diff --git a/clang/test/OpenMP/cancel_codegen.cpp b/clang/test/OpenMP/cancel_codegen.cpp
diff --git a/clang/test/OpenMP/cancellation_point_codegen.cpp b/clang/test/OpenMP/cancellation_point_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_codegen.cpp b/clang/test/OpenMP/distribute_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_firstprivate_codegen.cpp b/clang/test/OpenMP/distribute_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_lastprivate_codegen.cpp b/clang/test/OpenMP/distribute_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_firstprivate_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_if_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_if_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_lastprivate_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_num_threads_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_num_threads_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_private_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_private_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_proc_bind_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_proc_bind_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_reduction_task_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_simd_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_simd_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_simd_firstprivate_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_simd_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_simd_if_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_simd_if_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_simd_lastprivate_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_simd_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_simd_num_threads_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_simd_num_threads_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_simd_private_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_simd_private_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_parallel_for_simd_proc_bind_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_simd_proc_bind_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_private_codegen.cpp b/clang/test/OpenMP/distribute_private_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_simd_codegen.cpp b/clang/test/OpenMP/distribute_simd_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_simd_firstprivate_codegen.cpp b/clang/test/OpenMP/distribute_simd_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_simd_lastprivate_codegen.cpp b/clang/test/OpenMP/distribute_simd_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_simd_private_codegen.cpp b/clang/test/OpenMP/distribute_simd_private_codegen.cpp
diff --git a/clang/test/OpenMP/distribute_simd_reduction_codegen.cpp b/clang/test/OpenMP/distribute_simd_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/for_firstprivate_codegen.cpp b/clang/test/OpenMP/for_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/for_lastprivate_codegen.cpp b/clang/test/OpenMP/for_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/for_linear_codegen.cpp b/clang/test/OpenMP/for_linear_codegen.cpp
diff --git a/clang/test/OpenMP/for_private_codegen.cpp b/clang/test/OpenMP/for_private_codegen.cpp
diff --git a/clang/test/OpenMP/for_reduction_codegen.cpp b/clang/test/OpenMP/for_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/for_reduction_codegen_UDR.cpp b/clang/test/OpenMP/for_reduction_codegen_UDR.cpp
diff --git a/clang/test/OpenMP/for_reduction_task_codegen.cpp b/clang/test/OpenMP/for_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/master_taskloop_in_reduction_codegen.cpp b/clang/test/OpenMP/master_taskloop_in_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/master_taskloop_simd_in_reduction_codegen.cpp b/clang/test/OpenMP/master_taskloop_simd_in_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/nvptx_distribute_parallel_generic_mode_codegen.cpp b/clang/test/OpenMP/nvptx_distribute_parallel_generic_mode_codegen.cpp
diff --git a/clang/test/OpenMP/nvptx_lambda_capturing.cpp b/clang/test/OpenMP/nvptx_lambda_capturing.cpp
diff --git a/clang/test/OpenMP/nvptx_multi_target_parallel_codegen.cpp b/clang/test/OpenMP/nvptx_multi_target_parallel_codegen.cpp
@@ -4,7 +4,7 @@
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-cuda-mode -x c++ -triple nvptx64-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - -disable-llvm-optzns | FileCheck %s --check-prefix=CHECK1
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-cuda-mode -x c++ -triple i386-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm-bc %s -o %t-x86-host.bc
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-cuda-mode -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - -disable-llvm-optzns | FileCheck %s --check-prefix=CHECK2
-// RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-cuda-mode -fexceptions -fcxx-exceptions -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - -disable-llvm-optzns | FileCheck %s --check-prefix=CHECK3
+// RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-cuda-mode -fexceptions -fcxx-exceptions -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - -disable-llvm-optzns | FileCheck %s --check-prefix=CHECK2
 
 // expected-no-diagnostics
 #ifndef HEADER
@@ -180,81 +180,3 @@ int main() {
 // CHECK2-NEXT:    call void @__omp_outlined__1(i32* [[DOTADDR1]], i32* [[DOTZERO_ADDR]]) #[[ATTR3:[0-9]+]]
 // CHECK2-NEXT:    ret void
 //
-//
-// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l21
-// CHECK3-SAME: () #[[ATTR0:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = call i32 @__kmpc_target_init(%struct.ident_t* @[[GLOB1:[0-9]+]], i8 2, i1 false, i1 true)
-// CHECK3-NEXT:    [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
-// CHECK3-NEXT:    br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
-// CHECK3:       user_code.entry:
-// CHECK3-NEXT:    [[TMP1:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB2:[0-9]+]])
-// CHECK3-NEXT:    [[TMP2:%.*]] = bitcast [0 x i8*]* [[CAPTURED_VARS_ADDRS]] to i8**
-// CHECK3-NEXT:    call void @__kmpc_parallel_51(%struct.ident_t* @[[GLOB2]], i32 [[TMP1]], i32 1, i32 -1, i32 -1, i8* bitcast (void (i32*, i32*)* @__omp_outlined__ to i8*), i8* null, i8** [[TMP2]], i32 0)
-// CHECK3-NEXT:    call void @__kmpc_target_deinit(%struct.ident_t* @[[GLOB1]], i8 2, i1 true)
-// CHECK3-NEXT:    ret void
-// CHECK3:       worker.exit:
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR1:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    call void @_Z3usev() #[[ATTR8:[0-9]+]]
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@_Z3usev
-// CHECK3-SAME: () #[[ATTR2:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB3:[0-9]+]])
-// CHECK3-NEXT:    [[TMP1:%.*]] = bitcast [0 x i8*]* [[CAPTURED_VARS_ADDRS]] to i8**
-// CHECK3-NEXT:    call void @__kmpc_parallel_51(%struct.ident_t* @[[GLOB3]], i32 [[TMP0]], i32 1, i32 -1, i32 -1, i8* bitcast (void (i32*, i32*)* @__omp_outlined__1 to i8*), i8* bitcast (void (i16, i32)* @__omp_outlined__1_wrapper to i8*), i8** [[TMP1]], i32 0)
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l23
-// CHECK3-SAME: () #[[ATTR5:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[TMP0:%.*]] = call i32 @__kmpc_target_init(%struct.ident_t* @[[GLOB1]], i8 1, i1 true, i1 true)
-// CHECK3-NEXT:    [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
-// CHECK3-NEXT:    br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
-// CHECK3:       user_code.entry:
-// CHECK3-NEXT:    call void @_Z3usev() #[[ATTR8]]
-// CHECK3-NEXT:    call void @__kmpc_target_deinit(%struct.ident_t* @[[GLOB1]], i8 1, i1 true)
-// CHECK3-NEXT:    ret void
-// CHECK3:       worker.exit:
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__1
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR1]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    call void @_Z4workv() #[[ATTR8]]
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__1_wrapper
-// CHECK3-SAME: (i16 noundef zeroext [[TMP0:%.*]], i32 noundef [[TMP1:%.*]]) #[[ATTR7:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTADDR:%.*]] = alloca i16, align 2
-// CHECK3-NEXT:    [[DOTADDR1:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[GLOBAL_ARGS:%.*]] = alloca i8**, align 4
-// CHECK3-NEXT:    store i16 [[TMP0]], i16* [[DOTADDR]], align 2
-// CHECK3-NEXT:    store i32 [[TMP1]], i32* [[DOTADDR1]], align 4
-// CHECK3-NEXT:    store i32 0, i32* [[DOTZERO_ADDR]], align 4
-// CHECK3-NEXT:    call void @__kmpc_get_shared_variables(i8*** [[GLOBAL_ARGS]])
-// CHECK3-NEXT:    call void @__omp_outlined__1(i32* [[DOTADDR1]], i32* [[DOTZERO_ADDR]]) #[[ATTR3:[0-9]+]]
-// CHECK3-NEXT:    ret void
-//
diff --git a/clang/test/OpenMP/nvptx_nested_parallel_codegen.cpp b/clang/test/OpenMP/nvptx_nested_parallel_codegen.cpp
@@ -4,7 +4,7 @@
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-cuda-mode -x c++ -triple nvptx64-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - -disable-llvm-optzns | FileCheck %s --check-prefix=CHECK1
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-cuda-mode -x c++ -triple i386-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm-bc %s -o %t-x86-host.bc
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-cuda-mode -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - -disable-llvm-optzns | FileCheck %s --check-prefix=CHECK2
-// RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-cuda-mode -fexceptions -fcxx-exceptions -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - -disable-llvm-optzns | FileCheck %s --check-prefix=CHECK3
+// RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-cuda-mode -fexceptions -fcxx-exceptions -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - -disable-llvm-optzns | FileCheck %s --check-prefix=CHECK2
 
