gtbercea updated this revision to Diff 127865.
gtbercea added a comment.
Address comments.
Repository:
rL LLVM
https://reviews.llvm.org/D41486
Files:
lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp
test/OpenMP/nvptx_data_sharing.cpp
test/OpenMP/nvptx_target_teams_codegen.cpp
Index: test/OpenMP/nvptx_target_teams_codegen.cpp
===================================================================
--- test/OpenMP/nvptx_target_teams_codegen.cpp
+++ test/OpenMP/nvptx_target_teams_codegen.cpp
@@ -60,7 +60,7 @@
//
// CHECK: [[AWAIT_WORK]]
// CHECK: call void @llvm.nvvm.barrier0()
- // CHECK: [[KPR:%.+]] = call i1 @__kmpc_kernel_parallel(i8** [[OMP_WORK_FN]], i8*** %shared_args)
+ // CHECK: [[KPR:%.+]] = call i1 @__kmpc_kernel_parallel(i8** [[OMP_WORK_FN]], i16 1, i8*** %shared_args)
// CHECK: [[KPRB:%.+]] = zext i1 [[KPR]] to i8
// store i8 [[KPRB]], i8* [[OMP_EXEC_STATUS]], align 1
// CHECK: [[WORK:%.+]] = load i8*, i8** [[OMP_WORK_FN]],
@@ -148,7 +148,7 @@
//
// CHECK: [[AWAIT_WORK]]
// CHECK: call void @llvm.nvvm.barrier0()
- // CHECK: [[KPR:%.+]] = call i1 @__kmpc_kernel_parallel(i8** [[OMP_WORK_FN]], i8*** %shared_args)
+ // CHECK: [[KPR:%.+]] = call i1 @__kmpc_kernel_parallel(i8** [[OMP_WORK_FN]], i16 1, i8*** %shared_args)
// CHECK: [[KPRB:%.+]] = zext i1 [[KPR]] to i8
// store i8 [[KPRB]], i8* [[OMP_EXEC_STATUS]], align 1
// CHECK: [[WORK:%.+]] = load i8*, i8** [[OMP_WORK_FN]],
Index: test/OpenMP/nvptx_data_sharing.cpp
===================================================================
--- test/OpenMP/nvptx_data_sharing.cpp
+++ test/OpenMP/nvptx_data_sharing.cpp
@@ -24,15 +24,15 @@
// CK1: define internal void @__omp_offloading_{{.*}}test_ds{{.*}}worker() [[ATTR1:#.*]] {
// CK1: [[SHAREDARGS:%.+]] = alloca i8**
-// CK1: call i1 @__kmpc_kernel_parallel(i8** %work_fn, i8*** [[SHAREDARGS]])
+// CK1: call i1 @__kmpc_kernel_parallel(i8** %work_fn, i16 1, i8*** [[SHAREDARGS]])
// CK1: [[SHARGSTMP:%.+]] = load i8**, i8*** [[SHAREDARGS]]
// CK1: call void @__omp_outlined___wrapper{{.*}}({{.*}}, i8** [[SHARGSTMP]])
/// ========= In the kernel function ========= ///
// CK1: {{.*}}define void @__omp_offloading{{.*}}test_ds{{.*}}() [[ATTR2:#.*]] {
// CK1: [[SHAREDARGS1:%.+]] = alloca i8**
-// CK1: call void @__kmpc_kernel_prepare_parallel({{.*}}, i8*** [[SHAREDARGS1]], i32 1)
+// CK1: call void @__kmpc_kernel_prepare_parallel({{.*}}, i16 1, i8*** [[SHAREDARGS1]], i32 1)
// CK1: [[SHARGSTMP1:%.+]] = load i8**, i8*** [[SHAREDARGS1]]
// CK1: [[SHARGSTMP2:%.+]] = getelementptr inbounds i8*, i8** [[SHARGSTMP1]]
// CK1: [[SHAREDVAR:%.+]] = bitcast i32* {{.*}} to i8*
Index: lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp
===================================================================
--- lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp
+++ lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp
@@ -33,10 +33,11 @@
/// \brief Call to void __kmpc_spmd_kernel_deinit();
OMPRTL_NVPTX__kmpc_spmd_kernel_deinit,
/// \brief Call to void __kmpc_kernel_prepare_parallel(void
- /// *outlined_function, void ***args, kmp_int32 nArgs);
+ /// *outlined_function, int16_t IsOMPRuntimeInitialized,
+ /// void ***args, kmp_int32 nArgs);
OMPRTL_NVPTX__kmpc_kernel_prepare_parallel,
- /// \brief Call to bool __kmpc_kernel_parallel(void **outlined_function, void
- /// ***args);
+ /// \brief Call to bool __kmpc_kernel_parallel(void **outlined_function,
+ /// int16_t IsOMPRuntimeInitialized, void ***args);
OMPRTL_NVPTX__kmpc_kernel_parallel,
/// \brief Call to void __kmpc_kernel_end_parallel();
OMPRTL_NVPTX__kmpc_kernel_end_parallel,
@@ -521,7 +522,9 @@
// Set up shared arguments
Address SharedArgs =
CGF.CreateDefaultAlignTempAlloca(CGF.Int8PtrPtrTy, "shared_args");
- llvm::Value *Args[] = {WorkFn.getPointer(), SharedArgs.getPointer()};
+ llvm::Value *Args[] = {WorkFn.getPointer(),
+ /*IsOMPRuntimeInitialized*/ Bld.getInt16(1),
+ SharedArgs.getPointer()};
llvm::Value *Ret = CGF.EmitRuntimeCall(
createNVPTXRuntimeFunction(OMPRTL_NVPTX__kmpc_kernel_parallel), Args);
Bld.CreateStore(Bld.CreateZExt(Ret, CGF.Int8Ty), ExecStatus);
@@ -638,16 +641,16 @@
case OMPRTL_NVPTX__kmpc_kernel_prepare_parallel: {
/// Build void __kmpc_kernel_prepare_parallel(
/// void *outlined_function, void ***args, kmp_int32 nArgs);
- llvm::Type *TypeParams[] = {CGM.Int8PtrTy,
+ llvm::Type *TypeParams[] = {CGM.Int8PtrTy, CGM.Int16Ty,
CGM.Int8PtrPtrTy->getPointerTo(0), CGM.Int32Ty};
llvm::FunctionType *FnTy =
llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_kernel_prepare_parallel");
break;
}
case OMPRTL_NVPTX__kmpc_kernel_parallel: {
/// Build bool __kmpc_kernel_parallel(void **outlined_function, void ***args);
- llvm::Type *TypeParams[] = {CGM.Int8PtrPtrTy,
+ llvm::Type *TypeParams[] = {CGM.Int8PtrPtrTy, CGM.Int16Ty,
CGM.Int8PtrPtrTy->getPointerTo(0)};
llvm::Type *RetTy = CGM.getTypes().ConvertType(CGM.getContext().BoolTy);
llvm::FunctionType *FnTy =
@@ -949,8 +952,9 @@
CGF.CreateDefaultAlignTempAlloca(CGF.VoidPtrPtrTy,
"shared_args");
llvm::Value *SharedArgsPtr = SharedArgs.getPointer();
- llvm::Value *Args[] = {ID, SharedArgsPtr,
- Bld.getInt32(CapturedVars.size())};
+ llvm::Value *Args[] = {ID,
+ /*IsOMPRuntimeInitialized*/ Bld.getInt16(1), SharedArgsPtr,
+ Bld.getInt32(CapturedVars.size())};
CGF.EmitRuntimeCall(
createNVPTXRuntimeFunction(OMPRTL_NVPTX__kmpc_kernel_prepare_parallel),
@@ -970,7 +974,7 @@
Idx++;
}
} else {
- llvm::Value *Args[] = {ID,
+ llvm::Value *Args[] = {ID, /*IsOMPRuntimeInitialized*/ Bld.getInt16(1),
llvm::ConstantPointerNull::get(CGF.VoidPtrPtrTy->getPointerTo(0)),
/*nArgs=*/Bld.getInt32(0)};
CGF.EmitRuntimeCall(
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
