Skip to content

[flang][OpenMP] simplify getReductionName #85666

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Mar 20, 2024
Merged

[flang][OpenMP] simplify getReductionName #85666

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
75 changes: 22 additions & 53 deletions flang/lib/Lower/OpenMP/ReductionProcessor.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -81,8 +81,10 @@ bool ReductionProcessor::supportedIntrinsicProcReduction(
return redType;
}

std::string ReductionProcessor::getReductionName(llvm::StringRef name,
mlir::Type ty, bool isByRef) {
std::string
ReductionProcessor::getReductionName(llvm::StringRef name,
const fir::KindMapping &kindMap,
mlir::Type ty, bool isByRef) {
ty = fir::unwrapRefType(ty);

// extra string to distinguish reduction functions for variables passed by
Expand All @@ -91,47 +93,12 @@ std::string ReductionProcessor::getReductionName(llvm::StringRef name,
if (isByRef)
byrefAddition = "_byref";

if (fir::isa_trivial(ty))
return (llvm::Twine(name) +
(ty.isIntOrIndex() ? llvm::Twine("_i_") : llvm::Twine("_f_")) +
llvm::Twine(ty.getIntOrFloatBitWidth()) + byrefAddition)
.str();

// creates a name like reduction_i_64_box_ux4x3
if (auto boxTy = mlir::dyn_cast_or_null<fir::BaseBoxType>(ty)) {
// TODO: support for allocatable boxes:
// !fir.box<!fir.heap<!fir.array<...>>>
fir::SequenceType seqTy = fir::unwrapRefType(boxTy.getEleTy())
.dyn_cast_or_null<fir::SequenceType>();
if (!seqTy)
return {};

std::string prefix = getReductionName(
name, fir::unwrapSeqOrBoxedSeqType(ty), /*isByRef=*/false);
if (prefix.empty())
return {};
std::stringstream tyStr;
tyStr << prefix << "_box_";
bool first = true;
for (std::int64_t extent : seqTy.getShape()) {
if (first)
first = false;
else
tyStr << "x";
if (extent == seqTy.getUnknownExtent())
tyStr << 'u'; // I'm not sure that '?' is safe in symbol names
else
tyStr << extent;
}
return (tyStr.str() + byrefAddition).str();
}

return {};
return fir::getTypeAsString(ty, kindMap, (name + byrefAddition).str());
}

std::string ReductionProcessor::getReductionName(
omp::clause::DefinedOperator::IntrinsicOperator intrinsicOp, mlir::Type ty,
bool isByRef) {
omp::clause::DefinedOperator::IntrinsicOperator intrinsicOp,
const fir::KindMapping &kindMap, mlir::Type ty, bool isByRef) {
std::string reductionName;

switch (intrinsicOp) {
Expand All @@ -154,17 +121,17 @@ std::string ReductionProcessor::getReductionName(
break;
}

return getReductionName(reductionName, ty, isByRef);
return getReductionName(reductionName, kindMap, ty, isByRef);
}

mlir::Value
ReductionProcessor::getReductionInitValue(mlir::Location loc, mlir::Type type,
ReductionIdentifier redId,
fir::FirOpBuilder &builder) {
type = fir::unwrapRefType(type);
assert((fir::isa_integer(type) || fir::isa_real(type) ||
type.isa<fir::LogicalType>()) &&
"only integer, logical and real types are currently supported");
if (!fir::isa_integer(type) && !fir::isa_real(type) &&
!mlir::isa<fir::LogicalType>(type))
TODO(loc, "Reduction of some types is not supported");
switch (redId) {
case ReductionIdentifier::MAX: {
if (auto ty = type.dyn_cast<mlir::FloatType>()) {
Expand Down Expand Up @@ -463,8 +430,7 @@ mlir::omp::DeclareReductionOp ReductionProcessor::createDeclareReduction(
mlir::OpBuilder::InsertionGuard guard(builder);
mlir::ModuleOp module = builder.getModule();

