[clang] cf624b2 - [SROA] `isVectorPromotionViable()`: memory intrinsics operate on vectors of bytes

[Roman Lebedev via cfe-commits]

Author: Roman Lebedev
Date: 2022-11-23T02:38:25+03:00
New Revision: cf624b23bc5d5a6161706d1663def49380ff816a

URL: https://github.com/llvm/llvm-project/commit/cf624b23bc5d5a6161706d1663def49380ff816a
DIFF: https://github.com/llvm/llvm-project/commit/cf624b23bc5d5a6161706d1663def49380ff816a.diff

LOG: [SROA] `isVectorPromotionViable()`: memory intrinsics operate on vectors of bytes

Now, there's a big caveat here - these bytes
are abstract bytes, not the i8 we have in LLVM,
so strictly speaking this is not exactly legal,
see e.g. https://github.com/AliveToolkit/alive2/issues/860
^ the "bytes" "could" have been a pointer,
and loading it as an integer inserts an implicit ptrtoint.

But at the same time,
InstCombine's `InstCombinerImpl::SimplifyAnyMemTransfer()`
would expand a memtransfer of 1/2/4/8 bytes
into integer-typed load+store,
so this isn't exactly a new problem.

Note that in memory, poison is byte-wise,
so we really can't widen elements,
but SROA seems to be inconsistent here.

Fixes #59116.

Added: 
    

Modified: 
    clang/test/CodeGenOpenCL/amdgpu-nullptr.cl
    llvm/lib/Transforms/Scalar/SROA.cpp
    llvm/test/CodeGen/AMDGPU/v1024.ll
    llvm/test/DebugInfo/X86/sroasplit-1.ll
    llvm/test/DebugInfo/X86/sroasplit-4.ll
    llvm/test/Transforms/PhaseOrdering/instcombine-sroa-inttoptr.ll
    llvm/test/Transforms/SROA/address-spaces.ll
    llvm/test/Transforms/SROA/alignment.ll
    llvm/test/Transforms/SROA/alloca-address-space.ll
    llvm/test/Transforms/SROA/basictest.ll
    llvm/test/Transforms/SROA/pointer-offset-size.ll
    llvm/test/Transforms/SROA/scalable-vectors.ll
    llvm/test/Transforms/SROA/slice-width.ll
    llvm/test/Transforms/SROA/sroa-common-type-fail-promotion.ll
    llvm/test/Transforms/SROA/tbaa-struct.ll
    llvm/test/Transforms/SROA/tbaa-struct2.ll
    llvm/test/Transforms/SROA/vector-promotion.ll

Removed: 
    


################################################################################
diff  --git a/clang/test/CodeGenOpenCL/amdgpu-nullptr.cl b/clang/test/CodeGenOpenCL/amdgpu-nullptr.cl
index 65f6f2e7d8c24..859e81f08d6bd 100644
--- a/clang/test/CodeGenOpenCL/amdgpu-nullptr.cl
+++ b/clang/test/CodeGenOpenCL/amdgpu-nullptr.cl
@@ -515,13 +515,17 @@ typedef struct {
   private char *p;
 } StructTy3;
 
-// CHECK-LABEL: test_memset_private
-// CHECK: call void @llvm.memset.p5i8.i64(i8 addrspace(5)* noundef align 8 {{.*}}, i8 0, i64 32, i1 false)
-// CHECK: [[GEP:%.*]] = getelementptr inbounds %struct.StructTy3, %struct.StructTy3 addrspace(5)* %ptr, i32 0, i32 4
-// CHECK: store i8 addrspace(5)* addrspacecast (i8* null to i8 addrspace(5)*), i8 addrspace(5)* addrspace(5)* [[GEP]]
-// CHECK: [[GEP1:%.*]] = getelementptr inbounds i8, i8 addrspace(5)* {{.*}}, i32 36
-// CHECK: [[GEP1_CAST:%.*]] = bitcast i8 addrspace(5)* [[GEP1]] to i32 addrspace(5)*
-// CHECK: store i32 0, i32 addrspace(5)* [[GEP1_CAST]], align 4
+// CHECK-LABEL: @test_memset_private(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = bitcast [[STRUCT_STRUCTTY3:%.*]] addrspace(5)* [[PTR:%.*]] to i8 addrspace(5)*
+// CHECK-NEXT:    [[S3_SROA_0_SROA_0_0_S3_SROA_0_0__SROA_CAST2_SROA_CAST:%.*]] = bitcast [[STRUCT_STRUCTTY3]] addrspace(5)* [[PTR]] to <32 x i8> addrspace(5)*
+// CHECK-NEXT:    store <32 x i8> zeroinitializer, <32 x i8> addrspace(5)* [[S3_SROA_0_SROA_0_0_S3_SROA_0_0__SROA_CAST2_SROA_CAST]], align 8, !tbaa.struct !9
+// CHECK-NEXT:    [[S3_SROA_4_0__SROA_IDX6:%.*]] = getelementptr inbounds [[STRUCT_STRUCTTY3]], [[STRUCT_STRUCTTY3]] addrspace(5)* [[PTR]], i32 0, i32 4
+// CHECK-NEXT:    store i8 addrspace(5)* addrspacecast (i8* null to i8 addrspace(5)*), i8 addrspace(5)* addrspace(5)* [[S3_SROA_4_0__SROA_IDX6]], align 8, !tbaa.struct !12
+// CHECK-NEXT:    [[S3_SROA_5_0__SROA_IDX:%.*]] = getelementptr inbounds i8, i8 addrspace(5)* [[TMP0]], i32 36
+// CHECK-NEXT:    [[S3_SROA_5_0__SROA_CAST8:%.*]] = bitcast i8 addrspace(5)* [[S3_SROA_5_0__SROA_IDX]] to i32 addrspace(5)*
+// CHECK-NEXT:    store i32 0, i32 addrspace(5)* [[S3_SROA_5_0__SROA_CAST8]], align 4, !tbaa.struct !13
+// CHECK-NEXT:    ret void
 void test_memset_private(private StructTy3 *ptr) {
   StructTy3 S3 = {0, 0, 0, 0, 0};
   *ptr = S3;

diff  --git a/llvm/lib/Transforms/Scalar/SROA.cpp b/llvm/lib/Transforms/Scalar/SROA.cpp
index 0d045656661a1..0fafa3af76367 100644
--- a/llvm/lib/Transforms/Scalar/SROA.cpp
+++ b/llvm/lib/Transforms/Scalar/SROA.cpp
@@ -1806,8 +1806,10 @@ static bool isVectorPromotionViableForSlice(Partition &P, const Slice &S,
                       ? Ty->getElementType()
                       : FixedVectorType::get(Ty->getElementType(), NumElements);
 
-  Type *SplitIntTy =
-      Type::getIntNTy(Ty->getContext(), NumElements * ElementSize * 8);
+  Type *SplitIntTy = nullptr;
+  if (uint64_t Bitwidth = NumElements * ElementSize * 8;
+      Bitwidth <= IntegerType::MAX_INT_BITS)
+    SplitIntTy = Type::getIntNTy(Ty->getContext(), Bitwidth);
 
   Use *U = S.getUse();
 
