[Flang][OpenMP][MLIR] Lowering of reduction,inreduction, nogroup and lastprivate clause to MLIR #166751
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
kaviya2510
requested review from
NimishMishra,
Thirumalai-Shaktivel,
kiranchandramohan,
kiranktp,
skatrak and
tblah
llvmbot
added
flang
…lastprivate clause to MLIR
Member
|
@llvm/pr-subscribers-flang-fir-hlfir Author: Kaviya Rajendiran (kaviya2510) ChangesThis patch add MLIR lowering support for nogroup, reduction, inreduction and lastprivate clauses of taskloop directive. Full diff: https://github.com/llvm/llvm-project/pull/166751.diff 9 Files Affected:
diff --git a/flang/lib/Lower/OpenMP/ClauseProcessor.cpp b/flang/lib/Lower/OpenMP/ClauseProcessor.cpp
index 1c163e6de7e5a..872f31fe45cca 100644
--- a/flang/lib/Lower/OpenMP/ClauseProcessor.cpp
+++ b/flang/lib/Lower/OpenMP/ClauseProcessor.cpp
@@ -406,6 +406,11 @@ bool ClauseProcessor::processMergeable(
return markClauseOccurrence<omp::clause::Mergeable>(result.mergeable);
}
+bool ClauseProcessor::processNogroup(
+ mlir::omp::NogroupClauseOps &result) const {
+ return markClauseOccurrence<omp::clause::Nogroup>(result.nogroup);
+}
+
bool ClauseProcessor::processNowait(mlir::omp::NowaitClauseOps &result) const {
return markClauseOccurrence<omp::clause::Nowait>(result.nowait);
}
diff --git a/flang/lib/Lower/OpenMP/ClauseProcessor.h b/flang/lib/Lower/OpenMP/ClauseProcessor.h
index 6452e39b97551..d524b4ddc8ac4 100644
--- a/flang/lib/Lower/OpenMP/ClauseProcessor.h
+++ b/flang/lib/Lower/OpenMP/ClauseProcessor.h
@@ -89,6 +89,7 @@ class ClauseProcessor {
bool processInclusive(mlir::Location currentLocation,
mlir::omp::InclusiveClauseOps &result) const;
bool processMergeable(mlir::omp::MergeableClauseOps &result) const;
+ bool processNogroup(mlir::omp::NogroupClauseOps &result) const;
bool processNowait(mlir::omp::NowaitClauseOps &result) const;
bool processNumTasks(lower::StatementContext &stmtCtx,
mlir::omp::NumTasksClauseOps &result) const;
diff --git a/flang/lib/Lower/OpenMP/DataSharingProcessor.cpp b/flang/lib/Lower/OpenMP/DataSharingProcessor.cpp
index 146a252b049ec..c5f7a0f710e8a 100644
--- a/flang/lib/Lower/OpenMP/DataSharingProcessor.cpp
+++ b/flang/lib/Lower/OpenMP/DataSharingProcessor.cpp
@@ -342,7 +342,8 @@ void DataSharingProcessor::insertLastPrivateCompare(mlir::Operation *op) {
if (!hasLastPrivate)
return;
- if (mlir::isa<mlir::omp::WsloopOp>(op) || mlir::isa<mlir::omp::SimdOp>(op)) {
+ if (mlir::isa<mlir::omp::WsloopOp>(op) || mlir::isa<mlir::omp::SimdOp>(op) ||
+ mlir::isa<mlir::omp::TaskloopOp>(op)) {
mlir::omp::LoopRelatedClauseOps result;
llvm::SmallVector<const semantics::Symbol *> iv;
collectLoopRelatedInfo(converter, converter.getCurrentLocation(), eval,
diff --git a/flang/lib/Lower/OpenMP/OpenMP.cpp b/flang/lib/Lower/OpenMP/OpenMP.cpp
index ad456d89bc432..cb006211cc208 100644
--- a/flang/lib/Lower/OpenMP/OpenMP.cpp
+++ b/flang/lib/Lower/OpenMP/OpenMP.cpp
@@ -1763,21 +1763,24 @@ static void genTaskgroupClauses(
cp.processTaskReduction(loc, clauseOps, taskReductionSyms);
}
-static void genTaskloopClauses(lower::AbstractConverter &converter,
- semantics::SemanticsContext &semaCtx,
- lower::StatementContext &stmtCtx,
- const List<Clause> &clauses, mlir::Location loc,
- mlir::omp::TaskloopOperands &clauseOps) {
+static void genTaskloopClauses(
+ lower::AbstractConverter &converter, semantics::SemanticsContext &semaCtx,
+ lower::StatementContext &stmtCtx, const List<Clause> &clauses,
+ mlir::Location loc, mlir::omp::TaskloopOperands &clauseOps,
+ llvm::SmallVectorImpl<const semantics::Symbol *> &reductionSyms,
+ llvm::SmallVectorImpl<const semantics::Symbol *> &inReductionSyms) {
ClauseProcessor cp(converter, semaCtx, clauses);