 // expected-no-diagnostics
 #ifndef HEADER
@@ -294,134 +294,3 @@ int main() {
 // CHECK2-NEXT:    call void @__omp_outlined__1(i32* [[DOTADDR1]], i32* [[DOTZERO_ADDR]], i32** [[TMP5]]) #[[ATTR4]]
 // CHECK2-NEXT:    ret void
 //
-//
-// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l25
-// CHECK3-SAME: (i32* noundef nonnull align 4 dereferenceable(4) [[C:%.*]]) #[[ATTR0:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[C_ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [1 x i8*], align 4
-// CHECK3-NEXT:    store i32* [[C]], i32** [[C_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = load i32*, i32** [[C_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP1:%.*]] = call i32 @__kmpc_target_init(%struct.ident_t* @[[GLOB1:[0-9]+]], i8 1, i1 true, i1 true)
-// CHECK3-NEXT:    [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP1]], -1
-// CHECK3-NEXT:    br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
-// CHECK3:       user_code.entry:
-// CHECK3-NEXT:    [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB1]])
-// CHECK3-NEXT:    call void @_Z3usePi(i32* noundef [[TMP0]]) #[[ATTR7:[0-9]+]]
-// CHECK3-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[CAPTURED_VARS_ADDRS]], i32 0, i32 0
-// CHECK3-NEXT:    [[TMP4:%.*]] = bitcast i32* [[TMP0]] to i8*
-// CHECK3-NEXT:    store i8* [[TMP4]], i8** [[TMP3]], align 4
-// CHECK3-NEXT:    [[TMP5:%.*]] = bitcast [1 x i8*]* [[CAPTURED_VARS_ADDRS]] to i8**
-// CHECK3-NEXT:    call void @__kmpc_parallel_51(%struct.ident_t* @[[GLOB1]], i32 [[TMP2]], i32 1, i32 2, i32 -1, i8* bitcast (void (i32*, i32*, i32*)* @__omp_outlined__ to i8*), i8* bitcast (void (i16, i32)* @__omp_outlined___wrapper to i8*), i8** [[TMP5]], i32 1)
-// CHECK3-NEXT:    call void @__kmpc_target_deinit(%struct.ident_t* @[[GLOB1]], i8 1, i1 true)
-// CHECK3-NEXT:    ret void
-// CHECK3:       worker.exit:
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@_Z3usePi
-// CHECK3-SAME: (i32* noundef [[C:%.*]]) #[[ATTR1:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[C_ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [1 x i8*], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB2:[0-9]+]])
-// CHECK3-NEXT:    store i32* [[C]], i32** [[C_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP1:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[CAPTURED_VARS_ADDRS]], i32 0, i32 0
-// CHECK3-NEXT:    [[TMP2:%.*]] = bitcast i32** [[C_ADDR]] to i8*
-// CHECK3-NEXT:    store i8* [[TMP2]], i8** [[TMP1]], align 4
-// CHECK3-NEXT:    [[TMP3:%.*]] = bitcast [1 x i8*]* [[CAPTURED_VARS_ADDRS]] to i8**
-// CHECK3-NEXT:    call void @__kmpc_parallel_51(%struct.ident_t* @[[GLOB2]], i32 [[TMP0]], i32 1, i32 2, i32 -1, i8* bitcast (void (i32*, i32*, i32**)* @__omp_outlined__1 to i8*), i8* bitcast (void (i16, i32)* @__omp_outlined__1_wrapper to i8*), i8** [[TMP3]], i32 1)
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i32* noundef nonnull align 4 dereferenceable(4) [[C:%.*]]) #[[ATTR2:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[C_ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[C]], i32** [[C_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = load i32*, i32** [[C_ADDR]], align 4
-// CHECK3-NEXT:    call void @_Z3usePi(i32* noundef [[TMP0]]) #[[ATTR7]]
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined___wrapper
-// CHECK3-SAME: (i16 noundef zeroext [[TMP0:%.*]], i32 noundef [[TMP1:%.*]]) #[[ATTR3:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTADDR:%.*]] = alloca i16, align 2
-// CHECK3-NEXT:    [[DOTADDR1:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[GLOBAL_ARGS:%.*]] = alloca i8**, align 4
-// CHECK3-NEXT:    store i16 [[TMP0]], i16* [[DOTADDR]], align 2
-// CHECK3-NEXT:    store i32 [[TMP1]], i32* [[DOTADDR1]], align 4
-// CHECK3-NEXT:    store i32 0, i32* [[DOTZERO_ADDR]], align 4
-// CHECK3-NEXT:    call void @__kmpc_get_shared_variables(i8*** [[GLOBAL_ARGS]])
-// CHECK3-NEXT:    [[TMP2:%.*]] = load i8**, i8*** [[GLOBAL_ARGS]], align 4
-// CHECK3-NEXT:    [[TMP3:%.*]] = getelementptr inbounds i8*, i8** [[TMP2]], i32 0
-// CHECK3-NEXT:    [[TMP4:%.*]] = bitcast i8** [[TMP3]] to i32**
-// CHECK3-NEXT:    [[TMP5:%.*]] = load i32*, i32** [[TMP4]], align 4
-// CHECK3-NEXT:    call void @__omp_outlined__(i32* [[DOTADDR1]], i32* [[DOTZERO_ADDR]], i32* [[TMP5]]) #[[ATTR4:[0-9]+]]
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__1
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i32** noundef nonnull align 4 dereferenceable(4) [[C:%.*]]) #[[ATTR2]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[C_ADDR:%.*]] = alloca i32**, align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32** [[C]], i32*** [[C_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = load i32**, i32*** [[C_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP1:%.*]] = load i32*, i32** [[TMP0]], align 4
-// CHECK3-NEXT:    call void @_Z4workPi(i32* noundef [[TMP1]]) #[[ATTR7]]
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@_Z4workPi
-// CHECK3-SAME: (i32* noundef [[C:%.*]]) #[[ATTR1]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[C_ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[ATOMIC_TEMP:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[ATOMIC_TEMP1:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    store i32* [[C]], i32** [[C_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = load i32*, i32** [[C_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP1:%.*]] = bitcast i32* [[TMP0]] to i8*
-// CHECK3-NEXT:    [[TMP2:%.*]] = bitcast i32* [[ATOMIC_TEMP]] to i8*
-// CHECK3-NEXT:    call void @__atomic_load(i32 noundef 4, i8* noundef [[TMP1]], i8* noundef [[TMP2]], i32 noundef 0) #[[ATTR7]]
-// CHECK3-NEXT:    br label [[ATOMIC_CONT:%.*]]
-// CHECK3:       atomic_cont:
-// CHECK3-NEXT:    [[TMP3:%.*]] = load i32, i32* [[ATOMIC_TEMP]], align 4
-// CHECK3-NEXT:    [[ADD:%.*]] = add nsw i32 [[TMP3]], 1
-// CHECK3-NEXT:    store i32 [[ADD]], i32* [[ATOMIC_TEMP1]], align 4
-// CHECK3-NEXT:    [[TMP4:%.*]] = bitcast i32* [[TMP0]] to i8*
-// CHECK3-NEXT:    [[TMP5:%.*]] = bitcast i32* [[ATOMIC_TEMP]] to i8*
-// CHECK3-NEXT:    [[TMP6:%.*]] = bitcast i32* [[ATOMIC_TEMP1]] to i8*
-// CHECK3-NEXT:    [[CALL:%.*]] = call noundef zeroext i1 @__atomic_compare_exchange(i32 noundef 4, i8* noundef [[TMP4]], i8* noundef [[TMP5]], i8* noundef [[TMP6]], i32 noundef 0, i32 noundef 0) #[[ATTR7]]
-// CHECK3-NEXT:    br i1 [[CALL]], label [[ATOMIC_EXIT:%.*]], label [[ATOMIC_CONT]]
-// CHECK3:       atomic_exit:
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__1_wrapper
-// CHECK3-SAME: (i16 noundef zeroext [[TMP0:%.*]], i32 noundef [[TMP1:%.*]]) #[[ATTR3]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTADDR:%.*]] = alloca i16, align 2
-// CHECK3-NEXT:    [[DOTADDR1:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[GLOBAL_ARGS:%.*]] = alloca i8**, align 4
-// CHECK3-NEXT:    store i16 [[TMP0]], i16* [[DOTADDR]], align 2
-// CHECK3-NEXT:    store i32 [[TMP1]], i32* [[DOTADDR1]], align 4
-// CHECK3-NEXT:    store i32 0, i32* [[DOTZERO_ADDR]], align 4
-// CHECK3-NEXT:    call void @__kmpc_get_shared_variables(i8*** [[GLOBAL_ARGS]])
-// CHECK3-NEXT:    [[TMP2:%.*]] = load i8**, i8*** [[GLOBAL_ARGS]], align 4
-// CHECK3-NEXT:    [[TMP3:%.*]] = getelementptr inbounds i8*, i8** [[TMP2]], i32 0
-// CHECK3-NEXT:    [[TMP4:%.*]] = bitcast i8** [[TMP3]] to i32***
-// CHECK3-NEXT:    [[TMP5:%.*]] = load i32**, i32*** [[TMP4]], align 4
-// CHECK3-NEXT:    call void @__omp_outlined__1(i32* [[DOTADDR1]], i32* [[DOTZERO_ADDR]], i32** [[TMP5]]) #[[ATTR4]]
-// CHECK3-NEXT:    ret void
-//
diff --git a/clang/test/OpenMP/nvptx_target_codegen.cpp b/clang/test/OpenMP/nvptx_target_codegen.cpp
diff --git a/clang/test/OpenMP/nvptx_target_parallel_codegen.cpp b/clang/test/OpenMP/nvptx_target_parallel_codegen.cpp
diff --git a/clang/test/OpenMP/nvptx_target_parallel_num_threads_codegen.cpp b/clang/test/OpenMP/nvptx_target_parallel_num_threads_codegen.cpp
diff --git a/clang/test/OpenMP/nvptx_target_parallel_reduction_codegen_tbaa_PR46146.cpp b/clang/test/OpenMP/nvptx_target_parallel_reduction_codegen_tbaa_PR46146.cpp
diff --git a/clang/test/OpenMP/nvptx_target_teams_codegen.cpp b/clang/test/OpenMP/nvptx_target_teams_codegen.cpp
@@ -4,7 +4,7 @@
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -x c++ -triple nvptx64-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s --check-prefix=CHECK1
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm-bc %s -o %t-x86-host.bc
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix=CHECK2
-// RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fexceptions -fcxx-exceptions -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix=CHECK3
+// RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fexceptions -fcxx-exceptions -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix=CHECK2
 // expected-no-diagnostics
 #ifndef HEADER
 #define HEADER
@@ -378,168 +378,3 @@ int bar(int n){
 // CHECK2-NEXT:    store i16 1, i16* [[TMP0]], align 2
 // CHECK2-NEXT:    ret void
 //
-//
-// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIcET_i_l23
-// CHECK3-SAME: (i32 noundef [[A:%.*]]) #[[ATTR0:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[A_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[A_CASTED:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    store i32 [[A]], i32* [[A_ADDR]], align 4
-// CHECK3-NEXT:    [[CONV:%.*]] = bitcast i32* [[A_ADDR]] to i8*
-// CHECK3-NEXT:    [[TMP0:%.*]] = call i32 @__kmpc_target_init(%struct.ident_t* @[[GLOB1:[0-9]+]], i8 1, i1 true, i1 true)
-// CHECK3-NEXT:    [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
-// CHECK3-NEXT:    br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
-// CHECK3:       user_code.entry:
-// CHECK3-NEXT:    [[TMP1:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB1]])
-// CHECK3-NEXT:    [[TMP2:%.*]] = load i8, i8* [[CONV]], align 1
-// CHECK3-NEXT:    [[CONV1:%.*]] = bitcast i32* [[A_CASTED]] to i8*
-// CHECK3-NEXT:    store i8 [[TMP2]], i8* [[CONV1]], align 1
-// CHECK3-NEXT:    [[TMP3:%.*]] = load i32, i32* [[A_CASTED]], align 4
-// CHECK3-NEXT:    store i32 0, i32* [[DOTZERO_ADDR]], align 4
-// CHECK3-NEXT:    store i32 [[TMP1]], i32* [[DOTTHREADID_TEMP_]], align 4
-// CHECK3-NEXT:    call void @__omp_outlined__(i32* [[DOTTHREADID_TEMP_]], i32* [[DOTZERO_ADDR]], i32 [[TMP3]]) #[[ATTR2:[0-9]+]]
-// CHECK3-NEXT:    call void @__kmpc_target_deinit(%struct.ident_t* @[[GLOB1]], i8 1, i1 true)
-// CHECK3-NEXT:    ret void
-// CHECK3:       worker.exit:
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i32 noundef [[A:%.*]]) #[[ATTR1:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[A_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32 [[A]], i32* [[A_ADDR]], align 4
-// CHECK3-NEXT:    [[CONV:%.*]] = bitcast i32* [[A_ADDR]] to i8*
-// CHECK3-NEXT:    store i8 49, i8* [[CONV]], align 1
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIcET_i_l28
-// CHECK3-SAME: (i32 noundef [[AA:%.*]]) #[[ATTR0]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[AA_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[AA_CASTED:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    store i32 [[AA]], i32* [[AA_ADDR]], align 4
-// CHECK3-NEXT:    [[CONV:%.*]] = bitcast i32* [[AA_ADDR]] to i16*
-// CHECK3-NEXT:    [[TMP0:%.*]] = call i32 @__kmpc_target_init(%struct.ident_t* @[[GLOB1]], i8 1, i1 true, i1 true)
-// CHECK3-NEXT:    [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
-// CHECK3-NEXT:    br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
-// CHECK3:       user_code.entry:
-// CHECK3-NEXT:    [[TMP1:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB1]])
-// CHECK3-NEXT:    [[TMP2:%.*]] = load i16, i16* [[CONV]], align 2
-// CHECK3-NEXT:    [[CONV1:%.*]] = bitcast i32* [[AA_CASTED]] to i16*
-// CHECK3-NEXT:    store i16 [[TMP2]], i16* [[CONV1]], align 2
-// CHECK3-NEXT:    [[TMP3:%.*]] = load i32, i32* [[AA_CASTED]], align 4
-// CHECK3-NEXT:    store i32 0, i32* [[DOTZERO_ADDR]], align 4
-// CHECK3-NEXT:    store i32 [[TMP1]], i32* [[DOTTHREADID_TEMP_]], align 4
-// CHECK3-NEXT:    call void @__omp_outlined__1(i32* [[DOTTHREADID_TEMP_]], i32* [[DOTZERO_ADDR]], i32 [[TMP3]]) #[[ATTR2]]
-// CHECK3-NEXT:    call void @__kmpc_target_deinit(%struct.ident_t* @[[GLOB1]], i8 1, i1 true)
-// CHECK3-NEXT:    ret void
-// CHECK3:       worker.exit:
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__1
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i32 noundef [[AA:%.*]]) #[[ATTR1]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[AA_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32 [[AA]], i32* [[AA_ADDR]], align 4
-// CHECK3-NEXT:    [[CONV:%.*]] = bitcast i32* [[AA_ADDR]] to i16*
-// CHECK3-NEXT:    store i16 1, i16* [[CONV]], align 2
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIcET_i_l33
-// CHECK3-SAME: (i32 noundef [[AA:%.*]]) #[[ATTR0]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[AA_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[AA_CASTED:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    store i32 [[AA]], i32* [[AA_ADDR]], align 4
-// CHECK3-NEXT:    [[CONV:%.*]] = bitcast i32* [[AA_ADDR]] to i16*
-// CHECK3-NEXT:    [[TMP0:%.*]] = call i32 @__kmpc_target_init(%struct.ident_t* @[[GLOB1]], i8 2, i1 false, i1 true)
-// CHECK3-NEXT:    [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
-// CHECK3-NEXT:    br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
-// CHECK3:       user_code.entry:
-// CHECK3-NEXT:    [[TMP1:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB2:[0-9]+]])
-// CHECK3-NEXT:    [[TMP2:%.*]] = load i16, i16* [[CONV]], align 2
-// CHECK3-NEXT:    [[CONV1:%.*]] = bitcast i32* [[AA_CASTED]] to i16*
-// CHECK3-NEXT:    store i16 [[TMP2]], i16* [[CONV1]], align 2
-// CHECK3-NEXT:    [[TMP3:%.*]] = load i32, i32* [[AA_CASTED]], align 4
-// CHECK3-NEXT:    store i32 0, i32* [[DOTZERO_ADDR]], align 4
-// CHECK3-NEXT:    store i32 [[TMP1]], i32* [[DOTTHREADID_TEMP_]], align 4
-// CHECK3-NEXT:    call void @__omp_outlined__2(i32* [[DOTTHREADID_TEMP_]], i32* [[DOTZERO_ADDR]], i32 [[TMP3]]) #[[ATTR2]]
-// CHECK3-NEXT:    call void @__kmpc_target_deinit(%struct.ident_t* @[[GLOB1]], i8 2, i1 true)
-// CHECK3-NEXT:    ret void
-// CHECK3:       worker.exit:
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__2
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i32 noundef [[AA:%.*]]) #[[ATTR1]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[AA_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [1 x i8*], align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32 [[AA]], i32* [[AA_ADDR]], align 4
-// CHECK3-NEXT:    [[CONV:%.*]] = bitcast i32* [[AA_ADDR]] to i16*
-// CHECK3-NEXT:    [[TMP0:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[CAPTURED_VARS_ADDRS]], i32 0, i32 0
-// CHECK3-NEXT:    [[TMP1:%.*]] = bitcast i16* [[CONV]] to i8*
-// CHECK3-NEXT:    store i8* [[TMP1]], i8** [[TMP0]], align 4
-// CHECK3-NEXT:    [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4
-// CHECK3-NEXT:    [[TMP4:%.*]] = bitcast [1 x i8*]* [[CAPTURED_VARS_ADDRS]] to i8**
-// CHECK3-NEXT:    call void @__kmpc_parallel_51(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 1, i32 -1, i32 -1, i8* bitcast (void (i32*, i32*, i16*)* @__omp_outlined__3 to i8*), i8* null, i8** [[TMP4]], i32 1)
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__3
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i16* noundef nonnull align 2 dereferenceable(2) [[AA:%.*]]) #[[ATTR1]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[AA_ADDR:%.*]] = alloca i16*, align 4
-// CHECK3-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [1 x i8*], align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i16* [[AA]], i16** [[AA_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = load i16*, i16** [[AA_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP1:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[CAPTURED_VARS_ADDRS]], i32 0, i32 0
-// CHECK3-NEXT:    [[TMP2:%.*]] = bitcast i16* [[TMP0]] to i8*
-// CHECK3-NEXT:    store i8* [[TMP2]], i8** [[TMP1]], align 4
-// CHECK3-NEXT:    [[TMP3:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP3]], align 4
-// CHECK3-NEXT:    [[TMP5:%.*]] = bitcast [1 x i8*]* [[CAPTURED_VARS_ADDRS]] to i8**
-// CHECK3-NEXT:    call void @__kmpc_parallel_51(%struct.ident_t* @[[GLOB2]], i32 [[TMP4]], i32 1, i32 -1, i32 -1, i8* bitcast (void (i32*, i32*, i16*)* @__omp_outlined__4 to i8*), i8* null, i8** [[TMP5]], i32 1)
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__4
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i16* noundef nonnull align 2 dereferenceable(2) [[AA:%.*]]) #[[ATTR1]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[AA_ADDR:%.*]] = alloca i16*, align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i16* [[AA]], i16** [[AA_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = load i16*, i16** [[AA_ADDR]], align 4
-// CHECK3-NEXT:    store i16 1, i16* [[TMP0]], align 2
-// CHECK3-NEXT:    ret void
-//
diff --git a/clang/test/OpenMP/nvptx_target_teams_distribute_codegen.cpp b/clang/test/OpenMP/nvptx_target_teams_distribute_codegen.cpp
@@ -4,7 +4,7 @@
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -x c++ -triple nvptx64-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s --check-prefix=CHECK1
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm-bc %s -o %t-x86-host.bc
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix=CHECK2
-// RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fexceptions -fcxx-exceptions -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix=CHECK3
+// RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fexceptions -fcxx-exceptions -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix=CHECK2
 // expected-no-diagnostics
 #ifndef HEADER
 #define HEADER
@@ -275,125 +275,3 @@ int bar(int n){
 // CHECK2-NEXT:    call void @__omp_outlined__1(i32* [[DOTADDR1]], i32* [[DOTZERO_ADDR]], i32* [[TMP5]]) #[[ATTR5]]
 // CHECK2-NEXT:    ret void
 //
-//
-// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIcET_i_l16
-// CHECK3-SAME: () #[[ATTR0:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = call i32 @__kmpc_target_init(%struct.ident_t* @[[GLOB1:[0-9]+]], i8 1, i1 true, i1 true)
-// CHECK3-NEXT:    [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
-// CHECK3-NEXT:    br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
-// CHECK3:       user_code.entry:
-// CHECK3-NEXT:    [[TMP1:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB1]])
-// CHECK3-NEXT:    store i32 0, i32* [[DOTZERO_ADDR]], align 4
-// CHECK3-NEXT:    store i32 [[TMP1]], i32* [[DOTTHREADID_TEMP_]], align 4
-// CHECK3-NEXT:    call void @__omp_outlined__(i32* [[DOTTHREADID_TEMP_]], i32* [[DOTZERO_ADDR]]) #[[ATTR5:[0-9]+]]
-// CHECK3-NEXT:    call void @__kmpc_target_deinit(%struct.ident_t* @[[GLOB1]], i8 1, i1 true)
-// CHECK3-NEXT:    ret void
-// CHECK3:       worker.exit:
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR1:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTOMP_IV:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[TMP:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTOMP_LB:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTOMP_UB:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [1 x i8*], align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    [[I:%.*]] = call align 8 i8* @__kmpc_alloc_shared(i32 4)
-// CHECK3-NEXT:    [[I_ON_STACK:%.*]] = bitcast i8* [[I]] to i32*
-// CHECK3-NEXT:    store i32 0, i32* [[DOTOMP_LB]], align 4
-// CHECK3-NEXT:    store i32 9, i32* [[DOTOMP_UB]], align 4
-// CHECK3-NEXT:    store i32 1, i32* [[DOTOMP_STRIDE]], align 4
-// CHECK3-NEXT:    store i32 0, i32* [[DOTOMP_IS_LAST]], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4
-// CHECK3-NEXT:    call void @__kmpc_distribute_static_init_4(%struct.ident_t* @[[GLOB2:[0-9]+]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1)
-// CHECK3-NEXT:    [[TMP2:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4
-// CHECK3-NEXT:    [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 9
-// CHECK3-NEXT:    br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]]
-// CHECK3:       cond.true:
-// CHECK3-NEXT:    br label [[COND_END:%.*]]
-// CHECK3:       cond.false:
-// CHECK3-NEXT:    [[TMP3:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4
-// CHECK3-NEXT:    br label [[COND_END]]
-// CHECK3:       cond.end:
-// CHECK3-NEXT:    [[COND:%.*]] = phi i32 [ 9, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ]
-// CHECK3-NEXT:    store i32 [[COND]], i32* [[DOTOMP_UB]], align 4
-// CHECK3-NEXT:    [[TMP4:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4
-// CHECK3-NEXT:    store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4
-// CHECK3-NEXT:    br label [[OMP_INNER_FOR_COND:%.*]]
-// CHECK3:       omp.inner.for.cond:
-// CHECK3-NEXT:    [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4
-// CHECK3-NEXT:    [[TMP6:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4
-// CHECK3-NEXT:    [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]]
-// CHECK3-NEXT:    br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]]
-// CHECK3:       omp.inner.for.body:
-// CHECK3-NEXT:    [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4
-// CHECK3-NEXT:    [[MUL:%.*]] = mul nsw i32 [[TMP7]], 1
-// CHECK3-NEXT:    [[ADD:%.*]] = add nsw i32 0, [[MUL]]
-// CHECK3-NEXT:    store i32 [[ADD]], i32* [[I_ON_STACK]], align 4
-// CHECK3-NEXT:    [[TMP8:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[CAPTURED_VARS_ADDRS]], i32 0, i32 0
-// CHECK3-NEXT:    [[TMP9:%.*]] = bitcast i32* [[I_ON_STACK]] to i8*
-// CHECK3-NEXT:    store i8* [[TMP9]], i8** [[TMP8]], align 4
-// CHECK3-NEXT:    [[TMP10:%.*]] = bitcast [1 x i8*]* [[CAPTURED_VARS_ADDRS]] to i8**
-// CHECK3-NEXT:    call void @__kmpc_parallel_51(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 1, i32 -1, i32 -1, i8* bitcast (void (i32*, i32*, i32*)* @__omp_outlined__1 to i8*), i8* bitcast (void (i16, i32)* @__omp_outlined__1_wrapper to i8*), i8** [[TMP10]], i32 1)
-// CHECK3-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
-// CHECK3:       omp.body.continue:
-// CHECK3-NEXT:    br label [[OMP_INNER_FOR_INC:%.*]]
-// CHECK3:       omp.inner.for.inc:
-// CHECK3-NEXT:    [[TMP11:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4
-// CHECK3-NEXT:    [[ADD2:%.*]] = add nsw i32 [[TMP11]], 1
-// CHECK3-NEXT:    store i32 [[ADD2]], i32* [[DOTOMP_IV]], align 4
-// CHECK3-NEXT:    br label [[OMP_INNER_FOR_COND]]
-// CHECK3:       omp.inner.for.end:
-// CHECK3-NEXT:    br label [[OMP_LOOP_EXIT:%.*]]
-// CHECK3:       omp.loop.exit:
-// CHECK3-NEXT:    call void @__kmpc_distribute_static_fini(%struct.ident_t* @[[GLOB2]], i32 [[TMP1]])
-// CHECK3-NEXT:    call void @__kmpc_free_shared(i8* [[I]], i32 4)
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__1
-// CHECK3-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i32* noundef nonnull align 4 dereferenceable(4) [[I:%.*]]) #[[ATTR1]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    [[I_ADDR:%.*]] = alloca i32*, align 4
-// CHECK3-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 4
-// CHECK3-NEXT:    store i32* [[I]], i32** [[I_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP0:%.*]] = load i32*, i32** [[I_ADDR]], align 4
-// CHECK3-NEXT:    [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4
-// CHECK3-NEXT:    [[INC:%.*]] = add nsw i32 [[TMP1]], 1
-// CHECK3-NEXT:    store i32 [[INC]], i32* [[TMP0]], align 4
-// CHECK3-NEXT:    ret void
-//
-//
-// CHECK3-LABEL: define {{[^@]+}}@__omp_outlined__1_wrapper
-// CHECK3-SAME: (i16 noundef zeroext [[TMP0:%.*]], i32 noundef [[TMP1:%.*]]) #[[ATTR3:[0-9]+]] {
-// CHECK3-NEXT:  entry:
-// CHECK3-NEXT:    [[DOTADDR:%.*]] = alloca i16, align 2
-// CHECK3-NEXT:    [[DOTADDR1:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
-// CHECK3-NEXT:    [[GLOBAL_ARGS:%.*]] = alloca i8**, align 4
-// CHECK3-NEXT:    store i16 [[TMP0]], i16* [[DOTADDR]], align 2
-// CHECK3-NEXT:    store i32 [[TMP1]], i32* [[DOTADDR1]], align 4
-// CHECK3-NEXT:    store i32 0, i32* [[DOTZERO_ADDR]], align 4
-// CHECK3-NEXT:    call void @__kmpc_get_shared_variables(i8*** [[GLOBAL_ARGS]])
-// CHECK3-NEXT:    [[TMP2:%.*]] = load i8**, i8*** [[GLOBAL_ARGS]], align 4
-// CHECK3-NEXT:    [[TMP3:%.*]] = getelementptr inbounds i8*, i8** [[TMP2]], i32 0
-// CHECK3-NEXT:    [[TMP4:%.*]] = bitcast i8** [[TMP3]] to i32**
-// CHECK3-NEXT:    [[TMP5:%.*]] = load i32*, i32** [[TMP4]], align 4
-// CHECK3-NEXT:    call void @__omp_outlined__1(i32* [[DOTADDR1]], i32* [[DOTZERO_ADDR]], i32* [[TMP5]]) #[[ATTR5]]
-// CHECK3-NEXT:    ret void
-//
diff --git a/clang/test/OpenMP/nvptx_target_teams_distribute_parallel_for_codegen.cpp b/clang/test/OpenMP/nvptx_target_teams_distribute_parallel_for_codegen.cpp
diff --git a/clang/test/OpenMP/nvptx_target_teams_distribute_parallel_for_generic_mode_codegen.cpp b/clang/test/OpenMP/nvptx_target_teams_distribute_parallel_for_generic_mode_codegen.cpp
diff --git a/clang/test/OpenMP/nvptx_target_teams_distribute_parallel_for_simd_codegen.cpp b/clang/test/OpenMP/nvptx_target_teams_distribute_parallel_for_simd_codegen.cpp
diff --git a/clang/test/OpenMP/ordered_codegen.cpp b/clang/test/OpenMP/ordered_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_copyin_codegen.cpp b/clang/test/OpenMP/parallel_copyin_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_firstprivate_codegen.cpp b/clang/test/OpenMP/parallel_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_for_codegen.cpp b/clang/test/OpenMP/parallel_for_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_for_lastprivate_conditional.cpp b/clang/test/OpenMP/parallel_for_lastprivate_conditional.cpp
diff --git a/clang/test/OpenMP/parallel_for_linear_codegen.cpp b/clang/test/OpenMP/parallel_for_linear_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_for_reduction_task_codegen.cpp b/clang/test/OpenMP/parallel_for_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_for_simd_aligned_codegen.cpp b/clang/test/OpenMP/parallel_for_simd_aligned_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_if_codegen.cpp b/clang/test/OpenMP/parallel_if_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_master_codegen.cpp b/clang/test/OpenMP/parallel_master_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_master_reduction_task_codegen.cpp b/clang/test/OpenMP/parallel_master_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_master_taskloop_codegen.cpp b/clang/test/OpenMP/parallel_master_taskloop_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_master_taskloop_lastprivate_codegen.cpp b/clang/test/OpenMP/parallel_master_taskloop_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_master_taskloop_simd_codegen.cpp b/clang/test/OpenMP/parallel_master_taskloop_simd_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_master_taskloop_simd_lastprivate_codegen.cpp b/clang/test/OpenMP/parallel_master_taskloop_simd_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_private_codegen.cpp b/clang/test/OpenMP/parallel_private_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_reduction_codegen.cpp b/clang/test/OpenMP/parallel_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_reduction_task_codegen.cpp b/clang/test/OpenMP/parallel_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/parallel_sections_codegen.cpp b/clang/test/OpenMP/parallel_sections_codegen.cpp
@@ -1,7 +1,7 @@
 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --function-signature --include-generated-funcs --replace-value-regex "__omp_offloading_[0-9a-z]+_[0-9a-z]+" "reduction_size[.].+[.]" "pl_cond[.].+[.|,]" --prefix-filecheck-ir-name _
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -x c++ -emit-llvm -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -o - %s | FileCheck %s --check-prefix=CHECK1
 // RUN: %clang_cc1 -no-opaque-pointers -fopenmp -x c++ -std=c++11 -fexceptions -fcxx-exceptions -triple x86_64-unknown-unknown -emit-pch -o %t %s
-// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -x c++ -std=c++11 -include-pch %t -fsyntax-only -verify %s -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -emit-llvm -o - | FileCheck %s --check-prefix=CHECK2
+// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -x c++ -std=c++11 -include-pch %t -fsyntax-only -verify %s -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -emit-llvm -o - | FileCheck %s --check-prefix=CHECK1
 