if (reductionOpName.empty())
TODO(loc, "Reduction of some types is not supported");
assert(!reductionOpName.empty());

auto decl =
module.lookupSymbol<mlir::omp::DeclareReductionOp>(reductionOpName);
Expand Down Expand Up @@ -601,15 +567,18 @@ void ReductionProcessor::addDeclareReduction(

for (mlir::Value symVal : reductionVars) {
auto redType = mlir::cast<fir::ReferenceType>(symVal.getType());
const auto &kindMap = firOpBuilder.getKindMap();
if (redType.getEleTy().isa<fir::LogicalType>())
decl = createDeclareReduction(
firOpBuilder,
getReductionName(intrinsicOp, firOpBuilder.getI1Type(), isByRef),
redId, redType, currentLocation, isByRef);
decl = createDeclareReduction(firOpBuilder,
getReductionName(intrinsicOp, kindMap,
firOpBuilder.getI1Type(),
isByRef),
redId, redType, currentLocation, isByRef);
else
decl = createDeclareReduction(
firOpBuilder, getReductionName(intrinsicOp, redType, isByRef),
redId, redType, currentLocation, isByRef);
firOpBuilder,
getReductionName(intrinsicOp, kindMap, redType, isByRef), redId,
redType, currentLocation, isByRef);
reductionDeclSymbols.push_back(mlir::SymbolRefAttr::get(
firOpBuilder.getContext(), decl.getSymName()));
}
Expand All @@ -631,7 +600,7 @@ void ReductionProcessor::addDeclareReduction(
decl = createDeclareReduction(
firOpBuilder,
getReductionName(getRealName(*reductionIntrinsic).ToString(),
redType, isByRef),
firOpBuilder.getKindMap(), redType, isByRef),
redId, redType, currentLocation, isByRef);
reductionDeclSymbols.push_back(mlir::SymbolRefAttr::get(
firOpBuilder.getContext(), decl.getSymName()));
Expand Down
8 changes: 5 additions & 3 deletions flang/lib/Lower/OpenMP/ReductionProcessor.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,12 +76,14 @@ class ReductionProcessor {
static bool
doReductionByRef(const llvm::SmallVectorImpl<mlir::Value> &reductionVars);

static std::string getReductionName(llvm::StringRef name, mlir::Type ty,
bool isByRef);
static std::string getReductionName(llvm::StringRef name,
const fir::KindMapping &kindMap,
mlir::Type ty, bool isByRef);

static std::string
getReductionName(omp::clause::DefinedOperator::IntrinsicOperator intrinsicOp,
mlir::Type ty, bool isByRef);
const fir::KindMapping &kindMap, mlir::Type ty,
bool isByRef);

/// This function returns the identity value of the operator \p
/// reductionOpName. For example:
Expand Down
22 changes: 11 additions & 11 deletions flang/test/Lower/OpenMP/FIR/wsloop-reduction-add-byref.f90
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
! RUN: %flang_fc1 -emit-fir -flang-deprecated-no-hlfir -fopenmp -mmlir --force-byref-reduction %s -o - | FileCheck %s
! NOTE: Assertions have been autogenerated by utils/generate-test-checks.py

! CHECK-LABEL: omp.declare_reduction @add_reduction_f_64_byref : !fir.ref<f64>
! CHECK-LABEL: omp.declare_reduction @add_reduction_byref_f64 : !fir.ref<f64>
! CHECK-SAME: init {
! CHECK: ^bb0(%[[VAL_0:.*]]: !fir.ref<f64>):
! CHECK: %[[C0_1:.*]] = arith.constant 0.000000e+00 : f64
Expand All @@ -19,7 +19,7 @@
! CHECK: omp.yield(%[[ARG0]] : !fir.ref<f64>)
! CHECK: }