cp.processGrainsize(stmtCtx, clauseOps);
cp.processNumTasks(stmtCtx, clauseOps);
+ cp.processNogroup(clauseOps);
+ cp.processReduction(loc, clauseOps, reductionSyms);
+ cp.processInReduction(loc, clauseOps, inReductionSyms);
cp.processTODO<clause::Allocate, clause::Collapse, clause::Default,
- clause::Final, clause::If, clause::InReduction,
- clause::Lastprivate, clause::Mergeable, clause::Nogroup,
- clause::Priority, clause::Reduction, clause::Shared,
- clause::Untied>(loc, llvm::omp::Directive::OMPD_taskloop);
+ clause::Final, clause::If, clause::Mergeable, clause::Priority,
+ clause::Shared, clause::Untied>(
+ loc, llvm::omp::Directive::OMPD_taskloop);
}
static void genTaskwaitClauses(lower::AbstractConverter &converter,
@@ -2979,8 +2982,11 @@ static mlir::omp::TaskloopOp genStandaloneTaskloop(
lower::pft::Evaluation &eval, mlir::Location loc,
const ConstructQueue &queue, ConstructQueue::const_iterator item) {
mlir::omp::TaskloopOperands taskloopClauseOps;
+ llvm::SmallVector<const semantics::Symbol *> reductionSyms;
+ llvm::SmallVector<const semantics::Symbol *> inReductionSyms;
+
genTaskloopClauses(converter, semaCtx, stmtCtx, item->clauses, loc,
- taskloopClauseOps);
+ taskloopClauseOps, reductionSyms, inReductionSyms);
DataSharingProcessor dsp(converter, semaCtx, item->clauses, eval,
/*shouldCollectPreDeterminedSymbols=*/true,
enableDelayedPrivatization, symTable);
@@ -2994,6 +3000,10 @@ static mlir::omp::TaskloopOp genStandaloneTaskloop(
EntryBlockArgs taskloopArgs;
taskloopArgs.priv.syms = dsp.getDelayedPrivSymbols();
taskloopArgs.priv.vars = taskloopClauseOps.privateVars;
+ taskloopArgs.reduction.syms = reductionSyms;
+ taskloopArgs.reduction.vars = taskloopClauseOps.reductionVars;
+ taskloopArgs.inReduction.syms = inReductionSyms;
+ taskloopArgs.inReduction.vars = taskloopClauseOps.inReductionVars;
auto taskLoopOp = genWrapperOp<mlir::omp::TaskloopOp>(
converter, loc, taskloopClauseOps, taskloopArgs);
diff --git a/flang/test/Lower/OpenMP/Todo/taskloop-inreduction.f90 b/flang/test/Lower/OpenMP/Todo/taskloop-inreduction.f90
deleted file mode 100644
index 8acc399a92abe..0000000000000
--- a/flang/test/Lower/OpenMP/Todo/taskloop-inreduction.f90
+++ /dev/null
@@ -1,13 +0,0 @@
-! RUN: %not_todo_cmd bbc -emit-fir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
-! RUN: %not_todo_cmd %flang_fc1 -emit-fir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
-
-! CHECK: not yet implemented: Unhandled clause IN_REDUCTION in TASKLOOP construct
-subroutine omp_taskloop_inreduction()
- integer x
- x = 0
- !$omp taskloop in_reduction(+:x)
- do i = 1, 100
- x = x + 1
- end do
- !$omp end taskloop
-end subroutine omp_taskloop_inreduction
diff --git a/flang/test/Lower/OpenMP/Todo/taskloop-reduction.f90 b/flang/test/Lower/OpenMP/Todo/taskloop-reduction.f90
deleted file mode 100644
index 0c16bd227257f..0000000000000
--- a/flang/test/Lower/OpenMP/Todo/taskloop-reduction.f90
+++ /dev/null
@@ -1,13 +0,0 @@
-! RUN: %not_todo_cmd bbc -emit-fir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
-! RUN: %not_todo_cmd %flang_fc1 -emit-fir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
-
-! CHECK: not yet implemented: Unhandled clause REDUCTION in TASKLOOP construct
-subroutine omp_taskloop_reduction()
- integer x
- x = 0
- !$omp taskloop reduction(+:x)
- do i = 1, 100
- x = x + 1
- end do
- !$omp end taskloop
-end subroutine omp_taskloop_reduction
diff --git a/flang/test/Lower/OpenMP/taskloop.f90 b/flang/test/Lower/OpenMP/taskloop.f90
index 79b0c20e176c0..5f20b0bbd60b9 100644
--- a/flang/test/Lower/OpenMP/taskloop.f90
+++ b/flang/test/Lower/OpenMP/taskloop.f90
@@ -1,6 +1,15 @@
! RUN: bbc -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
