[clang] Revert "[clang] Add nuw attribute to GEPs" (PR #106343)

Vitaly Buka via cfe-commits cfe-commits at lists.llvm.org
Wed Aug 28 00:49:59 PDT 2024


https://github.com/vitalybuka created https://github.com/llvm/llvm-project/pull/106343

Reverts llvm/llvm-project#105496

This patch breaks:
https://lab.llvm.org/buildbot/#/builders/25/builds/1952
https://lab.llvm.org/buildbot/#/builders/52/builds/1775

Somehow output is different with sanitizers.
Maybe non-determinism in the code?

>From f20e99acb05d5c014f934e81205c87178349ed67 Mon Sep 17 00:00:00 2001
From: Vitaly Buka <vitalybuka at gmail.com>
Date: Wed, 28 Aug 2024 00:48:35 -0700
Subject: [PATCH] Revert "[clang] Add nuw attribute to GEPs (#105496)"

This reverts commit 3d2fd31c8f362b1054985197a65194fcbbca57bb.
---
 clang/lib/CodeGen/CGBuilder.h                 |   6 +-
 clang/lib/CodeGen/CGExprScalar.cpp            |  17 +-
 clang/test/CodeGen/2005-01-02-ConstantInits.c |   4 +-
 clang/test/CodeGen/PowerPC/ppc-emmintrin.c    |  12 +-
 clang/test/CodeGen/PowerPC/ppc-xmmintrin.c    |  16 +-
 clang/test/CodeGen/attr-counted-by.c          |  16 +-
 ...d-nonzero-offset-when-nullptr-is-defined.c |   2 +-
 .../catch-nullptr-and-nonzero-offset.c        |  64 ++---
 .../CodeGen/catch-pointer-overflow-volatile.c |   2 +-
 clang/test/CodeGen/catch-pointer-overflow.c   |   6 +-
 clang/test/CodeGen/ext-int.c                  |   2 +-
 .../test/CodeGen/hexagon-brev-ld-ptr-incdec.c |   6 +-
 clang/test/CodeGen/integer-overflow.c         |   6 +-
 clang/test/CodeGen/ms-intrinsics.c            |  12 +-
 clang/test/CodeGen/ubsan-pointer-overflow.m   |   2 +-
 clang/test/CodeGen/vla.c                      |   2 +-
 .../attr-likelihood-iteration-stmt.cpp        |  36 +--
 clang/test/CodeGenCXX/for-range.cpp           |  12 +-
 .../CodeGenCXX/pr45964-decomp-transform.cpp   |   2 +-
 clang/test/CodeGenCXX/vla.cpp                 |   4 +-
 .../CodeGenHLSL/buffer-array-operator.hlsl    |   4 +-
 .../CodeGenSYCL/address-space-deduction.cpp   |  48 ++--
 clang/test/Headers/__clang_hip_math.hip       |  54 ++--
 clang/test/OpenMP/bug60602.cpp                |   8 +-
 clang/test/OpenMP/declare_mapper_codegen.cpp  |   4 +-
 clang/test/OpenMP/distribute_codegen.cpp      |  32 +--
 ...te_parallel_for_reduction_task_codegen.cpp |  24 +-
 clang/test/OpenMP/distribute_simd_codegen.cpp |  96 +++----
 clang/test/OpenMP/for_linear_codegen.cpp      |   2 +-
 clang/test/OpenMP/for_reduction_codegen.cpp   |  64 ++---
 .../test/OpenMP/for_reduction_codegen_UDR.cpp |  32 +--
 .../OpenMP/for_reduction_task_codegen.cpp     |  24 +-
 clang/test/OpenMP/for_scan_codegen.cpp        |  40 +--
 clang/test/OpenMP/for_simd_scan_codegen.cpp   |  40 +--
 clang/test/OpenMP/irbuilder_for_iterator.cpp  |   6 +-
 clang/test/OpenMP/irbuilder_for_rangefor.cpp  |   6 +-
 clang/test/OpenMP/irbuilder_for_unsigned.c    |   8 +-
 .../test/OpenMP/irbuilder_for_unsigned_auto.c |   8 +-
 .../test/OpenMP/irbuilder_for_unsigned_down.c |   2 +-
 .../OpenMP/irbuilder_for_unsigned_dynamic.c   |   8 +-
 .../irbuilder_for_unsigned_dynamic_chunked.c  |   8 +-
 .../OpenMP/irbuilder_for_unsigned_runtime.c   |   8 +-
 .../irbuilder_for_unsigned_static_chunked.c   |   8 +-
 clang/test/OpenMP/map_struct_ordering.cpp     |   2 +-
 .../master_taskloop_in_reduction_codegen.cpp  |  10 +-
 .../master_taskloop_reduction_codegen.cpp     |  12 +-
 ...ter_taskloop_simd_in_reduction_codegen.cpp |  10 +-
 ...master_taskloop_simd_reduction_codegen.cpp |  12 +-
 clang/test/OpenMP/ordered_codegen.cpp         |  80 +++---
 clang/test/OpenMP/parallel_for_codegen.cpp    | 144 +++++-----
 .../OpenMP/parallel_for_linear_codegen.cpp    |   2 +-
 .../parallel_for_reduction_task_codegen.cpp   |  24 +-
 .../test/OpenMP/parallel_for_scan_codegen.cpp |  44 ++--
 .../OpenMP/parallel_for_simd_scan_codegen.cpp |  40 +--
 ...parallel_master_reduction_task_codegen.cpp |  24 +-
 ...llel_master_taskloop_reduction_codegen.cpp |  12 +-
 ...master_taskloop_simd_reduction_codegen.cpp |  12 +-
 .../OpenMP/parallel_reduction_codegen.cpp     |  12 +-
 .../parallel_reduction_task_codegen.cpp       |  24 +-
 ...rallel_sections_reduction_task_codegen.cpp |  24 +-
 clang/test/OpenMP/reduction_implicit_map.cpp  |  50 ++--
 .../sections_reduction_task_codegen.cpp       |  24 +-
 .../target_data_use_device_addr_codegen.cpp   |  10 +-
 .../target_data_use_device_ptr_codegen.cpp    |  82 +++---
 .../OpenMP/target_has_device_addr_codegen.cpp |  24 +-
 .../OpenMP/target_in_reduction_codegen.cpp    |  10 +-
 .../OpenMP/target_is_device_ptr_codegen.cpp   | 208 +++++++--------
 ...arget_map_both_pointer_pointee_codegen.cpp |   2 +-
 clang/test/OpenMP/target_map_codegen_01.cpp   |   2 +-
 clang/test/OpenMP/target_map_codegen_21.cpp   |   6 +-
 clang/test/OpenMP/target_map_codegen_27.cpp   |   2 +-
 clang/test/OpenMP/target_map_codegen_28.cpp   |   6 +-
 clang/test/OpenMP/target_map_codegen_29.cpp   |   4 +-
 .../OpenMP/target_map_deref_array_codegen.cpp |   2 +-
 ..._map_member_expr_array_section_codegen.cpp |   8 +-
 .../OpenMP/target_map_member_expr_codegen.cpp |   6 +-
 ...target_map_nest_defalut_mapper_codegen.cpp |   2 +-
 ...et_parallel_for_reduction_task_codegen.cpp |  24 +-
 ...target_parallel_reduction_task_codegen.cpp |  24 +-
 .../OpenMP/target_task_affinity_codegen.cpp   |  24 +-
 ...te_parallel_for_reduction_task_codegen.cpp |  44 ++--
 clang/test/OpenMP/target_update_codegen.cpp   |  34 +--
 clang/test/OpenMP/task_codegen.c              |   2 +-
 clang/test/OpenMP/task_codegen.cpp            | 248 +++++++++---------
 .../test/OpenMP/task_in_reduction_codegen.cpp |  10 +-
 .../taskgroup_task_reduction_codegen.cpp      |  10 +-
 .../OpenMP/taskloop_in_reduction_codegen.cpp  |  10 +-
 .../OpenMP/taskloop_reduction_codegen.cpp     |  12 +-
 .../taskloop_simd_in_reduction_codegen.cpp    |  10 +-
 .../taskloop_simd_reduction_codegen.cpp       |  12 +-
 ...te_parallel_for_reduction_task_codegen.cpp |  44 ++--
 91 files changed, 1097 insertions(+), 1110 deletions(-)

diff --git a/clang/lib/CodeGen/CGBuilder.h b/clang/lib/CodeGen/CGBuilder.h
index b8036cf6e6a306..08730a6a6672a1 100644
--- a/clang/lib/CodeGen/CGBuilder.h
+++ b/clang/lib/CodeGen/CGBuilder.h
@@ -14,7 +14,6 @@
 #include "CodeGenTypeCache.h"
 #include "llvm/Analysis/Utils/Local.h"
 #include "llvm/IR/DataLayout.h"
-#include "llvm/IR/GEPNoWrapFlags.h"
 #include "llvm/IR/IRBuilder.h"
 #include "llvm/IR/Type.h"
 
@@ -335,10 +334,9 @@ class CGBuilderTy : public CGBuilderBaseTy {
 
   Address CreateGEP(Address Addr, ArrayRef<llvm::Value *> IdxList,
                     llvm::Type *ElementType, CharUnits Align,
-                    const Twine &Name = "",
-                    llvm::GEPNoWrapFlags NW = llvm::GEPNoWrapFlags::none()) {
+                    const Twine &Name = "") {
     llvm::Value *Ptr = emitRawPointerFromAddress(Addr);
-    return RawAddress(CreateGEP(Addr.getElementType(), Ptr, IdxList, Name, NW),
+    return RawAddress(CreateGEP(Addr.getElementType(), Ptr, IdxList, Name),
                       ElementType, Align);
   }
 
diff --git a/clang/lib/CodeGen/CGExprScalar.cpp b/clang/lib/CodeGen/CGExprScalar.cpp
index 5b55ec9d8064e0..af11bc20a3b639 100644
--- a/clang/lib/CodeGen/CGExprScalar.cpp
+++ b/clang/lib/CodeGen/CGExprScalar.cpp
@@ -36,7 +36,6 @@
 #include "llvm/IR/DerivedTypes.h"
 #include "llvm/IR/FixedPointBuilder.h"
 #include "llvm/IR/Function.h"
-#include "llvm/IR/GEPNoWrapFlags.h"
 #include "llvm/IR/GetElementPtrTypeIterator.h"
 #include "llvm/IR/GlobalVariable.h"
 #include "llvm/IR/Intrinsics.h"
@@ -5757,12 +5756,7 @@ CodeGenFunction::EmitCheckedInBoundsGEP(llvm::Type *ElemTy, Value *Ptr,
                                         bool SignedIndices, bool IsSubtraction,
                                         SourceLocation Loc, const Twine &Name) {
   llvm::Type *PtrTy = Ptr->getType();
-
-  llvm::GEPNoWrapFlags NWFlags = llvm::GEPNoWrapFlags::inBounds();
-  if (!SignedIndices && !IsSubtraction)
-    NWFlags |= llvm::GEPNoWrapFlags::noUnsignedWrap();
-
-  Value *GEPVal = Builder.CreateGEP(ElemTy, Ptr, IdxList, Name, NWFlags);
+  Value *GEPVal = Builder.CreateInBoundsGEP(ElemTy, Ptr, IdxList, Name);
 
   // If the pointer overflow sanitizer isn't enabled, do nothing.
   if (!SanOpts.has(SanitizerKind::PointerOverflow))
@@ -5877,13 +5871,8 @@ Address CodeGenFunction::EmitCheckedInBoundsGEP(
     Address Addr, ArrayRef<Value *> IdxList, llvm::Type *elementType,
     bool SignedIndices, bool IsSubtraction, SourceLocation Loc, CharUnits Align,
     const Twine &Name) {
-  if (!SanOpts.has(SanitizerKind::PointerOverflow)) {
-    llvm::GEPNoWrapFlags NWFlags = llvm::GEPNoWrapFlags::inBounds();
-    if (!SignedIndices && !IsSubtraction)
-      NWFlags |= llvm::GEPNoWrapFlags::noUnsignedWrap();
-
-    return Builder.CreateGEP(Addr, IdxList, elementType, Align, Name, NWFlags);
-  }
+  if (!SanOpts.has(SanitizerKind::PointerOverflow))
+    return Builder.CreateInBoundsGEP(Addr, IdxList, elementType, Align, Name);
 
   return RawAddress(
       EmitCheckedInBoundsGEP(Addr.getElementType(), Addr.emitRawPointer(*this),
diff --git a/clang/test/CodeGen/2005-01-02-ConstantInits.c b/clang/test/CodeGen/2005-01-02-ConstantInits.c
index d90c2ea42da613..7772a64331ffb7 100644
--- a/clang/test/CodeGen/2005-01-02-ConstantInits.c
+++ b/clang/test/CodeGen/2005-01-02-ConstantInits.c
@@ -1,4 +1,4 @@
-// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --function-signature --check-globals --global-value-regex "[A-Za-z].*"
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --function-signature --check-globals --global-value-regex "@.+"
 // RUN: %clang_cc1 -triple=x86_64-unknown-linux %s -emit-llvm -o - | FileCheck %s
 
 // This tests all kinds of hard cases with initializers and
@@ -51,7 +51,7 @@ int foo(int i) { return bar(&Arr[49])+bar(&Arr[i]); }
 // CHECK-NEXT:    store i32 [[I]], ptr [[I_ADDR]], align 4
 // CHECK-NEXT:    store ptr @Arr, ptr [[P]], align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[P]], align 8
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP0]], i32 1
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP0]], i32 1
 // CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[P]], align 8
 // CHECK-NEXT:    [[TMP1:%.*]] = load i32, ptr [[I_ADDR]], align 4
 // CHECK-NEXT:    [[IDX_EXT:%.*]] = sext i32 [[TMP1]] to i64
diff --git a/clang/test/CodeGen/PowerPC/ppc-emmintrin.c b/clang/test/CodeGen/PowerPC/ppc-emmintrin.c
index 4c4d0dfce05eaf..a3650beec625f2 100644
--- a/clang/test/CodeGen/PowerPC/ppc-emmintrin.c
+++ b/clang/test/CodeGen/PowerPC/ppc-emmintrin.c
@@ -1012,14 +1012,14 @@ test_shuffle() {
 // CHECK: %[[SHR:[0-9a-zA-Z_.]+]] = ashr i32 %{{[0-9a-zA-Z_.]+}}, 6
 // CHECK: %[[AND4:[0-9a-zA-Z_.]+]] = and i32 %[[SHR]], 3
 // CHECK: sext i32 %[[AND4]] to i64
-// CHECK: getelementptr inbounds nuw [4 x i32], ptr @_mm_shuffle_epi32.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
+// CHECK: getelementptr inbounds [4 x i32], ptr @_mm_shuffle_epi32.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
 // CHECK: insertelement <4 x i32> %{{[0-9a-zA-Z_.]+}}, i32 %{{[0-9a-zA-Z_.]+}}, i32 0
-// CHECK: getelementptr inbounds nuw [4 x i32], ptr @_mm_shuffle_epi32.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
+// CHECK: getelementptr inbounds [4 x i32], ptr @_mm_shuffle_epi32.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
 // CHECK: insertelement <4 x i32> %{{[0-9a-zA-Z_.]+}}, i32 %{{[0-9a-zA-Z_.]+}}, i32 1
-// CHECK: getelementptr inbounds nuw [4 x i32], ptr @_mm_shuffle_epi32.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
+// CHECK: getelementptr inbounds [4 x i32], ptr @_mm_shuffle_epi32.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
 // CHECK: %[[ADD:[0-9a-zA-Z_.]+]] = add i32 %{{[0-9a-zA-Z_.]+}}, 269488144
 // CHECK: insertelement <4 x i32> %{{[0-9a-zA-Z_.]+}}, i32 %[[ADD]], i32 2
-// CHECK: getelementptr inbounds nuw [4 x i32], ptr @_mm_shuffle_epi32.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
+// CHECK: getelementptr inbounds [4 x i32], ptr @_mm_shuffle_epi32.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
 // CHECK: add i32 %{{[0-9a-zA-Z_.]+}}, 269488144
 // CHECK: call <4 x i32> @vec_perm(int vector[4], int vector[4], unsigned char vector[16])
 
@@ -1050,7 +1050,7 @@ test_shuffle() {
 // CHECK: sext i32 %[[AND4]] to i64
 // CHECK-LE: store <2 x i64> <i64 1663540288323457296, i64 0>, ptr %{{[0-9a-zA-Z_.]+}}, align 16
 // CHECK-BE: store <2 x i64> <i64 1157726452361532951, i64 0>, ptr %{{[0-9a-zA-Z_.]+}}, align 16
-// CHECK-COUNT-4: getelementptr inbounds nuw [4 x i16], ptr @_mm_shufflehi_epi16.__permute_selectors, i64 0, i64 {{[0-9a-zA-Z_%.]+}}
+// CHECK-COUNT-4: getelementptr inbounds [4 x i16], ptr @_mm_shufflehi_epi16.__permute_selectors, i64 0, i64 {{[0-9a-zA-Z_%.]+}}
 // CHECK: call <2 x i64> @vec_perm(unsigned long long vector[2], unsigned long long vector[2], unsigned char vector[16])
 
 // CHECK-LABEL: define available_externally <2 x i64> @_mm_shufflelo_epi16
@@ -1067,7 +1067,7 @@ test_shuffle() {
 // CHECK: sext i32 %[[AND4]] to i64
 // CHECK-LE: store <2 x i64> <i64 0, i64 2242261671028070680>, ptr %{{[0-9a-zA-Z_.]+}}, align 16
 // CHECK-BE: store <2 x i64> <i64 0, i64 1736447835066146335>, ptr %{{[0-9a-zA-Z_.]+}}, align 16
-// CHECK-COUNT-4: getelementptr inbounds nuw [4 x i16], ptr @_mm_shufflelo_epi16.__permute_selectors, i64 0, i64 {{[0-9a-zA-Z_%.]+}}
+// CHECK-COUNT-4: getelementptr inbounds [4 x i16], ptr @_mm_shufflelo_epi16.__permute_selectors, i64 0, i64 {{[0-9a-zA-Z_%.]+}}
 // CHECK: call <2 x i64> @vec_perm(unsigned long long vector[2], unsigned long long vector[2], unsigned char vector[16])
 
 void __attribute__((noinline))
diff --git a/clang/test/CodeGen/PowerPC/ppc-xmmintrin.c b/clang/test/CodeGen/PowerPC/ppc-xmmintrin.c
index 4a15fa9f76ceea..95dfd1202f1575 100644
--- a/clang/test/CodeGen/PowerPC/ppc-xmmintrin.c
+++ b/clang/test/CodeGen/PowerPC/ppc-xmmintrin.c
@@ -894,16 +894,16 @@ test_shuffle() {
 // CHECK: %[[SHR3:[0-9a-zA-Z_.]+]] = ashr i32 %{{[0-9a-zA-Z_.]+}}, 6
 // CHECK: %[[AND4:[0-9a-zA-Z_.]+]] = and i32 %[[SHR3]], 3
 // CHECK: sext i32 %[[AND4]] to i64
-// CHECK: getelementptr inbounds nuw [4 x i16], ptr @_mm_shuffle_pi16.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
+// CHECK: getelementptr inbounds [4 x i16], ptr @_mm_shuffle_pi16.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
 // CHECK-LE: getelementptr inbounds [4 x i16], ptr %{{[0-9a-zA-Z_.]+}}, i64 0, i64 0
 // CHECK-BE: getelementptr inbounds [4 x i16], ptr %{{[0-9a-zA-Z_.]+}}, i64 0, i64 3
-// CHECK: getelementptr inbounds nuw [4 x i16], ptr @_mm_shuffle_pi16.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
+// CHECK: getelementptr inbounds [4 x i16], ptr @_mm_shuffle_pi16.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
 // CHECK-LE: getelementptr inbounds [4 x i16], ptr %{{[0-9a-zA-Z_.]+}}, i64 0, i64 1
 // CHECK-BE: getelementptr inbounds [4 x i16], ptr %{{[0-9a-zA-Z_.]+}}, i64 0, i64 2
-// CHECK: getelementptr inbounds nuw [4 x i16], ptr @_mm_shuffle_pi16.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
+// CHECK: getelementptr inbounds [4 x i16], ptr @_mm_shuffle_pi16.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
 // CHECK-LE: getelementptr inbounds [4 x i16], ptr %{{[0-9a-zA-Z_.]+}}, i64 0, i64 2
 // CHECK-BE: getelementptr inbounds [4 x i16], ptr %{{[0-9a-zA-Z_.]+}}, i64 0, i64 1
-// CHECK: getelementptr inbounds nuw [4 x i16], ptr @_mm_shuffle_pi16.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
+// CHECK: getelementptr inbounds [4 x i16], ptr @_mm_shuffle_pi16.__permute_selectors, i64 0, i64 %{{[0-9a-zA-Z_.]+}}
 // CHECK-LE: getelementptr inbounds [4 x i16], ptr %{{[0-9a-zA-Z_.]+}}, i64 0, i64 3
 // CHECK-BE: getelementptr inbounds [4 x i16], ptr %{{[0-9a-zA-Z_.]+}}, i64 0, i64 0
 // CHECK: call <2 x i64> @vec_splats(unsigned long long)
@@ -923,14 +923,14 @@ test_shuffle() {
 // CHECK: %[[SHR3:[0-9a-zA-Z_.]+]] = ashr i32 %{{[0-9a-zA-Z_.]+}}, 6
 // CHECK: %[[AND4:[0-9a-zA-Z_.]+]] = and i32 %[[SHR3]], 3
 // CHECK: sext i32 %[[AND4]] to i64
-// CHECK: getelementptr inbounds nuw [4 x i32], ptr @_mm_shuffle_ps.__permute_selectors, i64 0, i64
+// CHECK: getelementptr inbounds [4 x i32], ptr @_mm_shuffle_ps.__permute_selectors, i64 0, i64
 // CHECK: insertelement <4 x i32> %{{[0-9a-zA-Z_.]+}}, i32 %{{[0-9a-zA-Z_.]+}}, i32 0
-// CHECK: getelementptr inbounds nuw [4 x i32], ptr @_mm_shuffle_ps.__permute_selectors, i64 0, i64
+// CHECK: getelementptr inbounds [4 x i32], ptr @_mm_shuffle_ps.__permute_selectors, i64 0, i64
 // CHECK: insertelement <4 x i32> %{{[0-9a-zA-Z_.]+}}, i32 %{{[0-9a-zA-Z_.]+}}, i32 1
-// CHECK: getelementptr inbounds nuw [4 x i32], ptr @_mm_shuffle_ps.__permute_selectors, i64 0, i64
+// CHECK: getelementptr inbounds [4 x i32], ptr @_mm_shuffle_ps.__permute_selectors, i64 0, i64
 // CHECK: %[[ADD:[0-9a-zA-Z_.]+]] = add i32 %{{[0-9a-zA-Z_.]+}}, 269488144
 // CHECK: insertelement <4 x i32> %{{[0-9a-zA-Z_.]+}}, i32 %[[ADD]], i32 2
-// CHECK: getelementptr inbounds nuw [4 x i32], ptr @_mm_shuffle_ps.__permute_selectors, i64 0, i64
+// CHECK: getelementptr inbounds [4 x i32], ptr @_mm_shuffle_ps.__permute_selectors, i64 0, i64
 // CHECK: %[[ADD2:[0-9a-zA-Z_.]+]] = add i32 %{{[0-9a-zA-Z_.]+}}, 269488144
 // CHECK: insertelement <4 x i32> %{{[0-9a-zA-Z_.]+}}, i32 %[[ADD2]], i32 3
 // CHECK: call <4 x float> @vec_perm(float vector[4], float vector[4], unsigned char vector[16])
diff --git a/clang/test/CodeGen/attr-counted-by.c b/clang/test/CodeGen/attr-counted-by.c
index f3b1653d3632af..3ed8b6f0c71861 100644
--- a/clang/test/CodeGen/attr-counted-by.c
+++ b/clang/test/CodeGen/attr-counted-by.c
@@ -118,7 +118,7 @@ void test1(struct annotated *p, int index, int val) {
 // SANITIZE-WITH-ATTR-NEXT:    unreachable, !nosanitize [[META2]]
 // SANITIZE-WITH-ATTR:       cont3:
 // SANITIZE-WITH-ATTR-NEXT:    [[ARRAY:%.*]] = getelementptr inbounds nuw i8, ptr [[P]], i64 12
-// SANITIZE-WITH-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
+// SANITIZE-WITH-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
 // SANITIZE-WITH-ATTR-NEXT:    [[TMP2:%.*]] = shl i32 [[DOT_COUNTED_BY_LOAD]], 2
 // SANITIZE-WITH-ATTR-NEXT:    [[DOTINV:%.*]] = icmp slt i32 [[DOT_COUNTED_BY_LOAD]], 0
 // SANITIZE-WITH-ATTR-NEXT:    [[CONV:%.*]] = select i1 [[DOTINV]], i32 0, i32 [[TMP2]]
@@ -134,7 +134,7 @@ void test1(struct annotated *p, int index, int val) {
 // NO-SANITIZE-WITH-ATTR-NEXT:    [[DOTINV:%.*]] = icmp slt i32 [[DOT_COUNTED_BY_LOAD]], 0
 // NO-SANITIZE-WITH-ATTR-NEXT:    [[CONV:%.*]] = select i1 [[DOTINV]], i32 0, i32 [[TMP0]]
 // NO-SANITIZE-WITH-ATTR-NEXT:    [[ARRAY:%.*]] = getelementptr inbounds nuw i8, ptr [[P]], i64 12
-// NO-SANITIZE-WITH-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
+// NO-SANITIZE-WITH-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
 // NO-SANITIZE-WITH-ATTR-NEXT:    store i32 [[CONV]], ptr [[ARRAYIDX]], align 4, !tbaa [[TBAA2]]
 // NO-SANITIZE-WITH-ATTR-NEXT:    ret void
 //
@@ -142,7 +142,7 @@ void test1(struct annotated *p, int index, int val) {
 // SANITIZE-WITHOUT-ATTR-SAME: ptr noundef [[P:%.*]], i64 noundef [[INDEX:%.*]]) local_unnamed_addr #[[ATTR0]] {
 // SANITIZE-WITHOUT-ATTR-NEXT:  entry:
 // SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAY:%.*]] = getelementptr inbounds nuw i8, ptr [[P]], i64 12
-// SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
+// SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
 // SANITIZE-WITHOUT-ATTR-NEXT:    store i32 -1, ptr [[ARRAYIDX]], align 4, !tbaa [[TBAA2]]
 // SANITIZE-WITHOUT-ATTR-NEXT:    ret void
 //
@@ -150,7 +150,7 @@ void test1(struct annotated *p, int index, int val) {
 // NO-SANITIZE-WITHOUT-ATTR-SAME: ptr noundef [[P:%.*]], i64 noundef [[INDEX:%.*]]) local_unnamed_addr #[[ATTR0]] {
 // NO-SANITIZE-WITHOUT-ATTR-NEXT:  entry:
 // NO-SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAY:%.*]] = getelementptr inbounds nuw i8, ptr [[P]], i64 12
-// NO-SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
+// NO-SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
 // NO-SANITIZE-WITHOUT-ATTR-NEXT:    store i32 -1, ptr [[ARRAYIDX]], align 4, !tbaa [[TBAA2]]
 // NO-SANITIZE-WITHOUT-ATTR-NEXT:    ret void
 //
@@ -207,7 +207,7 @@ size_t test2_bdos(struct annotated *p) {
 // SANITIZE-WITH-ATTR-NEXT:    unreachable, !nosanitize [[META2]]
 // SANITIZE-WITH-ATTR:       cont3:
 // SANITIZE-WITH-ATTR-NEXT:    [[ARRAY:%.*]] = getelementptr inbounds nuw i8, ptr [[P]], i64 12
-// SANITIZE-WITH-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
+// SANITIZE-WITH-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
 // SANITIZE-WITH-ATTR-NEXT:    [[TMP2:%.*]] = sext i32 [[DOT_COUNTED_BY_LOAD]] to i64
 // SANITIZE-WITH-ATTR-NEXT:    [[TMP3:%.*]] = shl nsw i64 [[TMP2]], 2
 // SANITIZE-WITH-ATTR-NEXT:    [[TMP4:%.*]] = tail call i64 @llvm.smax.i64(i64 [[TMP3]], i64 4)
@@ -231,7 +231,7 @@ size_t test2_bdos(struct annotated *p) {
 // NO-SANITIZE-WITH-ATTR-NEXT:    [[DOTINV:%.*]] = icmp slt i32 [[DOT_COUNTED_BY_LOAD]], 0
 // NO-SANITIZE-WITH-ATTR-NEXT:    [[CONV:%.*]] = select i1 [[DOTINV]], i32 0, i32 [[TMP4]]
 // NO-SANITIZE-WITH-ATTR-NEXT:    [[ARRAY:%.*]] = getelementptr inbounds nuw i8, ptr [[P]], i64 12
-// NO-SANITIZE-WITH-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
+// NO-SANITIZE-WITH-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
 // NO-SANITIZE-WITH-ATTR-NEXT:    store i32 [[CONV]], ptr [[ARRAYIDX]], align 4, !tbaa [[TBAA2]]
 // NO-SANITIZE-WITH-ATTR-NEXT:    ret void
 //
@@ -239,7 +239,7 @@ size_t test2_bdos(struct annotated *p) {
 // SANITIZE-WITHOUT-ATTR-SAME: ptr noundef [[P:%.*]], i64 noundef [[INDEX:%.*]]) local_unnamed_addr #[[ATTR0]] {
 // SANITIZE-WITHOUT-ATTR-NEXT:  entry:
 // SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAY:%.*]] = getelementptr inbounds nuw i8, ptr [[P]], i64 12
-// SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
+// SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
 // SANITIZE-WITHOUT-ATTR-NEXT:    store i32 -1, ptr [[ARRAYIDX]], align 4, !tbaa [[TBAA2]]
 // SANITIZE-WITHOUT-ATTR-NEXT:    ret void
 //
@@ -247,7 +247,7 @@ size_t test2_bdos(struct annotated *p) {
 // NO-SANITIZE-WITHOUT-ATTR-SAME: ptr noundef [[P:%.*]], i64 noundef [[INDEX:%.*]]) local_unnamed_addr #[[ATTR0]] {
 // NO-SANITIZE-WITHOUT-ATTR-NEXT:  entry:
 // NO-SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAY:%.*]] = getelementptr inbounds nuw i8, ptr [[P]], i64 12
-// NO-SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
+// NO-SANITIZE-WITHOUT-ATTR-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [0 x i32], ptr [[ARRAY]], i64 0, i64 [[INDEX]]
 // NO-SANITIZE-WITHOUT-ATTR-NEXT:    store i32 -1, ptr [[ARRAYIDX]], align 4, !tbaa [[TBAA2]]
 // NO-SANITIZE-WITHOUT-ATTR-NEXT:    ret void
 //
diff --git a/clang/test/CodeGen/catch-nullptr-and-nonzero-offset-when-nullptr-is-defined.c b/clang/test/CodeGen/catch-nullptr-and-nonzero-offset-when-nullptr-is-defined.c
index 8a560a47ad1e10..39ede01d6e3b83 100644
--- a/clang/test/CodeGen/catch-nullptr-and-nonzero-offset-when-nullptr-is-defined.c
+++ b/clang/test/CodeGen/catch-nullptr-and-nonzero-offset-when-nullptr-is-defined.c
@@ -33,7 +33,7 @@ char *add_unsigned(char *base, unsigned long offset) {
   // CHECK-NEXT:                        store i64 %[[OFFSET]], ptr %[[OFFSET_ADDR]], align 8
   // CHECK-NEXT:                        %[[BASE_RELOADED:.*]] = load ptr, ptr %[[BASE_ADDR]], align 8
   // CHECK-NEXT:                        %[[OFFSET_RELOADED:.*]] = load i64, ptr %[[OFFSET_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr %[[BASE_RELOADED]], i64 %[[OFFSET_RELOADED]]
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr %[[BASE_RELOADED]], i64 %[[OFFSET_RELOADED]]
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_AGGREGATE:.*]] = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 1, i64 %[[OFFSET_RELOADED]]), !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_OVERFLOWED:.*]] = extractvalue { i64, i1 } %[[COMPUTED_OFFSET_AGGREGATE]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[OR_OV:.+]] = or i1 %[[COMPUTED_OFFSET_OVERFLOWED]], false, !nosanitize
diff --git a/clang/test/CodeGen/catch-nullptr-and-nonzero-offset.c b/clang/test/CodeGen/catch-nullptr-and-nonzero-offset.c
index d884993ffb2b30..e93dbcb9f647bf 100644
--- a/clang/test/CodeGen/catch-nullptr-and-nonzero-offset.c
+++ b/clang/test/CodeGen/catch-nullptr-and-nonzero-offset.c
@@ -50,7 +50,7 @@ char *var_var(char *base, unsigned long offset) {
   // CHECK-NEXT:                        store i64 %[[OFFSET]], ptr %[[OFFSET_ADDR]], align 8
   // CHECK-NEXT:                        %[[BASE_RELOADED:.*]] = load ptr, ptr %[[BASE_ADDR]], align 8
   // CHECK-NEXT:                        %[[OFFSET_RELOADED:.*]] = load i64, ptr %[[OFFSET_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr %[[BASE_RELOADED]], i64 %[[OFFSET_RELOADED]]
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr %[[BASE_RELOADED]], i64 %[[OFFSET_RELOADED]]
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_AGGREGATE:.*]] = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 1, i64 %[[OFFSET_RELOADED]]), !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_OVERFLOWED:.*]] = extractvalue { i64, i1 } %[[COMPUTED_OFFSET_AGGREGATE]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[OR_OV:.+]] = or i1 %[[COMPUTED_OFFSET_OVERFLOWED]], false, !nosanitize
@@ -83,7 +83,7 @@ char *var_zero(char *base) {
   // CHECK-NEXT:                          %[[BASE_ADDR:.*]] = alloca ptr, align 8
   // CHECK-NEXT:                          store ptr %[[BASE]], ptr %[[BASE_ADDR]], align 8
   // CHECK-NEXT:                          %[[BASE_RELOADED:.*]] = load ptr, ptr %[[BASE_ADDR]], align 8
-  // CHECK-NEXT:                          %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr %[[BASE_RELOADED]], i64 0
+  // CHECK-NEXT:                          %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr %[[BASE_RELOADED]], i64 0
   // CHECK-SANITIZE-C-NEXT:               %[[BASE_RELOADED_INT:.*]] = ptrtoint ptr %[[BASE_RELOADED]] to i64, !nosanitize
   // CHECK-SANITIZE-C-NEXT:               %[[COMPUTED_GEP:.*]] = add i64 %[[BASE_RELOADED_INT]], 0, !nosanitize
   // CHECK-SANITIZE-C-NEXT:               %[[BASE_IS_NOT_NULLPTR:.*]] = icmp ne ptr %[[BASE_RELOADED]], null, !nosanitize
@@ -111,7 +111,7 @@ char *var_one(char *base) {
   // CHECK-NEXT:                        %[[BASE_ADDR:.*]] = alloca ptr, align 8
   // CHECK-NEXT:                        store ptr %[[BASE]], ptr %[[BASE_ADDR]], align 8
   // CHECK-NEXT:                        %[[BASE_RELOADED:.*]] = load ptr, ptr %[[BASE_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr %[[BASE_RELOADED]], i64 1
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr %[[BASE_RELOADED]], i64 1
   // CHECK-SANITIZE-NEXT:               %[[BASE_RELOADED_INT:.*]] = ptrtoint ptr %[[BASE_RELOADED]] to i64, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_GEP:.*]] = add i64 %[[BASE_RELOADED_INT]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[BASE_IS_NOT_NULLPTR:.*]] = icmp ne ptr %[[BASE_RELOADED]], null, !nosanitize
@@ -140,7 +140,7 @@ char *var_allones(char *base) {
   // CHECK-NEXT:                        %[[BASE_ADDR:.*]] = alloca ptr, align 8
   // CHECK-NEXT:                        store ptr %[[BASE]], ptr %[[BASE_ADDR]], align 8
   // CHECK-NEXT:                        %[[BASE_RELOADED:.*]] = load ptr, ptr %[[BASE_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr %[[BASE_RELOADED]], i64 -1
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr %[[BASE_RELOADED]], i64 -1
   // CHECK-SANITIZE-NEXT:               %[[BASE_RELOADED_INT:.*]] = ptrtoint ptr %[[BASE_RELOADED]] to i64, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_GEP:.*]] = add i64 %[[BASE_RELOADED_INT]], -1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[BASE_IS_NOT_NULLPTR:.*]] = icmp ne ptr %[[BASE_RELOADED]], null, !nosanitize
@@ -171,7 +171,7 @@ char *nullptr_var(unsigned long offset) {
   // CHECK-NEXT:                        %[[OFFSET_ADDR:.*]] = alloca i64, align 8
   // CHECK-NEXT:                        store i64 %[[OFFSET]], ptr %[[OFFSET_ADDR]], align 8
   // CHECK-NEXT:                        %[[OFFSET_RELOADED:.*]] = load i64, ptr %[[OFFSET_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr null, i64 %[[OFFSET_RELOADED]]
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr null, i64 %[[OFFSET_RELOADED]]
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_AGGREGATE:.*]] = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 1, i64 %[[OFFSET_RELOADED]]), !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_OVERFLOWED:.*]] = extractvalue { i64, i1 } %[[COMPUTED_OFFSET_AGGREGATE]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[OR_OV:.+]] = or i1 %[[COMPUTED_OFFSET_OVERFLOWED]], false, !nosanitize
@@ -217,17 +217,17 @@ char *nullptr_zero(void) {
 char *nullptr_one_BAD(void) {
   // CHECK:                           define{{.*}} ptr @nullptr_one_BAD()
   // CHECK-NEXT:                      [[ENTRY:.*]]:
-  // CHECK-SANITIZE-NEXT:             %[[CMP:.*]] = icmp ne i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr null, i64 1) to i64), 0, !nosanitize
+  // CHECK-SANITIZE-NEXT:             %[[CMP:.*]] = icmp ne i64 ptrtoint (ptr getelementptr inbounds (i8, ptr null, i64 1) to i64), 0, !nosanitize
   // CHECK-SANITIZE-C-NEXT:           %[[COND:.*]] = and i1 false, %[[CMP]], !nosanitize
   // CHECK-SANITIZE-CPP-NEXT:         %[[COND:.*]] = icmp eq i1 false, %[[CMP]], !nosanitize
   // CHECK-SANITIZE-NEXT:             br i1 %[[COND]], label %[[CONT:.*]], label %[[HANDLER_POINTER_OVERFLOW:[^,]+]],{{.*}} !nosanitize
   // CHECK-SANITIZE:                  [[HANDLER_POINTER_OVERFLOW]]:
-  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_700]], i64 0, i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr null, i64 1) to i64))
-  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_700]], i64 0, i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr null, i64 1) to i64))
+  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_700]], i64 0, i64 ptrtoint (ptr getelementptr inbounds (i8, ptr null, i64 1) to i64))
+  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_700]], i64 0, i64 ptrtoint (ptr getelementptr inbounds (i8, ptr null, i64 1) to i64))
   // CHECK-SANITIZE-TRAP-NEXT:          call void @llvm.ubsantrap(i8 19){{.*}}, !nosanitize
   // CHECK-SANITIZE-UNREACHABLE-NEXT:   unreachable, !nosanitize
   // CHECK-SANITIZE:                  [[CONT]]:
-  // CHECK-NEXT:                        ret ptr getelementptr inbounds nuw (i8, ptr null, i64 1)
+  // CHECK-NEXT:                        ret ptr getelementptr inbounds (i8, ptr null, i64 1)
   static char *const base = (char *)0;
   static const unsigned long offset = 1;
 #line 700
@@ -237,17 +237,17 @@ char *nullptr_one_BAD(void) {
 char *nullptr_allones_BAD(void) {
   // CHECK:                           define{{.*}} ptr @nullptr_allones_BAD()
   // CHECK-NEXT:                      [[ENTRY:.*]]:
-  // CHECK-SANITIZE-NEXT:             %[[CMP:.*]] = icmp ne i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr null, i64 -1) to i64), 0, !nosanitize
+  // CHECK-SANITIZE-NEXT:             %[[CMP:.*]] = icmp ne i64 ptrtoint (ptr getelementptr inbounds (i8, ptr null, i64 -1) to i64), 0, !nosanitize
   // CHECK-SANITIZE-C-NEXT:           %[[COND:.*]] = and i1 false, %[[CMP]], !nosanitize
   // CHECK-SANITIZE-CPP-NEXT:         %[[COND:.*]] = icmp eq i1 false, %[[CMP]], !nosanitize
   // CHECK-SANITIZE-NEXT:             br i1 %[[COND]], label %[[CONT:.*]], label %[[HANDLER_POINTER_OVERFLOW:[^,]+]],{{.*}} !nosanitize
   // CHECK-SANITIZE:                  [[HANDLER_POINTER_OVERFLOW]]:
-  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_800]], i64 0, i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr null, i64 -1) to i64))
-  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_800]], i64 0, i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr null, i64 -1) to i64))
+  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_800]], i64 0, i64 ptrtoint (ptr getelementptr inbounds (i8, ptr null, i64 -1) to i64))
+  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_800]], i64 0, i64 ptrtoint (ptr getelementptr inbounds (i8, ptr null, i64 -1) to i64))
   // CHECK-SANITIZE-TRAP-NEXT:          call void @llvm.ubsantrap(i8 19){{.*}}, !nosanitize
   // CHECK-SANITIZE-UNREACHABLE-NEXT:   unreachable, !nosanitize
   // CHECK-SANITIZE:                  [[CONT]]:
-  // CHECK-NEXT:                        ret ptr getelementptr inbounds nuw (i8, ptr null, i64 -1)
+  // CHECK-NEXT:                        ret ptr getelementptr inbounds (i8, ptr null, i64 -1)
   static char *const base = (char *)0;
   static const unsigned long offset = -1;
 #line 800
@@ -262,7 +262,7 @@ char *one_var(unsigned long offset) {
   // CHECK-NEXT:                        %[[OFFSET_ADDR:.*]] = alloca i64, align 8
   // CHECK-NEXT:                        store i64 %[[OFFSET]], ptr %[[OFFSET_ADDR]], align 8
   // CHECK-NEXT:                        %[[OFFSET_RELOADED:.*]] = load i64, ptr %[[OFFSET_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr inttoptr (i64 1 to ptr), i64 %[[OFFSET_RELOADED]]
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr inttoptr (i64 1 to ptr), i64 %[[OFFSET_RELOADED]]
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_AGGREGATE:.*]] = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 1, i64 %[[OFFSET_RELOADED]]), !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_OVERFLOWED:.*]] = extractvalue { i64, i1 } %[[COMPUTED_OFFSET_AGGREGATE]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[OR_OV:.+]] = or i1 %[[COMPUTED_OFFSET_OVERFLOWED]], false, !nosanitize
@@ -312,17 +312,17 @@ char *one_one_OK(void) {
   // CHECK:                           define{{.*}} ptr @one_one_OK()
   // CHECK-NEXT:                      [[ENTRY:.*]]:
   // CHECK-SANITIZE-NEXT:               %[[CMP1:.*]] = icmp ne ptr inttoptr (i64 1 to ptr), null, !nosanitize
-  // CHECK-SANITIZE-NEXT:               %[[CMP2:.*]] = icmp ne i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 1 to ptr), i64 1) to i64), i64 1), i64 1), 0, !nosanitize
+  // CHECK-SANITIZE-NEXT:               %[[CMP2:.*]] = icmp ne i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 1 to ptr), i64 1) to i64), i64 1), i64 1), 0, !nosanitize
   // CHECK-SANITIZE-C-NEXT:             %[[COND:.*]] = and i1 %[[CMP1]], %[[CMP2]], !nosanitize
   // CHECK-SANITIZE-CPP-NEXT:           %[[COND:.*]] = icmp eq i1 %[[CMP1]], %[[CMP2]], !nosanitize
   // CHECK-SANITIZE-NEXT:               br i1 %[[COND]], label %[[CONT:.*]], label %[[HANDLER_POINTER_OVERFLOW:[^,]+]],{{.*}} !nosanitize
   // CHECK-SANITIZE:                  [[HANDLER_POINTER_OVERFLOW]]:
-  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_1100]], i64 1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 1 to ptr), i64 1) to i64), i64 1), i64 1))
-  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_1100]], i64 1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 1 to ptr), i64 1) to i64), i64 1), i64 1))
+  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_1100]], i64 1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 1 to ptr), i64 1) to i64), i64 1), i64 1))
+  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_1100]], i64 1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 1 to ptr), i64 1) to i64), i64 1), i64 1))
   // CHECK-SANITIZE-TRAP-NEXT:          call void @llvm.ubsantrap(i8 19){{.*}}, !nosanitize
   // CHECK-SANITIZE-UNREACHABLE-NEXT:   unreachable, !nosanitize
   // CHECK-SANITIZE:                  [[CONT]]:
-  // CHECK-NEXT:                        ret ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 1 to ptr), i64 1)
+  // CHECK-NEXT:                        ret ptr getelementptr inbounds (i8, ptr inttoptr (i64 1 to ptr), i64 1)
   static char *const base = (char *)1;
   static const unsigned long offset = 1;
 #line 1100
@@ -333,17 +333,17 @@ char *one_allones_BAD(void) {
   // CHECK:                           define{{.*}} ptr @one_allones_BAD()
   // CHECK-NEXT:                      [[ENTRY:.*]]:
   // CHECK-SANITIZE-NEXT:               %[[CMP1:.*]] = icmp ne ptr inttoptr (i64 1 to ptr), null, !nosanitize
-  // CHECK-SANITIZE-NEXT:               %[[CMP2:.*]] = icmp ne i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 1 to ptr), i64 -1) to i64), i64 1), i64 1), 0, !nosanitize
+  // CHECK-SANITIZE-NEXT:               %[[CMP2:.*]] = icmp ne i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 1 to ptr), i64 -1) to i64), i64 1), i64 1), 0, !nosanitize
   // CHECK-SANITIZE-C-NEXT:             %[[COND:.*]] = and i1 %[[CMP1]], %[[CMP2]], !nosanitize
   // CHECK-SANITIZE-CPP-NEXT:           %[[COND:.*]] = icmp eq i1 %[[CMP1]], %[[CMP2]], !nosanitize
   // CHECK-SANITIZE-NEXT:               br i1 %[[COND]], label %[[CONT:.*]], label %[[HANDLER_POINTER_OVERFLOW:[^,]+]],{{.*}} !nosanitize
   // CHECK-SANITIZE:                  [[HANDLER_POINTER_OVERFLOW]]:
-  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_1200]], i64 1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 1 to ptr), i64 -1) to i64), i64 1), i64 1))
-  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_1200]], i64 1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 1 to ptr), i64 -1) to i64), i64 1), i64 1))
+  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_1200]], i64 1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 1 to ptr), i64 -1) to i64), i64 1), i64 1))
+  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_1200]], i64 1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 1 to ptr), i64 -1) to i64), i64 1), i64 1))
   // CHECK-SANITIZE-TRAP-NEXT:          call void @llvm.ubsantrap(i8 19){{.*}}, !nosanitize
   // CHECK-SANITIZE-UNREACHABLE-NEXT:   unreachable, !nosanitize
   // CHECK-SANITIZE:                  [[CONT]]:
-  // CHECK-NEXT:                        ret ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 1 to ptr), i64 -1)
+  // CHECK-NEXT:                        ret ptr getelementptr inbounds (i8, ptr inttoptr (i64 1 to ptr), i64 -1)
   static char *const base = (char *)1;
   static const unsigned long offset = -1;
 #line 1200
@@ -358,7 +358,7 @@ char *allones_var(unsigned long offset) {
   // CHECK-NEXT:                        %[[OFFSET_ADDR:.*]] = alloca i64, align 8
   // CHECK-NEXT:                        store i64 %[[OFFSET]], ptr %[[OFFSET_ADDR]], align 8
   // CHECK-NEXT:                        %[[OFFSET_RELOADED:.*]] = load i64, ptr %[[OFFSET_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr inttoptr (i64 -1 to ptr), i64 %[[OFFSET_RELOADED]]
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr inttoptr (i64 -1 to ptr), i64 %[[OFFSET_RELOADED]]
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_AGGREGATE:.*]] = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 1, i64 %[[OFFSET_RELOADED]]), !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_OVERFLOWED:.*]] = extractvalue { i64, i1 } %[[COMPUTED_OFFSET_AGGREGATE]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[OR_OV:.+]] = or i1 %[[COMPUTED_OFFSET_OVERFLOWED]], false, !nosanitize
@@ -408,17 +408,17 @@ char *allones_one_BAD(void) {
   // CHECK: define{{.*}} ptr @allones_one_BAD()
   // CHECK-NEXT: [[ENTRY:.*]]:
   // CHECK-SANITIZE-NEXT:               %[[CMP1:.*]] = icmp ne ptr inttoptr (i64 -1 to ptr), null, !nosanitize
-  // CHECK-SANITIZE-NEXT:               %[[CMP2:.*]] = icmp ne i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 -1 to ptr), i64 1) to i64), i64 -1), i64 -1), 0, !nosanitize
+  // CHECK-SANITIZE-NEXT:               %[[CMP2:.*]] = icmp ne i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 -1 to ptr), i64 1) to i64), i64 -1), i64 -1), 0, !nosanitize
   // CHECK-SANITIZE-C-NEXT:             %[[COND:.*]] = and i1 %[[CMP1]], %[[CMP2]], !nosanitize
   // CHECK-SANITIZE-CPP-NEXT:           %[[COND:.*]] = icmp eq i1 %[[CMP1]], %[[CMP2]], !nosanitize
   // CHECK-SANITIZE-NEXT:               br i1 %[[COND]], label %[[CONT:.*]], label %[[HANDLER_POINTER_OVERFLOW:[^,]+]],{{.*}} !nosanitize
   // CHECK-SANITIZE:                  [[HANDLER_POINTER_OVERFLOW]]:
-  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_1500]], i64 -1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 -1 to ptr), i64 1) to i64), i64 -1), i64 -1))
-  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_1500]], i64 -1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 -1 to ptr), i64 1) to i64), i64 -1), i64 -1))
+  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_1500]], i64 -1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 -1 to ptr), i64 1) to i64), i64 -1), i64 -1))
+  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_1500]], i64 -1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 -1 to ptr), i64 1) to i64), i64 -1), i64 -1))
   // CHECK-SANITIZE-TRAP-NEXT:          call void @llvm.ubsantrap(i8 19){{.*}}, !nosanitize
   // CHECK-SANITIZE-UNREACHABLE-NEXT:   unreachable, !nosanitize
   // CHECK-SANITIZE:                  [[CONT]]:
-  // CHECK-NEXT:                        ret ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 -1 to ptr), i64 1)
+  // CHECK-NEXT:                        ret ptr getelementptr inbounds (i8, ptr inttoptr (i64 -1 to ptr), i64 1)
   static char *const base = (char *)-1;
   static const unsigned long offset = 1;
 #line 1500
@@ -429,17 +429,17 @@ char *allones_allones_OK(void) {
   // CHECK: define{{.*}} ptr @allones_allones_OK()
   // CHECK-NEXT: [[ENTRY:.*]]:
   // CHECK-SANITIZE-NEXT:               %[[CMP1:.*]] = icmp ne ptr inttoptr (i64 -1 to ptr), null, !nosanitize
-  // CHECK-SANITIZE-NEXT:               %[[CMP2:.*]] = icmp ne i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 -1 to ptr), i64 -1) to i64), i64 -1), i64 -1), 0, !nosanitize
+  // CHECK-SANITIZE-NEXT:               %[[CMP2:.*]] = icmp ne i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 -1 to ptr), i64 -1) to i64), i64 -1), i64 -1), 0, !nosanitize
   // CHECK-SANITIZE-C-NEXT:             %[[COND:.*]] = and i1 %[[CMP1]], %[[CMP2]], !nosanitize
   // CHECK-SANITIZE-CPP-NEXT:           %[[COND:.*]] = icmp eq i1 %[[CMP1]], %[[CMP2]], !nosanitize
   // CHECK-SANITIZE-NEXT:               br i1 %[[COND]], label %[[CONT:.*]], label %[[HANDLER_POINTER_OVERFLOW:[^,]+]],{{.*}} !nosanitize
   // CHECK-SANITIZE:                  [[HANDLER_POINTER_OVERFLOW]]:
-  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_1600]], i64 -1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 -1 to ptr), i64 -1) to i64), i64 -1), i64 -1))
-  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_1600]], i64 -1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 -1 to ptr), i64 -1) to i64), i64 -1), i64 -1))
+  // CHECK-SANITIZE-NORECOVER-NEXT:     call void @__ubsan_handle_pointer_overflow_abort(ptr @[[LINE_1600]], i64 -1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 -1 to ptr), i64 -1) to i64), i64 -1), i64 -1))
+  // CHECK-SANITIZE-RECOVER-NEXT:       call void @__ubsan_handle_pointer_overflow(ptr @[[LINE_1600]], i64 -1, i64 add (i64 sub (i64 ptrtoint (ptr getelementptr inbounds (i8, ptr inttoptr (i64 -1 to ptr), i64 -1) to i64), i64 -1), i64 -1))
   // CHECK-SANITIZE-TRAP-NEXT:          call void @llvm.ubsantrap(i8 19){{.*}}, !nosanitize
   // CHECK-SANITIZE-UNREACHABLE-NEXT:   unreachable, !nosanitize
   // CHECK-SANITIZE:                  [[CONT]]:
-  // CHECK-NEXT:                        ret ptr getelementptr inbounds nuw (i8, ptr inttoptr (i64 -1 to ptr), i64 -1)
+  // CHECK-NEXT:                        ret ptr getelementptr inbounds (i8, ptr inttoptr (i64 -1 to ptr), i64 -1)
   static char *const base = (char *)-1;
   static const unsigned long offset = -1;
 #line 1600
@@ -461,7 +461,7 @@ char *void_ptr(void *base, unsigned long offset) {
   // CHECK-NEXT:                        store i64 %[[OFFSET]], ptr %[[OFFSET_ADDR]], align 8
   // CHECK-NEXT:                        %[[BASE_RELOADED:.*]] = load ptr, ptr %[[BASE_ADDR]], align 8
   // CHECK-NEXT:                        %[[OFFSET_RELOADED:.*]] = load i64, ptr %[[OFFSET_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr %[[BASE_RELOADED]], i64 %[[OFFSET_RELOADED]]
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr %[[BASE_RELOADED]], i64 %[[OFFSET_RELOADED]]
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_AGGREGATE:.*]] = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 1, i64 %[[OFFSET_RELOADED]]), !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_OVERFLOWED:.*]] = extractvalue { i64, i1 } %[[COMPUTED_OFFSET_AGGREGATE]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[OR_OV:.+]] = or i1 %[[COMPUTED_OFFSET_OVERFLOWED]], false, !nosanitize
diff --git a/clang/test/CodeGen/catch-pointer-overflow-volatile.c b/clang/test/CodeGen/catch-pointer-overflow-volatile.c
index 626bbc0db7afb4..4b0653a0ae59ec 100644
--- a/clang/test/CodeGen/catch-pointer-overflow-volatile.c
+++ b/clang/test/CodeGen/catch-pointer-overflow-volatile.c
@@ -23,7 +23,7 @@ char *volatile_ptr(char *volatile base, unsigned long offset) {
   // CHECK-NEXT:                        store i64 %[[OFFSET]], ptr %[[OFFSET_ADDR]], align 8
   // CHECK-NEXT:                        %[[BASE_RELOADED:.*]] = load volatile ptr, ptr %[[BASE_ADDR]], align 8
   // CHECK-NEXT:                        %[[OFFSET_RELOADED:.*]] = load i64, ptr %[[OFFSET_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr %[[BASE_RELOADED]], i64 %[[OFFSET_RELOADED]]
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr %[[BASE_RELOADED]], i64 %[[OFFSET_RELOADED]]
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_AGGREGATE:.*]] = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 1, i64 %[[OFFSET_RELOADED]]), !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_OVERFLOWED:.*]] = extractvalue { i64, i1 } %[[COMPUTED_OFFSET_AGGREGATE]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[OR_OV:.+]] = or i1 %[[COMPUTED_OFFSET_OVERFLOWED]], false, !nosanitize
diff --git a/clang/test/CodeGen/catch-pointer-overflow.c b/clang/test/CodeGen/catch-pointer-overflow.c
index 1f7f1729098c79..899af73bd81e00 100644
--- a/clang/test/CodeGen/catch-pointer-overflow.c
+++ b/clang/test/CodeGen/catch-pointer-overflow.c
@@ -30,7 +30,7 @@ char *add_unsigned(char *base, unsigned long offset) {
   // CHECK-NEXT:                        store i64 %[[OFFSET]], ptr %[[OFFSET_ADDR]], align 8
   // CHECK-NEXT:                        %[[BASE_RELOADED:.*]] = load ptr, ptr %[[BASE_ADDR]], align 8
   // CHECK-NEXT:                        %[[OFFSET_RELOADED:.*]] = load i64, ptr %[[OFFSET_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr %[[BASE_RELOADED]], i64 %[[OFFSET_RELOADED]]
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr %[[BASE_RELOADED]], i64 %[[OFFSET_RELOADED]]
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_AGGREGATE:.*]] = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 1, i64 %[[OFFSET_RELOADED]]), !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_OFFSET_OVERFLOWED:.*]] = extractvalue { i64, i1 } %[[COMPUTED_OFFSET_AGGREGATE]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[OR_OV:.+]] = or i1 %[[COMPUTED_OFFSET_OVERFLOWED]], false, !nosanitize
@@ -179,7 +179,7 @@ char *postinc(char *base) {
   // CHECK-NEXT:                        %[[BASE_ADDR:.*]] = alloca ptr, align 8
   // CHECK-NEXT:                        store ptr %[[BASE]], ptr %[[BASE_ADDR]], align 8
   // CHECK-NEXT:                        %[[BASE_RELOADED:.*]] = load ptr, ptr %[[BASE_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr %[[BASE_RELOADED]], i32 1
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr %[[BASE_RELOADED]], i32 1
   // CHECK-SANITIZE-NEXT:               %[[BASE_RELOADED_INT:.*]] = ptrtoint ptr %[[BASE_RELOADED]] to i64, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_GEP:.*]] = add i64 %[[BASE_RELOADED_INT]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[BASE_IS_NOT_NULLPTR:.*]] = icmp ne ptr %[[BASE_RELOADED]], null, !nosanitize
@@ -241,7 +241,7 @@ char *preinc(char *base) {
   // CHECK-NEXT:                        %[[BASE_ADDR:.*]] = alloca ptr, align 8
   // CHECK-NEXT:                        store ptr %[[BASE]], ptr %[[BASE_ADDR]], align 8
   // CHECK-NEXT:                        %[[BASE_RELOADED:.*]] = load ptr, ptr %[[BASE_ADDR]], align 8
-  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds nuw i8, ptr %[[BASE_RELOADED]], i32 1
+  // CHECK-NEXT:                        %[[ADD_PTR:.*]] = getelementptr inbounds i8, ptr %[[BASE_RELOADED]], i32 1
   // CHECK-SANITIZE-NEXT:               %[[BASE_RELOADED_INT:.*]] = ptrtoint ptr %[[BASE_RELOADED]] to i64, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[COMPUTED_GEP:.*]] = add i64 %[[BASE_RELOADED_INT]], 1, !nosanitize
   // CHECK-SANITIZE-NEXT:               %[[BASE_IS_NOT_NULLPTR:.*]] = icmp ne ptr %[[BASE_RELOADED]], null, !nosanitize
diff --git a/clang/test/CodeGen/ext-int.c b/clang/test/CodeGen/ext-int.c
index e3d609a4ba4a2e..714b7e122a706f 100644
--- a/clang/test/CodeGen/ext-int.c
+++ b/clang/test/CodeGen/ext-int.c
@@ -154,7 +154,7 @@ _BitInt(129) *f1(_BitInt(129) *p) {
 }
 
 char *f2(char *p) {
-  // CHECK64: getelementptr inbounds nuw i8, {{.*}} i64 24
+  // CHECK64: getelementptr inbounds i8, {{.*}} i64 24
   return p + sizeof(_BitInt(129));
 }
 
diff --git a/clang/test/CodeGen/hexagon-brev-ld-ptr-incdec.c b/clang/test/CodeGen/hexagon-brev-ld-ptr-incdec.c
index d25d1e04f15fab..7802168de4d75a 100644
--- a/clang/test/CodeGen/hexagon-brev-ld-ptr-incdec.c
+++ b/clang/test/CodeGen/hexagon-brev-ld-ptr-incdec.c
@@ -6,9 +6,9 @@
 // the return value will be the value in A[2]
 // CHECK: @brev_ptr_inc
 // CHECK-DAG: llvm.hexagon.L2.loadri.pbr
-// CHECK-DAG: getelementptr inbounds nuw i8, {{.*}}i32 4
-// CHECK-NOT: getelementptr inbounds nuw i8, {{.*}}i32 8
-// CHECK-NOT: getelementptr inbounds nuw i8, {{.*}}i32 4
+// CHECK-DAG: getelementptr inbounds i8, {{.*}}i32 4
+// CHECK-NOT: getelementptr inbounds i8, {{.*}}i32 8
+// CHECK-NOT: getelementptr inbounds i8, {{.*}}i32 4
 int brev_ptr_inc(int A[], int B[]) {
   int *p0 = &B[0];
   int *p1 = &A[0];
diff --git a/clang/test/CodeGen/integer-overflow.c b/clang/test/CodeGen/integer-overflow.c
index 9e8cde8b33b16e..461b026d39615b 100644
--- a/clang/test/CodeGen/integer-overflow.c
+++ b/clang/test/CodeGen/integer-overflow.c
@@ -60,10 +60,10 @@ void test1(void) {
   // -fwrapv should turn off inbounds for GEP's, PR9256
   extern int* P;
   ++P;
-  // DEFAULT: getelementptr inbounds nuw i32, ptr
+  // DEFAULT: getelementptr inbounds i32, ptr
   // WRAPV: getelementptr i32, ptr
-  // TRAPV: getelementptr inbounds nuw i32, ptr
-  // CATCH_UB_POINTER: getelementptr inbounds nuw i32, ptr
+  // TRAPV: getelementptr inbounds i32, ptr
+  // CATCH_UB_POINTER: getelementptr inbounds i32, ptr
   // NOCATCH_UB_POINTER: getelementptr i32, ptr
 
