[clang] 76e6c8d - Codegen changes for strict modifier with grainsize/num_tasks of taskloop construct (#117196)
via cfe-commits
cfe-commits at lists.llvm.org
Thu Nov 28 00:49:04 PST 2024
Author: CHANDRA GHALE
Date: 2024-11-28T14:18:59+05:30
New Revision: 76e6c8d3fc3dc097a3bd96e8959f73a809493976
URL: https://github.com/llvm/llvm-project/commit/76e6c8d3fc3dc097a3bd96e8959f73a809493976
DIFF: https://github.com/llvm/llvm-project/commit/76e6c8d3fc3dc097a3bd96e8959f73a809493976.diff
LOG: Codegen changes for strict modifier with grainsize/num_tasks of taskloop construct (#117196)
Initial parsing/sema for 'strict' modifier with 'num_tasks' and
‘grainsize’ clause is present in these commits
[grainsize_parsing](https://github.com/llvm/llvm-project/commit/ab9eac762c35068e77f57795e660d06f578c9614)
and
[num_tasks_parsing](https://github.com/llvm/llvm-project/commit/56c166017055595a9f26933e85bfd89e30c528d0#diff-4184486638e85284c3a2c961a81e7752231022daf97e411007c13a6732b50db9R6545)
. However, this implementation appears incomplete as it lacks code
generation support. A runtime patch was introduced in this runtime
commit
[runtime_patch](https://github.com/llvm/llvm-project/commit/540007b42701b5ac9adba076824bfd648a265413#diff-5e95f9319910d6965d09c301359dbe6b23f3eef5ce4d262ef2c2d2137875b5c4R374)
, which adds a new API, _kmpc_taskloop_5, to accommodate the strict
modifier.
In this patch I have added codegen support. When the strict modifier is
present alongside the grainsize or num_tasks clauses of taskloop
construct, the code now emits a call to _kmpc_taskloop_5, which includes
an additional parameter of type i32 with the value 1 to indicate the
strict modifier. If the strict modifier is not present, it falls back to
the existing _kmpc_taskloop API call.
---------
Co-authored-by: Chandra Ghale <ghale at pe31.hpc.amslabs.hpecorp.net>
Added:
clang/test/OpenMP/taskloop_strictmodifier_codegen.cpp
Modified:
clang/lib/CodeGen/CGOpenMPRuntime.cpp
clang/lib/CodeGen/CGOpenMPRuntime.h
clang/lib/CodeGen/CGStmtOpenMP.cpp
llvm/include/llvm/Frontend/OpenMP/OMPKinds.def
Removed:
################################################################################
diff --git a/clang/lib/CodeGen/CGOpenMPRuntime.cpp b/clang/lib/CodeGen/CGOpenMPRuntime.cpp
index cc389974e04081..6a5860242035b2 100644
--- a/clang/lib/CodeGen/CGOpenMPRuntime.cpp
+++ b/clang/lib/CodeGen/CGOpenMPRuntime.cpp
@@ -4666,7 +4666,7 @@ void CGOpenMPRuntime::emitTaskLoopCall(CodeGenFunction &CGF, SourceLocation Loc,
CGF.getContext().VoidPtrTy);
}
enum { NoSchedule = 0, Grainsize = 1, NumTasks = 2 };
- llvm::Value *TaskArgs[] = {
+ llvm::SmallVector<llvm::Value *, 12> TaskArgs{
UpLoc,
ThreadID,
Result.NewTask,
@@ -4683,12 +4683,18 @@ void CGOpenMPRuntime::emitTaskLoopCall(CodeGenFunction &CGF, SourceLocation Loc,
Data.Schedule.getPointer()
? CGF.Builder.CreateIntCast(Data.Schedule.getPointer(), CGF.Int64Ty,
/*isSigned=*/false)
