AuroraMiddleware/SPIRV-Tools
1
0
Fork 0
mirror of https://github.com/KhronosGroup/SPIRV-Tools synced 2024-10-18 19:20:05 +00:00
Code Issues Packages Projects Releases Wiki Activity

Opt: Add constant folding for FToI and IToF

Browse Source
This commit is contained in:
GregF 2018-02-23 15:46:30 -07:00 committed by Steven Perron
parent 9457cabbce
commit bdaf8d56fb
2 changed files with 190 additions and 28 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

178
source/opt/const_folding_rules.cpp
View File

@ -131,7 +131,14 @@ ConstantFoldingRule FoldCompositeWithConstants() {
// The interface for a function that returns the result of applying a scalar
// floating-point binary operation on |a| and |b|. The type of the return value
// will be |type|. The input constants must also be of type |type|.
using FloatScalarFoldingRule = std::function<const analysis::Constant*(
using UnaryScalarFoldingRule = std::function<const analysis::Constant*(
const analysis::Type* result_type, const analysis::Constant* a,
analysis::ConstantManager*)>;
// The interface for a function that returns the result of applying a scalar
// floating-point binary operation on |a| and |b|. The type of the return value
// will be |type|. The input constants must also be of type |type|.
using BinaryScalarFoldingRule = std::function<const analysis::Constant*(
const analysis::Type* result_type, const analysis::Constant* a,
const analysis::Constant* b, analysis::ConstantManager*)>;
@ -158,12 +165,63 @@ std::vector<const analysis::Constant*> GetVectorComponents(
return components;
}
// Returns a |ConstantFoldingRule| that folds unary floating point scalar ops
// using |scalar_rule| and unary float point vectors ops by applying
// |scalar_rule| to the elements of the vector. The |ConstantFoldingRule|
// that is returned assumes that |constants| contains 1 entry. If they are
// not |nullptr|, then their type is either |Float| or |Integer| or a |Vector|
// whose element type is |Float| or |Integer|.
ConstantFoldingRule FoldFPUnaryOp(UnaryScalarFoldingRule scalar_rule) {
return [scalar_rule](ir::Instruction* inst,
const std::vector<const analysis::Constant*>& constants)
-> const analysis::Constant* {
ir::IRContext* context = inst->context();
analysis::ConstantManager* const_mgr = context->get_constant_mgr();
analysis::TypeManager* type_mgr = context->get_type_mgr();
const analysis::Type* result_type = type_mgr->GetType(inst->type_id());
const analysis::Vector* vector_type = result_type->AsVector();
if (!inst->IsFloatingPointFoldingAllowed()) {
return nullptr;
}
if (constants[0] == nullptr) {
return nullptr;
}
if (vector_type != nullptr) {
std::vector<const analysis::Constant*> a_components;
std::vector<const analysis::Constant*> results_components;
a_components = GetVectorComponents(constants[0], const_mgr);
// Fold each component of the vector.
for (uint32_t i = 0; i < a_components.size(); ++i) {
results_components.push_back(scalar_rule(vector_type->element_type(),
a_components[i], const_mgr));
if (results_components[i] == nullptr) {
return nullptr;
}
}
// Build the constant object and return it.
std::vector<uint32_t> ids;
for (const analysis::Constant* member : results_components) {
ids.push_back(const_mgr->GetDefiningInstruction(member)->result_id());
}
return const_mgr->GetConstant(vector_type, ids);
} else {
return scalar_rule(result_type, constants[0], const_mgr);
}
};
}
// Returns a |ConstantFoldingRule| that folds floating point scalars using
// |scalar_rule| and vectors of floating point by applying |scalar_rule| to the
// elements of the vector. The |ConstantFoldingRule| that is returned assumes
// that |constants| contains 2 entries. If they are not |nullptr|, then their
// type is either |Float| or a |Vector| whose element type is |Float|.
ConstantFoldingRule FoldFloatingPointOp(FloatScalarFoldingRule scalar_rule) {
ConstantFoldingRule FoldFPBinaryOp(BinaryScalarFoldingRule scalar_rule) {
return [scalar_rule](ir::Instruction* inst,
const std::vector<const analysis::Constant*>& constants)
-> const analysis::Constant* {
@ -211,7 +269,70 @@ ConstantFoldingRule FoldFloatingPointOp(FloatScalarFoldingRule scalar_rule) {
};
}
// This macro defines a |FloatScalarFoldingRule| that applies |op|. The
// This macro defines a |UnaryScalarFoldingRule| that performs float to
// integer conversion.
// TODO(greg-lunarg): Support for 64-bit integer types.
UnaryScalarFoldingRule FoldFToIOp() {
return [](const analysis::Type* result_type, const analysis::Constant* a,
analysis::ConstantManager* const_mgr) -> const analysis::Constant* {
assert(result_type != nullptr && a != nullptr);
const analysis::Integer* integer_type = result_type->AsInteger();
const analysis::Float* float_type = a->type()->AsFloat();
assert(float_type != nullptr);
assert(integer_type != nullptr);
if (integer_type->width() != 32) return nullptr;
if (float_type->width() == 32) {
float fa = a->GetFloat();
uint32_t result = integer_type->IsSigned()
? static_cast<uint32_t>(static_cast<int32_t>(fa))
: static_cast<uint32_t>(fa);
std::vector<uint32_t> words = {result};
return const_mgr->GetConstant(result_type, words);
} else if (float_type->width() == 64) {
double fa = a->GetDouble();
uint32_t result = integer_type->IsSigned()
? static_cast<uint32_t>(static_cast<int32_t>(fa))
: static_cast<uint32_t>(fa);
std::vector<uint32_t> words = {result};
return const_mgr->GetConstant(result_type, words);
}
return nullptr;
};
}
// This macro defines a |UnaryScalarFoldingRule| that performs integer to
// float conversion.