   // PR9350: char pre-increment never overflows.
diff --git a/clang/test/CodeGen/ms-intrinsics.c b/clang/test/CodeGen/ms-intrinsics.c
index 459a708d9b2e0a..c3d64fda0b901f 100644
--- a/clang/test/CodeGen/ms-intrinsics.c
+++ b/clang/test/CodeGen/ms-intrinsics.c
@@ -156,7 +156,7 @@ unsigned char test_BitScanForward(unsigned long *Index, unsigned long Mask) {
 // CHECK:   [[RESULT:%[a-z0-9._]+]] = phi i8 [ 0, %[[ISZERO_LABEL:[a-z0-9._]+]] ], [ 1, %[[ISNOTZERO_LABEL]] ]
 // CHECK:   ret i8 [[RESULT]]
 // CHECK:   [[ISNOTZERO_LABEL]]:
-// CHECK:   [[IDXGEP:%[a-z0-9._]+]] = getelementptr inbounds nuw i8, ptr %Index, {{i64|i32}} 4
+// CHECK:   [[IDXGEP:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr %Index, {{i64|i32}} 4
 // CHECK:   [[INDEX:%[0-9]+]] = tail call range(i32 0, 33) i32 @llvm.cttz.i32(i32 %Mask, i1 true)
 // CHECK:   store i32 [[INDEX]], ptr [[IDXGEP]], align 4
 // CHECK:   br label %[[END_LABEL]]
@@ -171,7 +171,7 @@ unsigned char test_BitScanReverse(unsigned long *Index, unsigned long Mask) {
 // CHECK:   [[RESULT:%[a-z0-9._]+]] = phi i8 [ 0, %[[ISZERO_LABEL:[a-z0-9._]+]] ], [ 1, %[[ISNOTZERO_LABEL]] ]
 // CHECK:   ret i8 [[RESULT]]
 // CHECK:   [[ISNOTZERO_LABEL]]:
-// CHECK:   [[IDXGEP:%[a-z0-9._]+]] = getelementptr inbounds nuw i8, ptr %Index, {{i64|i32}} 4
+// CHECK:   [[IDXGEP:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr %Index, {{i64|i32}} 4
 // CHECK:   [[REVINDEX:%[0-9]+]] = tail call range(i32 0, 33) i32 @llvm.ctlz.i32(i32 %Mask, i1 true)
 // CHECK:   [[INDEX:%[0-9]+]] = xor i32 [[REVINDEX]], 31
 // CHECK:   store i32 [[INDEX]], ptr [[IDXGEP]], align 4
@@ -437,10 +437,10 @@ unsigned char test_InterlockedCompareExchange128(
                                         ++ExchangeLow, ++ComparandResult);
 }
 // CHECK-64: define{{.*}}i8 @test_InterlockedCompareExchange128(ptr{{[a-z_ ]*}}%Destination, i64{{[a-z_ ]*}}%ExchangeHigh, i64{{[a-z_ ]*}}%ExchangeLow, ptr{{[a-z_ ]*}}%ComparandResult){{.*}}{
-// CHECK-64: %incdec.ptr = getelementptr inbounds nuw i8, ptr %Destination, i64 8
+// CHECK-64: %incdec.ptr = getelementptr inbounds i8, ptr %Destination, i64 8
 // CHECK-64: %inc = add nsw i64 %ExchangeHigh, 1
 // CHECK-64: %inc1 = add nsw i64 %ExchangeLow, 1
-// CHECK-64: %incdec.ptr2 = getelementptr inbounds nuw i8, ptr %ComparandResult, i64 8
+// CHECK-64: %incdec.ptr2 = getelementptr inbounds i8, ptr %ComparandResult, i64 8
 // CHECK-64: [[EH:%[0-9]+]] = zext i64 %inc to i128
 // CHECK-64: [[EL:%[0-9]+]] = zext i64 %inc1 to i128
 // CHECK-64: [[EHS:%[0-9]+]] = shl nuw i128 [[EH]], 64
@@ -486,7 +486,7 @@ short test_InterlockedIncrement16(short volatile *Addend) {
   return _InterlockedIncrement16(++Addend);
 }
 // CHECK: define{{.*}}i16 @test_InterlockedIncrement16(ptr{{[a-z_ ]*}}%Addend){{.*}}{
-// CHECK: %incdec.ptr = getelementptr inbounds nuw i8, ptr %Addend, {{i64|i32}} 2
+// CHECK: %incdec.ptr = getelementptr inbounds i8, ptr %Addend, {{i64|i32}} 2
 // CHECK: [[TMP:%[0-9]+]] = atomicrmw add ptr %incdec.ptr, i16 1 seq_cst, align 2
 // CHECK: [[RESULT:%[0-9]+]] = add i16 [[TMP]], 1
 // CHECK: ret i16 [[RESULT]]
@@ -496,7 +496,7 @@ long test_InterlockedIncrement(long volatile *Addend) {
   return _InterlockedIncrement(++Addend);
 }
 // CHECK: define{{.*}}i32 @test_InterlockedIncrement(ptr{{[a-z_ ]*}}%Addend){{.*}}{
-// CHECK: %incdec.ptr = getelementptr inbounds nuw i8, ptr %Addend, {{i64|i32}} 4
+// CHECK: %incdec.ptr = getelementptr inbounds i8, ptr %Addend, {{i64|i32}} 4
 // CHECK: [[TMP:%[0-9]+]] = atomicrmw add ptr %incdec.ptr, i32 1 seq_cst, align 4
 // CHECK: [[RESULT:%[0-9]+]] = add i32 [[TMP]], 1
 // CHECK: ret i32 [[RESULT]]
diff --git a/clang/test/CodeGen/ubsan-pointer-overflow.m b/clang/test/CodeGen/ubsan-pointer-overflow.m
index 4ecdac655669f9..9192598da92fc8 100644
--- a/clang/test/CodeGen/ubsan-pointer-overflow.m
+++ b/clang/test/CodeGen/ubsan-pointer-overflow.m
@@ -5,7 +5,7 @@ void variable_len_array_arith(int n, int k) {
   int vla[n];
   int (*p)[n] = &vla;
 
-  // CHECK: getelementptr inbounds nuw i32, ptr {{.*}}, i64 [[INC:%.*]]
+  // CHECK: getelementptr inbounds i32, ptr {{.*}}, i64 [[INC:%.*]]
   // CHECK: @llvm.smul.with.overflow.i64(i64 4, i64 [[INC]]), !nosanitize
   // CHECK-NOT: select
   // CHECK: call void @__ubsan_handle_pointer_overflow{{.*}}
diff --git a/clang/test/CodeGen/vla.c b/clang/test/CodeGen/vla.c
index a22ba727df2fec..33621c5dd7a29d 100644
--- a/clang/test/CodeGen/vla.c
+++ b/clang/test/CodeGen/vla.c
@@ -120,7 +120,7 @@ int test4(unsigned n, char (*p)[n][n+1][6]) {
   // CHECK-NEXT: [[T2:%.*]] = udiv i32 [[T1]], 2
   // CHECK-NEXT: [[T3:%.*]] = mul nuw i32 [[DIM0]], [[DIM1]]
   // CHECK-NEXT: [[T4:%.*]] = mul nsw i32 [[T2]], [[T3]]
-  // CHECK-NEXT: [[T5:%.*]] = getelementptr inbounds nuw [6 x i8], ptr [[T0]], i32 [[T4]]
+  // CHECK-NEXT: [[T5:%.*]] = getelementptr inbounds [6 x i8], ptr [[T0]], i32 [[T4]]
   // CHECK-NEXT: [[T6:%.*]] = load i32, ptr [[N]], align 4
   // CHECK-NEXT: [[T7:%.*]] = udiv i32 [[T6]], 4
   // CHECK-NEXT: [[T8:%.*]] = sub i32 0, [[T7]]
diff --git a/clang/test/CodeGenCXX/attr-likelihood-iteration-stmt.cpp b/clang/test/CodeGenCXX/attr-likelihood-iteration-stmt.cpp
index fd9786de3a9499..e842de6335046f 100644
--- a/clang/test/CodeGenCXX/attr-likelihood-iteration-stmt.cpp
+++ b/clang/test/CodeGenCXX/attr-likelihood-iteration-stmt.cpp
@@ -152,16 +152,16 @@ void f_branch_elided()
 // CHECK-NEXT:    call void @llvm.lifetime.start.p0(i64 8, ptr [[__BEGIN1]]) #[[ATTR3]]
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[__RANGE1]], align 8, !tbaa [[TBAA14]]
 // CHECK-NEXT:    [[ARRAYDECAY:%.*]] = getelementptr inbounds [4 x i32], ptr [[TMP1]], i64 0, i64 0
-// CHECK-NEXT:    store ptr [[ARRAYDECAY]], ptr [[__BEGIN1]], align 8, !tbaa [[TBAA14]]
+// CHECK-NEXT:    store ptr [[ARRAYDECAY]], ptr [[__BEGIN1]], align 8, !tbaa [[TBAA16:![0-9]+]]
 // CHECK-NEXT:    call void @llvm.lifetime.start.p0(i64 8, ptr [[__END1]]) #[[ATTR3]]
 // CHECK-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[__RANGE1]], align 8, !tbaa [[TBAA14]]
 // CHECK-NEXT:    [[ARRAYDECAY1:%.*]] = getelementptr inbounds [4 x i32], ptr [[TMP2]], i64 0, i64 0
 // CHECK-NEXT:    [[ADD_PTR:%.*]] = getelementptr inbounds i32, ptr [[ARRAYDECAY1]], i64 4
-// CHECK-NEXT:    store ptr [[ADD_PTR]], ptr [[__END1]], align 8, !tbaa [[TBAA14]]
+// CHECK-NEXT:    store ptr [[ADD_PTR]], ptr [[__END1]], align 8, !tbaa [[TBAA16]]
 // CHECK-NEXT:    br label [[FOR_COND:%.*]]
 // CHECK:       for.cond:
-// CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA14]]
-// CHECK-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[__END1]], align 8, !tbaa [[TBAA14]]
+// CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA16]]
+// CHECK-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[__END1]], align 8, !tbaa [[TBAA16]]
 // CHECK-NEXT:    [[CMP:%.*]] = icmp ne ptr [[TMP3]], [[TMP4]]
 // CHECK-NEXT:    [[CMP_EXPVAL:%.*]] = call i1 @llvm.expect.i1(i1 [[CMP]], i1 true)
 // CHECK-NEXT:    br i1 [[CMP_EXPVAL]], label [[FOR_BODY:%.*]], label [[FOR_COND_CLEANUP:%.*]]
@@ -172,16 +172,16 @@ void f_branch_elided()
 // CHECK-NEXT:    br label [[FOR_END:%.*]]
 // CHECK:       for.body:
 // CHECK-NEXT:    call void @llvm.lifetime.start.p0(i64 4, ptr [[I]]) #[[ATTR3]]
-// CHECK-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA14]]
+// CHECK-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA16]]
 // CHECK-NEXT:    [[TMP6:%.*]] = load i32, ptr [[TMP5]], align 4, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i32 [[TMP6]], ptr [[I]], align 4, !tbaa [[TBAA2]]
 // CHECK-NEXT:    call void @llvm.lifetime.end.p0(i64 4, ptr [[I]]) #[[ATTR3]]
 // CHECK-NEXT:    br label [[FOR_INC:%.*]]
 // CHECK:       for.inc:
-// CHECK-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA14]]
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP7]], i32 1
-// CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[__BEGIN1]], align 8, !tbaa [[TBAA14]]
-// CHECK-NEXT:    br label [[FOR_COND]], !llvm.loop [[LOOP16:![0-9]+]]
+// CHECK-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA16]]
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP7]], i32 1
+// CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[__BEGIN1]], align 8, !tbaa [[TBAA16]]
+// CHECK-NEXT:    br label [[FOR_COND]], !llvm.loop [[LOOP18:![0-9]+]]
 // CHECK:       for.end:
 // CHECK-NEXT:    ret void
 //
@@ -204,16 +204,16 @@ void frl(int (&&e) [4])
 // CHECK-NEXT:    call void @llvm.lifetime.start.p0(i64 8, ptr [[__BEGIN1]]) #[[ATTR3]]
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[__RANGE1]], align 8, !tbaa [[TBAA14]]
 // CHECK-NEXT:    [[ARRAYDECAY:%.*]] = getelementptr inbounds [4 x i32], ptr [[TMP1]], i64 0, i64 0
-// CHECK-NEXT:    store ptr [[ARRAYDECAY]], ptr [[__BEGIN1]], align 8, !tbaa [[TBAA14]]
+// CHECK-NEXT:    store ptr [[ARRAYDECAY]], ptr [[__BEGIN1]], align 8, !tbaa [[TBAA16]]
 // CHECK-NEXT:    call void @llvm.lifetime.start.p0(i64 8, ptr [[__END1]]) #[[ATTR3]]
 // CHECK-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[__RANGE1]], align 8, !tbaa [[TBAA14]]
 // CHECK-NEXT:    [[ARRAYDECAY1:%.*]] = getelementptr inbounds [4 x i32], ptr [[TMP2]], i64 0, i64 0
 // CHECK-NEXT:    [[ADD_PTR:%.*]] = getelementptr inbounds i32, ptr [[ARRAYDECAY1]], i64 4
-// CHECK-NEXT:    store ptr [[ADD_PTR]], ptr [[__END1]], align 8, !tbaa [[TBAA14]]
+// CHECK-NEXT:    store ptr [[ADD_PTR]], ptr [[__END1]], align 8, !tbaa [[TBAA16]]
 // CHECK-NEXT:    br label [[FOR_COND:%.*]]
 // CHECK:       for.cond:
-// CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA14]]
-// CHECK-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[__END1]], align 8, !tbaa [[TBAA14]]
+// CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA16]]
+// CHECK-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[__END1]], align 8, !tbaa [[TBAA16]]
 // CHECK-NEXT:    [[CMP:%.*]] = icmp ne ptr [[TMP3]], [[TMP4]]
 // CHECK-NEXT:    [[CMP_EXPVAL:%.*]] = call i1 @llvm.expect.i1(i1 [[CMP]], i1 false)
 // CHECK-NEXT:    br i1 [[CMP_EXPVAL]], label [[FOR_BODY:%.*]], label [[FOR_COND_CLEANUP:%.*]]
@@ -224,16 +224,16 @@ void frl(int (&&e) [4])
 // CHECK-NEXT:    br label [[FOR_END:%.*]]
 // CHECK:       for.body:
 // CHECK-NEXT:    call void @llvm.lifetime.start.p0(i64 4, ptr [[I]]) #[[ATTR3]]
-// CHECK-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA14]]
+// CHECK-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA16]]
 // CHECK-NEXT:    [[TMP6:%.*]] = load i32, ptr [[TMP5]], align 4, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i32 [[TMP6]], ptr [[I]], align 4, !tbaa [[TBAA2]]
 // CHECK-NEXT:    call void @llvm.lifetime.end.p0(i64 4, ptr [[I]]) #[[ATTR3]]
 // CHECK-NEXT:    br label [[FOR_INC:%.*]]
 // CHECK:       for.inc:
-// CHECK-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA14]]
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP7]], i32 1
-// CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[__BEGIN1]], align 8, !tbaa [[TBAA14]]
-// CHECK-NEXT:    br label [[FOR_COND]], !llvm.loop [[LOOP17:![0-9]+]]
+// CHECK-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[__BEGIN1]], align 8, !tbaa [[TBAA16]]
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP7]], i32 1
+// CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[__BEGIN1]], align 8, !tbaa [[TBAA16]]
+// CHECK-NEXT:    br label [[FOR_COND]], !llvm.loop [[LOOP19:![0-9]+]]
 // CHECK:       for.end:
 // CHECK-NEXT:    ret void
 //
diff --git a/clang/test/CodeGenCXX/for-range.cpp b/clang/test/CodeGenCXX/for-range.cpp
index 088a34647c3745..10d27206d12e43 100644
--- a/clang/test/CodeGenCXX/for-range.cpp
+++ b/clang/test/CodeGenCXX/for-range.cpp
@@ -33,7 +33,7 @@ B *end(C&);
 
 extern B array[5];
 
-// CHECK-LABEL: @_Z9for_arrayv(
+// CHECK-LABEL: define {{[^@]+}}@_Z9for_arrayv(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[A:%.*]] = alloca [[STRUCT_A:%.*]], align 1
 // CHECK-NEXT:    [[__RANGE1:%.*]] = alloca ptr, align 8
@@ -57,7 +57,7 @@ extern B array[5];
 // CHECK-NEXT:    br label [[FOR_INC:%.*]]
 // CHECK:       for.inc:
 // CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[__BEGIN1]], align 8
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw [[STRUCT_B]], ptr [[TMP3]], i32 1
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds [[STRUCT_B]], ptr [[TMP3]], i32 1
 // CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[__BEGIN1]], align 8
 // CHECK-NEXT:    br label [[FOR_COND]]
 // CHECK:       for.end:
@@ -70,7 +70,7 @@ void for_array() {
   }
 }
 
-// CHECK-LABEL: @_Z9for_rangev(
+// CHECK-LABEL: define {{[^@]+}}@_Z9for_rangev(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[A:%.*]] = alloca [[STRUCT_A:%.*]], align 1
 // CHECK-NEXT:    [[__RANGE1:%.*]] = alloca ptr, align 8
@@ -103,7 +103,7 @@ void for_array() {
 // CHECK-NEXT:    br label [[FOR_INC:%.*]]
 // CHECK:       for.inc:
 // CHECK-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[__BEGIN1]], align 8
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw [[STRUCT_B]], ptr [[TMP5]], i32 1
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds [[STRUCT_B]], ptr [[TMP5]], i32 1
 // CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[__BEGIN1]], align 8
 // CHECK-NEXT:    br label [[FOR_COND]]
 // CHECK:       for.end:
@@ -116,7 +116,7 @@ void for_range() {
   }
 }
 
-// CHECK-LABEL: @_Z16for_member_rangev(
+// CHECK-LABEL: define {{[^@]+}}@_Z16for_member_rangev(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[A:%.*]] = alloca [[STRUCT_A:%.*]], align 1
 // CHECK-NEXT:    [[__RANGE1:%.*]] = alloca ptr, align 8
@@ -149,7 +149,7 @@ void for_range() {
 // CHECK-NEXT:    br label [[FOR_INC:%.*]]
 // CHECK:       for.inc:
 // CHECK-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[__BEGIN1]], align 8
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw [[STRUCT_B]], ptr [[TMP5]], i32 1
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds [[STRUCT_B]], ptr [[TMP5]], i32 1
 // CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[__BEGIN1]], align 8
 // CHECK-NEXT:    br label [[FOR_COND]]
 // CHECK:       for.end:
diff --git a/clang/test/CodeGenCXX/pr45964-decomp-transform.cpp b/clang/test/CodeGenCXX/pr45964-decomp-transform.cpp
index bcb2d875dce663..f7df110ec01298 100644
--- a/clang/test/CodeGenCXX/pr45964-decomp-transform.cpp
+++ b/clang/test/CodeGenCXX/pr45964-decomp-transform.cpp
@@ -16,7 +16,7 @@ void (*d)(){test_transform<0>};
 // CHECK-NEXT:  [[BODY]]:
 // CHECK-NEXT:  [[CUR:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[NEXT:%.*]], %[[BODY]] ]
 // CHECK-NEXT:  [[DEST:%.*]] = getelementptr inbounds i32, ptr [[BEGIN]], i64 [[CUR]]
-// CHECK-NEXT:  [[SRC:%.*]] = getelementptr inbounds nuw [1 x i32], ptr @a, i64 0, i64 [[CUR]]
+// CHECK-NEXT:  [[SRC:%.*]] = getelementptr inbounds [1 x i32], ptr @a, i64 0, i64 [[CUR]]
 // CHECK-NEXT:  [[X:%.*]] = load i32, ptr [[SRC]]
 // CHECK-NEXT:  store i32 [[X]], ptr [[DEST]]
 // CHECK-NEXT:  [[NEXT]] = add nuw i64 [[CUR]], 1
diff --git a/clang/test/CodeGenCXX/vla.cpp b/clang/test/CodeGenCXX/vla.cpp
index aadf51fce3a44c..4cf2b3b445b406 100644
--- a/clang/test/CodeGenCXX/vla.cpp
+++ b/clang/test/CodeGenCXX/vla.cpp
@@ -83,7 +83,7 @@ void test2(int b) {
   
   //CHECK: [[VLA_SIZEOF:%.*]] = mul nuw i64 4, [[VLA_NUM_ELEMENTS_PRE]]
   //CHECK-NEXT: [[VLA_NUM_ELEMENTS_POST:%.*]] = udiv i64 [[VLA_SIZEOF]], 4
-  //CHECK-NEXT: [[VLA_END_PTR:%.*]] = getelementptr inbounds nuw i32, ptr {{%.*}}, i64 [[VLA_NUM_ELEMENTS_POST]]
+  //CHECK-NEXT: [[VLA_END_PTR:%.*]] = getelementptr inbounds i32, ptr {{%.*}}, i64 [[VLA_NUM_ELEMENTS_POST]]
   //X64-NEXT: store ptr [[VLA_END_PTR]], ptr %__end1
   //AMDGCN-NEXT: store ptr [[VLA_END_PTR]], ptr [[END]]
   for (int d : varr) 0;
@@ -116,7 +116,7 @@ void test3(int b, int c) {
   //CHECK-NEXT: [[VLA_SIZEOF_DIM2:%.*]] = mul nuw i64 4, [[VLA_DIM2_PRE]]
   //CHECK-NEXT: [[VLA_NUM_ELEMENTS:%.*]] = udiv i64 [[VLA_SIZEOF]], [[VLA_SIZEOF_DIM2]]
   //CHECK-NEXT: [[VLA_END_INDEX:%.*]] = mul nsw i64 [[VLA_NUM_ELEMENTS]], [[VLA_DIM2_PRE]]
-  //CHECK-NEXT: [[VLA_END_PTR:%.*]] = getelementptr inbounds nuw i32, ptr {{%.*}}, i64 [[VLA_END_INDEX]]
+  //CHECK-NEXT: [[VLA_END_PTR:%.*]] = getelementptr inbounds i32, ptr {{%.*}}, i64 [[VLA_END_INDEX]]
   //X64-NEXT: store ptr [[VLA_END_PTR]], ptr %__end
   //AMDGCN-NEXT: store ptr [[VLA_END_PTR]], ptr [[END]]
  
diff --git a/clang/test/CodeGenHLSL/buffer-array-operator.hlsl b/clang/test/CodeGenHLSL/buffer-array-operator.hlsl
index 02e570ebdcb4fc..f5556df30871c4 100644
--- a/clang/test/CodeGenHLSL/buffer-array-operator.hlsl
+++ b/clang/test/CodeGenHLSL/buffer-array-operator.hlsl
@@ -17,7 +17,7 @@ void fn(int Idx) {
 // CHECK-NEXT: %h = getelementptr inbounds nuw %"class.hlsl::RWBuffer", ptr %this1, i32 0, i32 0
 // CHECK-NEXT: %0 = load ptr, ptr %h, align 4
 // CHECK-NEXT: %1 = load i32, ptr %Idx.addr, align 4
-// CHECK-NEXT: %arrayidx = getelementptr inbounds nuw float, ptr %0, i32 %1
+// CHECK-NEXT: %arrayidx = getelementptr inbounds float, ptr %0, i32 %1
 // CHECK-NEXT: ret ptr %arrayidx
 
