Index: lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp
===================================================================
--- lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp
+++ lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp
@@ -521,7 +521,8 @@
   // Set up shared arguments
   Address SharedArgs =
       CGF.CreateDefaultAlignTempAlloca(CGF.Int8PtrPtrTy, "shared_args");
-  llvm::Value *Args[] = {WorkFn.getPointer(), SharedArgs.getPointer()};
+  llvm::Value *Args[] = {WorkFn.getPointer(), 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);
@@ -638,7 +639,7 @@
   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);
@@ -647,7 +648,7 @@
   }
   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 =
@@ -949,7 +950,7 @@
           CGF.CreateDefaultAlignTempAlloca(CGF.VoidPtrPtrTy,
               "shared_args");
       llvm::Value *SharedArgsPtr = SharedArgs.getPointer();
-      llvm::Value *Args[] = {ID, SharedArgsPtr,
+      llvm::Value *Args[] = {ID, Bld.getInt16(1), SharedArgsPtr,
                              Bld.getInt32(CapturedVars.size())};
 
       CGF.EmitRuntimeCall(
@@ -970,7 +971,7 @@
         Idx++;
       }
     } else {
-      llvm::Value *Args[] = {ID,
+      llvm::Value *Args[] = {ID, Bld.getInt16(1),
           llvm::ConstantPointerNull::get(CGF.VoidPtrPtrTy->getPointerTo(0)),
           /*nArgs=*/Bld.getInt32(0)};
       CGF.EmitRuntimeCall(
Index: test/OpenMP/nvptx_data_sharing.cpp
===================================================================
--- test/OpenMP/nvptx_data_sharing.cpp
+++ test/OpenMP/nvptx_data_sharing.cpp
@@ -24,7 +24,7 @@
 
 // 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, {{.*}} i8*** [[SHAREDARGS]])
 // CK1: [[SHARGSTMP:%.+]] = load i8**, i8*** [[SHAREDARGS]]
 // CK1: call void @__omp_outlined___wrapper{{.*}}({{.*}}, i8** [[SHARGSTMP]])
 
@@ -32,7 +32,7 @@
 
 // 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({{.*}}, {{.*}} i8*** [[SHAREDARGS1]], i32 1)
 // CK1: [[SHARGSTMP1:%.+]] = load i8**, i8*** [[SHAREDARGS1]]
 // CK1: [[SHARGSTMP2:%.+]] = getelementptr inbounds i8*, i8** [[SHARGSTMP1]]
 // CK1: [[SHAREDVAR:%.+]] = bitcast i32* {{.*}} to i8*
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]], {{.*}} 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]], {{.*}} 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]],