@@ -1826,7 +1828,8 @@ static bool isVectorPromotionViableForSlice(Partition &P, const Slice &S,
     // Disable vector promotion when there are loads or stores of an FCA.
     if (LTy->isStructTy())
       return false;
-    if (P.beginOffset() > S.beginOffset() || P.endOffset() < S.endOffset()) {
+    if (SplitIntTy &&
+        (P.beginOffset() > S.beginOffset() || P.endOffset() < S.endOffset())) {
       assert(LTy->isIntegerTy());
       LTy = SplitIntTy;
     }
@@ -1839,7 +1842,8 @@ static bool isVectorPromotionViableForSlice(Partition &P, const Slice &S,
     // Disable vector promotion when there are loads or stores of an FCA.
     if (STy->isStructTy())
       return false;
-    if (P.beginOffset() > S.beginOffset() || P.endOffset() < S.endOffset()) {
+    if (SplitIntTy &&
+        (P.beginOffset() > S.beginOffset() || P.endOffset() < S.endOffset())) {
       assert(STy->isIntegerTy());
       STy = SplitIntTy;
     }
@@ -1934,6 +1938,9 @@ static VectorType *isVectorPromotionViable(Partition &P, const DataLayout &DL) {
         CheckCandidateType(LI->getType());
       else if (auto *SI = dyn_cast<StoreInst>(S.getUse()->getUser()))
         CheckCandidateType(SI->getValueOperand()->getType());
+      else if (auto *MTI = dyn_cast<MemIntrinsic>(S.getUse()->getUser()))
+        CheckCandidateType(FixedVectorType::get(
+            IntegerType::getInt8Ty(MTI->getContext()), P.size()));
     }
 
   // If we didn't find a vector type, nothing to do here.

diff  --git a/llvm/test/CodeGen/AMDGPU/v1024.ll b/llvm/test/CodeGen/AMDGPU/v1024.ll
index 1326ba437f94f..6dbb9443fd5cf 100644
--- a/llvm/test/CodeGen/AMDGPU/v1024.ll
+++ b/llvm/test/CodeGen/AMDGPU/v1024.ll
@@ -4,7 +4,7 @@
 
 ; GCN-LABEL: {{^}}test_v1024:
 ; GCN-NOT: v_accvgpr
-; GCN-COUNT-32: v_mov_b32_e32
+; GCN-COUNT-10: v_mov_b32_e32
 ; GCN-NOT: v_accvgpr
 define amdgpu_kernel void @test_v1024() {
 entry:

diff  --git a/llvm/test/DebugInfo/X86/sroasplit-1.ll b/llvm/test/DebugInfo/X86/sroasplit-1.ll
index 0ec368130da28..5a80b56950122 100644
--- a/llvm/test/DebugInfo/X86/sroasplit-1.ll
+++ b/llvm/test/DebugInfo/X86/sroasplit-1.ll
@@ -20,10 +20,8 @@
 ;
 
 ; Verify that SROA creates a variable piece when splitting i1.
-; CHECK: %[[I1:.*]] = alloca [12 x i8], align 4
-; CHECK: call void @llvm.dbg.declare(metadata [12 x i8]* %[[I1]], metadata ![[VAR:[0-9]+]], metadata !DIExpression(DW_OP_LLVM_fragment, 32, 96))
-; CHECK: call void @llvm.dbg.value(metadata i32 %[[A:.*]], metadata ![[VAR]], metadata !DIExpression(DW_OP_LLVM_fragment, 0, 32))
-; CHECK: ret i32 %[[A]]
+; CHECK: %[[I1:.*]] = load <12 x i8>,
+; CHECK: call void @llvm.dbg.value(metadata <12 x i8> %[[I1]], metadata ![[VAR:.*]], metadata !DIExpression(DW_OP_LLVM_fragment, 32, 96))
 ; Read Var and Piece:
 ; CHECK: ![[VAR]] = !DILocalVariable(name: "i1",{{.*}} line: 11,
 

diff  --git a/llvm/test/DebugInfo/X86/sroasplit-4.ll b/llvm/test/DebugInfo/X86/sroasplit-4.ll
index 0d5594ef867dd..a3b35b820a96c 100644
--- a/llvm/test/DebugInfo/X86/sroasplit-4.ll
+++ b/llvm/test/DebugInfo/X86/sroasplit-4.ll
@@ -1,28 +1,28 @@
 ; RUN: opt -sroa < %s -S -o - | FileCheck %s
 ;
 ; Test that recursively splitting an alloca updates the debug info correctly.
-; CHECK: %[[T:.*]] = load i64, i64* @t, align 8
-; CHECK: call void @llvm.dbg.value(metadata i64 %[[T]], metadata ![[Y:.*]], metadata !DIExpression(DW_OP_LLVM_fragment, 0, 64))
-; CHECK: %[[T1:.*]] = load i64, i64* @t, align 8
-; CHECK: call void @llvm.dbg.value(metadata i64 %[[T1]], metadata ![[Y]], metadata !DIExpression(DW_OP_LLVM_fragment, 64, 64))
-; CHECK: call void @llvm.dbg.value(metadata i64 %[[T]], metadata ![[R:.*]], metadata !DIExpression(DW_OP_LLVM_fragment, 192, 64))
-; CHECK: call void @llvm.dbg.value(metadata i64 %[[T1]], metadata ![[R]], metadata !DIExpression(DW_OP_LLVM_fragment, 256, 64))
-; 
+; CHECK: call void @llvm.dbg.value(metadata <16 x i8> %[[Y_VEC:.*]], metadata ![[Y:.*]], metadata !DIExpression())
+; CHECK: call void @llvm.dbg.value(metadata <16 x i8> %[[Y_VEC1:.*]], metadata ![[Y]], metadata !DIExpression())
+; CHECK: call void @llvm.dbg.value(metadata i32 0, metadata ![[R:.*]], metadata !DIExpression(DW_OP_LLVM_fragment, 0, 32))
+; CHECK: call void @llvm.dbg.value(metadata i64 0, metadata ![[R]], metadata !DIExpression(DW_OP_LLVM_fragment, 64, 64))
+; CHECK: call void @llvm.dbg.value(metadata i64 0, metadata ![[R]], metadata !DIExpression(DW_OP_LLVM_fragment, 128, 64))
+; CHECK: call void @llvm.dbg.value(metadata <16 x i8> %[[Y_VEC1]], metadata ![[R]], metadata !DIExpression(DW_OP_LLVM_fragment, 192, 128))
+;
 ; struct p {
 ;   __SIZE_TYPE__ s;
 ;   __SIZE_TYPE__ t;
 ; };
-;  
+;
 ; struct r {
 ;   int i;
 ;   struct p x;
 ;   struct p y;
 ; };
-;  
+;
 ; extern int call_me(struct r);
 ; extern int maybe();
 ; extern __SIZE_TYPE__ t;
-;  
+;
 ; int test() {
 ;   if (maybe())
 ;     return 0;