 // Const comes next, and returns the pointer instead of the value.
@@ -26,5 +26,5 @@ void fn(int Idx) {
 // CHECK-NEXT: %h = getelementptr inbounds nuw %"class.hlsl::RWBuffer", ptr %this1, i32 0, i32 0
 // CHECK-NEXT: %0 = load ptr, ptr %h, align 4
 // CHECK-NEXT: %1 = load i32, ptr %Idx.addr, align 4
-// CHECK-NEXT: %arrayidx = getelementptr inbounds nuw float, ptr %0, i32 %1
+// CHECK-NEXT: %arrayidx = getelementptr inbounds float, ptr %0, i32 %1
 // CHECK-NEXT: ret ptr %arrayidx
diff --git a/clang/test/CodeGenSYCL/address-space-deduction.cpp b/clang/test/CodeGenSYCL/address-space-deduction.cpp
index 5910ec3bfc305a..96075a47343fe9 100644
--- a/clang/test/CodeGenSYCL/address-space-deduction.cpp
+++ b/clang/test/CodeGenSYCL/address-space-deduction.cpp
@@ -33,55 +33,55 @@
 // CHECK-NEXT:    store ptr addrspace(4) [[I_ASCAST]], ptr addrspace(4) [[PPTR_ASCAST]], align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[PPTR_ASCAST]], align 8
 // CHECK-NEXT:    [[CMP:%.*]] = icmp eq ptr addrspace(4) [[TMP0]], [[I_ASCAST]]
-// CHECK-NEXT:    [[STOREDV:%.*]] = zext i1 [[CMP]] to i8
-// CHECK-NEXT:    store i8 [[STOREDV]], ptr addrspace(4) [[IS_I_PTR_ASCAST]], align 1
+// CHECK-NEXT:    [[FROMBOOL:%.*]] = zext i1 [[CMP]] to i8
+// CHECK-NEXT:    store i8 [[FROMBOOL]], ptr addrspace(4) [[IS_I_PTR_ASCAST]], align 1
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[PPTR_ASCAST]], align 8
 // CHECK-NEXT:    store i32 66, ptr addrspace(4) [[TMP1]], align 4
 // CHECK-NEXT:    store i32 23, ptr addrspace(4) [[VAR23_ASCAST]], align 4
 // CHECK-NEXT:    store ptr addrspace(4) [[VAR23_ASCAST]], ptr addrspace(4) [[CP_ASCAST]], align 8
-// CHECK-NEXT:    [[TMP2:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[CP_ASCAST]], align 8
-// CHECK-NEXT:    store i8 41, ptr addrspace(4) [[TMP2]], align 1
+// CHECK-NEXT:    [[TMP3:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[CP_ASCAST]], align 8
+// CHECK-NEXT:    store i8 41, ptr addrspace(4) [[TMP3]], align 1
 // CHECK-NEXT:    [[ARRAYDECAY:%.*]] = getelementptr inbounds [42 x i32], ptr addrspace(4) [[ARR_ASCAST]], i64 0, i64 0
 // CHECK-NEXT:    store ptr addrspace(4) [[ARRAYDECAY]], ptr addrspace(4) [[CPP_ASCAST]], align 8
-// CHECK-NEXT:    [[TMP3:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[CPP_ASCAST]], align 8
-// CHECK-NEXT:    store i8 43, ptr addrspace(4) [[TMP3]], align 1
+// CHECK-NEXT:    [[TMP5:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[CPP_ASCAST]], align 8
+// CHECK-NEXT:    store i8 43, ptr addrspace(4) [[TMP5]], align 1
 // CHECK-NEXT:    [[ARRAYDECAY1:%.*]] = getelementptr inbounds [42 x i32], ptr addrspace(4) [[ARR_ASCAST]], i64 0, i64 0
 // CHECK-NEXT:    [[ADD_PTR:%.*]] = getelementptr inbounds i32, ptr addrspace(4) [[ARRAYDECAY1]], i64 10
 // CHECK-NEXT:    store ptr addrspace(4) [[ADD_PTR]], ptr addrspace(4) [[APTR_ASCAST]], align 8
-// CHECK-NEXT:    [[TMP4:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[APTR_ASCAST]], align 8
+// CHECK-NEXT:    [[TMP6:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[APTR_ASCAST]], align 8
 // CHECK-NEXT:    [[ARRAYDECAY2:%.*]] = getelementptr inbounds [42 x i32], ptr addrspace(4) [[ARR_ASCAST]], i64 0, i64 0
-// CHECK-NEXT:    [[ADD_PTR3:%.*]] = getelementptr inbounds nuw i32, ptr addrspace(4) [[ARRAYDECAY2]], i64 168
-// CHECK-NEXT:    [[CMP4:%.*]] = icmp ult ptr addrspace(4) [[TMP4]], [[ADD_PTR3]]
+// CHECK-NEXT:    [[ADD_PTR3:%.*]] = getelementptr inbounds i32, ptr addrspace(4) [[ARRAYDECAY2]], i64 168
+// CHECK-NEXT:    [[CMP4:%.*]] = icmp ult ptr addrspace(4) [[TMP6]], [[ADD_PTR3]]
 // CHECK-NEXT:    br i1 [[CMP4]], label [[IF_THEN:%.*]], label [[IF_END:%.*]]
 // CHECK:       if.then:
-// CHECK-NEXT:    [[TMP5:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[APTR_ASCAST]], align 8
-// CHECK-NEXT:    store i32 44, ptr addrspace(4) [[TMP5]], align 4
+// CHECK-NEXT:    [[TMP7:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[APTR_ASCAST]], align 8
+// CHECK-NEXT:    store i32 44, ptr addrspace(4) [[TMP7]], align 4
 // CHECK-NEXT:    br label [[IF_END]]
 // CHECK:       if.end:
 // CHECK-NEXT:    store ptr addrspace(4) addrspacecast (ptr addrspace(1) @.str to ptr addrspace(4)), ptr addrspace(4) [[STR_ASCAST]], align 8
-// CHECK-NEXT:    [[TMP6:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[STR_ASCAST]], align 8
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i8, ptr addrspace(4) [[TMP6]], i64 0
-// CHECK-NEXT:    [[TMP7:%.*]] = load i8, ptr addrspace(4) [[ARRAYIDX]], align 1
-// CHECK-NEXT:    [[CONV:%.*]] = sext i8 [[TMP7]] to i32
+// CHECK-NEXT:    [[TMP8:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[STR_ASCAST]], align 8
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i8, ptr addrspace(4) [[TMP8]], i64 0
+// CHECK-NEXT:    [[TMP9:%.*]] = load i8, ptr addrspace(4) [[ARRAYIDX]], align 1
+// CHECK-NEXT:    [[CONV:%.*]] = sext i8 [[TMP9]] to i32
 // CHECK-NEXT:    store i32 [[CONV]], ptr addrspace(4) [[I_ASCAST]], align 4
-// CHECK-NEXT:    [[TMP8:%.*]] = load i32, ptr addrspace(4) [[I_ASCAST]], align 4
-// CHECK-NEXT:    [[CMP5:%.*]] = icmp sgt i32 [[TMP8]], 2
+// CHECK-NEXT:    [[TMP10:%.*]] = load i32, ptr addrspace(4) [[I_ASCAST]], align 4
+// CHECK-NEXT:    [[CMP5:%.*]] = icmp sgt i32 [[TMP10]], 2
 // CHECK-NEXT:    br i1 [[CMP5]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]]
 // CHECK:       cond.true:
-// CHECK-NEXT:    [[TMP9:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[STR_ASCAST]], align 8
+// CHECK-NEXT:    [[TMP11:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[STR_ASCAST]], align 8
 // CHECK-NEXT:    br label [[COND_END:%.*]]
 // CHECK:       cond.false:
 // CHECK-NEXT:    br label [[COND_END]]
 // CHECK:       cond.end:
-// CHECK-NEXT:    [[COND:%.*]] = phi ptr addrspace(4) [ [[TMP9]], [[COND_TRUE]] ], [ addrspacecast (ptr addrspace(1) @.str.1 to ptr addrspace(4)), [[COND_FALSE]] ]
+// CHECK-NEXT:    [[COND:%.*]] = phi ptr addrspace(4) [ [[TMP11]], [[COND_TRUE]] ], [ addrspacecast (ptr addrspace(1) @.str.1 to ptr addrspace(4)), [[COND_FALSE]] ]
 // CHECK-NEXT:    store ptr addrspace(4) [[COND]], ptr addrspace(4) [[PHI_STR_ASCAST]], align 8
-// CHECK-NEXT:    [[TMP10:%.*]] = load i32, ptr addrspace(4) [[I_ASCAST]], align 4
-// CHECK-NEXT:    [[CMP6:%.*]] = icmp sgt i32 [[TMP10]], 2
-// CHECK-NEXT:    [[TMP11:%.*]] = zext i1 [[CMP6]] to i64
+// CHECK-NEXT:    [[TMP12:%.*]] = load i32, ptr addrspace(4) [[I_ASCAST]], align 4
+// CHECK-NEXT:    [[CMP6:%.*]] = icmp sgt i32 [[TMP12]], 2
+// CHECK-NEXT:    [[TMP13:%.*]] = zext i1 [[CMP6]] to i64
 // CHECK-NEXT:    [[COND7:%.*]] = select i1 [[CMP6]], ptr addrspace(4) addrspacecast (ptr addrspace(1) @.str.2 to ptr addrspace(4)), ptr addrspace(4) null
 // CHECK-NEXT:    store ptr addrspace(4) [[COND7]], ptr addrspace(4) [[SELECT_NULL_ASCAST]], align 8
-// CHECK-NEXT:    [[TMP12:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[STR_ASCAST]], align 8
-// CHECK-NEXT:    store ptr addrspace(4) [[TMP12]], ptr addrspace(4) [[SELECT_STR_TRIVIAL1_ASCAST]], align 8
+// CHECK-NEXT:    [[TMP14:%.*]] = load ptr addrspace(4), ptr addrspace(4) [[STR_ASCAST]], align 8
+// CHECK-NEXT:    store ptr addrspace(4) [[TMP14]], ptr addrspace(4) [[SELECT_STR_TRIVIAL1_ASCAST]], align 8
 // CHECK-NEXT:    store ptr addrspace(4) addrspacecast (ptr addrspace(1) @.str.1 to ptr addrspace(4)), ptr addrspace(4) [[SELECT_STR_TRIVIAL2_ASCAST]], align 8
 // CHECK-NEXT:    ret void
 //
diff --git a/clang/test/Headers/__clang_hip_math.hip b/clang/test/Headers/__clang_hip_math.hip
index bc0c3796ca0f60..6ee10976f12079 100644
--- a/clang/test/Headers/__clang_hip_math.hip
+++ b/clang/test/Headers/__clang_hip_math.hip
@@ -47,7 +47,7 @@ typedef unsigned long long uint64_t;
 // CHECK-NEXT:    [[CONV5_I:%.*]] = zext nneg i8 [[TMP0]] to i64
 // CHECK-NEXT:    [[ADD_I:%.*]] = add i64 [[MUL_I]], -48
 // CHECK-NEXT:    [[SUB_I:%.*]] = add i64 [[ADD_I]], [[CONV5_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I]]
 // CHECK:       cleanup.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I]] = phi ptr [ [[INCDEC_PTR_I]], [[IF_THEN_I]] ], [ [[__TAGP_ADDR_0_I]], [[WHILE_BODY_I]] ]
@@ -79,7 +79,7 @@ extern "C" __device__ uint64_t test___make_mantissa_base8(const char *p) {
 // CHECK-NEXT:    [[CONV5_I:%.*]] = zext nneg i8 [[TMP0]] to i64
 // CHECK-NEXT:    [[ADD_I:%.*]] = add i64 [[MUL_I]], -48
 // CHECK-NEXT:    [[SUB_I:%.*]] = add i64 [[ADD_I]], [[CONV5_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I]]
 // CHECK:       cleanup.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I]] = phi ptr [ [[INCDEC_PTR_I]], [[IF_THEN_I]] ], [ [[__TAGP_ADDR_0_I]], [[WHILE_BODY_I]] ]
@@ -120,7 +120,7 @@ extern "C" __device__ uint64_t test___make_mantissa_base10(const char *p) {
 // CHECK-NEXT:    [[CONV25_I:%.*]] = zext nneg i8 [[TMP0]] to i64
 // CHECK-NEXT:    [[ADD26_I:%.*]] = add i64 [[MUL24_I]], [[DOTSINK]]
 // CHECK-NEXT:    [[ADD28_I:%.*]] = add i64 [[ADD26_I]], [[CONV25_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I]]
 // CHECK:       cleanup.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I]] = phi ptr [ [[INCDEC_PTR_I]], [[IF_END31_I]] ], [ [[__TAGP_ADDR_0_I]], [[IF_ELSE17_I]] ]
@@ -141,7 +141,7 @@ extern "C" __device__ uint64_t test___make_mantissa_base16(const char *p) {
 // CHECK-NEXT:    [[CMP_I:%.*]] = icmp eq i8 [[TMP0]], 48
 // CHECK-NEXT:    br i1 [[CMP_I]], label [[IF_THEN_I:%.*]], label [[WHILE_COND_I14_I:%.*]]
 // CHECK:       if.then.i:
-// CHECK-NEXT:    [[INCDEC_PTR_I:%.*]] = getelementptr inbounds nuw i8, ptr [[P]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I:%.*]] = getelementptr inbounds i8, ptr [[P]], i64 1
 // CHECK-NEXT:    [[TMP1:%.*]] = load i8, ptr [[INCDEC_PTR_I]], align 1, !tbaa [[TBAA4]]
 // CHECK-NEXT:    switch i8 [[TMP1]], label [[WHILE_COND_I_I:%.*]] [
 // CHECK-NEXT:      i8 120, label [[WHILE_COND_I30_I_PREHEADER:%.*]]
@@ -173,7 +173,7 @@ extern "C" __device__ uint64_t test___make_mantissa_base16(const char *p) {
 // CHECK-NEXT:    [[CONV25_I_I:%.*]] = zext nneg i8 [[TMP2]] to i64
 // CHECK-NEXT:    [[ADD26_I_I:%.*]] = add i64 [[MUL24_I_I]], [[DOTSINK]]
 // CHECK-NEXT:    [[ADD28_I_I:%.*]] = add i64 [[ADD26_I_I]], [[CONV25_I_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I40_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I31_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I40_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I31_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I36_I]]
 // CHECK:       cleanup.i36.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I37_I]] = phi ptr [ [[INCDEC_PTR_I40_I]], [[IF_END31_I_I]] ], [ [[__TAGP_ADDR_0_I31_I]], [[IF_ELSE17_I_I]] ]
@@ -195,7 +195,7 @@ extern "C" __device__ uint64_t test___make_mantissa_base16(const char *p) {
 // CHECK-NEXT:    [[CONV5_I_I:%.*]] = zext nneg i8 [[TMP6]] to i64
 // CHECK-NEXT:    [[ADD_I_I:%.*]] = add i64 [[MUL_I_I]], -48
 // CHECK-NEXT:    [[SUB_I_I:%.*]] = add i64 [[ADD_I_I]], [[CONV5_I_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I_I]]
 // CHECK:       cleanup.i.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I_I]] = phi ptr [ [[INCDEC_PTR_I_I]], [[IF_THEN_I_I]] ], [ [[__TAGP_ADDR_0_I_I]], [[WHILE_BODY_I_I]] ]
@@ -216,7 +216,7 @@ extern "C" __device__ uint64_t test___make_mantissa_base16(const char *p) {
 // CHECK-NEXT:    [[CONV5_I26_I:%.*]] = zext nneg i8 [[TMP8]] to i64
 // CHECK-NEXT:    [[ADD_I27_I:%.*]] = add i64 [[MUL_I25_I]], -48
 // CHECK-NEXT:    [[SUB_I28_I:%.*]] = add i64 [[ADD_I27_I]], [[CONV5_I26_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I29_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I15_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I29_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I15_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I20_I]]
 // CHECK:       cleanup.i20.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I21_I]] = phi ptr [ [[INCDEC_PTR_I29_I]], [[IF_THEN_I24_I]] ], [ [[__TAGP_ADDR_0_I15_I]], [[WHILE_BODY_I18_I]] ]
@@ -2367,7 +2367,7 @@ extern "C" __device__ double test_modf(double x, double* y) {
 // CHECK-NEXT:    [[CMP_I_I:%.*]] = icmp eq i8 [[TMP0]], 48
 // CHECK-NEXT:    br i1 [[CMP_I_I]], label [[IF_THEN_I_I:%.*]], label [[WHILE_COND_I14_I_I:%.*]]
 // CHECK:       if.then.i.i:
-// CHECK-NEXT:    [[INCDEC_PTR_I_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TAG]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I_I:%.*]] = getelementptr inbounds i8, ptr [[TAG]], i64 1
 // CHECK-NEXT:    [[TMP1:%.*]] = load i8, ptr [[INCDEC_PTR_I_I]], align 1, !tbaa [[TBAA4]]
 // CHECK-NEXT:    switch i8 [[TMP1]], label [[WHILE_COND_I_I_I:%.*]] [
 // CHECK-NEXT:      i8 120, label [[WHILE_COND_I30_I_I_PREHEADER:%.*]]
@@ -2399,7 +2399,7 @@ extern "C" __device__ double test_modf(double x, double* y) {
 // CHECK-NEXT:    [[CONV25_I_I_I:%.*]] = zext nneg i8 [[TMP2]] to i64
 // CHECK-NEXT:    [[ADD26_I_I_I:%.*]] = add i64 [[MUL24_I_I_I]], [[DOTSINK]]
 // CHECK-NEXT:    [[ADD28_I_I_I:%.*]] = add i64 [[ADD26_I_I_I]], [[CONV25_I_I_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I40_I_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I31_I_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I40_I_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I31_I_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I36_I_I]]
 // CHECK:       cleanup.i36.i.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I37_I_I]] = phi ptr [ [[INCDEC_PTR_I40_I_I]], [[IF_END31_I_I_I]] ], [ [[__TAGP_ADDR_0_I31_I_I]], [[IF_ELSE17_I_I_I]] ]
@@ -2421,7 +2421,7 @@ extern "C" __device__ double test_modf(double x, double* y) {
 // CHECK-NEXT:    [[CONV5_I_I_I:%.*]] = zext nneg i8 [[TMP6]] to i64
 // CHECK-NEXT:    [[ADD_I_I_I:%.*]] = add i64 [[MUL_I_I_I]], -48
 // CHECK-NEXT:    [[SUB_I_I_I:%.*]] = add i64 [[ADD_I_I_I]], [[CONV5_I_I_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I_I_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I_I_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I_I_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I_I_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I_I_I]]
 // CHECK:       cleanup.i.i.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I_I_I]] = phi ptr [ [[INCDEC_PTR_I_I_I]], [[IF_THEN_I_I_I]] ], [ [[__TAGP_ADDR_0_I_I_I]], [[WHILE_BODY_I_I_I]] ]
@@ -2442,7 +2442,7 @@ extern "C" __device__ double test_modf(double x, double* y) {
 // CHECK-NEXT:    [[CONV5_I26_I_I:%.*]] = zext nneg i8 [[TMP8]] to i64
 // CHECK-NEXT:    [[ADD_I27_I_I:%.*]] = add i64 [[MUL_I25_I_I]], -48
 // CHECK-NEXT:    [[SUB_I28_I_I:%.*]] = add i64 [[ADD_I27_I_I]], [[CONV5_I26_I_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I29_I_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I15_I_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I29_I_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I15_I_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I20_I_I]]
 // CHECK:       cleanup.i20.i.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I21_I_I]] = phi ptr [ [[INCDEC_PTR_I29_I_I]], [[IF_THEN_I24_I_I]] ], [ [[__TAGP_ADDR_0_I15_I_I]], [[WHILE_BODY_I18_I_I]] ]
@@ -2466,7 +2466,7 @@ extern "C" __device__ float test_nanf(const char *tag) {
 // CHECK-NEXT:    [[CMP_I_I:%.*]] = icmp eq i8 [[TMP0]], 48
 // CHECK-NEXT:    br i1 [[CMP_I_I]], label [[IF_THEN_I_I:%.*]], label [[WHILE_COND_I14_I_I:%.*]]
 // CHECK:       if.then.i.i:
-// CHECK-NEXT:    [[INCDEC_PTR_I_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TAG]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I_I:%.*]] = getelementptr inbounds i8, ptr [[TAG]], i64 1
 // CHECK-NEXT:    [[TMP1:%.*]] = load i8, ptr [[INCDEC_PTR_I_I]], align 1, !tbaa [[TBAA4]]
 // CHECK-NEXT:    switch i8 [[TMP1]], label [[WHILE_COND_I_I_I:%.*]] [
 // CHECK-NEXT:      i8 120, label [[WHILE_COND_I30_I_I_PREHEADER:%.*]]
@@ -2498,7 +2498,7 @@ extern "C" __device__ float test_nanf(const char *tag) {
 // CHECK-NEXT:    [[CONV25_I_I_I:%.*]] = zext nneg i8 [[TMP2]] to i64
 // CHECK-NEXT:    [[ADD26_I_I_I:%.*]] = add i64 [[MUL24_I_I_I]], [[DOTSINK]]
 // CHECK-NEXT:    [[ADD28_I_I_I:%.*]] = add i64 [[ADD26_I_I_I]], [[CONV25_I_I_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I40_I_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I31_I_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I40_I_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I31_I_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I36_I_I]]
 // CHECK:       cleanup.i36.i.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I37_I_I]] = phi ptr [ [[INCDEC_PTR_I40_I_I]], [[IF_END31_I_I_I]] ], [ [[__TAGP_ADDR_0_I31_I_I]], [[IF_ELSE17_I_I_I]] ]
@@ -2520,7 +2520,7 @@ extern "C" __device__ float test_nanf(const char *tag) {
 // CHECK-NEXT:    [[CONV5_I_I_I:%.*]] = zext nneg i8 [[TMP6]] to i64
 // CHECK-NEXT:    [[ADD_I_I_I:%.*]] = add i64 [[MUL_I_I_I]], -48
 // CHECK-NEXT:    [[SUB_I_I_I:%.*]] = add i64 [[ADD_I_I_I]], [[CONV5_I_I_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I_I_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I_I_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I_I_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I_I_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I_I_I]]
 // CHECK:       cleanup.i.i.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I_I_I]] = phi ptr [ [[INCDEC_PTR_I_I_I]], [[IF_THEN_I_I_I]] ], [ [[__TAGP_ADDR_0_I_I_I]], [[WHILE_BODY_I_I_I]] ]
@@ -2541,7 +2541,7 @@ extern "C" __device__ float test_nanf(const char *tag) {
 // CHECK-NEXT:    [[CONV5_I26_I_I:%.*]] = zext nneg i8 [[TMP8]] to i64
 // CHECK-NEXT:    [[ADD_I27_I_I:%.*]] = add i64 [[MUL_I25_I_I]], -48
 // CHECK-NEXT:    [[SUB_I28_I_I:%.*]] = add i64 [[ADD_I27_I_I]], [[CONV5_I26_I_I]]
-// CHECK-NEXT:    [[INCDEC_PTR_I29_I_I:%.*]] = getelementptr inbounds nuw i8, ptr [[__TAGP_ADDR_0_I15_I_I]], i64 1
+// CHECK-NEXT:    [[INCDEC_PTR_I29_I_I:%.*]] = getelementptr inbounds i8, ptr [[__TAGP_ADDR_0_I15_I_I]], i64 1
 // CHECK-NEXT:    br label [[CLEANUP_I20_I_I]]
 // CHECK:       cleanup.i20.i.i:
 // CHECK-NEXT:    [[__TAGP_ADDR_1_I21_I_I]] = phi ptr [ [[INCDEC_PTR_I29_I_I]], [[IF_THEN_I24_I_I]] ], [ [[__TAGP_ADDR_0_I15_I_I]], [[WHILE_BODY_I18_I_I]] ]
@@ -2862,7 +2862,7 @@ extern "C" __device__ double test_normcdfinv(double x) {
 // DEFAULT-NEXT:    [[TMP0:%.*]] = load float, ptr [[__A_ADDR_0_I3]], align 4, !tbaa [[TBAA16]]
 // DEFAULT-NEXT:    [[MUL_I:%.*]] = fmul contract float [[TMP0]], [[TMP0]]
 // DEFAULT-NEXT:    [[ADD_I]] = fadd contract float [[__R_0_I4]], [[MUL_I]]
-// DEFAULT-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 4
+// DEFAULT-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 4
 // DEFAULT-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // DEFAULT-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL5NORMFIPKF_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP20:![0-9]+]]
 // DEFAULT:       _ZL5normfiPKf.exit:
@@ -2882,7 +2882,7 @@ extern "C" __device__ double test_normcdfinv(double x) {
 // FINITEONLY-NEXT:    [[TMP0:%.*]] = load float, ptr [[__A_ADDR_0_I3]], align 4, !tbaa [[TBAA16]]
 // FINITEONLY-NEXT:    [[MUL_I:%.*]] = fmul nnan ninf contract float [[TMP0]], [[TMP0]]
 // FINITEONLY-NEXT:    [[ADD_I]] = fadd nnan ninf contract float [[__R_0_I4]], [[MUL_I]]
-// FINITEONLY-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 4
+// FINITEONLY-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 4
 // FINITEONLY-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // FINITEONLY-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL5NORMFIPKF_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP20:![0-9]+]]
 // FINITEONLY:       _ZL5normfiPKf.exit:
@@ -2902,7 +2902,7 @@ extern "C" __device__ double test_normcdfinv(double x) {
 // APPROX-NEXT:    [[TMP0:%.*]] = load float, ptr [[__A_ADDR_0_I3]], align 4, !tbaa [[TBAA16]]
 // APPROX-NEXT:    [[MUL_I:%.*]] = fmul contract float [[TMP0]], [[TMP0]]
 // APPROX-NEXT:    [[ADD_I]] = fadd contract float [[__R_0_I4]], [[MUL_I]]
-// APPROX-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 4
+// APPROX-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 4
 // APPROX-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // APPROX-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL5NORMFIPKF_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP20:![0-9]+]]
 // APPROX:       _ZL5normfiPKf.exit:
@@ -2926,7 +2926,7 @@ extern "C" __device__ float test_normf(int x, const float *y) {
 // DEFAULT-NEXT:    [[TMP0:%.*]] = load double, ptr [[__A_ADDR_0_I3]], align 8, !tbaa [[TBAA18]]
 // DEFAULT-NEXT:    [[MUL_I:%.*]] = fmul contract double [[TMP0]], [[TMP0]]
 // DEFAULT-NEXT:    [[ADD_I]] = fadd contract double [[__R_0_I4]], [[MUL_I]]
-// DEFAULT-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 8
+// DEFAULT-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 8
 // DEFAULT-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // DEFAULT-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL4NORMIPKD_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP21:![0-9]+]]
 // DEFAULT:       _ZL4normiPKd.exit:
@@ -2946,7 +2946,7 @@ extern "C" __device__ float test_normf(int x, const float *y) {
 // FINITEONLY-NEXT:    [[TMP0:%.*]] = load double, ptr [[__A_ADDR_0_I3]], align 8, !tbaa [[TBAA18]]
 // FINITEONLY-NEXT:    [[MUL_I:%.*]] = fmul nnan ninf contract double [[TMP0]], [[TMP0]]
 // FINITEONLY-NEXT:    [[ADD_I]] = fadd nnan ninf contract double [[__R_0_I4]], [[MUL_I]]
-// FINITEONLY-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 8
+// FINITEONLY-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 8
 // FINITEONLY-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // FINITEONLY-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL4NORMIPKD_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP21:![0-9]+]]
 // FINITEONLY:       _ZL4normiPKd.exit:
@@ -2966,7 +2966,7 @@ extern "C" __device__ float test_normf(int x, const float *y) {
 // APPROX-NEXT:    [[TMP0:%.*]] = load double, ptr [[__A_ADDR_0_I3]], align 8, !tbaa [[TBAA18]]
 // APPROX-NEXT:    [[MUL_I:%.*]] = fmul contract double [[TMP0]], [[TMP0]]
 // APPROX-NEXT:    [[ADD_I]] = fadd contract double [[__R_0_I4]], [[MUL_I]]
-// APPROX-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 8
+// APPROX-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 8
 // APPROX-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // APPROX-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL4NORMIPKD_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP21:![0-9]+]]
 // APPROX:       _ZL4normiPKd.exit:
@@ -3286,7 +3286,7 @@ extern "C" __device__ double test_rint(double x) {
 // DEFAULT-NEXT:    [[TMP0:%.*]] = load float, ptr [[__A_ADDR_0_I3]], align 4, !tbaa [[TBAA16]]
 // DEFAULT-NEXT:    [[MUL_I:%.*]] = fmul contract float [[TMP0]], [[TMP0]]
 // DEFAULT-NEXT:    [[ADD_I]] = fadd contract float [[__R_0_I4]], [[MUL_I]]
-// DEFAULT-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 4
+// DEFAULT-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 4
 // DEFAULT-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // DEFAULT-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL6RNORMFIPKF_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP22:![0-9]+]]
 // DEFAULT:       _ZL6rnormfiPKf.exit:
@@ -3306,7 +3306,7 @@ extern "C" __device__ double test_rint(double x) {
 // FINITEONLY-NEXT:    [[TMP0:%.*]] = load float, ptr [[__A_ADDR_0_I3]], align 4, !tbaa [[TBAA16]]
 // FINITEONLY-NEXT:    [[MUL_I:%.*]] = fmul nnan ninf contract float [[TMP0]], [[TMP0]]
 // FINITEONLY-NEXT:    [[ADD_I]] = fadd nnan ninf contract float [[__R_0_I4]], [[MUL_I]]
-// FINITEONLY-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 4
+// FINITEONLY-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 4
 // FINITEONLY-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // FINITEONLY-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL6RNORMFIPKF_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP22:![0-9]+]]
 // FINITEONLY:       _ZL6rnormfiPKf.exit:
@@ -3326,7 +3326,7 @@ extern "C" __device__ double test_rint(double x) {
 // APPROX-NEXT:    [[TMP0:%.*]] = load float, ptr [[__A_ADDR_0_I3]], align 4, !tbaa [[TBAA16]]
 // APPROX-NEXT:    [[MUL_I:%.*]] = fmul contract float [[TMP0]], [[TMP0]]
 // APPROX-NEXT:    [[ADD_I]] = fadd contract float [[__R_0_I4]], [[MUL_I]]
-// APPROX-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 4
+// APPROX-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 4
 // APPROX-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // APPROX-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL6RNORMFIPKF_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP22:![0-9]+]]
 // APPROX:       _ZL6rnormfiPKf.exit:
@@ -3350,7 +3350,7 @@ extern "C" __device__ float test_rnormf(int x, const float* y) {
 // DEFAULT-NEXT:    [[TMP0:%.*]] = load double, ptr [[__A_ADDR_0_I3]], align 8, !tbaa [[TBAA18]]
 // DEFAULT-NEXT:    [[MUL_I:%.*]] = fmul contract double [[TMP0]], [[TMP0]]
 // DEFAULT-NEXT:    [[ADD_I]] = fadd contract double [[__R_0_I4]], [[MUL_I]]
-// DEFAULT-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 8
+// DEFAULT-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 8
 // DEFAULT-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // DEFAULT-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL5RNORMIPKD_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP23:![0-9]+]]
 // DEFAULT:       _ZL5rnormiPKd.exit:
@@ -3370,7 +3370,7 @@ extern "C" __device__ float test_rnormf(int x, const float* y) {
 // FINITEONLY-NEXT:    [[TMP0:%.*]] = load double, ptr [[__A_ADDR_0_I3]], align 8, !tbaa [[TBAA18]]
 // FINITEONLY-NEXT:    [[MUL_I:%.*]] = fmul nnan ninf contract double [[TMP0]], [[TMP0]]
 // FINITEONLY-NEXT:    [[ADD_I]] = fadd nnan ninf contract double [[__R_0_I4]], [[MUL_I]]
-// FINITEONLY-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 8
+// FINITEONLY-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 8
 // FINITEONLY-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // FINITEONLY-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL5RNORMIPKD_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP23:![0-9]+]]
 // FINITEONLY:       _ZL5rnormiPKd.exit:
@@ -3390,7 +3390,7 @@ extern "C" __device__ float test_rnormf(int x, const float* y) {
 // APPROX-NEXT:    [[TMP0:%.*]] = load double, ptr [[__A_ADDR_0_I3]], align 8, !tbaa [[TBAA18]]
 // APPROX-NEXT:    [[MUL_I:%.*]] = fmul contract double [[TMP0]], [[TMP0]]
 // APPROX-NEXT:    [[ADD_I]] = fadd contract double [[__R_0_I4]], [[MUL_I]]
-// APPROX-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds nuw i8, ptr [[__A_ADDR_0_I3]], i64 8
+// APPROX-NEXT:    [[INCDEC_PTR_I]] = getelementptr inbounds i8, ptr [[__A_ADDR_0_I3]], i64 8
 // APPROX-NEXT:    [[TOBOOL_NOT_I:%.*]] = icmp eq i32 [[DEC_I]], 0
 // APPROX-NEXT:    br i1 [[TOBOOL_NOT_I]], label [[_ZL5RNORMIPKD_EXIT]], label [[WHILE_BODY_I]], !llvm.loop [[LOOP23:![0-9]+]]
 // APPROX:       _ZL5rnormiPKd.exit:
diff --git a/clang/test/OpenMP/bug60602.cpp b/clang/test/OpenMP/bug60602.cpp
index 0789ef958e523d..cb2e4e5b11e330 100644
--- a/clang/test/OpenMP/bug60602.cpp
+++ b/clang/test/OpenMP/bug60602.cpp
@@ -58,13 +58,13 @@ int kernel_within_loop(int *a, int *b, int N, int num_iters) {
 // CHECK-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[A_ADDR]], align 8
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP7]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP7]], i64 0
 // CHECK-NEXT:    [[TMP8:%.*]] = load i32, ptr [[N_ADDR]], align 4
 // CHECK-NEXT:    [[CONV:%.*]] = sext i32 [[TMP8]] to i64
 // CHECK-NEXT:    [[TMP9:%.*]] = mul nuw i64 [[CONV]], 4
 // CHECK-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[B_ADDR]], align 8
-// CHECK-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP11]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i32, ptr [[TMP11]], i64 0
 // CHECK-NEXT:    [[TMP12:%.*]] = load i32, ptr [[N_ADDR]], align 4
 // CHECK-NEXT:    [[CONV2:%.*]] = sext i32 [[TMP12]] to i64
 // CHECK-NEXT:    [[TMP13:%.*]] = mul nuw i64 [[CONV2]], 4
@@ -134,13 +134,13 @@ int kernel_within_loop(int *a, int *b, int N, int num_iters) {
 // CHECK-NEXT:    [[TMP46:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP47:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP48:%.*]] = load ptr, ptr [[A_ADDR]], align 8
-// CHECK-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP48]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i32, ptr [[TMP48]], i64 0
 // CHECK-NEXT:    [[TMP49:%.*]] = load i32, ptr [[N_ADDR]], align 4
 // CHECK-NEXT:    [[CONV5:%.*]] = sext i32 [[TMP49]] to i64
 // CHECK-NEXT:    [[TMP50:%.*]] = mul nuw i64 [[CONV5]], 4
 // CHECK-NEXT:    [[TMP51:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP52:%.*]] = load ptr, ptr [[B_ADDR]], align 8
-// CHECK-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP52]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds i32, ptr [[TMP52]], i64 0
 // CHECK-NEXT:    [[TMP53:%.*]] = load i32, ptr [[N_ADDR]], align 4
 // CHECK-NEXT:    [[CONV7:%.*]] = sext i32 [[TMP53]] to i64
 // CHECK-NEXT:    [[TMP54:%.*]] = mul nuw i64 [[CONV7]], 4
diff --git a/clang/test/OpenMP/declare_mapper_codegen.cpp b/clang/test/OpenMP/declare_mapper_codegen.cpp
index d2954b7a748217..52d5ceffa14715 100644
--- a/clang/test/OpenMP/declare_mapper_codegen.cpp
+++ b/clang/test/OpenMP/declare_mapper_codegen.cpp
@@ -129,7 +129,7 @@ class C {
 // CK0-DAG: [[BBEGIN:%.+]] = getelementptr inbounds nuw %class.C, ptr [[PTR]], i32 0, i32 1
 // CK0-DAG: [[BBEGIN2:%.+]] = getelementptr inbounds nuw %class.C, ptr [[PTR]], i32 0, i32 1
 // CK0-DAG: [[BARRBEGIN:%.+]] = load ptr, ptr [[BBEGIN2]]
-// CK0-DAG: [[BARRBEGINGEP:%.+]] = getelementptr inbounds nuw double, ptr [[BARRBEGIN]], i[[sz:64|32]] 0
+// CK0-DAG: [[BARRBEGINGEP:%.+]] = getelementptr inbounds double, ptr [[BARRBEGIN]], i[[sz:64|32]] 0
 // CK0-DAG: [[BEND:%.+]] = getelementptr ptr, ptr [[BBEGIN]], i32 1
 // CK0-DAG: [[ABEGINI:%.+]] = ptrtoint ptr [[ABEGIN]] to i64
 // CK0-DAG: [[BENDI:%.+]] = ptrtoint ptr [[BEND]] to i64
@@ -965,7 +965,7 @@ class C {
 // CK4-DAG: [[BBEGIN:%.+]] = getelementptr inbounds nuw %class.C, ptr [[PTR]], i32 0, i32 1
 // CK4-DAG: [[BBEGIN2:%.+]] = getelementptr inbounds nuw %class.C, ptr [[PTR]], i32 0, i32 1
 // CK4-DAG: [[BARRBEGIN:%.+]] = load ptr, ptr [[BBEGIN2]]
-// CK4-DAG: [[BARRBEGINGEP:%.+]] = getelementptr inbounds nuw double, ptr [[BARRBEGIN]], i[[sz:64|32]] 0
+// CK4-DAG: [[BARRBEGINGEP:%.+]] = getelementptr inbounds double, ptr [[BARRBEGIN]], i[[sz:64|32]] 0
 // CK4-DAG: [[BEND:%.+]] = getelementptr ptr, ptr [[BBEGIN]], i32 1
 // CK4-DAG: [[ABEGINI:%.+]] = ptrtoint ptr [[ABEGIN]] to i64
 // CK4-DAG: [[BENDI:%.+]] = ptrtoint ptr [[BEND]] to i64
diff --git a/clang/test/OpenMP/distribute_codegen.cpp b/clang/test/OpenMP/distribute_codegen.cpp
index 6c588ba25db301..ea619cb6e0f261 100644
--- a/clang/test/OpenMP/distribute_codegen.cpp
+++ b/clang/test/OpenMP/distribute_codegen.cpp
@@ -662,24 +662,24 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK1-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP15]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM]]
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM]]
 // CHECK1-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[IDXPROM3:%.*]] = zext i32 [[TMP18]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM3]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM3]]
 // CHECK1-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[MUL5:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK1-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[IDXPROM6:%.*]] = zext i32 [[TMP21]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM6]]
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM6]]
 // CHECK1-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[MUL8:%.*]] = fmul float [[MUL5]], [[TMP22]]
 // CHECK1-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    [[IDXPROM9:%.*]] = zext i32 [[TMP24]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM9]]
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM9]]
 // CHECK1-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK1-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK1:       omp.body.continue:
@@ -1574,21 +1574,21 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK3-NEXT:    store i32 [[ADD]], ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK3-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK3-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
-// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i32 [[TMP15]]
+// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i32 [[TMP15]]
 // CHECK3-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK3-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK3-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
-// CHECK3-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i32 [[TMP18]]
+// CHECK3-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i32 [[TMP18]]
 // CHECK3-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX3]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK3-NEXT:    [[MUL4:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK3-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK3-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
-// CHECK3-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i32 [[TMP21]]
+// CHECK3-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i32 [[TMP21]]
 // CHECK3-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX5]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK3-NEXT:    [[MUL6:%.*]] = fmul float [[MUL4]], [[TMP22]]
 // CHECK3-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK3-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
-// CHECK3-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i32 [[TMP24]]
+// CHECK3-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i32 [[TMP24]]
 // CHECK3-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK3-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK3:       omp.body.continue:
@@ -2252,24 +2252,24 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK17-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP15]] to i64
-// CHECK17-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM]]
+// CHECK17-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM]]
 // CHECK17-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[IDXPROM3:%.*]] = zext i32 [[TMP18]] to i64
-// CHECK17-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM3]]
+// CHECK17-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM3]]
 // CHECK17-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[MUL5:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK17-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[IDXPROM6:%.*]] = zext i32 [[TMP21]] to i64
-// CHECK17-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM6]]
+// CHECK17-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM6]]
 // CHECK17-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[MUL8:%.*]] = fmul float [[MUL5]], [[TMP22]]
 // CHECK17-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    [[IDXPROM9:%.*]] = zext i32 [[TMP24]] to i64
-// CHECK17-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM9]]
+// CHECK17-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM9]]
 // CHECK17-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK17-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK17:       omp.body.continue:
@@ -2790,21 +2790,21 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK19-NEXT:    store i32 [[ADD]], ptr [[I]], align 4, !llvm.access.group [[ACC_GRP12]]
 // CHECK19-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 4, !llvm.access.group [[ACC_GRP12]]
 // CHECK19-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP12]]
-// CHECK19-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i32 [[TMP15]]
+// CHECK19-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i32 [[TMP15]]
 // CHECK19-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP12]]
 // CHECK19-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 4, !llvm.access.group [[ACC_GRP12]]
 // CHECK19-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP12]]
-// CHECK19-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i32 [[TMP18]]
+// CHECK19-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i32 [[TMP18]]
 // CHECK19-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX3]], align 4, !llvm.access.group [[ACC_GRP12]]
 // CHECK19-NEXT:    [[MUL4:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK19-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 4, !llvm.access.group [[ACC_GRP12]]
 // CHECK19-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP12]]
-// CHECK19-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i32 [[TMP21]]
+// CHECK19-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i32 [[TMP21]]
 // CHECK19-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX5]], align 4, !llvm.access.group [[ACC_GRP12]]
 // CHECK19-NEXT:    [[MUL6:%.*]] = fmul float [[MUL4]], [[TMP22]]
 // CHECK19-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 4, !llvm.access.group [[ACC_GRP12]]
 // CHECK19-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP12]]
-// CHECK19-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i32 [[TMP24]]
+// CHECK19-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i32 [[TMP24]]
 // CHECK19-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP12]]
 // CHECK19-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK19:       omp.body.continue:
diff --git a/clang/test/OpenMP/distribute_parallel_for_reduction_task_codegen.cpp b/clang/test/OpenMP/distribute_parallel_for_reduction_task_codegen.cpp
index 93a6779ac02e83..b019b4ff92ad58 100644
--- a/clang/test/OpenMP/distribute_parallel_for_reduction_task_codegen.cpp
+++ b/clang/test/OpenMP/distribute_parallel_for_reduction_task_codegen.cpp
@@ -175,16 +175,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP3]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP3]], i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP4]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP4]], i64 0
 // CHECK1-NEXT:    [[TMP5:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP6:%.*]] = sext i32 [[TMP5]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP6]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP7]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP7]], i64 9
 // CHECK1-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP8]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP8]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP9:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP10:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP11:%.*]] = sub i64 [[TMP9]], [[TMP10]]
@@ -214,7 +214,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP22]]
 // CHECK1-NEXT:    store ptr [[_TMP6]], ptr [[_TMP5]], align 8
 // CHECK1-NEXT:    store ptr [[TMP23]], ptr [[_TMP6]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP24:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP24]], align 8
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -229,19 +229,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP30:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP30]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP31:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP32]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds ptr, ptr [[TMP32]], i64 0
 // CHECK1-NEXT:    [[TMP33:%.*]] = load ptr, ptr [[ARRAYIDX8]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP33]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds i8, ptr [[TMP33]], i64 0
 // CHECK1-NEXT:    [[TMP34:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP35:%.*]] = sext i32 [[TMP34]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN10:%.*]] = add nsw i64 -1, [[TMP35]]
 // CHECK1-NEXT:    [[TMP36:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP36]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds ptr, ptr [[TMP36]], i64 9
 // CHECK1-NEXT:    [[TMP37:%.*]] = load ptr, ptr [[ARRAYIDX11]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP37]], i64 [[LB_ADD_LEN10]]
+// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds i8, ptr [[TMP37]], i64 [[LB_ADD_LEN10]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP31]], align 8
 // CHECK1-NEXT:    [[TMP38:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX9]], ptr [[TMP38]], align 8
@@ -562,9 +562,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]
diff --git a/clang/test/OpenMP/distribute_simd_codegen.cpp b/clang/test/OpenMP/distribute_simd_codegen.cpp
index ad93fd6030ac7d..f7353172e235ca 100644
--- a/clang/test/OpenMP/distribute_simd_codegen.cpp
+++ b/clang/test/OpenMP/distribute_simd_codegen.cpp
@@ -706,24 +706,24 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK1-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP15]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM]]
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM]]
 // CHECK1-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[IDXPROM3:%.*]] = zext i32 [[TMP18]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM3]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM3]]
 // CHECK1-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[MUL5:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK1-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[IDXPROM6:%.*]] = zext i32 [[TMP21]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM6]]
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM6]]
 // CHECK1-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[MUL8:%.*]] = fmul float [[MUL5]], [[TMP22]]
 // CHECK1-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    [[IDXPROM9:%.*]] = zext i32 [[TMP24]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM9]]
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM9]]
 // CHECK1-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4, !llvm.access.group [[ACC_GRP17]]
 // CHECK1-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK1:       omp.body.continue:
@@ -1682,21 +1682,21 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK3-NEXT:    store i32 [[ADD]], ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK3-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK3-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
-// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i32 [[TMP15]]
+// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i32 [[TMP15]]
 // CHECK3-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK3-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK3-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
-// CHECK3-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i32 [[TMP18]]
+// CHECK3-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i32 [[TMP18]]
 // CHECK3-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX3]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK3-NEXT:    [[MUL4:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK3-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK3-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
-// CHECK3-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i32 [[TMP21]]
+// CHECK3-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i32 [[TMP21]]
 // CHECK3-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX5]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK3-NEXT:    [[MUL6:%.*]] = fmul float [[MUL4]], [[TMP22]]
 // CHECK3-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK3-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
-// CHECK3-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i32 [[TMP24]]
+// CHECK3-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i32 [[TMP24]]
 // CHECK3-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK3-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK3:       omp.body.continue:
@@ -2664,24 +2664,24 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK5-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP15]] to i64
-// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM]]
+// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM]]
 // CHECK5-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[IDXPROM3:%.*]] = zext i32 [[TMP18]] to i64
-// CHECK5-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM3]]
+// CHECK5-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM3]]
 // CHECK5-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[MUL5:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK5-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[IDXPROM6:%.*]] = zext i32 [[TMP21]] to i64
-// CHECK5-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM6]]
+// CHECK5-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM6]]
 // CHECK5-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[MUL8:%.*]] = fmul float [[MUL5]], [[TMP22]]
 // CHECK5-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    [[IDXPROM9:%.*]] = zext i32 [[TMP24]] to i64
-// CHECK5-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM9]]
+// CHECK5-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM9]]
 // CHECK5-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK5-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK5:       omp.body.continue:
@@ -3671,21 +3671,21 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK7-NEXT:    store i32 [[ADD]], ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK7-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK7-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
-// CHECK7-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i32 [[TMP15]]
+// CHECK7-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i32 [[TMP15]]
 // CHECK7-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK7-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK7-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
-// CHECK7-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i32 [[TMP18]]
+// CHECK7-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i32 [[TMP18]]
 // CHECK7-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX3]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK7-NEXT:    [[MUL4:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK7-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK7-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
-// CHECK7-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i32 [[TMP21]]
+// CHECK7-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i32 [[TMP21]]
 // CHECK7-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX5]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK7-NEXT:    [[MUL6:%.*]] = fmul float [[MUL4]], [[TMP22]]
 // CHECK7-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK7-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
-// CHECK7-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i32 [[TMP24]]
+// CHECK7-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i32 [[TMP24]]
 // CHECK7-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK7-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK7:       omp.body.continue:
@@ -4290,24 +4290,24 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK9-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[B_ADDR]], align 8, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[TMP5:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP5]] to i64
-// CHECK9-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP4]], i64 [[IDXPROM]]
+// CHECK9-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP4]], i64 [[IDXPROM]]
 // CHECK9-NEXT:    [[TMP6:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[C_ADDR]], align 8, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[TMP8:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[IDXPROM1:%.*]] = zext i32 [[TMP8]] to i64
-// CHECK9-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP7]], i64 [[IDXPROM1]]
+// CHECK9-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP7]], i64 [[IDXPROM1]]
 // CHECK9-NEXT:    [[TMP9:%.*]] = load float, ptr [[ARRAYIDX2]], align 4, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[MUL3:%.*]] = fmul float [[TMP6]], [[TMP9]]
 // CHECK9-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[D_ADDR]], align 8, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[TMP11:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[IDXPROM4:%.*]] = zext i32 [[TMP11]] to i64
-// CHECK9-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP10]], i64 [[IDXPROM4]]
+// CHECK9-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP10]], i64 [[IDXPROM4]]
 // CHECK9-NEXT:    [[TMP12:%.*]] = load float, ptr [[ARRAYIDX5]], align 4, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[MUL6:%.*]] = fmul float [[MUL3]], [[TMP12]]
 // CHECK9-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[A_ADDR]], align 8, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[TMP14:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP14]] to i64
-// CHECK9-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i64 [[IDXPROM7]]
+// CHECK9-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i64 [[IDXPROM7]]
 // CHECK9-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX8]], align 4, !llvm.access.group [[ACC_GRP9]]
 // CHECK9-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK9:       omp.body.continue:
@@ -4606,21 +4606,21 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK11-NEXT:    store i32 [[ADD]], ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK11-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[B_ADDR]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK11-NEXT:    [[TMP5:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
-// CHECK11-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP4]], i32 [[TMP5]]
+// CHECK11-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP4]], i32 [[TMP5]]
 // CHECK11-NEXT:    [[TMP6:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK11-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[C_ADDR]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK11-NEXT:    [[TMP8:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
-// CHECK11-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw float, ptr [[TMP7]], i32 [[TMP8]]
+// CHECK11-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds float, ptr [[TMP7]], i32 [[TMP8]]
 // CHECK11-NEXT:    [[TMP9:%.*]] = load float, ptr [[ARRAYIDX1]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK11-NEXT:    [[MUL2:%.*]] = fmul float [[TMP6]], [[TMP9]]
 // CHECK11-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[D_ADDR]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK11-NEXT:    [[TMP11:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
-// CHECK11-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP10]], i32 [[TMP11]]
+// CHECK11-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP10]], i32 [[TMP11]]
 // CHECK11-NEXT:    [[TMP12:%.*]] = load float, ptr [[ARRAYIDX3]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK11-NEXT:    [[MUL4:%.*]] = fmul float [[MUL2]], [[TMP12]]
 // CHECK11-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[A_ADDR]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK11-NEXT:    [[TMP14:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
-// CHECK11-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i32 [[TMP14]]
+// CHECK11-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i32 [[TMP14]]
 // CHECK11-NEXT:    store float [[MUL4]], ptr [[ARRAYIDX5]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK11-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK11:       omp.body.continue:
@@ -4928,24 +4928,24 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK13-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[B_ADDR]], align 8, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[TMP5:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP5]] to i64
-// CHECK13-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP4]], i64 [[IDXPROM]]
+// CHECK13-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP4]], i64 [[IDXPROM]]
 // CHECK13-NEXT:    [[TMP6:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[C_ADDR]], align 8, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[TMP8:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[IDXPROM1:%.*]] = zext i32 [[TMP8]] to i64
-// CHECK13-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP7]], i64 [[IDXPROM1]]
+// CHECK13-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP7]], i64 [[IDXPROM1]]
 // CHECK13-NEXT:    [[TMP9:%.*]] = load float, ptr [[ARRAYIDX2]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[MUL3:%.*]] = fmul float [[TMP6]], [[TMP9]]
 // CHECK13-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[D_ADDR]], align 8, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[TMP11:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[IDXPROM4:%.*]] = zext i32 [[TMP11]] to i64
-// CHECK13-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP10]], i64 [[IDXPROM4]]
+// CHECK13-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP10]], i64 [[IDXPROM4]]
 // CHECK13-NEXT:    [[TMP12:%.*]] = load float, ptr [[ARRAYIDX5]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[MUL6:%.*]] = fmul float [[MUL3]], [[TMP12]]
 // CHECK13-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[A_ADDR]], align 8, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[TMP14:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP14]] to i64
-// CHECK13-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i64 [[IDXPROM7]]
+// CHECK13-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i64 [[IDXPROM7]]
 // CHECK13-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX8]], align 4, !llvm.access.group [[ACC_GRP10]]
 // CHECK13-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK13:       omp.body.continue:
@@ -5275,21 +5275,21 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK15-NEXT:    store i32 [[ADD]], ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK15-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[B_ADDR]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK15-NEXT:    [[TMP5:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
-// CHECK15-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP4]], i32 [[TMP5]]
+// CHECK15-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP4]], i32 [[TMP5]]
 // CHECK15-NEXT:    [[TMP6:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK15-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[C_ADDR]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK15-NEXT:    [[TMP8:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
-// CHECK15-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw float, ptr [[TMP7]], i32 [[TMP8]]
+// CHECK15-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds float, ptr [[TMP7]], i32 [[TMP8]]
 // CHECK15-NEXT:    [[TMP9:%.*]] = load float, ptr [[ARRAYIDX1]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK15-NEXT:    [[MUL2:%.*]] = fmul float [[TMP6]], [[TMP9]]
 // CHECK15-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[D_ADDR]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK15-NEXT:    [[TMP11:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
-// CHECK15-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP10]], i32 [[TMP11]]
+// CHECK15-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP10]], i32 [[TMP11]]
 // CHECK15-NEXT:    [[TMP12:%.*]] = load float, ptr [[ARRAYIDX3]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK15-NEXT:    [[MUL4:%.*]] = fmul float [[MUL2]], [[TMP12]]
 // CHECK15-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[A_ADDR]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK15-NEXT:    [[TMP14:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP11]]
-// CHECK15-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i32 [[TMP14]]
+// CHECK15-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i32 [[TMP14]]
 // CHECK15-NEXT:    store float [[MUL4]], ptr [[ARRAYIDX5]], align 4, !llvm.access.group [[ACC_GRP11]]
 // CHECK15-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK15:       omp.body.continue:
@@ -5782,24 +5782,24 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK17-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP15]] to i64
-// CHECK17-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM]]
+// CHECK17-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM]]
 // CHECK17-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[IDXPROM3:%.*]] = zext i32 [[TMP18]] to i64
-// CHECK17-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM3]]
+// CHECK17-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM3]]
 // CHECK17-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[MUL5:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK17-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[IDXPROM6:%.*]] = zext i32 [[TMP21]] to i64
-// CHECK17-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM6]]
+// CHECK17-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM6]]
 // CHECK17-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[MUL8:%.*]] = fmul float [[MUL5]], [[TMP22]]
 // CHECK17-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    [[IDXPROM9:%.*]] = zext i32 [[TMP24]] to i64
-// CHECK17-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM9]]
+// CHECK17-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM9]]
 // CHECK17-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4, !llvm.access.group [[ACC_GRP18]]
 // CHECK17-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK17:       omp.body.continue:
@@ -6373,21 +6373,21 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK19-NEXT:    store i32 [[ADD]], ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK19-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK19-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
-// CHECK19-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i32 [[TMP15]]
+// CHECK19-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i32 [[TMP15]]
 // CHECK19-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK19-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK19-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
-// CHECK19-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i32 [[TMP18]]
+// CHECK19-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i32 [[TMP18]]
 // CHECK19-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX3]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK19-NEXT:    [[MUL4:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK19-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK19-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
-// CHECK19-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i32 [[TMP21]]
+// CHECK19-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i32 [[TMP21]]
 // CHECK19-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX5]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK19-NEXT:    [[MUL6:%.*]] = fmul float [[MUL4]], [[TMP22]]
 // CHECK19-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK19-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
-// CHECK19-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i32 [[TMP24]]
+// CHECK19-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i32 [[TMP24]]
 // CHECK19-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK19-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK19:       omp.body.continue:
@@ -6970,24 +6970,24 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK21-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP15]] to i64
-// CHECK21-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM]]
+// CHECK21-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM]]
 // CHECK21-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[IDXPROM3:%.*]] = zext i32 [[TMP18]] to i64
-// CHECK21-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM3]]
+// CHECK21-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM3]]
 // CHECK21-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[MUL5:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK21-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[IDXPROM6:%.*]] = zext i32 [[TMP21]] to i64
-// CHECK21-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM6]]
+// CHECK21-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM6]]
 // CHECK21-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[MUL8:%.*]] = fmul float [[MUL5]], [[TMP22]]
 // CHECK21-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    [[IDXPROM9:%.*]] = zext i32 [[TMP24]] to i64
-// CHECK21-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM9]]
+// CHECK21-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM9]]
 // CHECK21-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4, !llvm.access.group [[ACC_GRP19]]
 // CHECK21-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK21:       omp.body.continue:
@@ -7592,21 +7592,21 @@ int fint(void) { return ftemplate<int>(); }
 // CHECK23-NEXT:    store i32 [[ADD]], ptr [[I]], align 4, !llvm.access.group [[ACC_GRP20]]
 // CHECK23-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 4, !llvm.access.group [[ACC_GRP20]]
 // CHECK23-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP20]]
-// CHECK23-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i32 [[TMP15]]
+// CHECK23-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i32 [[TMP15]]
 // CHECK23-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP20]]
 // CHECK23-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 4, !llvm.access.group [[ACC_GRP20]]
 // CHECK23-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP20]]
-// CHECK23-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i32 [[TMP18]]
+// CHECK23-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i32 [[TMP18]]
 // CHECK23-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX3]], align 4, !llvm.access.group [[ACC_GRP20]]
 // CHECK23-NEXT:    [[MUL4:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK23-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 4, !llvm.access.group [[ACC_GRP20]]
 // CHECK23-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP20]]
-// CHECK23-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i32 [[TMP21]]
+// CHECK23-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i32 [[TMP21]]
 // CHECK23-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX5]], align 4, !llvm.access.group [[ACC_GRP20]]
 // CHECK23-NEXT:    [[MUL6:%.*]] = fmul float [[MUL4]], [[TMP22]]
 // CHECK23-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 4, !llvm.access.group [[ACC_GRP20]]
 // CHECK23-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !llvm.access.group [[ACC_GRP20]]
-// CHECK23-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i32 [[TMP24]]
+// CHECK23-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i32 [[TMP24]]
 // CHECK23-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP20]]
 // CHECK23-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK23:       omp.body.continue:
diff --git a/clang/test/OpenMP/for_linear_codegen.cpp b/clang/test/OpenMP/for_linear_codegen.cpp
index 5a21fe8509fd36..395ccdbeed7634 100644
--- a/clang/test/OpenMP/for_linear_codegen.cpp
+++ b/clang/test/OpenMP/for_linear_codegen.cpp
@@ -650,7 +650,7 @@ int main() {
 // CHECK1-NEXT:    [[ADD10:%.*]] = add nsw i32 [[TMP16]], [[MUL9]]
 // CHECK1-NEXT:    store i32 [[ADD10]], ptr [[LVAR5]], align 4
 // CHECK1-NEXT:    [[TMP18:%.*]] = load ptr, ptr [[PVAR4]], align 8
-// CHECK1-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP18]], i32 1
+// CHECK1-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP18]], i32 1
 // CHECK1-NEXT:    store ptr [[INCDEC_PTR]], ptr [[PVAR4]], align 8
 // CHECK1-NEXT:    [[TMP19:%.*]] = load ptr, ptr [[_TMP6]], align 8
 // CHECK1-NEXT:    [[TMP20:%.*]] = load i32, ptr [[TMP19]], align 4
diff --git a/clang/test/OpenMP/for_reduction_codegen.cpp b/clang/test/OpenMP/for_reduction_codegen.cpp
index 83632db2384843..ea32e98bf1423a 100644
--- a/clang/test/OpenMP/for_reduction_codegen.cpp
+++ b/clang/test/OpenMP/for_reduction_codegen.cpp
@@ -1021,14 +1021,14 @@ int main() {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP5:%.*]] = mul nsw i64 1, [[TMP1]]
 // CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i64 [[TMP5]]
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX]], i64 0
 // CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds [2 x i32], ptr [[TMP3]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP6:%.*]] = load i32, ptr [[ARRAYIDX4]], align 4
 // CHECK1-NEXT:    [[TMP7:%.*]] = sext i32 [[TMP6]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP7]]
 // CHECK1-NEXT:    [[TMP8:%.*]] = mul nsw i64 1, [[TMP1]]
 // CHECK1-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i64 [[TMP8]]
-// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX5]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX5]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP9:%.*]] = ptrtoint ptr [[ARRAYIDX6]] to i64
 // CHECK1-NEXT:    [[TMP10:%.*]] = ptrtoint ptr [[ARRAYIDX3]] to i64
 // CHECK1-NEXT:    [[TMP11:%.*]] = sub i64 [[TMP9]], [[TMP10]]
@@ -1054,16 +1054,16 @@ int main() {
 // CHECK1-NEXT:    [[TMP19:%.*]] = sub i64 [[TMP17]], [[TMP18]]
 // CHECK1-NEXT:    [[TMP20:%.*]] = sdiv exact i64 [[TMP19]], ptrtoint (ptr getelementptr (i32, ptr null, i32 1) to i64)
 // CHECK1-NEXT:    [[TMP21:%.*]] = getelementptr i32, ptr [[VLA7]], i64 [[TMP20]]
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw [10 x [4 x %struct.S]], ptr [[TMP4]], i64 0, i64 1
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds [10 x [4 x %struct.S]], ptr [[TMP4]], i64 0, i64 1
 // CHECK1-NEXT:    [[ARRAYDECAY:%.*]] = getelementptr inbounds [4 x %struct.S], ptr [[ARRAYIDX8]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw [[STRUCT_S:%.*]], ptr [[ARRAYDECAY]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds [[STRUCT_S:%.*]], ptr [[ARRAYDECAY]], i64 1
 // CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds [2 x i32], ptr [[TMP3]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP22:%.*]] = load i32, ptr [[ARRAYIDX10]], align 4
 // CHECK1-NEXT:    [[TMP23:%.*]] = sext i32 [[TMP22]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN11:%.*]] = add nsw i64 0, [[TMP23]]
-// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds nuw [10 x [4 x %struct.S]], ptr [[TMP4]], i64 0, i64 [[LB_ADD_LEN11]]
+// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds [10 x [4 x %struct.S]], ptr [[TMP4]], i64 0, i64 [[LB_ADD_LEN11]]
 // CHECK1-NEXT:    [[ARRAYDECAY13:%.*]] = getelementptr inbounds [4 x %struct.S], ptr [[ARRAYIDX12]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds nuw [[STRUCT_S]], ptr [[ARRAYDECAY13]], i64 2
+// CHECK1-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds [[STRUCT_S]], ptr [[ARRAYDECAY13]], i64 2
 // CHECK1-NEXT:    [[TMP24:%.*]] = ptrtoint ptr [[ARRAYIDX14]] to i64
 // CHECK1-NEXT:    [[TMP25:%.*]] = ptrtoint ptr [[ARRAYIDX9]] to i64
 // CHECK1-NEXT:    [[TMP26:%.*]] = sub i64 [[TMP24]], [[TMP25]]
@@ -1580,10 +1580,10 @@ int main() {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP3:%.*]] = mul nsw i64 1, [[TMP1]]
 // CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i64 [[TMP3]]
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX]], i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = mul nsw i64 1, [[TMP1]]
 // CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i64 [[TMP4]]
-// CHECK1-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX4]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX4]], i64 1
 // CHECK1-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [1 x [2 x i32]], ptr [[ARR6]], i32 0, i32 0, i32 0
 // CHECK1-NEXT:    [[TMP5:%.*]] = getelementptr i32, ptr [[ARRAY_BEGIN]], i64 2
 // CHECK1-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP5]]
@@ -1757,13 +1757,13 @@ int main() {
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw [[STRUCT_S:%.*]], ptr [[TMP2]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [[STRUCT_S:%.*]], ptr [[TMP2]], i64 1
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP3]], i64 4
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds ptr, ptr [[TMP3]], i64 4
 // CHECK1-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[ARRAYIDX2]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [[STRUCT_S]], ptr [[TMP4]], i64 6
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [[STRUCT_S]], ptr [[TMP4]], i64 6
 // CHECK1-NEXT:    [[TMP5:%.*]] = ptrtoint ptr [[ARRAYIDX3]] to i64
 // CHECK1-NEXT:    [[TMP6:%.*]] = ptrtoint ptr [[ARRAYIDX1]] to i64
 // CHECK1-NEXT:    [[TMP7:%.*]] = sub i64 [[TMP5]], [[TMP6]]
@@ -1963,11 +1963,11 @@ int main() {
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
 // CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 1
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw [[STRUCT_S:%.*]], ptr [[TMP2]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [[STRUCT_S:%.*]], ptr [[TMP2]], i64 1
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP0]], align 8
 // CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds ptr, ptr [[TMP3]], i64 1
 // CHECK1-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[ARRAYIDX2]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [[STRUCT_S]], ptr [[TMP4]], i64 6
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [[STRUCT_S]], ptr [[TMP4]], i64 6
 // CHECK1-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [1 x [6 x %struct.S]], ptr [[VAR24]], i32 0, i32 0, i32 0
 // CHECK1-NEXT:    [[TMP5:%.*]] = getelementptr [[STRUCT_S]], ptr [[ARRAY_BEGIN]], i64 6
 // CHECK1-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP5]]
@@ -2148,13 +2148,13 @@ int main() {
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 1
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw [[STRUCT_S:%.*]], ptr [[TMP2]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [[STRUCT_S:%.*]], ptr [[TMP2]], i64 1
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP3]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds ptr, ptr [[TMP3]], i64 1
 // CHECK1-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[ARRAYIDX2]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [[STRUCT_S]], ptr [[TMP4]], i64 6
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [[STRUCT_S]], ptr [[TMP4]], i64 6
 // CHECK1-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [1 x [6 x %struct.S]], ptr [[VAR24]], i32 0, i32 0, i32 0
 // CHECK1-NEXT:    [[TMP5:%.*]] = getelementptr [[STRUCT_S]], ptr [[ARRAY_BEGIN]], i64 6
 // CHECK1-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP5]]
@@ -2335,13 +2335,13 @@ int main() {
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 1
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw [[STRUCT_S]], ptr [[TMP2]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [[STRUCT_S]], ptr [[TMP2]], i64 1
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP3]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds ptr, ptr [[TMP3]], i64 1
 // CHECK1-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[ARRAYIDX2]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [[STRUCT_S]], ptr [[TMP4]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [[STRUCT_S]], ptr [[TMP4]], i64 1
 // CHECK1-NEXT:    call void @_ZN1SIfEC1Ev(ptr noundef nonnull align 4 dereferenceable(4) [[VAR24]])
 // CHECK1-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP0]], align 8
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[TMP5]], align 8
@@ -2459,8 +2459,8 @@ int main() {
 // CHECK1-NEXT:    store i32 9, ptr [[DOTOMP_UB]], align 4
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [5 x %struct.S], ptr [[TMP0]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw [5 x %struct.S], ptr [[TMP0]], i64 0, i64 4
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [5 x %struct.S], ptr [[TMP0]], i64 0, i64 0
+// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [5 x %struct.S], ptr [[TMP0]], i64 0, i64 4
 // CHECK1-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [5 x %struct.S], ptr [[VVAR22]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP1:%.*]] = getelementptr [[STRUCT_S:%.*]], ptr [[ARRAY_BEGIN]], i64 5
 // CHECK1-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP1]]
@@ -2641,9 +2641,9 @@ int main() {
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[_TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [4 x %struct.S], ptr [[TMP2]], i64 0, i64 1
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [4 x %struct.S], ptr [[TMP2]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[_TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [4 x %struct.S], ptr [[TMP3]], i64 0, i64 2
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [4 x %struct.S], ptr [[TMP3]], i64 0, i64 2
 // CHECK1-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [2 x %struct.S], ptr [[VAR34]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = getelementptr [[STRUCT_S:%.*]], ptr [[ARRAY_BEGIN]], i64 2
 // CHECK1-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP4]]
@@ -2826,9 +2826,9 @@ int main() {
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[_TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [4 x %struct.S], ptr [[TMP2]], i64 0, i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [4 x %struct.S], ptr [[TMP2]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[_TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [4 x %struct.S], ptr [[TMP3]], i64 0, i64 1
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [4 x %struct.S], ptr [[TMP3]], i64 0, i64 1
 // CHECK1-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [2 x %struct.S], ptr [[VAR34]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = getelementptr [[STRUCT_S:%.*]], ptr [[ARRAY_BEGIN]], i64 2
 // CHECK1-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP4]]
@@ -3012,9 +3012,9 @@ int main() {
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[_TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [4 x %struct.S], ptr [[TMP2]], i64 0, i64 2
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [4 x %struct.S], ptr [[TMP2]], i64 0, i64 2
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[_TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [4 x %struct.S], ptr [[TMP3]], i64 0, i64 3
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [4 x %struct.S], ptr [[TMP3]], i64 0, i64 3
 // CHECK1-NEXT:    [[TMP4:%.*]] = ptrtoint ptr [[ARRAYIDX3]] to i64
 // CHECK1-NEXT:    [[TMP5:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK1-NEXT:    [[TMP6:%.*]] = sub i64 [[TMP4]], [[TMP5]]
@@ -3974,8 +3974,8 @@ int main() {
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_UB]], align 4
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [42 x %struct.S.0], ptr [[TMP0]], i64 0, i64 1
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [42 x %struct.S.0], ptr [[TMP0]], i64 0, i64 40
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [42 x %struct.S.0], ptr [[TMP0]], i64 0, i64 1
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [42 x %struct.S.0], ptr [[TMP0]], i64 0, i64 40
 // CHECK1-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [40 x %struct.S.0], ptr [[ARR4]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP6:%.*]] = getelementptr [[STRUCT_S_0]], ptr [[ARRAY_BEGIN]], i64 40
 // CHECK1-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP6]]
diff --git a/clang/test/OpenMP/for_reduction_codegen_UDR.cpp b/clang/test/OpenMP/for_reduction_codegen_UDR.cpp
index 82f94c949eea64..16d6c23542fce8 100644
--- a/clang/test/OpenMP/for_reduction_codegen_UDR.cpp
+++ b/clang/test/OpenMP/for_reduction_codegen_UDR.cpp
@@ -1074,14 +1074,14 @@ int main() {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP5:%.*]] = mul nsw i64 1, [[TMP1]]
 // CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i64 [[TMP5]]
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX]], i64 0
 // CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds [2 x i32], ptr [[TMP3]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP6:%.*]] = load i32, ptr [[ARRAYIDX4]], align 4
 // CHECK1-NEXT:    [[TMP7:%.*]] = sext i32 [[TMP6]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP7]]
 // CHECK1-NEXT:    [[TMP8:%.*]] = mul nsw i64 1, [[TMP1]]
 // CHECK1-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i64 [[TMP8]]
-// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX5]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX5]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP9:%.*]] = ptrtoint ptr [[ARRAYIDX6]] to i64
 // CHECK1-NEXT:    [[TMP10:%.*]] = ptrtoint ptr [[ARRAYIDX3]] to i64
 // CHECK1-NEXT:    [[TMP11:%.*]] = sub i64 [[TMP9]], [[TMP10]]
@@ -1109,16 +1109,16 @@ int main() {
 // CHECK1-NEXT:    [[TMP19:%.*]] = sub i64 [[TMP17]], [[TMP18]]
 // CHECK1-NEXT:    [[TMP20:%.*]] = sdiv exact i64 [[TMP19]], ptrtoint (ptr getelementptr (i32, ptr null, i32 1) to i64)
 // CHECK1-NEXT:    [[TMP21:%.*]] = getelementptr i32, ptr [[VLA7]], i64 [[TMP20]]
-// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw [10 x [4 x %struct.S.0]], ptr [[TMP4]], i64 0, i64 1
+// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds [10 x [4 x %struct.S.0]], ptr [[TMP4]], i64 0, i64 1
 // CHECK1-NEXT:    [[ARRAYDECAY:%.*]] = getelementptr inbounds [4 x %struct.S.0], ptr [[ARRAYIDX9]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw [[STRUCT_S_0:%.*]], ptr [[ARRAYDECAY]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds [[STRUCT_S_0:%.*]], ptr [[ARRAYDECAY]], i64 1
 // CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds [2 x i32], ptr [[TMP3]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP22:%.*]] = load i32, ptr [[ARRAYIDX11]], align 4
 // CHECK1-NEXT:    [[TMP23:%.*]] = sext i32 [[TMP22]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN12:%.*]] = add nsw i64 0, [[TMP23]]
-// CHECK1-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw [10 x [4 x %struct.S.0]], ptr [[TMP4]], i64 0, i64 [[LB_ADD_LEN12]]
+// CHECK1-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds [10 x [4 x %struct.S.0]], ptr [[TMP4]], i64 0, i64 [[LB_ADD_LEN12]]
 // CHECK1-NEXT:    [[ARRAYDECAY14:%.*]] = getelementptr inbounds [4 x %struct.S.0], ptr [[ARRAYIDX13]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds nuw [[STRUCT_S_0]], ptr [[ARRAYDECAY14]], i64 2
+// CHECK1-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds [[STRUCT_S_0]], ptr [[ARRAYDECAY14]], i64 2
 // CHECK1-NEXT:    [[TMP24:%.*]] = ptrtoint ptr [[ARRAYIDX15]] to i64
 // CHECK1-NEXT:    [[TMP25:%.*]] = ptrtoint ptr [[ARRAYIDX10]] to i64
 // CHECK1-NEXT:    [[TMP26:%.*]] = sub i64 [[TMP24]], [[TMP25]]
@@ -1669,13 +1669,13 @@ int main() {
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw [[STRUCT_S_0:%.*]], ptr [[TMP2]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [[STRUCT_S_0:%.*]], ptr [[TMP2]], i64 1
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP3]], i64 4
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds ptr, ptr [[TMP3]], i64 4
 // CHECK1-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[ARRAYIDX2]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [[STRUCT_S_0]], ptr [[TMP4]], i64 6
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [[STRUCT_S_0]], ptr [[TMP4]], i64 6
 // CHECK1-NEXT:    [[TMP5:%.*]] = ptrtoint ptr [[ARRAYIDX3]] to i64
 // CHECK1-NEXT:    [[TMP6:%.*]] = ptrtoint ptr [[ARRAYIDX1]] to i64
 // CHECK1-NEXT:    [[TMP7:%.*]] = sub i64 [[TMP5]], [[TMP6]]
@@ -1877,8 +1877,8 @@ int main() {
 // CHECK1-NEXT:    store i32 9, ptr [[DOTOMP_UB]], align 4
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [5 x %struct.S.0], ptr [[TMP0]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw [5 x %struct.S.0], ptr [[TMP0]], i64 0, i64 4
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [5 x %struct.S.0], ptr [[TMP0]], i64 0, i64 0
+// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [5 x %struct.S.0], ptr [[TMP0]], i64 0, i64 4
 // CHECK1-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [5 x %struct.S.0], ptr [[VVAR22]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP1:%.*]] = getelementptr [[STRUCT_S_0:%.*]], ptr [[ARRAY_BEGIN]], i64 5
 // CHECK1-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP1]]
@@ -2066,9 +2066,9 @@ int main() {
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[_TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [4 x %struct.S.0], ptr [[TMP2]], i64 0, i64 1
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [4 x %struct.S.0], ptr [[TMP2]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[_TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [4 x %struct.S.0], ptr [[TMP3]], i64 0, i64 2
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [4 x %struct.S.0], ptr [[TMP3]], i64 0, i64 2
 // CHECK1-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [2 x %struct.S.0], ptr [[VAR34]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = getelementptr [[STRUCT_S_0:%.*]], ptr [[ARRAY_BEGIN]], i64 2
 // CHECK1-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP4]]
@@ -2979,8 +2979,8 @@ int main() {
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_UB]], align 4
 // CHECK1-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [42 x %struct.S], ptr [[TMP0]], i64 0, i64 1
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw [42 x %struct.S], ptr [[TMP0]], i64 0, i64 40
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [42 x %struct.S], ptr [[TMP0]], i64 0, i64 1
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds [42 x %struct.S], ptr [[TMP0]], i64 0, i64 40
 // CHECK1-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [40 x %struct.S], ptr [[ARR4]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP6:%.*]] = getelementptr [[STRUCT_S:%.*]], ptr [[ARRAY_BEGIN]], i64 40
 // CHECK1-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP6]]
diff --git a/clang/test/OpenMP/for_reduction_task_codegen.cpp b/clang/test/OpenMP/for_reduction_task_codegen.cpp
index b875279c2a1440..ea93323de77d0f 100644
--- a/clang/test/OpenMP/for_reduction_task_codegen.cpp
+++ b/clang/test/OpenMP/for_reduction_task_codegen.cpp
@@ -68,16 +68,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP3]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP3]], i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP5:%.*]] = sext i32 [[TMP4]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP5]]
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP6]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP6]], i64 9
 // CHECK1-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP7]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP7]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP9:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP10:%.*]] = sub i64 [[TMP8]], [[TMP9]]
@@ -107,7 +107,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP21]]
 // CHECK1-NEXT:    store ptr [[_TMP6]], ptr [[_TMP5]], align 8
 // CHECK1-NEXT:    store ptr [[TMP22]], ptr [[_TMP6]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP23]], align 8
 // CHECK1-NEXT:    [[TMP24:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -122,19 +122,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP28]], align 8
 // CHECK1-NEXT:    [[TMP29:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP29]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP30:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP31:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP31]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds ptr, ptr [[TMP31]], i64 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load ptr, ptr [[ARRAYIDX8]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP32]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds i8, ptr [[TMP32]], i64 0
 // CHECK1-NEXT:    [[TMP33:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP34:%.*]] = sext i32 [[TMP33]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN10:%.*]] = add nsw i64 -1, [[TMP34]]
 // CHECK1-NEXT:    [[TMP35:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP35]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds ptr, ptr [[TMP35]], i64 9
 // CHECK1-NEXT:    [[TMP36:%.*]] = load ptr, ptr [[ARRAYIDX11]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP36]], i64 [[LB_ADD_LEN10]]
+// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds i8, ptr [[TMP36]], i64 [[LB_ADD_LEN10]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP30]], align 8
 // CHECK1-NEXT:    [[TMP37:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX9]], ptr [[TMP37]], align 8
@@ -459,9 +459,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]
diff --git a/clang/test/OpenMP/for_scan_codegen.cpp b/clang/test/OpenMP/for_scan_codegen.cpp
index 61e6534db471e7..4cf18a76fbfefa 100644
--- a/clang/test/OpenMP/for_scan_codegen.cpp
+++ b/clang/test/OpenMP/for_scan_codegen.cpp
@@ -39,13 +39,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_BUF_IDX]], ptr {{.*}}[[A_PRIV]], i64 [[BYTES]], i1 false)
 
     // b_buffer[i] = b_priv;
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX]]
     // CHECK: [[B_PRIV:%.+]] = load double, ptr [[B_PRIV_ADDR]],
     // CHECK: store double [[B_PRIV]], ptr [[B_BUF_IDX]],
     // CHECK: br label %[[LOOP_CONTINUE:.+]]
@@ -72,13 +72,13 @@ void baz(int n) {
 
     // a_buffer[i] += a_buffer[i-pow(2, k)];
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[I]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[IDX_SUB_K2POW]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[I]]
+    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[I]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
-    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
+    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
     // CHECK: [[A_BUF_END:%.+]] = getelementptr float, ptr [[A_BUF_IDX]], i64 [[NUM_ELEMS]]
     // CHECK: [[ISEMPTY:%.+]] = icmp eq ptr [[A_BUF_IDX]], [[A_BUF_END]]
     // CHECK: br i1 [[ISEMPTY]], label %[[RED_DONE:[^,]+]], label %[[RED_BODY:[^,]+]]
@@ -132,13 +132,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_PRIV]], ptr {{.*}}[[A_BUF_IDX]], i64 [[BYTES]], i1 false)
 
     // b_priv = b_buffer[i];
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX]]
     // CHECK: [[B_BUF_IDX_VAL:%.+]] = load double, ptr [[B_BUF_IDX]],
     // CHECK: store double [[B_BUF_IDX_VAL]], ptr [[B_PRIV_ADDR]],
     // CHECK: br label %[[SCAN_PHASE:[^,]+]]
@@ -179,13 +179,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_BUF_IDX]], ptr {{.*}}[[A_PRIV]], i64 [[BYTES]], i1 false)
 
     // b_buffer[i] = b_priv;
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX]]
     // CHECK: [[B_PRIV:%.+]] = load double, ptr [[B_PRIV_ADDR]],
     // CHECK: store double [[B_PRIV]], ptr [[B_BUF_IDX]],
     // CHECK: br label %[[LOOP_CONTINUE:[^,]+]]
@@ -217,13 +217,13 @@ void baz(int n) {
 
     // a_buffer[i] += a_buffer[i-pow(2, k)];
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[I]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[IDX_SUB_K2POW]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[I]]
+    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[I]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
-    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
+    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
     // CHECK: [[A_BUF_END:%.+]] = getelementptr float, ptr [[A_BUF_IDX]], i64 [[NUM_ELEMS]]
     // CHECK: [[ISEMPTY:%.+]] = icmp eq ptr [[A_BUF_IDX]], [[A_BUF_END]]
     // CHECK: br i1 [[ISEMPTY]], label %[[RED_DONE:[^,]+]], label %[[RED_BODY:[^,]+]]
@@ -280,13 +280,13 @@ void baz(int n) {
     // CHECK: [[IF_THEN]]:
     // CHECK: [[BASE_IDX_SUB_1:%.+]] = sub nuw i64 [[BASE_IDX]], 1
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX_SUB_1]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_PRIV]], ptr {{.*}}[[A_BUF_IDX]], i64 [[BYTES]], i1 false)
 
     // b_priv = b_buffer[i];
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX_SUB_1]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX_SUB_1]]
     // CHECK: [[B_BUF_IDX_VAL:%.+]] = load double, ptr [[B_BUF_IDX]],
     // CHECK: store double [[B_BUF_IDX_VAL]], ptr [[B_PRIV_ADDR]],
     // CHECK: br label %[[SCAN_PHASE]]
diff --git a/clang/test/OpenMP/for_simd_scan_codegen.cpp b/clang/test/OpenMP/for_simd_scan_codegen.cpp
index 829f2656042fb4..29af5f74c5b5bd 100644
--- a/clang/test/OpenMP/for_simd_scan_codegen.cpp
+++ b/clang/test/OpenMP/for_simd_scan_codegen.cpp
@@ -39,13 +39,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_BUF_IDX]], ptr {{.*}}[[A_PRIV]], i64 [[BYTES]], i1 false)
 
     // b_buffer[i] = b_priv;
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX]]
     // CHECK: [[B_PRIV:%.+]] = load double, ptr [[B_PRIV_ADDR]],
     // CHECK: store double [[B_PRIV]], ptr [[B_BUF_IDX]],
     // CHECK: br label %[[LOOP_CONTINUE:.+]]
@@ -72,13 +72,13 @@ void baz(int n) {
 
     // a_buffer[i] += a_buffer[i-pow(2, k)];
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[I]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[IDX_SUB_K2POW]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[I]]
+    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[I]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
-    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
+    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
     // CHECK: [[A_BUF_END:%.+]] = getelementptr float, ptr [[A_BUF_IDX]], i64 [[NUM_ELEMS]]
     // CHECK: [[ISEMPTY:%.+]] = icmp eq ptr [[A_BUF_IDX]], [[A_BUF_END]]
     // CHECK: br i1 [[ISEMPTY]], label %[[RED_DONE:[^,]+]], label %[[RED_BODY:[^,]+]]
@@ -132,13 +132,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_PRIV]], ptr {{.*}}[[A_BUF_IDX]], i64 [[BYTES]], i1 false)
 
     // b_priv = b_buffer[i];
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX]]
     // CHECK: [[B_BUF_IDX_VAL:%.+]] = load double, ptr [[B_BUF_IDX]],
     // CHECK: store double [[B_BUF_IDX_VAL]], ptr [[B_PRIV_ADDR]],
     // CHECK: br label %[[SCAN_PHASE:[^,]+]]
@@ -179,13 +179,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_BUF_IDX]], ptr {{.*}}[[A_PRIV]], i64 [[BYTES]], i1 false)
 
     // b_buffer[i] = b_priv;
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX]]
     // CHECK: [[B_PRIV:%.+]] = load double, ptr [[B_PRIV_ADDR]],
     // CHECK: store double [[B_PRIV]], ptr [[B_BUF_IDX]],
     // CHECK: br label %[[LOOP_CONTINUE:[^,]+]]
@@ -217,13 +217,13 @@ void baz(int n) {
 
     // a_buffer[i] += a_buffer[i-pow(2, k)];
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[I]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[IDX_SUB_K2POW]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[I]]
+    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[I]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
-    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
+    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
     // CHECK: [[A_BUF_END:%.+]] = getelementptr float, ptr [[A_BUF_IDX]], i64 [[NUM_ELEMS]]
     // CHECK: [[ISEMPTY:%.+]] = icmp eq ptr [[A_BUF_IDX]], [[A_BUF_END]]
     // CHECK: br i1 [[ISEMPTY]], label %[[RED_DONE:[^,]+]], label %[[RED_BODY:[^,]+]]
@@ -280,13 +280,13 @@ void baz(int n) {
     // CHECK: [[IF_THEN]]:
     // CHECK: [[BASE_IDX_SUB_1:%.+]] = sub nuw i64 [[BASE_IDX]], 1
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX_SUB_1]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_PRIV]], ptr {{.*}}[[A_BUF_IDX]], i64 [[BYTES]], i1 false)
 