! RUN: %flang_fc1 -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
+! CHECK-LABEL: omp.private
+! CHECK-SAME: {type = private} @[[LAST_PRIVATE_I:.*]] : i32
+
+! CHECK-LABEL: omp.private
+! CHECK-SAME: {type = private} @[[LAST_PRIVATE_X:.*]] : i32
+
+! CHECK-LABEL: omp.private
+! CHECK-SAME: {type = private} @[[QFOMP_TASKLOOP_NOGROUPEI_PRIVATE_I32:.*]] : i32
+
! CHECK-LABEL: omp.private
! CHECK-SAME: {type = private} @[[I_PRIVATE_TEST2:.*]] : i32
@@ -70,3 +79,46 @@ subroutine omp_taskloop_private
! CHECK: }
!$omp end taskloop
end subroutine omp_taskloop_private
+
+!===============================================================================
+! `nogroup` clause
+!===============================================================================
+
+subroutine omp_taskloop_nogroup()
+ ! CHECK: omp.taskloop nogroup
+ !$omp taskloop nogroup
+ do i = 1, 10
+ call foo()
+ end do
+ !$omp end taskloop
+end subroutine
+
+!===============================================================================
+! `lastprivate` clause
+!===============================================================================
+
+! CHECK-LABEL: func.func @_QPomp_taskloop_lastprivate
+! CHECK: %[[ALLOCA_I:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFomp_taskloop_lastprivateEi"}
+! CHECK: %[[DECL_I:.*]]:2 = hlfir.declare %[[ALLOCA_I]] {uniq_name = "_QFomp_taskloop_lastprivateEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+! CHECK: %[[ALLOCA_X:.*]] = fir.alloca i32 {bindc_name = "x", uniq_name = "_QFomp_taskloop_lastprivateEx"}
+! CHECK: %[[DECL_X:.*]]:2 = hlfir.declare %[[ALLOCA_X]] {uniq_name = "_QFomp_taskloop_lastprivateEx"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+subroutine omp_taskloop_lastprivate()
+ integer x
+ x = 0
+ ! CHECK: omp.taskloop private(@[[LAST_PRIVATE_X]] %[[DECL_X]]#0 -> %[[ARG0]], @[[LAST_PRIVATE_I]] %[[DECL_I]]#0 -> %[[ARG1]] : !fir.ref<i32>, !fir.ref<i32>) {
+ !$omp taskloop lastprivate(x)
+ do i = 1, 100
+ ! CHECK: %[[DECL_ARG0:.*]]:2 = hlfir.declare %[[ARG0]] {uniq_name = "_QFomp_taskloop_lastprivateEx"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+ ! CHECK: %[[LOAD_ARG0:.*]] = fir.load %[[DECL_ARG0]]#0 : !fir.ref<i32>
+ ! CHECK: %[[RES_ADD:.*]] = arith.addi %[[LOAD_ARG0]], %{{.*}} : i32
+ ! CHECK: hlfir.assign %[[RES_ADD]] to %[[DECL_ARG0]]#0 : i32, !fir.ref<i32>
+ x = x + 1
+ ! CHECK: %[[SELCT_RESULT:.*]] = arith.select %{{.*}}, %{{.*}}, %{{.*}} : i1
+ ! CHECK: fir.if %[[SELCT_RESULT]] {
+ ! CHECK: %[[LOADED_SUM:.*]] = fir.load %[[DECL_ARG0]]#0 : !fir.ref<i32>
+ ! CHECK: hlfir.assign %[[LOADED_SUM]] to %[[DECL_X]]#0 : i32, !fir.ref<i32>
+ ! CHECK: }
+ ! CHECK: omp.yield
+ end do
+ !$omp end taskloop
+end subroutine omp_taskloop_lastprivate
\ No newline at end of file
diff --git a/flang/test/Lower/taskloop-inreduction.f90 b/flang/test/Lower/taskloop-inreduction.f90
new file mode 100644
index 0000000000000..9973382cb60ee
--- /dev/null
+++ b/flang/test/Lower/taskloop-inreduction.f90
@@ -0,0 +1,40 @@
+! RUN: bbc -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
+! RUN: %flang_fc1 -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
+
+! CHECK-LABEL: omp.private
+! CHECK-SAME: {type = private} @[[PRIVATE_I:.*]] : i32
+
+! CHECK-LABEL: omp.declare_reduction
+! CHECK-SAME: @[[ADD_RED_I32:.*]] : i32 init {
+! CHECK: ^bb0(%{{.*}}: i32):
+! CHECK: %[[C0_I32:.*]] = arith.constant 0 : i32
+! CHECK: omp.yield(%[[C0_I32]] : i32)
+! CHECK: } combiner {
+! CHECK: ^bb0(%{{.*}}: i32, %{{.*}}: i32):
+! CHECK: %[[RES:.*]] = arith.addi %{{.*}}, %{{.*}} : i32
+! CHECK: omp.yield(%[[RES]] : i32)
+! CHECK: }
+
+! CHECK-LABEL: func.func @_QPomp_taskloop_inreduction