- : llvm::ConstantInt::get(CGF.Int64Ty, /*V=*/0),
- Result.TaskDupFn ? CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
- Result.TaskDupFn, CGF.VoidPtrTy)
- : llvm::ConstantPointerNull::get(CGF.VoidPtrTy)};
+ : llvm::ConstantInt::get(CGF.Int64Ty, /*V=*/0)};
+ if (Data.HasModifier)
+ TaskArgs.push_back(llvm::ConstantInt::get(CGF.Int32Ty, 1));
+
+ TaskArgs.push_back(Result.TaskDupFn
+ ? CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
+ Result.TaskDupFn, CGF.VoidPtrTy)
+ : llvm::ConstantPointerNull::get(CGF.VoidPtrTy));
CGF.EmitRuntimeCall(OMPBuilder.getOrCreateRuntimeFunction(
- CGM.getModule(), OMPRTL___kmpc_taskloop),
+ CGM.getModule(), Data.HasModifier
+ ? OMPRTL___kmpc_taskloop_5
+ : OMPRTL___kmpc_taskloop),
TaskArgs);
}
diff --git a/clang/lib/CodeGen/CGOpenMPRuntime.h b/clang/lib/CodeGen/CGOpenMPRuntime.h
index 5e7715743afb58..56d502d92806eb 100644
--- a/clang/lib/CodeGen/CGOpenMPRuntime.h
+++ b/clang/lib/CodeGen/CGOpenMPRuntime.h
@@ -122,6 +122,7 @@ struct OMPTaskDataTy final {
bool IsReductionWithTaskMod = false;
bool IsWorksharingReduction = false;
bool HasNowaitClause = false;
+ bool HasModifier = false;
};
/// Class intended to support codegen of all kind of the reduction clauses.
diff --git a/clang/lib/CodeGen/CGStmtOpenMP.cpp b/clang/lib/CodeGen/CGStmtOpenMP.cpp
index 390516fea38498..6cb37b20b7aeee 100644
--- a/clang/lib/CodeGen/CGStmtOpenMP.cpp
+++ b/clang/lib/CodeGen/CGStmtOpenMP.cpp
@@ -7831,10 +7831,14 @@ void CodeGenFunction::EmitOMPTaskLoopBasedDirective(const OMPLoopDirective &S) {
// grainsize clause
Data.Schedule.setInt(/*IntVal=*/false);
Data.Schedule.setPointer(EmitScalarExpr(Clause->getGrainsize()));
+ Data.HasModifier =
+ (Clause->getModifier() == OMPC_GRAINSIZE_strict) ? true : false;
} else if (const auto *Clause = S.getSingleClause<OMPNumTasksClause>()) {
// num_tasks clause
Data.Schedule.setInt(/*IntVal=*/true);
Data.Schedule.setPointer(EmitScalarExpr(Clause->getNumTasks()));
+ Data.HasModifier =
+ (Clause->getModifier() == OMPC_NUMTASKS_strict) ? true : false;
}
auto &&BodyGen = [CS, &S](CodeGenFunction &CGF, PrePostActionTy &) {
diff --git a/clang/test/OpenMP/taskloop_strictmodifier_codegen.cpp b/clang/test/OpenMP/taskloop_strictmodifier_codegen.cpp
new file mode 100644
index 00000000000000..107db947d015f3
--- /dev/null
+++ b/clang/test/OpenMP/taskloop_strictmodifier_codegen.cpp
@@ -0,0 +1,165 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --include-generated-funcs --prefix-filecheck-ir-name _ --version 5
+// RUN: %clang_cc1 -fopenmp -O1 -x c++ -triple x86_64-apple-darwin10 -emit-llvm %s -o - | FileCheck %s --check-prefix=CHECK
+// expected-no-diagnostics
+#ifndef HEADER
+#define HEADER
+
+int main(int argc, char **argv) {
+#pragma omp task
+#pragma omp master taskloop priority(argc)
+ for (int i = 0; i < 10; ++i)
+ ;
+#pragma omp master taskloop nogroup grainsize(strict:argc)
+ for (int i = 0; i < 10; ++i)
