const;
1050 std::unordered_set<LocationComponentPair, InternalHasher> masked_output_locations;
1051 std::unordered_set<uint32_t> masked_output_builtins;
1052
1053 FloatFormatter *float_formatter = nullptr;
1054 std::string format_float(float value) const;
1055 std::string format_double(double value) const;
1056
1057private:
1058 void init();
1059
1060 SmallVector<ConstantID> get_composite_constant_ids(ConstantID const_id);
1061 void fill_composite_constant(SPIRConstant &constant, TypeID type_id, const SmallVector<ConstantID> &initializers);
1062 void set_composite_constant(ConstantID const_id, TypeID type_id, const SmallVector<ConstantID> &initializers);
1063 TypeID get_composite_member_type(TypeID type_id, uint32_t member_idx);
1064 std::unordered_map<uint32_t, SmallVector<ConstantID>> const_composite_insert_ids;
1065};
1066} // namespace SPIRV_CROSS_NAMESPACE
1067
1068#endif
1069

source code of qtshadertools/src/3rdparty/SPIRV-Cross/spirv_glsl.hpp