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp-simd -x c++ -emit-llvm -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -o - %s | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
 // RUN: %clang_cc1 -no-opaque-pointers -fopenmp-simd -x c++ -std=c++11 -fexceptions -fcxx-exceptions -triple x86_64-unknown-unknown -emit-pch -o %t %s
@@ -200,165 +200,3 @@ int main() {
 // CHECK1-NEXT:    call void @__clang_call_terminate(i8* [[TMP11]]) #[[ATTR7]]
 // CHECK1-NEXT:    unreachable
 //
-//
-// CHECK2-LABEL: define {{[^@]+}}@_Z3foov
-// CHECK2-SAME: () #[[ATTR0:[0-9]+]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    call void @_Z8mayThrowv()
-// CHECK2-NEXT:    ret void
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@_Z3barv
-// CHECK2-SAME: () #[[ATTR0]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    call void @_Z8mayThrowv()
-// CHECK2-NEXT:    ret void
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@main
-// CHECK2-SAME: () #[[ATTR2:[0-9]+]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[RETVAL:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    store i32 0, i32* [[RETVAL]], align 4
-// CHECK2-NEXT:    call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB2:[0-9]+]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined. to void (i32*, i32*, ...)*))
-// CHECK2-NEXT:    [[CALL:%.*]] = call noundef i32 @_Z5tmainIiET_v()
-// CHECK2-NEXT:    ret i32 [[CALL]]
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@.omp_outlined.
-// CHECK2-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR3:[0-9]+]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    [[DOTOMP_SECTIONS_LB_:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_SECTIONS_UB_:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_SECTIONS_ST_:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_SECTIONS_IL_:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_SECTIONS_IV_:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8
-// CHECK2-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8
-// CHECK2-NEXT:    store i32 0, i32* [[DOTOMP_SECTIONS_LB_]], align 4
-// CHECK2-NEXT:    store i32 1, i32* [[DOTOMP_SECTIONS_UB_]], align 4
-// CHECK2-NEXT:    store i32 1, i32* [[DOTOMP_SECTIONS_ST_]], align 4
-// CHECK2-NEXT:    store i32 0, i32* [[DOTOMP_SECTIONS_IL_]], align 4
-// CHECK2-NEXT:    [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8
-// CHECK2-NEXT:    [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4
-// CHECK2-NEXT:    call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1:[0-9]+]], i32 [[TMP1]], i32 34, i32* [[DOTOMP_SECTIONS_IL_]], i32* [[DOTOMP_SECTIONS_LB_]], i32* [[DOTOMP_SECTIONS_UB_]], i32* [[DOTOMP_SECTIONS_ST_]], i32 1, i32 1)
-// CHECK2-NEXT:    [[TMP2:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_UB_]], align 4
-// CHECK2-NEXT:    [[TMP3:%.*]] = icmp slt i32 [[TMP2]], 1
-// CHECK2-NEXT:    [[TMP4:%.*]] = select i1 [[TMP3]], i32 [[TMP2]], i32 1
-// CHECK2-NEXT:    store i32 [[TMP4]], i32* [[DOTOMP_SECTIONS_UB_]], align 4
-// CHECK2-NEXT:    [[TMP5:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_LB_]], align 4
-// CHECK2-NEXT:    store i32 [[TMP5]], i32* [[DOTOMP_SECTIONS_IV_]], align 4
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_COND:%.*]]
-// CHECK2:       omp.inner.for.cond:
-// CHECK2-NEXT:    [[TMP6:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_IV_]], align 4
-// CHECK2-NEXT:    [[TMP7:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_UB_]], align 4
-// CHECK2-NEXT:    [[CMP:%.*]] = icmp sle i32 [[TMP6]], [[TMP7]]
-// CHECK2-NEXT:    br i1 [[CMP]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]]
-// CHECK2:       omp.inner.for.body:
-// CHECK2-NEXT:    [[TMP8:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_IV_]], align 4
-// CHECK2-NEXT:    switch i32 [[TMP8]], label [[DOTOMP_SECTIONS_EXIT:%.*]] [
-// CHECK2-NEXT:    i32 0, label [[DOTOMP_SECTIONS_CASE:%.*]]
-// CHECK2-NEXT:    i32 1, label [[DOTOMP_SECTIONS_CASE1:%.*]]
-// CHECK2-NEXT:    ]
-// CHECK2:       .omp.sections.case:
-// CHECK2-NEXT:    invoke void @_Z3foov()
-// CHECK2-NEXT:    to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]]
-// CHECK2:       invoke.cont:
-// CHECK2-NEXT:    br label [[DOTOMP_SECTIONS_EXIT]]
-// CHECK2:       .omp.sections.case1:
-// CHECK2-NEXT:    invoke void @_Z3barv()
-// CHECK2-NEXT:    to label [[INVOKE_CONT2:%.*]] unwind label [[TERMINATE_LPAD]]
-// CHECK2:       invoke.cont2:
-// CHECK2-NEXT:    br label [[DOTOMP_SECTIONS_EXIT]]
-// CHECK2:       .omp.sections.exit:
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_INC:%.*]]
-// CHECK2:       omp.inner.for.inc:
-// CHECK2-NEXT:    [[TMP9:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_IV_]], align 4
-// CHECK2-NEXT:    [[INC:%.*]] = add nsw i32 [[TMP9]], 1
-// CHECK2-NEXT:    store i32 [[INC]], i32* [[DOTOMP_SECTIONS_IV_]], align 4
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_COND]]
-// CHECK2:       omp.inner.for.end:
-// CHECK2-NEXT:    call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]])
-// CHECK2-NEXT:    ret void
-// CHECK2:       terminate.lpad:
-// CHECK2-NEXT:    [[TMP10:%.*]] = landingpad { i8*, i32 }
-// CHECK2-NEXT:    catch i8* null
-// CHECK2-NEXT:    [[TMP11:%.*]] = extractvalue { i8*, i32 } [[TMP10]], 0
-// CHECK2-NEXT:    call void @__clang_call_terminate(i8* [[TMP11]]) #[[ATTR7:[0-9]+]]
-// CHECK2-NEXT:    unreachable
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@__clang_call_terminate
-// CHECK2-SAME: (i8* [[TMP0:%.*]]) #[[ATTR4:[0-9]+]] comdat {
-// CHECK2-NEXT:    [[TMP2:%.*]] = call i8* @__cxa_begin_catch(i8* [[TMP0]]) #[[ATTR5:[0-9]+]]
-// CHECK2-NEXT:    call void @_ZSt9terminatev() #[[ATTR7]]
-// CHECK2-NEXT:    unreachable
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@_Z5tmainIiET_v
-// CHECK2-SAME: () #[[ATTR6:[0-9]+]] comdat {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB2]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined..1 to void (i32*, i32*, ...)*))
-// CHECK2-NEXT:    ret i32 0
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@.omp_outlined..1
-// CHECK2-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    [[DOTOMP_SECTIONS_LB_:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_SECTIONS_UB_:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_SECTIONS_ST_:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_SECTIONS_IL_:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_SECTIONS_IV_:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8
-// CHECK2-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8
-// CHECK2-NEXT:    store i32 0, i32* [[DOTOMP_SECTIONS_LB_]], align 4
-// CHECK2-NEXT:    store i32 0, i32* [[DOTOMP_SECTIONS_UB_]], align 4
-// CHECK2-NEXT:    store i32 1, i32* [[DOTOMP_SECTIONS_ST_]], align 4
-// CHECK2-NEXT:    store i32 0, i32* [[DOTOMP_SECTIONS_IL_]], align 4
-// CHECK2-NEXT:    [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8
-// CHECK2-NEXT:    [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4
-// CHECK2-NEXT:    call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 34, i32* [[DOTOMP_SECTIONS_IL_]], i32* [[DOTOMP_SECTIONS_LB_]], i32* [[DOTOMP_SECTIONS_UB_]], i32* [[DOTOMP_SECTIONS_ST_]], i32 1, i32 1)
-// CHECK2-NEXT:    [[TMP2:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_UB_]], align 4
-// CHECK2-NEXT:    [[TMP3:%.*]] = icmp slt i32 [[TMP2]], 0
-// CHECK2-NEXT:    [[TMP4:%.*]] = select i1 [[TMP3]], i32 [[TMP2]], i32 0
-// CHECK2-NEXT:    store i32 [[TMP4]], i32* [[DOTOMP_SECTIONS_UB_]], align 4
-// CHECK2-NEXT:    [[TMP5:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_LB_]], align 4
-// CHECK2-NEXT:    store i32 [[TMP5]], i32* [[DOTOMP_SECTIONS_IV_]], align 4
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_COND:%.*]]
-// CHECK2:       omp.inner.for.cond:
-// CHECK2-NEXT:    [[TMP6:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_IV_]], align 4
-// CHECK2-NEXT:    [[TMP7:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_UB_]], align 4
-// CHECK2-NEXT:    [[CMP:%.*]] = icmp sle i32 [[TMP6]], [[TMP7]]
-// CHECK2-NEXT:    br i1 [[CMP]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]]
-// CHECK2:       omp.inner.for.body:
-// CHECK2-NEXT:    [[TMP8:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_IV_]], align 4
-// CHECK2-NEXT:    switch i32 [[TMP8]], label [[DOTOMP_SECTIONS_EXIT:%.*]] [
-// CHECK2-NEXT:    i32 0, label [[DOTOMP_SECTIONS_CASE:%.*]]
-// CHECK2-NEXT:    ]
-// CHECK2:       .omp.sections.case:
-// CHECK2-NEXT:    invoke void @_Z3foov()
-// CHECK2-NEXT:    to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]]
-// CHECK2:       invoke.cont:
-// CHECK2-NEXT:    br label [[DOTOMP_SECTIONS_EXIT]]
-// CHECK2:       .omp.sections.exit:
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_INC:%.*]]
-// CHECK2:       omp.inner.for.inc:
-// CHECK2-NEXT:    [[TMP9:%.*]] = load i32, i32* [[DOTOMP_SECTIONS_IV_]], align 4
-// CHECK2-NEXT:    [[INC:%.*]] = add nsw i32 [[TMP9]], 1
-// CHECK2-NEXT:    store i32 [[INC]], i32* [[DOTOMP_SECTIONS_IV_]], align 4
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_COND]]
-// CHECK2:       omp.inner.for.end:
-// CHECK2-NEXT:    call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]])
-// CHECK2-NEXT:    ret void
-// CHECK2:       terminate.lpad:
-// CHECK2-NEXT:    [[TMP10:%.*]] = landingpad { i8*, i32 }
-// CHECK2-NEXT:    catch i8* null
-// CHECK2-NEXT:    [[TMP11:%.*]] = extractvalue { i8*, i32 } [[TMP10]], 0
-// CHECK2-NEXT:    call void @__clang_call_terminate(i8* [[TMP11]]) #[[ATTR7]]
-// CHECK2-NEXT:    unreachable
-//
diff --git a/clang/test/OpenMP/parallel_sections_reduction_task_codegen.cpp b/clang/test/OpenMP/parallel_sections_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/sections_firstprivate_codegen.cpp b/clang/test/OpenMP/sections_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/sections_lastprivate_codegen.cpp b/clang/test/OpenMP/sections_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/sections_private_codegen.cpp b/clang/test/OpenMP/sections_private_codegen.cpp
diff --git a/clang/test/OpenMP/sections_reduction_codegen.cpp b/clang/test/OpenMP/sections_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/sections_reduction_task_codegen.cpp b/clang/test/OpenMP/sections_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/single_codegen.cpp b/clang/test/OpenMP/single_codegen.cpp
diff --git a/clang/test/OpenMP/single_firstprivate_codegen.cpp b/clang/test/OpenMP/single_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/single_private_codegen.cpp b/clang/test/OpenMP/single_private_codegen.cpp
diff --git a/clang/test/OpenMP/target_codegen_global_capture.cpp b/clang/test/OpenMP/target_codegen_global_capture.cpp
diff --git a/clang/test/OpenMP/target_map_codegen_03.cpp b/clang/test/OpenMP/target_map_codegen_03.cpp
diff --git a/clang/test/OpenMP/target_parallel_codegen.cpp b/clang/test/OpenMP/target_parallel_codegen.cpp
diff --git a/clang/test/OpenMP/target_parallel_for_codegen.cpp b/clang/test/OpenMP/target_parallel_for_codegen.cpp
diff --git a/clang/test/OpenMP/target_parallel_for_reduction_task_codegen.cpp b/clang/test/OpenMP/target_parallel_for_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/target_parallel_for_simd_codegen.cpp b/clang/test/OpenMP/target_parallel_for_simd_codegen.cpp
diff --git a/clang/test/OpenMP/target_parallel_if_codegen.cpp b/clang/test/OpenMP/target_parallel_if_codegen.cpp
diff --git a/clang/test/OpenMP/target_parallel_num_threads_codegen.cpp b/clang/test/OpenMP/target_parallel_num_threads_codegen.cpp
diff --git a/clang/test/OpenMP/target_parallel_reduction_task_codegen.cpp b/clang/test/OpenMP/target_parallel_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_codegen.cpp b/clang/test/OpenMP/target_teams_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_collapse_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_collapse_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_dist_schedule_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_dist_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_firstprivate_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_lastprivate_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_collapse_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_collapse_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_dist_schedule_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_dist_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_firstprivate_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_if_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_if_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_lastprivate_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_order_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_order_codegen.cpp
@@ -1,7 +1,7 @@
 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --function-signature --include-generated-funcs --replace-value-regex "__omp_offloading_[0-9a-z]+_[0-9a-z]+" "reduction_size[.].+[.]" "pl_cond[.].+[.|,]" --prefix-filecheck-ir-name _
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-version=50 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix=CHECK1
 // RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-version=50 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
-// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-version=50 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK2
+// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-version=50 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK1
 