+ ;
+ int i;
+#pragma omp master taskloop if(argc) shared(argc, argv) collapse(2) num_tasks(strict: 4)
+ for (i = 0; i < argc; ++i)
+ for (int j = argc; j < argv[argc][argc]; ++j)
+ ;
+#pragma omp master taskloop
+ for (int i = 0; i < 10; ++i) {
+#pragma omp cancel taskgroup
+#pragma omp cancellation point taskgroup
+ }
+}
+struct S {
+ int a;
+ S(int c) {
+#pragma omp master taskloop shared(c) num_tasks(strict:a)
+ for (a = 0; a < c; ++a)
+ ;
+ }
+} s(1);
+
+#endif
+
+// CHECK-LABEL: define noundef i32 @main(
+// CHECK-SAME: i32 noundef [[ARGC:%.*]], ptr noundef [[ARGV:%.*]]) local_unnamed_addr #[[ATTR0:[0-9]+]] {
+// CHECK-NEXT: [[ENTRY:.*:]]
+// CHECK-NEXT: [[ARGC_ADDR:%.*]] = alloca i32, align 4
+// CHECK-NEXT: [[ARGV_ADDR:%.*]] = alloca ptr, align 8
+// CHECK-NEXT: [[TMP0:%.*]] = tail call i32 @__kmpc_global_thread_num(ptr nonnull @[[GLOB1:[0-9]+]])
+// CHECK-NEXT: store i32 [[ARGC]], ptr [[ARGC_ADDR]], align 4, !tbaa [[TBAA3:![0-9]+]]
+// CHECK-NEXT: store ptr [[ARGV]], ptr [[ARGV_ADDR]], align 8, !tbaa [[TBAA7:![0-9]+]]
+// CHECK-NEXT: [[TMP1:%.*]] = tail call ptr @__kmpc_omp_task_alloc(ptr nonnull @[[GLOB1]], i32 [[TMP0]], i32 1, i64 48, i64 1, ptr nonnull @.omp_task_entry..2)
+// CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP1]], i64 40
+// CHECK-NEXT: [[TMP3:%.*]] = load i32, ptr [[ARGC_ADDR]], align 4, !tbaa [[TBAA3]]
+// CHECK-NEXT: store i32 [[TMP3]], ptr [[TMP2]], align 8, !tbaa [[TBAA9:![0-9]+]]
+// CHECK-NEXT: [[TMP4:%.*]] = tail call i32 @__kmpc_omp_task(ptr nonnull @[[GLOB1]], i32 [[TMP0]], ptr [[TMP1]])
+// CHECK-NEXT: [[TMP5:%.*]] = tail call i32 @__kmpc_master(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: [[DOTNOT:%.*]] = icmp eq i32 [[TMP5]], 0
+// CHECK-NEXT: br i1 [[DOTNOT]], label %[[OMP_IF_END:.*]], label %[[OMP_IF_THEN:.*]]
+// CHECK: [[OMP_IF_THEN]]:
+// CHECK-NEXT: [[TMP6:%.*]] = load i32, ptr [[ARGC_ADDR]], align 4, !tbaa [[TBAA3]]
+// CHECK-NEXT: [[TMP7:%.*]] = tail call ptr @__kmpc_omp_task_alloc(ptr nonnull @[[GLOB1]], i32 [[TMP0]], i32 1, i64 80, i64 1, ptr nonnull @.omp_task_entry..4)
+// CHECK-NEXT: [[TMP8:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP7]], i64 40
+// CHECK-NEXT: store i64 0, ptr [[TMP8]], align 8, !tbaa [[TBAA13:![0-9]+]]
+// CHECK-NEXT: [[TMP9:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP7]], i64 48
+// CHECK-NEXT: store i64 9, ptr [[TMP9]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP10:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP7]], i64 56
+// CHECK-NEXT: store i64 1, ptr [[TMP10]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP11:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP7]], i64 72
+// CHECK-NEXT: store i64 0, ptr [[TMP11]], align 8
+// CHECK-NEXT: [[TMP12:%.*]] = zext i32 [[TMP6]] to i64
+// CHECK-NEXT: tail call void @__kmpc_taskloop_5(ptr nonnull @[[GLOB1]], i32 [[TMP0]], ptr [[TMP7]], i32 1, ptr nonnull [[TMP8]], ptr nonnull [[TMP9]], i64 1, i32 1, i32 1, i64 [[TMP12]], i32 1, ptr null) #[[ATTR1:[0-9]+]]