diff  --git a/llvm/test/Transforms/PhaseOrdering/instcombine-sroa-inttoptr.ll b/llvm/test/Transforms/PhaseOrdering/instcombine-sroa-inttoptr.ll
index 886bde2686bd9..3361ad11c244b 100644
--- a/llvm/test/Transforms/PhaseOrdering/instcombine-sroa-inttoptr.ll
+++ b/llvm/test/Transforms/PhaseOrdering/instcombine-sroa-inttoptr.ll
@@ -68,12 +68,13 @@ define dso_local i32* @_Z3foo1S(%0* byval(%0) align 8 %arg) {
 ; CHECK-LABEL: @_Z3foo1S(
 ; CHECK-NEXT:  bb:
 ; CHECK-NEXT:    [[I2:%.*]] = alloca [[TMP0:%.*]], align 8
-; CHECK-NEXT:    [[I1_SROA_0_0_I5_SROA_IDX:%.*]] = getelementptr inbounds [[TMP0]], %0* [[ARG:%.*]], i64 0, i32 0
-; CHECK-NEXT:    [[I1_SROA_0_0_COPYLOAD:%.*]] = load i32*, i32** [[I1_SROA_0_0_I5_SROA_IDX]], align 8
+; CHECK-NEXT:    [[TMP0]] = bitcast %0* [[ARG:%.*]] to i64*
+; CHECK-NEXT:    [[I11_SROA_0_0_VEC_EXTRACT_EXTRACT:%.*]] = load i64, i64* [[TMP0]], align 8
+; CHECK-NEXT:    [[TMP1:%.*]] = inttoptr i64 [[I11_SROA_0_0_VEC_EXTRACT_EXTRACT]] to i32*
 ; CHECK-NEXT:    [[I_SROA_0_0_I6_SROA_IDX:%.*]] = getelementptr inbounds [[TMP0]], %0* [[I2]], i64 0, i32 0
-; CHECK-NEXT:    store i32* [[I1_SROA_0_0_COPYLOAD]], i32** [[I_SROA_0_0_I6_SROA_IDX]], align 8
+; CHECK-NEXT:    store i32* [[TMP1]], i32** [[I_SROA_0_0_I6_SROA_IDX]], align 8
 ; CHECK-NEXT:    tail call void @_Z7escape01S(%0* nonnull byval([[TMP0]]) align 8 [[I2]])
-; CHECK-NEXT:    ret i32* [[I1_SROA_0_0_COPYLOAD]]
+; CHECK-NEXT:    ret i32* [[TMP1]]
 ;
 bb:
   %i = alloca %0, align 8
@@ -107,21 +108,22 @@ declare void @llvm.lifetime.end.p0i8(i64 immarg, i8* nocapture)
 define dso_local i32* @_Z3bar1S(%0* byval(%0) align 8 %arg) {
 ; CHECK-LABEL: @_Z3bar1S(
 ; CHECK-NEXT:  bb:
-; CHECK-NEXT:    [[I1_SROA_0_0_I4_SROA_IDX:%.*]] = getelementptr inbounds [[TMP0:%.*]], %0* [[ARG:%.*]], i64 0, i32 0
-; CHECK-NEXT:    [[I1_SROA_0_0_COPYLOAD:%.*]] = load i32*, i32** [[I1_SROA_0_0_I4_SROA_IDX]], align 8
+; CHECK-NEXT:    [[TMP0:%.*]] = bitcast %0* [[ARG:%.*]] to i64*
+; CHECK-NEXT:    [[I13_SROA_0_0_VEC_EXTRACT_EXTRACT:%.*]] = load i64, i64* [[TMP0]], align 8
+; CHECK-NEXT:    [[TMP1:%.*]] = inttoptr i64 [[I13_SROA_0_0_VEC_EXTRACT_EXTRACT]] to i32*
 ; CHECK-NEXT:    [[I5:%.*]] = tail call i32 @_Z4condv()
 ; CHECK-NEXT:    [[I6_NOT:%.*]] = icmp eq i32 [[I5]], 0
 ; CHECK-NEXT:    br i1 [[I6_NOT]], label [[BB10:%.*]], label [[BB7:%.*]]
 ; CHECK:       bb7:
 ; CHECK-NEXT:    tail call void @_Z5sync0v()
-; CHECK-NEXT:    tail call void @_Z7escape0Pi(i32* [[I1_SROA_0_0_COPYLOAD]])
+; CHECK-NEXT:    tail call void @_Z7escape0Pi(i32* [[TMP1]])
 ; CHECK-NEXT:    br label [[BB13:%.*]]
 ; CHECK:       bb10:
 ; CHECK-NEXT:    tail call void @_Z5sync1v()
-; CHECK-NEXT:    tail call void @_Z7escape1Pi(i32* [[I1_SROA_0_0_COPYLOAD]])
+; CHECK-NEXT:    tail call void @_Z7escape1Pi(i32* [[TMP1]])
 ; CHECK-NEXT:    br label [[BB13]]
 ; CHECK:       bb13:
-; CHECK-NEXT:    ret i32* [[I1_SROA_0_0_COPYLOAD]]
+; CHECK-NEXT:    ret i32* [[TMP1]]
 ;
 bb:
   %i = alloca %0, align 8

diff  --git a/llvm/test/Transforms/SROA/address-spaces.ll b/llvm/test/Transforms/SROA/address-spaces.ll
index 70e1a682d7bfc..0300e99f9a217 100644
--- a/llvm/test/Transforms/SROA/address-spaces.ll
+++ b/llvm/test/Transforms/SROA/address-spaces.ll
@@ -11,8 +11,8 @@ declare void @llvm.memcpy.p1.p1.i32(ptr addrspace(1) nocapture, ptr addrspace(1)
 ; Make sure an illegal bitcast isn't introduced
 define void @test_address_space_1_1(ptr addrspace(1) %a, ptr addrspace(1) %b) {
 ; CHECK-LABEL: @test_address_space_1_1(
-; CHECK-NEXT:    [[AA_0_COPYLOAD:%.*]] = load <2 x i64>, ptr addrspace(1) [[A:%.*]], align 2
-; CHECK-NEXT:    store <2 x i64> [[AA_0_COPYLOAD]], ptr addrspace(1) [[B:%.*]], align 2
+; CHECK-NEXT:    [[AA_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr addrspace(1) [[A:%.*]], align 2
+; CHECK-NEXT:    store <16 x i8> [[AA_SROA_0_0_COPYLOAD]], ptr addrspace(1) [[B:%.*]], align 2
 ; CHECK-NEXT:    ret void
 ;
   %aa = alloca <2 x i64>, align 16
@@ -23,8 +23,8 @@ define void @test_address_space_1_1(ptr addrspace(1) %a, ptr addrspace(1) %b) {
 
 define void @test_address_space_1_0(ptr addrspace(1) %a, ptr %b) {
 ; CHECK-LABEL: @test_address_space_1_0(
-; CHECK-NEXT:    [[AA_0_COPYLOAD:%.*]] = load <2 x i64>, ptr addrspace(1) [[A:%.*]], align 2
-; CHECK-NEXT:    store <2 x i64> [[AA_0_COPYLOAD]], ptr [[B:%.*]], align 2
+; CHECK-NEXT:    [[AA_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr addrspace(1) [[A:%.*]], align 2
+; CHECK-NEXT:    store <16 x i8> [[AA_SROA_0_0_COPYLOAD]], ptr [[B:%.*]], align 2
 ; CHECK-NEXT:    ret void
 ;
   %aa = alloca <2 x i64>, align 16
@@ -35,8 +35,8 @@ define void @test_address_space_1_0(ptr addrspace(1) %a, ptr %b) {
 
 define void @test_address_space_0_1(ptr %a, ptr addrspace(1) %b) {
 ; CHECK-LABEL: @test_address_space_0_1(
-; CHECK-NEXT:    [[AA_0_COPYLOAD:%.*]] = load <2 x i64>, ptr [[A:%.*]], align 2
-; CHECK-NEXT:    store <2 x i64> [[AA_0_COPYLOAD]], ptr addrspace(1) [[B:%.*]], align 2
+; CHECK-NEXT:    [[AA_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr [[A:%.*]], align 2
+; CHECK-NEXT:    store <16 x i8> [[AA_SROA_0_0_COPYLOAD]], ptr addrspace(1) [[B:%.*]], align 2
 ; CHECK-NEXT:    ret void
 ;
   %aa = alloca <2 x i64>, align 16