     // b_priv = b_buffer[i];
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX_SUB_1]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX_SUB_1]]
     // CHECK: [[B_BUF_IDX_VAL:%.+]] = load double, ptr [[B_BUF_IDX]],
     // CHECK: store double [[B_BUF_IDX_VAL]], ptr [[B_PRIV_ADDR]],
     // CHECK: br label %[[SCAN_PHASE]]
diff --git a/clang/test/OpenMP/irbuilder_for_iterator.cpp b/clang/test/OpenMP/irbuilder_for_iterator.cpp
index ec1c3af744b49a..0098a7db575c32 100644
--- a/clang/test/OpenMP/irbuilder_for_iterator.cpp
+++ b/clang/test/OpenMP/irbuilder_for_iterator.cpp
@@ -78,18 +78,18 @@ extern "C" void workshareloop_iterator(float *a, float *b, float *c) {
 // CHECK-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP8:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP8]] to i64
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP7]], i64 [[IDXPROM]]
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP7]], i64 [[IDXPROM]]
 // CHECK-NEXT:    [[TMP9:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK-NEXT:    [[TMP11:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM2:%.*]] = zext i32 [[TMP11]] to i64
-// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP10]], i64 [[IDXPROM2]]
+// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP10]], i64 [[IDXPROM2]]
 // CHECK-NEXT:    [[TMP12:%.*]] = load float, ptr [[ARRAYIDX3]], align 4
 // CHECK-NEXT:    [[MUL:%.*]] = fmul float [[TMP9]], [[TMP12]]
 // CHECK-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP14:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM4:%.*]] = zext i32 [[TMP14]] to i64
-// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i64 [[IDXPROM4]]
+// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i64 [[IDXPROM4]]
 // CHECK-NEXT:    store float [[MUL]], ptr [[ARRAYIDX5]], align 4
 // CHECK-NEXT:    br label [[OMP_LOOP_INC]]
 // CHECK:       omp_loop.inc:
diff --git a/clang/test/OpenMP/irbuilder_for_rangefor.cpp b/clang/test/OpenMP/irbuilder_for_rangefor.cpp
index 86a043e638bc3a..45b34621afbb95 100644
--- a/clang/test/OpenMP/irbuilder_for_rangefor.cpp
+++ b/clang/test/OpenMP/irbuilder_for_rangefor.cpp
@@ -94,18 +94,18 @@ extern "C" void workshareloop_rangefor(float *a, float *b, float *c) {
 // CHECK-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP11:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP11]] to i64
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP10]], i64 [[IDXPROM]]
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP10]], i64 [[IDXPROM]]
 // CHECK-NEXT:    [[TMP12:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK-NEXT:    [[TMP14:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM2:%.*]] = zext i32 [[TMP14]] to i64
-// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i64 [[IDXPROM2]]
+// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i64 [[IDXPROM2]]
 // CHECK-NEXT:    [[TMP15:%.*]] = load float, ptr [[ARRAYIDX3]], align 4
 // CHECK-NEXT:    [[MUL:%.*]] = fmul float [[TMP12]], [[TMP15]]
 // CHECK-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP17:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM4:%.*]] = zext i32 [[TMP17]] to i64
-// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM4]]
+// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM4]]
 // CHECK-NEXT:    store float [[MUL]], ptr [[ARRAYIDX5]], align 4
 // CHECK-NEXT:    br label [[OMP_LOOP_INC]]
 // CHECK:       omp_loop.inc:
diff --git a/clang/test/OpenMP/irbuilder_for_unsigned.c b/clang/test/OpenMP/irbuilder_for_unsigned.c
index 675871a87b3bd4..b0043b823ac85f 100644
--- a/clang/test/OpenMP/irbuilder_for_unsigned.c
+++ b/clang/test/OpenMP/irbuilder_for_unsigned.c
@@ -65,24 +65,24 @@ extern "C" void workshareloop_unsigned(float *a, float *b, float *c, float *d) {
 // CHECK-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP10:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP10]] to i64
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP9]], i64 [[IDXPROM]]
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP9]], i64 [[IDXPROM]]
 // CHECK-NEXT:    [[TMP11:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK-NEXT:    [[TMP13:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM2:%.*]] = zext i32 [[TMP13]] to i64
-// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP12]], i64 [[IDXPROM2]]
+// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP12]], i64 [[IDXPROM2]]
 // CHECK-NEXT:    [[TMP14:%.*]] = load float, ptr [[ARRAYIDX3]], align 4
 // CHECK-NEXT:    [[MUL:%.*]] = fmul float [[TMP11]], [[TMP14]]
 // CHECK-NEXT:    [[TMP15:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK-NEXT:    [[TMP16:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM4:%.*]] = zext i32 [[TMP16]] to i64
-// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP15]], i64 [[IDXPROM4]]
+// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP15]], i64 [[IDXPROM4]]
 // CHECK-NEXT:    [[TMP17:%.*]] = load float, ptr [[ARRAYIDX5]], align 4
 // CHECK-NEXT:    [[MUL6:%.*]] = fmul float [[MUL]], [[TMP17]]
 // CHECK-NEXT:    [[TMP18:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP19:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP19]] to i64
-// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP18]], i64 [[IDXPROM7]]
+// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP18]], i64 [[IDXPROM7]]
 // CHECK-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX8]], align 4
 // CHECK-NEXT:    br label [[OMP_LOOP_INC]]
 // CHECK:       omp_loop.inc:
diff --git a/clang/test/OpenMP/irbuilder_for_unsigned_auto.c b/clang/test/OpenMP/irbuilder_for_unsigned_auto.c
index 39ede3ef971d0b..19b2770bfa2df3 100644
--- a/clang/test/OpenMP/irbuilder_for_unsigned_auto.c
+++ b/clang/test/OpenMP/irbuilder_for_unsigned_auto.c
@@ -66,24 +66,24 @@ extern "C" void workshareloop_unsigned_auto(float *a, float *b, float *c, float
 // CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP4:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP4]] to i64
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP3]], i64 [[IDXPROM]]
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP3]], i64 [[IDXPROM]]
 // CHECK-NEXT:    [[TMP5:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK-NEXT:    [[TMP7:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM2:%.*]] = zext i32 [[TMP7]] to i64
-// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP6]], i64 [[IDXPROM2]]
+// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i64 [[IDXPROM2]]
 // CHECK-NEXT:    [[TMP8:%.*]] = load float, ptr [[ARRAYIDX3]], align 4
 // CHECK-NEXT:    [[MUL:%.*]] = fmul float [[TMP5]], [[TMP8]]
 // CHECK-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK-NEXT:    [[TMP10:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM4:%.*]] = zext i32 [[TMP10]] to i64
-// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP9]], i64 [[IDXPROM4]]
+// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP9]], i64 [[IDXPROM4]]
 // CHECK-NEXT:    [[TMP11:%.*]] = load float, ptr [[ARRAYIDX5]], align 4
 // CHECK-NEXT:    [[MUL6:%.*]] = fmul float [[MUL]], [[TMP11]]
 // CHECK-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP13:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP13]] to i64
-// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP12]], i64 [[IDXPROM7]]
+// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP12]], i64 [[IDXPROM7]]
 // CHECK-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX8]], align 4
 // CHECK-NEXT:    br label [[OMP_LOOP_INC]]
 // CHECK:       omp_loop.inc:
diff --git a/clang/test/OpenMP/irbuilder_for_unsigned_down.c b/clang/test/OpenMP/irbuilder_for_unsigned_down.c
index 5515f086c34a72..6e179826a6efa8 100644
--- a/clang/test/OpenMP/irbuilder_for_unsigned_down.c
+++ b/clang/test/OpenMP/irbuilder_for_unsigned_down.c
@@ -67,7 +67,7 @@ extern "C" void workshareloop_unsigned_down(float *a) {
 // CHECK-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP11:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP11]] to i64
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP10]], i64 [[IDXPROM]]
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP10]], i64 [[IDXPROM]]
 // CHECK-NEXT:    store float [[CONV]], ptr [[ARRAYIDX]], align 4
 // CHECK-NEXT:    br label [[OMP_LOOP_INC]]
 // CHECK:       omp_loop.inc:
diff --git a/clang/test/OpenMP/irbuilder_for_unsigned_dynamic.c b/clang/test/OpenMP/irbuilder_for_unsigned_dynamic.c
index f20b60e608d2f7..8f3297061938b3 100644
--- a/clang/test/OpenMP/irbuilder_for_unsigned_dynamic.c
+++ b/clang/test/OpenMP/irbuilder_for_unsigned_dynamic.c
@@ -66,24 +66,24 @@ extern "C" void workshareloop_unsigned_dynamic(float *a, float *b, float *c, flo
 // CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP4:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP4]] to i64
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP3]], i64 [[IDXPROM]]
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP3]], i64 [[IDXPROM]]
 // CHECK-NEXT:    [[TMP5:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK-NEXT:    [[TMP7:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM2:%.*]] = zext i32 [[TMP7]] to i64
-// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP6]], i64 [[IDXPROM2]]
+// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i64 [[IDXPROM2]]
 // CHECK-NEXT:    [[TMP8:%.*]] = load float, ptr [[ARRAYIDX3]], align 4
 // CHECK-NEXT:    [[MUL:%.*]] = fmul float [[TMP5]], [[TMP8]]
 // CHECK-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK-NEXT:    [[TMP10:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM4:%.*]] = zext i32 [[TMP10]] to i64
-// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP9]], i64 [[IDXPROM4]]
+// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP9]], i64 [[IDXPROM4]]
 // CHECK-NEXT:    [[TMP11:%.*]] = load float, ptr [[ARRAYIDX5]], align 4
 // CHECK-NEXT:    [[MUL6:%.*]] = fmul float [[MUL]], [[TMP11]]
 // CHECK-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP13:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP13]] to i64
-// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP12]], i64 [[IDXPROM7]]
+// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP12]], i64 [[IDXPROM7]]
 // CHECK-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX8]], align 4
 // CHECK-NEXT:    br label [[OMP_LOOP_INC]]
 // CHECK:       omp_loop.inc:
diff --git a/clang/test/OpenMP/irbuilder_for_unsigned_dynamic_chunked.c b/clang/test/OpenMP/irbuilder_for_unsigned_dynamic_chunked.c
index 599f256243b117..c2b0948bf7aeb0 100644
--- a/clang/test/OpenMP/irbuilder_for_unsigned_dynamic_chunked.c
+++ b/clang/test/OpenMP/irbuilder_for_unsigned_dynamic_chunked.c
@@ -66,24 +66,24 @@ extern "C" void workshareloop_unsigned_dynamic_chunked(float *a, float *b, float
 // CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP4:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP4]] to i64
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP3]], i64 [[IDXPROM]]
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP3]], i64 [[IDXPROM]]
 // CHECK-NEXT:    [[TMP5:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK-NEXT:    [[TMP7:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM2:%.*]] = zext i32 [[TMP7]] to i64
-// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP6]], i64 [[IDXPROM2]]
+// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i64 [[IDXPROM2]]
 // CHECK-NEXT:    [[TMP8:%.*]] = load float, ptr [[ARRAYIDX3]], align 4
 // CHECK-NEXT:    [[MUL:%.*]] = fmul float [[TMP5]], [[TMP8]]
 // CHECK-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK-NEXT:    [[TMP10:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM4:%.*]] = zext i32 [[TMP10]] to i64
-// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP9]], i64 [[IDXPROM4]]
+// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP9]], i64 [[IDXPROM4]]
 // CHECK-NEXT:    [[TMP11:%.*]] = load float, ptr [[ARRAYIDX5]], align 4
 // CHECK-NEXT:    [[MUL6:%.*]] = fmul float [[MUL]], [[TMP11]]
 // CHECK-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP13:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP13]] to i64
-// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP12]], i64 [[IDXPROM7]]
+// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP12]], i64 [[IDXPROM7]]
 // CHECK-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX8]], align 4
 // CHECK-NEXT:    br label [[OMP_LOOP_INC]]
 // CHECK:       omp_loop.inc:
diff --git a/clang/test/OpenMP/irbuilder_for_unsigned_runtime.c b/clang/test/OpenMP/irbuilder_for_unsigned_runtime.c
index c27bcba1559108..68becf9f694adb 100644
--- a/clang/test/OpenMP/irbuilder_for_unsigned_runtime.c
+++ b/clang/test/OpenMP/irbuilder_for_unsigned_runtime.c
@@ -66,24 +66,24 @@ extern "C" void workshareloop_unsigned_runtime(float *a, float *b, float *c, flo
 // CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP4:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP4]] to i64
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP3]], i64 [[IDXPROM]]
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP3]], i64 [[IDXPROM]]
 // CHECK-NEXT:    [[TMP5:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK-NEXT:    [[TMP7:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM2:%.*]] = zext i32 [[TMP7]] to i64
-// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP6]], i64 [[IDXPROM2]]
+// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i64 [[IDXPROM2]]
 // CHECK-NEXT:    [[TMP8:%.*]] = load float, ptr [[ARRAYIDX3]], align 4
 // CHECK-NEXT:    [[MUL:%.*]] = fmul float [[TMP5]], [[TMP8]]
 // CHECK-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK-NEXT:    [[TMP10:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM4:%.*]] = zext i32 [[TMP10]] to i64
-// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP9]], i64 [[IDXPROM4]]
+// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP9]], i64 [[IDXPROM4]]
 // CHECK-NEXT:    [[TMP11:%.*]] = load float, ptr [[ARRAYIDX5]], align 4
 // CHECK-NEXT:    [[MUL6:%.*]] = fmul float [[MUL]], [[TMP11]]
 // CHECK-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP13:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP13]] to i64
-// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP12]], i64 [[IDXPROM7]]
+// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP12]], i64 [[IDXPROM7]]
 // CHECK-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX8]], align 4
 // CHECK-NEXT:    br label [[OMP_LOOP_INC]]
 // CHECK:       omp_loop.inc:
diff --git a/clang/test/OpenMP/irbuilder_for_unsigned_static_chunked.c b/clang/test/OpenMP/irbuilder_for_unsigned_static_chunked.c
index b937568ca9f118..71fb6b5473da8b 100644
--- a/clang/test/OpenMP/irbuilder_for_unsigned_static_chunked.c
+++ b/clang/test/OpenMP/irbuilder_for_unsigned_static_chunked.c
@@ -108,24 +108,24 @@ extern "C" void workshareloop_unsigned_static_chunked(float *a, float *b, float
 // CHECK-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP18]] to i64
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM]]
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM]]
 // CHECK-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM2:%.*]] = zext i32 [[TMP21]] to i64
-// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM2]]
+// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM2]]
 // CHECK-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX3]], align 4
 // CHECK-NEXT:    [[MUL:%.*]] = fmul float [[TMP19]], [[TMP22]]
 // CHECK-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM4:%.*]] = zext i32 [[TMP24]] to i64
-// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM4]]
+// CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM4]]
 // CHECK-NEXT:    [[TMP25:%.*]] = load float, ptr [[ARRAYIDX5]], align 4
 // CHECK-NEXT:    [[MUL6:%.*]] = fmul float [[MUL]], [[TMP25]]
 // CHECK-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK-NEXT:    [[TMP27:%.*]] = load i32, ptr [[I]], align 4
 // CHECK-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP27]] to i64
-// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP26]], i64 [[IDXPROM7]]
+// CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP26]], i64 [[IDXPROM7]]
 // CHECK-NEXT:    store float [[MUL6]], ptr [[ARRAYIDX8]], align 4
 // CHECK-NEXT:    br label [[OMP_LOOP_INC]]
 // CHECK:       omp_loop.inc:
diff --git a/clang/test/OpenMP/map_struct_ordering.cpp b/clang/test/OpenMP/map_struct_ordering.cpp
index a52ddad465f37b..d5b22d8ff2a4d4 100644
--- a/clang/test/OpenMP/map_struct_ordering.cpp
+++ b/clang/test/OpenMP/map_struct_ordering.cpp
@@ -57,7 +57,7 @@ int map_struct() {
 // CHECK-NEXT:    [[DATUM:%.*]] = getelementptr inbounds nuw [[STRUCT_DESCRIPTOR]], ptr [[DAT]], i32 0, i32 0
 // CHECK-NEXT:    [[DATUM2:%.*]] = getelementptr inbounds nuw [[STRUCT_DESCRIPTOR]], ptr [[DAT]], i32 0, i32 0
 // CHECK-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[DATUM2]], align 8
-// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP0]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds i32, ptr [[TMP0]], i64 0
 // CHECK-NEXT:    [[TMP1:%.*]] = getelementptr [[STRUCT_DESCRIPTOR]], ptr [[DAT]], i32 1
 // CHECK-NEXT:    [[TMP2:%.*]] = ptrtoint ptr [[TMP1]] to i64
 // CHECK-NEXT:    [[TMP3:%.*]] = ptrtoint ptr [[DAT]] to i64
diff --git a/clang/test/OpenMP/master_taskloop_in_reduction_codegen.cpp b/clang/test/OpenMP/master_taskloop_in_reduction_codegen.cpp
index 7d467293d0c8f1..e90f0783787c05 100644
--- a/clang/test/OpenMP/master_taskloop_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/master_taskloop_in_reduction_codegen.cpp
@@ -76,7 +76,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[VLA:%.*]] = alloca i16, i64 [[TMP2]], align 16
 // CHECK1-NEXT:    store i64 [[TMP2]], ptr [[__VLA_EXPR0]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[A]], ptr [[TMP4]], align 8
 // CHECK1-NEXT:    [[TMP5:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -91,7 +91,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP9]], align 8
 // CHECK1-NEXT:    [[TMP10:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP10]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP11:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[B]], ptr [[TMP11]], align 8
 // CHECK1-NEXT:    [[TMP12:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 1
@@ -106,7 +106,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..2, ptr [[TMP16]], align 8
 // CHECK1-NEXT:    [[TMP17:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP17]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
 // CHECK1-NEXT:    [[TMP18:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC_ADDR]], ptr [[TMP18]], align 8
 // CHECK1-NEXT:    [[TMP19:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 1
@@ -124,7 +124,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP25:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0]], i32 3, ptr [[DOTRD_INPUT_]])
 // CHECK1-NEXT:    store ptr [[TMP25]], ptr [[DOTTASK_RED_]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP26:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0:%.*]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[C]], ptr [[TMP26]], align 8
 // CHECK1-NEXT:    [[TMP27:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 1
@@ -139,7 +139,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..6, ptr [[TMP31]], align 8
 // CHECK1-NEXT:    [[TMP32:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP32]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP33:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP33]], align 8
 // CHECK1-NEXT:    [[TMP34:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 1
diff --git a/clang/test/OpenMP/master_taskloop_reduction_codegen.cpp b/clang/test/OpenMP/master_taskloop_reduction_codegen.cpp
index b0652c843845c7..180ff3a94d24ca 100644
--- a/clang/test/OpenMP/master_taskloop_reduction_codegen.cpp
+++ b/clang/test/OpenMP/master_taskloop_reduction_codegen.cpp
@@ -84,9 +84,9 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB1:.+]], ptr [[TMP25]],
 // CHECK-DAG:    [[TMP26:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK-DAG:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP26]], i8 0, i64 4, i1 false)
-// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C]], i64 0, i64 0
+// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C]], i64 0, i64 0
 // CHECK-DAG:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, %
-// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
+// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
 // CHECK-DAG:    [[TMP28]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_4:%.+]], i32 0, i32 0
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
@@ -138,10 +138,10 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB4:.+]], ptr [[TMP59]],
 // CHECK-DAG:    [[TMP60:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_8]], i32 0, i32 6
 // CHECK-DAG:    store i32 1, ptr [[TMP60]],
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
 // CHECK:    [[TMP62:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0]], i32 4, ptr [[DOTRD_INPUT_]])
 // CHECK:    [[TMP63:%.*]] = load i32, ptr [[N]],
 // CHECK:    store i32 [[TMP63]], ptr [[DOTCAPTURE_EXPR_]],
diff --git a/clang/test/OpenMP/master_taskloop_simd_in_reduction_codegen.cpp b/clang/test/OpenMP/master_taskloop_simd_in_reduction_codegen.cpp
index b0d00c5f539b17..2061da7c4d781c 100644
--- a/clang/test/OpenMP/master_taskloop_simd_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/master_taskloop_simd_in_reduction_codegen.cpp
@@ -76,7 +76,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[VLA:%.*]] = alloca i16, i64 [[TMP2]], align 16
 // CHECK1-NEXT:    store i64 [[TMP2]], ptr [[__VLA_EXPR0]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[A]], ptr [[TMP4]], align 8
 // CHECK1-NEXT:    [[TMP5:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -91,7 +91,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP9]], align 8
 // CHECK1-NEXT:    [[TMP10:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP10]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP11:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[B]], ptr [[TMP11]], align 8
 // CHECK1-NEXT:    [[TMP12:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 1
@@ -106,7 +106,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..2, ptr [[TMP16]], align 8
 // CHECK1-NEXT:    [[TMP17:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP17]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
 // CHECK1-NEXT:    [[TMP18:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC_ADDR]], ptr [[TMP18]], align 8
 // CHECK1-NEXT:    [[TMP19:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 1
@@ -124,7 +124,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP25:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0]], i32 3, ptr [[DOTRD_INPUT_]])
 // CHECK1-NEXT:    store ptr [[TMP25]], ptr [[DOTTASK_RED_]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP26:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0:%.*]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[C]], ptr [[TMP26]], align 8
 // CHECK1-NEXT:    [[TMP27:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 1
@@ -139,7 +139,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..6, ptr [[TMP31]], align 8
 // CHECK1-NEXT:    [[TMP32:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP32]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP33:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP33]], align 8
 // CHECK1-NEXT:    [[TMP34:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 1
diff --git a/clang/test/OpenMP/master_taskloop_simd_reduction_codegen.cpp b/clang/test/OpenMP/master_taskloop_simd_reduction_codegen.cpp
index 7def61251b24ea..a69844dc4dee26 100644
--- a/clang/test/OpenMP/master_taskloop_simd_reduction_codegen.cpp
+++ b/clang/test/OpenMP/master_taskloop_simd_reduction_codegen.cpp
@@ -80,9 +80,9 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB1:.+]], ptr [[TMP25]],
 // CHECK-DAG:    [[TMP26:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK-DAG:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP26]], i8 0, i64 4, i1 false)
-// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C]], i64 0, i64 0
+// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C]], i64 0, i64 0
 // CHECK-DAG:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, %
-// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
+// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
 // CHECK-DAG:    [[TMP28]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_4:%.+]], i32 0, i32 0
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
@@ -134,10 +134,10 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB4:.+]], ptr [[TMP59]],
 // CHECK-DAG:    [[TMP60:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_8]], i32 0, i32 6
 // CHECK-DAG:    store i32 1, ptr [[TMP60]],
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
 // CHECK:    [[TMP62:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0]], i32 4, ptr [[DOTRD_INPUT_]])
 // CHECK:    [[TMP63:%.*]] = load i32, ptr [[N]],
 // CHECK:    store i32 [[TMP63]], ptr [[DOTCAPTURE_EXPR_]],
diff --git a/clang/test/OpenMP/ordered_codegen.cpp b/clang/test/OpenMP/ordered_codegen.cpp
index 67285cfaef34d5..0a73eaefc98085 100644
--- a/clang/test/OpenMP/ordered_codegen.cpp
+++ b/clang/test/OpenMP/ordered_codegen.cpp
@@ -255,21 +255,21 @@ void foo_simd(int low, int up) {
 // CHECK1-NEXT:    call void @__kmpc_ordered(ptr @[[GLOB1]], i32 [[TMP0]])
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK1-NEXT:    [[TMP7:%.*]] = load i64, ptr [[I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP6]], i64 [[TMP7]]
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i64 [[TMP7]]
 // CHECK1-NEXT:    [[TMP8:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK1-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK1-NEXT:    [[TMP10:%.*]] = load i64, ptr [[I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP9]], i64 [[TMP10]]
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP9]], i64 [[TMP10]]
 // CHECK1-NEXT:    [[TMP11:%.*]] = load float, ptr [[ARRAYIDX2]], align 4
 // CHECK1-NEXT:    [[MUL3:%.*]] = fmul float [[TMP8]], [[TMP11]]
 // CHECK1-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK1-NEXT:    [[TMP13:%.*]] = load i64, ptr [[I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP12]], i64 [[TMP13]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP12]], i64 [[TMP13]]
 // CHECK1-NEXT:    [[TMP14:%.*]] = load float, ptr [[ARRAYIDX4]], align 4
 // CHECK1-NEXT:    [[MUL5:%.*]] = fmul float [[MUL3]], [[TMP14]]
 // CHECK1-NEXT:    [[TMP15:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK1-NEXT:    [[TMP16:%.*]] = load i64, ptr [[I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP15]], i64 [[TMP16]]
+// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP15]], i64 [[TMP16]]
 // CHECK1-NEXT:    store float [[MUL5]], ptr [[ARRAYIDX6]], align 4
 // CHECK1-NEXT:    call void @__kmpc_end_ordered(ptr @[[GLOB1]], i32 [[TMP0]])
 // CHECK1-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
@@ -485,24 +485,24 @@ void foo_simd(int low, int up) {
 // CHECK1-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK1-NEXT:    [[TMP9:%.*]] = load i8, ptr [[I]], align 1
 // CHECK1-NEXT:    [[IDXPROM:%.*]] = zext i8 [[TMP9]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP8]], i64 [[IDXPROM]]
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i64 [[IDXPROM]]
 // CHECK1-NEXT:    [[TMP10:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK1-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK1-NEXT:    [[TMP12:%.*]] = load i8, ptr [[I]], align 1
 // CHECK1-NEXT:    [[IDXPROM7:%.*]] = zext i8 [[TMP12]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[IDXPROM7]]
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[IDXPROM7]]
 // CHECK1-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX8]], align 4
 // CHECK1-NEXT:    [[MUL9:%.*]] = fmul float [[TMP10]], [[TMP13]]
 // CHECK1-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK1-NEXT:    [[TMP15:%.*]] = load i8, ptr [[I]], align 1
 // CHECK1-NEXT:    [[IDXPROM10:%.*]] = zext i8 [[TMP15]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM10]]
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM10]]
 // CHECK1-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX11]], align 4
 // CHECK1-NEXT:    [[MUL12:%.*]] = fmul float [[MUL9]], [[TMP16]]
 // CHECK1-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK1-NEXT:    [[TMP18:%.*]] = load i8, ptr [[I]], align 1
 // CHECK1-NEXT:    [[IDXPROM13:%.*]] = zext i8 [[TMP18]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM13]]
+// CHECK1-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM13]]
 // CHECK1-NEXT:    store float [[MUL12]], ptr [[ARRAYIDX14]], align 4
 // CHECK1-NEXT:    call void @__kmpc_end_ordered(ptr @[[GLOB1]], i32 [[TMP0]])
 // CHECK1-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
@@ -866,21 +866,21 @@ void foo_simd(int low, int up) {
 // CHECK1-IRBUILDER-NEXT:    call void @__kmpc_ordered(ptr @[[GLOB1]], i32 [[OMP_GLOBAL_THREAD_NUM3]])
 // CHECK1-IRBUILDER-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK1-IRBUILDER-NEXT:    [[TMP6:%.*]] = load i64, ptr [[I]], align 8
-// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP5]], i64 [[TMP6]]
+// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP5]], i64 [[TMP6]]
 // CHECK1-IRBUILDER-NEXT:    [[TMP7:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK1-IRBUILDER-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK1-IRBUILDER-NEXT:    [[TMP9:%.*]] = load i64, ptr [[I]], align 8
-// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP8]], i64 [[TMP9]]
+// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i64 [[TMP9]]
 // CHECK1-IRBUILDER-NEXT:    [[TMP10:%.*]] = load float, ptr [[ARRAYIDX4]], align 4
 // CHECK1-IRBUILDER-NEXT:    [[MUL5:%.*]] = fmul float [[TMP7]], [[TMP10]]
 // CHECK1-IRBUILDER-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK1-IRBUILDER-NEXT:    [[TMP12:%.*]] = load i64, ptr [[I]], align 8
-// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[TMP12]]
+// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[TMP12]]
 // CHECK1-IRBUILDER-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX6]], align 4
 // CHECK1-IRBUILDER-NEXT:    [[MUL7:%.*]] = fmul float [[MUL5]], [[TMP13]]
 // CHECK1-IRBUILDER-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK1-IRBUILDER-NEXT:    [[TMP15:%.*]] = load i64, ptr [[I]], align 8
-// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[TMP15]]
+// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[TMP15]]
 // CHECK1-IRBUILDER-NEXT:    store float [[MUL7]], ptr [[ARRAYIDX8]], align 4
 // CHECK1-IRBUILDER-NEXT:    br label [[OMP_INNER_FOR_BODY_ORDERED_AFTER:%.*]]
 // CHECK1-IRBUILDER:       omp.inner.for.body.ordered.after:
@@ -1110,24 +1110,24 @@ void foo_simd(int low, int up) {
 // CHECK1-IRBUILDER-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK1-IRBUILDER-NEXT:    [[TMP8:%.*]] = load i8, ptr [[I]], align 1
 // CHECK1-IRBUILDER-NEXT:    [[IDXPROM:%.*]] = zext i8 [[TMP8]] to i64
-// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP7]], i64 [[IDXPROM]]
+// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP7]], i64 [[IDXPROM]]
 // CHECK1-IRBUILDER-NEXT:    [[TMP9:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK1-IRBUILDER-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK1-IRBUILDER-NEXT:    [[TMP11:%.*]] = load i8, ptr [[I]], align 1
 // CHECK1-IRBUILDER-NEXT:    [[IDXPROM9:%.*]] = zext i8 [[TMP11]] to i64
-// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP10]], i64 [[IDXPROM9]]
+// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP10]], i64 [[IDXPROM9]]
 // CHECK1-IRBUILDER-NEXT:    [[TMP12:%.*]] = load float, ptr [[ARRAYIDX10]], align 4
 // CHECK1-IRBUILDER-NEXT:    [[MUL11:%.*]] = fmul float [[TMP9]], [[TMP12]]
 // CHECK1-IRBUILDER-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK1-IRBUILDER-NEXT:    [[TMP14:%.*]] = load i8, ptr [[I]], align 1
 // CHECK1-IRBUILDER-NEXT:    [[IDXPROM12:%.*]] = zext i8 [[TMP14]] to i64
-// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i64 [[IDXPROM12]]
+// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i64 [[IDXPROM12]]
 // CHECK1-IRBUILDER-NEXT:    [[TMP15:%.*]] = load float, ptr [[ARRAYIDX13]], align 4
 // CHECK1-IRBUILDER-NEXT:    [[MUL14:%.*]] = fmul float [[MUL11]], [[TMP15]]
 // CHECK1-IRBUILDER-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK1-IRBUILDER-NEXT:    [[TMP17:%.*]] = load i8, ptr [[I]], align 1
 // CHECK1-IRBUILDER-NEXT:    [[IDXPROM15:%.*]] = zext i8 [[TMP17]] to i64
-// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM15]]
+// CHECK1-IRBUILDER-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM15]]
 // CHECK1-IRBUILDER-NEXT:    store float [[MUL14]], ptr [[ARRAYIDX16]], align 4
 // CHECK1-IRBUILDER-NEXT:    br label [[OMP_INNER_FOR_BODY_ORDERED_AFTER:%.*]]
 // CHECK1-IRBUILDER:       omp.inner.for.body.ordered.after:
@@ -1495,21 +1495,21 @@ void foo_simd(int low, int up) {
 // CHECK3-NEXT:    call void @__kmpc_ordered(ptr @[[GLOB1]], i32 [[TMP0]])
 // CHECK3-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK3-NEXT:    [[TMP7:%.*]] = load i64, ptr [[I]], align 8
-// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP6]], i64 [[TMP7]]
+// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i64 [[TMP7]]
 // CHECK3-NEXT:    [[TMP8:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK3-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK3-NEXT:    [[TMP10:%.*]] = load i64, ptr [[I]], align 8
-// CHECK3-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP9]], i64 [[TMP10]]
+// CHECK3-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP9]], i64 [[TMP10]]
 // CHECK3-NEXT:    [[TMP11:%.*]] = load float, ptr [[ARRAYIDX2]], align 4
 // CHECK3-NEXT:    [[MUL3:%.*]] = fmul float [[TMP8]], [[TMP11]]
 // CHECK3-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK3-NEXT:    [[TMP13:%.*]] = load i64, ptr [[I]], align 8
-// CHECK3-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP12]], i64 [[TMP13]]
+// CHECK3-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP12]], i64 [[TMP13]]
 // CHECK3-NEXT:    [[TMP14:%.*]] = load float, ptr [[ARRAYIDX4]], align 4
 // CHECK3-NEXT:    [[MUL5:%.*]] = fmul float [[MUL3]], [[TMP14]]
 // CHECK3-NEXT:    [[TMP15:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK3-NEXT:    [[TMP16:%.*]] = load i64, ptr [[I]], align 8
-// CHECK3-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP15]], i64 [[TMP16]]
+// CHECK3-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP15]], i64 [[TMP16]]
 // CHECK3-NEXT:    store float [[MUL5]], ptr [[ARRAYIDX6]], align 4
 // CHECK3-NEXT:    call void @__kmpc_end_ordered(ptr @[[GLOB1]], i32 [[TMP0]])
 // CHECK3-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
@@ -1725,24 +1725,24 @@ void foo_simd(int low, int up) {
 // CHECK3-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK3-NEXT:    [[TMP9:%.*]] = load i8, ptr [[I]], align 1
 // CHECK3-NEXT:    [[IDXPROM:%.*]] = zext i8 [[TMP9]] to i64
-// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP8]], i64 [[IDXPROM]]
+// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i64 [[IDXPROM]]
 // CHECK3-NEXT:    [[TMP10:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK3-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK3-NEXT:    [[TMP12:%.*]] = load i8, ptr [[I]], align 1
 // CHECK3-NEXT:    [[IDXPROM7:%.*]] = zext i8 [[TMP12]] to i64
-// CHECK3-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[IDXPROM7]]
+// CHECK3-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[IDXPROM7]]
 // CHECK3-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX8]], align 4
 // CHECK3-NEXT:    [[MUL9:%.*]] = fmul float [[TMP10]], [[TMP13]]
 // CHECK3-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK3-NEXT:    [[TMP15:%.*]] = load i8, ptr [[I]], align 1
 // CHECK3-NEXT:    [[IDXPROM10:%.*]] = zext i8 [[TMP15]] to i64
-// CHECK3-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM10]]
+// CHECK3-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM10]]
 // CHECK3-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX11]], align 4
 // CHECK3-NEXT:    [[MUL12:%.*]] = fmul float [[MUL9]], [[TMP16]]
 // CHECK3-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK3-NEXT:    [[TMP18:%.*]] = load i8, ptr [[I]], align 1
 // CHECK3-NEXT:    [[IDXPROM13:%.*]] = zext i8 [[TMP18]] to i64
-// CHECK3-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM13]]
+// CHECK3-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM13]]
 // CHECK3-NEXT:    store float [[MUL12]], ptr [[ARRAYIDX14]], align 4
 // CHECK3-NEXT:    call void @__kmpc_end_ordered(ptr @[[GLOB1]], i32 [[TMP0]])
 // CHECK3-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
@@ -2106,21 +2106,21 @@ void foo_simd(int low, int up) {
 // CHECK3-IRBUILDER-NEXT:    call void @__kmpc_ordered(ptr @[[GLOB1]], i32 [[OMP_GLOBAL_THREAD_NUM3]])
 // CHECK3-IRBUILDER-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK3-IRBUILDER-NEXT:    [[TMP6:%.*]] = load i64, ptr [[I]], align 8
-// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP5]], i64 [[TMP6]]
+// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP5]], i64 [[TMP6]]
 // CHECK3-IRBUILDER-NEXT:    [[TMP7:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK3-IRBUILDER-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK3-IRBUILDER-NEXT:    [[TMP9:%.*]] = load i64, ptr [[I]], align 8
-// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP8]], i64 [[TMP9]]
+// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i64 [[TMP9]]
 // CHECK3-IRBUILDER-NEXT:    [[TMP10:%.*]] = load float, ptr [[ARRAYIDX4]], align 4
 // CHECK3-IRBUILDER-NEXT:    [[MUL5:%.*]] = fmul float [[TMP7]], [[TMP10]]
 // CHECK3-IRBUILDER-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK3-IRBUILDER-NEXT:    [[TMP12:%.*]] = load i64, ptr [[I]], align 8
-// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[TMP12]]
+// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[TMP12]]
 // CHECK3-IRBUILDER-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX6]], align 4
 // CHECK3-IRBUILDER-NEXT:    [[MUL7:%.*]] = fmul float [[MUL5]], [[TMP13]]
 // CHECK3-IRBUILDER-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK3-IRBUILDER-NEXT:    [[TMP15:%.*]] = load i64, ptr [[I]], align 8
-// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[TMP15]]
+// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[TMP15]]
 // CHECK3-IRBUILDER-NEXT:    store float [[MUL7]], ptr [[ARRAYIDX8]], align 4
 // CHECK3-IRBUILDER-NEXT:    br label [[OMP_INNER_FOR_BODY_ORDERED_AFTER:%.*]]
 // CHECK3-IRBUILDER:       omp.inner.for.body.ordered.after:
@@ -2350,24 +2350,24 @@ void foo_simd(int low, int up) {
 // CHECK3-IRBUILDER-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK3-IRBUILDER-NEXT:    [[TMP8:%.*]] = load i8, ptr [[I]], align 1
 // CHECK3-IRBUILDER-NEXT:    [[IDXPROM:%.*]] = zext i8 [[TMP8]] to i64
-// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP7]], i64 [[IDXPROM]]
+// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP7]], i64 [[IDXPROM]]
 // CHECK3-IRBUILDER-NEXT:    [[TMP9:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK3-IRBUILDER-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK3-IRBUILDER-NEXT:    [[TMP11:%.*]] = load i8, ptr [[I]], align 1
 // CHECK3-IRBUILDER-NEXT:    [[IDXPROM9:%.*]] = zext i8 [[TMP11]] to i64
-// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP10]], i64 [[IDXPROM9]]
+// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP10]], i64 [[IDXPROM9]]
 // CHECK3-IRBUILDER-NEXT:    [[TMP12:%.*]] = load float, ptr [[ARRAYIDX10]], align 4
 // CHECK3-IRBUILDER-NEXT:    [[MUL11:%.*]] = fmul float [[TMP9]], [[TMP12]]
 // CHECK3-IRBUILDER-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK3-IRBUILDER-NEXT:    [[TMP14:%.*]] = load i8, ptr [[I]], align 1
 // CHECK3-IRBUILDER-NEXT:    [[IDXPROM12:%.*]] = zext i8 [[TMP14]] to i64
-// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i64 [[IDXPROM12]]
+// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i64 [[IDXPROM12]]
 // CHECK3-IRBUILDER-NEXT:    [[TMP15:%.*]] = load float, ptr [[ARRAYIDX13]], align 4
 // CHECK3-IRBUILDER-NEXT:    [[MUL14:%.*]] = fmul float [[MUL11]], [[TMP15]]
 // CHECK3-IRBUILDER-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK3-IRBUILDER-NEXT:    [[TMP17:%.*]] = load i8, ptr [[I]], align 1
 // CHECK3-IRBUILDER-NEXT:    [[IDXPROM15:%.*]] = zext i8 [[TMP17]] to i64
-// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM15]]
+// CHECK3-IRBUILDER-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM15]]
 // CHECK3-IRBUILDER-NEXT:    store float [[MUL14]], ptr [[ARRAYIDX16]], align 4
 // CHECK3-IRBUILDER-NEXT:    br label [[OMP_INNER_FOR_BODY_ORDERED_AFTER:%.*]]
 // CHECK3-IRBUILDER:       omp.inner.for.body.ordered.after:
@@ -2674,21 +2674,21 @@ void foo_simd(int low, int up) {
 // CHECK5:       for.body:
 // CHECK5-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK5-NEXT:    [[TMP2:%.*]] = load i64, ptr [[I]], align 8
-// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP1]], i64 [[TMP2]]
+// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i64 [[TMP2]]
 // CHECK5-NEXT:    [[TMP3:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK5-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK5-NEXT:    [[TMP5:%.*]] = load i64, ptr [[I]], align 8
-// CHECK5-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw float, ptr [[TMP4]], i64 [[TMP5]]
+// CHECK5-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds float, ptr [[TMP4]], i64 [[TMP5]]
 // CHECK5-NEXT:    [[TMP6:%.*]] = load float, ptr [[ARRAYIDX1]], align 4
 // CHECK5-NEXT:    [[MUL:%.*]] = fmul float [[TMP3]], [[TMP6]]
 // CHECK5-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK5-NEXT:    [[TMP8:%.*]] = load i64, ptr [[I]], align 8
-// CHECK5-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP7]], i64 [[TMP8]]
+// CHECK5-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP7]], i64 [[TMP8]]
 // CHECK5-NEXT:    [[TMP9:%.*]] = load float, ptr [[ARRAYIDX2]], align 4
 // CHECK5-NEXT:    [[MUL3:%.*]] = fmul float [[MUL]], [[TMP9]]
 // CHECK5-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK5-NEXT:    [[TMP11:%.*]] = load i64, ptr [[I]], align 8
-// CHECK5-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP10]], i64 [[TMP11]]
+// CHECK5-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP10]], i64 [[TMP11]]
 // CHECK5-NEXT:    store float [[MUL3]], ptr [[ARRAYIDX4]], align 4
 // CHECK5-NEXT:    br label [[FOR_INC:%.*]]
 // CHECK5:       for.inc:
@@ -2804,24 +2804,24 @@ void foo_simd(int low, int up) {
 // CHECK5-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK5-NEXT:    [[TMP3:%.*]] = load i8, ptr [[I]], align 1
 // CHECK5-NEXT:    [[IDXPROM:%.*]] = zext i8 [[TMP3]] to i64
-// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP2]], i64 [[IDXPROM]]
+// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP2]], i64 [[IDXPROM]]
 // CHECK5-NEXT:    [[TMP4:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK5-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[C_ADDR]], align 8
 // CHECK5-NEXT:    [[TMP6:%.*]] = load i8, ptr [[I]], align 1
 // CHECK5-NEXT:    [[IDXPROM4:%.*]] = zext i8 [[TMP6]] to i64
-// CHECK5-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw float, ptr [[TMP5]], i64 [[IDXPROM4]]
+// CHECK5-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[TMP5]], i64 [[IDXPROM4]]
 // CHECK5-NEXT:    [[TMP7:%.*]] = load float, ptr [[ARRAYIDX5]], align 4
 // CHECK5-NEXT:    [[MUL:%.*]] = fmul float [[TMP4]], [[TMP7]]
 // CHECK5-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK5-NEXT:    [[TMP9:%.*]] = load i8, ptr [[I]], align 1
 // CHECK5-NEXT:    [[IDXPROM6:%.*]] = zext i8 [[TMP9]] to i64
-// CHECK5-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP8]], i64 [[IDXPROM6]]
+// CHECK5-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i64 [[IDXPROM6]]
 // CHECK5-NEXT:    [[TMP10:%.*]] = load float, ptr [[ARRAYIDX7]], align 4
 // CHECK5-NEXT:    [[MUL8:%.*]] = fmul float [[MUL]], [[TMP10]]
 // CHECK5-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[A_ADDR]], align 8
 // CHECK5-NEXT:    [[TMP12:%.*]] = load i8, ptr [[I]], align 1
 // CHECK5-NEXT:    [[IDXPROM9:%.*]] = zext i8 [[TMP12]] to i64
-// CHECK5-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[IDXPROM9]]
+// CHECK5-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[IDXPROM9]]
 // CHECK5-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4
 // CHECK5-NEXT:    br label [[FOR_INC:%.*]]
 // CHECK5:       for.inc:
diff --git a/clang/test/OpenMP/parallel_for_codegen.cpp b/clang/test/OpenMP/parallel_for_codegen.cpp
index c7afae419509b7..2dec32d71b91a1 100644
--- a/clang/test/OpenMP/parallel_for_codegen.cpp
+++ b/clang/test/OpenMP/parallel_for_codegen.cpp
@@ -665,24 +665,24 @@ void range_for_collapsed() {
 // CHECK1-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 8
 // CHECK1-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4
 // CHECK1-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP15]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM]]
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM]]
 // CHECK1-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK1-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 8
 // CHECK1-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4
 // CHECK1-NEXT:    [[IDXPROM3:%.*]] = zext i32 [[TMP18]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM3]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM3]]
 // CHECK1-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4
 // CHECK1-NEXT:    [[MUL5:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK1-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 8
 // CHECK1-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4
 // CHECK1-NEXT:    [[IDXPROM6:%.*]] = zext i32 [[TMP21]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM6]]
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM6]]
 // CHECK1-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX7]], align 4
 // CHECK1-NEXT:    [[MUL8:%.*]] = fmul float [[MUL5]], [[TMP22]]
 // CHECK1-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 8
 // CHECK1-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4
 // CHECK1-NEXT:    [[IDXPROM9:%.*]] = zext i32 [[TMP24]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM9]]
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM9]]
 // CHECK1-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4
 // CHECK1-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK1:       omp.body.continue:
@@ -779,21 +779,21 @@ void range_for_collapsed() {
 // CHECK1-NEXT:    store i64 [[ADD1]], ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK1-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK1-NEXT:    [[TMP12:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[TMP12]]
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[TMP12]]
 // CHECK1-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK1-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK1-NEXT:    [[TMP15:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[TMP15]]
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[TMP15]]
 // CHECK1-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX2]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK1-NEXT:    [[MUL3:%.*]] = fmul float [[TMP13]], [[TMP16]]
 // CHECK1-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK1-NEXT:    [[TMP18:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[TMP18]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[TMP18]]
 // CHECK1-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK1-NEXT:    [[MUL5:%.*]] = fmul float [[MUL3]], [[TMP19]]
 // CHECK1-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK1-NEXT:    [[TMP21:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[TMP21]]
+// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[TMP21]]
 // CHECK1-NEXT:    store float [[MUL5]], ptr [[ARRAYIDX6]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK1-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK1:       omp.body.continue:
@@ -882,21 +882,21 @@ void range_for_collapsed() {
 // CHECK1-NEXT:    store i64 [[ADD1]], ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK1-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK1-NEXT:    [[TMP12:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[TMP12]]
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[TMP12]]
 // CHECK1-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK1-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK1-NEXT:    [[TMP15:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[TMP15]]
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[TMP15]]
 // CHECK1-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX2]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK1-NEXT:    [[MUL3:%.*]] = fmul float [[TMP13]], [[TMP16]]
 // CHECK1-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK1-NEXT:    [[TMP18:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[TMP18]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[TMP18]]
 // CHECK1-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK1-NEXT:    [[MUL5:%.*]] = fmul float [[MUL3]], [[TMP19]]
 // CHECK1-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK1-NEXT:    [[TMP21:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[TMP21]]
+// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[TMP21]]
 // CHECK1-NEXT:    store float [[MUL5]], ptr [[ARRAYIDX6]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK1-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK1:       omp.body.continue:
@@ -1159,24 +1159,24 @@ void range_for_collapsed() {
 // CHECK1-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[TMP14:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[IDXPROM:%.*]] = zext i8 [[TMP14]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i64 [[IDXPROM]]
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i64 [[IDXPROM]]
 // CHECK1-NEXT:    [[TMP15:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[TMP17:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[IDXPROM6:%.*]] = zext i8 [[TMP17]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM6]]
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM6]]
 // CHECK1-NEXT:    [[TMP18:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[MUL8:%.*]] = fmul float [[TMP15]], [[TMP18]]
 // CHECK1-NEXT:    [[TMP19:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[TMP20:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[IDXPROM9:%.*]] = zext i8 [[TMP20]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP19]], i64 [[IDXPROM9]]
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP19]], i64 [[IDXPROM9]]
 // CHECK1-NEXT:    [[TMP21:%.*]] = load float, ptr [[ARRAYIDX10]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[MUL11:%.*]] = fmul float [[MUL8]], [[TMP21]]
 // CHECK1-NEXT:    [[TMP22:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[TMP23:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    [[IDXPROM12:%.*]] = zext i8 [[TMP23]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw float, ptr [[TMP22]], i64 [[IDXPROM12]]
+// CHECK1-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds float, ptr [[TMP22]], i64 [[IDXPROM12]]
 // CHECK1-NEXT:    store float [[MUL11]], ptr [[ARRAYIDX13]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK1-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK1:       omp.body.continue:
@@ -1303,7 +1303,7 @@ void range_for_collapsed() {
 // CHECK1-NEXT:    [[CONV:%.*]] = sitofp i32 [[CALL]] to float
 // CHECK1-NEXT:    [[TMP13:%.*]] = load i32, ptr [[I]], align 4
 // CHECK1-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP13]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[VLA1]], i64 [[IDXPROM]]
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[VLA1]], i64 [[IDXPROM]]
 // CHECK1-NEXT:    [[TMP14:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK1-NEXT:    [[ADD4:%.*]] = fadd float [[CONV]], [[TMP14]]
 // CHECK1-NEXT:    [[TMP15:%.*]] = load i32, ptr [[N_ADDR]], align 4
@@ -1312,7 +1312,7 @@ void range_for_collapsed() {
 // CHECK1-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[TMP0]], align 8
 // CHECK1-NEXT:    [[TMP17:%.*]] = load i32, ptr [[I]], align 4
 // CHECK1-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP17]] to i64
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM7]]
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM7]]
 // CHECK1-NEXT:    [[TMP18:%.*]] = load float, ptr [[ARRAYIDX8]], align 4
 // CHECK1-NEXT:    [[ADD9:%.*]] = fadd float [[TMP18]], [[ADD6]]
 // CHECK1-NEXT:    store float [[ADD9]], ptr [[ARRAYIDX8]], align 4
@@ -1781,24 +1781,24 @@ void range_for_collapsed() {
 // CHECK2-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 8
 // CHECK2-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4
 // CHECK2-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP15]] to i64
-// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM]]
+// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM]]
 // CHECK2-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK2-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 8
 // CHECK2-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4
 // CHECK2-NEXT:    [[IDXPROM3:%.*]] = zext i32 [[TMP18]] to i64
-// CHECK2-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM3]]
+// CHECK2-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM3]]
 // CHECK2-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4
 // CHECK2-NEXT:    [[MUL5:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK2-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 8
 // CHECK2-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4
 // CHECK2-NEXT:    [[IDXPROM6:%.*]] = zext i32 [[TMP21]] to i64
-// CHECK2-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM6]]
+// CHECK2-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM6]]
 // CHECK2-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX7]], align 4
 // CHECK2-NEXT:    [[MUL8:%.*]] = fmul float [[MUL5]], [[TMP22]]
 // CHECK2-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 8
 // CHECK2-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4
 // CHECK2-NEXT:    [[IDXPROM9:%.*]] = zext i32 [[TMP24]] to i64
-// CHECK2-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM9]]
+// CHECK2-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM9]]
 // CHECK2-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4
 // CHECK2-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK2:       omp.body.continue:
@@ -1895,21 +1895,21 @@ void range_for_collapsed() {
 // CHECK2-NEXT:    store i64 [[ADD1]], ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK2-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK2-NEXT:    [[TMP12:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[TMP12]]
+// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[TMP12]]
 // CHECK2-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK2-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK2-NEXT:    [[TMP15:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK2-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[TMP15]]
+// CHECK2-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[TMP15]]
 // CHECK2-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX2]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK2-NEXT:    [[MUL3:%.*]] = fmul float [[TMP13]], [[TMP16]]
 // CHECK2-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK2-NEXT:    [[TMP18:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK2-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[TMP18]]
+// CHECK2-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[TMP18]]
 // CHECK2-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK2-NEXT:    [[MUL5:%.*]] = fmul float [[MUL3]], [[TMP19]]
 // CHECK2-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK2-NEXT:    [[TMP21:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK2-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[TMP21]]
+// CHECK2-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[TMP21]]
 // CHECK2-NEXT:    store float [[MUL5]], ptr [[ARRAYIDX6]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK2-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK2:       omp.body.continue:
@@ -1998,21 +1998,21 @@ void range_for_collapsed() {
 // CHECK2-NEXT:    store i64 [[ADD1]], ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK2-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK2-NEXT:    [[TMP12:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[TMP12]]
+// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[TMP12]]
 // CHECK2-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK2-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK2-NEXT:    [[TMP15:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK2-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[TMP15]]
+// CHECK2-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[TMP15]]
 // CHECK2-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX2]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK2-NEXT:    [[MUL3:%.*]] = fmul float [[TMP13]], [[TMP16]]
 // CHECK2-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK2-NEXT:    [[TMP18:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK2-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[TMP18]]
+// CHECK2-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[TMP18]]
 // CHECK2-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK2-NEXT:    [[MUL5:%.*]] = fmul float [[MUL3]], [[TMP19]]
 // CHECK2-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK2-NEXT:    [[TMP21:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK2-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[TMP21]]
+// CHECK2-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[TMP21]]
 // CHECK2-NEXT:    store float [[MUL5]], ptr [[ARRAYIDX6]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK2-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK2:       omp.body.continue:
@@ -2275,24 +2275,24 @@ void range_for_collapsed() {
 // CHECK2-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[TMP14:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[IDXPROM:%.*]] = zext i8 [[TMP14]] to i64
-// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i64 [[IDXPROM]]
+// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i64 [[IDXPROM]]
 // CHECK2-NEXT:    [[TMP15:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[TMP17:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[IDXPROM6:%.*]] = zext i8 [[TMP17]] to i64
-// CHECK2-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM6]]
+// CHECK2-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM6]]
 // CHECK2-NEXT:    [[TMP18:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[MUL8:%.*]] = fmul float [[TMP15]], [[TMP18]]
 // CHECK2-NEXT:    [[TMP19:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[TMP20:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[IDXPROM9:%.*]] = zext i8 [[TMP20]] to i64
-// CHECK2-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP19]], i64 [[IDXPROM9]]
+// CHECK2-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP19]], i64 [[IDXPROM9]]
 // CHECK2-NEXT:    [[TMP21:%.*]] = load float, ptr [[ARRAYIDX10]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[MUL11:%.*]] = fmul float [[MUL8]], [[TMP21]]
 // CHECK2-NEXT:    [[TMP22:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[TMP23:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    [[IDXPROM12:%.*]] = zext i8 [[TMP23]] to i64
-// CHECK2-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw float, ptr [[TMP22]], i64 [[IDXPROM12]]
+// CHECK2-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds float, ptr [[TMP22]], i64 [[IDXPROM12]]
 // CHECK2-NEXT:    store float [[MUL11]], ptr [[ARRAYIDX13]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK2-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK2:       omp.body.continue:
@@ -2419,7 +2419,7 @@ void range_for_collapsed() {
 // CHECK2-NEXT:    [[CONV:%.*]] = sitofp i32 [[CALL]] to float
 // CHECK2-NEXT:    [[TMP13:%.*]] = load i32, ptr [[I]], align 4
 // CHECK2-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP13]] to i64
-// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[VLA1]], i64 [[IDXPROM]]
+// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[VLA1]], i64 [[IDXPROM]]
 // CHECK2-NEXT:    [[TMP14:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK2-NEXT:    [[ADD4:%.*]] = fadd float [[CONV]], [[TMP14]]
 // CHECK2-NEXT:    [[TMP15:%.*]] = load i32, ptr [[N_ADDR]], align 4
@@ -2428,7 +2428,7 @@ void range_for_collapsed() {
 // CHECK2-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[TMP0]], align 8
 // CHECK2-NEXT:    [[TMP17:%.*]] = load i32, ptr [[I]], align 4
 // CHECK2-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP17]] to i64
-// CHECK2-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM7]]
+// CHECK2-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM7]]
 // CHECK2-NEXT:    [[TMP18:%.*]] = load float, ptr [[ARRAYIDX8]], align 4
 // CHECK2-NEXT:    [[ADD9:%.*]] = fadd float [[TMP18]], [[ADD6]]
 // CHECK2-NEXT:    store float [[ADD9]], ptr [[ARRAYIDX8]], align 4
@@ -2897,24 +2897,24 @@ void range_for_collapsed() {
 // CHECK5-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 8, !dbg [[DBG45:![0-9]+]]
 // CHECK5-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4, !dbg [[DBG45]]
 // CHECK5-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP15]] to i64, !dbg [[DBG45]]
-// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM]], !dbg [[DBG45]]
+// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM]], !dbg [[DBG45]]
 // CHECK5-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !dbg [[DBG45]]
 // CHECK5-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 8, !dbg [[DBG45]]
 // CHECK5-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4, !dbg [[DBG45]]
 // CHECK5-NEXT:    [[IDXPROM3:%.*]] = zext i32 [[TMP18]] to i64, !dbg [[DBG45]]
-// CHECK5-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM3]], !dbg [[DBG45]]
+// CHECK5-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM3]], !dbg [[DBG45]]
 // CHECK5-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !dbg [[DBG45]]
 // CHECK5-NEXT:    [[MUL5:%.*]] = fmul float [[TMP16]], [[TMP19]], !dbg [[DBG45]]
 // CHECK5-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 8, !dbg [[DBG45]]
 // CHECK5-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4, !dbg [[DBG45]]
 // CHECK5-NEXT:    [[IDXPROM6:%.*]] = zext i32 [[TMP21]] to i64, !dbg [[DBG45]]
-// CHECK5-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM6]], !dbg [[DBG45]]
+// CHECK5-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM6]], !dbg [[DBG45]]
 // CHECK5-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !dbg [[DBG45]]
 // CHECK5-NEXT:    [[MUL8:%.*]] = fmul float [[MUL5]], [[TMP22]], !dbg [[DBG45]]
 // CHECK5-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 8, !dbg [[DBG45]]
 // CHECK5-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4, !dbg [[DBG45]]
 // CHECK5-NEXT:    [[IDXPROM9:%.*]] = zext i32 [[TMP24]] to i64, !dbg [[DBG45]]
-// CHECK5-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM9]], !dbg [[DBG45]]
+// CHECK5-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM9]], !dbg [[DBG45]]
 // CHECK5-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4, !dbg [[DBG45]]
 // CHECK5-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]], !dbg [[DBG46:![0-9]+]]
 // CHECK5:       omp.body.continue:
@@ -3011,21 +3011,21 @@ void range_for_collapsed() {
 // CHECK5-NEXT:    store i64 [[ADD1]], ptr [[I]], align 8, !dbg [[DBG54]], !llvm.access.group [[ACC_GRP55]]
 // CHECK5-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TMP1]], align 8, !dbg [[DBG56:![0-9]+]], !llvm.access.group [[ACC_GRP55]]
 // CHECK5-NEXT:    [[TMP12:%.*]] = load i64, ptr [[I]], align 8, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
-// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[TMP12]], !dbg [[DBG56]]
+// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[TMP12]], !dbg [[DBG56]]
 // CHECK5-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
 // CHECK5-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP2]], align 8, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
 // CHECK5-NEXT:    [[TMP15:%.*]] = load i64, ptr [[I]], align 8, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
-// CHECK5-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[TMP15]], !dbg [[DBG56]]
+// CHECK5-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[TMP15]], !dbg [[DBG56]]
 // CHECK5-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX2]], align 4, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
 // CHECK5-NEXT:    [[MUL3:%.*]] = fmul float [[TMP13]], [[TMP16]], !dbg [[DBG56]]
 // CHECK5-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP3]], align 8, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
 // CHECK5-NEXT:    [[TMP18:%.*]] = load i64, ptr [[I]], align 8, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
-// CHECK5-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[TMP18]], !dbg [[DBG56]]
+// CHECK5-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[TMP18]], !dbg [[DBG56]]
 // CHECK5-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
 // CHECK5-NEXT:    [[MUL5:%.*]] = fmul float [[MUL3]], [[TMP19]], !dbg [[DBG56]]
 // CHECK5-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP0]], align 8, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
 // CHECK5-NEXT:    [[TMP21:%.*]] = load i64, ptr [[I]], align 8, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
-// CHECK5-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[TMP21]], !dbg [[DBG56]]
+// CHECK5-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[TMP21]], !dbg [[DBG56]]
 // CHECK5-NEXT:    store float [[MUL5]], ptr [[ARRAYIDX6]], align 4, !dbg [[DBG56]], !llvm.access.group [[ACC_GRP55]]
 // CHECK5-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]], !dbg [[DBG57:![0-9]+]]
 // CHECK5:       omp.body.continue:
@@ -3114,21 +3114,21 @@ void range_for_collapsed() {
 // CHECK5-NEXT:    store i64 [[ADD1]], ptr [[I]], align 8, !dbg [[DBG66]], !llvm.access.group [[ACC_GRP67]]
 // CHECK5-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TMP1]], align 8, !dbg [[DBG68:![0-9]+]], !llvm.access.group [[ACC_GRP67]]
 // CHECK5-NEXT:    [[TMP12:%.*]] = load i64, ptr [[I]], align 8, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
-// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[TMP12]], !dbg [[DBG68]]
+// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[TMP12]], !dbg [[DBG68]]
 // CHECK5-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
 // CHECK5-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP2]], align 8, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
 // CHECK5-NEXT:    [[TMP15:%.*]] = load i64, ptr [[I]], align 8, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
-// CHECK5-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[TMP15]], !dbg [[DBG68]]
+// CHECK5-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[TMP15]], !dbg [[DBG68]]
 // CHECK5-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX2]], align 4, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
 // CHECK5-NEXT:    [[MUL3:%.*]] = fmul float [[TMP13]], [[TMP16]], !dbg [[DBG68]]
 // CHECK5-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP3]], align 8, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
 // CHECK5-NEXT:    [[TMP18:%.*]] = load i64, ptr [[I]], align 8, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
-// CHECK5-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[TMP18]], !dbg [[DBG68]]
+// CHECK5-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[TMP18]], !dbg [[DBG68]]
 // CHECK5-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
 // CHECK5-NEXT:    [[MUL5:%.*]] = fmul float [[MUL3]], [[TMP19]], !dbg [[DBG68]]
 // CHECK5-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP0]], align 8, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
 // CHECK5-NEXT:    [[TMP21:%.*]] = load i64, ptr [[I]], align 8, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
-// CHECK5-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[TMP21]], !dbg [[DBG68]]
+// CHECK5-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[TMP21]], !dbg [[DBG68]]
 // CHECK5-NEXT:    store float [[MUL5]], ptr [[ARRAYIDX6]], align 4, !dbg [[DBG68]], !llvm.access.group [[ACC_GRP67]]
 // CHECK5-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]], !dbg [[DBG69:![0-9]+]]
 // CHECK5:       omp.body.continue:
@@ -3391,24 +3391,24 @@ void range_for_collapsed() {
 // CHECK5-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[TMP1]], align 8, !dbg [[DBG97:![0-9]+]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[TMP14:%.*]] = load i8, ptr [[I]], align 1, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[IDXPROM:%.*]] = zext i8 [[TMP14]] to i64, !dbg [[DBG97]]
-// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i64 [[IDXPROM]], !dbg [[DBG97]]
+// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i64 [[IDXPROM]], !dbg [[DBG97]]
 // CHECK5-NEXT:    [[TMP15:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[TMP2]], align 8, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[TMP17:%.*]] = load i8, ptr [[I]], align 1, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[IDXPROM6:%.*]] = zext i8 [[TMP17]] to i64, !dbg [[DBG97]]
-// CHECK5-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM6]], !dbg [[DBG97]]
+// CHECK5-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM6]], !dbg [[DBG97]]
 // CHECK5-NEXT:    [[TMP18:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[MUL8:%.*]] = fmul float [[TMP15]], [[TMP18]], !dbg [[DBG97]]
 // CHECK5-NEXT:    [[TMP19:%.*]] = load ptr, ptr [[TMP3]], align 8, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[TMP20:%.*]] = load i8, ptr [[I]], align 1, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[IDXPROM9:%.*]] = zext i8 [[TMP20]] to i64, !dbg [[DBG97]]
-// CHECK5-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP19]], i64 [[IDXPROM9]], !dbg [[DBG97]]
+// CHECK5-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP19]], i64 [[IDXPROM9]], !dbg [[DBG97]]
 // CHECK5-NEXT:    [[TMP21:%.*]] = load float, ptr [[ARRAYIDX10]], align 4, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[MUL11:%.*]] = fmul float [[MUL8]], [[TMP21]], !dbg [[DBG97]]
 // CHECK5-NEXT:    [[TMP22:%.*]] = load ptr, ptr [[TMP0]], align 8, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[TMP23:%.*]] = load i8, ptr [[I]], align 1, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    [[IDXPROM12:%.*]] = zext i8 [[TMP23]] to i64, !dbg [[DBG97]]
-// CHECK5-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw float, ptr [[TMP22]], i64 [[IDXPROM12]], !dbg [[DBG97]]
+// CHECK5-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds float, ptr [[TMP22]], i64 [[IDXPROM12]], !dbg [[DBG97]]
 // CHECK5-NEXT:    store float [[MUL11]], ptr [[ARRAYIDX13]], align 4, !dbg [[DBG97]], !llvm.access.group [[ACC_GRP95]]
 // CHECK5-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]], !dbg [[DBG98:![0-9]+]]
 // CHECK5:       omp.body.continue:
@@ -3535,7 +3535,7 @@ void range_for_collapsed() {
 // CHECK5-NEXT:    [[CONV:%.*]] = sitofp i32 [[CALL]] to float, !dbg [[DBG111]]
 // CHECK5-NEXT:    [[TMP13:%.*]] = load i32, ptr [[I]], align 4, !dbg [[DBG111]]
 // CHECK5-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP13]] to i64, !dbg [[DBG111]]
-// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[VLA1]], i64 [[IDXPROM]], !dbg [[DBG111]]
+// CHECK5-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[VLA1]], i64 [[IDXPROM]], !dbg [[DBG111]]
 // CHECK5-NEXT:    [[TMP14:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !dbg [[DBG111]]
 // CHECK5-NEXT:    [[ADD4:%.*]] = fadd float [[CONV]], [[TMP14]], !dbg [[DBG111]]
 // CHECK5-NEXT:    [[TMP15:%.*]] = load i32, ptr [[N_ADDR]], align 4, !dbg [[DBG111]]
@@ -3544,7 +3544,7 @@ void range_for_collapsed() {
 // CHECK5-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[TMP0]], align 8, !dbg [[DBG111]]
 // CHECK5-NEXT:    [[TMP17:%.*]] = load i32, ptr [[I]], align 4, !dbg [[DBG111]]
 // CHECK5-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP17]] to i64, !dbg [[DBG111]]
-// CHECK5-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM7]], !dbg [[DBG111]]
+// CHECK5-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM7]], !dbg [[DBG111]]
 // CHECK5-NEXT:    [[TMP18:%.*]] = load float, ptr [[ARRAYIDX8]], align 4, !dbg [[DBG111]]
 // CHECK5-NEXT:    [[ADD9:%.*]] = fadd float [[TMP18]], [[ADD6]], !dbg [[DBG111]]
 // CHECK5-NEXT:    store float [[ADD9]], ptr [[ARRAYIDX8]], align 4, !dbg [[DBG111]]
@@ -4013,24 +4013,24 @@ void range_for_collapsed() {
 // CHECK6-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP1]], align 8
 // CHECK6-NEXT:    [[TMP15:%.*]] = load i32, ptr [[I]], align 4
 // CHECK6-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP15]] to i64
-// CHECK6-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[IDXPROM]]
+// CHECK6-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[IDXPROM]]
 // CHECK6-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK6-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP2]], align 8
 // CHECK6-NEXT:    [[TMP18:%.*]] = load i32, ptr [[I]], align 4
 // CHECK6-NEXT:    [[IDXPROM3:%.*]] = zext i32 [[TMP18]] to i64
-// CHECK6-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[IDXPROM3]]
+// CHECK6-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[IDXPROM3]]
 // CHECK6-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4
 // CHECK6-NEXT:    [[MUL5:%.*]] = fmul float [[TMP16]], [[TMP19]]
 // CHECK6-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP3]], align 8
 // CHECK6-NEXT:    [[TMP21:%.*]] = load i32, ptr [[I]], align 4
 // CHECK6-NEXT:    [[IDXPROM6:%.*]] = zext i32 [[TMP21]] to i64
-// CHECK6-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[IDXPROM6]]
+// CHECK6-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[IDXPROM6]]
 // CHECK6-NEXT:    [[TMP22:%.*]] = load float, ptr [[ARRAYIDX7]], align 4
 // CHECK6-NEXT:    [[MUL8:%.*]] = fmul float [[MUL5]], [[TMP22]]
 // CHECK6-NEXT:    [[TMP23:%.*]] = load ptr, ptr [[TMP0]], align 8
 // CHECK6-NEXT:    [[TMP24:%.*]] = load i32, ptr [[I]], align 4
 // CHECK6-NEXT:    [[IDXPROM9:%.*]] = zext i32 [[TMP24]] to i64
-// CHECK6-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP23]], i64 [[IDXPROM9]]
+// CHECK6-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP23]], i64 [[IDXPROM9]]
 // CHECK6-NEXT:    store float [[MUL8]], ptr [[ARRAYIDX10]], align 4
 // CHECK6-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK6:       omp.body.continue:
@@ -4127,21 +4127,21 @@ void range_for_collapsed() {
 // CHECK6-NEXT:    store i64 [[ADD1]], ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK6-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK6-NEXT:    [[TMP12:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK6-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[TMP12]]
+// CHECK6-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[TMP12]]
 // CHECK6-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK6-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK6-NEXT:    [[TMP15:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK6-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[TMP15]]
+// CHECK6-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[TMP15]]
 // CHECK6-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX2]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK6-NEXT:    [[MUL3:%.*]] = fmul float [[TMP13]], [[TMP16]]
 // CHECK6-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK6-NEXT:    [[TMP18:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK6-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[TMP18]]
+// CHECK6-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[TMP18]]
 // CHECK6-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK6-NEXT:    [[MUL5:%.*]] = fmul float [[MUL3]], [[TMP19]]
 // CHECK6-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP5]]
 // CHECK6-NEXT:    [[TMP21:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP5]]
-// CHECK6-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[TMP21]]
+// CHECK6-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[TMP21]]
 // CHECK6-NEXT:    store float [[MUL5]], ptr [[ARRAYIDX6]], align 4, !llvm.access.group [[ACC_GRP5]]
 // CHECK6-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK6:       omp.body.continue:
@@ -4230,21 +4230,21 @@ void range_for_collapsed() {
 // CHECK6-NEXT:    store i64 [[ADD1]], ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK6-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK6-NEXT:    [[TMP12:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK6-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 [[TMP12]]
+// CHECK6-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 [[TMP12]]
 // CHECK6-NEXT:    [[TMP13:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK6-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK6-NEXT:    [[TMP15:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK6-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP14]], i64 [[TMP15]]
+// CHECK6-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[TMP14]], i64 [[TMP15]]
 // CHECK6-NEXT:    [[TMP16:%.*]] = load float, ptr [[ARRAYIDX2]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK6-NEXT:    [[MUL3:%.*]] = fmul float [[TMP13]], [[TMP16]]
 // CHECK6-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK6-NEXT:    [[TMP18:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK6-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 [[TMP18]]
+// CHECK6-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 [[TMP18]]
 // CHECK6-NEXT:    [[TMP19:%.*]] = load float, ptr [[ARRAYIDX4]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK6-NEXT:    [[MUL5:%.*]] = fmul float [[MUL3]], [[TMP19]]
 // CHECK6-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP8]]
 // CHECK6-NEXT:    [[TMP21:%.*]] = load i64, ptr [[I]], align 8, !llvm.access.group [[ACC_GRP8]]
-// CHECK6-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP20]], i64 [[TMP21]]
+// CHECK6-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP20]], i64 [[TMP21]]
 // CHECK6-NEXT:    store float [[MUL5]], ptr [[ARRAYIDX6]], align 4, !llvm.access.group [[ACC_GRP8]]
 // CHECK6-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK6:       omp.body.continue:
@@ -4507,24 +4507,24 @@ void range_for_collapsed() {
 // CHECK6-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[TMP1]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[TMP14:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[IDXPROM:%.*]] = zext i8 [[TMP14]] to i64
-// CHECK6-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP13]], i64 [[IDXPROM]]
+// CHECK6-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP13]], i64 [[IDXPROM]]
 // CHECK6-NEXT:    [[TMP15:%.*]] = load float, ptr [[ARRAYIDX]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[TMP2]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[TMP17:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[IDXPROM6:%.*]] = zext i8 [[TMP17]] to i64
-// CHECK6-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM6]]
+// CHECK6-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM6]]
 // CHECK6-NEXT:    [[TMP18:%.*]] = load float, ptr [[ARRAYIDX7]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[MUL8:%.*]] = fmul float [[TMP15]], [[TMP18]]
 // CHECK6-NEXT:    [[TMP19:%.*]] = load ptr, ptr [[TMP3]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[TMP20:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[IDXPROM9:%.*]] = zext i8 [[TMP20]] to i64
-// CHECK6-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP19]], i64 [[IDXPROM9]]
+// CHECK6-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, ptr [[TMP19]], i64 [[IDXPROM9]]
 // CHECK6-NEXT:    [[TMP21:%.*]] = load float, ptr [[ARRAYIDX10]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[MUL11:%.*]] = fmul float [[MUL8]], [[TMP21]]
 // CHECK6-NEXT:    [[TMP22:%.*]] = load ptr, ptr [[TMP0]], align 8, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[TMP23:%.*]] = load i8, ptr [[I]], align 1, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    [[IDXPROM12:%.*]] = zext i8 [[TMP23]] to i64
-// CHECK6-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw float, ptr [[TMP22]], i64 [[IDXPROM12]]
+// CHECK6-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds float, ptr [[TMP22]], i64 [[IDXPROM12]]
 // CHECK6-NEXT:    store float [[MUL11]], ptr [[ARRAYIDX13]], align 4, !llvm.access.group [[ACC_GRP14]]
 // CHECK6-NEXT:    br label [[OMP_BODY_CONTINUE:%.*]]
 // CHECK6:       omp.body.continue:
@@ -4649,7 +4649,7 @@ void range_for_collapsed() {
 // CHECK6-NEXT:    [[CONV:%.*]] = sitofp i32 [[CALL]] to float
 // CHECK6-NEXT:    [[TMP13:%.*]] = load i32, ptr [[I]], align 4
 // CHECK6-NEXT:    [[IDXPROM:%.*]] = zext i32 [[TMP13]] to i64
-// CHECK6-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[VLA1]], i64 [[IDXPROM]]
+// CHECK6-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[VLA1]], i64 [[IDXPROM]]
 // CHECK6-NEXT:    [[TMP14:%.*]] = load float, ptr [[ARRAYIDX]], align 4
 // CHECK6-NEXT:    [[ADD4:%.*]] = fadd float [[CONV]], [[TMP14]]
 // CHECK6-NEXT:    [[TMP15:%.*]] = load i32, ptr [[N_ADDR]], align 4
@@ -4658,7 +4658,7 @@ void range_for_collapsed() {
 // CHECK6-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[TMP0]], align 8
 // CHECK6-NEXT:    [[TMP17:%.*]] = load i32, ptr [[I]], align 4
 // CHECK6-NEXT:    [[IDXPROM7:%.*]] = zext i32 [[TMP17]] to i64
-// CHECK6-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw float, ptr [[TMP16]], i64 [[IDXPROM7]]
+// CHECK6-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, ptr [[TMP16]], i64 [[IDXPROM7]]
 // CHECK6-NEXT:    [[TMP18:%.*]] = load float, ptr [[ARRAYIDX8]], align 4
 // CHECK6-NEXT:    [[ADD9:%.*]] = fadd float [[TMP18]], [[ADD6]]
 // CHECK6-NEXT:    store float [[ADD9]], ptr [[ARRAYIDX8]], align 4
diff --git a/clang/test/OpenMP/parallel_for_linear_codegen.cpp b/clang/test/OpenMP/parallel_for_linear_codegen.cpp
index 15eb0dfa42af59..8b46797ae253f9 100644
--- a/clang/test/OpenMP/parallel_for_linear_codegen.cpp
+++ b/clang/test/OpenMP/parallel_for_linear_codegen.cpp
@@ -337,7 +337,7 @@ int main() {
 // CHECK1-NEXT:    [[ADD7:%.*]] = add nsw i32 [[TMP14]], [[MUL6]]
 // CHECK1-NEXT:    store i32 [[ADD7]], ptr [[LVAR3]], align 4
 // CHECK1-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[PVAR2]], align 8
-// CHECK1-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP16]], i32 1
+// CHECK1-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP16]], i32 1
 // CHECK1-NEXT:    store ptr [[INCDEC_PTR]], ptr [[PVAR2]], align 8
 // CHECK1-NEXT:    [[TMP17:%.*]] = load i32, ptr [[LVAR3]], align 4
 // CHECK1-NEXT:    [[INC:%.*]] = add nsw i32 [[TMP17]], 1
diff --git a/clang/test/OpenMP/parallel_for_reduction_task_codegen.cpp b/clang/test/OpenMP/parallel_for_reduction_task_codegen.cpp
index 59d169d7a17387..752aec788bf349 100644
--- a/clang/test/OpenMP/parallel_for_reduction_task_codegen.cpp
+++ b/clang/test/OpenMP/parallel_for_reduction_task_codegen.cpp
@@ -83,16 +83,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP4:%.*]] = sext i32 [[TMP3]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP4]]
 // CHECK1-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP5]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP5]], i64 9
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP9:%.*]] = sub i64 [[TMP7]], [[TMP8]]
@@ -122,7 +122,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP21:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP20]]
 // CHECK1-NEXT:    store ptr [[_TMP6]], ptr [[_TMP5]], align 8
 // CHECK1-NEXT:    store ptr [[TMP21]], ptr [[_TMP6]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP22]], align 8
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -137,19 +137,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP27]], align 8
 // CHECK1-NEXT:    [[TMP28:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP28]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP29:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP30:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP30]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds ptr, ptr [[TMP30]], i64 0
 // CHECK1-NEXT:    [[TMP31:%.*]] = load ptr, ptr [[ARRAYIDX8]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP31]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds i8, ptr [[TMP31]], i64 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP33:%.*]] = sext i32 [[TMP32]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN10:%.*]] = add nsw i64 -1, [[TMP33]]
 // CHECK1-NEXT:    [[TMP34:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP34]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds ptr, ptr [[TMP34]], i64 9
 // CHECK1-NEXT:    [[TMP35:%.*]] = load ptr, ptr [[ARRAYIDX11]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP35]], i64 [[LB_ADD_LEN10]]
+// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds i8, ptr [[TMP35]], i64 [[LB_ADD_LEN10]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP36:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX9]], ptr [[TMP36]], align 8
@@ -470,9 +470,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]
diff --git a/clang/test/OpenMP/parallel_for_scan_codegen.cpp b/clang/test/OpenMP/parallel_for_scan_codegen.cpp
index 67b32407c712fd..161534814a793d 100644
--- a/clang/test/OpenMP/parallel_for_scan_codegen.cpp
+++ b/clang/test/OpenMP/parallel_for_scan_codegen.cpp
@@ -28,9 +28,9 @@ void baz(int n) {
 
