[Flang][OpenMP] DISTRIBUTE PARALLEL DO SIMD lowering #106211
Merged
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
skatrak
requested review from
ergawy,
Meinersbur,
jsjodin,
tblah,
kiranchandramohan,
abidh,
mjklemm,
raghavendhra,
agozillon,
DominikAdamski,
TIFitis,
kparzysz,
anchuraj and
bhandarkar-pranav
llvmbot
added
flang
|
@llvm/pr-subscribers-flang-fir-hlfir @llvm/pr-subscribers-flang-openmp Author: Sergio Afonso (skatrak) ChangesThis patch adds PFT to MLIR lowering support for Patch is 21.97 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/106211.diff 4 Files Affected:
diff --git a/flang/lib/Lower/OpenMP/OpenMP.cpp b/flang/lib/Lower/OpenMP/OpenMP.cpp
index 072ed71b8c9958..28bcaf19743754 100644
--- a/flang/lib/Lower/OpenMP/OpenMP.cpp
+++ b/flang/lib/Lower/OpenMP/OpenMP.cpp
@@ -2144,8 +2144,78 @@ 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(¶llelClauseOps);
+
+ 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,
diff --git a/flang/test/Lower/OpenMP/distribute-parallel-do-simd.f90 b/flang/test/Lower/OpenMP/distribute-parallel-do-simd.f90
new file mode 100644
index 00000000000000..711d4dc4ba1773
--- /dev/null
+++ b/flang/test/Lower/OpenMP/distribute-parallel-do-simd.f90
@@ -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
diff --git a/flang/test/Lower/OpenMP/if-clause.f90 b/flang/test/Lower/OpenMP/if-clause.f90
index fce4efc7a5c2e2..3ae9018ae4d5d0 100644
--- a/flang/test/Lower/OpenMP/if-clause.f90
+++ b/flang/test/Lower/OpenMP/if-clause.f90
@@ -9,14 +9,11 @@ program main
integer :: i
! TODO When they are supported, add tests for:
- ! - DISTRIBUTE PARALLEL DO SIMD
! - PARALLEL SECTIONS
! - PARALLEL WORKSHARE
- ! - TARGET TEAMS DISTRIBUTE PARALLEL DO SIMD
! - TARGET UPDATE
! - TASKLOOP
! - TASKLOOP SIMD
- ! - TEAMS DISTRIBUTE PARALLEL DO SIMD
! ----------------------------------------------------------------------------
! DISTRIBUTE PARALLEL DO
@@ -68,6 +65,97 @@ program main
!$omp end teams
+ ! ----------------------------------------------------------------------------
+ ! DISTRIBUTE PARALLEL DO SIMD
+ ! ----------------------------------------------------------------------------
+ !$omp teams
+
+ ! CHECK: omp.parallel
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp distribute parallel do simd
+ do i = 1, 10
+ end do
+ !$omp end distribute parallel do simd
+
+ ! CHECK: omp.parallel
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp distribute parallel do simd if(.true.)
+ do i = 1, 10
+ end do
+ !$omp end distribute parallel do simd
+
+ ! CHECK: omp.parallel
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp distribute parallel do simd if(parallel: .true.) if(simd: .false.)
+ do i = 1, 10
+ end do
+ !$omp end distribute parallel do simd
+
+ ! CHECK: omp.parallel
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp distribute parallel do simd if(parallel: .true.)
+ do i = 1, 10
+ end do
+ !$omp end distribute parallel do simd
+
+ ! CHECK: omp.parallel
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp distribute parallel do simd if(simd: .true.)
+ do i = 1, 10
+ end do
+ !$omp end distribute parallel do simd
+
+ !$omp end teams
+
! ----------------------------------------------------------------------------
! DISTRIBUTE SIMD
! ----------------------------------------------------------------------------
@@ -828,6 +916,165 @@ program main
end do
!$omp end target teams distribute parallel do
+ ! ----------------------------------------------------------------------------
+ ! TARGET TEAMS DISTRIBUTE PARALLEL DO SIMD
+ ! ----------------------------------------------------------------------------
+ ! CHECK: omp.target
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.teams
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.parallel
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp target teams distribute parallel do simd
+ do i = 1, 10
+ end do
+ !$omp end target teams distribute parallel do simd
+
+ ! CHECK: omp.target
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.teams
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.parallel
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp target teams distribute parallel do simd if(.true.)
+ do i = 1, 10
+ end do
+ !$omp end target teams distribute parallel do simd
+
+ ! CHECK: omp.target
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.teams
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.parallel
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp target teams distribute parallel do simd if(target: .true.) if(teams: .false.) if(parallel: .true.) if(simd: .false.)
+ do i = 1, 10
+ end do
+ !$omp end target teams distribute parallel do simd
+
+ ! CHECK: omp.target
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.teams
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.parallel
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp target teams distribute parallel do simd if(target: .true.)
+ do i = 1, 10
+ end do
+ !$omp end target teams distribute parallel do simd
+
+ ! CHECK: omp.target
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.teams
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.parallel
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp target teams distribute parallel do simd if(teams: .true.)