+! CHECK: %[[ALLOCA_I:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFomp_taskloop_inreductionEi"}
+! CHECK: %[[DECL_I:.*]]:2 = hlfir.declare %[[ALLOCA_I]] {uniq_name = "_QFomp_taskloop_inreductionEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+! CHECK: %[[ALLOCA_X:.*]] = fir.alloca i32 {bindc_name = "x", uniq_name = "_QFomp_taskloop_inreductionEx"}
+! CHECK: %[[DECL_X:.*]]:2 = hlfir.declare %[[ALLOCA_X]] {uniq_name = "_QFomp_taskloop_inreductionEx"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+! CHECK: %[[INIT_X:.*]] = arith.constant 0 : i32
+! CHECK: hlfir.assign %[[INIT_X]] to %[[DECL_X]]#0 : i32, !fir.ref<i32>
+subroutine omp_taskloop_inreduction()
+ integer x
+ x = 0
+ ! CHECK: omp.taskloop in_reduction(@[[ADD_RED_I32]]
+ ! CHECK: %[[DECL_X]]#0 -> %[[ARG0:.*]] : !fir.ref<i32>) private(@[[PRIVATE_I]] %[[DECL_I]]#0 -> %[[ARG1:.*]] : !fir.ref<i32>) {
+ ! CHECK: %[[VAL_ARG1:.*]]:2 = hlfir.declare %[[ARG0]]
+ ! CHECK-SAME: {uniq_name = "_QFomp_taskloop_inreductionEx"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+ !$omp taskloop in_reduction(+:x)
+ do i = 1, 100
+ ! CHECK: %[[X_VAL:.*]] = fir.load %[[VAL_ARG1]]#0 : !fir.ref<i32>
+ ! CHECK: %[[ADD_VAL:.*]] = arith.addi %[[X_VAL]], %{{.*}} : i32
+ x = x + 1
+ ! CHECK: hlfir.assign %[[ADD_VAL]] to %[[VAL_ARG1]]#0 : i32, !fir.ref<i32>
+ end do
+ !$omp end taskloop
+end subroutine omp_taskloop_inreduction
\ No newline at end of file
diff --git a/flang/test/Lower/taskloop-reduction.f90 b/flang/test/Lower/taskloop-reduction.f90
new file mode 100644
index 0000000000000..bf7511c4c1abc
--- /dev/null
+++ b/flang/test/Lower/taskloop-reduction.f90
@@ -0,0 +1,39 @@
+! RUN: bbc -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
+! RUN: %flang_fc1 -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
+
+! CHECK-LABEL: omp.private
+! CHECK-SAME: {type = private} @[[PRIVATE_I:.*]] : i32
+
+! CHECK-LABEL: omp.declare_reduction
+! CHECK-SAME: @[[ADD_RED_I32:.*]] : i32 init {
+! CHECK: ^bb0(%{{.*}}: i32):
+! CHECK: %[[C0_I32:.*]] = arith.constant 0 : i32
+! CHECK: omp.yield(%[[C0_I32]] : i32)
+! CHECK: } combiner {
+! CHECK: ^bb0(%{{.*}}: i32, %{{.*}}: i32):
+! CHECK: %[[RES:.*]] = arith.addi %{{.*}}, %{{.*}} : i32
+! CHECK: omp.yield(%[[RES]] : i32)
+! CHECK: }
+
+! CHECK-LABEL: func.func @_QPomp_taskloop_reduction
+! CHECK: %[[ALLOCA_I:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFomp_taskloop_reductionEi"}
+! CHECK: %[[DECL_I:.*]]:2 = hlfir.declare %[[ALLOCA_I]] {uniq_name = "_QFomp_taskloop_reductionEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+! CHECK: %[[ALLOCA_X:.*]] = fir.alloca i32 {bindc_name = "x", uniq_name = "_QFomp_taskloop_reductionEx"}
+! CHECK: %[[DECL_X:.*]]:2 = hlfir.declare %[[ALLOCA_X]] {uniq_name = "_QFomp_taskloop_reductionEx"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+! CHECK: %[[INIT_X:.*]] = arith.constant 0 : i32
+! CHECK: hlfir.assign %[[INIT_X]] to %[[DECL_X]]#0 : i32, !fir.ref<i32>
+subroutine omp_taskloop_reduction()
+ integer x
+ x = 0
+ ! CHECK: omp.taskloop private(@[[PRIVATE_I]]
+ ! CHECK-SAME: %[[DECL_I]]#0 -> %[[ARG0:.*]] : !fir.ref<i32>) reduction(@[[ADD_RED_I32]] %[[DECL_X]]#0 -> %[[ARG1:.*]] : !fir.ref<i32>) {
+ ! CHECK: %[[VAL_ARG1:.*]]:2 = hlfir.declare %[[ARG1]]
+ !$omp taskloop reduction(+:x)
+ do i = 1, 100
+ ! CHECK: %[[X_VAL:.*]] = fir.load %[[VAL_ARG1]]#0 : !fir.ref<i32>
+ ! CHECK: %[[ADD_VAL:.*]] = arith.addi %[[X_VAL]], %{{.*}} : i32
+ x = x + 1
+ ! CHECK: hlfir.assign %[[ADD_VAL]] to %[[VAL_ARG1]]#0 : i32, !fir.ref<i32>
+ end do
+ !$omp end taskloop
+end subroutine omp_taskloop_reduction
\ No newline at end of file
|
Member
|