   // CHECK: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(
   // CHECK: [[LAST:%.+]] = mul nsw i64 9, %
-  // CHECK: [[LAST_REF:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[LAST]]
+  // CHECK: [[LAST_REF:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[LAST]]
   // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 16 @_ZZ3baziE1a, ptr align 4 [[LAST_REF]], i64 %{{.+}}, i1 false)
-  // CHECK: [[LAST_REF_B:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 9
+  // CHECK: [[LAST_REF_B:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 9
   // CHECK: [[LAST_VAL:%.+]] = load double, ptr [[LAST_REF_B]],
   // CHECK: store double [[LAST_VAL]], ptr @_ZZ3baziE1b,
 
@@ -58,13 +58,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS:%.+]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF:%.+]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF:%.+]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_BUF_IDX]], ptr {{.*}}[[A_PRIV]], i64 [[BYTES]], i1 false)
 
     // b_buffer[i] = b_priv;
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF:%.+]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF:%.+]], i64 [[BASE_IDX]]
     // CHECK: [[B_PRIV:%.+]] = load double, ptr [[B_PRIV_ADDR]],
     // CHECK: store double [[B_PRIV]], ptr [[B_BUF_IDX]],
     // CHECK: br label %[[LOOP_CONTINUE:.+]]
@@ -91,13 +91,13 @@ void baz(int n) {
 
     // a_buffer[i] += a_buffer[i-pow(2, k)];
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[I]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[IDX_SUB_K2POW]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[I]]
+    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[I]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
-    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
+    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
     // CHECK: [[A_BUF_END:%.+]] = getelementptr float, ptr [[A_BUF_IDX]], i64 [[NUM_ELEMS]]
     // CHECK: [[ISEMPTY:%.+]] = icmp eq ptr [[A_BUF_IDX]], [[A_BUF_END]]
     // CHECK: br i1 [[ISEMPTY]], label %[[RED_DONE:[^,]+]], label %[[RED_BODY:[^,]+]]
@@ -151,13 +151,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_PRIV]], ptr {{.*}}[[A_BUF_IDX]], i64 [[BYTES]], i1 false)
 
     // b_priv = b_buffer[i];
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX]]
     // CHECK: [[B_BUF_IDX_VAL:%.+]] = load double, ptr [[B_BUF_IDX]],
     // CHECK: store double [[B_BUF_IDX_VAL]], ptr [[B_PRIV_ADDR]],
     // CHECK: br label %[[SCAN_PHASE:[^,]+]]
@@ -188,13 +188,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS:%.+]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF:%.+]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF:%.+]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_BUF_IDX]], ptr {{.*}}[[A_PRIV]], i64 [[BYTES]], i1 false)
 
     // b_buffer[i] = b_priv;
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF:%.+]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF:%.+]], i64 [[BASE_IDX]]
     // CHECK: [[B_PRIV:%.+]] = load double, ptr [[B_PRIV_ADDR]],
     // CHECK: store double [[B_PRIV]], ptr [[B_BUF_IDX]],
     // CHECK: br label %[[LOOP_CONTINUE:[^,]+]]
@@ -226,13 +226,13 @@ void baz(int n) {
 
     // a_buffer[i] += a_buffer[i-pow(2, k)];
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[I]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[IDX_SUB_K2POW]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[I]]
+    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[I]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
-    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
+    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
     // CHECK: [[A_BUF_END:%.+]] = getelementptr float, ptr [[A_BUF_IDX]], i64 [[NUM_ELEMS]]
     // CHECK: [[ISEMPTY:%.+]] = icmp eq ptr [[A_BUF_IDX]], [[A_BUF_END]]
     // CHECK: br i1 [[ISEMPTY]], label %[[RED_DONE:[^,]+]], label %[[RED_BODY:[^,]+]]
@@ -289,13 +289,13 @@ void baz(int n) {
     // CHECK: [[IF_THEN]]:
     // CHECK: [[BASE_IDX_SUB_1:%.+]] = sub nuw i64 [[BASE_IDX]], 1
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX_SUB_1]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_PRIV]], ptr {{.*}}[[A_BUF_IDX]], i64 [[BYTES]], i1 false)
 
     // b_priv = b_buffer[i];
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX_SUB_1]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX_SUB_1]]
     // CHECK: [[B_BUF_IDX_VAL:%.+]] = load double, ptr [[B_BUF_IDX]],
     // CHECK: store double [[B_BUF_IDX_VAL]], ptr [[B_PRIV_ADDR]],
     // CHECK: br label %[[SCAN_PHASE]]
diff --git a/clang/test/OpenMP/parallel_for_simd_scan_codegen.cpp b/clang/test/OpenMP/parallel_for_simd_scan_codegen.cpp
index cac997753d4801..7e973a602a65cc 100644
--- a/clang/test/OpenMP/parallel_for_simd_scan_codegen.cpp
+++ b/clang/test/OpenMP/parallel_for_simd_scan_codegen.cpp
@@ -51,13 +51,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS:%.+]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF:%.+]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF:%.+]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_BUF_IDX]], ptr {{.*}}[[A_PRIV]], i64 [[BYTES]], i1 false)
 
     // b_buffer[i] = b_priv;
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF:%.+]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF:%.+]], i64 [[BASE_IDX]]
     // CHECK: [[B_PRIV:%.+]] = load double, ptr [[B_PRIV_ADDR]],
     // CHECK: store double [[B_PRIV]], ptr [[B_BUF_IDX]],
     // CHECK: br label %[[LOOP_CONTINUE:.+]]
@@ -84,13 +84,13 @@ void baz(int n) {
 
     // a_buffer[i] += a_buffer[i-pow(2, k)];
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[I]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[IDX_SUB_K2POW]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[I]]
+    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[I]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
-    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
+    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
     // CHECK: [[A_BUF_END:%.+]] = getelementptr float, ptr [[A_BUF_IDX]], i64 [[NUM_ELEMS]]
     // CHECK: [[ISEMPTY:%.+]] = icmp eq ptr [[A_BUF_IDX]], [[A_BUF_END]]
     // CHECK: br i1 [[ISEMPTY]], label %[[RED_DONE:[^,]+]], label %[[RED_BODY:[^,]+]]
@@ -144,13 +144,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_PRIV]], ptr {{.*}}[[A_BUF_IDX]], i64 [[BYTES]], i1 false)
 
     // b_priv = b_buffer[i];
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX]]
     // CHECK: [[B_BUF_IDX_VAL:%.+]] = load double, ptr [[B_BUF_IDX]],
     // CHECK: store double [[B_BUF_IDX_VAL]], ptr [[B_PRIV_ADDR]],
     // CHECK: br label %[[SCAN_PHASE:[^,]+]]
@@ -181,13 +181,13 @@ void baz(int n) {
     // CHECK: [[BASE_IDX_I:%.+]] = load i32, ptr [[IV_ADDR:%.+]],
     // CHECK: [[BASE_IDX:%.+]] = zext i32 [[BASE_IDX_I]] to i64
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX]], [[NUM_ELEMS:%.+]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF:%.+]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF:%.+]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_BUF_IDX]], ptr {{.*}}[[A_PRIV]], i64 [[BYTES]], i1 false)
 
     // b_buffer[i] = b_priv;
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF:%.+]], i64 [[BASE_IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF:%.+]], i64 [[BASE_IDX]]
     // CHECK: [[B_PRIV:%.+]] = load double, ptr [[B_PRIV_ADDR]],
     // CHECK: store double [[B_PRIV]], ptr [[B_BUF_IDX]],
     // CHECK: br label %[[LOOP_CONTINUE:[^,]+]]
@@ -219,13 +219,13 @@ void baz(int n) {
 
     // a_buffer[i] += a_buffer[i-pow(2, k)];
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[I]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[IDX_SUB_K2POW]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[I]]
+    // CHECK: [[A_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[I]]
     // CHECK: [[IDX_SUB_K2POW:%.+]] = sub nuw i64 [[I]], [[K2POW]]
-    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
+    // CHECK: [[B_BUF_IDX_SUB_K2POW:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[IDX_SUB_K2POW]]
     // CHECK: [[A_BUF_END:%.+]] = getelementptr float, ptr [[A_BUF_IDX]], i64 [[NUM_ELEMS]]
     // CHECK: [[ISEMPTY:%.+]] = icmp eq ptr [[A_BUF_IDX]], [[A_BUF_END]]
     // CHECK: br i1 [[ISEMPTY]], label %[[RED_DONE:[^,]+]], label %[[RED_BODY:[^,]+]]
@@ -282,13 +282,13 @@ void baz(int n) {
     // CHECK: [[IF_THEN]]:
     // CHECK: [[BASE_IDX_SUB_1:%.+]] = sub nuw i64 [[BASE_IDX]], 1
     // CHECK: [[IDX:%.+]] = mul nsw i64 [[BASE_IDX_SUB_1]], [[NUM_ELEMS]]
-    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[A_BUF]], i64 [[IDX]]
-    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds nuw [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
+    // CHECK: [[A_BUF_IDX:%.+]] = getelementptr inbounds float, ptr [[A_BUF]], i64 [[IDX]]
+    // CHECK: [[A_PRIV:%.+]] = getelementptr inbounds [10 x float], ptr [[A_PRIV_ADDR:%.+]], i64 0, i64 0
     // CHECK: [[BYTES:%.+]] = mul nuw i64 [[NUM_ELEMS:%.+]], 4
     // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr {{.*}}[[A_PRIV]], ptr {{.*}}[[A_BUF_IDX]], i64 [[BYTES]], i1 false)
 
     // b_priv = b_buffer[i];
-    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds nuw double, ptr [[B_BUF]], i64 [[BASE_IDX_SUB_1]]
+    // CHECK: [[B_BUF_IDX:%.+]] = getelementptr inbounds double, ptr [[B_BUF]], i64 [[BASE_IDX_SUB_1]]
     // CHECK: [[B_BUF_IDX_VAL:%.+]] = load double, ptr [[B_BUF_IDX]],
     // CHECK: store double [[B_BUF_IDX_VAL]], ptr [[B_PRIV_ADDR]],
     // CHECK: br label %[[SCAN_PHASE]]
diff --git a/clang/test/OpenMP/parallel_master_reduction_task_codegen.cpp b/clang/test/OpenMP/parallel_master_reduction_task_codegen.cpp
index c1fe00f2380018..b17757a5f978ab 100644
--- a/clang/test/OpenMP/parallel_master_reduction_task_codegen.cpp
+++ b/clang/test/OpenMP/parallel_master_reduction_task_codegen.cpp
@@ -72,16 +72,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP4:%.*]] = sext i32 [[TMP3]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP4]]
 // CHECK1-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP5]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP5]], i64 9
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP9:%.*]] = sub i64 [[TMP7]], [[TMP8]]
@@ -111,7 +111,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP21:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP20]]
 // CHECK1-NEXT:    store ptr [[_TMP5]], ptr [[TMP]], align 8
 // CHECK1-NEXT:    store ptr [[TMP21]], ptr [[_TMP5]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP22]], align 8
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -126,19 +126,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP27]], align 8
 // CHECK1-NEXT:    [[TMP28:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP28]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP29:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP30:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP30]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds ptr, ptr [[TMP30]], i64 0
 // CHECK1-NEXT:    [[TMP31:%.*]] = load ptr, ptr [[ARRAYIDX7]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP31]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i8, ptr [[TMP31]], i64 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP33:%.*]] = sext i32 [[TMP32]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN9:%.*]] = add nsw i64 -1, [[TMP33]]
 // CHECK1-NEXT:    [[TMP34:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP34]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds ptr, ptr [[TMP34]], i64 9
 // CHECK1-NEXT:    [[TMP35:%.*]] = load ptr, ptr [[ARRAYIDX10]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP36:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX8]], ptr [[TMP36]], align 8
@@ -425,9 +425,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]
diff --git a/clang/test/OpenMP/parallel_master_taskloop_reduction_codegen.cpp b/clang/test/OpenMP/parallel_master_taskloop_reduction_codegen.cpp
index 1d106922435d5c..3c74aaca3f46d8 100644
--- a/clang/test/OpenMP/parallel_master_taskloop_reduction_codegen.cpp
+++ b/clang/test/OpenMP/parallel_master_taskloop_reduction_codegen.cpp
@@ -84,9 +84,9 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB1:.+]], ptr [[TMP25]],
 // CHECK-DAG:    [[TMP26:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK-DAG:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP26]], i8 0, i64 4, i1 false)
-// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C:%.+]], i64 0, i64 0
+// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C:%.+]], i64 0, i64 0
 // CHECK-DAG:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, %
-// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
+// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
 // CHECK-DAG:    [[TMP28]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_4:%.+]], i32 0, i32 0
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
@@ -138,10 +138,10 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB4:.+]], ptr [[TMP59]],
 // CHECK-DAG:    [[TMP60:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_8]], i32 0, i32 6
 // CHECK-DAG:    store i32 1, ptr [[TMP60]],
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
 // CHECK:    [[TMP62:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0:%.+]], i32 4, ptr [[DOTRD_INPUT_]])
 // CHECK:    [[TMP63:%.*]] = load i32, ptr [[N:%.+]],
 // CHECK:    store i32 [[TMP63]], ptr [[DOTCAPTURE_EXPR_]],
diff --git a/clang/test/OpenMP/parallel_master_taskloop_simd_reduction_codegen.cpp b/clang/test/OpenMP/parallel_master_taskloop_simd_reduction_codegen.cpp
index 9a524c3b94c6e8..7d4216ddde6a37 100644
--- a/clang/test/OpenMP/parallel_master_taskloop_simd_reduction_codegen.cpp
+++ b/clang/test/OpenMP/parallel_master_taskloop_simd_reduction_codegen.cpp
@@ -84,9 +84,9 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB1:.+]], ptr [[TMP25]],
 // CHECK-DAG:    [[TMP26:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK-DAG:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP26]], i8 0, i64 4, i1 false)
-// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C:%.+]], i64 0, i64 0
+// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C:%.+]], i64 0, i64 0
 // CHECK-DAG:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, %
-// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
+// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
 // CHECK-DAG:    [[TMP28]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_4:%.+]], i32 0, i32 0
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
@@ -138,10 +138,10 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB4:.+]], ptr [[TMP59]],
 // CHECK-DAG:    [[TMP60:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_8]], i32 0, i32 6
 // CHECK-DAG:    store i32 1, ptr [[TMP60]],
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
 // CHECK:    [[TMP62:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0:%.+]], i32 4, ptr [[DOTRD_INPUT_]])
 // CHECK:    [[TMP63:%.*]] = load i32, ptr [[N:%.+]],
 // CHECK:    store i32 [[TMP63]], ptr [[DOTCAPTURE_EXPR_]],
diff --git a/clang/test/OpenMP/parallel_reduction_codegen.cpp b/clang/test/OpenMP/parallel_reduction_codegen.cpp
index ce76429b871fe0..f49faa6b89deb4 100644
--- a/clang/test/OpenMP/parallel_reduction_codegen.cpp
+++ b/clang/test/OpenMP/parallel_reduction_codegen.cpp
@@ -354,9 +354,9 @@ int main() {
 // CHECK1-NEXT:    store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
 // CHECK1-NEXT:    store ptr [[X]], ptr [[X_ADDR]], align 8
 // CHECK1-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[X_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i16, ptr [[TMP0]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i16, ptr [[TMP0]], i64 0
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[X_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i16, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i16, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = ptrtoint ptr [[ARRAYIDX1]] to i64
 // CHECK1-NEXT:    [[TMP3:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK1-NEXT:    [[TMP4:%.*]] = sub i64 [[TMP2]], [[TMP3]]
@@ -1632,9 +1632,9 @@ int main() {
 // CHECK3-NEXT:    store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
 // CHECK3-NEXT:    store ptr [[X]], ptr [[X_ADDR]], align 8
 // CHECK3-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[X_ADDR]], align 8
-// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i16, ptr [[TMP0]], i64 0
+// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i16, ptr [[TMP0]], i64 0
 // CHECK3-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[X_ADDR]], align 8
-// CHECK3-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i16, ptr [[TMP1]], i64 0
+// CHECK3-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i16, ptr [[TMP1]], i64 0
 // CHECK3-NEXT:    [[TMP2:%.*]] = ptrtoint ptr [[ARRAYIDX1]] to i64
 // CHECK3-NEXT:    [[TMP3:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK3-NEXT:    [[TMP4:%.*]] = sub i64 [[TMP2]], [[TMP3]]
@@ -2142,9 +2142,9 @@ int main() {
 // CHECK4-NEXT:    store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
 // CHECK4-NEXT:    store ptr [[X]], ptr [[X_ADDR]], align 8
 // CHECK4-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[X_ADDR]], align 8
-// CHECK4-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i16, ptr [[TMP0]], i64 0
+// CHECK4-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i16, ptr [[TMP0]], i64 0
 // CHECK4-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[X_ADDR]], align 8
-// CHECK4-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i16, ptr [[TMP1]], i64 0
+// CHECK4-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i16, ptr [[TMP1]], i64 0
 // CHECK4-NEXT:    [[TMP2:%.*]] = ptrtoint ptr [[ARRAYIDX1]] to i64
 // CHECK4-NEXT:    [[TMP3:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK4-NEXT:    [[TMP4:%.*]] = sub i64 [[TMP2]], [[TMP3]]
diff --git a/clang/test/OpenMP/parallel_reduction_task_codegen.cpp b/clang/test/OpenMP/parallel_reduction_task_codegen.cpp
index 40cc3103b1c0fd..208f7a41aa3db7 100644
--- a/clang/test/OpenMP/parallel_reduction_task_codegen.cpp
+++ b/clang/test/OpenMP/parallel_reduction_task_codegen.cpp
@@ -72,16 +72,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP4:%.*]] = sext i32 [[TMP3]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP4]]
 // CHECK1-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP5]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP5]], i64 9
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP9:%.*]] = sub i64 [[TMP7]], [[TMP8]]
@@ -111,7 +111,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP21:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP20]]
 // CHECK1-NEXT:    store ptr [[_TMP5]], ptr [[TMP]], align 8
 // CHECK1-NEXT:    store ptr [[TMP21]], ptr [[_TMP5]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP22]], align 8
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -126,19 +126,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP27]], align 8
 // CHECK1-NEXT:    [[TMP28:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP28]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP29:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP30:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP30]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds ptr, ptr [[TMP30]], i64 0
 // CHECK1-NEXT:    [[TMP31:%.*]] = load ptr, ptr [[ARRAYIDX7]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP31]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i8, ptr [[TMP31]], i64 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP33:%.*]] = sext i32 [[TMP32]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN9:%.*]] = add nsw i64 -1, [[TMP33]]
 // CHECK1-NEXT:    [[TMP34:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP34]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds ptr, ptr [[TMP34]], i64 9
 // CHECK1-NEXT:    [[TMP35:%.*]] = load ptr, ptr [[ARRAYIDX10]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP36:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX8]], ptr [[TMP36]], align 8
@@ -416,9 +416,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]
diff --git a/clang/test/OpenMP/parallel_sections_reduction_task_codegen.cpp b/clang/test/OpenMP/parallel_sections_reduction_task_codegen.cpp
index 61597a074cf59d..6d73652c3ea279 100644
--- a/clang/test/OpenMP/parallel_sections_reduction_task_codegen.cpp
+++ b/clang/test/OpenMP/parallel_sections_reduction_task_codegen.cpp
@@ -81,16 +81,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_SECTIONS_IL_]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP4:%.*]] = sext i32 [[TMP3]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP4]]
 // CHECK1-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP5]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP5]], i64 9
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP9:%.*]] = sub i64 [[TMP7]], [[TMP8]]
@@ -120,7 +120,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP21:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP20]]
 // CHECK1-NEXT:    store ptr [[_TMP5]], ptr [[TMP]], align 8
 // CHECK1-NEXT:    store ptr [[TMP21]], ptr [[_TMP5]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP22]], align 8
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -135,19 +135,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP27]], align 8
 // CHECK1-NEXT:    [[TMP28:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP28]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP29:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP30:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP30]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds ptr, ptr [[TMP30]], i64 0
 // CHECK1-NEXT:    [[TMP31:%.*]] = load ptr, ptr [[ARRAYIDX7]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP31]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i8, ptr [[TMP31]], i64 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP33:%.*]] = sext i32 [[TMP32]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN9:%.*]] = add nsw i64 -1, [[TMP33]]
 // CHECK1-NEXT:    [[TMP34:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP34]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds ptr, ptr [[TMP34]], i64 9
 // CHECK1-NEXT:    [[TMP35:%.*]] = load ptr, ptr [[ARRAYIDX10]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP36:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX8]], ptr [[TMP36]], align 8
@@ -458,9 +458,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]
diff --git a/clang/test/OpenMP/reduction_implicit_map.cpp b/clang/test/OpenMP/reduction_implicit_map.cpp
index a7db3da7d1f868..4d2b93ffd47120 100644
--- a/clang/test/OpenMP/reduction_implicit_map.cpp
+++ b/clang/test/OpenMP/reduction_implicit_map.cpp
@@ -133,9 +133,9 @@ int main()
 // CHECK-NEXT:    store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
 // CHECK-NEXT:    store ptr [[E]], ptr [[E_ADDR]], align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[E_ADDR]], align 8
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i64 0
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[E_ADDR]], align 8
-// CHECK-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw double, ptr [[TMP1]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds double, ptr [[TMP1]], i64 0
 // CHECK-NEXT:    store double 0.000000e+00, ptr [[E2]], align 8
 // CHECK-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[E_ADDR]], align 8
 // CHECK-NEXT:    [[TMP3:%.*]] = ptrtoint ptr [[TMP2]] to i64
@@ -529,16 +529,16 @@ int main()
 // CHECK1-NEXT:    store i64 9, ptr [[DOTOMP_UB]], align 8
 // CHECK1-NEXT:    store i64 1, ptr [[DOTOMP_STRIDE]], align 8
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [10 x [10 x [10 x double]]], ptr [[TMP0]], i64 0, i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [10 x [10 x [10 x double]]], ptr [[TMP0]], i64 0, i64 0
 // CHECK1-NEXT:    [[ARRAYDECAY:%.*]] = getelementptr inbounds [10 x [10 x double]], ptr [[ARRAYIDX]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw [10 x double], ptr [[ARRAYDECAY]], i64 2
+// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [10 x double], ptr [[ARRAYDECAY]], i64 2
 // CHECK1-NEXT:    [[ARRAYDECAY2:%.*]] = getelementptr inbounds [10 x double], ptr [[ARRAYIDX1]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw double, ptr [[ARRAYDECAY2]], i64 1
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw [10 x [10 x [10 x double]]], ptr [[TMP0]], i64 0, i64 1
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds double, ptr [[ARRAYDECAY2]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds [10 x [10 x [10 x double]]], ptr [[TMP0]], i64 0, i64 1
 // CHECK1-NEXT:    [[ARRAYDECAY5:%.*]] = getelementptr inbounds [10 x [10 x double]], ptr [[ARRAYIDX4]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw [10 x double], ptr [[ARRAYDECAY5]], i64 5
+// CHECK1-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds [10 x double], ptr [[ARRAYDECAY5]], i64 5
 // CHECK1-NEXT:    [[ARRAYDECAY7:%.*]] = getelementptr inbounds [10 x double], ptr [[ARRAYIDX6]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw double, ptr [[ARRAYDECAY7]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds double, ptr [[ARRAYDECAY7]], i64 1
 // CHECK1-NEXT:    [[TMP1:%.*]] = ptrtoint ptr [[ARRAYIDX8]] to i64
 // CHECK1-NEXT:    [[TMP2:%.*]] = ptrtoint ptr [[ARRAYIDX3]] to i64
 // CHECK1-NEXT:    [[TMP3:%.*]] = sub i64 [[TMP1]], [[TMP2]]
@@ -564,18 +564,18 @@ int main()
 // CHECK1-NEXT:    [[TMP11:%.*]] = sub i64 [[TMP9]], [[TMP10]]
 // CHECK1-NEXT:    [[TMP12:%.*]] = sdiv exact i64 [[TMP11]], ptrtoint (ptr getelementptr (double, ptr null, i32 1) to i64)
 // CHECK1-NEXT:    [[TMP13:%.*]] = getelementptr double, ptr [[VLA]], i64 [[TMP12]]
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [1 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [1 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP14:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
-// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw [10 x [10 x [10 x double]]], ptr [[TMP0]], i64 0, i64 0
+// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds [10 x [10 x [10 x double]]], ptr [[TMP0]], i64 0, i64 0
 // CHECK1-NEXT:    [[ARRAYDECAY10:%.*]] = getelementptr inbounds [10 x [10 x double]], ptr [[ARRAYIDX9]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw [10 x double], ptr [[ARRAYDECAY10]], i64 2
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds [10 x double], ptr [[ARRAYDECAY10]], i64 2
 // CHECK1-NEXT:    [[ARRAYDECAY12:%.*]] = getelementptr inbounds [10 x double], ptr [[ARRAYIDX11]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw double, ptr [[ARRAYDECAY12]], i64 1
-// CHECK1-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds nuw [10 x [10 x [10 x double]]], ptr [[TMP0]], i64 0, i64 1
+// CHECK1-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds double, ptr [[ARRAYDECAY12]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds [10 x [10 x [10 x double]]], ptr [[TMP0]], i64 0, i64 1
 // CHECK1-NEXT:    [[ARRAYDECAY15:%.*]] = getelementptr inbounds [10 x [10 x double]], ptr [[ARRAYIDX14]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw [10 x double], ptr [[ARRAYDECAY15]], i64 5
+// CHECK1-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds [10 x double], ptr [[ARRAYDECAY15]], i64 5
 // CHECK1-NEXT:    [[ARRAYDECAY17:%.*]] = getelementptr inbounds [10 x double], ptr [[ARRAYIDX16]], i64 0, i64 0
-// CHECK1-NEXT:    [[ARRAYIDX18:%.*]] = getelementptr inbounds nuw double, ptr [[ARRAYDECAY17]], i64 1
+// CHECK1-NEXT:    [[ARRAYIDX18:%.*]] = getelementptr inbounds double, ptr [[ARRAYDECAY17]], i64 1
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP14]], align 8
 // CHECK1-NEXT:    [[TMP15:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX13]], ptr [[TMP15]], align 8
@@ -852,7 +852,7 @@ int main()
 // CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 0
 // CHECK2-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[INPUT_ADDR]], align 4
 // CHECK2-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[INPUT_ADDR]], align 4
-// CHECK2-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP7]], i32 0
+// CHECK2-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i32, ptr [[TMP7]], i32 0
 // CHECK2-NEXT:    [[TMP8:%.*]] = load i32, ptr [[SIZE_ADDR]], align 4
 // CHECK2-NEXT:    [[TMP9:%.*]] = mul nuw i32 [[TMP8]], 4
 // CHECK2-NEXT:    [[TMP10:%.*]] = sext i32 [[TMP9]] to i64
@@ -930,10 +930,10 @@ int main()
 // CHECK2-NEXT:    [[TMP46:%.*]] = load ptr, ptr [[INPUT_ADDR]], align 4
 // CHECK2-NEXT:    [[TMP47:%.*]] = load ptr, ptr [[OUTPUT_ADDR]], align 4
 // CHECK2-NEXT:    [[TMP48:%.*]] = load ptr, ptr [[OUTPUT_ADDR]], align 4
-// CHECK2-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP48]], i32 0
+// CHECK2-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds i32, ptr [[TMP48]], i32 0
 // CHECK2-NEXT:    [[TMP49:%.*]] = load ptr, ptr [[INPUT_ADDR]], align 4
 // CHECK2-NEXT:    [[TMP50:%.*]] = load ptr, ptr [[INPUT_ADDR]], align 4
-// CHECK2-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP50]], i32 0
+// CHECK2-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds i32, ptr [[TMP50]], i32 0
 // CHECK2-NEXT:    [[TMP51:%.*]] = load i32, ptr [[SIZE_ADDR]], align 4
 // CHECK2-NEXT:    [[TMP52:%.*]] = mul nuw i32 [[TMP51]], 4
 // CHECK2-NEXT:    [[TMP53:%.*]] = sext i32 [[TMP52]] to i64
@@ -1007,7 +1007,7 @@ int main()
 // CHECK2-NEXT:    [[TMP86:%.*]] = load i32, ptr [[SIZE_ADDR]], align 4
 // CHECK2-NEXT:    store i32 [[TMP86]], ptr [[SIZE_CASTED21]], align 4
 // CHECK2-NEXT:    [[TMP87:%.*]] = load i32, ptr [[SIZE_CASTED21]], align 4
-// CHECK2-NEXT:    [[ARRAYIDX22:%.*]] = getelementptr inbounds nuw [10 x i32], ptr [[A]], i32 0, i32 0
+// CHECK2-NEXT:    [[ARRAYIDX22:%.*]] = getelementptr inbounds [10 x i32], ptr [[A]], i32 0, i32 0
 // CHECK2-NEXT:    [[TMP88:%.*]] = getelementptr inbounds [2 x ptr], ptr [[DOTOFFLOAD_BASEPTRS23]], i32 0, i32 0
 // CHECK2-NEXT:    store i32 [[TMP87]], ptr [[TMP88]], align 4
 // CHECK2-NEXT:    [[TMP89:%.*]] = getelementptr inbounds [2 x ptr], ptr [[DOTOFFLOAD_PTRS24]], i32 0, i32 0
@@ -1480,9 +1480,9 @@ int main()
 // CHECK2-NEXT:    store ptr [[OUTPUT]], ptr [[OUTPUT_ADDR]], align 4
 // CHECK2-NEXT:    store ptr [[INPUT]], ptr [[INPUT_ADDR]], align 4
 // CHECK2-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[OUTPUT_ADDR]], align 4
-// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP0]], i32 0
+// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP0]], i32 0
 // CHECK2-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[OUTPUT_ADDR]], align 4
-// CHECK2-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 2
+// CHECK2-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 2
 // CHECK2-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [3 x i32], ptr [[OUTPUT2]], i32 0, i32 0
 // CHECK2-NEXT:    [[TMP2:%.*]] = getelementptr i32, ptr [[ARRAY_BEGIN]], i32 3
 // CHECK2-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP2]]
@@ -1664,9 +1664,9 @@ int main()
 // CHECK2-NEXT:    store i32 1, ptr [[DOTOMP_STRIDE]], align 4
 // CHECK2-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK2-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[OUTPUT_ADDR]], align 4
-// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP6]], i32 0
+// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP6]], i32 0
 // CHECK2-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[OUTPUT_ADDR]], align 4
-// CHECK2-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP7]], i32 2
+// CHECK2-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds i32, ptr [[TMP7]], i32 2
 // CHECK2-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [3 x i32], ptr [[OUTPUT4]], i32 0, i32 0
 // CHECK2-NEXT:    [[TMP8:%.*]] = getelementptr i32, ptr [[ARRAY_BEGIN]], i32 3
 // CHECK2-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP8]]
@@ -1878,8 +1878,8 @@ int main()
 // CHECK2-NEXT:    store i32 [[SIZE]], ptr [[SIZE_ADDR]], align 4
 // CHECK2-NEXT:    store ptr [[A]], ptr [[A_ADDR]], align 4
 // CHECK2-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[A_ADDR]], align 4
-// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw [10 x i32], ptr [[TMP0]], i32 0, i32 0
-// CHECK2-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw [10 x i32], ptr [[TMP0]], i32 0, i32 1
+// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [10 x i32], ptr [[TMP0]], i32 0, i32 0
+// CHECK2-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [10 x i32], ptr [[TMP0]], i32 0, i32 1
 // CHECK2-NEXT:    [[ARRAY_BEGIN:%.*]] = getelementptr inbounds [2 x i32], ptr [[A2]], i32 0, i32 0
 // CHECK2-NEXT:    [[TMP1:%.*]] = getelementptr i32, ptr [[ARRAY_BEGIN]], i32 2
 // CHECK2-NEXT:    [[OMP_ARRAYINIT_ISEMPTY:%.*]] = icmp eq ptr [[ARRAY_BEGIN]], [[TMP1]]
diff --git a/clang/test/OpenMP/sections_reduction_task_codegen.cpp b/clang/test/OpenMP/sections_reduction_task_codegen.cpp
index 5d749eeb817767..1a2cf7aede3218 100644
--- a/clang/test/OpenMP/sections_reduction_task_codegen.cpp
+++ b/clang/test/OpenMP/sections_reduction_task_codegen.cpp
@@ -82,16 +82,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_SECTIONS_IL_]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP3]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP3]], i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP5:%.*]] = sext i32 [[TMP4]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP5]]
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP6]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP6]], i64 9
 // CHECK1-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP7]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP7]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP9:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP10:%.*]] = sub i64 [[TMP8]], [[TMP9]]
@@ -121,7 +121,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP21]]
 // CHECK1-NEXT:    store ptr [[_TMP5]], ptr [[TMP]], align 8
 // CHECK1-NEXT:    store ptr [[TMP22]], ptr [[_TMP5]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP23]], align 8
 // CHECK1-NEXT:    [[TMP24:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -136,19 +136,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP28]], align 8
 // CHECK1-NEXT:    [[TMP29:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP29]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP30:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP31:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP31]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds ptr, ptr [[TMP31]], i64 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load ptr, ptr [[ARRAYIDX7]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP32]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i8, ptr [[TMP32]], i64 0
 // CHECK1-NEXT:    [[TMP33:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP34:%.*]] = sext i32 [[TMP33]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN9:%.*]] = add nsw i64 -1, [[TMP34]]
 // CHECK1-NEXT:    [[TMP35:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP35]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds ptr, ptr [[TMP35]], i64 9
 // CHECK1-NEXT:    [[TMP36:%.*]] = load ptr, ptr [[ARRAYIDX10]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP36]], i64 [[LB_ADD_LEN9]]
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds i8, ptr [[TMP36]], i64 [[LB_ADD_LEN9]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP30]], align 8
 // CHECK1-NEXT:    [[TMP37:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX8]], ptr [[TMP37]], align 8
@@ -463,9 +463,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]
diff --git a/clang/test/OpenMP/target_data_use_device_addr_codegen.cpp b/clang/test/OpenMP/target_data_use_device_addr_codegen.cpp
index 19a523ee165c8d..d912f801a33dbb 100644
--- a/clang/test/OpenMP/target_data_use_device_addr_codegen.cpp
+++ b/clang/test/OpenMP/target_data_use_device_addr_codegen.cpp
@@ -54,12 +54,12 @@ int main() {
 // CHECK: [[SIZES:%.+]] = alloca [6 x i64],
 // CHECK: [[VLA_ADDR:%.+]] = alloca float, i64 %{{.+}},
 // CHECK: [[PTR:%.+]] = load ptr, ptr [[PTR_ADDR]],
-// CHECK-NEXT: [[ARR_IDX:%.+]] = getelementptr inbounds nuw float, ptr [[PTR]], i64 3
+// CHECK-NEXT: [[ARR_IDX:%.+]] = getelementptr inbounds float, ptr [[PTR]], i64 3
 // CHECK: [[P5:%.+]] = load ptr, ptr [[PTR_ADDR]], align 8
 // CHECK-NEXT: [[ARR_IDX1:%.+]] = getelementptr inbounds float, ptr [[P5]], i64 0
 // CHECK: [[P7:%.+]] = load ptr, ptr [[REF_ADDR]],
 // CHECK-NEXT: [[REF:%.+]] = load ptr, ptr [[REF_ADDR]],
-// CHECK-NEXT: [[ARR_IDX2:%.+]] = getelementptr inbounds nuw [4 x float], ptr [[ARR_ADDR]], i64 0, i64 0
+// CHECK-NEXT: [[ARR_IDX2:%.+]] = getelementptr inbounds [4 x float], ptr [[ARR_ADDR]], i64 0, i64 0
 // CHECK: [[P10:%.+]] = mul nuw i64 {{.+}}, 4
 // CHECK-NEXT: [[ARR_IDX5:%.+]] = getelementptr inbounds float, ptr [[VLA_ADDR]], i64 0
 // CHECK-NEXT: call void @llvm.memcpy.p0.p0.i64(ptr align 8 [[SIZES]], ptr align 8 [[SIZES1]], i64 48, i1 false)
@@ -132,14 +132,14 @@ int main() {
 // CHECK: [[SIZES:%.+]] = alloca [6 x i64],
 // CHECK: [[A_ADDR:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[THIS:%.+]], i32 0, i32 0
 // CHECK: [[PTR_ADDR:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[THIS]], i32 0, i32 1
-// CHECK: [[ARR_IDX:%.+]] = getelementptr inbounds nuw i32, ptr %{{.+}}, i64 3
+// CHECK: [[ARR_IDX:%.+]] = getelementptr inbounds i32, ptr %{{.+}}, i64 3
 // CHECK: [[REF_REF:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[THIS]], i32 0, i32 2
 // CHECK: [[REF_PTR:%.+]] = load ptr, ptr [[REF_REF]],
 // CHECK-NEXT: [[P3:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[THIS]], i32 0, i32 1
 // CHECK: [[ARR_IDX5:%.+]] = getelementptr inbounds i32, ptr {{.+}}, i64 0
 // CHECK: [[ARR_ADDR:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[THIS]], i32 0, i32 3
 
-// CHECK: [[ARR_IDX6:%.+]] = getelementptr inbounds nuw [4 x i32], ptr [[ARR_ADDR]], i64 0, i64 0
+// CHECK: [[ARR_IDX6:%.+]] = getelementptr inbounds [4 x i32], ptr [[ARR_ADDR]], i64 0, i64 0
 // CHECK: [[A_ADDR2:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[THIS]], i32 0, i32 0
 // CHECK: [[P4:%.+]] = mul nuw i64 [[CONV:%.+]], 4
 // CHECK: [[A_ADDR3:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[THIS]], i32 0, i32 0
@@ -147,7 +147,7 @@ int main() {
 // CHECK: [[L6:%.+]] = sext i32 [[L5]] to i64
 // CHECK: [[LB_ADD_LEN:%lb_add_len]] = add nsw i64 -1, [[L6]]
 // CHECK: [[ARR_ADDR9:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[THIS]], i32 0, i32 3
-// CHECK: [[ARR_IDX10:%arrayidx.+]] = getelementptr inbounds nuw [4 x i32], ptr [[ARR_ADDR9]], i64 0, i64 %lb_add_len
+// CHECK: [[ARR_IDX10:%arrayidx.+]] = getelementptr inbounds [4 x i32], ptr [[ARR_ADDR9]], i64 0, i64 %lb_add_len
 // CHECK: [[ARR_END:%.+]] = getelementptr i32, ptr [[ARR_IDX10]], i32 1
 // CHECK: [[E:%.+]] = ptrtoint ptr [[ARR_END]] to i64
 // CHECK: [[B:%.+]] = ptrtoint ptr [[A_ADDR]] to i64
diff --git a/clang/test/OpenMP/target_data_use_device_ptr_codegen.cpp b/clang/test/OpenMP/target_data_use_device_ptr_codegen.cpp
index 6d1c0213d648ce..c90819dc2a22f4 100644
--- a/clang/test/OpenMP/target_data_use_device_ptr_codegen.cpp
+++ b/clang/test/OpenMP/target_data_use_device_ptr_codegen.cpp
@@ -49,14 +49,14 @@ void foo(float *&lr, T *&tr) {
   // CK1-NOT: store ptr [[VAL]], ptr [[DECL]],
   // CK1:     store ptr [[VAL]], ptr [[PVT:%.+]],
   // CK1:     [[TT:%.+]] = load ptr, ptr [[PVT]],
-  // CK1:     getelementptr inbounds nuw double, ptr [[TT]], i32 1
+  // CK1:     getelementptr inbounds double, ptr [[TT]], i32 1
   #pragma omp target data map(g[:10]) use_device_ptr(g)
   {
     ++g;
   }
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE00]]
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-  // CK1:     getelementptr inbounds nuw double, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds double, ptr [[TTT]], i32 1
   ++g;
 
   // CK1:     [[T1:%.+]] = load ptr, ptr [[DECL:%.+]],
@@ -67,26 +67,26 @@ void foo(float *&lr, T *&tr) {
   // CK1-NOT: store ptr [[VAL]], ptr [[DECL]],
   // CK1:     store ptr [[VAL]], ptr [[PVT:%.+]],
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TT1]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TT1]], i32 1
   #pragma omp target data map(l[:10]) use_device_ptr(l)
   {
     ++l;
   }
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE01]]
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TTT]], i32 1
   ++l;
 
   // CK1-NOT: call void @__tgt_target
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TTT]], i32 1
   #pragma omp target data map(l[:10]) use_device_ptr(l) if(0)
   {
     ++l;
   }
   // CK1-NOT: call void @__tgt_target
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TTT]], i32 1
   ++l;
 
   // CK1:     [[T1:%.+]] = load ptr, ptr [[DECL:%.+]],
@@ -97,14 +97,14 @@ void foo(float *&lr, T *&tr) {
   // CK1-NOT: store ptr [[VAL]], ptr [[DECL]],
   // CK1:     store ptr [[VAL]], ptr [[PVT:%.+]],
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TT1]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TT1]], i32 1
   #pragma omp target data map(l[:10]) use_device_ptr(l) if(1)
   {
     ++l;
   }
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE03]]
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TTT]], i32 1
   ++l;
 
   // CK1:     [[CMP:%.+]] = icmp ne ptr %{{.+}}, null
@@ -119,12 +119,12 @@ void foo(float *&lr, T *&tr) {
   // CK1-NOT: store ptr [[VAL]], ptr [[DECL]],
   // CK1:     store ptr [[VAL]], ptr [[PVT:%.+]],
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TT1]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TT1]], i32 1
   // CK1:     br label %[[BEND:.+]]
 
   // CK1:     [[BELSE]]:
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TTT]], i32 1
   // CK1:     br label %[[BEND]]
   #pragma omp target data map(l[:10]) use_device_ptr(l) if(lr != 0)
   {
@@ -142,7 +142,7 @@ void foo(float *&lr, T *&tr) {
 
   // CK1:     [[BEND]]:
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TTT]], i32 1
   ++l;
 
   // CK1:     [[T2:%.+]] = load ptr, ptr [[DECL:%.+]],
@@ -156,7 +156,7 @@ void foo(float *&lr, T *&tr) {
   // CK1:     store ptr [[PVTV]], ptr [[PVT:%.+]],
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
   // CK1:     [[TT2:%.+]] = load ptr, ptr [[TT1]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TT2]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TT2]], i32 1
   #pragma omp target data map(lr[:10]) use_device_ptr(lr)
   {
     ++lr;
@@ -164,7 +164,7 @@ void foo(float *&lr, T *&tr) {
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE05]]
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
   // CK1:     [[TTTT:%.+]] = load ptr, ptr [[TTT]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TTTT]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TTTT]], i32 1
   ++lr;
 
   // CK1:     [[T1:%.+]] = load ptr, ptr [[DECL:%.+]],
@@ -175,14 +175,14 @@ void foo(float *&lr, T *&tr) {
   // CK1-NOT: store ptr [[VAL]], ptr [[DECL]],
   // CK1:     store ptr [[VAL]], ptr [[PVT:%.+]],
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TT1]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TT1]], i32 1
   #pragma omp target data map(t[:10]) use_device_ptr(t)
   {
     ++t;
   }
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE06]]
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TTT]], i32 1
   ++t;
 
   // CK1:     [[T2:%.+]] = load ptr, ptr [[DECL:%.+]],
@@ -196,7 +196,7 @@ void foo(float *&lr, T *&tr) {
   // CK1:     store ptr [[PVTV]], ptr [[PVT:%.+]],
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
   // CK1:     [[TT2:%.+]] = load ptr, ptr [[TT1]],
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TT2]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TT2]], i32 1
   #pragma omp target data map(tr[:10]) use_device_ptr(tr)
   {
     ++tr;
@@ -204,7 +204,7 @@ void foo(float *&lr, T *&tr) {
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE07]]
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
   // CK1:     [[TTTT:%.+]] = load ptr, ptr [[TTT]],
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TTTT]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TTTT]], i32 1
   ++tr;
 
   // CK1:     [[T1:%.+]] = load ptr, ptr [[DECL:%.+]],
@@ -215,14 +215,14 @@ void foo(float *&lr, T *&tr) {
   // CK1-NOT: store ptr [[VAL]], ptr [[DECL]],
   // CK1:     store ptr [[VAL]], ptr [[PVT:%.+]],
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TT1]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TT1]], i32 1
   #pragma omp target data map(l[:10], t[:10]) use_device_ptr(l)
   {
     ++l; ++t;
   }
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE08]]
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[TTT]], i32 1
   ++l; ++t;
 
 
@@ -232,18 +232,18 @@ void foo(float *&lr, T *&tr) {
   // CK1:     [[VAL:%.+]] = load ptr, ptr {{%.+}},
   // CK1:     store ptr [[VAL]], ptr [[PVT:%.+]],
   // CK1:     [[_TT1:%.+]] = load ptr, ptr [[_PVT]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[_TT1]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[_TT1]], i32 1
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TT1]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TT1]], i32 1
   #pragma omp target data map(l[:10], t[:10]) use_device_ptr(l) use_device_ptr(t)
   {
     ++l; ++t;
   }
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE09]]
   // CK1:     [[_TTT:%.+]] = load ptr, ptr {{%.+}},
-  // CK1:     getelementptr inbounds nuw float, ptr [[_TTT]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[_TTT]], i32 1
   // CK1:     [[TTT:%.+]] = load ptr, ptr {{%.+}},
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TTT]], i32 1
   ++l; ++t;
 
   // CK1:     call void @__tgt_target_data_begin{{.+}}[[MTYPE10]]
@@ -252,18 +252,18 @@ void foo(float *&lr, T *&tr) {
   // CK1:     [[VAL:%.+]] = load ptr, ptr {{%.+}},
   // CK1:     store ptr [[VAL]], ptr [[PVT:%.+]],
   // CK1:     [[_TT1:%.+]] = load ptr, ptr [[_PVT]],
-  // CK1:     getelementptr inbounds nuw float, ptr [[_TT1]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[_TT1]], i32 1
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TT1]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TT1]], i32 1
   #pragma omp target data map(l[:10], t[:10]) use_device_ptr(l,t)
   {
     ++l; ++t;
   }
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE10]]
   // CK1:     [[_TTT:%.+]] = load ptr, ptr {{%.+}},
-  // CK1:     getelementptr inbounds nuw float, ptr [[_TTT]], i32 1
+  // CK1:     getelementptr inbounds float, ptr [[_TTT]], i32 1
   // CK1:     [[TTT:%.+]] = load ptr, ptr {{%.+}},
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TTT]], i32 1
   ++l; ++t;
 
   // CK1:     [[T1:%.+]] = load ptr, ptr [[DECL:%.+]],
@@ -274,14 +274,14 @@ void foo(float *&lr, T *&tr) {
   // CK1-NOT: store ptr [[VAL]], ptr [[DECL]],
   // CK1:     store ptr [[VAL]], ptr [[PVT:%.+]],
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TT1]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TT1]], i32 1
   #pragma omp target data map(l[:10]) use_device_ptr(t)
   {
     ++l; ++t;
   }
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE11]]
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TTT]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TTT]], i32 1
   ++l; ++t;
 
   // CK1:     [[T2:%.+]] = load ptr, ptr [[DECL:%.+]],
@@ -295,7 +295,7 @@ void foo(float *&lr, T *&tr) {
   // CK1:     store ptr [[PVTV]], ptr [[PVT:%.+]],
   // CK1:     [[TT1:%.+]] = load ptr, ptr [[PVT]],
   // CK1:     [[TT2:%.+]] = load ptr, ptr [[TT1]],
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TT2]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TT2]], i32 1
   #pragma omp target data map(l[:10]) use_device_ptr(tr)
   {
     ++l; ++tr;
@@ -303,7 +303,7 @@ void foo(float *&lr, T *&tr) {
   // CK1:     call void @__tgt_target_data_end{{.+}}[[MTYPE12]]
   // CK1:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
   // CK1:     [[TTTT:%.+]] = load ptr, ptr [[TTT]],
-  // CK1:     getelementptr inbounds nuw i32, ptr [[TTTT]], i32 1
+  // CK1:     getelementptr inbounds i32, ptr [[TTTT]], i32 1
   ++l; ++tr;
 
 }
@@ -354,7 +354,7 @@ struct ST {
     // CK2:     store ptr [[PVT]], ptr [[PVT2:%.+]],
     // CK2:     [[TT1:%.+]] = load ptr, ptr [[PVT2]],
     // CK2:     [[TT2:%.+]] = load ptr, ptr [[TT1]],
-    // CK2:     getelementptr inbounds nuw double, ptr [[TT2]], i32 1
+    // CK2:     getelementptr inbounds double, ptr [[TT2]], i32 1
     #pragma omp target data map(a[:10]) use_device_ptr(a)
     {
       a++;
@@ -362,7 +362,7 @@ struct ST {
     // CK2:     call void @__tgt_target_data_end{{.+}}[[MTYPE00]]
     // CK2:     [[DECL:%.+]] = getelementptr inbounds nuw [[ST]], ptr %this1, i32 0, i32 0
     // CK2:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-    // CK2:     getelementptr inbounds nuw double, ptr [[TTT]], i32 1
+    // CK2:     getelementptr inbounds double, ptr [[TTT]], i32 1
     a++;
 
     // CK2:     [[BP:%.+]] = getelementptr inbounds [2 x ptr], ptr %{{.+}}, i32 0, i32 1
@@ -373,7 +373,7 @@ struct ST {
     // CK2:     store ptr [[PVT]], ptr [[PVT2:%.+]],
     // CK2:     [[TT1:%.+]] = load ptr, ptr [[PVT2]],
     // CK2:     [[TT2:%.+]] = load ptr, ptr [[TT1]],
-    // CK2:     getelementptr inbounds nuw double, ptr [[TT2]], i32 1
+    // CK2:     getelementptr inbounds double, ptr [[TT2]], i32 1
     #pragma omp target data map(b[:10]) use_device_ptr(b)
     {
       b++;
@@ -382,7 +382,7 @@ struct ST {
     // CK2:     [[DECL:%.+]] = getelementptr inbounds nuw [[ST]], ptr %{{.+}}, i32 0, i32 1
     // CK2:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
     // CK2:     [[TTTT:%.+]] = load ptr, ptr [[TTT]],
-    // CK2:     getelementptr inbounds nuw double, ptr [[TTTT]], i32 1
+    // CK2:     getelementptr inbounds double, ptr [[TTTT]], i32 1
     b++;
 