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp-simd -fopenmp-version=50 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
 // RUN: %clang_cc1 -no-opaque-pointers -fopenmp-simd -fopenmp-version=50 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
@@ -179,162 +179,3 @@ void gtid_test() {
 // CHECK1-NEXT:    call void @__tgt_register_requires(i64 1)
 // CHECK1-NEXT:    ret void
 //
-//
-// CHECK2-LABEL: define {{[^@]+}}@_Z9gtid_testv
-// CHECK2-SAME: () #[[ATTR0:[0-9]+]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[TMP:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    call void @__kmpc_push_target_tripcount_mapper(%struct.ident_t* @[[GLOB3:[0-9]+]], i64 -1, i64 100)
-// CHECK2-NEXT:    [[TMP0:%.*]] = call i32 @__tgt_target_teams_mapper(%struct.ident_t* @[[GLOB3]], i64 -1, i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9gtid_testv_l16.region_id, i32 0, i8** null, i8** null, i64* null, i64* null, i8** null, i8** null, i32 0, i32 0)
-// CHECK2-NEXT:    [[TMP1:%.*]] = icmp ne i32 [[TMP0]], 0
-// CHECK2-NEXT:    br i1 [[TMP1]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]]
-// CHECK2:       omp_offload.failed:
-// CHECK2-NEXT:    call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9gtid_testv_l16() #[[ATTR2:[0-9]+]]
-// CHECK2-NEXT:    br label [[OMP_OFFLOAD_CONT]]
-// CHECK2:       omp_offload.cont:
-// CHECK2-NEXT:    ret void
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9gtid_testv_l16
-// CHECK2-SAME: () #[[ATTR1:[0-9]+]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined. to void (i32*, i32*, ...)*))
-// CHECK2-NEXT:    ret void
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@.omp_outlined.
-// CHECK2-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR1]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    [[DOTOMP_IV:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[TMP:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[I:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8
-// CHECK2-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8
-// CHECK2-NEXT:    store i32 0, i32* [[DOTOMP_COMB_LB]], align 4
-// CHECK2-NEXT:    store i32 99, i32* [[DOTOMP_COMB_UB]], align 4
-// CHECK2-NEXT:    store i32 1, i32* [[DOTOMP_STRIDE]], align 4
-// CHECK2-NEXT:    store i32 0, i32* [[DOTOMP_IS_LAST]], align 4
-// CHECK2-NEXT:    [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8
-// CHECK2-NEXT:    [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4
-// CHECK2-NEXT:    call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1:[0-9]+]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1)
-// CHECK2-NEXT:    [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4
-// CHECK2-NEXT:    [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99
-// CHECK2-NEXT:    br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]]
-// CHECK2:       cond.true:
-// CHECK2-NEXT:    br label [[COND_END:%.*]]
-// CHECK2:       cond.false:
-// CHECK2-NEXT:    [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4
-// CHECK2-NEXT:    br label [[COND_END]]
-// CHECK2:       cond.end:
-// CHECK2-NEXT:    [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ]
-// CHECK2-NEXT:    store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4
-// CHECK2-NEXT:    [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4
-// CHECK2-NEXT:    store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_COND:%.*]]
-// CHECK2:       omp.inner.for.cond:
-// CHECK2-NEXT:    [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4
-// CHECK2-NEXT:    [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4
-// CHECK2-NEXT:    [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]]
-// CHECK2-NEXT:    br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]]
-// CHECK2:       omp.inner.for.body:
-// CHECK2-NEXT:    [[TMP7:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4
-// CHECK2-NEXT:    [[TMP8:%.*]] = zext i32 [[TMP7]] to i64
-// CHECK2-NEXT:    [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4
-// CHECK2-NEXT:    [[TMP10:%.*]] = zext i32 [[TMP9]] to i64
-// CHECK2-NEXT:    call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..1 to void (i32*, i32*, ...)*), i64 [[TMP8]], i64 [[TMP10]])
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_INC:%.*]]
-// CHECK2:       omp.inner.for.inc:
-// CHECK2-NEXT:    [[TMP11:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4
-// CHECK2-NEXT:    [[TMP12:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4
-// CHECK2-NEXT:    [[ADD:%.*]] = add nsw i32 [[TMP11]], [[TMP12]]
-// CHECK2-NEXT:    store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_COND]]
-// CHECK2:       omp.inner.for.end:
-// CHECK2-NEXT:    br label [[OMP_LOOP_EXIT:%.*]]
-// CHECK2:       omp.loop.exit:
-// CHECK2-NEXT:    call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]])
-// CHECK2-NEXT:    ret void
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@.omp_outlined..1
-// CHECK2-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR1]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8
-// CHECK2-NEXT:    [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8
-// CHECK2-NEXT:    [[DOTOMP_IV:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[TMP:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_LB:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_UB:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[I:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8
-// CHECK2-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8
-// CHECK2-NEXT:    store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8
-// CHECK2-NEXT:    store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8
-// CHECK2-NEXT:    store i32 0, i32* [[DOTOMP_LB]], align 4
-// CHECK2-NEXT:    store i32 99, i32* [[DOTOMP_UB]], align 4
-// CHECK2-NEXT:    [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8
-// CHECK2-NEXT:    [[CONV:%.*]] = trunc i64 [[TMP0]] to i32
-// CHECK2-NEXT:    [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8
-// CHECK2-NEXT:    [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32
-// CHECK2-NEXT:    store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4
-// CHECK2-NEXT:    store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4
-// CHECK2-NEXT:    store i32 1, i32* [[DOTOMP_STRIDE]], align 4
-// CHECK2-NEXT:    store i32 0, i32* [[DOTOMP_IS_LAST]], align 4
-// CHECK2-NEXT:    [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8
-// CHECK2-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4
-// CHECK2-NEXT:    call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2:[0-9]+]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1)
-// CHECK2-NEXT:    [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4
-// CHECK2-NEXT:    [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99
-// CHECK2-NEXT:    br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]]
-// CHECK2:       cond.true:
-// CHECK2-NEXT:    br label [[COND_END:%.*]]
-// CHECK2:       cond.false:
-// CHECK2-NEXT:    [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4
-// CHECK2-NEXT:    br label [[COND_END]]
-// CHECK2:       cond.end:
-// CHECK2-NEXT:    [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ]
-// CHECK2-NEXT:    store i32 [[COND]], i32* [[DOTOMP_UB]], align 4
-// CHECK2-NEXT:    [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4
-// CHECK2-NEXT:    store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_COND:%.*]]
-// CHECK2:       omp.inner.for.cond:
-// CHECK2-NEXT:    [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4, !llvm.access.group !4
-// CHECK2-NEXT:    [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4, !llvm.access.group !4
-// CHECK2-NEXT:    [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]]
-// CHECK2-NEXT:    br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]]
-// CHECK2:       omp.inner.for.body:
-// CHECK2-NEXT:    [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4, !llvm.access.group !4
-// CHECK2-NEXT:    [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1
-// CHECK2-NEXT:    [[ADD:%.*]] = add nsw i32 0, [[MUL]]
-// CHECK2-NEXT:    store i32 [[ADD]], i32* [[I]], align 4, !llvm.access.group !4
-// CHECK2-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
-// CHECK2:       omp.body.continue:
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_INC:%.*]]
-// CHECK2:       omp.inner.for.inc:
-// CHECK2-NEXT:    [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4, !llvm.access.group !4
-// CHECK2-NEXT:    [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1
-// CHECK2-NEXT:    store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4, !llvm.access.group !4
-// CHECK2-NEXT:    br label [[OMP_INNER_FOR_COND]], !llvm.loop [[LOOP5:![0-9]+]]
-// CHECK2:       omp.inner.for.end:
-// CHECK2-NEXT:    br label [[OMP_LOOP_EXIT:%.*]]
-// CHECK2:       omp.loop.exit:
-// CHECK2-NEXT:    call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]])
-// CHECK2-NEXT:    ret void
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@.omp_offloading.requires_reg
-// CHECK2-SAME: () #[[ATTR3:[0-9]+]] section ".text.startup" {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    call void @__tgt_register_requires(i64 1)
-// CHECK2-NEXT:    ret void
-//
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_private_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_private_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_proc_bind_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_proc_bind_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_reduction_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_reduction_task_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_schedule_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_collapse_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_collapse_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_dist_schedule_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_dist_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_firstprivate_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_if_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_if_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_lastprivate_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_private_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_private_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_proc_bind_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_proc_bind_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_reduction_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_schedule_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_parallel_for_simd_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_private_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_private_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_reduction_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_simd_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_simd_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_simd_collapse_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_simd_collapse_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_simd_dist_schedule_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_simd_dist_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_simd_firstprivate_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_simd_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_simd_lastprivate_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_simd_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_simd_private_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_simd_private_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_distribute_simd_reduction_codegen.cpp b/clang/test/OpenMP/target_teams_distribute_simd_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_map_codegen.cpp b/clang/test/OpenMP/target_teams_map_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_num_teams_codegen.cpp b/clang/test/OpenMP/target_teams_num_teams_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_thread_limit_codegen.cpp b/clang/test/OpenMP/target_teams_thread_limit_codegen.cpp
diff --git a/clang/test/OpenMP/task_if_codegen.cpp b/clang/test/OpenMP/task_if_codegen.cpp
diff --git a/clang/test/OpenMP/task_in_reduction_codegen.cpp b/clang/test/OpenMP/task_in_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/task_member_call_codegen.cpp b/clang/test/OpenMP/task_member_call_codegen.cpp
@@ -1,10 +1,10 @@
 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --function-signature --include-generated-funcs --replace-value-regex "__omp_offloading_[0-9a-z]+_[0-9a-z]+" "reduction_size[.].+[.]" "pl_cond[.].+[.|,]" --prefix-filecheck-ir-name _
 // RUN: %clang_cc1 -no-opaque-pointers -verify -triple x86_64-apple-darwin10 -fopenmp -x c++ -emit-llvm %s -o - | FileCheck %s --check-prefix=CHECK1
 // RUN: %clang_cc1 -no-opaque-pointers -fopenmp -x c++ -triple x86_64-apple-darwin10 -emit-pch -o %t %s
-// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -x c++ -triple x86_64-apple-darwin10 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK2
+// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -x c++ -triple x86_64-apple-darwin10 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK1
 // RUN: %clang_cc1 -no-opaque-pointers -verify -triple x86_64-apple-darwin10 -fopenmp -fopenmp-enable-irbuilder -x c++ -emit-llvm %s -o - | FileCheck %s --check-prefix=CHECK3
 // RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-enable-irbuilder -x c++ -triple x86_64-apple-darwin10 -emit-pch -o %t %s
-// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-enable-irbuilder -x c++ -triple x86_64-apple-darwin10 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK4
+// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-enable-irbuilder -x c++ -triple x86_64-apple-darwin10 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK3
 
