Skip to content

[Flang][OpenMP] DISTRIBUTE PARALLEL DO SIMD lowering #106211

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
Aug 29, 2024
Merged

[Flang][OpenMP] DISTRIBUTE PARALLEL DO SIMD lowering #106211

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
72 changes: 71 additions & 1 deletion flang/lib/Lower/OpenMP/OpenMP.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2150,8 +2150,78 @@ static void genCompositeDistributeParallelDoSimd(
semantics::SemanticsContext &semaCtx, lower::pft::Evaluation &eval,
mlir::Location loc, const ConstructQueue &queue,
ConstructQueue::const_iterator item) {
lower::StatementContext stmtCtx;

assert(std::distance(item, queue.end()) == 4 && "Invalid leaf constructs");
TODO(loc, "Composite DISTRIBUTE PARALLEL DO SIMD");
ConstructQueue::const_iterator distributeItem = item;
ConstructQueue::const_iterator parallelItem = std::next(distributeItem);
ConstructQueue::const_iterator doItem = std::next(parallelItem);
ConstructQueue::const_iterator simdItem = std::next(doItem);

// Create parent omp.parallel first.
mlir::omp::ParallelOperands parallelClauseOps;
llvm::SmallVector<const semantics::Symbol *> parallelReductionSyms;
llvm::SmallVector<mlir::Type> parallelReductionTypes;
genParallelClauses(converter, semaCtx, stmtCtx, parallelItem->clauses, loc,
parallelClauseOps, parallelReductionTypes,
parallelReductionSyms);

DataSharingProcessor dsp(converter, semaCtx, simdItem->clauses, eval,
/*shouldCollectPreDeterminedSymbols=*/true,
/*useDelayedPrivatization=*/true, &symTable);
dsp.processStep1(&parallelClauseOps);

genParallelOp(converter, symTable, semaCtx, eval, loc, queue, parallelItem,
parallelClauseOps, parallelReductionSyms,
parallelReductionTypes, &dsp, /*isComposite=*/true);

// Clause processing.
mlir::omp::DistributeOperands distributeClauseOps;
genDistributeClauses(converter, semaCtx, stmtCtx, distributeItem->clauses,
loc, distributeClauseOps);

mlir::omp::WsloopOperands wsloopClauseOps;
llvm::SmallVector<const semantics::Symbol *> wsloopReductionSyms;
llvm::SmallVector<mlir::Type> wsloopReductionTypes;
genWsloopClauses(converter, semaCtx, stmtCtx, doItem->clauses, loc,
wsloopClauseOps, wsloopReductionTypes, wsloopReductionSyms);

mlir::omp::SimdOperands simdClauseOps;
genSimdClauses(converter, semaCtx, simdItem->clauses, loc, simdClauseOps);

mlir::omp::LoopNestOperands loopNestClauseOps;
llvm::SmallVector<const semantics::Symbol *> iv;
genLoopNestClauses(converter, semaCtx, eval, simdItem->clauses, loc,
loopNestClauseOps, iv);

// Operation creation.
// TODO: Populate entry block arguments with private variables.
auto distributeOp = genWrapperOp<mlir::omp::DistributeOp>(
converter, loc, distributeClauseOps, /*blockArgTypes=*/{});
distributeOp.setComposite(/*val=*/true);

// TODO: Add private variables to entry block arguments.
auto wsloopOp = genWrapperOp<mlir::omp::WsloopOp>(
converter, loc, wsloopClauseOps, wsloopReductionTypes);
wsloopOp.setComposite(/*val=*/true);

// TODO: Populate entry block arguments with reduction and private variables.
auto simdOp = genWrapperOp<mlir::omp::SimdOp>(converter, loc, simdClauseOps,
/*blockArgTypes=*/{});
simdOp.setComposite(/*val=*/true);

// Construct wrapper entry block list and associated symbols. It is important
// that the symbol order and the block argument order match, so that the
// symbol-value bindings created are correct.
auto &wrapperSyms = wsloopReductionSyms;

auto wrapperArgs = llvm::to_vector(llvm::concat<mlir::BlockArgument>(
distributeOp.getRegion().getArguments(),
wsloopOp.getRegion().getArguments(), simdOp.getRegion().getArguments()));

genLoopNestOp(converter, symTable, semaCtx, eval, loc, queue, simdItem,
loopNestClauseOps, iv, wrapperSyms, wrapperArgs,
llvm::omp::Directive::OMPD_distribute_parallel_do_simd, dsp);
}

static void genCompositeDistributeSimd(lower::AbstractConverter &converter,
Expand Down
100 changes: 100 additions & 0 deletions flang/test/Lower/OpenMP/distribute-parallel-do-simd.f90
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,100 @@
! This test checks lowering of OpenMP DISTRIBUTE PARALLEL DO SIMD composite
! constructs.