! CHECK-LABEL: omp.declare_reduction @add_reduction_i_64_byref : !fir.ref<i64>
! CHECK-LABEL: omp.declare_reduction @add_reduction_byref_i64 : !fir.ref<i64>
! CHECK-SAME: init {
! CHECK: ^bb0(%[[VAL_0:.*]]: !fir.ref<i64>):
! CHECK: %[[C0_1:.*]] = arith.constant 0 : i64
Expand All @@ -36,7 +36,7 @@
! CHECK: omp.yield(%[[ARG0]] : !fir.ref<i64>)
! CHECK: }

! CHECK-LABEL: omp.declare_reduction @add_reduction_f_32_byref : !fir.ref<f32>
! CHECK-LABEL: omp.declare_reduction @add_reduction_byref_f32 : !fir.ref<f32>
! CHECK-SAME: init {
! CHECK: ^bb0(%[[VAL_0:.*]]: !fir.ref<f32>):
! CHECK: %[[C0_1:.*]] = arith.constant 0.000000e+00 : f32
Expand All @@ -53,7 +53,7 @@
! CHECK: omp.yield(%[[ARG0]] : !fir.ref<f32>)
! CHECK: }

! CHECK-LABEL: omp.declare_reduction @add_reduction_i_32_byref : !fir.ref<i32>
! CHECK-LABEL: omp.declare_reduction @add_reduction_byref_i32 : !fir.ref<i32>
! CHECK-SAME: init {
! CHECK: ^bb0(%[[VAL_0:.*]]: !fir.ref<i32>):
! CHECK: %[[C0_1:.*]] = arith.constant 0 : i32
Expand All @@ -80,7 +80,7 @@
! CHECK: %[[VAL_4:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_5:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_6:.*]] = arith.constant 1 : i32
! CHECK: omp.wsloop byref reduction(@add_reduction_i_32_byref %[[VAL_1]] -> %[[VAL_7:.*]] : !fir.ref<i32>) for (%[[VAL_8:.*]]) : i32 = (%[[VAL_4]]) to (%[[VAL_5]]) inclusive step (%[[VAL_6]]) {
! CHECK: omp.wsloop byref reduction(@add_reduction_byref_i32 %[[VAL_1]] -> %[[VAL_7:.*]] : !fir.ref<i32>) for (%[[VAL_8:.*]]) : i32 = (%[[VAL_4]]) to (%[[VAL_5]]) inclusive step (%[[VAL_6]]) {
! CHECK: fir.store %[[VAL_8]] to %[[VAL_3]] : !fir.ref<i32>
! CHECK: %[[VAL_9:.*]] = fir.load %[[VAL_7]] : !fir.ref<i32>
! CHECK: %[[VAL_10:.*]] = fir.load %[[VAL_3]] : !fir.ref<i32>
Expand Down Expand Up @@ -116,7 +116,7 @@ subroutine simple_int_reduction
! CHECK: %[[VAL_4:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_5:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_6:.*]] = arith.constant 1 : i32
! CHECK: omp.wsloop byref reduction(@add_reduction_f_32_byref %[[VAL_1]] -> %[[VAL_7:.*]] : !fir.ref<f32>) for (%[[VAL_8:.*]]) : i32 = (%[[VAL_4]]) to (%[[VAL_5]]) inclusive step (%[[VAL_6]]) {
! CHECK: omp.wsloop byref reduction(@add_reduction_byref_f32 %[[VAL_1]] -> %[[VAL_7:.*]] : !fir.ref<f32>) for (%[[VAL_8:.*]]) : i32 = (%[[VAL_4]]) to (%[[VAL_5]]) inclusive step (%[[VAL_6]]) {
! CHECK: fir.store %[[VAL_8]] to %[[VAL_3]] : !fir.ref<i32>
! CHECK: %[[VAL_9:.*]] = fir.load %[[VAL_7]] : !fir.ref<f32>
! CHECK: %[[VAL_10:.*]] = fir.load %[[VAL_3]] : !fir.ref<i32>