 // RUN: %clang_cc1 -no-opaque-pointers -verify -triple x86_64-apple-darwin10 -fopenmp-simd -x c++ -emit-llvm %s -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
 // RUN: %clang_cc1 -no-opaque-pointers -fopenmp-simd -x c++ -triple x86_64-apple-darwin10 -emit-pch -o %t %s
@@ -97,79 +97,6 @@ void c() {
 // CHECK1-NEXT:    ret i32 0
 //
 //
-// CHECK2-LABEL: define {{[^@]+}}@_Z1cv
-// CHECK2-SAME: () #[[ATTR0:[0-9]+]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[D:%.*]] = alloca [[CLASS_A:%.*]], align 1
-// CHECK2-NEXT:    [[AGG_CAPTURED:%.*]] = alloca [[STRUCT_ANON:%.*]], align 1
-// CHECK2-NEXT:    [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB1:[0-9]+]])
-// CHECK2-NEXT:    [[TMP1:%.*]] = call i8* @__kmpc_omp_task_alloc(%struct.ident_t* @[[GLOB1]], i32 [[TMP0]], i32 1, i64 48, i64 1, i32 (i32, i8*)* bitcast (i32 (i32, %struct.kmp_task_t_with_privates*)* @.omp_task_entry. to i32 (i32, i8*)*))
-// CHECK2-NEXT:    [[TMP2:%.*]] = bitcast i8* [[TMP1]] to %struct.kmp_task_t_with_privates*
-// CHECK2-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T_WITH_PRIVATES:%.*]], %struct.kmp_task_t_with_privates* [[TMP2]], i32 0, i32 0
-// CHECK2-NEXT:    [[TMP4:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T_WITH_PRIVATES]], %struct.kmp_task_t_with_privates* [[TMP2]], i32 0, i32 1
-// CHECK2-NEXT:    [[TMP5:%.*]] = getelementptr inbounds [[STRUCT__KMP_PRIVATES_T:%.*]], %struct..kmp_privates.t* [[TMP4]], i32 0, i32 0
-// CHECK2-NEXT:    [[TMP6:%.*]] = call i32 @__kmpc_omp_task(%struct.ident_t* @[[GLOB1]], i32 [[TMP0]], i8* [[TMP1]])
-// CHECK2-NEXT:    ret void
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@.omp_task_privates_map.
-// CHECK2-SAME: (%struct..kmp_privates.t* noalias noundef [[TMP0:%.*]], %class.a** noalias noundef [[TMP1:%.*]]) #[[ATTR2:[0-9]+]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[DOTADDR:%.*]] = alloca %struct..kmp_privates.t*, align 8
-// CHECK2-NEXT:    [[DOTADDR1:%.*]] = alloca %class.a**, align 8
-// CHECK2-NEXT:    store %struct..kmp_privates.t* [[TMP0]], %struct..kmp_privates.t** [[DOTADDR]], align 8
-// CHECK2-NEXT:    store %class.a** [[TMP1]], %class.a*** [[DOTADDR1]], align 8
-// CHECK2-NEXT:    [[TMP2:%.*]] = load %struct..kmp_privates.t*, %struct..kmp_privates.t** [[DOTADDR]], align 8
-// CHECK2-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [[STRUCT__KMP_PRIVATES_T:%.*]], %struct..kmp_privates.t* [[TMP2]], i32 0, i32 0
-// CHECK2-NEXT:    [[TMP4:%.*]] = load %class.a**, %class.a*** [[DOTADDR1]], align 8
-// CHECK2-NEXT:    store %class.a* [[TMP3]], %class.a** [[TMP4]], align 8
-// CHECK2-NEXT:    ret void
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@.omp_task_entry.
-// CHECK2-SAME: (i32 noundef [[TMP0:%.*]], %struct.kmp_task_t_with_privates* noalias noundef [[TMP1:%.*]]) #[[ATTR3:[0-9]+]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[DOTGLOBAL_TID__ADDR_I:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTPART_ID__ADDR_I:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    [[DOTPRIVATES__ADDR_I:%.*]] = alloca i8*, align 8
-// CHECK2-NEXT:    [[DOTCOPY_FN__ADDR_I:%.*]] = alloca void (i8*, ...)*, align 8
-// CHECK2-NEXT:    [[DOTTASK_T__ADDR_I:%.*]] = alloca i8*, align 8
-// CHECK2-NEXT:    [[__CONTEXT_ADDR_I:%.*]] = alloca %struct.anon*, align 8
-// CHECK2-NEXT:    [[DOTFIRSTPRIV_PTR_ADDR_I:%.*]] = alloca %class.a*, align 8
-// CHECK2-NEXT:    [[DOTADDR:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[DOTADDR1:%.*]] = alloca %struct.kmp_task_t_with_privates*, align 8
-// CHECK2-NEXT:    store i32 [[TMP0]], i32* [[DOTADDR]], align 4
-// CHECK2-NEXT:    store %struct.kmp_task_t_with_privates* [[TMP1]], %struct.kmp_task_t_with_privates** [[DOTADDR1]], align 8
-// CHECK2-NEXT:    [[TMP2:%.*]] = load i32, i32* [[DOTADDR]], align 4
-// CHECK2-NEXT:    [[TMP3:%.*]] = load %struct.kmp_task_t_with_privates*, %struct.kmp_task_t_with_privates** [[DOTADDR1]], align 8
-// CHECK2-NEXT:    [[TMP4:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T_WITH_PRIVATES:%.*]], %struct.kmp_task_t_with_privates* [[TMP3]], i32 0, i32 0
-// CHECK2-NEXT:    [[TMP5:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T:%.*]], %struct.kmp_task_t* [[TMP4]], i32 0, i32 2
-// CHECK2-NEXT:    [[TMP6:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T]], %struct.kmp_task_t* [[TMP4]], i32 0, i32 0
-// CHECK2-NEXT:    [[TMP7:%.*]] = load i8*, i8** [[TMP6]], align 8
-// CHECK2-NEXT:    [[TMP8:%.*]] = bitcast i8* [[TMP7]] to %struct.anon*
-// CHECK2-NEXT:    [[TMP9:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T_WITH_PRIVATES]], %struct.kmp_task_t_with_privates* [[TMP3]], i32 0, i32 1
-// CHECK2-NEXT:    [[TMP10:%.*]] = bitcast %struct..kmp_privates.t* [[TMP9]] to i8*
-// CHECK2-NEXT:    [[TMP11:%.*]] = bitcast %struct.kmp_task_t_with_privates* [[TMP3]] to i8*
-// CHECK2-NEXT:    call void @llvm.experimental.noalias.scope.decl(metadata [[META3:![0-9]+]])
-// CHECK2-NEXT:    call void @llvm.experimental.noalias.scope.decl(metadata [[META6:![0-9]+]])
-// CHECK2-NEXT:    call void @llvm.experimental.noalias.scope.decl(metadata [[META8:![0-9]+]])
-// CHECK2-NEXT:    call void @llvm.experimental.noalias.scope.decl(metadata [[META10:![0-9]+]])
-// CHECK2-NEXT:    store i32 [[TMP2]], i32* [[DOTGLOBAL_TID__ADDR_I]], align 4, !noalias !12
-// CHECK2-NEXT:    store i32* [[TMP5]], i32** [[DOTPART_ID__ADDR_I]], align 8, !noalias !12
-// CHECK2-NEXT:    store i8* [[TMP10]], i8** [[DOTPRIVATES__ADDR_I]], align 8, !noalias !12
-// CHECK2-NEXT:    store void (i8*, ...)* bitcast (void (%struct..kmp_privates.t*, %class.a**)* @.omp_task_privates_map. to void (i8*, ...)*), void (i8*, ...)** [[DOTCOPY_FN__ADDR_I]], align 8, !noalias !12
-// CHECK2-NEXT:    store i8* [[TMP11]], i8** [[DOTTASK_T__ADDR_I]], align 8, !noalias !12
-// CHECK2-NEXT:    store %struct.anon* [[TMP8]], %struct.anon** [[__CONTEXT_ADDR_I]], align 8, !noalias !12
-// CHECK2-NEXT:    [[TMP12:%.*]] = load %struct.anon*, %struct.anon** [[__CONTEXT_ADDR_I]], align 8, !noalias !12
-// CHECK2-NEXT:    [[TMP13:%.*]] = load void (i8*, ...)*, void (i8*, ...)** [[DOTCOPY_FN__ADDR_I]], align 8, !noalias !12
-// CHECK2-NEXT:    [[TMP14:%.*]] = load i8*, i8** [[DOTPRIVATES__ADDR_I]], align 8, !noalias !12
-// CHECK2-NEXT:    [[TMP15:%.*]] = bitcast void (i8*, ...)* [[TMP13]] to void (i8*, %class.a**)*
-// CHECK2-NEXT:    call void [[TMP15]](i8* [[TMP14]], %class.a** [[DOTFIRSTPRIV_PTR_ADDR_I]]) #[[ATTR4:[0-9]+]]
-// CHECK2-NEXT:    [[TMP16:%.*]] = load %class.a*, %class.a** [[DOTFIRSTPRIV_PTR_ADDR_I]], align 8, !noalias !12
-// CHECK2-NEXT:    call void @_ZN1a1bEv(%class.a* noundef nonnull align 1 dereferenceable(1) [[TMP16]]) #[[ATTR4]]
-// CHECK2-NEXT:    ret i32 0
-//
-//
 // CHECK3-LABEL: define {{[^@]+}}@_Z1cv
 // CHECK3-SAME: () #[[ATTR0:[0-9]+]] {
 // CHECK3-NEXT:  entry:
@@ -243,77 +170,3 @@ void c() {
 // CHECK3-NEXT:    call void @_ZN1a1bEv(%class.a* noundef nonnull align 1 dereferenceable(1) [[TMP16]]) #[[ATTR4]]
 // CHECK3-NEXT:    ret i32 0
 //
-//
-// CHECK4-LABEL: define {{[^@]+}}@_Z1cv
-// CHECK4-SAME: () #[[ATTR0:[0-9]+]] {
-// CHECK4-NEXT:  entry:
-// CHECK4-NEXT:    [[D:%.*]] = alloca [[CLASS_A:%.*]], align 1
-// CHECK4-NEXT:    [[AGG_CAPTURED:%.*]] = alloca [[STRUCT_ANON:%.*]], align 1
-// CHECK4-NEXT:    [[OMP_GLOBAL_THREAD_NUM:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB3:[0-9]+]])
-// CHECK4-NEXT:    [[TMP0:%.*]] = call i8* @__kmpc_omp_task_alloc(%struct.ident_t* @[[GLOB1:[0-9]+]], i32 [[OMP_GLOBAL_THREAD_NUM]], i32 1, i64 48, i64 1, i32 (i32, i8*)* bitcast (i32 (i32, %struct.kmp_task_t_with_privates*)* @.omp_task_entry. to i32 (i32, i8*)*))
-// CHECK4-NEXT:    [[TMP1:%.*]] = bitcast i8* [[TMP0]] to %struct.kmp_task_t_with_privates*
-// CHECK4-NEXT:    [[TMP2:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T_WITH_PRIVATES:%.*]], %struct.kmp_task_t_with_privates* [[TMP1]], i32 0, i32 0
-// CHECK4-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T_WITH_PRIVATES]], %struct.kmp_task_t_with_privates* [[TMP1]], i32 0, i32 1
-// CHECK4-NEXT:    [[TMP4:%.*]] = getelementptr inbounds [[STRUCT__KMP_PRIVATES_T:%.*]], %struct..kmp_privates.t* [[TMP3]], i32 0, i32 0
-// CHECK4-NEXT:    [[OMP_GLOBAL_THREAD_NUM1:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB3]])
-// CHECK4-NEXT:    [[TMP5:%.*]] = call i32 @__kmpc_omp_task(%struct.ident_t* @[[GLOB1]], i32 [[OMP_GLOBAL_THREAD_NUM1]], i8* [[TMP0]])
-// CHECK4-NEXT:    ret void
-//
-//
-// CHECK4-LABEL: define {{[^@]+}}@.omp_task_privates_map.
-// CHECK4-SAME: (%struct..kmp_privates.t* noalias noundef [[TMP0:%.*]], %class.a** noalias noundef [[TMP1:%.*]]) #[[ATTR2:[0-9]+]] {
-// CHECK4-NEXT:  entry:
-// CHECK4-NEXT:    [[DOTADDR:%.*]] = alloca %struct..kmp_privates.t*, align 8
-// CHECK4-NEXT:    [[DOTADDR1:%.*]] = alloca %class.a**, align 8
-// CHECK4-NEXT:    store %struct..kmp_privates.t* [[TMP0]], %struct..kmp_privates.t** [[DOTADDR]], align 8
-// CHECK4-NEXT:    store %class.a** [[TMP1]], %class.a*** [[DOTADDR1]], align 8
-// CHECK4-NEXT:    [[TMP2:%.*]] = load %struct..kmp_privates.t*, %struct..kmp_privates.t** [[DOTADDR]], align 8
-// CHECK4-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [[STRUCT__KMP_PRIVATES_T:%.*]], %struct..kmp_privates.t* [[TMP2]], i32 0, i32 0
-// CHECK4-NEXT:    [[TMP4:%.*]] = load %class.a**, %class.a*** [[DOTADDR1]], align 8
-// CHECK4-NEXT:    store %class.a* [[TMP3]], %class.a** [[TMP4]], align 8
-// CHECK4-NEXT:    ret void
-//
-//
-// CHECK4-LABEL: define {{[^@]+}}@.omp_task_entry.
-// CHECK4-SAME: (i32 noundef [[TMP0:%.*]], %struct.kmp_task_t_with_privates* noalias noundef [[TMP1:%.*]]) #[[ATTR3:[0-9]+]] {
-// CHECK4-NEXT:  entry:
-// CHECK4-NEXT:    [[DOTGLOBAL_TID__ADDR_I:%.*]] = alloca i32, align 4
-// CHECK4-NEXT:    [[DOTPART_ID__ADDR_I:%.*]] = alloca i32*, align 8
-// CHECK4-NEXT:    [[DOTPRIVATES__ADDR_I:%.*]] = alloca i8*, align 8
-// CHECK4-NEXT:    [[DOTCOPY_FN__ADDR_I:%.*]] = alloca void (i8*, ...)*, align 8
-// CHECK4-NEXT:    [[DOTTASK_T__ADDR_I:%.*]] = alloca i8*, align 8
-// CHECK4-NEXT:    [[__CONTEXT_ADDR_I:%.*]] = alloca %struct.anon*, align 8
-// CHECK4-NEXT:    [[DOTFIRSTPRIV_PTR_ADDR_I:%.*]] = alloca %class.a*, align 8
-// CHECK4-NEXT:    [[DOTADDR:%.*]] = alloca i32, align 4
-// CHECK4-NEXT:    [[DOTADDR1:%.*]] = alloca %struct.kmp_task_t_with_privates*, align 8
-// CHECK4-NEXT:    store i32 [[TMP0]], i32* [[DOTADDR]], align 4
-// CHECK4-NEXT:    store %struct.kmp_task_t_with_privates* [[TMP1]], %struct.kmp_task_t_with_privates** [[DOTADDR1]], align 8
-// CHECK4-NEXT:    [[TMP2:%.*]] = load i32, i32* [[DOTADDR]], align 4
-// CHECK4-NEXT:    [[TMP3:%.*]] = load %struct.kmp_task_t_with_privates*, %struct.kmp_task_t_with_privates** [[DOTADDR1]], align 8
-// CHECK4-NEXT:    [[TMP4:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T_WITH_PRIVATES:%.*]], %struct.kmp_task_t_with_privates* [[TMP3]], i32 0, i32 0
-// CHECK4-NEXT:    [[TMP5:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T:%.*]], %struct.kmp_task_t* [[TMP4]], i32 0, i32 2
-// CHECK4-NEXT:    [[TMP6:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T]], %struct.kmp_task_t* [[TMP4]], i32 0, i32 0
-// CHECK4-NEXT:    [[TMP7:%.*]] = load i8*, i8** [[TMP6]], align 8
-// CHECK4-NEXT:    [[TMP8:%.*]] = bitcast i8* [[TMP7]] to %struct.anon*
-// CHECK4-NEXT:    [[TMP9:%.*]] = getelementptr inbounds [[STRUCT_KMP_TASK_T_WITH_PRIVATES]], %struct.kmp_task_t_with_privates* [[TMP3]], i32 0, i32 1
-// CHECK4-NEXT:    [[TMP10:%.*]] = bitcast %struct..kmp_privates.t* [[TMP9]] to i8*
-// CHECK4-NEXT:    [[TMP11:%.*]] = bitcast %struct.kmp_task_t_with_privates* [[TMP3]] to i8*
-// CHECK4-NEXT:    call void @llvm.experimental.noalias.scope.decl(metadata [[META3:![0-9]+]])
-// CHECK4-NEXT:    call void @llvm.experimental.noalias.scope.decl(metadata [[META6:![0-9]+]])
-// CHECK4-NEXT:    call void @llvm.experimental.noalias.scope.decl(metadata [[META8:![0-9]+]])
-// CHECK4-NEXT:    call void @llvm.experimental.noalias.scope.decl(metadata [[META10:![0-9]+]])
-// CHECK4-NEXT:    store i32 [[TMP2]], i32* [[DOTGLOBAL_TID__ADDR_I]], align 4, !noalias !12
-// CHECK4-NEXT:    store i32* [[TMP5]], i32** [[DOTPART_ID__ADDR_I]], align 8, !noalias !12
-// CHECK4-NEXT:    store i8* [[TMP10]], i8** [[DOTPRIVATES__ADDR_I]], align 8, !noalias !12
-// CHECK4-NEXT:    store void (i8*, ...)* bitcast (void (%struct..kmp_privates.t*, %class.a**)* @.omp_task_privates_map. to void (i8*, ...)*), void (i8*, ...)** [[DOTCOPY_FN__ADDR_I]], align 8, !noalias !12
-// CHECK4-NEXT:    store i8* [[TMP11]], i8** [[DOTTASK_T__ADDR_I]], align 8, !noalias !12
-// CHECK4-NEXT:    store %struct.anon* [[TMP8]], %struct.anon** [[__CONTEXT_ADDR_I]], align 8, !noalias !12
-// CHECK4-NEXT:    [[TMP12:%.*]] = load %struct.anon*, %struct.anon** [[__CONTEXT_ADDR_I]], align 8, !noalias !12
-// CHECK4-NEXT:    [[TMP13:%.*]] = load void (i8*, ...)*, void (i8*, ...)** [[DOTCOPY_FN__ADDR_I]], align 8, !noalias !12
-// CHECK4-NEXT:    [[TMP14:%.*]] = load i8*, i8** [[DOTPRIVATES__ADDR_I]], align 8, !noalias !12
-// CHECK4-NEXT:    [[TMP15:%.*]] = bitcast void (i8*, ...)* [[TMP13]] to void (i8*, %class.a**)*
-// CHECK4-NEXT:    call void [[TMP15]](i8* [[TMP14]], %class.a** [[DOTFIRSTPRIV_PTR_ADDR_I]]) #[[ATTR4:[0-9]+]]
-// CHECK4-NEXT:    [[TMP16:%.*]] = load %class.a*, %class.a** [[DOTFIRSTPRIV_PTR_ADDR_I]], align 8, !noalias !12
-// CHECK4-NEXT:    call void @_ZN1a1bEv(%class.a* noundef nonnull align 1 dereferenceable(1) [[TMP16]]) #[[ATTR4]]
-// CHECK4-NEXT:    ret i32 0
-//
diff --git a/clang/test/OpenMP/taskgroup_codegen.cpp b/clang/test/OpenMP/taskgroup_codegen.cpp
@@ -1,7 +1,7 @@
 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --function-signature --include-generated-funcs --replace-value-regex "__omp_offloading_[0-9a-z]+_[0-9a-z]+" "reduction_size[.].+[.]" "pl_cond[.].+[.|,]" --prefix-filecheck-ir-name _
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -x c++ -triple x86_64-unknown-unknown -emit-llvm %s -fexceptions -fcxx-exceptions -o - | FileCheck %s --check-prefix=CHECK1
 // RUN: %clang_cc1 -no-opaque-pointers -fopenmp -x c++ -std=c++11 -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -emit-pch -o %t %s
-// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -x c++ -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK2
+// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -x c++ -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK1
 // RUN: %clang_cc1 -no-opaque-pointers -verify -triple x86_64-apple-darwin10 -fopenmp -fexceptions -fcxx-exceptions -debug-info-kind=line-tables-only -x c++ -emit-llvm %s -o - | FileCheck %s --check-prefix=DEBUG1
 
 // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp-simd -x c++ -emit-llvm %s -fexceptions -fcxx-exceptions -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
@@ -99,76 +99,6 @@ void parallel_taskgroup() {
 // CHECK1-NEXT:    unreachable
 //
 //
-// CHECK2-LABEL: define {{[^@]+}}@_Z3foov
-// CHECK2-SAME: () #[[ATTR0:[0-9]+]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    call void @_Z8mayThrowv()
-// CHECK2-NEXT:    ret void
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@main
-// CHECK2-SAME: () #[[ATTR2:[0-9]+]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[RETVAL:%.*]] = alloca i32, align 4
-// CHECK2-NEXT:    [[A:%.*]] = alloca i8, align 1
-// CHECK2-NEXT:    [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @[[GLOB1:[0-9]+]])
-// CHECK2-NEXT:    store i32 0, i32* [[RETVAL]], align 4
-// CHECK2-NEXT:    call void @__kmpc_taskgroup(%struct.ident_t* @[[GLOB1]], i32 [[TMP0]])
-// CHECK2-NEXT:    store i8 2, i8* [[A]], align 1
-// CHECK2-NEXT:    call void @__kmpc_end_taskgroup(%struct.ident_t* @[[GLOB1]], i32 [[TMP0]])
-// CHECK2-NEXT:    call void @__kmpc_taskgroup(%struct.ident_t* @[[GLOB1]], i32 [[TMP0]])
-// CHECK2-NEXT:    invoke void @_Z3foov()
-// CHECK2-NEXT:    to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]]
-// CHECK2:       invoke.cont:
-// CHECK2-NEXT:    call void @__kmpc_end_taskgroup(%struct.ident_t* @[[GLOB1]], i32 [[TMP0]])
-// CHECK2-NEXT:    [[TMP1:%.*]] = load i8, i8* [[A]], align 1
-// CHECK2-NEXT:    [[CONV:%.*]] = sext i8 [[TMP1]] to i32
-// CHECK2-NEXT:    ret i32 [[CONV]]
-// CHECK2:       terminate.lpad:
-// CHECK2-NEXT:    [[TMP2:%.*]] = landingpad { i8*, i32 }
-// CHECK2-NEXT:    catch i8* null
-// CHECK2-NEXT:    [[TMP3:%.*]] = extractvalue { i8*, i32 } [[TMP2]], 0
-// CHECK2-NEXT:    call void @__clang_call_terminate(i8* [[TMP3]]) #[[ATTR8:[0-9]+]]
-// CHECK2-NEXT:    unreachable
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@__clang_call_terminate
-// CHECK2-SAME: (i8* [[TMP0:%.*]]) #[[ATTR5:[0-9]+]] comdat {
-// CHECK2-NEXT:    [[TMP2:%.*]] = call i8* @__cxa_begin_catch(i8* [[TMP0]]) #[[ATTR3:[0-9]+]]
-// CHECK2-NEXT:    call void @_ZSt9terminatev() #[[ATTR8]]
-// CHECK2-NEXT:    unreachable
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@_Z18parallel_taskgroupv
-// CHECK2-SAME: () #[[ATTR6:[0-9]+]] {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB1]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined. to void (i32*, i32*, ...)*))
-// CHECK2-NEXT:    ret void
-//
-//
-// CHECK2-LABEL: define {{[^@]+}}@.omp_outlined.
-// CHECK2-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR7:[0-9]+]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) {
-// CHECK2-NEXT:  entry:
-// CHECK2-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8
-// CHECK2-NEXT:    store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8
-// CHECK2-NEXT:    store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8
-// CHECK2-NEXT:    [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8
-// CHECK2-NEXT:    [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4
-// CHECK2-NEXT:    call void @__kmpc_taskgroup(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]])
-// CHECK2-NEXT:    invoke void @_Z3foov()
-// CHECK2-NEXT:    to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]]
-// CHECK2:       invoke.cont:
-// CHECK2-NEXT:    call void @__kmpc_end_taskgroup(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]])
-// CHECK2-NEXT:    ret void
-// CHECK2:       terminate.lpad:
-// CHECK2-NEXT:    [[TMP2:%.*]] = landingpad { i8*, i32 }
-// CHECK2-NEXT:    catch i8* null
-// CHECK2-NEXT:    [[TMP3:%.*]] = extractvalue { i8*, i32 } [[TMP2]], 0
-// CHECK2-NEXT:    call void @__clang_call_terminate(i8* [[TMP3]]) #[[ATTR8]]
-// CHECK2-NEXT:    unreachable
-//
-//
 // DEBUG1-LABEL: define {{[^@]+}}@_Z3foov
 // DEBUG1-SAME: () #[[ATTR0:[0-9]+]] !dbg [[DBG6:![0-9]+]] {
 // DEBUG1-NEXT:  entry:
@@ -217,7 +147,7 @@ void parallel_taskgroup() {
 //
 //
 // DEBUG1-LABEL: define {{[^@]+}}@.omp_outlined.
-// DEBUG1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR7:[0-9]+]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) !dbg [[DBG24:![0-9]+]] {
+// DEBUG1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR7:[0-9]+]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) !dbg [[DBG23:![0-9]+]] {
 // DEBUG1-NEXT:  entry:
 // DEBUG1-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8
 // DEBUG1-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8