diff  --git a/llvm/test/Transforms/SROA/alignment.ll b/llvm/test/Transforms/SROA/alignment.ll
index 66da09cd52d27..ba673b1d16465 100644
--- a/llvm/test/Transforms/SROA/alignment.ll
+++ b/llvm/test/Transforms/SROA/alignment.ll
@@ -92,15 +92,15 @@ define void @PR13920(ptr %a, ptr %b) {
 ; Test that alignments on memcpy intrinsics get propagated to loads and stores.
 ; CHECK-LABEL: @PR13920(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[AA_0_COPYLOAD:%.*]] = load <2 x i64>, ptr [[A:%.*]], align 2
-; CHECK-NEXT:    store <2 x i64> [[AA_0_COPYLOAD]], ptr [[B:%.*]], align 2
+; CHECK-NEXT:    [[AA_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr [[A:%.*]], align 2
+; CHECK-NEXT:    store <16 x i8> [[AA_SROA_0_0_COPYLOAD]], ptr [[B:%.*]], align 2
 ; CHECK-NEXT:    ret void
 ;
 ; DEBUGLOC-LABEL: @PR13920(
 ; DEBUGLOC-NEXT:  entry:
 ; DEBUGLOC-NEXT:    call void @llvm.dbg.value(metadata ptr undef, metadata [[META37:![0-9]+]], metadata !DIExpression()), !dbg [[DBG38:![0-9]+]]
-; DEBUGLOC-NEXT:    [[AA_0_COPYLOAD:%.*]] = load <2 x i64>, ptr [[A:%.*]], align 2, !dbg [[DBG39:![0-9]+]]
-; DEBUGLOC-NEXT:    store <2 x i64> [[AA_0_COPYLOAD]], ptr [[B:%.*]], align 2, !dbg [[DBG40:![0-9]+]]
+; DEBUGLOC-NEXT:    [[AA_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr [[A:%.*]], align 2, !dbg [[DBG39:![0-9]+]]
+; DEBUGLOC-NEXT:    store <16 x i8> [[AA_SROA_0_0_COPYLOAD]], ptr [[B:%.*]], align 2, !dbg [[DBG40:![0-9]+]]
 ; DEBUGLOC-NEXT:    ret void, !dbg [[DBG41:![0-9]+]]
 ;
 
@@ -118,21 +118,17 @@ define void @test3(ptr %x) {
 ; reduce the alignment.
 ; CHECK-LABEL: @test3(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[A_SROA_0:%.*]] = alloca [22 x i8], align 8
-; CHECK-NEXT:    [[B_SROA_0:%.*]] = alloca [18 x i8], align 2
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 8 [[A_SROA_0]], ptr align 8 [[X:%.*]], i32 22, i1 false)
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 2 [[B_SROA_0]], ptr align 2 [[X]], i32 18, i1 false)
+; CHECK-NEXT:    [[A_SROA_0_0_COPYLOAD:%.*]] = load <22 x i8>, ptr [[X:%.*]], align 8
+; CHECK-NEXT:    [[B_SROA_0_6_COPYLOAD:%.*]] = load <18 x i8>, ptr [[X]], align 2
 ; CHECK-NEXT:    ret void
 ;
 ; DEBUGLOC-LABEL: @test3(
 ; DEBUGLOC-NEXT:  entry:
-; DEBUGLOC-NEXT:    [[A_SROA_0:%.*]] = alloca [22 x i8], align 8, !dbg [[DBG47:![0-9]+]]
-; DEBUGLOC-NEXT:    call void @llvm.dbg.value(metadata ptr undef, metadata [[META44:![0-9]+]], metadata !DIExpression()), !dbg [[DBG47]]
-; DEBUGLOC-NEXT:    [[B_SROA_0:%.*]] = alloca [18 x i8], align 2, !dbg [[DBG48:![0-9]+]]
-; DEBUGLOC-NEXT:    call void @llvm.dbg.value(metadata ptr undef, metadata [[META45:![0-9]+]], metadata !DIExpression()), !dbg [[DBG48]]
-; DEBUGLOC-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 8 [[A_SROA_0]], ptr align 8 [[X:%.*]], i32 22, i1 false), !dbg [[DBG49:![0-9]+]]
+; DEBUGLOC-NEXT:    call void @llvm.dbg.value(metadata ptr undef, metadata [[META44:![0-9]+]], metadata !DIExpression()), !dbg [[DBG47:![0-9]+]]
+; DEBUGLOC-NEXT:    call void @llvm.dbg.value(metadata ptr undef, metadata [[META45:![0-9]+]], metadata !DIExpression()), !dbg [[DBG48:![0-9]+]]
+; DEBUGLOC-NEXT:    [[A_SROA_0_0_COPYLOAD:%.*]] = load <22 x i8>, ptr [[X:%.*]], align 8, !dbg [[DBG49:![0-9]+]]
 ; DEBUGLOC-NEXT:    call void @llvm.dbg.value(metadata ptr undef, metadata [[META46:![0-9]+]], metadata !DIExpression()), !dbg [[DBG50:![0-9]+]]
-; DEBUGLOC-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 2 [[B_SROA_0]], ptr align 2 [[X]], i32 18, i1 false), !dbg [[DBG51:![0-9]+]]
+; DEBUGLOC-NEXT:    [[B_SROA_0_6_COPYLOAD:%.*]] = load <18 x i8>, ptr [[X]], align 2, !dbg [[DBG51:![0-9]+]]
 ; DEBUGLOC-NEXT:    ret void, !dbg [[DBG52:![0-9]+]]
 ;
 

diff  --git a/llvm/test/Transforms/SROA/alloca-address-space.ll b/llvm/test/Transforms/SROA/alloca-address-space.ll
index d4f305c39c8fe..b06f269d806a5 100644
--- a/llvm/test/Transforms/SROA/alloca-address-space.ll
+++ b/llvm/test/Transforms/SROA/alloca-address-space.ll
@@ -10,8 +10,8 @@ declare void @llvm.memcpy.p1.p1.i32(ptr addrspace(1) nocapture, ptr addrspace(1)
 
 define void @test_address_space_1_1(ptr addrspace(1) %a, ptr addrspace(1) %b) {
 ; CHECK-LABEL: @test_address_space_1_1(
-; CHECK-NEXT:    [[AA_0_COPYLOAD:%.*]] = load <2 x i64>, ptr addrspace(1) [[A:%.*]], align 2
-; CHECK-NEXT:    store <2 x i64> [[AA_0_COPYLOAD]], ptr addrspace(1) [[B:%.*]], align 2
+; CHECK-NEXT:    [[AA_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr addrspace(1) [[A:%.*]], align 2
+; CHECK-NEXT:    store <16 x i8> [[AA_SROA_0_0_COPYLOAD]], ptr addrspace(1) [[B:%.*]], align 2
 ; CHECK-NEXT:    ret void
 ;
   %aa = alloca <2 x i64>, align 16, addrspace(2)
@@ -22,8 +22,8 @@ define void @test_address_space_1_1(ptr addrspace(1) %a, ptr addrspace(1) %b) {
 
 define void @test_address_space_1_0(ptr addrspace(1) %a, ptr addrspace(2) %b) {
 ; CHECK-LABEL: @test_address_space_1_0(
-; CHECK-NEXT:    [[AA_0_COPYLOAD:%.*]] = load <2 x i64>, ptr addrspace(1) [[A:%.*]], align 2
-; CHECK-NEXT:    store <2 x i64> [[AA_0_COPYLOAD]], ptr addrspace(2) [[B:%.*]], align 2
+; CHECK-NEXT:    [[AA_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr addrspace(1) [[A:%.*]], align 2
+; CHECK-NEXT:    store <16 x i8> [[AA_SROA_0_0_COPYLOAD]], ptr addrspace(2) [[B:%.*]], align 2
 ; CHECK-NEXT:    ret void
 ;
   %aa = alloca <2 x i64>, align 16, addrspace(2)
@@ -34,8 +34,8 @@ define void @test_address_space_1_0(ptr addrspace(1) %a, ptr addrspace(2) %b) {
 
 define void @test_address_space_0_1(ptr addrspace(2) %a, ptr addrspace(1) %b) {
 ; CHECK-LABEL: @test_address_space_0_1(
-; CHECK-NEXT:    [[AA_0_COPYLOAD:%.*]] = load <2 x i64>, ptr addrspace(2) [[A:%.*]], align 2
-; CHECK-NEXT:    store <2 x i64> [[AA_0_COPYLOAD]], ptr addrspace(1) [[B:%.*]], align 2
+; CHECK-NEXT:    [[AA_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr addrspace(2) [[A:%.*]], align 2
+; CHECK-NEXT:    store <16 x i8> [[AA_SROA_0_0_COPYLOAD]], ptr addrspace(1) [[B:%.*]], align 2
 ; CHECK-NEXT:    ret void
 ;
   %aa = alloca <2 x i64>, align 16, addrspace(2)