@llvm/pr-subscribers-flang-openmp Author: Kaviya Rajendiran (kaviya2510) ChangesThis patch add MLIR lowering support for nogroup, reduction, inreduction and lastprivate clauses of taskloop directive. Full diff: https://github.com/llvm/llvm-project/pull/166751.diff 9 Files Affected:
diff --git a/flang/lib/Lower/OpenMP/ClauseProcessor.cpp b/flang/lib/Lower/OpenMP/ClauseProcessor.cpp
index 1c163e6de7e5a..872f31fe45cca 100644
--- a/flang/lib/Lower/OpenMP/ClauseProcessor.cpp
+++ b/flang/lib/Lower/OpenMP/ClauseProcessor.cpp
@@ -406,6 +406,11 @@ bool ClauseProcessor::processMergeable(
return markClauseOccurrence<omp::clause::Mergeable>(result.mergeable);
}
+bool ClauseProcessor::processNogroup(
+ mlir::omp::NogroupClauseOps &result) const {
+ return markClauseOccurrence<omp::clause::Nogroup>(result.nogroup);
+}
+
bool ClauseProcessor::processNowait(mlir::omp::NowaitClauseOps &result) const {
return markClauseOccurrence<omp::clause::Nowait>(result.nowait);
}
diff --git a/flang/lib/Lower/OpenMP/ClauseProcessor.h b/flang/lib/Lower/OpenMP/ClauseProcessor.h
index 6452e39b97551..d524b4ddc8ac4 100644
--- a/flang/lib/Lower/OpenMP/ClauseProcessor.h
+++ b/flang/lib/Lower/OpenMP/ClauseProcessor.h
@@ -89,6 +89,7 @@ class ClauseProcessor {
bool processInclusive(mlir::Location currentLocation,
mlir::omp::InclusiveClauseOps &result) const;
bool processMergeable(mlir::omp::MergeableClauseOps &result) const;
+ bool processNogroup(mlir::omp::NogroupClauseOps &result) const;
bool processNowait(mlir::omp::NowaitClauseOps &result) const;
bool processNumTasks(lower::StatementContext &stmtCtx,
mlir::omp::NumTasksClauseOps &result) const;
diff --git a/flang/lib/Lower/OpenMP/DataSharingProcessor.cpp b/flang/lib/Lower/OpenMP/DataSharingProcessor.cpp
index 146a252b049ec..c5f7a0f710e8a 100644
--- a/flang/lib/Lower/OpenMP/DataSharingProcessor.cpp
+++ b/flang/lib/Lower/OpenMP/DataSharingProcessor.cpp
@@ -342,7 +342,8 @@ void DataSharingProcessor::insertLastPrivateCompare(mlir::Operation *op) {
if (!hasLastPrivate)
return;
- if (mlir::isa<mlir::omp::WsloopOp>(op) || mlir::isa<mlir::omp::SimdOp>(op)) {
+ if (mlir::isa<mlir::omp::WsloopOp>(op) || mlir::isa<mlir::omp::SimdOp>(op) ||
+ mlir::isa<mlir::omp::TaskloopOp>(op)) {
mlir::omp::LoopRelatedClauseOps result;
llvm::SmallVector<const semantics::Symbol *> iv;
collectLoopRelatedInfo(converter, converter.getCurrentLocation(), eval,
diff --git a/flang/lib/Lower/OpenMP/OpenMP.cpp b/flang/lib/Lower/OpenMP/OpenMP.cpp
index ad456d89bc432..cb006211cc208 100644
--- a/flang/lib/Lower/OpenMP/OpenMP.cpp
+++ b/flang/lib/Lower/OpenMP/OpenMP.cpp
@@ -1763,21 +1763,24 @@ static void genTaskgroupClauses(
cp.processTaskReduction(loc, clauseOps, taskReductionSyms);
}
-static void genTaskloopClauses(lower::AbstractConverter &converter,
- semantics::SemanticsContext &semaCtx,
- lower::StatementContext &stmtCtx,
- const List<Clause> &clauses, mlir::Location loc,
- mlir::omp::TaskloopOperands &clauseOps) {
+static void genTaskloopClauses(
+ lower::AbstractConverter &converter, semantics::SemanticsContext &semaCtx,
+ lower::StatementContext &stmtCtx, const List<Clause> &clauses,
+ mlir::Location loc, mlir::omp::TaskloopOperands &clauseOps,
+ llvm::SmallVectorImpl<const semantics::Symbol *> &reductionSyms,
+ llvm::SmallVectorImpl<const semantics::Symbol *> &inReductionSyms) {
ClauseProcessor cp(converter, semaCtx, clauses);
cp.processGrainsize(stmtCtx, clauseOps);
cp.processNumTasks(stmtCtx, clauseOps);
+ cp.processNogroup(clauseOps);
+ cp.processReduction(loc, clauseOps, reductionSyms);
+ cp.processInReduction(loc, clauseOps, inReductionSyms);
cp.processTODO<clause::Allocate, clause::Collapse, clause::Default,
- clause::Final, clause::If, clause::InReduction,
- clause::Lastprivate, clause::Mergeable, clause::Nogroup,
- clause::Priority, clause::Reduction, clause::Shared,