! RUN: bbc -fopenmp -emit-hlfir %s -o - | FileCheck %s
! RUN: %flang_fc1 -fopenmp -emit-hlfir %s -o - | FileCheck %s

! CHECK-LABEL: func.func @_QPdistribute_parallel_do_simd_num_threads(
subroutine distribute_parallel_do_simd_num_threads()
!$omp teams

! CHECK: omp.parallel num_threads({{.*}}) private({{.*}}) {
! CHECK: omp.distribute {
! CHECK-NEXT: omp.wsloop {
! CHECK-NEXT: omp.simd {
! CHECK-NEXT: omp.loop_nest
!$omp distribute parallel do simd num_threads(10)
do index_ = 1, 10
end do
!$omp end distribute parallel do simd

!$omp end teams
end subroutine distribute_parallel_do_simd_num_threads

! CHECK-LABEL: func.func @_QPdistribute_parallel_do_simd_dist_schedule(
subroutine distribute_parallel_do_simd_dist_schedule()
!$omp teams

! CHECK: omp.parallel private({{.*}}) {
! CHECK: omp.distribute dist_schedule_static dist_schedule_chunk_size({{.*}}) {
! CHECK-NEXT: omp.wsloop {
! CHECK-NEXT: omp.simd {
! CHECK-NEXT: omp.loop_nest
!$omp distribute parallel do simd dist_schedule(static, 4)
do index_ = 1, 10
end do
!$omp end distribute parallel do simd

!$omp end teams
end subroutine distribute_parallel_do_simd_dist_schedule

! CHECK-LABEL: func.func @_QPdistribute_parallel_do_simd_schedule(
subroutine distribute_parallel_do_simd_schedule()
!$omp teams

! CHECK: omp.parallel private({{.*}}) {
! CHECK: omp.distribute {
! CHECK-NEXT: omp.wsloop schedule(static = {{.*}}) {
! CHECK-NEXT: omp.simd {
! CHECK-NEXT: omp.loop_nest
!$omp distribute parallel do simd schedule(static, 4)
do index_ = 1, 10
end do
!$omp end distribute parallel do simd

!$omp end teams
end subroutine distribute_parallel_do_simd_schedule

! CHECK-LABEL: func.func @_QPdistribute_parallel_do_simd_simdlen(
subroutine distribute_parallel_do_simd_simdlen()
!$omp teams

! CHECK: omp.parallel private({{.*}}) {
! CHECK: omp.distribute {
! CHECK-NEXT: omp.wsloop {
! CHECK-NEXT: omp.simd simdlen(4) {
! CHECK-NEXT: omp.loop_nest
!$omp distribute parallel do simd simdlen(4)
do index_ = 1, 10
end do
!$omp end distribute parallel do simd

!$omp end teams
end subroutine distribute_parallel_do_simd_simdlen

! CHECK-LABEL: func.func @_QPdistribute_parallel_do_simd_private(
subroutine distribute_parallel_do_simd_private()
! CHECK: %[[INDEX_ALLOC:.*]] = fir.alloca i32
! CHECK: %[[INDEX:.*]]:2 = hlfir.declare %[[INDEX_ALLOC]]
! CHECK: %[[X_ALLOC:.*]] = fir.alloca i64
! CHECK: %[[X:.*]]:2 = hlfir.declare %[[X_ALLOC]]
integer(8) :: x

! CHECK: omp.teams {
!$omp teams

! CHECK: omp.parallel private(@{{.*}} %[[X]]#0 -> %[[X_ARG:.*]] : !fir.ref<i64>,
! CHECK-SAME: @{{.*}} %[[INDEX]]#0 -> %[[INDEX_ARG:.*]] : !fir.ref<i32>) {
! CHECK: %[[X_PRIV:.*]]:2 = hlfir.declare %[[X_ARG]]
! CHECK: %[[INDEX_PRIV:.*]]:2 = hlfir.declare %[[INDEX_ARG]]
! CHECK: omp.distribute {
! CHECK-NEXT: omp.wsloop {
! CHECK-NEXT: omp.simd {
! CHECK-NEXT: omp.loop_nest
!$omp distribute parallel do simd private(x)
do index_ = 1, 10
end do
!$omp end distribute parallel do simd

!$omp end teams
end subroutine distribute_parallel_do_simd_private
Loading
Loading