diff  --git a/llvm/test/Transforms/SROA/basictest.ll b/llvm/test/Transforms/SROA/basictest.ll
index 5ac8ed8c6e6a3..18874617e0b26 100644
--- a/llvm/test/Transforms/SROA/basictest.ll
+++ b/llvm/test/Transforms/SROA/basictest.ll
@@ -139,100 +139,83 @@ L2:
 define void @test3(ptr %dst, ptr align 8 %src) {
 ; CHECK-LABEL: @test3(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[A_SROA_0:%.*]] = alloca [42 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_3:%.*]] = alloca [99 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_32:%.*]] = alloca [16 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_15:%.*]] = alloca [42 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_16:%.*]] = alloca [7 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_235:%.*]] = alloca [7 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_31:%.*]] = alloca [85 x i8], align 1
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_0]], ptr align 8 [[SRC:%.*]], i32 42, i1 false), !tbaa [[TBAA0:![0-9]+]]
+; CHECK-NEXT:    [[A_SROA_0_SROA_0_0_COPYLOAD:%.*]] = load <42 x i8>, ptr [[SRC:%.*]], align 8, !tbaa [[TBAA0:![0-9]+]]
 ; CHECK-NEXT:    [[A_SROA_2_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 42
 ; CHECK-NEXT:    [[A_SROA_2_0_COPYLOAD:%.*]] = load i8, ptr [[A_SROA_2_0_SRC_SROA_IDX]], align 2, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_3_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 43
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_3]], ptr align 1 [[A_SROA_3_0_SRC_SROA_IDX]], i32 99, i1 false), !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_3_SROA_0_0_COPYLOAD:%.*]] = load <99 x i8>, ptr [[A_SROA_3_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_32_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 142
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_32]], ptr align 2 [[A_SROA_32_0_SRC_SROA_IDX]], i32 16, i1 false), !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr [[A_SROA_32_0_SRC_SROA_IDX]], align 2, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_15_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 158
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_15]], ptr align 2 [[A_SROA_15_0_SRC_SROA_IDX]], i32 42, i1 false), !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_15_SROA_0_0_COPYLOAD:%.*]] = load <42 x i8>, ptr [[A_SROA_15_0_SRC_SROA_IDX]], align 2, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_16_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 200
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_16]], ptr align 8 [[A_SROA_16_0_SRC_SROA_IDX]], i32 7, i1 false), !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_0_COPYLOAD:%.*]] = load <7 x i8>, ptr [[A_SROA_16_0_SRC_SROA_IDX]], align 8, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_23_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 207
 ; CHECK-NEXT:    [[A_SROA_23_0_COPYLOAD:%.*]] = load i8, ptr [[A_SROA_23_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_235_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 208
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_235]], ptr align 8 [[A_SROA_235_0_SRC_SROA_IDX]], i32 7, i1 false), !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_0_COPYLOAD:%.*]] = load <7 x i8>, ptr [[A_SROA_235_0_SRC_SROA_IDX]], align 8, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_31_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 215
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_31]], ptr align 1 [[A_SROA_31_0_SRC_SROA_IDX]], i32 85, i1 false), !tbaa [[TBAA0]]
-; CHECK-NEXT:    store i8 1, ptr [[A_SROA_32]], align 1, !tbaa [[TBAA3:![0-9]+]]
-; CHECK-NEXT:    store i16 1, ptr [[A_SROA_32]], align 1, !tbaa [[TBAA5:![0-9]+]]
-; CHECK-NEXT:    store i32 1, ptr [[A_SROA_32]], align 1, !tbaa [[TBAA7:![0-9]+]]
-; CHECK-NEXT:    store i64 1, ptr [[A_SROA_32]], align 1, !tbaa [[TBAA9:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_32_1_OVERLAP_2_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_32]], i64 1
-; CHECK-NEXT:    store i64 2, ptr [[A_SROA_32_1_OVERLAP_2_I8_SROA_IDX]], align 1, !tbaa [[TBAA11:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_32_2_OVERLAP_3_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_32]], i64 2
-; CHECK-NEXT:    store i64 3, ptr [[A_SROA_32_2_OVERLAP_3_I8_SROA_IDX]], align 1, !tbaa [[TBAA13:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_32_3_OVERLAP_4_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_32]], i64 3
-; CHECK-NEXT:    store i64 4, ptr [[A_SROA_32_3_OVERLAP_4_I8_SROA_IDX]], align 1, !tbaa [[TBAA15:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_32_4_OVERLAP_5_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_32]], i64 4
-; CHECK-NEXT:    store i64 5, ptr [[A_SROA_32_4_OVERLAP_5_I8_SROA_IDX]], align 1, !tbaa [[TBAA17:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_32_5_OVERLAP_6_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_32]], i64 5
-; CHECK-NEXT:    store i64 6, ptr [[A_SROA_32_5_OVERLAP_6_I8_SROA_IDX]], align 1, !tbaa [[TBAA19:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_32_6_OVERLAP_7_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_32]], i64 6
-; CHECK-NEXT:    store i64 7, ptr [[A_SROA_32_6_OVERLAP_7_I8_SROA_IDX]], align 1, !tbaa [[TBAA21:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_32_7_OVERLAP_8_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_32]], i64 7
-; CHECK-NEXT:    store i64 8, ptr [[A_SROA_32_7_OVERLAP_8_I8_SROA_IDX]], align 1, !tbaa [[TBAA23:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_32_8_OVERLAP_9_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_32]], i64 8
-; CHECK-NEXT:    store i64 9, ptr [[A_SROA_32_8_OVERLAP_9_I8_SROA_IDX]], align 1, !tbaa [[TBAA25:![0-9]+]]
-; CHECK-NEXT:    store i8 1, ptr [[A_SROA_16]], align 1, !tbaa [[TBAA27:![0-9]+]]
-; CHECK-NEXT:    store i16 1, ptr [[A_SROA_16]], align 1, !tbaa [[TBAA29:![0-9]+]]
-; CHECK-NEXT:    store i32 1, ptr [[A_SROA_16]], align 1, !tbaa [[TBAA31:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_16_1_OVERLAP2_1_1_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_16]], i64 1
-; CHECK-NEXT:    store i32 2, ptr [[A_SROA_16_1_OVERLAP2_1_1_I8_SROA_IDX]], align 1, !tbaa [[TBAA33:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_16_2_OVERLAP2_1_2_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_16]], i64 2
-; CHECK-NEXT:    store i32 3, ptr [[A_SROA_16_2_OVERLAP2_1_2_I8_SROA_IDX]], align 1, !tbaa [[TBAA35:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_16_3_OVERLAP2_1_3_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_16]], i64 3
-; CHECK-NEXT:    store i32 4, ptr [[A_SROA_16_3_OVERLAP2_1_3_I8_SROA_IDX]], align 1, !tbaa [[TBAA37:![0-9]+]]
-; CHECK-NEXT:    store i32 1, ptr [[A_SROA_235]], align 1, !tbaa [[TBAA39:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_235_1_OVERLAP2_2_1_I8_SROA_IDX11:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_235]], i64 1
-; CHECK-NEXT:    store i8 1, ptr [[A_SROA_235_1_OVERLAP2_2_1_I8_SROA_IDX11]], align 1, !tbaa [[TBAA41:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_235_1_OVERLAP2_2_1_I8_SROA_IDX10:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_235]], i64 1
-; CHECK-NEXT:    store i16 1, ptr [[A_SROA_235_1_OVERLAP2_2_1_I8_SROA_IDX10]], align 1, !tbaa [[TBAA43:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_235_1_OVERLAP2_2_1_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_235]], i64 1
-; CHECK-NEXT:    store i32 1, ptr [[A_SROA_235_1_OVERLAP2_2_1_I8_SROA_IDX]], align 1, !tbaa [[TBAA45:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_235_2_OVERLAP2_2_2_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_235]], i64 2
-; CHECK-NEXT:    store i32 3, ptr [[A_SROA_235_2_OVERLAP2_2_2_I8_SROA_IDX]], align 1, !tbaa [[TBAA47:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_235_3_OVERLAP2_2_3_I8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_235]], i64 3
-; CHECK-NEXT:    store i32 4, ptr [[A_SROA_235_3_OVERLAP2_2_3_I8_SROA_IDX]], align 1, !tbaa [[TBAA49:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_15_197_OVERLAP2_PREFIX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_15]], i64 39
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_15_197_OVERLAP2_PREFIX_SROA_IDX]], ptr align 1 [[SRC]], i32 3, i1 false), !tbaa [[TBAA51:![0-9]+]]
+; CHECK-NEXT:    [[A_SROA_31_SROA_0_0_COPYLOAD:%.*]] = load <85 x i8>, ptr [[A_SROA_31_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_0_VEC_INSERT:%.*]] = insertelement <16 x i8> [[A_SROA_32_SROA_0_0_COPYLOAD]], i8 1, i32 0
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_0_VECBLEND30:%.*]] = select <16 x i1> <i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false>, <16 x i8> <i8 extractelement (<2 x i8> bitcast (<1 x i16> <i16 1> to <2 x i8>), i32 0), i8 extractelement (<2 x i8> bitcast (<1 x i16> <i16 1> to <2 x i8>), i32 1), i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef>, <16 x i8> [[A_SROA_32_SROA_0_0_VEC_INSERT]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_0_VECBLEND28:%.*]] = select <16 x i1> <i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false>, <16 x i8> <i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 0), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 1), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 2), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 3), i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef>, <16 x i8> [[A_SROA_32_SROA_0_0_VECBLEND30]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_0_VECBLEND:%.*]] = select <16 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false>, <16 x i8> <i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 1> to <8 x i8>), i32 0), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 1> to <8 x i8>), i32 1), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 1> to <8 x i8>), i32 2), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 1> to <8 x i8>), i32 3), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 1> to <8 x i8>), i32 4), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 1> to <8 x i8>), i32 5), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 1> to <8 x i8>), i32 6), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 1> to <8 x i8>), i32 7), i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef>, <16 x i8> [[A_SROA_32_SROA_0_0_VECBLEND28]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_1_VECBLEND:%.