+// CHECK-NEXT: tail call void @__kmpc_end_master(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: br label %[[OMP_IF_END]]
+// CHECK: [[OMP_IF_END]]:
+// CHECK-NEXT: [[TMP13:%.*]] = tail call i32 @__kmpc_master(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: [[DOTNOT22:%.*]] = icmp eq i32 [[TMP13]], 0
+// CHECK-NEXT: br i1 [[DOTNOT22]], label %[[OMP_IF_END17:.*]], label %[[OMP_IF_THEN2:.*]]
+// CHECK: [[OMP_IF_THEN2]]:
+// CHECK-NEXT: tail call void @__kmpc_taskgroup(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: [[TMP14:%.*]] = load i32, ptr [[ARGC_ADDR]], align 4, !tbaa [[TBAA3]]
+// CHECK-NEXT: [[TMP15:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8, !tbaa [[TBAA7]]
+// CHECK-NEXT: [[IDXPROM:%.*]] = sext i32 [[TMP14]] to i64
+// CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP15]], i64 [[IDXPROM]]
+// CHECK-NEXT: [[TMP16:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8, !tbaa [[TBAA7]]
+// CHECK-NEXT: [[ARRAYIDX9:%.*]] = getelementptr inbounds i8, ptr [[TMP16]], i64 [[IDXPROM]]
+// CHECK-NEXT: [[TMP17:%.*]] = load i8, ptr [[ARRAYIDX9]], align 1, !tbaa [[TBAA15:![0-9]+]]
+// CHECK-NEXT: [[CONV:%.*]] = sext i8 [[TMP17]] to i32
+// CHECK-NEXT: [[SUB12:%.*]] = sub i32 [[CONV]], [[TMP14]]
+// CHECK-NEXT: [[CONV15:%.*]] = zext i32 [[SUB12]] to i64
+// CHECK-NEXT: [[MUL:%.*]] = mul nsw i64 [[CONV15]], [[IDXPROM]]
+// CHECK-NEXT: [[SUB16:%.*]] = add nsw i64 [[MUL]], -1
+// CHECK-NEXT: [[TMP18:%.*]] = tail call ptr @__kmpc_omp_task_alloc(ptr nonnull @[[GLOB1]], i32 [[TMP0]], i32 1, i64 80, i64 16, ptr nonnull @.omp_task_entry..6)
+// CHECK-NEXT: [[TMP19:%.*]] = load ptr, ptr [[TMP18]], align 8, !tbaa [[TBAA16:![0-9]+]]
+// CHECK-NEXT: store ptr [[ARGC_ADDR]], ptr [[TMP19]], align 8, !tbaa [[TBAA7]]
+// CHECK-NEXT: [[AGG_CAPTURED3_SROA_2_0__SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[TMP19]], i64 8
+// CHECK-NEXT: store ptr [[ARGV_ADDR]], ptr [[AGG_CAPTURED3_SROA_2_0__SROA_IDX]], align 8, !tbaa [[TBAA7]]
+// CHECK-NEXT: [[TMP20:%.*]] = load i32, ptr [[ARGC_ADDR]], align 4, !tbaa [[TBAA3]]
+// CHECK-NEXT: [[TOBOOL:%.*]] = icmp ne i32 [[TMP20]], 0
+// CHECK-NEXT: [[TMP21:%.*]] = sext i1 [[TOBOOL]] to i32
+// CHECK-NEXT: [[TMP22:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP18]], i64 40
+// CHECK-NEXT: store i64 0, ptr [[TMP22]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP23:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP18]], i64 48
+// CHECK-NEXT: store i64 [[SUB16]], ptr [[TMP23]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP24:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP18]], i64 56
+// CHECK-NEXT: store i64 1, ptr [[TMP24]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP25:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP18]], i64 72
+// CHECK-NEXT: store i64 0, ptr [[TMP25]], align 8