     // CK2:     [[BP:%.+]] = getelementptr inbounds [3 x ptr], ptr %{{.+}}, i32 0, i32 2
@@ -393,7 +393,7 @@ struct ST {
     // CK2:     store ptr [[PVT]], ptr [[PVT2:%.+]],
     // CK2:     [[TT1:%.+]] = load ptr, ptr [[PVT2]],
     // CK2:     [[TT2:%.+]] = load ptr, ptr [[TT1]],
-    // CK2:     getelementptr inbounds nuw double, ptr [[TT2]], i32 1
+    // CK2:     getelementptr inbounds double, ptr [[TT2]], i32 1
     #pragma omp target data map(la[:10]) use_device_ptr(a)
     {
       a++;
@@ -402,7 +402,7 @@ struct ST {
     // CK2:     call void @__tgt_target_data_end{{.+}}[[MTYPE02]]
     // CK2:     [[DECL:%.+]] = getelementptr inbounds nuw [[ST]], ptr %this1, i32 0, i32 0
     // CK2:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-    // CK2:     getelementptr inbounds nuw double, ptr [[TTT]], i32 1
+    // CK2:     getelementptr inbounds double, ptr [[TTT]], i32 1
     a++;
     la++;
 
@@ -419,10 +419,10 @@ struct ST {
     // CK2:     store ptr [[PVT1]], ptr [[_PVT1:%.+]],
     // CK2:     [[TT2:%.+]] = load ptr, ptr [[_PVT2]],
     // CK2:     [[_TT2:%.+]] = load ptr, ptr [[TT2]],
-    // CK2:     getelementptr inbounds nuw double, ptr [[_TT2]], i32 1
+    // CK2:     getelementptr inbounds double, ptr [[_TT2]], i32 1
     // CK2:     [[TT1:%.+]] = load ptr, ptr [[_PVT1]],
     // CK2:     [[_TT1:%.+]] = load ptr, ptr [[TT1]],
-    // CK2:     getelementptr inbounds nuw double, ptr [[_TT1]], i32 1
+    // CK2:     getelementptr inbounds double, ptr [[_TT1]], i32 1
     #pragma omp target data map(b[:10]) use_device_ptr(a, b)
     {
       a++;
@@ -431,11 +431,11 @@ struct ST {
     // CK2:     call void @__tgt_target_data_end{{.+}}[[MTYPE03]]
     // CK2:     [[DECL:%.+]] = getelementptr inbounds nuw [[ST]], ptr %this1, i32 0, i32 0
     // CK2:     [[TTT:%.+]] = load ptr, ptr [[DECL]],
-    // CK2:     getelementptr inbounds nuw double, ptr [[TTT]], i32 1
+    // CK2:     getelementptr inbounds double, ptr [[TTT]], i32 1
     // CK2:     [[_DECL:%.+]] = getelementptr inbounds nuw [[ST]], ptr %this1, i32 0, i32 1
     // CK2:     [[_TTT:%.+]] = load ptr, ptr [[_DECL]],
     // CK2:     [[_TTTT:%.+]] = load ptr, ptr [[_TTT]],
-    // CK2:     getelementptr inbounds nuw double, ptr [[_TTTT]], i32 1
+    // CK2:     getelementptr inbounds double, ptr [[_TTTT]], i32 1
     a++;
     b++;
   }
diff --git a/clang/test/OpenMP/target_has_device_addr_codegen.cpp b/clang/test/OpenMP/target_has_device_addr_codegen.cpp
index 39eaedb0e48d1c..08bcc87ca5f0a9 100644
--- a/clang/test/OpenMP/target_has_device_addr_codegen.cpp
+++ b/clang/test/OpenMP/target_has_device_addr_codegen.cpp
@@ -586,7 +586,7 @@ void use_template() {
 // CHECK-NEXT:    store ptr [[K]], ptr [[K_ADDR]], align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[K_ADDR]], align 8
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CHECK-NEXT:    ret void
 //
@@ -601,7 +601,7 @@ void use_template() {
 // CHECK-NEXT:    store ptr [[TMP0]], ptr [[TMP]], align 8
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP]], align 8
 // CHECK-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP2]], i32 1
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i32 1
 // CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP1]], align 8
 // CHECK-NEXT:    ret void
 //
@@ -1079,7 +1079,7 @@ void use_template() {
 // CHECK-NEXT:    store ptr [[K]], ptr [[K_ADDR]], align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[K_ADDR]], align 8
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CHECK-NEXT:    ret void
 //
@@ -1094,7 +1094,7 @@ void use_template() {
 // CHECK-NEXT:    store ptr [[TMP0]], ptr [[TMP]], align 8
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP]], align 8
 // CHECK-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP2]], i32 1
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i32 1
 // CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP1]], align 8
 // CHECK-NEXT:    ret void
 //
@@ -1133,7 +1133,7 @@ void use_template() {
 // CHECK-NEXT:    store ptr [[K]], ptr [[K_ADDR]], align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[K_ADDR]], align 8
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i32 1
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i32 1
 // CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CHECK-NEXT:    ret void
 //
@@ -1148,7 +1148,7 @@ void use_template() {
 // CHECK-NEXT:    store ptr [[TMP0]], ptr [[TMP]], align 8
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP]], align 8
 // CHECK-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP2]], i32 1
+// CHECK-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds ptr, ptr [[TMP2]], i32 1
 // CHECK-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP1]], align 8
 // CHECK-NEXT:    ret void
 //
@@ -1422,14 +1422,14 @@ void use_template() {
 // SIMD-ONLY0-NEXT:    store ptr [[K]], ptr [[Z]], align 8
 // SIMD-ONLY0-NEXT:    store ptr [[AA]], ptr [[RAA]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[K]], align 8
-// SIMD-ONLY0-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// SIMD-ONLY0-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // SIMD-ONLY0-NEXT:    store ptr [[INCDEC_PTR]], ptr [[K]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[Z]], align 8
 // SIMD-ONLY0-NEXT:    store ptr [[TMP2]], ptr [[TMP]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[Z]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[TMP]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP4]], align 8
-// SIMD-ONLY0-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP5]], i32 1
+// SIMD-ONLY0-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 1
 // SIMD-ONLY0-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[TMP4]], align 8
 // SIMD-ONLY0-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [10 x i32], ptr [[AA]], i64 0, i64 0
 // SIMD-ONLY0-NEXT:    store i32 1, ptr [[ARRAYIDX]], align 4
@@ -1478,14 +1478,14 @@ void use_template() {
 // SIMD-ONLY0-NEXT:    store ptr [[TMP0]], ptr [[K]], align 8
 // SIMD-ONLY0-NEXT:    store ptr [[K]], ptr [[Z]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[K]], align 8
-// SIMD-ONLY0-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// SIMD-ONLY0-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // SIMD-ONLY0-NEXT:    store ptr [[INCDEC_PTR]], ptr [[K]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[Z]], align 8
 // SIMD-ONLY0-NEXT:    store ptr [[TMP2]], ptr [[TMP]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[Z]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[TMP]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP4]], align 8
-// SIMD-ONLY0-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP5]], i32 1
+// SIMD-ONLY0-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 1
 // SIMD-ONLY0-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[TMP4]], align 8
 // SIMD-ONLY0-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [10 x i32], ptr [[AA]], i64 0, i64 0
 // SIMD-ONLY0-NEXT:    [[TMP6:%.*]] = load i32, ptr [[ARRAYIDX]], align 4
@@ -1520,14 +1520,14 @@ void use_template() {
 // SIMD-ONLY0-NEXT:    store ptr [[TMP0]], ptr [[K]], align 8
 // SIMD-ONLY0-NEXT:    store ptr [[K]], ptr [[Z]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[K]], align 8
-// SIMD-ONLY0-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i32 1
+// SIMD-ONLY0-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i32 1
 // SIMD-ONLY0-NEXT:    store ptr [[INCDEC_PTR]], ptr [[K]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[Z]], align 8
 // SIMD-ONLY0-NEXT:    store ptr [[TMP2]], ptr [[TMP]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[Z]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[TMP]], align 8
 // SIMD-ONLY0-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP4]], align 8
-// SIMD-ONLY0-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP5]], i32 1
+// SIMD-ONLY0-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds ptr, ptr [[TMP5]], i32 1
 // SIMD-ONLY0-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[TMP4]], align 8
 // SIMD-ONLY0-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [10 x ptr], ptr [[AA]], i64 0, i64 0
 // SIMD-ONLY0-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
diff --git a/clang/test/OpenMP/target_in_reduction_codegen.cpp b/clang/test/OpenMP/target_in_reduction_codegen.cpp
index 56191ee5751366..fb715e2de2a592 100644
--- a/clang/test/OpenMP/target_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/target_in_reduction_codegen.cpp
@@ -70,7 +70,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[VLA:%.*]] = alloca i16, i64 [[TMP2]], align 16
 // CHECK1-NEXT:    store i64 [[TMP2]], ptr [[__VLA_EXPR0]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[A]], ptr [[TMP4]], align 8
 // CHECK1-NEXT:    [[TMP6:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -85,7 +85,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP11]], align 8
 // CHECK1-NEXT:    [[TMP12:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP12]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP14:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[B]], ptr [[TMP14]], align 8
 // CHECK1-NEXT:    [[TMP16:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 1
@@ -100,7 +100,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..2, ptr [[TMP21]], align 8
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP22]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
 // CHECK1-NEXT:    [[TMP24:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC_ADDR]], ptr [[TMP24]], align 8
 // CHECK1-NEXT:    [[TMP26:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 1
@@ -118,7 +118,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP35:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0]], i32 3, ptr [[DOTRD_INPUT_]])
 // CHECK1-NEXT:    store ptr [[TMP35]], ptr [[DOTTASK_RED_]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP36:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0:%.*]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[C]], ptr [[TMP36]], align 8
 // CHECK1-NEXT:    [[TMP38:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 1
@@ -133,7 +133,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..6, ptr [[TMP43]], align 8
 // CHECK1-NEXT:    [[TMP44:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP44]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP46:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP46]], align 8
 // CHECK1-NEXT:    [[TMP48:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 1
diff --git a/clang/test/OpenMP/target_is_device_ptr_codegen.cpp b/clang/test/OpenMP/target_is_device_ptr_codegen.cpp
index 505c34e21733ca..3a1c168533c377 100644
--- a/clang/test/OpenMP/target_is_device_ptr_codegen.cpp
+++ b/clang/test/OpenMP/target_is_device_ptr_codegen.cpp
@@ -2142,7 +2142,7 @@ void bar() {
 // CK10-NEXT:    [[G_ADDR:%.*]] = alloca ptr, align 8
 // CK10-NEXT:    store ptr [[G]], ptr [[G_ADDR]], align 8
 // CK10-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[G_ADDR]], align 8
-// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // CK10-NEXT:    store ptr [[INCDEC_PTR]], ptr [[G_ADDR]], align 8
 // CK10-NEXT:    ret void
 //
@@ -2153,7 +2153,7 @@ void bar() {
 // CK10-NEXT:    [[L_ADDR:%.*]] = alloca ptr, align 8
 // CK10-NEXT:    store ptr [[L]], ptr [[L_ADDR]], align 8
 // CK10-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[L_ADDR]], align 8
-// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw float, ptr [[TMP0]], i32 1
+// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds float, ptr [[TMP0]], i32 1
 // CK10-NEXT:    store ptr [[INCDEC_PTR]], ptr [[L_ADDR]], align 8
 // CK10-NEXT:    ret void
 //
@@ -2164,7 +2164,7 @@ void bar() {
 // CK10-NEXT:    [[T_ADDR:%.*]] = alloca ptr, align 8
 // CK10-NEXT:    store ptr [[T]], ptr [[T_ADDR]], align 8
 // CK10-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[T_ADDR]], align 8
-// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP0]], i32 1
+// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP0]], i32 1
 // CK10-NEXT:    store ptr [[INCDEC_PTR]], ptr [[T_ADDR]], align 8
 // CK10-NEXT:    ret void
 //
@@ -2178,7 +2178,7 @@ void bar() {
 // CK10-NEXT:    store ptr [[LR_ADDR]], ptr [[TMP]], align 8
 // CK10-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 8
 // CK10-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw float, ptr [[TMP1]], i32 1
+// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 1
 // CK10-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CK10-NEXT:    ret void
 //
@@ -2192,7 +2192,7 @@ void bar() {
 // CK10-NEXT:    store ptr [[TR_ADDR]], ptr [[TMP]], align 8
 // CK10-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 8
 // CK10-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK10-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CK10-NEXT:    ret void
 //
@@ -2206,7 +2206,7 @@ void bar() {
 // CK10-NEXT:    store ptr [[TR_ADDR]], ptr [[TMP]], align 8
 // CK10-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 8
 // CK10-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK10-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CK10-NEXT:    ret void
 //
@@ -2224,11 +2224,11 @@ void bar() {
 // CK10-NEXT:    store ptr [[LR_ADDR]], ptr [[_TMP1]], align 8
 // CK10-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 8
 // CK10-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK10-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CK10-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[_TMP1]], align 8
 // CK10-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP2]], align 8
-// CK10-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP3]], i32 1
+// CK10-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds float, ptr [[TMP3]], i32 1
 // CK10-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[TMP2]], align 8
 // CK10-NEXT:    ret void
 //
@@ -2613,7 +2613,7 @@ void bar() {
 // CK11-NEXT:    [[G_ADDR:%.*]] = alloca ptr, align 8
 // CK11-NEXT:    store ptr [[G]], ptr [[G_ADDR]], align 8
 // CK11-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[G_ADDR]], align 8
-// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // CK11-NEXT:    store ptr [[INCDEC_PTR]], ptr [[G_ADDR]], align 8
 // CK11-NEXT:    ret void
 //
@@ -2624,7 +2624,7 @@ void bar() {
 // CK11-NEXT:    [[L_ADDR:%.*]] = alloca ptr, align 8
 // CK11-NEXT:    store ptr [[L]], ptr [[L_ADDR]], align 8
 // CK11-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[L_ADDR]], align 8
-// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw float, ptr [[TMP0]], i32 1
+// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds float, ptr [[TMP0]], i32 1
 // CK11-NEXT:    store ptr [[INCDEC_PTR]], ptr [[L_ADDR]], align 8
 // CK11-NEXT:    ret void
 //
@@ -2635,7 +2635,7 @@ void bar() {
 // CK11-NEXT:    [[T_ADDR:%.*]] = alloca ptr, align 8
 // CK11-NEXT:    store ptr [[T]], ptr [[T_ADDR]], align 8
 // CK11-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[T_ADDR]], align 8
-// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP0]], i32 1
+// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP0]], i32 1
 // CK11-NEXT:    store ptr [[INCDEC_PTR]], ptr [[T_ADDR]], align 8
 // CK11-NEXT:    ret void
 //
@@ -2649,7 +2649,7 @@ void bar() {
 // CK11-NEXT:    store ptr [[LR_ADDR]], ptr [[TMP]], align 8
 // CK11-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 8
 // CK11-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw float, ptr [[TMP1]], i32 1
+// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 1
 // CK11-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CK11-NEXT:    ret void
 //
@@ -2663,7 +2663,7 @@ void bar() {
 // CK11-NEXT:    store ptr [[TR_ADDR]], ptr [[TMP]], align 8
 // CK11-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 8
 // CK11-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK11-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CK11-NEXT:    ret void
 //
@@ -2677,7 +2677,7 @@ void bar() {
 // CK11-NEXT:    store ptr [[TR_ADDR]], ptr [[TMP]], align 8
 // CK11-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 8
 // CK11-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK11-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CK11-NEXT:    ret void
 //
@@ -2695,11 +2695,11 @@ void bar() {
 // CK11-NEXT:    store ptr [[LR_ADDR]], ptr [[_TMP1]], align 8
 // CK11-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 8
 // CK11-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 8
-// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK11-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 8
 // CK11-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[_TMP1]], align 8
 // CK11-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP2]], align 8
-// CK11-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP3]], i32 1
+// CK11-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds float, ptr [[TMP3]], i32 1
 // CK11-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[TMP2]], align 8
 // CK11-NEXT:    ret void
 //
@@ -3084,7 +3084,7 @@ void bar() {
 // CK12-NEXT:    [[G_ADDR:%.*]] = alloca ptr, align 4
 // CK12-NEXT:    store ptr [[G]], ptr [[G_ADDR]], align 4
 // CK12-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[G_ADDR]], align 4
-// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // CK12-NEXT:    store ptr [[INCDEC_PTR]], ptr [[G_ADDR]], align 4
 // CK12-NEXT:    ret void
 //
@@ -3095,7 +3095,7 @@ void bar() {
 // CK12-NEXT:    [[L_ADDR:%.*]] = alloca ptr, align 4
 // CK12-NEXT:    store ptr [[L]], ptr [[L_ADDR]], align 4
 // CK12-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[L_ADDR]], align 4
-// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw float, ptr [[TMP0]], i32 1
+// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds float, ptr [[TMP0]], i32 1
 // CK12-NEXT:    store ptr [[INCDEC_PTR]], ptr [[L_ADDR]], align 4
 // CK12-NEXT:    ret void
 //
@@ -3106,7 +3106,7 @@ void bar() {
 // CK12-NEXT:    [[T_ADDR:%.*]] = alloca ptr, align 4
 // CK12-NEXT:    store ptr [[T]], ptr [[T_ADDR]], align 4
 // CK12-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[T_ADDR]], align 4
-// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP0]], i32 1
+// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP0]], i32 1
 // CK12-NEXT:    store ptr [[INCDEC_PTR]], ptr [[T_ADDR]], align 4
 // CK12-NEXT:    ret void
 //
@@ -3120,7 +3120,7 @@ void bar() {
 // CK12-NEXT:    store ptr [[LR_ADDR]], ptr [[TMP]], align 4
 // CK12-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 4
 // CK12-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 4
-// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw float, ptr [[TMP1]], i32 1
+// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 1
 // CK12-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 4
 // CK12-NEXT:    ret void
 //
@@ -3134,7 +3134,7 @@ void bar() {
 // CK12-NEXT:    store ptr [[TR_ADDR]], ptr [[TMP]], align 4
 // CK12-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 4
 // CK12-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 4
-// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK12-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 4
 // CK12-NEXT:    ret void
 //
@@ -3148,7 +3148,7 @@ void bar() {
 // CK12-NEXT:    store ptr [[TR_ADDR]], ptr [[TMP]], align 4
 // CK12-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 4
 // CK12-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 4
-// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK12-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 4
 // CK12-NEXT:    ret void
 //
@@ -3166,11 +3166,11 @@ void bar() {
 // CK12-NEXT:    store ptr [[LR_ADDR]], ptr [[_TMP1]], align 4
 // CK12-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 4
 // CK12-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 4
-// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK12-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 4
 // CK12-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[_TMP1]], align 4
 // CK12-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP2]], align 4
-// CK12-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP3]], i32 1
+// CK12-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds float, ptr [[TMP3]], i32 1
 // CK12-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[TMP2]], align 4
 // CK12-NEXT:    ret void
 //
@@ -3555,7 +3555,7 @@ void bar() {
 // CK13-NEXT:    [[G_ADDR:%.*]] = alloca ptr, align 4
 // CK13-NEXT:    store ptr [[G]], ptr [[G_ADDR]], align 4
 // CK13-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[G_ADDR]], align 4
-// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // CK13-NEXT:    store ptr [[INCDEC_PTR]], ptr [[G_ADDR]], align 4
 // CK13-NEXT:    ret void
 //
@@ -3566,7 +3566,7 @@ void bar() {
 // CK13-NEXT:    [[L_ADDR:%.*]] = alloca ptr, align 4
 // CK13-NEXT:    store ptr [[L]], ptr [[L_ADDR]], align 4
 // CK13-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[L_ADDR]], align 4
-// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw float, ptr [[TMP0]], i32 1
+// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds float, ptr [[TMP0]], i32 1
 // CK13-NEXT:    store ptr [[INCDEC_PTR]], ptr [[L_ADDR]], align 4
 // CK13-NEXT:    ret void
 //
@@ -3577,7 +3577,7 @@ void bar() {
 // CK13-NEXT:    [[T_ADDR:%.*]] = alloca ptr, align 4
 // CK13-NEXT:    store ptr [[T]], ptr [[T_ADDR]], align 4
 // CK13-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[T_ADDR]], align 4
-// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP0]], i32 1
+// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP0]], i32 1
 // CK13-NEXT:    store ptr [[INCDEC_PTR]], ptr [[T_ADDR]], align 4
 // CK13-NEXT:    ret void
 //
@@ -3591,7 +3591,7 @@ void bar() {
 // CK13-NEXT:    store ptr [[LR_ADDR]], ptr [[TMP]], align 4
 // CK13-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 4
 // CK13-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 4
-// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw float, ptr [[TMP1]], i32 1
+// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 1
 // CK13-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 4
 // CK13-NEXT:    ret void
 //
@@ -3605,7 +3605,7 @@ void bar() {
 // CK13-NEXT:    store ptr [[TR_ADDR]], ptr [[TMP]], align 4
 // CK13-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 4
 // CK13-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 4
-// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK13-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 4
 // CK13-NEXT:    ret void
 //
@@ -3619,7 +3619,7 @@ void bar() {
 // CK13-NEXT:    store ptr [[TR_ADDR]], ptr [[TMP]], align 4
 // CK13-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 4
 // CK13-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 4
-// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK13-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 4
 // CK13-NEXT:    ret void
 //
@@ -3637,11 +3637,11 @@ void bar() {
 // CK13-NEXT:    store ptr [[LR_ADDR]], ptr [[_TMP1]], align 4
 // CK13-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[TMP]], align 4
 // CK13-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[TMP0]], align 4
-// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i32 1
+// CK13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 1
 // CK13-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP0]], align 4
 // CK13-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[_TMP1]], align 4
 // CK13-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP2]], align 4
-// CK13-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw float, ptr [[TMP3]], i32 1
+// CK13-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds float, ptr [[TMP3]], i32 1
 // CK13-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[TMP2]], align 4
 // CK13-NEXT:    ret void
 //
@@ -3674,34 +3674,34 @@ void bar() {
 // SIMD-ONLY00-NEXT:    store ptr [[LR]], ptr [[LR_ADDR]], align 8
 // SIMD-ONLY00-NEXT:    store ptr [[TR]], ptr [[TR_ADDR]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP0:%.*]] = load ptr, ptr @g, align 8
-// SIMD-ONLY00-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY00-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY00-NEXT:    store ptr [[INCDEC_PTR]], ptr @g, align 8
 // SIMD-ONLY00-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[L]], align 8
-// SIMD-ONLY00-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw float, ptr [[TMP1]], i32 1
+// SIMD-ONLY00-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 1
 // SIMD-ONLY00-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[L]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[T]], align 8
-// SIMD-ONLY00-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP2]], i32 1
+// SIMD-ONLY00-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i32 1
 // SIMD-ONLY00-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[T]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[LR_ADDR]], align 8
 // SIMD-ONLY00-NEXT:    store ptr [[TMP3]], ptr [[TMP]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[LR_ADDR]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[TMP5]], align 8
-// SIMD-ONLY00-NEXT:    [[INCDEC_PTR3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP6]], i32 1
+// SIMD-ONLY00-NEXT:    [[INCDEC_PTR3:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i32 1
 // SIMD-ONLY00-NEXT:    store ptr [[INCDEC_PTR3]], ptr [[TMP5]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[TR_ADDR]], align 8
 // SIMD-ONLY00-NEXT:    store ptr [[TMP7]], ptr [[_TMP4]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[TR_ADDR]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[_TMP4]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[TMP9]], align 8
-// SIMD-ONLY00-NEXT:    [[INCDEC_PTR5:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP10]], i32 1
+// SIMD-ONLY00-NEXT:    [[INCDEC_PTR5:%.*]] = getelementptr inbounds i32, ptr [[TMP10]], i32 1
 // SIMD-ONLY00-NEXT:    store ptr [[INCDEC_PTR5]], ptr [[TMP9]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TR_ADDR]], align 8
 // SIMD-ONLY00-NEXT:    store ptr [[TMP11]], ptr [[_TMP6]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[TR_ADDR]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[_TMP6]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP13]], align 8
-// SIMD-ONLY00-NEXT:    [[INCDEC_PTR7:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP14]], i32 1
+// SIMD-ONLY00-NEXT:    [[INCDEC_PTR7:%.*]] = getelementptr inbounds i32, ptr [[TMP14]], i32 1
 // SIMD-ONLY00-NEXT:    store ptr [[INCDEC_PTR7]], ptr [[TMP13]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP15:%.*]] = load ptr, ptr [[TR_ADDR]], align 8
 // SIMD-ONLY00-NEXT:    store ptr [[TMP15]], ptr [[_TMP8]], align 8
@@ -3711,11 +3711,11 @@ void bar() {
 // SIMD-ONLY00-NEXT:    [[TMP18:%.*]] = load ptr, ptr [[LR_ADDR]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP19:%.*]] = load ptr, ptr [[_TMP8]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP19]], align 8
-// SIMD-ONLY00-NEXT:    [[INCDEC_PTR10:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP20]], i32 1
+// SIMD-ONLY00-NEXT:    [[INCDEC_PTR10:%.*]] = getelementptr inbounds i32, ptr [[TMP20]], i32 1
 // SIMD-ONLY00-NEXT:    store ptr [[INCDEC_PTR10]], ptr [[TMP19]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP21:%.*]] = load ptr, ptr [[_TMP9]], align 8
 // SIMD-ONLY00-NEXT:    [[TMP22:%.*]] = load ptr, ptr [[TMP21]], align 8
-// SIMD-ONLY00-NEXT:    [[INCDEC_PTR11:%.*]] = getelementptr inbounds nuw float, ptr [[TMP22]], i32 1
+// SIMD-ONLY00-NEXT:    [[INCDEC_PTR11:%.*]] = getelementptr inbounds float, ptr [[TMP22]], i32 1
 // SIMD-ONLY00-NEXT:    store ptr [[INCDEC_PTR11]], ptr [[TMP21]], align 8
 // SIMD-ONLY00-NEXT:    ret void
 //
@@ -3748,34 +3748,34 @@ void bar() {
 // SIMD-ONLY01-NEXT:    store ptr [[LR]], ptr [[LR_ADDR]], align 8
 // SIMD-ONLY01-NEXT:    store ptr [[TR]], ptr [[TR_ADDR]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP0:%.*]] = load ptr, ptr @g, align 8
-// SIMD-ONLY01-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY01-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY01-NEXT:    store ptr [[INCDEC_PTR]], ptr @g, align 8
 // SIMD-ONLY01-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[L]], align 8
-// SIMD-ONLY01-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw float, ptr [[TMP1]], i32 1
+// SIMD-ONLY01-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 1
 // SIMD-ONLY01-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[L]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[T]], align 8
-// SIMD-ONLY01-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP2]], i32 1
+// SIMD-ONLY01-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i32 1
 // SIMD-ONLY01-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[T]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[LR_ADDR]], align 8
 // SIMD-ONLY01-NEXT:    store ptr [[TMP3]], ptr [[TMP]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[LR_ADDR]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[TMP5]], align 8
-// SIMD-ONLY01-NEXT:    [[INCDEC_PTR3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP6]], i32 1
+// SIMD-ONLY01-NEXT:    [[INCDEC_PTR3:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i32 1
 // SIMD-ONLY01-NEXT:    store ptr [[INCDEC_PTR3]], ptr [[TMP5]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[TR_ADDR]], align 8
 // SIMD-ONLY01-NEXT:    store ptr [[TMP7]], ptr [[_TMP4]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[TR_ADDR]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[_TMP4]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[TMP9]], align 8
-// SIMD-ONLY01-NEXT:    [[INCDEC_PTR5:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP10]], i32 1
+// SIMD-ONLY01-NEXT:    [[INCDEC_PTR5:%.*]] = getelementptr inbounds i32, ptr [[TMP10]], i32 1
 // SIMD-ONLY01-NEXT:    store ptr [[INCDEC_PTR5]], ptr [[TMP9]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TR_ADDR]], align 8
 // SIMD-ONLY01-NEXT:    store ptr [[TMP11]], ptr [[_TMP6]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[TR_ADDR]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[_TMP6]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP13]], align 8
-// SIMD-ONLY01-NEXT:    [[INCDEC_PTR7:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP14]], i32 1
+// SIMD-ONLY01-NEXT:    [[INCDEC_PTR7:%.*]] = getelementptr inbounds i32, ptr [[TMP14]], i32 1
 // SIMD-ONLY01-NEXT:    store ptr [[INCDEC_PTR7]], ptr [[TMP13]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP15:%.*]] = load ptr, ptr [[TR_ADDR]], align 8
 // SIMD-ONLY01-NEXT:    store ptr [[TMP15]], ptr [[_TMP8]], align 8
@@ -3785,11 +3785,11 @@ void bar() {
 // SIMD-ONLY01-NEXT:    [[TMP18:%.*]] = load ptr, ptr [[LR_ADDR]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP19:%.*]] = load ptr, ptr [[_TMP8]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP19]], align 8
-// SIMD-ONLY01-NEXT:    [[INCDEC_PTR10:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP20]], i32 1
+// SIMD-ONLY01-NEXT:    [[INCDEC_PTR10:%.*]] = getelementptr inbounds i32, ptr [[TMP20]], i32 1
 // SIMD-ONLY01-NEXT:    store ptr [[INCDEC_PTR10]], ptr [[TMP19]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP21:%.*]] = load ptr, ptr [[_TMP9]], align 8
 // SIMD-ONLY01-NEXT:    [[TMP22:%.*]] = load ptr, ptr [[TMP21]], align 8
-// SIMD-ONLY01-NEXT:    [[INCDEC_PTR11:%.*]] = getelementptr inbounds nuw float, ptr [[TMP22]], i32 1
+// SIMD-ONLY01-NEXT:    [[INCDEC_PTR11:%.*]] = getelementptr inbounds float, ptr [[TMP22]], i32 1
 // SIMD-ONLY01-NEXT:    store ptr [[INCDEC_PTR11]], ptr [[TMP21]], align 8
 // SIMD-ONLY01-NEXT:    ret void
 //
@@ -3822,34 +3822,34 @@ void bar() {
 // SIMD-ONLY02-NEXT:    store ptr [[LR]], ptr [[LR_ADDR]], align 4
 // SIMD-ONLY02-NEXT:    store ptr [[TR]], ptr [[TR_ADDR]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP0:%.*]] = load ptr, ptr @g, align 4
-// SIMD-ONLY02-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY02-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY02-NEXT:    store ptr [[INCDEC_PTR]], ptr @g, align 4
 // SIMD-ONLY02-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[L]], align 4
-// SIMD-ONLY02-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw float, ptr [[TMP1]], i32 1
+// SIMD-ONLY02-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 1
 // SIMD-ONLY02-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[L]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[T]], align 4
-// SIMD-ONLY02-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP2]], i32 1
+// SIMD-ONLY02-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i32 1
 // SIMD-ONLY02-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[T]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[LR_ADDR]], align 4
 // SIMD-ONLY02-NEXT:    store ptr [[TMP3]], ptr [[TMP]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[LR_ADDR]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[TMP5]], align 4
-// SIMD-ONLY02-NEXT:    [[INCDEC_PTR3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP6]], i32 1
+// SIMD-ONLY02-NEXT:    [[INCDEC_PTR3:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i32 1
 // SIMD-ONLY02-NEXT:    store ptr [[INCDEC_PTR3]], ptr [[TMP5]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[TR_ADDR]], align 4
 // SIMD-ONLY02-NEXT:    store ptr [[TMP7]], ptr [[_TMP4]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[TR_ADDR]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[_TMP4]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[TMP9]], align 4
-// SIMD-ONLY02-NEXT:    [[INCDEC_PTR5:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP10]], i32 1
+// SIMD-ONLY02-NEXT:    [[INCDEC_PTR5:%.*]] = getelementptr inbounds i32, ptr [[TMP10]], i32 1
 // SIMD-ONLY02-NEXT:    store ptr [[INCDEC_PTR5]], ptr [[TMP9]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TR_ADDR]], align 4
 // SIMD-ONLY02-NEXT:    store ptr [[TMP11]], ptr [[_TMP6]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[TR_ADDR]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[_TMP6]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP13]], align 4
-// SIMD-ONLY02-NEXT:    [[INCDEC_PTR7:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP14]], i32 1
+// SIMD-ONLY02-NEXT:    [[INCDEC_PTR7:%.*]] = getelementptr inbounds i32, ptr [[TMP14]], i32 1
 // SIMD-ONLY02-NEXT:    store ptr [[INCDEC_PTR7]], ptr [[TMP13]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP15:%.*]] = load ptr, ptr [[TR_ADDR]], align 4
 // SIMD-ONLY02-NEXT:    store ptr [[TMP15]], ptr [[_TMP8]], align 4
@@ -3859,11 +3859,11 @@ void bar() {
 // SIMD-ONLY02-NEXT:    [[TMP18:%.*]] = load ptr, ptr [[LR_ADDR]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP19:%.*]] = load ptr, ptr [[_TMP8]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP19]], align 4
-// SIMD-ONLY02-NEXT:    [[INCDEC_PTR10:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP20]], i32 1
+// SIMD-ONLY02-NEXT:    [[INCDEC_PTR10:%.*]] = getelementptr inbounds i32, ptr [[TMP20]], i32 1
 // SIMD-ONLY02-NEXT:    store ptr [[INCDEC_PTR10]], ptr [[TMP19]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP21:%.*]] = load ptr, ptr [[_TMP9]], align 4
 // SIMD-ONLY02-NEXT:    [[TMP22:%.*]] = load ptr, ptr [[TMP21]], align 4
-// SIMD-ONLY02-NEXT:    [[INCDEC_PTR11:%.*]] = getelementptr inbounds nuw float, ptr [[TMP22]], i32 1
+// SIMD-ONLY02-NEXT:    [[INCDEC_PTR11:%.*]] = getelementptr inbounds float, ptr [[TMP22]], i32 1
 // SIMD-ONLY02-NEXT:    store ptr [[INCDEC_PTR11]], ptr [[TMP21]], align 4
 // SIMD-ONLY02-NEXT:    ret void
 //
@@ -3896,34 +3896,34 @@ void bar() {
 // SIMD-ONLY03-NEXT:    store ptr [[LR]], ptr [[LR_ADDR]], align 4
 // SIMD-ONLY03-NEXT:    store ptr [[TR]], ptr [[TR_ADDR]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP0:%.*]] = load ptr, ptr @g, align 4
-// SIMD-ONLY03-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY03-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY03-NEXT:    store ptr [[INCDEC_PTR]], ptr @g, align 4
 // SIMD-ONLY03-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[L]], align 4
-// SIMD-ONLY03-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw float, ptr [[TMP1]], i32 1
+// SIMD-ONLY03-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 1
 // SIMD-ONLY03-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[L]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[T]], align 4
-// SIMD-ONLY03-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP2]], i32 1
+// SIMD-ONLY03-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i32 1
 // SIMD-ONLY03-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[T]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[LR_ADDR]], align 4
 // SIMD-ONLY03-NEXT:    store ptr [[TMP3]], ptr [[TMP]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[LR_ADDR]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[TMP5]], align 4
-// SIMD-ONLY03-NEXT:    [[INCDEC_PTR3:%.*]] = getelementptr inbounds nuw float, ptr [[TMP6]], i32 1
+// SIMD-ONLY03-NEXT:    [[INCDEC_PTR3:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i32 1
 // SIMD-ONLY03-NEXT:    store ptr [[INCDEC_PTR3]], ptr [[TMP5]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[TR_ADDR]], align 4
 // SIMD-ONLY03-NEXT:    store ptr [[TMP7]], ptr [[_TMP4]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[TR_ADDR]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[_TMP4]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[TMP9]], align 4
-// SIMD-ONLY03-NEXT:    [[INCDEC_PTR5:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP10]], i32 1
+// SIMD-ONLY03-NEXT:    [[INCDEC_PTR5:%.*]] = getelementptr inbounds i32, ptr [[TMP10]], i32 1
 // SIMD-ONLY03-NEXT:    store ptr [[INCDEC_PTR5]], ptr [[TMP9]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[TR_ADDR]], align 4
 // SIMD-ONLY03-NEXT:    store ptr [[TMP11]], ptr [[_TMP6]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP12:%.*]] = load ptr, ptr [[TR_ADDR]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[_TMP6]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP14:%.*]] = load ptr, ptr [[TMP13]], align 4
-// SIMD-ONLY03-NEXT:    [[INCDEC_PTR7:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP14]], i32 1
+// SIMD-ONLY03-NEXT:    [[INCDEC_PTR7:%.*]] = getelementptr inbounds i32, ptr [[TMP14]], i32 1
 // SIMD-ONLY03-NEXT:    store ptr [[INCDEC_PTR7]], ptr [[TMP13]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP15:%.*]] = load ptr, ptr [[TR_ADDR]], align 4
 // SIMD-ONLY03-NEXT:    store ptr [[TMP15]], ptr [[_TMP8]], align 4
@@ -3933,11 +3933,11 @@ void bar() {
 // SIMD-ONLY03-NEXT:    [[TMP18:%.*]] = load ptr, ptr [[LR_ADDR]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP19:%.*]] = load ptr, ptr [[_TMP8]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP20:%.*]] = load ptr, ptr [[TMP19]], align 4
-// SIMD-ONLY03-NEXT:    [[INCDEC_PTR10:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP20]], i32 1
+// SIMD-ONLY03-NEXT:    [[INCDEC_PTR10:%.*]] = getelementptr inbounds i32, ptr [[TMP20]], i32 1
 // SIMD-ONLY03-NEXT:    store ptr [[INCDEC_PTR10]], ptr [[TMP19]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP21:%.*]] = load ptr, ptr [[_TMP9]], align 4
 // SIMD-ONLY03-NEXT:    [[TMP22:%.*]] = load ptr, ptr [[TMP21]], align 4
-// SIMD-ONLY03-NEXT:    [[INCDEC_PTR11:%.*]] = getelementptr inbounds nuw float, ptr [[TMP22]], i32 1
+// SIMD-ONLY03-NEXT:    [[INCDEC_PTR11:%.*]] = getelementptr inbounds float, ptr [[TMP22]], i32 1
 // SIMD-ONLY03-NEXT:    store ptr [[INCDEC_PTR11]], ptr [[TMP21]], align 4
 // SIMD-ONLY03-NEXT:    ret void
 //
@@ -3951,7 +3951,7 @@ void bar() {
 // CK20-NEXT:    call void @_ZN2STIdEC1ERPd(ptr noundef nonnull align 8 dereferenceable(16) [[A]], ptr noundef nonnull align 8 dereferenceable(8) [[ARG_ADDR]])
 // CK20-NEXT:    call void @_ZN2STIdE3fooERPd(ptr noundef nonnull align 8 dereferenceable(16) [[A]], ptr noundef nonnull align 8 dereferenceable(8) [[ARG_ADDR]])
 // CK20-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARG_ADDR]], align 8
-// CK20-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// CK20-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // CK20-NEXT:    store ptr [[INCDEC_PTR]], ptr [[ARG_ADDR]], align 8
 // CK20-NEXT:    ret void
 //
@@ -4185,7 +4185,7 @@ void bar() {
 // CK20-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
 // CK20-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 0
 // CK20-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[A]], align 8
-// CK20-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP1]], i32 1
+// CK20-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP1]], i32 1
 // CK20-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 8
 // CK20-NEXT:    ret void
 //
@@ -4199,7 +4199,7 @@ void bar() {
 // CK20-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 1
 // CK20-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[B]], align 8
 // CK20-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CK20-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP2]], i32 1
+// CK20-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP2]], i32 1
 // CK20-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP1]], align 8
 // CK20-NEXT:    ret void
 //
@@ -4212,12 +4212,12 @@ void bar() {
 // CK20-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
 // CK20-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 0
 // CK20-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[A]], align 8
-// CK20-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP1]], i32 1
+// CK20-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP1]], i32 1
 // CK20-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 8
 // CK20-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[TMP0]], i32 0, i32 1
 // CK20-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[B]], align 8
 // CK20-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP2]], align 8
-// CK20-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw double, ptr [[TMP3]], i32 1
+// CK20-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds double, ptr [[TMP3]], i32 1
 // CK20-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[TMP2]], align 8
 // CK20-NEXT:    ret void
 //
@@ -4231,7 +4231,7 @@ void bar() {
 // CK21-NEXT:    call void @_ZN2STIdEC1ERPd(ptr noundef nonnull align 8 dereferenceable(16) [[A]], ptr noundef nonnull align 8 dereferenceable(8) [[ARG_ADDR]])
 // CK21-NEXT:    call void @_ZN2STIdE3fooERPd(ptr noundef nonnull align 8 dereferenceable(16) [[A]], ptr noundef nonnull align 8 dereferenceable(8) [[ARG_ADDR]])
 // CK21-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARG_ADDR]], align 8
-// CK21-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// CK21-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // CK21-NEXT:    store ptr [[INCDEC_PTR]], ptr [[ARG_ADDR]], align 8
 // CK21-NEXT:    ret void
 //
@@ -4465,7 +4465,7 @@ void bar() {
 // CK21-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
 // CK21-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 0
 // CK21-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[A]], align 8
-// CK21-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP1]], i32 1
+// CK21-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP1]], i32 1
 // CK21-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 8
 // CK21-NEXT:    ret void
 //
@@ -4479,7 +4479,7 @@ void bar() {
 // CK21-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 1
 // CK21-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[B]], align 8
 // CK21-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 8
-// CK21-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP2]], i32 1
+// CK21-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP2]], i32 1
 // CK21-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP1]], align 8
 // CK21-NEXT:    ret void
 //
@@ -4492,12 +4492,12 @@ void bar() {
 // CK21-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
 // CK21-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 0
 // CK21-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[A]], align 8
-// CK21-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP1]], i32 1
+// CK21-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP1]], i32 1
 // CK21-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 8
 // CK21-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[TMP0]], i32 0, i32 1
 // CK21-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[B]], align 8
 // CK21-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP2]], align 8
-// CK21-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw double, ptr [[TMP3]], i32 1
+// CK21-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds double, ptr [[TMP3]], i32 1
 // CK21-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[TMP2]], align 8
 // CK21-NEXT:    ret void
 //
@@ -4511,7 +4511,7 @@ void bar() {
 // CK22-NEXT:    call void @_ZN2STIdEC1ERPd(ptr noundef nonnull align 4 dereferenceable(8) [[A]], ptr noundef nonnull align 4 dereferenceable(4) [[ARG_ADDR]])
 // CK22-NEXT:    call void @_ZN2STIdE3fooERPd(ptr noundef nonnull align 4 dereferenceable(8) [[A]], ptr noundef nonnull align 4 dereferenceable(4) [[ARG_ADDR]])
 // CK22-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARG_ADDR]], align 4
-// CK22-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// CK22-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // CK22-NEXT:    store ptr [[INCDEC_PTR]], ptr [[ARG_ADDR]], align 4
 // CK22-NEXT:    ret void
 //
@@ -4745,7 +4745,7 @@ void bar() {
 // CK22-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
 // CK22-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 0
 // CK22-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[A]], align 4
-// CK22-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP1]], i32 1
+// CK22-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP1]], i32 1
 // CK22-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 4
 // CK22-NEXT:    ret void
 //
@@ -4759,7 +4759,7 @@ void bar() {
 // CK22-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 1
 // CK22-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[B]], align 4
 // CK22-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 4
-// CK22-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP2]], i32 1
+// CK22-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP2]], i32 1
 // CK22-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP1]], align 4
 // CK22-NEXT:    ret void
 //
@@ -4772,12 +4772,12 @@ void bar() {
 // CK22-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
 // CK22-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 0
 // CK22-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[A]], align 4
-// CK22-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP1]], i32 1
+// CK22-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP1]], i32 1
 // CK22-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 4
 // CK22-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[TMP0]], i32 0, i32 1
 // CK22-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[B]], align 4
 // CK22-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP2]], align 4
-// CK22-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw double, ptr [[TMP3]], i32 1
+// CK22-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds double, ptr [[TMP3]], i32 1
 // CK22-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[TMP2]], align 4
 // CK22-NEXT:    ret void
 //
@@ -4791,7 +4791,7 @@ void bar() {
 // CK23-NEXT:    call void @_ZN2STIdEC1ERPd(ptr noundef nonnull align 4 dereferenceable(8) [[A]], ptr noundef nonnull align 4 dereferenceable(4) [[ARG_ADDR]])
 // CK23-NEXT:    call void @_ZN2STIdE3fooERPd(ptr noundef nonnull align 4 dereferenceable(8) [[A]], ptr noundef nonnull align 4 dereferenceable(4) [[ARG_ADDR]])
 // CK23-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARG_ADDR]], align 4
-// CK23-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// CK23-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // CK23-NEXT:    store ptr [[INCDEC_PTR]], ptr [[ARG_ADDR]], align 4
 // CK23-NEXT:    ret void
 //
@@ -5025,7 +5025,7 @@ void bar() {
 // CK23-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
 // CK23-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 0
 // CK23-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[A]], align 4
-// CK23-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP1]], i32 1
+// CK23-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP1]], i32 1
 // CK23-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 4
 // CK23-NEXT:    ret void
 //
@@ -5039,7 +5039,7 @@ void bar() {
 // CK23-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 1
 // CK23-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[B]], align 4
 // CK23-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 4
-// CK23-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP2]], i32 1
+// CK23-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP2]], i32 1
 // CK23-NEXT:    store ptr [[INCDEC_PTR]], ptr [[TMP1]], align 4
 // CK23-NEXT:    ret void
 //
@@ -5052,12 +5052,12 @@ void bar() {
 // CK23-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
 // CK23-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[TMP0]], i32 0, i32 0
 // CK23-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[A]], align 4
-// CK23-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP1]], i32 1
+// CK23-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP1]], i32 1
 // CK23-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 4
 // CK23-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[TMP0]], i32 0, i32 1
 // CK23-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[B]], align 4
 // CK23-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[TMP2]], align 4
-// CK23-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds nuw double, ptr [[TMP3]], i32 1
+// CK23-NEXT:    [[INCDEC_PTR1:%.*]] = getelementptr inbounds double, ptr [[TMP3]], i32 1
 // CK23-NEXT:    store ptr [[INCDEC_PTR1]], ptr [[TMP2]], align 4
 // CK23-NEXT:    ret void
 //
@@ -5071,7 +5071,7 @@ void bar() {
 // SIMD-ONLY10-NEXT:    call void @_ZN2STIdEC1ERPd(ptr noundef nonnull align 8 dereferenceable(16) [[A]], ptr noundef nonnull align 8 dereferenceable(8) [[ARG_ADDR]])
 // SIMD-ONLY10-NEXT:    call void @_ZN2STIdE3fooERPd(ptr noundef nonnull align 8 dereferenceable(16) [[A]], ptr noundef nonnull align 8 dereferenceable(8) [[ARG_ADDR]])
 // SIMD-ONLY10-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARG_ADDR]], align 8
-// SIMD-ONLY10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY10-NEXT:    store ptr [[INCDEC_PTR]], ptr [[ARG_ADDR]], align 8
 // SIMD-ONLY10-NEXT:    ret void
 //
@@ -5101,21 +5101,21 @@ void bar() {
 // SIMD-ONLY10-NEXT:    store ptr null, ptr [[LA]], align 8
 // SIMD-ONLY10-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[THIS1]], i32 0, i32 0
 // SIMD-ONLY10-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[A]], align 8
-// SIMD-ONLY10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY10-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY10-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 8
 // SIMD-ONLY10-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 1
 // SIMD-ONLY10-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[B]], align 8
 // SIMD-ONLY10-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 8
-// SIMD-ONLY10-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw double, ptr [[TMP2]], i32 1
+// SIMD-ONLY10-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds double, ptr [[TMP2]], i32 1
 // SIMD-ONLY10-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[TMP1]], align 8
 // SIMD-ONLY10-NEXT:    [[A3:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 0
 // SIMD-ONLY10-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[A3]], align 8
-// SIMD-ONLY10-NEXT:    [[INCDEC_PTR4:%.*]] = getelementptr inbounds nuw double, ptr [[TMP3]], i32 1
+// SIMD-ONLY10-NEXT:    [[INCDEC_PTR4:%.*]] = getelementptr inbounds double, ptr [[TMP3]], i32 1
 // SIMD-ONLY10-NEXT:    store ptr [[INCDEC_PTR4]], ptr [[A3]], align 8
 // SIMD-ONLY10-NEXT:    [[B5:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 1
 // SIMD-ONLY10-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[B5]], align 8
 // SIMD-ONLY10-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP4]], align 8
-// SIMD-ONLY10-NEXT:    [[INCDEC_PTR6:%.*]] = getelementptr inbounds nuw double, ptr [[TMP5]], i32 1
+// SIMD-ONLY10-NEXT:    [[INCDEC_PTR6:%.*]] = getelementptr inbounds double, ptr [[TMP5]], i32 1
 // SIMD-ONLY10-NEXT:    store ptr [[INCDEC_PTR6]], ptr [[TMP4]], align 8
 // SIMD-ONLY10-NEXT:    ret void
 //
@@ -5145,7 +5145,7 @@ void bar() {
 // SIMD-ONLY11-NEXT:    call void @_ZN2STIdEC1ERPd(ptr noundef nonnull align 8 dereferenceable(16) [[A]], ptr noundef nonnull align 8 dereferenceable(8) [[ARG_ADDR]])
 // SIMD-ONLY11-NEXT:    call void @_ZN2STIdE3fooERPd(ptr noundef nonnull align 8 dereferenceable(16) [[A]], ptr noundef nonnull align 8 dereferenceable(8) [[ARG_ADDR]])
 // SIMD-ONLY11-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARG_ADDR]], align 8
-// SIMD-ONLY11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY11-NEXT:    store ptr [[INCDEC_PTR]], ptr [[ARG_ADDR]], align 8
 // SIMD-ONLY11-NEXT:    ret void
 //
@@ -5175,21 +5175,21 @@ void bar() {
 // SIMD-ONLY11-NEXT:    store ptr null, ptr [[LA]], align 8
 // SIMD-ONLY11-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[THIS1]], i32 0, i32 0
 // SIMD-ONLY11-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[A]], align 8
-// SIMD-ONLY11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY11-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY11-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 8
 // SIMD-ONLY11-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 1
 // SIMD-ONLY11-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[B]], align 8
 // SIMD-ONLY11-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 8
-// SIMD-ONLY11-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw double, ptr [[TMP2]], i32 1
+// SIMD-ONLY11-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds double, ptr [[TMP2]], i32 1
 // SIMD-ONLY11-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[TMP1]], align 8
 // SIMD-ONLY11-NEXT:    [[A3:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 0
 // SIMD-ONLY11-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[A3]], align 8
-// SIMD-ONLY11-NEXT:    [[INCDEC_PTR4:%.*]] = getelementptr inbounds nuw double, ptr [[TMP3]], i32 1
+// SIMD-ONLY11-NEXT:    [[INCDEC_PTR4:%.*]] = getelementptr inbounds double, ptr [[TMP3]], i32 1
 // SIMD-ONLY11-NEXT:    store ptr [[INCDEC_PTR4]], ptr [[A3]], align 8
 // SIMD-ONLY11-NEXT:    [[B5:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 1
 // SIMD-ONLY11-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[B5]], align 8
 // SIMD-ONLY11-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP4]], align 8
-// SIMD-ONLY11-NEXT:    [[INCDEC_PTR6:%.*]] = getelementptr inbounds nuw double, ptr [[TMP5]], i32 1
+// SIMD-ONLY11-NEXT:    [[INCDEC_PTR6:%.*]] = getelementptr inbounds double, ptr [[TMP5]], i32 1
 // SIMD-ONLY11-NEXT:    store ptr [[INCDEC_PTR6]], ptr [[TMP4]], align 8
 // SIMD-ONLY11-NEXT:    ret void
 //
@@ -5219,7 +5219,7 @@ void bar() {
 // SIMD-ONLY12-NEXT:    call void @_ZN2STIdEC1ERPd(ptr noundef nonnull align 4 dereferenceable(8) [[A]], ptr noundef nonnull align 4 dereferenceable(4) [[ARG_ADDR]])
 // SIMD-ONLY12-NEXT:    call void @_ZN2STIdE3fooERPd(ptr noundef nonnull align 4 dereferenceable(8) [[A]], ptr noundef nonnull align 4 dereferenceable(4) [[ARG_ADDR]])
 // SIMD-ONLY12-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARG_ADDR]], align 4
-// SIMD-ONLY12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY12-NEXT:    store ptr [[INCDEC_PTR]], ptr [[ARG_ADDR]], align 4
 // SIMD-ONLY12-NEXT:    ret void
 //
@@ -5249,21 +5249,21 @@ void bar() {
 // SIMD-ONLY12-NEXT:    store ptr null, ptr [[LA]], align 4
 // SIMD-ONLY12-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[THIS1]], i32 0, i32 0
 // SIMD-ONLY12-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[A]], align 4
-// SIMD-ONLY12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY12-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY12-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 4
 // SIMD-ONLY12-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 1
 // SIMD-ONLY12-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[B]], align 4
 // SIMD-ONLY12-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 4
-// SIMD-ONLY12-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw double, ptr [[TMP2]], i32 1
+// SIMD-ONLY12-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds double, ptr [[TMP2]], i32 1
 // SIMD-ONLY12-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[TMP1]], align 4
 // SIMD-ONLY12-NEXT:    [[A3:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 0
 // SIMD-ONLY12-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[A3]], align 4
-// SIMD-ONLY12-NEXT:    [[INCDEC_PTR4:%.*]] = getelementptr inbounds nuw double, ptr [[TMP3]], i32 1
+// SIMD-ONLY12-NEXT:    [[INCDEC_PTR4:%.*]] = getelementptr inbounds double, ptr [[TMP3]], i32 1
 // SIMD-ONLY12-NEXT:    store ptr [[INCDEC_PTR4]], ptr [[A3]], align 4
 // SIMD-ONLY12-NEXT:    [[B5:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 1
 // SIMD-ONLY12-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[B5]], align 4
 // SIMD-ONLY12-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP4]], align 4
-// SIMD-ONLY12-NEXT:    [[INCDEC_PTR6:%.*]] = getelementptr inbounds nuw double, ptr [[TMP5]], i32 1
+// SIMD-ONLY12-NEXT:    [[INCDEC_PTR6:%.*]] = getelementptr inbounds double, ptr [[TMP5]], i32 1
 // SIMD-ONLY12-NEXT:    store ptr [[INCDEC_PTR6]], ptr [[TMP4]], align 4
 // SIMD-ONLY12-NEXT:    ret void
 //
@@ -5293,7 +5293,7 @@ void bar() {
 // SIMD-ONLY13-NEXT:    call void @_ZN2STIdEC1ERPd(ptr noundef nonnull align 4 dereferenceable(8) [[A]], ptr noundef nonnull align 4 dereferenceable(4) [[ARG_ADDR]])
 // SIMD-ONLY13-NEXT:    call void @_ZN2STIdE3fooERPd(ptr noundef nonnull align 4 dereferenceable(8) [[A]], ptr noundef nonnull align 4 dereferenceable(4) [[ARG_ADDR]])
 // SIMD-ONLY13-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARG_ADDR]], align 4
-// SIMD-ONLY13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY13-NEXT:    store ptr [[INCDEC_PTR]], ptr [[ARG_ADDR]], align 4
 // SIMD-ONLY13-NEXT:    ret void
 //
@@ -5323,21 +5323,21 @@ void bar() {
 // SIMD-ONLY13-NEXT:    store ptr null, ptr [[LA]], align 4
 // SIMD-ONLY13-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_ST:%.*]], ptr [[THIS1]], i32 0, i32 0
 // SIMD-ONLY13-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[A]], align 4
-// SIMD-ONLY13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds nuw double, ptr [[TMP0]], i32 1
+// SIMD-ONLY13-NEXT:    [[INCDEC_PTR:%.*]] = getelementptr inbounds double, ptr [[TMP0]], i32 1
 // SIMD-ONLY13-NEXT:    store ptr [[INCDEC_PTR]], ptr [[A]], align 4
 // SIMD-ONLY13-NEXT:    [[B:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 1
 // SIMD-ONLY13-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[B]], align 4
 // SIMD-ONLY13-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[TMP1]], align 4
-// SIMD-ONLY13-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds nuw double, ptr [[TMP2]], i32 1
+// SIMD-ONLY13-NEXT:    [[INCDEC_PTR2:%.*]] = getelementptr inbounds double, ptr [[TMP2]], i32 1
 // SIMD-ONLY13-NEXT:    store ptr [[INCDEC_PTR2]], ptr [[TMP1]], align 4
 // SIMD-ONLY13-NEXT:    [[A3:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 0
 // SIMD-ONLY13-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[A3]], align 4
-// SIMD-ONLY13-NEXT:    [[INCDEC_PTR4:%.*]] = getelementptr inbounds nuw double, ptr [[TMP3]], i32 1
+// SIMD-ONLY13-NEXT:    [[INCDEC_PTR4:%.*]] = getelementptr inbounds double, ptr [[TMP3]], i32 1
 // SIMD-ONLY13-NEXT:    store ptr [[INCDEC_PTR4]], ptr [[A3]], align 4
 // SIMD-ONLY13-NEXT:    [[B5:%.*]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[THIS1]], i32 0, i32 1
 // SIMD-ONLY13-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[B5]], align 4
 // SIMD-ONLY13-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[TMP4]], align 4
-// SIMD-ONLY13-NEXT:    [[INCDEC_PTR6:%.*]] = getelementptr inbounds nuw double, ptr [[TMP5]], i32 1
+// SIMD-ONLY13-NEXT:    [[INCDEC_PTR6:%.*]] = getelementptr inbounds double, ptr [[TMP5]], i32 1
 // SIMD-ONLY13-NEXT:    store ptr [[INCDEC_PTR6]], ptr [[TMP4]], align 4
 // SIMD-ONLY13-NEXT:    ret void
 //
diff --git a/clang/test/OpenMP/target_map_both_pointer_pointee_codegen.cpp b/clang/test/OpenMP/target_map_both_pointer_pointee_codegen.cpp
index 87fa7fe462daa8..fcaceac7d34671 100644
--- a/clang/test/OpenMP/target_map_both_pointer_pointee_codegen.cpp
+++ b/clang/test/OpenMP/target_map_both_pointer_pointee_codegen.cpp
@@ -45,7 +45,7 @@ void foo() {
 // CHECK-NEXT:    store ptr [[CALL]], ptr [[PTR]], align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[PTR]], align 8
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[PTR]], align 8
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i64 0
 // CHECK-NEXT:    [[TMP2:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
 // CHECK-NEXT:    store ptr [[PTR]], ptr [[TMP2]], align 8
 // CHECK-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
diff --git a/clang/test/OpenMP/target_map_codegen_01.cpp b/clang/test/OpenMP/target_map_codegen_01.cpp
index 9f3553d2377cbb..d112500eb5fdd4 100644
--- a/clang/test/OpenMP/target_map_codegen_01.cpp
+++ b/clang/test/OpenMP/target_map_codegen_01.cpp
@@ -108,6 +108,6 @@ void implicit_maps_reference (int a, int *b){
 // CK2: store ptr [[ADDR]], ptr [[REF]],
 // CK2: [[T:%.+]] = load ptr, ptr [[REF]],
 // CK2: [[TT:%.+]] = load ptr, ptr [[T]],
-// CK2: getelementptr inbounds nuw i32, ptr [[TT]], i32 1
+// CK2: getelementptr inbounds i32, ptr [[TT]], i32 1
 #endif // CK2
 #endif
diff --git a/clang/test/OpenMP/target_map_codegen_21.cpp b/clang/test/OpenMP/target_map_codegen_21.cpp
index f5c517692d8c8e..a1419b7d4beb8c 100644
--- a/clang/test/OpenMP/target_map_codegen_21.cpp
+++ b/clang/test/OpenMP/target_map_codegen_21.cpp
@@ -185,7 +185,7 @@ int explicit_maps_globals(void){
 // CK22-DAG: [[BP0:%.+]] = getelementptr inbounds {{.+}}[[BP]], i{{.+}} 0, i{{.+}} 0
 // CK22-DAG: [[P0:%.+]] = getelementptr inbounds {{.+}}[[P]], i{{.+}} 0, i{{.+}} 0
 // CK22-DAG: store ptr @c, ptr [[BP0]]
-// CK22-DAG: store ptr getelementptr inbounds nuw ([100 x i32], ptr @c, i{{.+}} 0, i{{.+}} 1), ptr [[P0]]
+// CK22-DAG: store ptr getelementptr inbounds ([100 x i32], ptr @c, i{{.+}} 0, i{{.+}} 1), ptr [[P0]]
 