*]] = select <16 x i1> <i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false>, <16 x i8> <i8 undef, i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 2> to <8 x i8>), i32 0), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 2> to <8 x i8>), i32 1), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 2> to <8 x i8>), i32 2), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 2> to <8 x i8>), i32 3), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 2> to <8 x i8>), i32 4), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 2> to <8 x i8>), i32 5), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 2> to <8 x i8>), i32 6), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 2> to <8 x i8>), i32 7), i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef>, <16 x i8> [[A_SROA_32_SROA_0_0_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_2_VECBLEND:%.*]] = select <16 x i1> <i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false>, <16 x i8> <i8 undef, i8 undef, i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 3> to <8 x i8>), i32 0), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 3> to <8 x i8>), i32 1), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 3> to <8 x i8>), i32 2), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 3> to <8 x i8>), i32 3), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 3> to <8 x i8>), i32 4), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 3> to <8 x i8>), i32 5), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 3> to <8 x i8>), i32 6), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 3> to <8 x i8>), i32 7), i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef>, <16 x i8> [[A_SROA_32_SROA_0_1_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_3_VECBLEND:%.*]] = select <16 x i1> <i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false>, <16 x i8> <i8 undef, i8 undef, i8 undef, i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 4> to <8 x i8>), i32 0), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 4> to <8 x i8>), i32 1), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 4> to <8 x i8>), i32 2), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 4> to <8 x i8>), i32 3), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 4> to <8 x i8>), i32 4), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 4> to <8 x i8>), i32 5), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 4> to <8 x i8>), i32 6), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 4> to <8 x i8>), i32 7), i8 undef, i8 undef, i8 undef, i8 undef, i8 undef>, <16 x i8> [[A_SROA_32_SROA_0_2_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_4_VECBLEND:%.*]] = select <16 x i1> <i1 false, i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false>, <16 x i8> <i8 undef, i8 undef, i8 undef, i8 undef, i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 5> to <8 x i8>), i32 0), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 5> to <8 x i8>), i32 1), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 5> to <8 x i8>), i32 2), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 5> to <8 x i8>), i32 3), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 5> to <8 x i8>), i32 4), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 5> to <8 x i8>), i32 5), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 5> to <8 x i8>), i32 6), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 5> to <8 x i8>), i32 7), i8 undef, i8 undef, i8 undef, i8 undef>, <16 x i8> [[A_SROA_32_SROA_0_3_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_5_VECBLEND:%.*]] = select <16 x i1> <i1 false, i1 false, i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false>, <16 x i8> <i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 6> to <8 x i8>), i32 0), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 6> to <8 x i8>), i32 1), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 6> to <8 x i8>), i32 2), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 6> to <8 x i8>), i32 3), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 6> to <8 x i8>), i32 4), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 6> to <8 x i8>), i32 5), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 6> to <8 x i8>), i32 6), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 6> to <8 x i8>), i32 7), i8 undef, i8 undef, i8 undef>, <16 x i8> [[A_SROA_32_SROA_0_4_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_6_VECBLEND:%.*]] = select <16 x i1> <i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false>, <16 x i8> <i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 7> to <8 x i8>), i32 0), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 7> to <8 x i8>), i32 1), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 7> to <8 x i8>), i32 2), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 7> to <8 x i8>), i32 3), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 7> to <8 x i8>), i32 4), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 7> to <8 x i8>), i32 5), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 7> to <8 x i8>), i32 6), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 7> to <8 x i8>), i32 7), i8 undef, i8 undef>, <16 x i8> [[A_SROA_32_SROA_0_5_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_7_VECBLEND:%.*]] = select <16 x i1> <i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false>, <16 x i8> <i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 8> to <8 x i8>), i32 0), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 8> to <8 x i8>), i32 1), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 8> to <8 x i8>), i32 2), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 8> to <8 x i8>), i32 3), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 8> to <8 x i8>), i32 4), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 8> to <8 x i8>), i32 5), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 8> to <8 x i8>), i32 6), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 8> to <8 x i8>), i32 7), i8 undef>, <16 x i8> [[A_SROA_32_SROA_0_6_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_32_SROA_0_8_VECBLEND:%.*]] = select <16 x i1> <i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>, <16 x i8> <i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef, i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 9> to <8 x i8>), i32 0), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 9> to <8 x i8>), i32 1), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 9> to <8 x i8>), i32 2), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 9> to <8 x i8>), i32 3), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 9> to <8 x i8>), i32 4), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 9> to <8 x i8>), i32 5), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 9> to <8 x i8>), i32 6), i8 extractelement (<8 x i8> bitcast (<1 x i64> <i64 9> to <8 x i8>), i32 7)>, <16 x i8> [[A_SROA_32_SROA_0_7_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_0_VEC_INSERT:%.*]] = insertelement <7 x i8> [[A_SROA_16_SROA_0_0_COPYLOAD]], i8 1, i32 0
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_0_VECBLEND20:%.*]] = select <7 x i1> <i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false>, <7 x i8> <i8 extractelement (<2 x i8> bitcast (<1 x i16> <i16 1> to <2 x i8>), i32 0), i8 extractelement (<2 x i8> bitcast (<1 x i16> <i16 1> to <2 x i8>), i32 1), i8 undef, i8 undef, i8 undef, i8 undef, i8 undef>, <7 x i8> [[A_SROA_16_SROA_0_0_VEC_INSERT]]
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_0_VECBLEND:%.*]] = select <7 x i1> <i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false>, <7 x i8> <i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 0), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 1), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 2), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 3), i8 undef, i8 undef, i8 undef>, <7 x i8> [[A_SROA_16_SROA_0_0_VECBLEND20]]
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_1_VECBLEND:%.*]] = select <7 x i1> <i1 false, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false>, <7 x i8> <i8 undef, i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 2> to <4 x i8>), i32 0), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 2> to <4 x i8>), i32 1), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 2> to <4 x i8>), i32 2), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 2> to <4 x i8>), i32 3), i8 undef, i8 undef>, <7 x i8> [[A_SROA_16_SROA_0_0_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_2_VECBLEND:%.*]] = select <7 x i1> <i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 false>, <7 x i8> <i8 undef, i8 undef, i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 3> to <4 x i8>), i32 0), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 3> to <4 x i8>), i32 1), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 3> to <4 x i8>), i32 2), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 3> to <4 x i8>), i32 3), i8 undef>, <7 x i8> [[A_SROA_16_SROA_0_1_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_3_VECBLEND:%.*]] = select <7 x i1> <i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true>, <7 x i8> <i8 undef, i8 undef, i8 undef, i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 4> to <4 x i8>), i32 0), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 4> to <4 x i8>), i32 1), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 4> to <4 x i8>), i32 2), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 4> to <4 x i8>), i32 3)>, <7 x i8> [[A_SROA_16_SROA_0_2_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_0_VECBLEND:%.*]] = select <7 x i1> <i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false>, <7 x i8> <i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 0), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 1), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 2), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 3), i8 undef, i8 undef, i8 undef>, <7 x i8> [[A_SROA_235_SROA_0_0_COPYLOAD]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_1_VEC_INSERT:%.*]] = insertelement <7 x i8> [[A_SROA_235_SROA_0_0_VECBLEND]], i8 1, i32 1