- clause::Untied>(loc, llvm::omp::Directive::OMPD_taskloop);
+ clause::Final, clause::If, clause::Mergeable, clause::Priority,
+ clause::Shared, clause::Untied>(
+ loc, llvm::omp::Directive::OMPD_taskloop);
}
static void genTaskwaitClauses(lower::AbstractConverter &converter,
@@ -2979,8 +2982,11 @@ static mlir::omp::TaskloopOp genStandaloneTaskloop(
lower::pft::Evaluation &eval, mlir::Location loc,
const ConstructQueue &queue, ConstructQueue::const_iterator item) {
mlir::omp::TaskloopOperands taskloopClauseOps;
+ llvm::SmallVector<const semantics::Symbol *> reductionSyms;
+ llvm::SmallVector<const semantics::Symbol *> inReductionSyms;
+
genTaskloopClauses(converter, semaCtx, stmtCtx, item->clauses, loc,
- taskloopClauseOps);
+ taskloopClauseOps, reductionSyms, inReductionSyms);
DataSharingProcessor dsp(converter, semaCtx, item->clauses, eval,
/*shouldCollectPreDeterminedSymbols=*/true,
enableDelayedPrivatization, symTable);
@@ -2994,6 +3000,10 @@ static mlir::omp::TaskloopOp genStandaloneTaskloop(
EntryBlockArgs taskloopArgs;
taskloopArgs.priv.syms = dsp.getDelayedPrivSymbols();
taskloopArgs.priv.vars = taskloopClauseOps.privateVars;
+ taskloopArgs.reduction.syms = reductionSyms;
+ taskloopArgs.reduction.vars = taskloopClauseOps.reductionVars;
+ taskloopArgs.inReduction.syms = inReductionSyms;
+ taskloopArgs.inReduction.vars = taskloopClauseOps.inReductionVars;
auto taskLoopOp = genWrapperOp<mlir::omp::TaskloopOp>(
converter, loc, taskloopClauseOps, taskloopArgs);
diff --git a/flang/test/Lower/OpenMP/Todo/taskloop-inreduction.f90 b/flang/test/Lower/OpenMP/Todo/taskloop-inreduction.f90
deleted file mode 100644
index 8acc399a92abe..0000000000000
--- a/flang/test/Lower/OpenMP/Todo/taskloop-inreduction.f90
+++ /dev/null
@@ -1,13 +0,0 @@
-! RUN: %not_todo_cmd bbc -emit-fir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
-! RUN: %not_todo_cmd %flang_fc1 -emit-fir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
-
-! CHECK: not yet implemented: Unhandled clause IN_REDUCTION in TASKLOOP construct
-subroutine omp_taskloop_inreduction()
- integer x
- x = 0
- !$omp taskloop in_reduction(+:x)
- do i = 1, 100
- x = x + 1
- end do
- !$omp end taskloop
-end subroutine omp_taskloop_inreduction
diff --git a/flang/test/Lower/OpenMP/Todo/taskloop-reduction.f90 b/flang/test/Lower/OpenMP/Todo/taskloop-reduction.f90
deleted file mode 100644
index 0c16bd227257f..0000000000000
--- a/flang/test/Lower/OpenMP/Todo/taskloop-reduction.f90
+++ /dev/null
@@ -1,13 +0,0 @@
-! RUN: %not_todo_cmd bbc -emit-fir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
-! RUN: %not_todo_cmd %flang_fc1 -emit-fir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
-
-! CHECK: not yet implemented: Unhandled clause REDUCTION in TASKLOOP construct
-subroutine omp_taskloop_reduction()
- integer x
- x = 0
- !$omp taskloop reduction(+:x)
- do i = 1, 100
- x = x + 1
- end do
- !$omp end taskloop
-end subroutine omp_taskloop_reduction
diff --git a/flang/test/Lower/OpenMP/taskloop.f90 b/flang/test/Lower/OpenMP/taskloop.f90
index 79b0c20e176c0..5f20b0bbd60b9 100644
--- a/flang/test/Lower/OpenMP/taskloop.f90
+++ b/flang/test/Lower/OpenMP/taskloop.f90
@@ -1,6 +1,15 @@
! RUN: bbc -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
! RUN: %flang_fc1 -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
+! CHECK-LABEL: omp.private
+! CHECK-SAME: {type = private} @[[LAST_PRIVATE_I:.*]] : i32
+
+! CHECK-LABEL: omp.private
+! CHECK-SAME: {type = private} @[[LAST_PRIVATE_X:.*]] : i32
+
+! CHECK-LABEL: omp.private
+! CHECK-SAME: {type = private} @[[QFOMP_TASKLOOP_NOGROUPEI_PRIVATE_I32:.*]] : i32
+
! CHECK-LABEL: omp.private
! CHECK-SAME: {type = private} @[[I_PRIVATE_TEST2:.*]] : i32