// TODO(greg-lunarg): Support for 64-bit integer types.
UnaryScalarFoldingRule FoldIToFOp() {
return [](const analysis::Type* result_type, const analysis::Constant* a,
analysis::ConstantManager* const_mgr) -> const analysis::Constant* {
assert(result_type != nullptr && a != nullptr);
const analysis::Integer* integer_type = a->type()->AsInteger();
const analysis::Float* float_type = result_type->AsFloat();
assert(float_type != nullptr);
assert(integer_type != nullptr);
if (integer_type->width() != 32) return nullptr;
uint32_t ua = a->GetU32();
if (float_type->width() == 32) {
float result_val = integer_type->IsSigned()
? static_cast<float>(static_cast<int32_t>(ua))
: static_cast<float>(ua);
spvutils::FloatProxy<float> result(result_val);
std::vector<uint32_t> words = {result.data()};
return const_mgr->GetConstant(result_type, words);
} else if (float_type->width() == 64) {
double result_val = integer_type->IsSigned()
? static_cast<double>(static_cast<int32_t>(ua))
: static_cast<double>(ua);
spvutils::FloatProxy<double> result(result_val);
std::vector<uint32_t> words = result.GetWords();
return const_mgr->GetConstant(result_type, words);
}
return nullptr;
};
}
// This macro defines a |BinaryScalarFoldingRule| that applies |op|. The
// operator |op| must work for both float and double, and use syntax "f1 op f2".
#define FOLD_FPARITH_OP(op) \
[](const analysis::Type* result_type, const analysis::Constant* a, \
@ -237,20 +358,16 @@ ConstantFoldingRule FoldFloatingPointOp(FloatScalarFoldingRule scalar_rule) {
return nullptr; \
}
// Define the folding rule for conversion between floating point and integer
ConstantFoldingRule FoldFToI() { return FoldFPUnaryOp(FoldFToIOp()); }
ConstantFoldingRule FoldIToF() { return FoldFPUnaryOp(FoldIToFOp()); }
// Define the folding rules for subtraction, addition, multiplication, and
// division for floating point values.
ConstantFoldingRule FoldFSub() {
return FoldFloatingPointOp(FOLD_FPARITH_OP(-));
}
ConstantFoldingRule FoldFAdd() {
return FoldFloatingPointOp(FOLD_FPARITH_OP(+));
}
ConstantFoldingRule FoldFMul() {
return FoldFloatingPointOp(FOLD_FPARITH_OP(*));
}
ConstantFoldingRule FoldFDiv() {
return FoldFloatingPointOp(FOLD_FPARITH_OP(/));
}
ConstantFoldingRule FoldFSub() { return FoldFPBinaryOp(FOLD_FPARITH_OP(-)); }
ConstantFoldingRule FoldFAdd() { return FoldFPBinaryOp(FOLD_FPARITH_OP(+)); }
ConstantFoldingRule FoldFMul() { return FoldFPBinaryOp(FOLD_FPARITH_OP(*)); }
ConstantFoldingRule FoldFDiv() { return FoldFPBinaryOp(FOLD_FPARITH_OP(/)); }
bool CompareFloatingPoint(bool op_result, bool op_unordered,
bool need_ordered) {
@ -263,7 +380,7 @@ bool CompareFloatingPoint(bool op_result, bool op_unordered,
}
}
// This macro defines a |FloatScalarFoldingRule| that applies |op|. The
// This macro defines a |BinaryScalarFoldingRule| that applies |op|. The
// operator |op| must work for both float and double, and use syntax "f1 op f2".
#define FOLD_FPCMP_OP(op, ord) \
[](const analysis::Type* result_type, const analysis::Constant* a, \
@ -295,40 +412,40 @@ bool CompareFloatingPoint(bool op_result, bool op_unordered,
// Define the folding rules for ordered and unordered comparison for floating
// point values.
ConstantFoldingRule FoldFOrdEqual() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(==, true));
return FoldFPBinaryOp(FOLD_FPCMP_OP(==, true));
}
ConstantFoldingRule FoldFUnordEqual() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(==, false));
return FoldFPBinaryOp(FOLD_FPCMP_OP(==, false));
}
ConstantFoldingRule FoldFOrdNotEqual() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(!=, true));
return FoldFPBinaryOp(FOLD_FPCMP_OP(!=, true));
}
ConstantFoldingRule FoldFUnordNotEqual() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(!=, false));
return FoldFPBinaryOp(FOLD_FPCMP_OP(!=, false));
}
ConstantFoldingRule FoldFOrdLessThan() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(<, true));
return FoldFPBinaryOp(FOLD_FPCMP_OP(<, true));
}
ConstantFoldingRule FoldFUnordLessThan() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(<, false));
return FoldFPBinaryOp(FOLD_FPCMP_OP(<, false));
}
ConstantFoldingRule FoldFOrdGreaterThan() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(>, true));
return FoldFPBinaryOp(FOLD_FPCMP_OP(>, true));
}
ConstantFoldingRule FoldFUnordGreaterThan() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(>, false));
return FoldFPBinaryOp(FOLD_FPCMP_OP(>, false));
}
ConstantFoldingRule FoldFOrdLessThanEqual() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(<=, true));
return FoldFPBinaryOp(FOLD_FPCMP_OP(<=, true));
}
ConstantFoldingRule FoldFUnordLessThanEqual() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(<=, false));
return FoldFPBinaryOp(FOLD_FPCMP_OP(<=, false));
}
ConstantFoldingRule FoldFOrdGreaterThanEqual() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(>=, true));
return FoldFPBinaryOp(FOLD_FPCMP_OP(>=, true));
}
ConstantFoldingRule FoldFUnordGreaterThanEqual() {
return FoldFloatingPointOp(FOLD_FPCMP_OP(>=, false));
return FoldFPBinaryOp(FOLD_FPCMP_OP(>=, false));
}
} // namespace
@ -342,6 +459,11 @@ spvtools::opt::ConstantFoldingRules::ConstantFoldingRules() {
rules_[SpvOpCompositeExtract].push_back(FoldExtractWithConstants());
rules_[SpvOpConvertFToS].push_back(FoldFToI());
rules_[SpvOpConvertFToU].push_back(FoldFToI());
rules_[SpvOpConvertSToF].push_back(FoldIToF());
rules_[SpvOpConvertUToF].push_back(FoldIToF());
rules_[SpvOpFAdd].push_back(FoldFAdd());
rules_[SpvOpFDiv].push_back(FoldFDiv());
rules_[SpvOpFMul].push_back(FoldFMul());