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_1_VECBLEND15:%.*]] = select <7 x i1> <i1 false, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false>, <7 x i8> <i8 undef, i8 extractelement (<2 x i8> bitcast (<1 x i16> <i16 1> to <2 x i8>), i32 0), i8 extractelement (<2 x i8> bitcast (<1 x i16> <i16 1> to <2 x i8>), i32 1), i8 undef, i8 undef, i8 undef, i8 undef>, <7 x i8> [[A_SROA_235_SROA_0_1_VEC_INSERT]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_1_VECBLEND:%.*]] = select <7 x i1> <i1 false, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false>, <7 x i8> <i8 undef, i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 0), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 1), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 2), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 1> to <4 x i8>), i32 3), i8 undef, i8 undef>, <7 x i8> [[A_SROA_235_SROA_0_1_VECBLEND15]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_2_VECBLEND:%.*]] = select <7 x i1> <i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 false>, <7 x i8> <i8 undef, i8 undef, i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 3> to <4 x i8>), i32 0), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 3> to <4 x i8>), i32 1), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 3> to <4 x i8>), i32 2), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 3> to <4 x i8>), i32 3), i8 undef>, <7 x i8> [[A_SROA_235_SROA_0_1_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_3_VECBLEND:%.*]] = select <7 x i1> <i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true>, <7 x i8> <i8 undef, i8 undef, i8 undef, i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 4> to <4 x i8>), i32 0), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 4> to <4 x i8>), i32 1), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 4> to <4 x i8>), i32 2), i8 extractelement (<4 x i8> bitcast (<1 x i32> <i32 4> to <4 x i8>), i32 3)>, <7 x i8> [[A_SROA_235_SROA_0_2_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_15_SROA_0_39_COPYLOAD:%.*]] = load <3 x i8>, ptr [[SRC]], align 1, !tbaa [[TBAA3:![0-9]+]]
+; CHECK-NEXT:    [[A_SROA_15_SROA_0_39_VEC_EXPAND:%.*]] = shufflevector <3 x i8> [[A_SROA_15_SROA_0_39_COPYLOAD]], <3 x i8> poison, <42 x i32> <i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 0, i32 1, i32 2>
+; CHECK-NEXT:    [[A_SROA_15_SROA_0_39_VECBLEND:%.*]] = select <42 x i1> <i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 true, i1 true, i1 true>, <42 x i8> [[A_SROA_15_SROA_0_39_VEC_EXPAND]], <42 x i8> [[A_SROA_15_SROA_0_0_COPYLOAD]]
 ; CHECK-NEXT:    [[A_SROA_16_197_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 3
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_16]], ptr align 1 [[A_SROA_16_197_SRC_SROA_IDX]], i32 5, i1 false), !tbaa [[TBAA51]]
-; CHECK-NEXT:    [[A_SROA_16_2_OVERLAP2_1_2_I8_SROA_IDX12:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_16]], i64 2
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 1 [[A_SROA_16_2_OVERLAP2_1_2_I8_SROA_IDX12]], i8 42, i32 5, i1 false), !tbaa [[TBAA53:![0-9]+]]
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 1 [[A_SROA_235]], i8 42, i32 2, i1 false), !tbaa [[TBAA53]]
-; CHECK-NEXT:    [[A_SROA_235_209_OVERLAP2_2_1_I8_SROA_IDX8:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_235]], i64 1
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_235_209_OVERLAP2_2_1_I8_SROA_IDX8]], ptr align 1 [[SRC]], i32 5, i1 false), !tbaa [[TBAA55:![0-9]+]]
-; CHECK-NEXT:    [[A_SROA_235_210_OVERLAP2_2_2_I8_SROA_IDX9:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_235]], i64 2
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_235_210_OVERLAP2_2_2_I8_SROA_IDX9]], ptr align 1 [[SRC]], i32 5, i1 false), !tbaa [[TBAA57:![0-9]+]]
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_0_COPYLOAD23:%.*]] = load <5 x i8>, ptr [[A_SROA_16_197_SRC_SROA_IDX]], align 1, !tbaa [[TBAA3]]
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_0_VEC_EXPAND:%.*]] = shufflevector <5 x i8> [[A_SROA_16_SROA_0_0_COPYLOAD23]], <5 x i8> poison, <7 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 undef, i32 undef>
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_0_VECBLEND24:%.*]] = select <7 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false>, <7 x i8> [[A_SROA_16_SROA_0_0_VEC_EXPAND]], <7 x i8> [[A_SROA_16_SROA_0_3_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_16_SROA_0_2_VECBLEND25:%.*]] = select <7 x i1> <i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true>, <7 x i8> <i8 undef, i8 undef, i8 42, i8 42, i8 42, i8 42, i8 42>, <7 x i8> [[A_SROA_16_SROA_0_0_VECBLEND24]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_0_VECBLEND13:%.*]] = select <7 x i1> <i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false>, <7 x i8> <i8 42, i8 42, i8 undef, i8 undef, i8 undef, i8 undef, i8 undef>, <7 x i8> [[A_SROA_235_SROA_0_3_VECBLEND]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_1_COPYLOAD:%.*]] = load <5 x i8>, ptr [[SRC]], align 1, !tbaa [[TBAA5:![0-9]+]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_1_VEC_EXPAND:%.*]] = shufflevector <5 x i8> [[A_SROA_235_SROA_0_1_COPYLOAD]], <5 x i8> poison, <7 x i32> <i32 undef, i32 0, i32 1, i32 2, i32 3, i32 4, i32 undef>
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_1_VECBLEND17:%.*]] = select <7 x i1> <i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false>, <7 x i8> [[A_SROA_235_SROA_0_1_VEC_EXPAND]], <7 x i8> [[A_SROA_235_SROA_0_0_VECBLEND13]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_2_COPYLOAD:%.*]] = load <5 x i8>, ptr [[SRC]], align 1, !tbaa [[TBAA7:![0-9]+]]
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_2_VEC_EXPAND:%.*]] = shufflevector <5 x i8> [[A_SROA_235_SROA_0_2_COPYLOAD]], <5 x i8> poison, <7 x i32> <i32 undef, i32 undef, i32 0, i32 1, i32 2, i32 3, i32 4>
+; CHECK-NEXT:    [[A_SROA_235_SROA_0_2_VECBLEND18:%.*]] = select <7 x i1> <i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true>, <7 x i8> [[A_SROA_235_SROA_0_2_VEC_EXPAND]], <7 x i8> [[A_SROA_235_SROA_0_1_VECBLEND17]]
 ; CHECK-NEXT:    [[A_SROA_31_210_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 5
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_31]], ptr align 1 [[A_SROA_31_210_SRC_SROA_IDX]], i32 3, i1 false), !tbaa [[TBAA57]]
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[DST:%.*]], ptr align 1 [[A_SROA_0]], i32 42, i1 false), !tbaa [[TBAA59:![0-9]+]]
+; CHECK-NEXT:    [[A_SROA_31_SROA_0_0_COPYLOAD11:%.*]] = load <3 x i8>, ptr [[A_SROA_31_210_SRC_SROA_IDX]], align 1, !tbaa [[TBAA7]]
+; CHECK-NEXT:    [[A_SROA_31_SROA_0_0_VEC_EXPAND:%.*]] = shufflevector <3 x i8> [[A_SROA_31_SROA_0_0_COPYLOAD11]], <3 x i8> poison, <85 x i32> <i32 0, i32 1, i32 2, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+; CHECK-NEXT:    [[A_SROA_31_SROA_0_0_VECBLEND:%.*]] = select <85 x i1> <i1 true, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false>, <85 x i8> [[A_SROA_31_SROA_0_0_VEC_EXPAND]], <85 x i8> [[A_SROA_31_SROA_0_0_COPYLOAD]]
+; CHECK-NEXT:    store <42 x i8> [[A_SROA_0_SROA_0_0_COPYLOAD]], ptr [[DST:%.*]], align 1, !tbaa [[TBAA9:![0-9]+]]
 ; CHECK-NEXT:    [[A_SROA_2_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 42
-; CHECK-NEXT:    store i8 0, ptr [[A_SROA_2_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA59]]
+; CHECK-NEXT:    store i8 0, ptr [[A_SROA_2_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA9]]
 ; CHECK-NEXT:    [[A_SROA_3_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 43
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_3_0_DST_SROA_IDX]], ptr align 1 [[A_SROA_3]], i32 99, i1 false), !tbaa [[TBAA59]]
+; CHECK-NEXT:    store <99 x i8> [[A_SROA_3_SROA_0_0_COPYLOAD]], ptr [[A_SROA_3_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA9]]
 ; CHECK-NEXT:    [[A_SROA_32_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 142
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_32_0_DST_SROA_IDX]], ptr align 1 [[A_SROA_32]], i32 16, i1 false), !tbaa [[TBAA59]]
+; CHECK-NEXT:    store <16 x i8> [[A_SROA_32_SROA_0_8_VECBLEND]], ptr [[A_SROA_32_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA9]]
 ; CHECK-NEXT:    [[A_SROA_15_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 158
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_15_0_DST_SROA_IDX]], ptr align 1 [[A_SROA_15]], i32 42, i1 false), !tbaa [[TBAA59]]
+; CHECK-NEXT:    store <42 x i8> [[A_SROA_15_SROA_0_39_VECBLEND]], ptr [[A_SROA_15_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA9]]
 ; CHECK-NEXT:    [[A_SROA_16_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 200
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_16_0_DST_SROA_IDX]], ptr align 1 [[A_SROA_16]], i32 7, i1 false), !tbaa [[TBAA59]]
+; CHECK-NEXT:    store <7 x i8> [[A_SROA_16_SROA_0_2_VECBLEND25]], ptr [[A_SROA_16_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA9]]
 ; CHECK-NEXT:    [[A_SROA_23_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 207
-; CHECK-NEXT:    store i8 42, ptr [[A_SROA_23_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA59]]
+; CHECK-NEXT:    store i8 42, ptr [[A_SROA_23_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA9]]
 ; CHECK-NEXT:    [[A_SROA_235_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 208
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_235_0_DST_SROA_IDX]], ptr align 1 [[A_SROA_235]], i32 7, i1 false), !tbaa [[TBAA59]]
+; CHECK-NEXT:    store <7 x i8> [[A_SROA_235_SROA_0_2_VECBLEND18]], ptr [[A_SROA_235_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA9]]
 ; CHECK-NEXT:    [[A_SROA_31_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 215
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_31_0_DST_SROA_IDX]], ptr align 1 [[A_SROA_31]], i32 85, i1 false), !tbaa [[TBAA59]]
+; CHECK-NEXT:    store <85 x i8> [[A_SROA_31_SROA_0_0_VECBLEND]], ptr [[A_SROA_31_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA9]]
 ; CHECK-NEXT:    ret void
 ;
 