+// CHECK-NEXT: call void @__kmpc_taskloop_5(ptr nonnull @[[GLOB1]], i32 [[TMP0]], ptr nonnull [[TMP18]], i32 [[TMP21]], ptr nonnull [[TMP22]], ptr nonnull [[TMP23]], i64 1, i32 1, i32 2, i64 4, i32 1, ptr null) #[[ATTR1]]
+// CHECK-NEXT: call void @__kmpc_end_taskgroup(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: call void @__kmpc_end_master(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: br label %[[OMP_IF_END17]]
+// CHECK: [[OMP_IF_END17]]:
+// CHECK-NEXT: [[TMP26:%.*]] = call i32 @__kmpc_master(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: [[DOTNOT23:%.*]] = icmp eq i32 [[TMP26]], 0
+// CHECK-NEXT: br i1 [[DOTNOT23]], label %[[OMP_IF_END21:.*]], label %[[OMP_IF_THEN18:.*]]
+// CHECK: [[OMP_IF_THEN18]]:
+// CHECK-NEXT: call void @__kmpc_taskgroup(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: [[TMP27:%.*]] = call ptr @__kmpc_omp_task_alloc(ptr nonnull @[[GLOB1]], i32 [[TMP0]], i32 1, i64 80, i64 1, ptr nonnull @.omp_task_entry..8)
+// CHECK-NEXT: [[TMP28:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 40
+// CHECK-NEXT: store i64 0, ptr [[TMP28]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP29:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 48
+// CHECK-NEXT: store i64 9, ptr [[TMP29]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP30:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 56
+// CHECK-NEXT: store i64 1, ptr [[TMP30]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP31:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 72
+// CHECK-NEXT: store i64 0, ptr [[TMP31]], align 8
+// CHECK-NEXT: call void @__kmpc_taskloop(ptr nonnull @[[GLOB1]], i32 [[TMP0]], ptr [[TMP27]], i32 1, ptr nonnull [[TMP28]], ptr nonnull [[TMP29]], i64 1, i32 1, i32 0, i64 0, ptr null)
+// CHECK-NEXT: call void @__kmpc_end_taskgroup(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: call void @__kmpc_end_master(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: br label %[[OMP_IF_END21]]
+// CHECK: [[OMP_IF_END21]]:
+// CHECK-NEXT: ret i32 0
+//
+//
+//
+// CHECK-LABEL: define linkonce_odr void @_ZN1SC2Ei(
+// CHECK-SAME: ptr noundef nonnull align 4 dereferenceable(4) [[THIS:%.*]], i32 noundef [[C:%.*]]) unnamed_addr #[[ATTR6:[0-9]+]] align 2 {
+// CHECK-NEXT: [[ENTRY:.*:]]
+// CHECK-NEXT: [[C_ADDR:%.*]] = alloca i32, align 4
+// CHECK-NEXT: [[TMP0:%.*]] = tail call i32 @__kmpc_global_thread_num(ptr nonnull @[[GLOB1]])
+// CHECK-NEXT: store i32 [[C]], ptr [[C_ADDR]], align 4, !tbaa [[TBAA3]]
+// CHECK-NEXT: [[TMP1:%.*]] = tail call i32 @__kmpc_master(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: [[DOTNOT:%.*]] = icmp eq i32 [[TMP1]], 0
+// CHECK-NEXT: br i1 [[DOTNOT]], label %[[OMP_IF_END:.*]], label %[[OMP_IF_THEN:.*]]
+// CHECK: [[OMP_IF_THEN]]:
+// CHECK-NEXT: [[TMP2:%.*]] = load i32, ptr [[THIS]], align 4, !tbaa [[TBAA27:![0-9]+]]