diff --git a/clang/test/OpenMP/taskloop_in_reduction_codegen.cpp b/clang/test/OpenMP/taskloop_in_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/taskloop_simd_in_reduction_codegen.cpp b/clang/test/OpenMP/taskloop_simd_in_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/teams_codegen.cpp b/clang/test/OpenMP/teams_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_codegen.cpp b/clang/test/OpenMP/teams_distribute_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_collapse_codegen.cpp b/clang/test/OpenMP/teams_distribute_collapse_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_dist_schedule_codegen.cpp b/clang/test/OpenMP/teams_distribute_dist_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_firstprivate_codegen.cpp b/clang/test/OpenMP/teams_distribute_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_lastprivate_codegen.cpp b/clang/test/OpenMP/teams_distribute_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_collapse_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_collapse_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_copyin_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_copyin_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_dist_schedule_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_dist_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_firstprivate_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_if_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_if_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_lastprivate_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_num_threads_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_num_threads_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_private_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_private_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_proc_bind_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_proc_bind_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_reduction_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_reduction_task_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_reduction_task_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_schedule_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_collapse_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_collapse_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_dist_schedule_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_dist_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_firstprivate_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_if_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_if_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_lastprivate_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_num_threads_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_num_threads_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_private_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_private_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_proc_bind_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_proc_bind_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_reduction_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_parallel_for_simd_schedule_codegen.cpp b/clang/test/OpenMP/teams_distribute_parallel_for_simd_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_private_codegen.cpp b/clang/test/OpenMP/teams_distribute_private_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_reduction_codegen.cpp b/clang/test/OpenMP/teams_distribute_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_simd_codegen.cpp b/clang/test/OpenMP/teams_distribute_simd_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_simd_collapse_codegen.cpp b/clang/test/OpenMP/teams_distribute_simd_collapse_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_simd_dist_schedule_codegen.cpp b/clang/test/OpenMP/teams_distribute_simd_dist_schedule_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_simd_firstprivate_codegen.cpp b/clang/test/OpenMP/teams_distribute_simd_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_simd_lastprivate_codegen.cpp b/clang/test/OpenMP/teams_distribute_simd_lastprivate_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_simd_private_codegen.cpp b/clang/test/OpenMP/teams_distribute_simd_private_codegen.cpp
diff --git a/clang/test/OpenMP/teams_distribute_simd_reduction_codegen.cpp b/clang/test/OpenMP/teams_distribute_simd_reduction_codegen.cpp
diff --git a/clang/test/OpenMP/teams_firstprivate_codegen.cpp b/clang/test/OpenMP/teams_firstprivate_codegen.cpp
diff --git a/clang/test/OpenMP/teams_private_codegen.cpp b/clang/test/OpenMP/teams_private_codegen.cpp
diff --git a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
@@ -34,6 +34,7 @@
 #include "llvm/Analysis/ValueTracking.h"
 #include "llvm/IR/Argument.h"
 #include "llvm/IR/Assumptions.h"
+#include "llvm/IR/BasicBlock.h"
 #include "llvm/IR/Constants.h"
 #include "llvm/IR/DataLayout.h"
 #include "llvm/IR/DerivedTypes.h"
@@ -6241,6 +6242,17 @@ struct AAHeapToStackFunction final : public AAHeapToStack {
     Function *F = getAnchorScope();
     const auto *TLI = A.getInfoCache().getTargetLibraryInfoForFunction(*F);
 