+ do i = 1, 10
+ end do
+ !$omp end target teams distribute parallel do simd
+
+ ! CHECK: omp.target
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.teams
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.parallel
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp target teams distribute parallel do simd if(parallel: .true.)
+ do i = 1, 10
+ end do
+ !$omp end target teams distribute parallel do simd
+
+ ! CHECK: omp.target
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.teams
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.parallel
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp target teams distribute parallel do simd if(simd: .true.)
+ do i = 1, 10
+ end do
+ !$omp end target teams distribute parallel do simd
+
! ----------------------------------------------------------------------------
! TARGET TEAMS DISTRIBUTE SIMD
! ----------------------------------------------------------------------------
@@ -1120,6 +1367,126 @@ program main
end do
!$omp end teams distribute parallel do
+ ! ----------------------------------------------------------------------------
+ ! TEAMS DISTRIBUTE PARALLEL DO SIMD
+ ! ----------------------------------------------------------------------------
+ ! CHECK: omp.teams
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.parallel
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp teams distribute parallel do simd
+ do i = 1, 10
+ end do
+ !$omp end teams distribute parallel do simd
+
+ ! CHECK: omp.teams
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.parallel
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp teams distribute parallel do simd if(.true.)
+ do i = 1, 10
+ end do
+ !$omp end teams distribute parallel do simd
+
+ ! CHECK: omp.teams
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.parallel
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp teams distribute parallel do simd if(teams: .false.) if(parallel: .true.) if(simd: .false.)
+ do i = 1, 10
+ end do
+ !$omp end teams distribute parallel do simd
+
+ ! CHECK: omp.teams
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.parallel
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp teams distribute parallel do simd if(teams: .true.)
+ do i = 1, 10
+ end do
+ !$omp end teams distribute parallel do simd
+
+ ! CHECK: omp.teams
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.parallel
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp teams distribute parallel do simd if(parallel: .true.)
+ do i = 1, 10
+ end do
+ !$omp end teams distribute parallel do simd
+
+ ! CHECK: omp.teams
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.parallel
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK: omp.distribute
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.wsloop
+ ! CHECK-NOT: if({{.*}})
+ ! CHECK-SAME: {
+ ! CHECK-NEXT: omp.simd
+ ! CHECK-SAME: if({{.*}})
+ ! CHECK-NEXT: omp.loop_nest
+ !$omp teams distribute parallel do simd if(simd: .true.)
+ do i = 1, 10
+ end do
+ !$omp end teams distribute parallel do simd
+
! ----------------------------------------------------------------------------
! TEAMS DISTRIBUTE SIMD
! ----------------------------------------------------------------------------
diff --git a/flang/test/Lower/OpenMP/loop-compound.f90 b/flang/test/Lower/OpenMP/loop-compound.f90
index fb61621e10f0ac..e76edfe052f745 100644
--- a/flang/test/Lower/OpenMP/loop-compound.f90
+++ b/flang/test/Lower/OpenMP/loop-compound.f90
@@ -8,12 +8,7 @@ program main
integer :: i
! TODO When composite constructs are supported add:
- ! - DISTRIBUTE PARALLEL DO SIMD
- ! - TARGET TEAMS DISTRIBUTE PARALLEL DO SIMD
- ! - TARGET TEAMS DISTRIBUTE PARALLEL DO
! - TASKLOOP SIMD
- ! - TEAMS DISTRIBUTE PARALLEL DO SIMD
- ! - TEAMS DISTRIBUTE PARALLEL DO
! ----------------------------------------------------------------------------
! DISTRIBUTE PARALLEL DO
@@ -31,6 +26,23 @@ program main
!$omp end teams
+ ! ----------------------------------------------------------------------------
+ ! DISTRIBUTE PARALLEL DO SIMD
+ ! ----------------------------------------------------------------------------
+ !$omp teams
+
+ ! ...
[truncated]
|
|
This looks good on its own. How many of these composite constructs are there in total? I'm wondering if there are enough that we should consider a more elaborate solution to avoid duplicating code (e.g. what if clause processing changes at some point). |
Thank you for the quick reviews! There are just a few composite constructs: I think that when this is a bit more mature we should be able to move to a per-leaf composite lowering and avoid having all this similar code. But I think doing it before having composite privatization completely figured out may result in some bad abstractions we'll have to change later anyways. So maybe that improvement could come later, but let me know if you think I should address it as part of this PR. |
tblah
approved these changes
Aug 27, 2024
skatrak
force-pushed
the
users/skatrak/dpd-03-dist-par-do
branch
from
5b654d9 to
1ae4c05
Compare
skatrak
force-pushed
the
users/skatrak/dpd-04-dist-par-do-simd
branch
from
e16c878 to
3508bc0
Compare
skatrak
force-pushed
the
users/skatrak/dpd-03-dist-par-do
branch
from
1ae4c05 to
e63b7ab
Compare
This patch adds PFT to MLIR lowering support for `distribute parallel do simd` composite constructs.
skatrak
force-pushed
the
users/skatrak/dpd-04-dist-par-do-simd
branch
from
3508bc0 to
bbc23dc
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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 adds PFT to MLIR lowering support for
distribute parallel do simdcomposite constructs.