+// CHECK-NEXT: tail call void @__kmpc_taskgroup(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: [[TMP3:%.*]] = load i32, ptr [[C_ADDR]], align 4, !tbaa [[TBAA3]]
+// CHECK-NEXT: [[SUB4:%.*]] = add nsw i32 [[TMP3]], -1
+// CHECK-NEXT: [[TMP4:%.*]] = tail call ptr @__kmpc_omp_task_alloc(ptr nonnull @[[GLOB1]], i32 [[TMP0]], i32 1, i64 80, i64 16, ptr nonnull @.omp_task_entry..10)
+// CHECK-NEXT: [[TMP5:%.*]] = load ptr, ptr [[TMP4]], align 8, !tbaa [[TBAA16]]
+// CHECK-NEXT: store ptr [[THIS]], ptr [[TMP5]], align 8, !tbaa [[TBAA7]]
+// CHECK-NEXT: [[AGG_CAPTURED_SROA_2_0__SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[TMP5]], i64 8
+// CHECK-NEXT: store ptr [[C_ADDR]], ptr [[AGG_CAPTURED_SROA_2_0__SROA_IDX]], align 8, !tbaa [[TBAA7]]
+// CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP4]], i64 40
+// CHECK-NEXT: store i64 0, ptr [[TMP6]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP7:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP4]], i64 48
+// CHECK-NEXT: [[CONV:%.*]] = sext i32 [[SUB4]] to i64
+// CHECK-NEXT: store i64 [[CONV]], ptr [[TMP7]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP8:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP4]], i64 56
+// CHECK-NEXT: store i64 1, ptr [[TMP8]], align 8, !tbaa [[TBAA13]]
+// CHECK-NEXT: [[TMP9:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP4]], i64 72
+// CHECK-NEXT: store i64 0, ptr [[TMP9]], align 8
+// CHECK-NEXT: [[TMP10:%.*]] = zext i32 [[TMP2]] to i64
+// CHECK-NEXT: call void @__kmpc_taskloop_5(ptr nonnull @[[GLOB1]], i32 [[TMP0]], ptr nonnull [[TMP4]], i32 1, ptr nonnull [[TMP6]], ptr nonnull [[TMP7]], i64 1, i32 1, i32 2, i64 [[TMP10]], i32 1, ptr null) #[[ATTR1]]
+// CHECK-NEXT: call void @__kmpc_end_taskgroup(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: call void @__kmpc_end_master(ptr nonnull @[[GLOB1]], i32 [[TMP0]])
+// CHECK-NEXT: br label %[[OMP_IF_END]]
+// CHECK: [[OMP_IF_END]]:
+// CHECK-NEXT: ret void
+
diff --git a/llvm/include/llvm/Frontend/OpenMP/OMPKinds.def b/llvm/include/llvm/Frontend/OpenMP/OMPKinds.def
index 6f26f853eca032..928a03148c4165 100644
--- a/llvm/include/llvm/Frontend/OpenMP/OMPKinds.def
+++ b/llvm/include/llvm/Frontend/OpenMP/OMPKinds.def
@@ -365,6 +365,9 @@ __OMP_RTL(__kmpc_omp_task_with_deps, false, Int32, IdentPtr, Int32,
__OMP_RTL(__kmpc_taskloop, false, Void, IdentPtr, /* Int */ Int32, VoidPtr,
/* Int */ Int32, Int64Ptr, Int64Ptr, Int64, /* Int */ Int32,
/* Int */ Int32, Int64, VoidPtr)
+__OMP_RTL(__kmpc_taskloop_5, false, Void, IdentPtr, /* Int */ Int32, VoidPtr,
+ /* Int */ Int32, Int64Ptr, Int64Ptr, Int64, /* Int */ Int32,
+ /* Int */ Int32, Int64, Int32, VoidPtr)
__OMP_RTL(__kmpc_omp_target_task_alloc, false, /* kmp_task_t */ VoidPtr,
IdentPtr, Int32, Int32, SizeTy, SizeTy, TaskRoutineEntryPtr, Int64)
__OMP_RTL(__kmpc_taskred_modifier_init, false, /* kmp_taskgroup */ VoidPtr,
More information about the cfe-commits
mailing list