40
test/opt/fold_test.cpp
View File

@ -2756,6 +2756,46 @@ INSTANTIATE_TEST_CASE_P(DoubleVectorRedundantFoldingTest, GeneralInstructionFold
"OpFunctionEnd",
2, 3)
));
INSTANTIATE_TEST_CASE_P(FToIConstantFoldingTest, IntegerInstructionFoldingTest,
::testing::Values(
// Test case 0: Fold int(3.0)
InstructionFoldingCase<uint32_t>(
Header() + "%main = OpFunction %void None %void_func\n" +
"%main_lab = OpLabel\n" +
"%2 = OpConvertFToS %int %float_3\n" +
"OpReturn\n" +
"OpFunctionEnd",
2, 3),
// Test case 1: Fold uint(3.0)
InstructionFoldingCase<uint32_t>(
Header() + "%main = OpFunction %void None %void_func\n" +
"%main_lab = OpLabel\n" +
"%2 = OpConvertFToU %int %float_3\n" +
"OpReturn\n" +
"OpFunctionEnd",
2, 3)
));
INSTANTIATE_TEST_CASE_P(IToFConstantFoldingTest, FloatInstructionFoldingTest,
::testing::Values(
// Test case 0: Fold float(3)
InstructionFoldingCase<float>(
Header() + "%main = OpFunction %void None %void_func\n" +
"%main_lab = OpLabel\n" +
"%2 = OpConvertSToF %float %int_3\n" +
"OpReturn\n" +
"OpFunctionEnd",
2, 3.0),
// Test case 1: Fold float(3u)
InstructionFoldingCase<float>(
Header() + "%main = OpFunction %void None %void_func\n" +
"%main_lab = OpLabel\n" +
"%2 = OpConvertUToF %float %uint_3\n" +
"OpReturn\n" +
"OpFunctionEnd",
2, 3.0)
));
// clang-format on
using ToNegateFoldingTest =