@@ -70,3 +79,46 @@ subroutine omp_taskloop_private
! CHECK: }
!$omp end taskloop
end subroutine omp_taskloop_private
+
+!===============================================================================
+! `nogroup` clause
+!===============================================================================
+
+subroutine omp_taskloop_nogroup()
+ ! CHECK: omp.taskloop nogroup
+ !$omp taskloop nogroup
+ do i = 1, 10
+ call foo()
+ end do
+ !$omp end taskloop
+end subroutine
+
+!===============================================================================
+! `lastprivate` clause
+!===============================================================================
+
+! CHECK-LABEL: func.func @_QPomp_taskloop_lastprivate
+! CHECK: %[[ALLOCA_I:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFomp_taskloop_lastprivateEi"}
+! CHECK: %[[DECL_I:.*]]:2 = hlfir.declare %[[ALLOCA_I]] {uniq_name = "_QFomp_taskloop_lastprivateEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+! CHECK: %[[ALLOCA_X:.*]] = fir.alloca i32 {bindc_name = "x", uniq_name = "_QFomp_taskloop_lastprivateEx"}
+! CHECK: %[[DECL_X:.*]]:2 = hlfir.declare %[[ALLOCA_X]] {uniq_name = "_QFomp_taskloop_lastprivateEx"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+subroutine omp_taskloop_lastprivate()
+ integer x
+ x = 0
+ ! CHECK: omp.taskloop private(@[[LAST_PRIVATE_X]] %[[DECL_X]]#0 -> %[[ARG0]], @[[LAST_PRIVATE_I]] %[[DECL_I]]#0 -> %[[ARG1]] : !fir.ref<i32>, !fir.ref<i32>) {
+ !$omp taskloop lastprivate(x)
+ do i = 1, 100
+ ! CHECK: %[[DECL_ARG0:.*]]:2 = hlfir.declare %[[ARG0]] {uniq_name = "_QFomp_taskloop_lastprivateEx"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+ ! CHECK: %[[LOAD_ARG0:.*]] = fir.load %[[DECL_ARG0]]#0 : !fir.ref<i32>
+ ! CHECK: %[[RES_ADD:.*]] = arith.addi %[[LOAD_ARG0]], %{{.*}} : i32
+ ! CHECK: hlfir.assign %[[RES_ADD]] to %[[DECL_ARG0]]#0 : i32, !fir.ref<i32>
+ x = x + 1
+ ! CHECK: %[[SELCT_RESULT:.*]] = arith.select %{{.*}}, %{{.*}}, %{{.*}} : i1
+ ! CHECK: fir.if %[[SELCT_RESULT]] {
+ ! CHECK: %[[LOADED_SUM:.*]] = fir.load %[[DECL_ARG0]]#0 : !fir.ref<i32>
+ ! CHECK: hlfir.assign %[[LOADED_SUM]] to %[[DECL_X]]#0 : i32, !fir.ref<i32>
+ ! CHECK: }
+ ! CHECK: omp.yield
+ end do
+ !$omp end taskloop
+end subroutine omp_taskloop_lastprivate
\ No newline at end of file
diff --git a/flang/test/Lower/taskloop-inreduction.f90 b/flang/test/Lower/taskloop-inreduction.f90
new file mode 100644
index 0000000000000..9973382cb60ee
--- /dev/null
+++ b/flang/test/Lower/taskloop-inreduction.f90
@@ -0,0 +1,40 @@
+! RUN: bbc -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
+! RUN: %flang_fc1 -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
+
+! CHECK-LABEL: omp.private
+! CHECK-SAME: {type = private} @[[PRIVATE_I:.*]] : i32
+
+! CHECK-LABEL: omp.declare_reduction
+! CHECK-SAME: @[[ADD_RED_I32:.*]] : i32 init {
+! CHECK: ^bb0(%{{.*}}: i32):
+! CHECK: %[[C0_I32:.*]] = arith.constant 0 : i32
+! CHECK: omp.yield(%[[C0_I32]] : i32)
+! CHECK: } combiner {
+! CHECK: ^bb0(%{{.*}}: i32, %{{.*}}: i32):
+! CHECK: %[[RES:.*]] = arith.addi %{{.*}}, %{{.*}} : i32
+! CHECK: omp.yield(%[[RES]] : i32)
+! CHECK: }
+
+! CHECK-LABEL: func.func @_QPomp_taskloop_inreduction
+! CHECK: %[[ALLOCA_I:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFomp_taskloop_inreductionEi"}
+! CHECK: %[[DECL_I:.*]]:2 = hlfir.declare %[[ALLOCA_I]] {uniq_name = "_QFomp_taskloop_inreductionEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+! CHECK: %[[ALLOCA_X:.*]] = fir.alloca i32 {bindc_name = "x", uniq_name = "_QFomp_taskloop_inreductionEx"}
+! CHECK: %[[DECL_X:.*]]:2 = hlfir.declare %[[ALLOCA_X]] {uniq_name = "_QFomp_taskloop_inreductionEx"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+! CHECK: %[[INIT_X:.*]] = arith.constant 0 : i32