 // CK22: call void [[CALL03:@.+]](ptr {{[^,]+}})
 #pragma omp target map(c [1:4])
@@ -277,7 +277,7 @@ int explicit_maps_globals(void){
 // CK22-DAG: [[BP0:%.+]] = getelementptr inbounds {{.+}}[[BP]], i{{.+}} 0, i{{.+}} 0
 // CK22-DAG: [[P0:%.+]] = getelementptr inbounds {{.+}}[[P]], i{{.+}} 0, i{{.+}} 0
 // CK22-DAG: store ptr @sc, ptr [[BP0]]
-// CK22-DAG: store ptr getelementptr inbounds nuw ([100 x [[ST]]], ptr @sc, i{{.+}} 0, i{{.+}} 1), ptr [[P0]]
+// CK22-DAG: store ptr getelementptr inbounds ([100 x [[ST]]], ptr @sc, i{{.+}} 0, i{{.+}} 1), ptr [[P0]]
 
 // CK22: call void [[CALL08:@.+]](ptr {{[^,]+}})
 #pragma omp target map(sc [1:4])
@@ -369,7 +369,7 @@ int explicit_maps_globals(void){
 // CK22-DAG: [[BP0:%.+]] = getelementptr inbounds {{.+}}[[BP]], i{{.+}} 0, i{{.+}} 0
 // CK22-DAG: [[P0:%.+]] = getelementptr inbounds {{.+}}[[P]], i{{.+}} 0, i{{.+}} 0
 // CK22-DAG: store ptr @stc, ptr [[BP0]]
-// CK22-DAG: store ptr getelementptr inbounds nuw ([100 x [[STT]]], ptr @stc, i{{.+}} 0, i{{.+}} 1),  ptr [[P0]]
+// CK22-DAG: store ptr getelementptr inbounds ([100 x [[STT]]], ptr @stc, i{{.+}} 0, i{{.+}} 1),  ptr [[P0]]
 
 // CK22: call void [[CALL13:@.+]](ptr {{[^,]+}})
 #pragma omp target map(stc [1:4])
diff --git a/clang/test/OpenMP/target_map_codegen_27.cpp b/clang/test/OpenMP/target_map_codegen_27.cpp
index bfe75bca481be4..fe7ae12e00d137 100644
--- a/clang/test/OpenMP/target_map_codegen_27.cpp
+++ b/clang/test/OpenMP/target_map_codegen_27.cpp
@@ -82,7 +82,7 @@ void explicit_maps_pointer_references (int *p){
 // CK28-DAG: store ptr [[VAR1:%.+]], ptr [[P0]]
 // CK28-DAG: [[VAR0]] = load ptr, ptr [[VAR00:%.+]],
 // CK28-DAG: [[VAR00]] = load ptr, ptr [[VAR000:%.+]],
-// CK28-DAG: [[VAR1]] = getelementptr inbounds nuw i32, ptr [[VAR11:%.+]], i{{64|32}} 2
+// CK28-DAG: [[VAR1]] = getelementptr inbounds i32, ptr [[VAR11:%.+]], i{{64|32}} 2
 // CK28-DAG: [[VAR11]] = load ptr, ptr [[VAR111:%.+]],
 // CK28-DAG: [[VAR111]] = load ptr, ptr [[VAR1111:%.+]],
 
diff --git a/clang/test/OpenMP/target_map_codegen_28.cpp b/clang/test/OpenMP/target_map_codegen_28.cpp
index 67ea72d791d03c..e92f7e4773ecf1 100644
--- a/clang/test/OpenMP/target_map_codegen_28.cpp
+++ b/clang/test/OpenMP/target_map_codegen_28.cpp
@@ -89,7 +89,7 @@ struct SSB{
 // CK29-DAG: store ptr [[VAR1:%.+]], ptr [[BP2]]
 // CK29-DAG: store ptr [[VAR2:%.+]], ptr [[P2]]
 // CK29-DAG: [[VAR1]] = getelementptr inbounds nuw [[SSA]], ptr %{{.+}}, i32 0, i32 1
-// CK29-DAG: [[VAR2]] = getelementptr inbounds nuw double, ptr [[VAR22:%.+]], i{{.+}} 0
+// CK29-DAG: [[VAR2]] = getelementptr inbounds double, ptr [[VAR22:%.+]], i{{.+}} 0
 // CK29-DAG: [[VAR22]] = load ptr, ptr %{{.+}},
 
 // CK29: call void [[CALL00:@.+]](ptr {{[^,]+}})
@@ -129,7 +129,7 @@ struct SSB{
 // CK29-DAG: [[P2:%.+]] = getelementptr inbounds {{.+}}[[P]], i{{.+}} 0, i{{.+}} 2
 // CK29-DAG: store ptr [[VAR1]], ptr [[BP2]]
 // CK29-DAG: store ptr [[VAR2:%.+]], ptr [[P2]]
-// CK29-DAG: [[VAR2]] = getelementptr inbounds nuw double, ptr [[VAR22:%.+]], i{{.+}} 0
+// CK29-DAG: [[VAR2]] = getelementptr inbounds double, ptr [[VAR22:%.+]], i{{.+}} 0
 // CK29-DAG: [[VAR22]] = load ptr, ptr %{{.+}},
 
 // CK29: call void [[CALL00:@.+]](ptr {{[^,]+}})
@@ -164,7 +164,7 @@ struct SSB{
 // CK29-DAG: store ptr [[VAR1:%.+]], ptr [[BP2]]
 // CK29-DAG: store ptr [[VAR2:%.+]], ptr [[P2]]
 // CK29-DAG: [[VAR1]] = getelementptr inbounds nuw [[SSA]], ptr %{{.+}}, i32 0, i32 1
-// CK29-DAG: [[VAR2]] = getelementptr inbounds nuw double, ptr [[VAR22:%.+]], i{{.+}} 0
+// CK29-DAG: [[VAR2]] = getelementptr inbounds double, ptr [[VAR22:%.+]], i{{.+}} 0
 // CK29-DAG: [[VAR22]] = load ptr, ptr %{{.+}},
 
 // CK29: call void [[CALL00:@.+]](ptr {{[^,]+}})
diff --git a/clang/test/OpenMP/target_map_codegen_29.cpp b/clang/test/OpenMP/target_map_codegen_29.cpp
index 3ca7b228d26c2e..936a01573c2d2e 100644
--- a/clang/test/OpenMP/target_map_codegen_29.cpp
+++ b/clang/test/OpenMP/target_map_codegen_29.cpp
@@ -89,7 +89,7 @@ typedef struct StructWithPtrTag : public Base {
 // CK30-DAG: [[PTR:%.+]] = getelementptr inbounds [4 x ptr], ptr [[PTRS]], i32 0, i32 2
 // CK30-DAG: store ptr [[S_PTR1_BEGIN:%.+]], ptr [[PTR]],
 // CK30-DAG: [[S_PTR1]] = getelementptr inbounds nuw [[STRUCT]], ptr [[S]], i32 0, i32 4
-// CK30-DAG: [[S_PTR1_BEGIN]] = getelementptr inbounds nuw i32, ptr [[S_PTR1_BEGIN_REF:%.+]], i{{64|32}} 0
+// CK30-DAG: [[S_PTR1_BEGIN]] = getelementptr inbounds i32, ptr [[S_PTR1_BEGIN_REF:%.+]], i{{64|32}} 0
 // CK30-DAG: [[S_PTR1_BEGIN_REF]] = load ptr, ptr [[S_PTR1:%.+]],
 // CK30-DAG: [[S_PTR1]] = getelementptr inbounds nuw [[STRUCT]], ptr [[S]], i32 0, i32 4
 
@@ -98,7 +98,7 @@ typedef struct StructWithPtrTag : public Base {
 // CK30-DAG: [[PTR:%.+]] = getelementptr inbounds [4 x ptr], ptr [[PTRS]], i32 0, i32 3
 // CK30-DAG: store ptr [[S_PTRBASE1_BEGIN:%.+]], ptr [[PTR]],
 // CK30-DAG: [[S_PTRBASE1]] = getelementptr inbounds nuw [[BASE]], ptr [[S_BASE:%.+]], i32 0, i32 2
-// CK30-DAG: [[S_PTRBASE1_BEGIN]] = getelementptr inbounds nuw i32, ptr [[S_PTRBASE1_BEGIN_REF:%.+]], i{{64|32}} 0
+// CK30-DAG: [[S_PTRBASE1_BEGIN]] = getelementptr inbounds i32, ptr [[S_PTRBASE1_BEGIN_REF:%.+]], i{{64|32}} 0
 // CK30-DAG: [[S_PTRBASE1_BEGIN_REF]] = load ptr, ptr [[S_PTRBASE1:%.+]],
 // CK30-DAG: [[S_PTRBASE1]] = getelementptr inbounds nuw [[BASE]], ptr [[S_BASE:%.+]], i32 0, i32 2
 void map_with_deep_copy() {
diff --git a/clang/test/OpenMP/target_map_deref_array_codegen.cpp b/clang/test/OpenMP/target_map_deref_array_codegen.cpp
index e61fc7296332b1..9d395b0ab8cd8f 100644
--- a/clang/test/OpenMP/target_map_deref_array_codegen.cpp
+++ b/clang/test/OpenMP/target_map_deref_array_codegen.cpp
@@ -75,7 +75,7 @@ void foo(int **t1d)
 // CHECK-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[T1D_ADDR]], align 8
 // CHECK-NEXT:    [[TMP9:%.*]] = load ptr, ptr [[T1D_ADDR]], align 8
 // CHECK-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[TMP9]], align 8
-// CHECK-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP10]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i32, ptr [[TMP10]], i64 0
 // CHECK-NEXT:    [[TMP11:%.*]] = getelementptr inbounds [2 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
 // CHECK-NEXT:    store ptr [[TMP7]], ptr [[TMP11]], align 8
 // CHECK-NEXT:    [[TMP12:%.*]] = getelementptr inbounds [2 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
diff --git a/clang/test/OpenMP/target_map_member_expr_array_section_codegen.cpp b/clang/test/OpenMP/target_map_member_expr_array_section_codegen.cpp
index 692e3a4214c9d3..7a0da002fb944a 100644
--- a/clang/test/OpenMP/target_map_member_expr_array_section_codegen.cpp
+++ b/clang/test/OpenMP/target_map_member_expr_array_section_codegen.cpp
@@ -28,12 +28,12 @@ struct maptest {
     // CHECK: getelementptr inbounds
     // CHECK: [[S_ADDR:%.+]] = getelementptr inbounds nuw %struct.maptest, ptr [[THIS:%.+]], i32 0, i32 0
     // CHECK: [[S_DATA_ADDR:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[S_ADDR]], i32 0, i32 0
-    // CHECK: [[S_DATA_0_ADDR:%.+]] = getelementptr inbounds nuw [6 x float], ptr [[S_DATA_ADDR]], i64 0, i64 0
+    // CHECK: [[S_DATA_0_ADDR:%.+]] = getelementptr inbounds [6 x float], ptr [[S_DATA_ADDR]], i64 0, i64 0
 
     // SZ = &this->s.data[6]-&this->s.data[0]
     // CHECK: [[S_ADDR:%.+]] = getelementptr inbounds nuw %struct.maptest, ptr [[THIS]], i32 0, i32 0
     // CHECK: [[S_DATA_ADDR:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[S_ADDR]], i32 0, i32 0
-    // CHECK: [[S_DATA_5_ADDR:%.+]] = getelementptr inbounds nuw [6 x float], ptr [[S_DATA_ADDR]], i64 0, i64 5
+    // CHECK: [[S_DATA_5_ADDR:%.+]] = getelementptr inbounds [6 x float], ptr [[S_DATA_ADDR]], i64 0, i64 5
     // CHECK: [[S_DATA_6_ADDR:%.+]] = getelementptr float, ptr [[S_DATA_5_ADDR]], i32 1
     // CHECK: [[END_BC:%.+]] = ptrtoint ptr [[S_DATA_6_ADDR]] to i64
     // CHECK: [[BEG_BC:%.+]] = ptrtoint ptr [[S_DATA_0_ADDR]] to i64
@@ -64,12 +64,12 @@ struct maptest {
     // CHECK: [[SIZE:%.+]] = alloca [2 x i64],
     // CHECK: [[S_ADDR:%.+]] = getelementptr inbounds nuw %struct.maptest, ptr [[THIS:%.+]], i32 0, i32 0
     // CHECK: [[S_DATA_ADDR:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[S_ADDR]], i32 0, i32 0
-    // CHECK: [[S_DATA_0_ADDR:%.+]] = getelementptr inbounds nuw [6 x float], ptr [[S_DATA_ADDR]], i64 0, i64 0
+    // CHECK: [[S_DATA_0_ADDR:%.+]] = getelementptr inbounds [6 x float], ptr [[S_DATA_ADDR]], i64 0, i64 0
 
     // SZ = &this->s.data[6]-&this->s.data[0]
     // CHECK: [[S_ADDR:%.+]] = getelementptr inbounds nuw %struct.maptest, ptr [[THIS]], i32 0, i32 0
     // CHECK: [[S_DATA_ADDR:%.+]] = getelementptr inbounds nuw %struct.S, ptr [[S_ADDR]], i32 0, i32 0
-    // CHECK: [[S_DATA_5_ADDR:%.+]] = getelementptr inbounds nuw [6 x float], ptr [[S_DATA_ADDR]], i64 0, i64 5
+    // CHECK: [[S_DATA_5_ADDR:%.+]] = getelementptr inbounds [6 x float], ptr [[S_DATA_ADDR]], i64 0, i64 5
     // CHECK: [[S_DATA_6_ADDR:%.+]] = getelementptr float, ptr [[S_DATA_5_ADDR]], i32 1
     // CHECK: [[END_BC:%.+]] = ptrtoint ptr [[S_DATA_6_ADDR]] to i64
     // CHECK: [[BEG_BC:%.+]] = ptrtoint ptr [[S_DATA_0_ADDR]] to i64
diff --git a/clang/test/OpenMP/target_map_member_expr_codegen.cpp b/clang/test/OpenMP/target_map_member_expr_codegen.cpp
index fb36ba7b78d5bf..9b64647928a24e 100644
--- a/clang/test/OpenMP/target_map_member_expr_codegen.cpp
+++ b/clang/test/OpenMP/target_map_member_expr_codegen.cpp
@@ -223,7 +223,7 @@ void foo() {
 // CHECK-NEXT:    [[TMP10:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK-NEXT:    [[A4:%.*]] = getelementptr inbounds nuw [[STRUCT_DESCRIPTOR]], ptr [[TMP10]], i32 0, i32 0
 // CHECK-NEXT:    [[TMP11:%.*]] = load ptr, ptr [[A4]], align 8
-// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP11]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP11]], i64 0
 // CHECK-NEXT:    [[TMP12:%.*]] = load i32, ptr [[ASIZE]], align 4
 // CHECK-NEXT:    [[CONV:%.*]] = zext i32 [[TMP12]] to i64
 // CHECK-NEXT:    [[TMP13:%.*]] = mul nuw i64 [[CONV]], 4
@@ -233,7 +233,7 @@ void foo() {
 // CHECK-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[D_ADDR]], align 8
 // CHECK-NEXT:    [[C5:%.*]] = getelementptr inbounds nuw [[STRUCT_DESCRIPTOR]], ptr [[TMP16]], i32 0, i32 1
 // CHECK-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[C5]], align 8
-// CHECK-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw float, ptr [[TMP17]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, ptr [[TMP17]], i64 0
 // CHECK-NEXT:    [[TMP18:%.*]] = load i32, ptr [[CSIZE]], align 4
 // CHECK-NEXT:    [[CONV7:%.*]] = zext i32 [[TMP18]] to i64
 // CHECK-NEXT:    [[TMP19:%.*]] = mul nuw i64 [[CONV7]], 4
@@ -343,7 +343,7 @@ void foo() {
 // CHECK-NEXT:    [[TMP79:%.*]] = load ptr, ptr [[_TMP12]], align 8
 // CHECK-NEXT:    [[C15:%.*]] = getelementptr inbounds nuw [[STRUCT_DESCRIPTOR]], ptr [[TMP79]], i32 0, i32 1
 // CHECK-NEXT:    [[TMP80:%.*]] = load ptr, ptr [[C15]], align 8
-// CHECK-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw float, ptr [[TMP80]], i64 0
+// CHECK-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i64 0
 // CHECK-NEXT:    [[TMP81:%.*]] = load i32, ptr [[CSIZE]], align 4
 // CHECK-NEXT:    [[CONV17:%.*]] = zext i32 [[TMP81]] to i64
 // CHECK-NEXT:    [[TMP82:%.*]] = mul nuw i64 [[CONV17]], 4
diff --git a/clang/test/OpenMP/target_map_nest_defalut_mapper_codegen.cpp b/clang/test/OpenMP/target_map_nest_defalut_mapper_codegen.cpp
index 775f0b296b1b63..ffb145d8e50fc7 100644
--- a/clang/test/OpenMP/target_map_nest_defalut_mapper_codegen.cpp
+++ b/clang/test/OpenMP/target_map_nest_defalut_mapper_codegen.cpp
@@ -45,7 +45,7 @@ void foo() {
 // CHECK-NEXT:    [[F:%.*]] = getelementptr inbounds nuw [[STRUCT_D]], ptr [[ARRAYIDX1]], i32 0, i32 1
 // CHECK-NEXT:    [[A:%.*]] = getelementptr inbounds nuw [[STRUCT_C:%.*]], ptr [[F]], i32 0, i32 0
 // CHECK-NEXT:    store i32 222, ptr [[A]], align 4
-// CHECK-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw [10 x %struct.D], ptr [[SA]], i64 0, i64 0
+// CHECK-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds [10 x %struct.D], ptr [[SA]], i64 0, i64 0
 // CHECK-NEXT:    [[TMP0:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
 // CHECK-NEXT:    store ptr [[SA]], ptr [[TMP0]], align 8
 // CHECK-NEXT:    [[TMP1:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
diff --git a/clang/test/OpenMP/target_parallel_for_reduction_task_codegen.cpp b/clang/test/OpenMP/target_parallel_for_reduction_task_codegen.cpp
index 5cce677e88572d..3d0710acf0ee79 100644
--- a/clang/test/OpenMP/target_parallel_for_reduction_task_codegen.cpp
+++ b/clang/test/OpenMP/target_parallel_for_reduction_task_codegen.cpp
@@ -96,16 +96,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP4:%.*]] = sext i32 [[TMP3]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP4]]
 // CHECK1-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP5]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP5]], i64 9
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP9:%.*]] = sub i64 [[TMP7]], [[TMP8]]
@@ -135,7 +135,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP21:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP20]]
 // CHECK1-NEXT:    store ptr [[_TMP6]], ptr [[_TMP5]], align 8
 // CHECK1-NEXT:    store ptr [[TMP21]], ptr [[_TMP6]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP22]], align 8
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -150,19 +150,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP27]], align 8
 // CHECK1-NEXT:    [[TMP28:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP28]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP29:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP30:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP30]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds ptr, ptr [[TMP30]], i64 0
 // CHECK1-NEXT:    [[TMP31:%.*]] = load ptr, ptr [[ARRAYIDX8]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP31]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds i8, ptr [[TMP31]], i64 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP33:%.*]] = sext i32 [[TMP32]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN10:%.*]] = add nsw i64 -1, [[TMP33]]
 // CHECK1-NEXT:    [[TMP34:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP34]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds ptr, ptr [[TMP34]], i64 9
 // CHECK1-NEXT:    [[TMP35:%.*]] = load ptr, ptr [[ARRAYIDX11]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP35]], i64 [[LB_ADD_LEN10]]
+// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds i8, ptr [[TMP35]], i64 [[LB_ADD_LEN10]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP36:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX9]], ptr [[TMP36]], align 8
@@ -483,9 +483,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]
diff --git a/clang/test/OpenMP/target_parallel_reduction_task_codegen.cpp b/clang/test/OpenMP/target_parallel_reduction_task_codegen.cpp
index c0bb4a6d6cc82f..28d63dbf8c4a93 100644
--- a/clang/test/OpenMP/target_parallel_reduction_task_codegen.cpp
+++ b/clang/test/OpenMP/target_parallel_reduction_task_codegen.cpp
@@ -85,16 +85,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP4:%.*]] = sext i32 [[TMP3]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP4]]
 // CHECK1-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP5]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP5]], i64 9
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP9:%.*]] = sub i64 [[TMP7]], [[TMP8]]
@@ -124,7 +124,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP21:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP20]]
 // CHECK1-NEXT:    store ptr [[_TMP5]], ptr [[TMP]], align 8
 // CHECK1-NEXT:    store ptr [[TMP21]], ptr [[_TMP5]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP22]], align 8
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -139,19 +139,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP27]], align 8
 // CHECK1-NEXT:    [[TMP28:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP28]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP29:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP30:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP30]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds ptr, ptr [[TMP30]], i64 0
 // CHECK1-NEXT:    [[TMP31:%.*]] = load ptr, ptr [[ARRAYIDX7]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP31]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i8, ptr [[TMP31]], i64 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP33:%.*]] = sext i32 [[TMP32]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN9:%.*]] = add nsw i64 -1, [[TMP33]]
 // CHECK1-NEXT:    [[TMP34:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP34]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds ptr, ptr [[TMP34]], i64 9
 // CHECK1-NEXT:    [[TMP35:%.*]] = load ptr, ptr [[ARRAYIDX10]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP36:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX8]], ptr [[TMP36]], align 8
@@ -429,9 +429,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]
diff --git a/clang/test/OpenMP/target_task_affinity_codegen.cpp b/clang/test/OpenMP/target_task_affinity_codegen.cpp
index 53960cee4b730a..85c5d63a6cd9c8 100644
--- a/clang/test/OpenMP/target_task_affinity_codegen.cpp
+++ b/clang/test/OpenMP/target_task_affinity_codegen.cpp
@@ -76,7 +76,7 @@ int main() {
 // CHECK1-NEXT:    store i32 0, ptr [[RETVAL]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[A]], align 8
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[A]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[TMP1]], ptr [[TMP3]], align 8
 // CHECK1-NEXT:    [[TMP4:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
@@ -102,7 +102,7 @@ int main() {
 // CHECK1-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[TMP0]], align 8
 // CHECK1-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[B]], align 8
 // CHECK1-NEXT:    [[TMP18:%.*]] = load ptr, ptr [[B]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP18]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i32, ptr [[TMP18]], i64 0
 // CHECK1-NEXT:    [[TMP19:%.*]] = getelementptr inbounds [2 x ptr], ptr [[DOTOFFLOAD_BASEPTRS5]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[TMP17]], ptr [[TMP19]], align 8
 // CHECK1-NEXT:    [[TMP20:%.*]] = getelementptr inbounds [2 x ptr], ptr [[DOTOFFLOAD_PTRS6]], i32 0, i32 0
@@ -174,9 +174,9 @@ int main() {
 // CHECK1-NEXT:    [[TMP2:%.*]] = call ptr @__kmpc_omp_task_alloc(ptr @[[GLOB1]], i32 [[TMP0]], i32 1, i64 48, i64 8, ptr @.omp_task_entry.)
 // CHECK1-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [1 x %struct.kmp_task_affinity_info_t], ptr [[DOTAFFS_ARR_ADDR]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[A_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP4]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP4]], i64 0
 // CHECK1-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[A_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP5]], i64 1023
+// CHECK1-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i64 1023
 // CHECK1-NEXT:    [[TMP6:%.*]] = getelementptr i32, ptr [[ARRAYIDX1]], i32 1
 // CHECK1-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[TMP6]] to i64
@@ -299,7 +299,7 @@ int main() {
 // CHECK3-NEXT:    store i32 0, ptr [[RETVAL]], align 4
 // CHECK3-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[A]], align 4
 // CHECK3-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[A]], align 4
-// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP2]], i32 0
+// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i32 0
 // CHECK3-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
 // CHECK3-NEXT:    store ptr [[TMP1]], ptr [[TMP3]], align 4
 // CHECK3-NEXT:    [[TMP4:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
@@ -325,7 +325,7 @@ int main() {
 // CHECK3-NEXT:    [[TMP16:%.*]] = load ptr, ptr [[TMP0]], align 4
 // CHECK3-NEXT:    [[TMP17:%.*]] = load ptr, ptr [[B]], align 4
 // CHECK3-NEXT:    [[TMP18:%.*]] = load ptr, ptr [[B]], align 4
-// CHECK3-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP18]], i32 0
+// CHECK3-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i32, ptr [[TMP18]], i32 0
 // CHECK3-NEXT:    [[TMP19:%.*]] = getelementptr inbounds [2 x ptr], ptr [[DOTOFFLOAD_BASEPTRS5]], i32 0, i32 0
 // CHECK3-NEXT:    store ptr [[TMP17]], ptr [[TMP19]], align 4
 // CHECK3-NEXT:    [[TMP20:%.*]] = getelementptr inbounds [2 x ptr], ptr [[DOTOFFLOAD_PTRS6]], i32 0, i32 0
@@ -397,9 +397,9 @@ int main() {
 // CHECK3-NEXT:    [[TMP2:%.*]] = call ptr @__kmpc_omp_task_alloc(ptr @[[GLOB1]], i32 [[TMP0]], i32 1, i32 24, i32 4, ptr @.omp_task_entry.)
 // CHECK3-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [1 x %struct.kmp_task_affinity_info_t], ptr [[DOTAFFS_ARR_ADDR]], i32 0, i32 0
 // CHECK3-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[A_ADDR]], align 4
-// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP4]], i32 0
+// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP4]], i32 0
 // CHECK3-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[A_ADDR]], align 4
-// CHECK3-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP5]], i32 1023
+// CHECK3-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 1023
 // CHECK3-NEXT:    [[TMP6:%.*]] = getelementptr i32, ptr [[ARRAYIDX1]], i32 1
 // CHECK3-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i32
 // CHECK3-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[TMP6]] to i32
@@ -587,9 +587,9 @@ int main() {
 // CHECK9-NEXT:    [[TMP2:%.*]] = call ptr @__kmpc_omp_task_alloc(ptr @[[GLOB1]], i32 [[TMP0]], i32 1, i64 48, i64 8, ptr @.omp_task_entry.)
 // CHECK9-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [1 x %struct.kmp_task_affinity_info_t], ptr [[DOTAFFS_ARR_ADDR]], i64 0, i64 0
 // CHECK9-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[A_ADDR]], align 8
-// CHECK9-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP4]], i64 0
+// CHECK9-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP4]], i64 0
 // CHECK9-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[A_ADDR]], align 8
-// CHECK9-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP5]], i64 1023
+// CHECK9-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i64 1023
 // CHECK9-NEXT:    [[TMP6:%.*]] = getelementptr i32, ptr [[ARRAYIDX1]], i32 1
 // CHECK9-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK9-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[TMP6]] to i64
@@ -709,9 +709,9 @@ int main() {
 // CHECK11-NEXT:    [[TMP2:%.*]] = call ptr @__kmpc_omp_task_alloc(ptr @[[GLOB1]], i32 [[TMP0]], i32 1, i32 24, i32 4, ptr @.omp_task_entry.)
 // CHECK11-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [1 x %struct.kmp_task_affinity_info_t], ptr [[DOTAFFS_ARR_ADDR]], i32 0, i32 0
 // CHECK11-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[A_ADDR]], align 4
-// CHECK11-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP4]], i32 0
+// CHECK11-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TMP4]], i32 0
 // CHECK11-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[A_ADDR]], align 4
-// CHECK11-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP5]], i32 1023
+// CHECK11-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 1023
 // CHECK11-NEXT:    [[TMP6:%.*]] = getelementptr i32, ptr [[ARRAYIDX1]], i32 1
 // CHECK11-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i32
 // CHECK11-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[TMP6]] to i32
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
index 2c36b410af0642..6f671dbb27abba 100644
--- 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
@@ -91,16 +91,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP4:%.*]] = sext i32 [[TMP3]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP4]]
 // CHECK1-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP5]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP5]], i64 9
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP9:%.*]] = sub i64 [[TMP7]], [[TMP8]]
@@ -130,7 +130,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP21:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP20]]
 // CHECK1-NEXT:    store ptr [[_TMP5]], ptr [[TMP]], align 8
 // CHECK1-NEXT:    store ptr [[TMP21]], ptr [[_TMP5]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP22]], align 8
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -145,19 +145,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP27]], align 8
 // CHECK1-NEXT:    [[TMP28:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP28]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP29:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP30:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP30]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds ptr, ptr [[TMP30]], i64 0
 // CHECK1-NEXT:    [[TMP31:%.*]] = load ptr, ptr [[ARRAYIDX7]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP31]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i8, ptr [[TMP31]], i64 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP33:%.*]] = sext i32 [[TMP32]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN9:%.*]] = add nsw i64 -1, [[TMP33]]
 // CHECK1-NEXT:    [[TMP34:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP34]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds ptr, ptr [[TMP34]], i64 9
 // CHECK1-NEXT:    [[TMP35:%.*]] = load ptr, ptr [[ARRAYIDX10]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP36:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX8]], ptr [[TMP36]], align 8
@@ -435,16 +435,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP3]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP3]], i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP4]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP4]], i64 0
 // CHECK1-NEXT:    [[TMP5:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP6:%.*]] = sext i32 [[TMP5]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP6]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP7]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP7]], i64 9
 // CHECK1-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP8]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP8]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP9:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP10:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP11:%.*]] = sub i64 [[TMP9]], [[TMP10]]
@@ -474,7 +474,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP22]]
 // CHECK1-NEXT:    store ptr [[_TMP6]], ptr [[_TMP5]], align 8
 // CHECK1-NEXT:    store ptr [[TMP23]], ptr [[_TMP6]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP24:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP24]], align 8
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -489,19 +489,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..4, ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP30:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP30]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP31:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP32]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds ptr, ptr [[TMP32]], i64 0
 // CHECK1-NEXT:    [[TMP33:%.*]] = load ptr, ptr [[ARRAYIDX8]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP33]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds i8, ptr [[TMP33]], i64 0
 // CHECK1-NEXT:    [[TMP34:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP35:%.*]] = sext i32 [[TMP34]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN10:%.*]] = add nsw i64 -1, [[TMP35]]
 // CHECK1-NEXT:    [[TMP36:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP36]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds ptr, ptr [[TMP36]], i64 9
 // CHECK1-NEXT:    [[TMP37:%.*]] = load ptr, ptr [[ARRAYIDX11]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP37]], i64 [[LB_ADD_LEN10]]
+// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds i8, ptr [[TMP37]], i64 [[LB_ADD_LEN10]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP31]], align 8
 // CHECK1-NEXT:    [[TMP38:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX9]], ptr [[TMP38]], align 8
@@ -822,9 +822,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]
diff --git a/clang/test/OpenMP/target_update_codegen.cpp b/clang/test/OpenMP/target_update_codegen.cpp
index c8211f475c7fce..5e038989ab6dd4 100644
--- a/clang/test/OpenMP/target_update_codegen.cpp
+++ b/clang/test/OpenMP/target_update_codegen.cpp
@@ -1118,9 +1118,9 @@ struct ST {
 void foo(int arg) {
   ST arr[3][4];
   // CK20: [[DIMS:%.+]] = alloca [3 x [[STRUCT_DESCRIPTOR]]],
-  // CK20: [[ARRAY_IDX:%.+]] = getelementptr inbounds nuw [3 x [4 x [[STRUCT_ST]]]], ptr [[ARR:%.+]], {{.+}} 0, {{.+}} 0
+  // CK20: [[ARRAY_IDX:%.+]] = getelementptr inbounds [3 x [4 x [[STRUCT_ST]]]], ptr [[ARR:%.+]], {{.+}} 0, {{.+}} 0
   // CK20: [[ARRAY_DECAY:%.+]] = getelementptr inbounds [4 x [[STRUCT_ST]]], ptr [[ARRAY_IDX]], {{.+}} 0, {{.+}} 0
-  // CK20: [[ARRAY_IDX_1:%.+]] = getelementptr inbounds nuw [[STRUCT_ST]], ptr [[ARRAY_DECAY]], {{.+}}
+  // CK20: [[ARRAY_IDX_1:%.+]] = getelementptr inbounds [[STRUCT_ST]], ptr [[ARRAY_DECAY]], {{.+}}
   // CK20: [[BP0:%.+]] = getelementptr inbounds [1 x ptr], ptr [[BP:%.+]], {{.+}} 0, {{.+}} 0
   // CK20: store ptr [[ARR]], ptr [[BP0]],
   // CK20: [[P0:%.+]] = getelementptr inbounds [1 x ptr], ptr [[P:%.+]], {{.+}} 0, {{.+}} 0
@@ -1186,9 +1186,9 @@ struct ST {
   // CK21: _ZN2ST3fooEv
   void foo() {
     // CK21: [[DIMS:%.+]] = alloca [4 x [[STRUCT_DESCRIPTOR]]],
-    // CK21: [[ARRAY_IDX:%.+]] = getelementptr inbounds nuw [10 x [10 x [10 x ptr]]], ptr [[DPTR:%.+]], {{.+}} 0, {{.+}} 0
+    // CK21: [[ARRAY_IDX:%.+]] = getelementptr inbounds [10 x [10 x [10 x ptr]]], ptr [[DPTR:%.+]], {{.+}} 0, {{.+}} 0
     // CK21: [[ARRAY_DECAY:%.+]] = getelementptr inbounds [10 x [10 x ptr]], ptr [[ARRAY_IDX]], {{.+}} 0, {{.+}} 0
-    // CK21: [[ARRAY_IDX_1:%.+]] = getelementptr inbounds nuw [10 x ptr], ptr [[ARRAY_DECAY]], {{.+}} 1
+    // CK21: [[ARRAY_IDX_1:%.+]] = getelementptr inbounds [10 x ptr], ptr [[ARRAY_DECAY]], {{.+}} 1
     // CK21: [[ARRAY_DECAY_2:%.+]] = getelementptr inbounds [10 x ptr], ptr [[ARRAY_IDX_1]], {{.+}} 0, {{.+}} 0
     // CK21: [[ARRAY_IDX_3:%.+]] = getelementptr inbounds {{.+}}, ptr [[ARRAY_DECAY_2]], {{.+}} 0
     // CK21: [[BP0:%.+]] = getelementptr inbounds [2 x ptr], ptr [[BP:%.+]], {{.+}} 0, {{.+}} 0
@@ -1262,9 +1262,9 @@ struct ST {
   // CK22: _ZN2ST3fooEPA10_Pi
   void foo(int *arr[5][10]) {
     // CK22: [[DIMS:%.+]] = alloca [4 x [[STRUCT_DESCRIPTOR]]],
-    // CK22: [[ARRAY_IDX:%.+]] = getelementptr inbounds nuw [10 x ptr], ptr [[ARR:%.+]], {{.+}} 0
+    // CK22: [[ARRAY_IDX:%.+]] = getelementptr inbounds [10 x ptr], ptr [[ARR:%.+]], {{.+}} 0
     // CK22: [[ARRAY_DECAY:%.+]] = getelementptr inbounds [10 x ptr], ptr [[ARRAY_IDX]], {{.+}} 0, {{.+}} 0
-    // CK22: [[ARRAY_IDX_2:%.+]] = getelementptr inbounds nuw ptr, ptr [[ARRAY_DECAY:%.+]], {{.+}} 1
+    // CK22: [[ARRAY_IDX_2:%.+]] = getelementptr inbounds ptr, ptr [[ARRAY_DECAY:%.+]], {{.+}} 1
     // CK22: [[BP0:%.+]] = getelementptr inbounds [1 x ptr], ptr [[BP:%.+]], {{.+}} 0, {{.+}} 0
     // CK22: [[P0:%.+]] = getelementptr inbounds [1 x ptr], ptr [[P:%.+]], i{{.+}} 0, i{{.+}} 0
     // CK22: [[DIM_1:%.+]] = getelementptr inbounds [4 x [[STRUCT_DESCRIPTOR]]], ptr [[DIMS]], {{.+}} 0, {{.+}} 0
@@ -1338,11 +1338,11 @@ void foo(int arg) {
   float farr[5][5][5];
   // CK23: [[ARG_ADDR:%.+]] = alloca i32,
   // CK23: [[DIMS:%.+]] = alloca [4 x [[STRUCT_DESCRIPTOR]]],
-  // CK23: [[ARRAY_IDX:%.+]] = getelementptr inbounds nuw [5 x [5 x [5 x float]]], ptr [[ARR:%.+]], {{.+}} 0, {{.+}} 0
+  // CK23: [[ARRAY_IDX:%.+]] = getelementptr inbounds [5 x [5 x [5 x float]]], ptr [[ARR:%.+]], {{.+}} 0, {{.+}} 0
   // CK23: [[ARRAY_DECAY:%.+]] = getelementptr inbounds [5 x [5 x float]], ptr [[ARRAY_IDX]], {{.+}} 0, {{.+}} 0
-  // CK23: [[ARRAY_IDX_1:%.+]] = getelementptr inbounds nuw [5 x float], ptr [[ARRAY_DECAY]], {{.+}}
+  // CK23: [[ARRAY_IDX_1:%.+]] = getelementptr inbounds [5 x float], ptr [[ARRAY_DECAY]], {{.+}}
   // CK23: [[ARRAY_DECAY_2:%.+]] = getelementptr inbounds [5 x float], ptr [[ARRAY_IDX_1]], {{.+}} 0, {{.+}} 0
-  // CK23: [[ARRAY_IDX_2:%.+]] = getelementptr inbounds nuw float, ptr [[ARRAY_DECAY_2]], {{.+}}
+  // CK23: [[ARRAY_IDX_2:%.+]] = getelementptr inbounds float, ptr [[ARRAY_DECAY_2]], {{.+}}
   // CK23: [[MUL:%.+]] = mul nuw i64 4,
   // CK23: [[BP0:%.+]] = getelementptr inbounds [1 x ptr], ptr [[BP:%.+]], {{.+}} 0, {{.+}} 0
   // CK23: store ptr [[ARR]], ptr [[BP0]],
@@ -1411,11 +1411,11 @@ void foo(int arg) {
 void foo(int arg) {
   double darr[3][4][5];
   // CK24: [[DIMS:%.+]] = alloca [4 x [[STRUCT_DESCRIPTOR]]],
-  // CK24: [[ARRAY_IDX:%.+]] = getelementptr inbounds nuw [3 x [4 x [5 x double]]], ptr [[ARR:%.+]], {{.+}} 0, {{.+}} 0
+  // CK24: [[ARRAY_IDX:%.+]] = getelementptr inbounds [3 x [4 x [5 x double]]], ptr [[ARR:%.+]], {{.+}} 0, {{.+}} 0
   // CK24: [[ARRAY_DECAY:%.+]] = getelementptr inbounds [4 x [5 x double]], ptr [[ARRAY_IDX]], {{.+}} 0, {{.+}} 0
-  // CK24: [[ARRAY_IDX_1:%.+]] = getelementptr inbounds nuw [5 x double], ptr [[ARRAY_DECAY]], {{.+}}
+  // CK24: [[ARRAY_IDX_1:%.+]] = getelementptr inbounds [5 x double], ptr [[ARRAY_DECAY]], {{.+}}
   // CK24: [[ARRAY_DECAY_2:%.+]] = getelementptr inbounds [5 x double], ptr [[ARRAY_IDX_1]], {{.+}} 0, {{.+}} 0
-  // CK24: [[ARRAY_IDX_2:%.+]] = getelementptr inbounds nuw double, ptr [[ARRAY_DECAY_2]], {{.+}}
+  // CK24: [[ARRAY_IDX_2:%.+]] = getelementptr inbounds double, ptr [[ARRAY_DECAY_2]], {{.+}}
   // CK24: [[MUL:%.+]] = mul nuw i64 8,
   // CK24: [[SUB:%.+]] = sub nuw i64 4, [[ARG:%.+]]
   // CK24: [[LEN:%.+]] = udiv {{.+}} [[SUB]], 1
@@ -1488,15 +1488,15 @@ void foo(int arg) {
 