Expand Down Expand Up @@ -152,7 +152,7 @@ subroutine simple_real_reduction
! CHECK: %[[VAL_4:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_5:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_6:.*]] = arith.constant 1 : i32
! CHECK: omp.wsloop byref reduction(@add_reduction_i_32_byref %[[VAL_1]] -> %[[VAL_7:.*]] : !fir.ref<i32>) for (%[[VAL_8:.*]]) : i32 = (%[[VAL_4]]) to (%[[VAL_5]]) inclusive step (%[[VAL_6]]) {
! CHECK: omp.wsloop byref reduction(@add_reduction_byref_i32 %[[VAL_1]] -> %[[VAL_7:.*]] : !fir.ref<i32>) for (%[[VAL_8:.*]]) : i32 = (%[[VAL_4]]) to (%[[VAL_5]]) inclusive step (%[[VAL_6]]) {
! CHECK: fir.store %[[VAL_8]] to %[[VAL_3]] : !fir.ref<i32>
! CHECK: %[[VAL_9:.*]] = fir.load %[[VAL_3]] : !fir.ref<i32>
! CHECK: %[[VAL_10:.*]] = fir.load %[[VAL_7]] : !fir.ref<i32>
Expand Down Expand Up @@ -187,7 +187,7 @@ subroutine simple_int_reduction_switch_order
! CHECK: %[[VAL_4:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_5:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_6:.*]] = arith.constant 1 : i32
! CHECK: omp.wsloop byref reduction(@add_reduction_f_32_byref %[[VAL_1]] -> %[[VAL_7:.*]] : !fir.ref<f32>) for (%[[VAL_8:.*]]) : i32 = (%[[VAL_4]]) to (%[[VAL_5]]) inclusive step (%[[VAL_6]]) {
! CHECK: omp.wsloop byref reduction(@add_reduction_byref_f32 %[[VAL_1]] -> %[[VAL_7:.*]] : !fir.ref<f32>) for (%[[VAL_8:.*]]) : i32 = (%[[VAL_4]]) to (%[[VAL_5]]) inclusive step (%[[VAL_6]]) {
! CHECK: fir.store %[[VAL_8]] to %[[VAL_3]] : !fir.ref<i32>
! CHECK: %[[VAL_9:.*]] = fir.load %[[VAL_3]] : !fir.ref<i32>
! CHECK: %[[VAL_10:.*]] = fir.convert %[[VAL_9]] : (i32) -> f32
Expand Down Expand Up @@ -229,7 +229,7 @@ subroutine simple_real_reduction_switch_order
! CHECK: %[[VAL_8:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_9:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_10:.*]] = arith.constant 1 : i32
! CHECK: omp.wsloop byref reduction(@add_reduction_i_32_byref %[[VAL_1]] -> %[[VAL_11:.*]] : !fir.ref<i32>, @add_reduction_i_32_byref %[[VAL_2]] -> %[[VAL_12:.*]] : !fir.ref<i32>, @add_reduction_i_32_byref %[[VAL_3]] -> %[[VAL_13:.*]] : !fir.ref<i32>) for (%[[VAL_14:.*]]) : i32 = (%[[VAL_8]]) to (%[[VAL_9]]) inclusive step (%[[VAL_10]]) {
! CHECK: omp.wsloop byref reduction(@add_reduction_byref_i32 %[[VAL_1]] -> %[[VAL_11:.*]] : !fir.ref<i32>, @add_reduction_byref_i32 %[[VAL_2]] -> %[[VAL_12:.*]] : !fir.ref<i32>, @add_reduction_byref_i32 %[[VAL_3]] -> %[[VAL_13:.*]] : !fir.ref<i32>) for (%[[VAL_14:.*]]) : i32 = (%[[VAL_8]]) to (%[[VAL_9]]) inclusive step (%[[VAL_10]]) {