+! CHECK: hlfir.assign %[[INIT_X]] to %[[DECL_X]]#0 : i32, !fir.ref<i32>
+subroutine omp_taskloop_inreduction()
+ integer x
+ x = 0
+ ! CHECK: omp.taskloop in_reduction(@[[ADD_RED_I32]]
+ ! CHECK: %[[DECL_X]]#0 -> %[[ARG0:.*]] : !fir.ref<i32>) private(@[[PRIVATE_I]] %[[DECL_I]]#0 -> %[[ARG1:.*]] : !fir.ref<i32>) {
+ ! CHECK: %[[VAL_ARG1:.*]]:2 = hlfir.declare %[[ARG0]]
+ ! CHECK-SAME: {uniq_name = "_QFomp_taskloop_inreductionEx"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+ !$omp taskloop in_reduction(+:x)
+ do i = 1, 100
+ ! CHECK: %[[X_VAL:.*]] = fir.load %[[VAL_ARG1]]#0 : !fir.ref<i32>
+ ! CHECK: %[[ADD_VAL:.*]] = arith.addi %[[X_VAL]], %{{.*}} : i32
+ x = x + 1
+ ! CHECK: hlfir.assign %[[ADD_VAL]] to %[[VAL_ARG1]]#0 : i32, !fir.ref<i32>
+ end do
+ !$omp end taskloop
+end subroutine omp_taskloop_inreduction
\ No newline at end of file
diff --git a/flang/test/Lower/taskloop-reduction.f90 b/flang/test/Lower/taskloop-reduction.f90
new file mode 100644
index 0000000000000..bf7511c4c1abc
--- /dev/null
+++ b/flang/test/Lower/taskloop-reduction.f90
@@ -0,0 +1,39 @@
+! RUN: bbc -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
+! RUN: %flang_fc1 -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
+
+! CHECK-LABEL: omp.private
+! CHECK-SAME: {type = private} @[[PRIVATE_I:.*]] : i32
+
+! CHECK-LABEL: omp.declare_reduction
+! CHECK-SAME: @[[ADD_RED_I32:.*]] : i32 init {
+! CHECK: ^bb0(%{{.*}}: i32):
+! CHECK: %[[C0_I32:.*]] = arith.constant 0 : i32
+! CHECK: omp.yield(%[[C0_I32]] : i32)
+! CHECK: } combiner {
+! CHECK: ^bb0(%{{.*}}: i32, %{{.*}}: i32):
+! CHECK: %[[RES:.*]] = arith.addi %{{.*}}, %{{.*}} : i32
+! CHECK: omp.yield(%[[RES]] : i32)
+! CHECK: }
+
+! CHECK-LABEL: func.func @_QPomp_taskloop_reduction
+! CHECK: %[[ALLOCA_I:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFomp_taskloop_reductionEi"}
+! CHECK: %[[DECL_I:.*]]:2 = hlfir.declare %[[ALLOCA_I]] {uniq_name = "_QFomp_taskloop_reductionEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+! CHECK: %[[ALLOCA_X:.*]] = fir.alloca i32 {bindc_name = "x", uniq_name = "_QFomp_taskloop_reductionEx"}
+! CHECK: %[[DECL_X:.*]]:2 = hlfir.declare %[[ALLOCA_X]] {uniq_name = "_QFomp_taskloop_reductionEx"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+! CHECK: %[[INIT_X:.*]] = arith.constant 0 : i32
+! CHECK: hlfir.assign %[[INIT_X]] to %[[DECL_X]]#0 : i32, !fir.ref<i32>
+subroutine omp_taskloop_reduction()
+ integer x
+ x = 0
+ ! CHECK: omp.taskloop private(@[[PRIVATE_I]]
+ ! CHECK-SAME: %[[DECL_I]]#0 -> %[[ARG0:.*]] : !fir.ref<i32>) reduction(@[[ADD_RED_I32]] %[[DECL_X]]#0 -> %[[ARG1:.*]] : !fir.ref<i32>) {
+ ! CHECK: %[[VAL_ARG1:.*]]:2 = hlfir.declare %[[ARG1]]
+ !$omp taskloop reduction(+:x)
+ do i = 1, 100
+ ! CHECK: %[[X_VAL:.*]] = fir.load %[[VAL_ARG1]]#0 : !fir.ref<i32>
+ ! CHECK: %[[ADD_VAL:.*]] = arith.addi %[[X_VAL]], %{{.*}} : i32
+ x = x + 1
+ ! CHECK: hlfir.assign %[[ADD_VAL]] to %[[VAL_ARG1]]#0 : i32, !fir.ref<i32>
+ end do
+ !$omp end taskloop
+end subroutine omp_taskloop_reduction
\ No newline at end of file
|
kaviya2510
force-pushed
the
taskloop_mlir_lowering
branch
from
609a1c6 to
397edbf
Compare
This was referenced Nov 7, 2025
Stylie777
reviewed
Nov 7, 2025
Contributor
Stylie777
left a comment
Stylie777
left a comment
There was a problem hiding this comment.
One small comment from me, Thanks for posting Kaviya
Contributor
Author
Thankyou for the apporoval @tblah ! |
Contributor
Author
Thankyou @Stylie777 for the approval! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This patch add MLIR lowering support for nogroup, reduction, inreduction and lastprivate clauses of taskloop directive.