   // CK25: [[DIMS:%.+]] = alloca [4 x [[STRUCT_DESCRIPTOR]]],
   // CK25: [[DIMS_2:%.+]] = alloca [3 x [[STRUCT_DESCRIPTOR]]],
-  // CK25: [[ARRAY_IDX:%.+]] = getelementptr inbounds nuw [3 x [4 x [5 x i32]]], ptr [[ARR:%.+]], {{.+}} 0, {{.+}} 0
+  // CK25: [[ARRAY_IDX:%.+]] = getelementptr inbounds [3 x [4 x [5 x i32]]], ptr [[ARR:%.+]], {{.+}} 0, {{.+}} 0
   // CK25: [[ARRAY_DECAY:%.+]] = getelementptr inbounds [4 x [5 x i32]], ptr [[ARRAY_IDX]], {{.+}} 0, {{.+}} 0
-  // CK25: [[ARRAY_IDX_1:%.+]] = getelementptr inbounds nuw [5 x i32], ptr [[ARRAY_DECAY]], {{.+}}
+  // CK25: [[ARRAY_IDX_1:%.+]] = getelementptr inbounds [5 x i32], ptr [[ARRAY_DECAY]], {{.+}}
   // CK25: [[ARRAY_DECAY_2:%.+]] = getelementptr inbounds [5 x i32], ptr [[ARRAY_IDX_1]], {{.+}} 0, {{.+}} 0
-  // CK25: [[ARRAY_IDX_3:%.+]] = getelementptr inbounds nuw {{.+}}, ptr [[ARRAY_DECAY_2]], {{.+}} 1
+  // CK25: [[ARRAY_IDX_3:%.+]] = getelementptr inbounds {{.+}}, ptr [[ARRAY_DECAY_2]], {{.+}} 1
   // CK25: [[LEN:%.+]] = sub nuw i64 4, [[ARG_ADDR:%.+]]
-  // CK25: [[ARRAY_IDX_4:%.+]] = getelementptr inbounds nuw [4 x [3 x float]], ptr [[FARR:%.+]], {{.+}} 0, {{.+}} 0
+  // CK25: [[ARRAY_IDX_4:%.+]] = getelementptr inbounds [4 x [3 x float]], ptr [[FARR:%.+]], {{.+}} 0, {{.+}} 0
   // CK25: [[ARRAY_DECAY_5:%.+]] = getelementptr inbounds [3 x float], ptr [[ARRAY_IDX_4]], {{.+}} 0, {{.+}} 0
-  // CK25: [[ARRAY_IDX_6:%.+]] = getelementptr inbounds nuw float, ptr [[ARRAY_DECAY_5:%.+]], {{.+}} 1
+  // CK25: [[ARRAY_IDX_6:%.+]] = getelementptr inbounds float, ptr [[ARRAY_DECAY_5:%.+]], {{.+}} 1
   // CK25: [[BP0:%.+]] = getelementptr inbounds [3 x ptr], ptr [[BP:%.+]], i{{.+}} 0, i{{.+}} 0
   // CK25: [[P0:%.+]] = getelementptr inbounds [3 x ptr], ptr [[P:%.+]], i{{.+}} 0, i{{.+}} 0
   // CK25: [[DIM_1:%.+]] = getelementptr inbounds [4 x [[STRUCT_DESCRIPTOR]]], ptr [[DIMS]], {{.+}} 0, {{.+}} 0
diff --git a/clang/test/OpenMP/task_codegen.c b/clang/test/OpenMP/task_codegen.c
index d08eb3762d5c94..0d10cbce4aa807 100644
--- a/clang/test/OpenMP/task_codegen.c
+++ b/clang/test/OpenMP/task_codegen.c
@@ -183,7 +183,7 @@ for (int i = 0; i < 10; ++i)
   // CHECK: [[A:%.+]] = load ptr, ptr [[A_ADDR:%.+]],
   // CHECK: [[K:%.+]] = load i32, ptr [[K_ADDR]],
   // CHECK: [[IDX:%.+]] = zext i32 [[K]] to i64
-  // CHECK: [[AK_ADDR:%.+]] = getelementptr inbounds nuw ptr, ptr [[A]], i64 [[IDX]]
+  // CHECK: [[AK_ADDR:%.+]] = getelementptr inbounds ptr, ptr [[A]], i64 [[IDX]]
   // CHECK: [[AK:%.+]] = load ptr, ptr [[AK_ADDR]],
   // CHECK: [[I:%.+]] = load i32, ptr [[I_ADDR]],
   // CHECK: [[IDX:%.+]] = sext i32 [[I]] to i64
diff --git a/clang/test/OpenMP/task_codegen.cpp b/clang/test/OpenMP/task_codegen.cpp
index c3e6d9e6b1cf72..b256c41132ed39 100644
--- a/clang/test/OpenMP/task_codegen.cpp
+++ b/clang/test/OpenMP/task_codegen.cpp
@@ -309,9 +309,9 @@ void test_omp_all_memory()
 // CHECK1-NEXT:    [[TMP33:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP30]], i32 0, i32 2
 // CHECK1-NEXT:    store i8 1, ptr [[TMP33]], align 8
 // CHECK1-NEXT:    [[TMP34:%.*]] = mul nsw i64 0, [[TMP2]]
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP34]]
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP34]]
 // CHECK1-NEXT:    [[TMP35:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP35]]
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP35]]
 // CHECK1-NEXT:    [[TMP36:%.*]] = getelementptr i32, ptr [[ARRAYIDX2]], i32 1
 // CHECK1-NEXT:    [[TMP37:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK1-NEXT:    [[TMP38:%.*]] = ptrtoint ptr [[TMP36]] to i64
@@ -346,13 +346,13 @@ void test_omp_all_memory()
 // CHECK1-NEXT:    [[TMP58:%.*]] = load i8, ptr [[B]], align 1
 // CHECK1-NEXT:    [[TMP59:%.*]] = sext i8 [[TMP58]] to i64
 // CHECK1-NEXT:    [[TMP60:%.*]] = mul nsw i64 4, [[TMP2]]
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP60]]
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX7]], i64 [[TMP59]]
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP60]]
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX7]], i64 [[TMP59]]
 // CHECK1-NEXT:    [[TMP61:%.*]] = load i8, ptr [[B]], align 1
 // CHECK1-NEXT:    [[TMP62:%.*]] = sext i8 [[TMP61]] to i64
 // CHECK1-NEXT:    [[TMP63:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP63]]
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX9]], i64 [[TMP62]]
+// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP63]]
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX9]], i64 [[TMP62]]
 // CHECK1-NEXT:    [[TMP64:%.*]] = getelementptr i32, ptr [[ARRAYIDX10]], i32 1
 // CHECK1-NEXT:    [[TMP65:%.*]] = ptrtoint ptr [[ARRAYIDX8]] to i64
 // CHECK1-NEXT:    [[TMP66:%.*]] = ptrtoint ptr [[TMP64]] to i64
@@ -384,13 +384,13 @@ void test_omp_all_memory()
 // CHECK1-NEXT:    [[TMP83:%.*]] = load i8, ptr [[B]], align 1
 // CHECK1-NEXT:    [[TMP84:%.*]] = sext i8 [[TMP83]] to i64
 // CHECK1-NEXT:    [[TMP85:%.*]] = mul nsw i64 4, [[TMP2]]
-// CHECK1-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP85]]
-// CHECK1-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX15]], i64 [[TMP84]]
+// CHECK1-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP85]]
+// CHECK1-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX15]], i64 [[TMP84]]
 // CHECK1-NEXT:    [[TMP86:%.*]] = load i8, ptr [[B]], align 1
 // CHECK1-NEXT:    [[TMP87:%.*]] = sext i8 [[TMP86]] to i64
 // CHECK1-NEXT:    [[TMP88:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK1-NEXT:    [[ARRAYIDX17:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP88]]
-// CHECK1-NEXT:    [[ARRAYIDX18:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX17]], i64 [[TMP87]]
+// CHECK1-NEXT:    [[ARRAYIDX17:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP88]]
+// CHECK1-NEXT:    [[ARRAYIDX18:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX17]], i64 [[TMP87]]
 // CHECK1-NEXT:    [[TMP89:%.*]] = getelementptr i32, ptr [[ARRAYIDX18]], i32 1
 // CHECK1-NEXT:    [[TMP90:%.*]] = ptrtoint ptr [[ARRAYIDX16]] to i64
 // CHECK1-NEXT:    [[TMP91:%.*]] = ptrtoint ptr [[TMP89]] to i64
@@ -427,8 +427,8 @@ void test_omp_all_memory()
 // CHECK1-NEXT:    [[TMP111:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP108]], i32 0, i32 2
 // CHECK1-NEXT:    store i8 3, ptr [[TMP111]], align 8
 // CHECK1-NEXT:    [[TMP112:%.*]] = mul nsw i64 0, [[TMP2]]
-// CHECK1-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP112]]
-// CHECK1-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX23]], i64 3
+// CHECK1-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP112]]
+// CHECK1-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX23]], i64 3
 // CHECK1-NEXT:    [[TMP113:%.*]] = load i32, ptr @a, align 4
 // CHECK1-NEXT:    [[TMP114:%.*]] = sext i32 [[TMP113]] to i64
 // CHECK1-NEXT:    [[LEN_SUB_1:%.*]] = sub nsw i64 [[TMP114]], 1
@@ -436,8 +436,8 @@ void test_omp_all_memory()
 // CHECK1-NEXT:    [[TMP116:%.*]] = sext i32 [[TMP115]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP116]]
 // CHECK1-NEXT:    [[TMP117:%.*]] = mul nsw i64 [[LB_ADD_LEN]], [[TMP2]]
-// CHECK1-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP117]]
-// CHECK1-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX25]], i64 [[LEN_SUB_1]]
+// CHECK1-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP117]]
+// CHECK1-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX25]], i64 [[LEN_SUB_1]]
 // CHECK1-NEXT:    [[TMP118:%.*]] = getelementptr i32, ptr [[ARRAYIDX26]], i32 1
 // CHECK1-NEXT:    [[TMP119:%.*]] = ptrtoint ptr [[ARRAYIDX24]] to i64
 // CHECK1-NEXT:    [[TMP120:%.*]] = ptrtoint ptr [[TMP118]] to i64
@@ -1432,9 +1432,9 @@ void test_omp_all_memory()
 // CHECK1-51-NEXT:    [[TMP33:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP30]], i32 0, i32 2
 // CHECK1-51-NEXT:    store i8 1, ptr [[TMP33]], align 8
 // CHECK1-51-NEXT:    [[TMP34:%.*]] = mul nsw i64 0, [[TMP2]]
-// CHECK1-51-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP34]]
+// CHECK1-51-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP34]]
 // CHECK1-51-NEXT:    [[TMP35:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK1-51-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP35]]
+// CHECK1-51-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP35]]
 // CHECK1-51-NEXT:    [[TMP36:%.*]] = getelementptr i32, ptr [[ARRAYIDX2]], i32 1
 // CHECK1-51-NEXT:    [[TMP37:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK1-51-NEXT:    [[TMP38:%.*]] = ptrtoint ptr [[TMP36]] to i64
@@ -1469,13 +1469,13 @@ void test_omp_all_memory()
 // CHECK1-51-NEXT:    [[TMP58:%.*]] = load i8, ptr [[B]], align 1
 // CHECK1-51-NEXT:    [[TMP59:%.*]] = sext i8 [[TMP58]] to i64
 // CHECK1-51-NEXT:    [[TMP60:%.*]] = mul nsw i64 4, [[TMP2]]
-// CHECK1-51-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP60]]
-// CHECK1-51-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX7]], i64 [[TMP59]]
+// CHECK1-51-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP60]]
+// CHECK1-51-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX7]], i64 [[TMP59]]
 // CHECK1-51-NEXT:    [[TMP61:%.*]] = load i8, ptr [[B]], align 1
 // CHECK1-51-NEXT:    [[TMP62:%.*]] = sext i8 [[TMP61]] to i64
 // CHECK1-51-NEXT:    [[TMP63:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK1-51-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP63]]
-// CHECK1-51-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX9]], i64 [[TMP62]]
+// CHECK1-51-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP63]]
+// CHECK1-51-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX9]], i64 [[TMP62]]
 // CHECK1-51-NEXT:    [[TMP64:%.*]] = getelementptr i32, ptr [[ARRAYIDX10]], i32 1
 // CHECK1-51-NEXT:    [[TMP65:%.*]] = ptrtoint ptr [[ARRAYIDX8]] to i64
 // CHECK1-51-NEXT:    [[TMP66:%.*]] = ptrtoint ptr [[TMP64]] to i64
@@ -1507,13 +1507,13 @@ void test_omp_all_memory()
 // CHECK1-51-NEXT:    [[TMP83:%.*]] = load i8, ptr [[B]], align 1
 // CHECK1-51-NEXT:    [[TMP84:%.*]] = sext i8 [[TMP83]] to i64
 // CHECK1-51-NEXT:    [[TMP85:%.*]] = mul nsw i64 4, [[TMP2]]
-// CHECK1-51-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP85]]
-// CHECK1-51-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX15]], i64 [[TMP84]]
+// CHECK1-51-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP85]]
+// CHECK1-51-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX15]], i64 [[TMP84]]
 // CHECK1-51-NEXT:    [[TMP86:%.*]] = load i8, ptr [[B]], align 1
 // CHECK1-51-NEXT:    [[TMP87:%.*]] = sext i8 [[TMP86]] to i64
 // CHECK1-51-NEXT:    [[TMP88:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK1-51-NEXT:    [[ARRAYIDX17:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP88]]
-// CHECK1-51-NEXT:    [[ARRAYIDX18:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX17]], i64 [[TMP87]]
+// CHECK1-51-NEXT:    [[ARRAYIDX17:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP88]]
+// CHECK1-51-NEXT:    [[ARRAYIDX18:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX17]], i64 [[TMP87]]
 // CHECK1-51-NEXT:    [[TMP89:%.*]] = getelementptr i32, ptr [[ARRAYIDX18]], i32 1
 // CHECK1-51-NEXT:    [[TMP90:%.*]] = ptrtoint ptr [[ARRAYIDX16]] to i64
 // CHECK1-51-NEXT:    [[TMP91:%.*]] = ptrtoint ptr [[TMP89]] to i64
@@ -1550,8 +1550,8 @@ void test_omp_all_memory()
 // CHECK1-51-NEXT:    [[TMP111:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP108]], i32 0, i32 2
 // CHECK1-51-NEXT:    store i8 3, ptr [[TMP111]], align 8
 // CHECK1-51-NEXT:    [[TMP112:%.*]] = mul nsw i64 0, [[TMP2]]
-// CHECK1-51-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP112]]
-// CHECK1-51-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX23]], i64 3
+// CHECK1-51-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP112]]
+// CHECK1-51-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX23]], i64 3
 // CHECK1-51-NEXT:    [[TMP113:%.*]] = load i32, ptr @a, align 4
 // CHECK1-51-NEXT:    [[TMP114:%.*]] = sext i32 [[TMP113]] to i64
 // CHECK1-51-NEXT:    [[LEN_SUB_1:%.*]] = sub nsw i64 [[TMP114]], 1
@@ -1559,8 +1559,8 @@ void test_omp_all_memory()
 // CHECK1-51-NEXT:    [[TMP116:%.*]] = sext i32 [[TMP115]] to i64
 // CHECK1-51-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP116]]
 // CHECK1-51-NEXT:    [[TMP117:%.*]] = mul nsw i64 [[LB_ADD_LEN]], [[TMP2]]
-// CHECK1-51-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP117]]
-// CHECK1-51-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX25]], i64 [[LEN_SUB_1]]
+// CHECK1-51-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP117]]
+// CHECK1-51-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX25]], i64 [[LEN_SUB_1]]
 // CHECK1-51-NEXT:    [[TMP118:%.*]] = getelementptr i32, ptr [[ARRAYIDX26]], i32 1
 // CHECK1-51-NEXT:    [[TMP119:%.*]] = ptrtoint ptr [[ARRAYIDX24]] to i64
 // CHECK1-51-NEXT:    [[TMP120:%.*]] = ptrtoint ptr [[TMP118]] to i64
@@ -1595,8 +1595,8 @@ void test_omp_all_memory()
 // CHECK1-51-NEXT:    [[TMP139:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP136]], i32 0, i32 2
 // CHECK1-51-NEXT:    store i8 8, ptr [[TMP139]], align 8
 // CHECK1-51-NEXT:    [[TMP140:%.*]] = mul nsw i64 0, [[TMP2]]
-// CHECK1-51-NEXT:    [[ARRAYIDX31:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP140]]
-// CHECK1-51-NEXT:    [[ARRAYIDX32:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX31]], i64 3
+// CHECK1-51-NEXT:    [[ARRAYIDX31:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP140]]
+// CHECK1-51-NEXT:    [[ARRAYIDX32:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX31]], i64 3
 // CHECK1-51-NEXT:    [[TMP141:%.*]] = load i32, ptr @a, align 4
 // CHECK1-51-NEXT:    [[TMP142:%.*]] = sext i32 [[TMP141]] to i64
 // CHECK1-51-NEXT:    [[LEN_SUB_133:%.*]] = sub nsw i64 [[TMP142]], 1
@@ -1604,8 +1604,8 @@ void test_omp_all_memory()
 // CHECK1-51-NEXT:    [[TMP144:%.*]] = sext i32 [[TMP143]] to i64
 // CHECK1-51-NEXT:    [[LB_ADD_LEN34:%.*]] = add nsw i64 -1, [[TMP144]]
 // CHECK1-51-NEXT:    [[TMP145:%.*]] = mul nsw i64 [[LB_ADD_LEN34]], [[TMP2]]
-// CHECK1-51-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP145]]
-// CHECK1-51-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_133]]
+// CHECK1-51-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP145]]
+// CHECK1-51-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_133]]
 // CHECK1-51-NEXT:    [[TMP146:%.*]] = getelementptr i32, ptr [[ARRAYIDX36]], i32 1
 // CHECK1-51-NEXT:    [[TMP147:%.*]] = ptrtoint ptr [[ARRAYIDX32]] to i64
 // CHECK1-51-NEXT:    [[TMP148:%.*]] = ptrtoint ptr [[TMP146]] to i64
@@ -3040,9 +3040,9 @@ void test_omp_all_memory()
 // CHECK2-NEXT:    [[TMP33:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP30]], i32 0, i32 2
 // CHECK2-NEXT:    store i8 1, ptr [[TMP33]], align 8
 // CHECK2-NEXT:    [[TMP34:%.*]] = mul nsw i64 0, [[TMP2]]
-// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP34]]
+// CHECK2-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP34]]
 // CHECK2-NEXT:    [[TMP35:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK2-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP35]]
+// CHECK2-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP35]]
 // CHECK2-NEXT:    [[TMP36:%.*]] = getelementptr i32, ptr [[ARRAYIDX2]], i32 1
 // CHECK2-NEXT:    [[TMP37:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK2-NEXT:    [[TMP38:%.*]] = ptrtoint ptr [[TMP36]] to i64
@@ -3077,13 +3077,13 @@ void test_omp_all_memory()
 // CHECK2-NEXT:    [[TMP58:%.*]] = load i8, ptr [[B]], align 1
 // CHECK2-NEXT:    [[TMP59:%.*]] = sext i8 [[TMP58]] to i64
 // CHECK2-NEXT:    [[TMP60:%.*]] = mul nsw i64 4, [[TMP2]]
-// CHECK2-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP60]]
-// CHECK2-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX7]], i64 [[TMP59]]
+// CHECK2-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP60]]
+// CHECK2-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX7]], i64 [[TMP59]]
 // CHECK2-NEXT:    [[TMP61:%.*]] = load i8, ptr [[B]], align 1
 // CHECK2-NEXT:    [[TMP62:%.*]] = sext i8 [[TMP61]] to i64
 // CHECK2-NEXT:    [[TMP63:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK2-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP63]]
-// CHECK2-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX9]], i64 [[TMP62]]
+// CHECK2-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP63]]
+// CHECK2-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX9]], i64 [[TMP62]]
 // CHECK2-NEXT:    [[TMP64:%.*]] = getelementptr i32, ptr [[ARRAYIDX10]], i32 1
 // CHECK2-NEXT:    [[TMP65:%.*]] = ptrtoint ptr [[ARRAYIDX8]] to i64
 // CHECK2-NEXT:    [[TMP66:%.*]] = ptrtoint ptr [[TMP64]] to i64
@@ -3115,13 +3115,13 @@ void test_omp_all_memory()
 // CHECK2-NEXT:    [[TMP83:%.*]] = load i8, ptr [[B]], align 1
 // CHECK2-NEXT:    [[TMP84:%.*]] = sext i8 [[TMP83]] to i64
 // CHECK2-NEXT:    [[TMP85:%.*]] = mul nsw i64 4, [[TMP2]]
-// CHECK2-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP85]]
-// CHECK2-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX15]], i64 [[TMP84]]
+// CHECK2-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP85]]
+// CHECK2-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX15]], i64 [[TMP84]]
 // CHECK2-NEXT:    [[TMP86:%.*]] = load i8, ptr [[B]], align 1
 // CHECK2-NEXT:    [[TMP87:%.*]] = sext i8 [[TMP86]] to i64
 // CHECK2-NEXT:    [[TMP88:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK2-NEXT:    [[ARRAYIDX17:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP88]]
-// CHECK2-NEXT:    [[ARRAYIDX18:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX17]], i64 [[TMP87]]
+// CHECK2-NEXT:    [[ARRAYIDX17:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP88]]
+// CHECK2-NEXT:    [[ARRAYIDX18:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX17]], i64 [[TMP87]]
 // CHECK2-NEXT:    [[TMP89:%.*]] = getelementptr i32, ptr [[ARRAYIDX18]], i32 1
 // CHECK2-NEXT:    [[TMP90:%.*]] = ptrtoint ptr [[ARRAYIDX16]] to i64
 // CHECK2-NEXT:    [[TMP91:%.*]] = ptrtoint ptr [[TMP89]] to i64
@@ -3158,8 +3158,8 @@ void test_omp_all_memory()
 // CHECK2-NEXT:    [[TMP111:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP108]], i32 0, i32 2
 // CHECK2-NEXT:    store i8 3, ptr [[TMP111]], align 8
 // CHECK2-NEXT:    [[TMP112:%.*]] = mul nsw i64 0, [[TMP2]]
-// CHECK2-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP112]]
-// CHECK2-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX23]], i64 3
+// CHECK2-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP112]]
+// CHECK2-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX23]], i64 3
 // CHECK2-NEXT:    [[TMP113:%.*]] = load i32, ptr @a, align 4
 // CHECK2-NEXT:    [[TMP114:%.*]] = sext i32 [[TMP113]] to i64
 // CHECK2-NEXT:    [[LEN_SUB_1:%.*]] = sub nsw i64 [[TMP114]], 1
@@ -3167,8 +3167,8 @@ void test_omp_all_memory()
 // CHECK2-NEXT:    [[TMP116:%.*]] = sext i32 [[TMP115]] to i64
 // CHECK2-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP116]]
 // CHECK2-NEXT:    [[TMP117:%.*]] = mul nsw i64 [[LB_ADD_LEN]], [[TMP2]]
-// CHECK2-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP117]]
-// CHECK2-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX25]], i64 [[LEN_SUB_1]]
+// CHECK2-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP117]]
+// CHECK2-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX25]], i64 [[LEN_SUB_1]]
 // CHECK2-NEXT:    [[TMP118:%.*]] = getelementptr i32, ptr [[ARRAYIDX26]], i32 1
 // CHECK2-NEXT:    [[TMP119:%.*]] = ptrtoint ptr [[ARRAYIDX24]] to i64
 // CHECK2-NEXT:    [[TMP120:%.*]] = ptrtoint ptr [[TMP118]] to i64
@@ -4163,9 +4163,9 @@ void test_omp_all_memory()
 // CHECK2-51-NEXT:    [[TMP33:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP30]], i32 0, i32 2
 // CHECK2-51-NEXT:    store i8 1, ptr [[TMP33]], align 8
 // CHECK2-51-NEXT:    [[TMP34:%.*]] = mul nsw i64 0, [[TMP2]]
-// CHECK2-51-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP34]]
+// CHECK2-51-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP34]]
 // CHECK2-51-NEXT:    [[TMP35:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK2-51-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP35]]
+// CHECK2-51-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP35]]
 // CHECK2-51-NEXT:    [[TMP36:%.*]] = getelementptr i32, ptr [[ARRAYIDX2]], i32 1
 // CHECK2-51-NEXT:    [[TMP37:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK2-51-NEXT:    [[TMP38:%.*]] = ptrtoint ptr [[TMP36]] to i64
@@ -4200,13 +4200,13 @@ void test_omp_all_memory()
 // CHECK2-51-NEXT:    [[TMP58:%.*]] = load i8, ptr [[B]], align 1
 // CHECK2-51-NEXT:    [[TMP59:%.*]] = sext i8 [[TMP58]] to i64
 // CHECK2-51-NEXT:    [[TMP60:%.*]] = mul nsw i64 4, [[TMP2]]
-// CHECK2-51-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP60]]
-// CHECK2-51-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX7]], i64 [[TMP59]]
+// CHECK2-51-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP60]]
+// CHECK2-51-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX7]], i64 [[TMP59]]
 // CHECK2-51-NEXT:    [[TMP61:%.*]] = load i8, ptr [[B]], align 1
 // CHECK2-51-NEXT:    [[TMP62:%.*]] = sext i8 [[TMP61]] to i64
 // CHECK2-51-NEXT:    [[TMP63:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK2-51-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP63]]
-// CHECK2-51-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX9]], i64 [[TMP62]]
+// CHECK2-51-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP63]]
+// CHECK2-51-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX9]], i64 [[TMP62]]
 // CHECK2-51-NEXT:    [[TMP64:%.*]] = getelementptr i32, ptr [[ARRAYIDX10]], i32 1
 // CHECK2-51-NEXT:    [[TMP65:%.*]] = ptrtoint ptr [[ARRAYIDX8]] to i64
 // CHECK2-51-NEXT:    [[TMP66:%.*]] = ptrtoint ptr [[TMP64]] to i64
@@ -4238,13 +4238,13 @@ void test_omp_all_memory()
 // CHECK2-51-NEXT:    [[TMP83:%.*]] = load i8, ptr [[B]], align 1
 // CHECK2-51-NEXT:    [[TMP84:%.*]] = sext i8 [[TMP83]] to i64
 // CHECK2-51-NEXT:    [[TMP85:%.*]] = mul nsw i64 4, [[TMP2]]
-// CHECK2-51-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP85]]
-// CHECK2-51-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX15]], i64 [[TMP84]]
+// CHECK2-51-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP85]]
+// CHECK2-51-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX15]], i64 [[TMP84]]
 // CHECK2-51-NEXT:    [[TMP86:%.*]] = load i8, ptr [[B]], align 1
 // CHECK2-51-NEXT:    [[TMP87:%.*]] = sext i8 [[TMP86]] to i64
 // CHECK2-51-NEXT:    [[TMP88:%.*]] = mul nsw i64 9, [[TMP2]]
-// CHECK2-51-NEXT:    [[ARRAYIDX17:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP88]]
-// CHECK2-51-NEXT:    [[ARRAYIDX18:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX17]], i64 [[TMP87]]
+// CHECK2-51-NEXT:    [[ARRAYIDX17:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP88]]
+// CHECK2-51-NEXT:    [[ARRAYIDX18:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX17]], i64 [[TMP87]]
 // CHECK2-51-NEXT:    [[TMP89:%.*]] = getelementptr i32, ptr [[ARRAYIDX18]], i32 1
 // CHECK2-51-NEXT:    [[TMP90:%.*]] = ptrtoint ptr [[ARRAYIDX16]] to i64
 // CHECK2-51-NEXT:    [[TMP91:%.*]] = ptrtoint ptr [[TMP89]] to i64
@@ -4281,8 +4281,8 @@ void test_omp_all_memory()
 // CHECK2-51-NEXT:    [[TMP111:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP108]], i32 0, i32 2
 // CHECK2-51-NEXT:    store i8 3, ptr [[TMP111]], align 8
 // CHECK2-51-NEXT:    [[TMP112:%.*]] = mul nsw i64 0, [[TMP2]]
-// CHECK2-51-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP112]]
-// CHECK2-51-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX23]], i64 3
+// CHECK2-51-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP112]]
+// CHECK2-51-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX23]], i64 3
 // CHECK2-51-NEXT:    [[TMP113:%.*]] = load i32, ptr @a, align 4
 // CHECK2-51-NEXT:    [[TMP114:%.*]] = sext i32 [[TMP113]] to i64
 // CHECK2-51-NEXT:    [[LEN_SUB_1:%.*]] = sub nsw i64 [[TMP114]], 1
@@ -4290,8 +4290,8 @@ void test_omp_all_memory()
 // CHECK2-51-NEXT:    [[TMP116:%.*]] = sext i32 [[TMP115]] to i64
 // CHECK2-51-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP116]]
 // CHECK2-51-NEXT:    [[TMP117:%.*]] = mul nsw i64 [[LB_ADD_LEN]], [[TMP2]]
-// CHECK2-51-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP117]]
-// CHECK2-51-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX25]], i64 [[LEN_SUB_1]]
+// CHECK2-51-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP117]]
+// CHECK2-51-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX25]], i64 [[LEN_SUB_1]]
 // CHECK2-51-NEXT:    [[TMP118:%.*]] = getelementptr i32, ptr [[ARRAYIDX26]], i32 1
 // CHECK2-51-NEXT:    [[TMP119:%.*]] = ptrtoint ptr [[ARRAYIDX24]] to i64
 // CHECK2-51-NEXT:    [[TMP120:%.*]] = ptrtoint ptr [[TMP118]] to i64
@@ -4326,8 +4326,8 @@ void test_omp_all_memory()
 // CHECK2-51-NEXT:    [[TMP139:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP136]], i32 0, i32 2
 // CHECK2-51-NEXT:    store i8 8, ptr [[TMP139]], align 8
 // CHECK2-51-NEXT:    [[TMP140:%.*]] = mul nsw i64 0, [[TMP2]]
-// CHECK2-51-NEXT:    [[ARRAYIDX31:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP140]]
-// CHECK2-51-NEXT:    [[ARRAYIDX32:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX31]], i64 3
+// CHECK2-51-NEXT:    [[ARRAYIDX31:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP140]]
+// CHECK2-51-NEXT:    [[ARRAYIDX32:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX31]], i64 3
 // CHECK2-51-NEXT:    [[TMP141:%.*]] = load i32, ptr @a, align 4
 // CHECK2-51-NEXT:    [[TMP142:%.*]] = sext i32 [[TMP141]] to i64
 // CHECK2-51-NEXT:    [[LEN_SUB_133:%.*]] = sub nsw i64 [[TMP142]], 1
@@ -4335,8 +4335,8 @@ void test_omp_all_memory()
 // CHECK2-51-NEXT:    [[TMP144:%.*]] = sext i32 [[TMP143]] to i64
 // CHECK2-51-NEXT:    [[LB_ADD_LEN34:%.*]] = add nsw i64 -1, [[TMP144]]
 // CHECK2-51-NEXT:    [[TMP145:%.*]] = mul nsw i64 [[LB_ADD_LEN34]], [[TMP2]]
-// CHECK2-51-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP145]]
-// CHECK2-51-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_133]]
+// CHECK2-51-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP145]]
+// CHECK2-51-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_133]]
 // CHECK2-51-NEXT:    [[TMP146:%.*]] = getelementptr i32, ptr [[ARRAYIDX36]], i32 1
 // CHECK2-51-NEXT:    [[TMP147:%.*]] = ptrtoint ptr [[ARRAYIDX32]] to i64
 // CHECK2-51-NEXT:    [[TMP148:%.*]] = ptrtoint ptr [[TMP146]] to i64
@@ -5773,9 +5773,9 @@ void test_omp_all_memory()
 // CHECK3-NEXT:    [[TMP32:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP29]], i32 0, i32 2
 // CHECK3-NEXT:    store i8 1, ptr [[TMP32]], align 8
 // CHECK3-NEXT:    [[TMP33:%.*]] = mul nsw i64 0, [[TMP1]]
-// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP33]]
+// CHECK3-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP33]]
 // CHECK3-NEXT:    [[TMP34:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK3-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP34]]
+// CHECK3-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP34]]
 // CHECK3-NEXT:    [[TMP35:%.*]] = getelementptr i32, ptr [[ARRAYIDX4]], i32 1
 // CHECK3-NEXT:    [[TMP36:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK3-NEXT:    [[TMP37:%.*]] = ptrtoint ptr [[TMP35]] to i64
@@ -5814,13 +5814,13 @@ void test_omp_all_memory()
 // CHECK3-NEXT:    [[TMP57:%.*]] = load i8, ptr [[B]], align 1
 // CHECK3-NEXT:    [[TMP58:%.*]] = sext i8 [[TMP57]] to i64
 // CHECK3-NEXT:    [[TMP59:%.*]] = mul nsw i64 4, [[TMP1]]
-// CHECK3-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP59]]
-// CHECK3-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX13]], i64 [[TMP58]]
+// CHECK3-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP59]]
+// CHECK3-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX13]], i64 [[TMP58]]
 // CHECK3-NEXT:    [[TMP60:%.*]] = load i8, ptr [[B]], align 1
 // CHECK3-NEXT:    [[TMP61:%.*]] = sext i8 [[TMP60]] to i64
 // CHECK3-NEXT:    [[TMP62:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK3-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP62]]
-// CHECK3-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX15]], i64 [[TMP61]]
+// CHECK3-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP62]]
+// CHECK3-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX15]], i64 [[TMP61]]
 // CHECK3-NEXT:    [[TMP63:%.*]] = getelementptr i32, ptr [[ARRAYIDX16]], i32 1
 // CHECK3-NEXT:    [[TMP64:%.*]] = ptrtoint ptr [[ARRAYIDX14]] to i64
 // CHECK3-NEXT:    [[TMP65:%.*]] = ptrtoint ptr [[TMP63]] to i64
@@ -5854,13 +5854,13 @@ void test_omp_all_memory()
 // CHECK3-NEXT:    [[TMP82:%.*]] = load i8, ptr [[B]], align 1
 // CHECK3-NEXT:    [[TMP83:%.*]] = sext i8 [[TMP82]] to i64
 // CHECK3-NEXT:    [[TMP84:%.*]] = mul nsw i64 4, [[TMP1]]
-// CHECK3-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP84]]
-// CHECK3-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX23]], i64 [[TMP83]]
+// CHECK3-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP84]]
+// CHECK3-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX23]], i64 [[TMP83]]
 // CHECK3-NEXT:    [[TMP85:%.*]] = load i8, ptr [[B]], align 1
 // CHECK3-NEXT:    [[TMP86:%.*]] = sext i8 [[TMP85]] to i64
 // CHECK3-NEXT:    [[TMP87:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK3-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP87]]
-// CHECK3-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX25]], i64 [[TMP86]]
+// CHECK3-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP87]]
+// CHECK3-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX25]], i64 [[TMP86]]
 // CHECK3-NEXT:    [[TMP88:%.*]] = getelementptr i32, ptr [[ARRAYIDX26]], i32 1
 // CHECK3-NEXT:    [[TMP89:%.*]] = ptrtoint ptr [[ARRAYIDX24]] to i64
 // CHECK3-NEXT:    [[TMP90:%.*]] = ptrtoint ptr [[TMP88]] to i64
@@ -5899,8 +5899,8 @@ void test_omp_all_memory()
 // CHECK3-NEXT:    [[TMP110:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP107]], i32 0, i32 2
 // CHECK3-NEXT:    store i8 3, ptr [[TMP110]], align 8
 // CHECK3-NEXT:    [[TMP111:%.*]] = mul nsw i64 0, [[TMP1]]
-// CHECK3-NEXT:    [[ARRAYIDX33:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP111]]
-// CHECK3-NEXT:    [[ARRAYIDX34:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX33]], i64 3
+// CHECK3-NEXT:    [[ARRAYIDX33:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP111]]
+// CHECK3-NEXT:    [[ARRAYIDX34:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX33]], i64 3
 // CHECK3-NEXT:    [[TMP112:%.*]] = load i32, ptr @a, align 4
 // CHECK3-NEXT:    [[TMP113:%.*]] = sext i32 [[TMP112]] to i64
 // CHECK3-NEXT:    [[LEN_SUB_1:%.*]] = sub nsw i64 [[TMP113]], 1
@@ -5908,8 +5908,8 @@ void test_omp_all_memory()
 // CHECK3-NEXT:    [[TMP115:%.*]] = sext i32 [[TMP114]] to i64
 // CHECK3-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP115]]
 // CHECK3-NEXT:    [[TMP116:%.*]] = mul nsw i64 [[LB_ADD_LEN]], [[TMP1]]
-// CHECK3-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP116]]
-// CHECK3-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_1]]
+// CHECK3-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP116]]
+// CHECK3-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_1]]
 // CHECK3-NEXT:    [[TMP117:%.*]] = getelementptr i32, ptr [[ARRAYIDX36]], i32 1
 // CHECK3-NEXT:    [[TMP118:%.*]] = ptrtoint ptr [[ARRAYIDX34]] to i64
 // CHECK3-NEXT:    [[TMP119:%.*]] = ptrtoint ptr [[TMP117]] to i64
@@ -6789,9 +6789,9 @@ void test_omp_all_memory()
 // CHECK4-NEXT:    [[TMP32:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP29]], i32 0, i32 2
 // CHECK4-NEXT:    store i8 1, ptr [[TMP32]], align 8
 // CHECK4-NEXT:    [[TMP33:%.*]] = mul nsw i64 0, [[TMP1]]
-// CHECK4-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP33]]
+// CHECK4-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP33]]
 // CHECK4-NEXT:    [[TMP34:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK4-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP34]]
+// CHECK4-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP34]]
 // CHECK4-NEXT:    [[TMP35:%.*]] = getelementptr i32, ptr [[ARRAYIDX4]], i32 1
 // CHECK4-NEXT:    [[TMP36:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK4-NEXT:    [[TMP37:%.*]] = ptrtoint ptr [[TMP35]] to i64
@@ -6830,13 +6830,13 @@ void test_omp_all_memory()
 // CHECK4-NEXT:    [[TMP57:%.*]] = load i8, ptr [[B]], align 1
 // CHECK4-NEXT:    [[TMP58:%.*]] = sext i8 [[TMP57]] to i64
 // CHECK4-NEXT:    [[TMP59:%.*]] = mul nsw i64 4, [[TMP1]]
-// CHECK4-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP59]]
-// CHECK4-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX13]], i64 [[TMP58]]
+// CHECK4-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP59]]
+// CHECK4-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX13]], i64 [[TMP58]]
 // CHECK4-NEXT:    [[TMP60:%.*]] = load i8, ptr [[B]], align 1
 // CHECK4-NEXT:    [[TMP61:%.*]] = sext i8 [[TMP60]] to i64
 // CHECK4-NEXT:    [[TMP62:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK4-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP62]]
-// CHECK4-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX15]], i64 [[TMP61]]
+// CHECK4-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP62]]
+// CHECK4-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX15]], i64 [[TMP61]]
 // CHECK4-NEXT:    [[TMP63:%.*]] = getelementptr i32, ptr [[ARRAYIDX16]], i32 1
 // CHECK4-NEXT:    [[TMP64:%.*]] = ptrtoint ptr [[ARRAYIDX14]] to i64
 // CHECK4-NEXT:    [[TMP65:%.*]] = ptrtoint ptr [[TMP63]] to i64
@@ -6870,13 +6870,13 @@ void test_omp_all_memory()
 // CHECK4-NEXT:    [[TMP82:%.*]] = load i8, ptr [[B]], align 1
 // CHECK4-NEXT:    [[TMP83:%.*]] = sext i8 [[TMP82]] to i64
 // CHECK4-NEXT:    [[TMP84:%.*]] = mul nsw i64 4, [[TMP1]]
-// CHECK4-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP84]]
-// CHECK4-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX23]], i64 [[TMP83]]
+// CHECK4-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP84]]
+// CHECK4-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX23]], i64 [[TMP83]]
 // CHECK4-NEXT:    [[TMP85:%.*]] = load i8, ptr [[B]], align 1
 // CHECK4-NEXT:    [[TMP86:%.*]] = sext i8 [[TMP85]] to i64
 // CHECK4-NEXT:    [[TMP87:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK4-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP87]]
-// CHECK4-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX25]], i64 [[TMP86]]
+// CHECK4-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP87]]
+// CHECK4-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX25]], i64 [[TMP86]]
 // CHECK4-NEXT:    [[TMP88:%.*]] = getelementptr i32, ptr [[ARRAYIDX26]], i32 1
 // CHECK4-NEXT:    [[TMP89:%.*]] = ptrtoint ptr [[ARRAYIDX24]] to i64
 // CHECK4-NEXT:    [[TMP90:%.*]] = ptrtoint ptr [[TMP88]] to i64
@@ -6915,8 +6915,8 @@ void test_omp_all_memory()
 // CHECK4-NEXT:    [[TMP110:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP107]], i32 0, i32 2
 // CHECK4-NEXT:    store i8 3, ptr [[TMP110]], align 8
 // CHECK4-NEXT:    [[TMP111:%.*]] = mul nsw i64 0, [[TMP1]]
-// CHECK4-NEXT:    [[ARRAYIDX33:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP111]]
-// CHECK4-NEXT:    [[ARRAYIDX34:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX33]], i64 3
+// CHECK4-NEXT:    [[ARRAYIDX33:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP111]]
+// CHECK4-NEXT:    [[ARRAYIDX34:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX33]], i64 3
 // CHECK4-NEXT:    [[TMP112:%.*]] = load i32, ptr @a, align 4
 // CHECK4-NEXT:    [[TMP113:%.*]] = sext i32 [[TMP112]] to i64
 // CHECK4-NEXT:    [[LEN_SUB_1:%.*]] = sub nsw i64 [[TMP113]], 1
@@ -6924,8 +6924,8 @@ void test_omp_all_memory()
 // CHECK4-NEXT:    [[TMP115:%.*]] = sext i32 [[TMP114]] to i64
 // CHECK4-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP115]]
 // CHECK4-NEXT:    [[TMP116:%.*]] = mul nsw i64 [[LB_ADD_LEN]], [[TMP1]]
-// CHECK4-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP116]]
-// CHECK4-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_1]]
+// CHECK4-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP116]]
+// CHECK4-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_1]]
 // CHECK4-NEXT:    [[TMP117:%.*]] = getelementptr i32, ptr [[ARRAYIDX36]], i32 1
 // CHECK4-NEXT:    [[TMP118:%.*]] = ptrtoint ptr [[ARRAYIDX34]] to i64
 // CHECK4-NEXT:    [[TMP119:%.*]] = ptrtoint ptr [[TMP117]] to i64
@@ -7808,9 +7808,9 @@ void test_omp_all_memory()
 // CHECK3-51-NEXT:    [[TMP32:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP29]], i32 0, i32 2
 // CHECK3-51-NEXT:    store i8 1, ptr [[TMP32]], align 8
 // CHECK3-51-NEXT:    [[TMP33:%.*]] = mul nsw i64 0, [[TMP1]]
-// CHECK3-51-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP33]]
+// CHECK3-51-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP33]]
 // CHECK3-51-NEXT:    [[TMP34:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK3-51-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP34]]
+// CHECK3-51-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP34]]
 // CHECK3-51-NEXT:    [[TMP35:%.*]] = getelementptr i32, ptr [[ARRAYIDX4]], i32 1
 // CHECK3-51-NEXT:    [[TMP36:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK3-51-NEXT:    [[TMP37:%.*]] = ptrtoint ptr [[TMP35]] to i64
@@ -7849,13 +7849,13 @@ void test_omp_all_memory()
 // CHECK3-51-NEXT:    [[TMP57:%.*]] = load i8, ptr [[B]], align 1
 // CHECK3-51-NEXT:    [[TMP58:%.*]] = sext i8 [[TMP57]] to i64
 // CHECK3-51-NEXT:    [[TMP59:%.*]] = mul nsw i64 4, [[TMP1]]
-// CHECK3-51-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP59]]
-// CHECK3-51-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX13]], i64 [[TMP58]]
+// CHECK3-51-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP59]]
+// CHECK3-51-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX13]], i64 [[TMP58]]
 // CHECK3-51-NEXT:    [[TMP60:%.*]] = load i8, ptr [[B]], align 1
 // CHECK3-51-NEXT:    [[TMP61:%.*]] = sext i8 [[TMP60]] to i64
 // CHECK3-51-NEXT:    [[TMP62:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK3-51-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP62]]
-// CHECK3-51-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX15]], i64 [[TMP61]]
+// CHECK3-51-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP62]]
+// CHECK3-51-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX15]], i64 [[TMP61]]
 // CHECK3-51-NEXT:    [[TMP63:%.*]] = getelementptr i32, ptr [[ARRAYIDX16]], i32 1
 // CHECK3-51-NEXT:    [[TMP64:%.*]] = ptrtoint ptr [[ARRAYIDX14]] to i64
 // CHECK3-51-NEXT:    [[TMP65:%.*]] = ptrtoint ptr [[TMP63]] to i64
@@ -7889,13 +7889,13 @@ void test_omp_all_memory()
 // CHECK3-51-NEXT:    [[TMP82:%.*]] = load i8, ptr [[B]], align 1
 // CHECK3-51-NEXT:    [[TMP83:%.*]] = sext i8 [[TMP82]] to i64
 // CHECK3-51-NEXT:    [[TMP84:%.*]] = mul nsw i64 4, [[TMP1]]
-// CHECK3-51-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP84]]
-// CHECK3-51-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX23]], i64 [[TMP83]]
+// CHECK3-51-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP84]]
+// CHECK3-51-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX23]], i64 [[TMP83]]
 // CHECK3-51-NEXT:    [[TMP85:%.*]] = load i8, ptr [[B]], align 1
 // CHECK3-51-NEXT:    [[TMP86:%.*]] = sext i8 [[TMP85]] to i64
 // CHECK3-51-NEXT:    [[TMP87:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK3-51-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP87]]
-// CHECK3-51-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX25]], i64 [[TMP86]]
+// CHECK3-51-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP87]]
+// CHECK3-51-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX25]], i64 [[TMP86]]
 // CHECK3-51-NEXT:    [[TMP88:%.*]] = getelementptr i32, ptr [[ARRAYIDX26]], i32 1
 // CHECK3-51-NEXT:    [[TMP89:%.*]] = ptrtoint ptr [[ARRAYIDX24]] to i64
 // CHECK3-51-NEXT:    [[TMP90:%.*]] = ptrtoint ptr [[TMP88]] to i64
@@ -7934,8 +7934,8 @@ void test_omp_all_memory()
 // CHECK3-51-NEXT:    [[TMP110:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP107]], i32 0, i32 2
 // CHECK3-51-NEXT:    store i8 3, ptr [[TMP110]], align 8
 // CHECK3-51-NEXT:    [[TMP111:%.*]] = mul nsw i64 0, [[TMP1]]
-// CHECK3-51-NEXT:    [[ARRAYIDX33:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP111]]
-// CHECK3-51-NEXT:    [[ARRAYIDX34:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX33]], i64 3
+// CHECK3-51-NEXT:    [[ARRAYIDX33:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP111]]
+// CHECK3-51-NEXT:    [[ARRAYIDX34:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX33]], i64 3
 // CHECK3-51-NEXT:    [[TMP112:%.*]] = load i32, ptr @a, align 4
 // CHECK3-51-NEXT:    [[TMP113:%.*]] = sext i32 [[TMP112]] to i64
 // CHECK3-51-NEXT:    [[LEN_SUB_1:%.*]] = sub nsw i64 [[TMP113]], 1
@@ -7943,8 +7943,8 @@ void test_omp_all_memory()
 // CHECK3-51-NEXT:    [[TMP115:%.*]] = sext i32 [[TMP114]] to i64
 // CHECK3-51-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP115]]
 // CHECK3-51-NEXT:    [[TMP116:%.*]] = mul nsw i64 [[LB_ADD_LEN]], [[TMP1]]
-// CHECK3-51-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP116]]
-// CHECK3-51-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_1]]
+// CHECK3-51-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP116]]
+// CHECK3-51-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_1]]
 // CHECK3-51-NEXT:    [[TMP117:%.*]] = getelementptr i32, ptr [[ARRAYIDX36]], i32 1
 // CHECK3-51-NEXT:    [[TMP118:%.*]] = ptrtoint ptr [[ARRAYIDX34]] to i64
 // CHECK3-51-NEXT:    [[TMP119:%.*]] = ptrtoint ptr [[TMP117]] to i64
@@ -7981,8 +7981,8 @@ void test_omp_all_memory()
 // CHECK3-51-NEXT:    [[TMP138:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP135]], i32 0, i32 2
 // CHECK3-51-NEXT:    store i8 8, ptr [[TMP138]], align 8
 // CHECK3-51-NEXT:    [[TMP139:%.*]] = mul nsw i64 0, [[TMP1]]
-// CHECK3-51-NEXT:    [[ARRAYIDX43:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP139]]
-// CHECK3-51-NEXT:    [[ARRAYIDX44:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX43]], i64 3
+// CHECK3-51-NEXT:    [[ARRAYIDX43:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP139]]
+// CHECK3-51-NEXT:    [[ARRAYIDX44:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX43]], i64 3
 // CHECK3-51-NEXT:    [[TMP140:%.*]] = load i32, ptr @a, align 4
 // CHECK3-51-NEXT:    [[TMP141:%.*]] = sext i32 [[TMP140]] to i64
 // CHECK3-51-NEXT:    [[LEN_SUB_145:%.*]] = sub nsw i64 [[TMP141]], 1
@@ -7990,8 +7990,8 @@ void test_omp_all_memory()
 // CHECK3-51-NEXT:    [[TMP143:%.*]] = sext i32 [[TMP142]] to i64
 // CHECK3-51-NEXT:    [[LB_ADD_LEN46:%.*]] = add nsw i64 -1, [[TMP143]]
 // CHECK3-51-NEXT:    [[TMP144:%.*]] = mul nsw i64 [[LB_ADD_LEN46]], [[TMP1]]
-// CHECK3-51-NEXT:    [[ARRAYIDX47:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP144]]
-// CHECK3-51-NEXT:    [[ARRAYIDX48:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX47]], i64 [[LEN_SUB_145]]
+// CHECK3-51-NEXT:    [[ARRAYIDX47:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP144]]
+// CHECK3-51-NEXT:    [[ARRAYIDX48:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX47]], i64 [[LEN_SUB_145]]
 // CHECK3-51-NEXT:    [[TMP145:%.*]] = getelementptr i32, ptr [[ARRAYIDX48]], i32 1
 // CHECK3-51-NEXT:    [[TMP146:%.*]] = ptrtoint ptr [[ARRAYIDX44]] to i64
 // CHECK3-51-NEXT:    [[TMP147:%.*]] = ptrtoint ptr [[TMP145]] to i64
@@ -9323,9 +9323,9 @@ void test_omp_all_memory()
 // CHECK4-51-NEXT:    [[TMP32:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP29]], i32 0, i32 2
 // CHECK4-51-NEXT:    store i8 1, ptr [[TMP32]], align 8
 // CHECK4-51-NEXT:    [[TMP33:%.*]] = mul nsw i64 0, [[TMP1]]
-// CHECK4-51-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP33]]
+// CHECK4-51-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP33]]
 // CHECK4-51-NEXT:    [[TMP34:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK4-51-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP34]]
+// CHECK4-51-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP34]]
 // CHECK4-51-NEXT:    [[TMP35:%.*]] = getelementptr i32, ptr [[ARRAYIDX4]], i32 1
 // CHECK4-51-NEXT:    [[TMP36:%.*]] = ptrtoint ptr [[ARRAYIDX]] to i64
 // CHECK4-51-NEXT:    [[TMP37:%.*]] = ptrtoint ptr [[TMP35]] to i64
@@ -9364,13 +9364,13 @@ void test_omp_all_memory()
 // CHECK4-51-NEXT:    [[TMP57:%.*]] = load i8, ptr [[B]], align 1
 // CHECK4-51-NEXT:    [[TMP58:%.*]] = sext i8 [[TMP57]] to i64
 // CHECK4-51-NEXT:    [[TMP59:%.*]] = mul nsw i64 4, [[TMP1]]
-// CHECK4-51-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP59]]
-// CHECK4-51-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX13]], i64 [[TMP58]]
+// CHECK4-51-NEXT:    [[ARRAYIDX13:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP59]]
+// CHECK4-51-NEXT:    [[ARRAYIDX14:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX13]], i64 [[TMP58]]
 // CHECK4-51-NEXT:    [[TMP60:%.*]] = load i8, ptr [[B]], align 1
 // CHECK4-51-NEXT:    [[TMP61:%.*]] = sext i8 [[TMP60]] to i64
 // CHECK4-51-NEXT:    [[TMP62:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK4-51-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP62]]
-// CHECK4-51-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX15]], i64 [[TMP61]]
+// CHECK4-51-NEXT:    [[ARRAYIDX15:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP62]]
+// CHECK4-51-NEXT:    [[ARRAYIDX16:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX15]], i64 [[TMP61]]
 // CHECK4-51-NEXT:    [[TMP63:%.*]] = getelementptr i32, ptr [[ARRAYIDX16]], i32 1
 // CHECK4-51-NEXT:    [[TMP64:%.*]] = ptrtoint ptr [[ARRAYIDX14]] to i64
 // CHECK4-51-NEXT:    [[TMP65:%.*]] = ptrtoint ptr [[TMP63]] to i64
@@ -9404,13 +9404,13 @@ void test_omp_all_memory()
 // CHECK4-51-NEXT:    [[TMP82:%.*]] = load i8, ptr [[B]], align 1
 // CHECK4-51-NEXT:    [[TMP83:%.*]] = sext i8 [[TMP82]] to i64
 // CHECK4-51-NEXT:    [[TMP84:%.*]] = mul nsw i64 4, [[TMP1]]
-// CHECK4-51-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP84]]
-// CHECK4-51-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX23]], i64 [[TMP83]]
+// CHECK4-51-NEXT:    [[ARRAYIDX23:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP84]]
+// CHECK4-51-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX23]], i64 [[TMP83]]
 // CHECK4-51-NEXT:    [[TMP85:%.*]] = load i8, ptr [[B]], align 1
 // CHECK4-51-NEXT:    [[TMP86:%.*]] = sext i8 [[TMP85]] to i64
 // CHECK4-51-NEXT:    [[TMP87:%.*]] = mul nsw i64 9, [[TMP1]]
-// CHECK4-51-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP87]]
-// CHECK4-51-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX25]], i64 [[TMP86]]
+// CHECK4-51-NEXT:    [[ARRAYIDX25:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP87]]
+// CHECK4-51-NEXT:    [[ARRAYIDX26:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX25]], i64 [[TMP86]]
 // CHECK4-51-NEXT:    [[TMP88:%.*]] = getelementptr i32, ptr [[ARRAYIDX26]], i32 1
 // CHECK4-51-NEXT:    [[TMP89:%.*]] = ptrtoint ptr [[ARRAYIDX24]] to i64
 // CHECK4-51-NEXT:    [[TMP90:%.*]] = ptrtoint ptr [[TMP88]] to i64
@@ -9449,8 +9449,8 @@ void test_omp_all_memory()
 // CHECK4-51-NEXT:    [[TMP110:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_DEPEND_INFO]], ptr [[TMP107]], i32 0, i32 2
 // CHECK4-51-NEXT:    store i8 3, ptr [[TMP110]], align 8
 // CHECK4-51-NEXT:    [[TMP111:%.*]] = mul nsw i64 0, [[TMP1]]
-// CHECK4-51-NEXT:    [[ARRAYIDX33:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP111]]
-// CHECK4-51-NEXT:    [[ARRAYIDX34:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX33]], i64 3
+// CHECK4-51-NEXT:    [[ARRAYIDX33:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP111]]
+// CHECK4-51-NEXT:    [[ARRAYIDX34:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX33]], i64 3
 // CHECK4-51-NEXT:    [[TMP112:%.*]] = load i32, ptr @a, align 4
 // CHECK4-51-NEXT:    [[TMP113:%.*]] = sext i32 [[TMP112]] to i64
 // CHECK4-51-NEXT:    [[LEN_SUB_1:%.*]] = sub nsw i64 [[TMP113]], 1
@@ -9458,8 +9458,8 @@ void test_omp_all_memory()
 // CHECK4-51-NEXT:    [[TMP115:%.*]] = sext i32 [[TMP114]] to i64
 // CHECK4-51-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP115]]
 // CHECK4-51-NEXT:    [[TMP116:%.*]] = mul nsw i64 [[LB_ADD_LEN]], [[TMP1]]
-// CHECK4-51-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds nuw i32, ptr [[VLA]], i64 [[TMP116]]
-// CHECK4-51-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds nuw i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_1]]
+// CHECK4-51-NEXT:    [[ARRAYIDX35:%.*]] = getelementptr inbounds i32, ptr [[VLA]], i64 [[TMP116]]
+// CHECK4-51-NEXT:    [[ARRAYIDX36:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX35]], i64 [[LEN_SUB_1]]
 // CHECK4-51-NEXT:    [[TMP117:%.*]] = getelementptr i32, ptr [[ARRAYIDX36]], i32 1
 // CHECK4-51-NEXT:    [[TMP118:%.*]] = ptrtoint ptr [[ARRAYIDX34]] to i64
 // CHECK4-51-NEXT:    [[TMP119:%.*]] = ptrtoint ptr [[TMP117]] to i64
diff --git a/clang/test/OpenMP/task_in_reduction_codegen.cpp b/clang/test/OpenMP/task_in_reduction_codegen.cpp
index 29dc12978d7d90..aa2a4781379909 100644
--- a/clang/test/OpenMP/task_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/task_in_reduction_codegen.cpp
@@ -90,7 +90,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[VLA:%.*]] = alloca i16, i64 [[TMP2]], align 16
 // CHECK1-NEXT:    store i64 [[TMP2]], ptr [[__VLA_EXPR0]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[A]], ptr [[TMP4]], align 8
 // CHECK1-NEXT:    [[TMP5:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -105,7 +105,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP9]], align 8
 // CHECK1-NEXT:    [[TMP10:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP10]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP11:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[B]], ptr [[TMP11]], align 8
 // CHECK1-NEXT:    [[TMP12:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 1
@@ -120,7 +120,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..2, ptr [[TMP16]], align 8
 // CHECK1-NEXT:    [[TMP17:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP17]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
 // CHECK1-NEXT:    [[TMP18:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC_ADDR]], ptr [[TMP18]], align 8
 // CHECK1-NEXT:    [[TMP19:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 1
@@ -138,7 +138,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP25:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0]], i32 3, ptr [[DOTRD_INPUT_]])
 // CHECK1-NEXT:    store ptr [[TMP25]], ptr [[DOTTASK_RED_]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP26:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0:%.*]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[C]], ptr [[TMP26]], align 8
 // CHECK1-NEXT:    [[TMP27:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 1
@@ -153,7 +153,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..6, ptr [[TMP31]], align 8
 // CHECK1-NEXT:    [[TMP32:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP32]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP33:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP33]], align 8
 // CHECK1-NEXT:    [[TMP34:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 1
diff --git a/clang/test/OpenMP/taskgroup_task_reduction_codegen.cpp b/clang/test/OpenMP/taskgroup_task_reduction_codegen.cpp
index a8577c7a135793..faf86479dfdae1 100644
--- a/clang/test/OpenMP/taskgroup_task_reduction_codegen.cpp
+++ b/clang/test/OpenMP/taskgroup_task_reduction_codegen.cpp
@@ -67,7 +67,7 @@ int main(int argc, char **argv) {
 // CHECK-DAG:   [[A_REF]] = getelementptr inbounds nuw [[T1]], ptr [[GEPA:%[^,]+]], i32 0, i32 0
 // CHECK-DAG:   store ptr [[A]], ptr [[A_REF:[^,]+]],
 // CHECK-DAG:   [[A_REF]] = getelementptr inbounds nuw [[T1]], ptr [[GEPA]], i32 0, i32 1
-// CHECK-DAG:   [[GEPA]] = getelementptr inbounds nuw [3 x [[T1]]], ptr [[RD_IN1]], i64 0, i64
+// CHECK-DAG:   [[GEPA]] = getelementptr inbounds [3 x [[T1]]], ptr [[RD_IN1]], i64 0, i64
 // CHECK-DAG:   [[TMP6:%.+]] = getelementptr inbounds nuw [[T1]], ptr [[GEPA]], i32 0, i32 2
 // CHECK-DAG:   store i64 4, ptr [[TMP6]],
 // CHECK-DAG:   [[TMP7:%.+]] = getelementptr inbounds nuw [[T1]], ptr [[GEPA]], i32 0, i32 3
@@ -82,7 +82,7 @@ int main(int argc, char **argv) {
 // CHECK-DAG:   [[TMP12]] = getelementptr inbounds nuw [[T1]], ptr [[GEPB:%[^,]+]], i32 0, i32 0
 // CHECK-DAG:   store ptr [[B]], ptr [[TMP12:%[^,]+]],
 // CHECK-DAG:   [[TMP12]] = getelementptr inbounds nuw [[T1]], ptr [[GEPB]], i32 0, i32 1
-// CHECK-DAG:   [[GEPB]] = getelementptr inbounds nuw [3 x [[T1]]], ptr [[RD_IN1]], i64 0, i64
+// CHECK-DAG:   [[GEPB]] = getelementptr inbounds [3 x [[T1]]], ptr [[RD_IN1]], i64 0, i64
 // CHECK-DAG:   [[TMP14:%.+]] = getelementptr inbounds nuw [[T1]], ptr [[GEPB]], i32 0, i32 2
 // CHECK-DAG:   store i64 4, ptr [[TMP14]],
 // CHECK-DAG:   [[TMP15:%.+]] = getelementptr inbounds nuw [[T1]], ptr [[GEPB]], i32 0, i32 3
@@ -97,7 +97,7 @@ int main(int argc, char **argv) {
 // CHECK-DAG:   [[TMP20]] = getelementptr inbounds nuw [[T1]], ptr [[GEPARGC:%[^,]+]], i32 0, i32 0
 // CHECK-DAG:   store ptr [[ARGC_ADDR]], ptr [[TMP20:%[^,]+]],
 // CHECK-DAG:   [[TMP20]] = getelementptr inbounds nuw [[T1]], ptr [[GEPARGC]], i32 0, i32 1
-// CHECK-DAG:   [[GEPARGC]] = getelementptr inbounds nuw [3 x [[T1]]], ptr [[RD_IN1]], i64 0, i64
+// CHECK-DAG:   [[GEPARGC]] = getelementptr inbounds [3 x [[T1]]], ptr [[RD_IN1]], i64 0, i64
 // CHECK-DAG:   [[TMP22:%.+]] = getelementptr inbounds nuw [[T1]], ptr [[GEPARGC]], i32 0, i32 2
 // CHECK-DAG:   store i64 4, ptr [[TMP22]],
 // CHECK-DAG:   [[TMP23:%.+]] = getelementptr inbounds nuw [[T1]], ptr [[GEPARGC]], i32 0, i32 3
@@ -116,7 +116,7 @@ int main(int argc, char **argv) {
 // CHECK-DAG:   [[TMP30]] = getelementptr inbounds nuw [[T2]], ptr [[GEPC:%[^,]+]], i32 0, i32 0
 // CHECK-DAG:   store ptr [[C]], ptr [[TMP30:%[^,]+]],
 // CHECK-DAG:   [[TMP30]] = getelementptr inbounds nuw [[T2]], ptr [[GEPC]], i32 0, i32 1
-// CHECK-DAG:   [[GEPC]] = getelementptr inbounds nuw [2 x [[T2]]], ptr [[RD_IN2]], i64 0, i64
+// CHECK-DAG:   [[GEPC]] = getelementptr inbounds [2 x [[T2]]], ptr [[RD_IN2]], i64 0, i64
 // CHECK-DAG:   [[TMP32:%.+]] = getelementptr inbounds nuw [[T2]], ptr [[GEPC]], i32 0, i32 2
 // CHECK-DAG:   store i64 20, ptr [[TMP32]],
 // CHECK-DAG:   [[TMP33:%.+]] = getelementptr inbounds nuw [[T2]], ptr [[GEPC]], i32 0, i32 3
@@ -131,7 +131,7 @@ int main(int argc, char **argv) {
 // CHECK-DAG:   [[TMP38]] = getelementptr inbounds nuw [[T2]], ptr [[GEPVLA:%[^,]+]], i32 0, i32 0
 // CHECK-DAG:   store ptr [[VLA]], ptr [[TMP38:%[^,]+]],
 // CHECK-DAG:   [[TMP38]] = getelementptr inbounds nuw [[T2]], ptr [[GEPVLA]], i32 0, i32 1
-// CHECK-DAG:   [[GEPVLA]] = getelementptr inbounds nuw [2 x [[T2]]], ptr [[RD_IN2]], i64 0, i64
+// CHECK-DAG:   [[GEPVLA]] = getelementptr inbounds [2 x [[T2]]], ptr [[RD_IN2]], i64 0, i64
 // CHECK-DAG:   [[TMP40:%.+]] = mul nuw i64 [[VLA_SIZE]], 2
 // CHECK-DAG:   [[TMP41:%.+]] = udiv exact i64 [[TMP40]], ptrtoint (ptr getelementptr (i16, ptr null, i32 1) to i64)
 // CHECK-DAG:   [[TMP42:%.+]] = getelementptr inbounds nuw [[T2]], ptr [[GEPVLA]], i32 0, i32 2
diff --git a/clang/test/OpenMP/taskloop_in_reduction_codegen.cpp b/clang/test/OpenMP/taskloop_in_reduction_codegen.cpp
index 87b4cd2caf18ae..ae0d007756140f 100644
--- a/clang/test/OpenMP/taskloop_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/taskloop_in_reduction_codegen.cpp
@@ -76,7 +76,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[VLA:%.*]] = alloca i16, i64 [[TMP2]], align 16
 // CHECK1-NEXT:    store i64 [[TMP2]], ptr [[__VLA_EXPR0]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[A]], ptr [[TMP4]], align 8
 // CHECK1-NEXT:    [[TMP5:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -91,7 +91,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP9]], align 8
 // CHECK1-NEXT:    [[TMP10:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP10]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP11:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[B]], ptr [[TMP11]], align 8
 // CHECK1-NEXT:    [[TMP12:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 1
@@ -106,7 +106,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..2, ptr [[TMP16]], align 8
 // CHECK1-NEXT:    [[TMP17:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP17]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
 // CHECK1-NEXT:    [[TMP18:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC_ADDR]], ptr [[TMP18]], align 8
 // CHECK1-NEXT:    [[TMP19:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 1
@@ -124,7 +124,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP25:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0]], i32 3, ptr [[DOTRD_INPUT_]])
 // CHECK1-NEXT:    store ptr [[TMP25]], ptr [[DOTTASK_RED_]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP26:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0:%.*]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[C]], ptr [[TMP26]], align 8
 // CHECK1-NEXT:    [[TMP27:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 1
@@ -139,7 +139,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..6, ptr [[TMP31]], align 8
 // CHECK1-NEXT:    [[TMP32:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP32]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP33:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP33]], align 8
 // CHECK1-NEXT:    [[TMP34:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 1
diff --git a/clang/test/OpenMP/taskloop_reduction_codegen.cpp b/clang/test/OpenMP/taskloop_reduction_codegen.cpp
index 6eca033eca5515..3cdc88ba20b770 100644
--- a/clang/test/OpenMP/taskloop_reduction_codegen.cpp
+++ b/clang/test/OpenMP/taskloop_reduction_codegen.cpp
@@ -83,9 +83,9 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB1:.+]], ptr [[TMP25]],
 // CHECK-DAG:    [[TMP26:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK-DAG:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP26]], i8 0, i64 4, i1 false)
-// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C]], i64 0, i64 0
+// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C]], i64 0, i64 0
 // CHECK-DAG:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, %
-// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
+// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
 // CHECK-DAG:    [[TMP28]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_4:%.+]], i32 0, i32 0
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
@@ -137,10 +137,10 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB4:.+]], ptr [[TMP59]],
 // CHECK-DAG:    [[TMP60:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_8]], i32 0, i32 6
 // CHECK-DAG:    store i32 1, ptr [[TMP60]],
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
 // CHECK:    [[TMP62:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0]], i32 4, ptr [[DOTRD_INPUT_]])
 // CHECK:    [[TMP63:%.*]] = load i32, ptr [[N]],
 // CHECK:    store i32 [[TMP63]], ptr [[DOTCAPTURE_EXPR_]],
diff --git a/clang/test/OpenMP/taskloop_simd_in_reduction_codegen.cpp b/clang/test/OpenMP/taskloop_simd_in_reduction_codegen.cpp
index 9e4e51a442742b..6da28d2d973c98 100644
--- a/clang/test/OpenMP/taskloop_simd_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/taskloop_simd_in_reduction_codegen.cpp
@@ -76,7 +76,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[VLA:%.*]] = alloca i16, i64 [[TMP2]], align 16
 // CHECK1-NEXT:    store i64 [[TMP2]], ptr [[__VLA_EXPR0]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[A]], ptr [[TMP4]], align 8
 // CHECK1-NEXT:    [[TMP5:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -91,7 +91,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP9]], align 8
 // CHECK1-NEXT:    [[TMP10:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP10]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_1:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP11:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[B]], ptr [[TMP11]], align 8
 // CHECK1-NEXT:    [[TMP12:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 1
@@ -106,7 +106,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..2, ptr [[TMP16]], align 8
 // CHECK1-NEXT:    [[TMP17:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_1]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP17]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds nuw [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_2:%.*]] = getelementptr inbounds [3 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 2
 // CHECK1-NEXT:    [[TMP18:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC_ADDR]], ptr [[TMP18]], align 8
 // CHECK1-NEXT:    [[TMP19:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_2]], i32 0, i32 1
@@ -124,7 +124,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP25:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0]], i32 3, ptr [[DOTRD_INPUT_]])
 // CHECK1-NEXT:    store ptr [[TMP25]], ptr [[DOTTASK_RED_]], align 8
 // CHECK1-NEXT:    call void @__kmpc_taskgroup(ptr @[[GLOB1]], i32 [[TMP0]])
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_4:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP26:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0:%.*]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[C]], ptr [[TMP26]], align 8
 // CHECK1-NEXT:    [[TMP27:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 1
@@ -139,7 +139,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..6, ptr [[TMP31]], align 8
 // CHECK1-NEXT:    [[TMP32:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_4]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP32]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_5:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_3]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP33:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP33]], align 8
 // CHECK1-NEXT:    [[TMP34:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_5]], i32 0, i32 1
diff --git a/clang/test/OpenMP/taskloop_simd_reduction_codegen.cpp b/clang/test/OpenMP/taskloop_simd_reduction_codegen.cpp
index 83ae053cfd9bd2..d6e40831484aac 100644
--- a/clang/test/OpenMP/taskloop_simd_reduction_codegen.cpp
+++ b/clang/test/OpenMP/taskloop_simd_reduction_codegen.cpp
@@ -80,9 +80,9 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB1:.+]], ptr [[TMP25]],
 // CHECK-DAG:    [[TMP26:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK-DAG:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP26]], i8 0, i64 4, i1 false)
-// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C]], i64 0, i64 0
+// CHECK-DAG:    [[ARRAYIDX5:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C]], i64 0, i64 0
 // CHECK-DAG:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, %
-// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds nuw [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
+// CHECK-DAG:    [[ARRAYIDX6:%.*]] = getelementptr inbounds [100 x %struct.S], ptr [[C]], i64 0, i64 [[LB_ADD_LEN]]
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
 // CHECK-DAG:    [[TMP28]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_4:%.+]], i32 0, i32 0
 // CHECK-DAG:    store ptr [[ARRAYIDX5]], ptr [[TMP28:%[^,]+]],
@@ -134,10 +134,10 @@ sum = 0.0;
 // CHECK-DAG:    store ptr @[[RED_COMB4:.+]], ptr [[TMP59]],
 // CHECK-DAG:    [[TMP60:%.*]] = getelementptr inbounds nuw %struct.kmp_taskred_input_t, ptr [[DOTRD_INPUT_GEP_8]], i32 0, i32 6
 // CHECK-DAG:    store i32 1, ptr [[TMP60]],
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
-// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds nuw [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_4]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_7]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
+// CHECK-DAG:    [[DOTRD_INPUT_GEP_8]] = getelementptr inbounds [4 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64
 // CHECK:    [[TMP62:%.*]] = call ptr @__kmpc_taskred_init(i32 [[TMP0]], i32 4, ptr [[DOTRD_INPUT_]])
 // CHECK:    [[TMP63:%.*]] = load i32, ptr [[N]],
 // CHECK:    store i32 [[TMP63]], ptr [[DOTCAPTURE_EXPR_]],
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
index 7987c2de7dd8fa..be499e0b36548d 100644
--- a/clang/test/OpenMP/teams_distribute_parallel_for_reduction_task_codegen.cpp
+++ b/clang/test/OpenMP/teams_distribute_parallel_for_reduction_task_codegen.cpp
@@ -100,16 +100,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP1]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP1]], i64 0
 // CHECK1-NEXT:    [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP2]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP2]], i64 0
 // CHECK1-NEXT:    [[TMP3:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP4:%.*]] = sext i32 [[TMP3]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP4]]
 // CHECK1-NEXT:    [[TMP5:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP5]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP5]], i64 9
 // CHECK1-NEXT:    [[TMP6:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP6]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP8:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP9:%.*]] = sub i64 [[TMP7]], [[TMP8]]
@@ -139,7 +139,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP21:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP20]]
 // CHECK1-NEXT:    store ptr [[_TMP5]], ptr [[TMP]], align 8
 // CHECK1-NEXT:    store ptr [[TMP21]], ptr [[_TMP5]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP22:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP22]], align 8
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -154,19 +154,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb., ptr [[TMP27]], align 8
 // CHECK1-NEXT:    [[TMP28:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP28]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_6:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP29:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP30:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP30]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX7:%.*]] = getelementptr inbounds ptr, ptr [[TMP30]], i64 0
 // CHECK1-NEXT:    [[TMP31:%.*]] = load ptr, ptr [[ARRAYIDX7]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP31]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds i8, ptr [[TMP31]], i64 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP33:%.*]] = sext i32 [[TMP32]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN9:%.*]] = add nsw i64 -1, [[TMP33]]
 // CHECK1-NEXT:    [[TMP34:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP34]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds ptr, ptr [[TMP34]], i64 9
 // CHECK1-NEXT:    [[TMP35:%.*]] = load ptr, ptr [[ARRAYIDX10]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds i8, ptr [[TMP35]], i64 [[LB_ADD_LEN9]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP36:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T]], ptr [[DOTRD_INPUT_GEP_6]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX8]], ptr [[TMP36]], align 8
@@ -444,16 +444,16 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_IS_LAST]], align 4
 // CHECK1-NEXT:    store i32 0, ptr [[ARGC1]], align 4
 // CHECK1-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP3]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP3]], i64 0
 // CHECK1-NEXT:    [[TMP4:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP4]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i8, ptr [[TMP4]], i64 0
 // CHECK1-NEXT:    [[TMP5:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP6:%.*]] = sext i32 [[TMP5]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN:%.*]] = add nsw i64 -1, [[TMP6]]
 // CHECK1-NEXT:    [[TMP7:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP7]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP7]], i64 9
 // CHECK1-NEXT:    [[TMP8:%.*]] = load ptr, ptr [[ARRAYIDX3]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP8]], i64 [[LB_ADD_LEN]]
+// CHECK1-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds i8, ptr [[TMP8]], i64 [[LB_ADD_LEN]]
 // CHECK1-NEXT:    [[TMP9:%.*]] = ptrtoint ptr [[ARRAYIDX4]] to i64
 // CHECK1-NEXT:    [[TMP10:%.*]] = ptrtoint ptr [[ARRAYIDX2]] to i64
 // CHECK1-NEXT:    [[TMP11:%.*]] = sub i64 [[TMP9]], [[TMP10]]
@@ -483,7 +483,7 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[TMP23:%.*]] = getelementptr i8, ptr [[VLA]], i64 [[TMP22]]
 // CHECK1-NEXT:    store ptr [[_TMP6]], ptr [[_TMP5]], align 8
 // CHECK1-NEXT:    store ptr [[TMP23]], ptr [[_TMP6]], align 8
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_]], i64 0, i64 0
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_]], i64 0, i64 0
 // CHECK1-NEXT:    [[TMP24:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0:%.*]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 0
 // CHECK1-NEXT:    store ptr [[ARGC1]], ptr [[TMP24]], align 8
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 1
@@ -498,19 +498,19 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    store ptr @.red_comb..4, ptr [[TMP29]], align 8
 // CHECK1-NEXT:    [[TMP30:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_]], i32 0, i32 6
 // CHECK1-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[TMP30]], i8 0, i64 4, i1 false)
-// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds nuw [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_]], i64 0, i64 1
+// CHECK1-NEXT:    [[DOTRD_INPUT_GEP_7:%.*]] = getelementptr inbounds [2 x %struct.kmp_taskred_input_t.0], ptr [[DOTRD_INPUT_]], i64 0, i64 1
 // CHECK1-NEXT:    [[TMP31:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 0
 // CHECK1-NEXT:    [[TMP32:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP32]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds ptr, ptr [[TMP32]], i64 0
 // CHECK1-NEXT:    [[TMP33:%.*]] = load ptr, ptr [[ARRAYIDX8]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP33]], i64 0
+// CHECK1-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds i8, ptr [[TMP33]], i64 0
 // CHECK1-NEXT:    [[TMP34:%.*]] = load i32, ptr [[TMP0]], align 4
 // CHECK1-NEXT:    [[TMP35:%.*]] = sext i32 [[TMP34]] to i64
 // CHECK1-NEXT:    [[LB_ADD_LEN10:%.*]] = add nsw i64 -1, [[TMP35]]
 // CHECK1-NEXT:    [[TMP36:%.*]] = load ptr, ptr [[ARGV_ADDR]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP36]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX11:%.*]] = getelementptr inbounds ptr, ptr [[TMP36]], i64 9
 // CHECK1-NEXT:    [[TMP37:%.*]] = load ptr, ptr [[ARRAYIDX11]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP37]], i64 [[LB_ADD_LEN10]]
+// CHECK1-NEXT:    [[ARRAYIDX12:%.*]] = getelementptr inbounds i8, ptr [[TMP37]], i64 [[LB_ADD_LEN10]]
 // CHECK1-NEXT:    store ptr [[VLA]], ptr [[TMP31]], align 8
 // CHECK1-NEXT:    [[TMP38:%.*]] = getelementptr inbounds nuw [[STRUCT_KMP_TASKRED_INPUT_T_0]], ptr [[DOTRD_INPUT_GEP_7]], i32 0, i32 1
 // CHECK1-NEXT:    store ptr [[ARRAYIDX9]], ptr [[TMP38]], align 8
@@ -831,9 +831,9 @@ int main(int argc, char **argv) {
 // CHECK1-NEXT:    [[LB_ADD_LEN_I:%.*]] = add nsw i64 -1, [[TMP24]]
 // CHECK1-NEXT:    [[TMP25:%.*]] = getelementptr inbounds nuw [[STRUCT_ANON]], ptr [[TMP9]], i32 0, i32 2
 // CHECK1-NEXT:    [[TMP26:%.*]] = load ptr, ptr [[TMP25]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds nuw ptr, ptr [[TMP26]], i64 9
+// CHECK1-NEXT:    [[ARRAYIDX2_I:%.*]] = getelementptr inbounds ptr, ptr [[TMP26]], i64 9
 // CHECK1-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[ARRAYIDX2_I]], align 8
-// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
+// CHECK1-NEXT:    [[ARRAYIDX3_I:%.*]] = getelementptr inbounds i8, ptr [[TMP27]], i64 [[LB_ADD_LEN_I]]
 // CHECK1-NEXT:    [[TMP28:%.*]] = ptrtoint ptr [[ARRAYIDX3_I]] to i64
 // CHECK1-NEXT:    [[TMP29:%.*]] = ptrtoint ptr [[TMP20]] to i64
 // CHECK1-NEXT:    [[TMP30:%.*]] = sub i64 [[TMP28]], [[TMP29]]



More information about the cfe-commits mailing list