+    LoopInfo *LI =
+        A.getInfoCache().getAnalysisResultForFunction<LoopAnalysis>(*F);
+    Optional<bool> MayContainIrreducibleControl;
+    auto IsInLoop = [&](BasicBlock &BB) {
+      if (!MayContainIrreducibleControl.has_value())
+        MayContainIrreducibleControl = mayContainIrreducibleControl(*F, LI);
+      if (MayContainIrreducibleControl.value())
+        return true;
+      return LI->getLoopFor(&BB) != nullptr;
+    };
+
     for (auto &It : AllocationInfos) {
       AllocationInfo &AI = *It.second;
       if (AI.Status == AllocationInfo::INVALID)
@@ -6282,6 +6294,10 @@ struct AAHeapToStackFunction final : public AAHeapToStack {
         Size = SizeOffsetPair.first;
       }
 
+      Instruction *IP = (!SizeAPI.has_value() || IsInLoop(*AI.CB->getParent()))
+                            ? AI.CB
+                            : &F->getEntryBlock().front();
+
       Align Alignment(1);
       if (MaybeAlign RetAlign = AI.CB->getRetAlign())
         Alignment = std::max(Alignment, *RetAlign);
@@ -6296,7 +6312,7 @@ struct AAHeapToStackFunction final : public AAHeapToStack {
       // TODO: Hoist the alloca towards the function entry.
       unsigned AS = DL.getAllocaAddrSpace();
       Instruction *Alloca = new AllocaInst(Type::getInt8Ty(F->getContext()), AS,
-                                           Size, Alignment, "", AI.CB);
+                                           Size, Alignment, "", IP);
 
       if (Alloca->getType() != AI.CB->getType())
         Alloca = BitCastInst::CreatePointerBitCastOrAddrSpaceCast(

diff --git a/llvm/test/Transforms/Attributor/heap_to_stack_gpu.ll b/llvm/test/Transforms/Attributor/heap_to_stack_gpu.ll
@@ -704,6 +704,35 @@ define void @test17b() {
   ret void
 }
 
+define void @move_alloca() {
+; IS________OPM-LABEL: define {{[^@]+}}@move_alloca() {
+; IS________OPM-NEXT:  entry:
+; IS________OPM-NEXT:    br label [[NOT_ENTRY:%.*]]
+; IS________OPM:       not_entry:
+; IS________OPM-NEXT:    [[TMP0:%.*]] = tail call noalias i8* @__kmpc_alloc_shared(i64 noundef 4)
+; IS________OPM-NEXT:    tail call void @usei8(i8* noalias nocapture nofree [[TMP0]]) #[[ATTR6]]
+; IS________OPM-NEXT:    tail call void @__kmpc_free_shared(i8* noalias nocapture [[TMP0]], i64 noundef 4)
+; IS________OPM-NEXT:    ret void
+;
+; IS________NPM-LABEL: define {{[^@]+}}@move_alloca() {
+; IS________NPM-NEXT:  entry:
+; IS________NPM-NEXT:    [[TMP0:%.*]] = alloca i8, i64 4, align 1, addrspace(5)
+; IS________NPM-NEXT:    br label [[NOT_ENTRY:%.*]]
+; IS________NPM:       not_entry:
+; IS________NPM-NEXT:    [[MALLOC_CAST:%.*]] = addrspacecast i8 addrspace(5)* [[TMP0]] to i8*
+; IS________NPM-NEXT:    tail call void @usei8(i8* noalias nocapture nofree [[MALLOC_CAST]]) #[[ATTR6]]
+; IS________NPM-NEXT:    ret void
+;
+entry:
+  br label %not_entry
+
+not_entry:
+  %0 = tail call noalias i8* @__kmpc_alloc_shared(i64 4)
+  tail call void @usei8(i8* nocapture nofree %0) willreturn nounwind nosync
+  tail call void @__kmpc_free_shared(i8* %0, i64 4)
+  ret void
+}
+
 
 ;.
 ; CHECK: attributes #[[ATTR0:[0-9]+]] = { nounwind willreturn }

diff --git a/llvm/test/Transforms/Attributor/value-simplify-pointer-info.ll b/llvm/test/Transforms/Attributor/value-simplify-pointer-info.ll
@@ -3978,16 +3978,16 @@ define dso_local void @test_nested_memory(float* %dst, double* %src) {
 ; IS__TUNIT_NPM-LABEL: define {{[^@]+}}@test_nested_memory
 ; IS__TUNIT_NPM-SAME: (float* nocapture nofree writeonly [[DST:%.*]], double* nocapture nofree readonly [[SRC:%.*]]) {
 ; IS__TUNIT_NPM-NEXT:  entry:
+; IS__TUNIT_NPM-NEXT:    [[TMP0:%.*]] = alloca i8, i64 24, align 1
 ; IS__TUNIT_NPM-NEXT:    [[LOCAL:%.*]] = alloca [[STRUCT_STY:%.*]], align 8
-; IS__TUNIT_NPM-NEXT:    [[TMP0:%.*]] = bitcast %struct.STy* [[LOCAL]] to i8*
+; IS__TUNIT_NPM-NEXT:    [[TMP1:%.*]] = bitcast %struct.STy* [[LOCAL]] to i8*
 ; IS__TUNIT_NPM-NEXT:    [[INNER:%.*]] = getelementptr inbounds [[STRUCT_STY]], %struct.STy* [[LOCAL]], i64 0, i32 2
-; IS__TUNIT_NPM-NEXT:    [[TMP1:%.*]] = alloca i8, i64 24, align 1
-; IS__TUNIT_NPM-NEXT:    [[DST1:%.*]] = bitcast i8* [[TMP1]] to float**
+; IS__TUNIT_NPM-NEXT:    [[DST1:%.*]] = bitcast i8* [[TMP0]] to float**
 ; IS__TUNIT_NPM-NEXT:    store float* [[DST]], float** [[DST1]], align 8
-; IS__TUNIT_NPM-NEXT:    [[SRC2:%.*]] = getelementptr inbounds i8, i8* [[TMP1]], i64 8
+; IS__TUNIT_NPM-NEXT:    [[SRC2:%.*]] = getelementptr inbounds i8, i8* [[TMP0]], i64 8
 ; IS__TUNIT_NPM-NEXT:    [[TMP2:%.*]] = bitcast i8* [[SRC2]] to double**
 ; IS__TUNIT_NPM-NEXT:    store double* [[SRC]], double** [[TMP2]], align 8
-; IS__TUNIT_NPM-NEXT:    store i8* [[TMP1]], i8** bitcast (%struct.STy** getelementptr inbounds ([[STRUCT_STY]], %struct.STy* @global, i64 0, i32 2) to i8**), align 8
+; IS__TUNIT_NPM-NEXT:    store i8* [[TMP0]], i8** bitcast (%struct.STy** getelementptr inbounds ([[STRUCT_STY]], %struct.STy* @global, i64 0, i32 2) to i8**), align 8
 ; IS__TUNIT_NPM-NEXT:    call fastcc void @nested_memory_callee() #[[ATTR15:[0-9]+]]
 ; IS__TUNIT_NPM-NEXT:    ret void
 ;

diff --git a/llvm/test/Transforms/OpenMP/spmdization.ll b/llvm/test/Transforms/OpenMP/spmdization.ll
@@ -702,8 +702,8 @@ define internal void @__omp_outlined__2(i32* noalias %.global_tid., i32* noalias
 ; AMDGPU-LABEL: define {{[^@]+}}@__omp_outlined__2
 ; AMDGPU-SAME: (i32* noalias [[DOTGLOBAL_TID_:%.*]], i32* noalias [[DOTBOUND_TID_:%.*]]) #[[ATTR0]] {
 ; AMDGPU-NEXT:  entry:
-; AMDGPU-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 8
 ; AMDGPU-NEXT:    [[TMP0:%.*]] = alloca i8, i64 4, align 4, addrspace(5)
+; AMDGPU-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 8
 ; AMDGPU-NEXT:    [[MALLOC_CAST:%.*]] = addrspacecast i8 addrspace(5)* [[TMP0]] to i8*
 ; AMDGPU-NEXT:    [[X_ON_STACK:%.*]] = bitcast i8* [[MALLOC_CAST]] to i32*
 ; AMDGPU-NEXT:    call void @use(i32* nocapture [[X_ON_STACK]]) #[[ATTR7]]
@@ -725,8 +725,8 @@ define internal void @__omp_outlined__2(i32* noalias %.global_tid., i32* noalias
 ; NVPTX-LABEL: define {{[^@]+}}@__omp_outlined__2
 ; NVPTX-SAME: (i32* noalias [[DOTGLOBAL_TID_:%.*]], i32* noalias [[DOTBOUND_TID_:%.*]]) #[[ATTR0]] {
 ; NVPTX-NEXT:  entry:
-; NVPTX-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 8
 ; NVPTX-NEXT:    [[TMP0:%.*]] = alloca i8, i64 4, align 4
+; NVPTX-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 8
 ; NVPTX-NEXT:    [[X_ON_STACK:%.*]] = bitcast i8* [[TMP0]] to i32*
 ; NVPTX-NEXT:    call void @use(i32* nocapture [[X_ON_STACK]]) #[[ATTR7]]
 ; NVPTX-NEXT:    br label [[FOR_COND:%.*]]
@@ -747,8 +747,8 @@ define internal void @__omp_outlined__2(i32* noalias %.global_tid., i32* noalias
 ; AMDGPU-DISABLED-LABEL: define {{[^@]+}}@__omp_outlined__2
 ; AMDGPU-DISABLED-SAME: (i32* noalias [[DOTGLOBAL_TID_:%.*]], i32* noalias [[DOTBOUND_TID_:%.*]]) #[[ATTR0]] {
 ; AMDGPU-DISABLED-NEXT:  entry:
-; AMDGPU-DISABLED-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 8
 ; AMDGPU-DISABLED-NEXT:    [[TMP0:%.*]] = alloca i8, i64 4, align 4, addrspace(5)
+; AMDGPU-DISABLED-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 8
 ; AMDGPU-DISABLED-NEXT:    [[MALLOC_CAST:%.*]] = addrspacecast i8 addrspace(5)* [[TMP0]] to i8*
 ; AMDGPU-DISABLED-NEXT:    [[X_ON_STACK:%.*]] = bitcast i8* [[MALLOC_CAST]] to i32*
 ; AMDGPU-DISABLED-NEXT:    call void @use(i32* nocapture [[X_ON_STACK]]) #[[ATTR7]]
@@ -770,8 +770,8 @@ define internal void @__omp_outlined__2(i32* noalias %.global_tid., i32* noalias
 ; NVPTX-DISABLED-LABEL: define {{[^@]+}}@__omp_outlined__2
 ; NVPTX-DISABLED-SAME: (i32* noalias [[DOTGLOBAL_TID_:%.*]], i32* noalias [[DOTBOUND_TID_:%.*]]) #[[ATTR0]] {
 ; NVPTX-DISABLED-NEXT:  entry:
-; NVPTX-DISABLED-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 8
 ; NVPTX-DISABLED-NEXT:    [[TMP0:%.*]] = alloca i8, i64 4, align 4
+; NVPTX-DISABLED-NEXT:    [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 8
 ; NVPTX-DISABLED-NEXT:    [[X_ON_STACK:%.*]] = bitcast i8* [[TMP0]] to i32*
 ; NVPTX-DISABLED-NEXT:    call void @use(i32* nocapture [[X_ON_STACK]]) #[[ATTR7]]
 ; NVPTX-DISABLED-NEXT:    br label [[FOR_COND:%.*]]