! CHECK: fir.store %[[VAL_14]] to %[[VAL_7]] : !fir.ref<i32>
! CHECK: %[[VAL_15:.*]] = fir.load %[[VAL_11]] : !fir.ref<i32>
! CHECK: %[[VAL_16:.*]] = fir.load %[[VAL_7]] : !fir.ref<i32>
Expand Down Expand Up @@ -282,7 +282,7 @@ subroutine multiple_int_reductions_same_type
! CHECK: %[[VAL_8:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_9:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_10:.*]] = arith.constant 1 : i32
! CHECK: omp.wsloop byref reduction(@add_reduction_f_32_byref %[[VAL_1]] -> %[[VAL_11:.*]] : !fir.ref<f32>, @add_reduction_f_32_byref %[[VAL_2]] -> %[[VAL_12:.*]] : !fir.ref<f32>, @add_reduction_f_32_byref %[[VAL_3]] -> %[[VAL_13:.*]] : !fir.ref<f32>) for (%[[VAL_14:.*]]) : i32 = (%[[VAL_8]]) to (%[[VAL_9]]) inclusive step (%[[VAL_10]]) {
! CHECK: omp.wsloop byref reduction(@add_reduction_byref_f32 %[[VAL_1]] -> %[[VAL_11:.*]] : !fir.ref<f32>, @add_reduction_byref_f32 %[[VAL_2]] -> %[[VAL_12:.*]] : !fir.ref<f32>, @add_reduction_byref_f32 %[[VAL_3]] -> %[[VAL_13:.*]] : !fir.ref<f32>) for (%[[VAL_14:.*]]) : i32 = (%[[VAL_8]]) to (%[[VAL_9]]) inclusive step (%[[VAL_10]]) {
! CHECK: fir.store %[[VAL_14]] to %[[VAL_7]] : !fir.ref<i32>
! CHECK: %[[VAL_15:.*]] = fir.load %[[VAL_11]] : !fir.ref<f32>
! CHECK: %[[VAL_16:.*]] = fir.load %[[VAL_7]] : !fir.ref<i32>
Expand Down Expand Up @@ -341,7 +341,7 @@ subroutine multiple_real_reductions_same_type
! CHECK: %[[VAL_10:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_11:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_12:.*]] = arith.constant 1 : i32
! CHECK: omp.wsloop byref reduction(@add_reduction_i_32_byref %[[VAL_2]] -> %[[VAL_13:.*]] : !fir.ref<i32>, @add_reduction_i_64_byref %[[VAL_3]] -> %[[VAL_14:.*]] : !fir.ref<i64>, @add_reduction_f_32_byref %[[VAL_4]] -> %[[VAL_15:.*]] : !fir.ref<f32>, @add_reduction_f_64_byref %[[VAL_1]] -> %[[VAL_16:.*]] : !fir.ref<f64>) for (%[[VAL_17:.*]]) : i32 = (%[[VAL_10]]) to (%[[VAL_11]]) inclusive step (%[[VAL_12]]) {
! CHECK: omp.wsloop byref reduction(@add_reduction_byref_i32 %[[VAL_2]] -> %[[VAL_13:.*]] : !fir.ref<i32>, @add_reduction_byref_i64 %[[VAL_3]] -> %[[VAL_14:.*]] : !fir.ref<i64>, @add_reduction_byref_f32 %[[VAL_4]] -> %[[VAL_15:.*]] : !fir.ref<f32>, @add_reduction_byref_f64 %[[VAL_1]] -> %[[VAL_16:.*]] : !fir.ref<f64>) for (%[[VAL_17:.*]]) : i32 = (%[[VAL_10]]) to (%[[VAL_11]]) inclusive step (%[[VAL_12]]) {
! CHECK: fir.store %[[VAL_17]] to %[[VAL_9]] : !fir.ref<i32>
! CHECK: %[[VAL_18:.*]] = fir.load %[[VAL_13]] : !fir.ref<i32>
! CHECK: %[[VAL_19:.*]] = fir.load %[[VAL_9]] : !fir.ref<i32>
Expand Down
Loading