@@ -315,60 +298,29 @@ entry:
 define void @test4(ptr %dst, ptr %src) {
 ; CHECK-LABEL: @test4(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[A_SROA_0:%.*]] = alloca [20 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_2_SROA_4:%.*]] = alloca [7 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_3:%.*]] = alloca [10 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_31_SROA_5:%.*]] = alloca [7 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_6_SROA_4:%.*]] = alloca [7 x i8], align 1
-; CHECK-NEXT:    [[A_SROA_7:%.*]] = alloca [40 x i8], align 1
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_0]], ptr align 1 [[SRC:%.*]], i32 20, i1 false), !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_0_SROA_0_0_COPYLOAD:%.*]] = load <20 x i8>, ptr [[SRC:%.*]], align 1, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_2_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 20
-; CHECK-NEXT:    [[A_SROA_2_SROA_0_0_COPYLOAD:%.*]] = load i16, ptr [[A_SROA_2_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
-; CHECK-NEXT:    [[A_SROA_2_SROA_3_0_A_SROA_2_0_SRC_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_2_0_SRC_SROA_IDX]], i64 2
-; CHECK-NEXT:    [[A_SROA_2_SROA_3_0_COPYLOAD:%.*]] = load i8, ptr [[A_SROA_2_SROA_3_0_A_SROA_2_0_SRC_SROA_IDX_SROA_IDX]], align 1, !tbaa [[TBAA0]]
-; CHECK-NEXT:    [[A_SROA_2_SROA_4_0_A_SROA_2_0_SRC_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_2_0_SRC_SROA_IDX]], i64 3
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_2_SROA_4]], ptr align 1 [[A_SROA_2_SROA_4_0_A_SROA_2_0_SRC_SROA_IDX_SROA_IDX]], i32 7, i1 false), !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_2_SROA_0_0_COPYLOAD:%.*]] = load <10 x i8>, ptr [[A_SROA_2_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_3_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 30
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_3]], ptr align 1 [[A_SROA_3_0_SRC_SROA_IDX]], i32 10, i1 false), !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_3_SROA_0_0_COPYLOAD:%.*]] = load <10 x i8>, ptr [[A_SROA_3_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_31_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 40
-; CHECK-NEXT:    [[A_SROA_31_SROA_0_0_COPYLOAD:%.*]] = load i16, ptr [[A_SROA_31_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
-; CHECK-NEXT:    [[A_SROA_31_SROA_4_0_A_SROA_31_0_SRC_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_31_0_SRC_SROA_IDX]], i64 2
-; CHECK-NEXT:    [[A_SROA_31_SROA_4_0_COPYLOAD:%.*]] = load i8, ptr [[A_SROA_31_SROA_4_0_A_SROA_31_0_SRC_SROA_IDX_SROA_IDX]], align 1, !tbaa [[TBAA0]]
-; CHECK-NEXT:    [[A_SROA_31_SROA_5_0_A_SROA_31_0_SRC_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_31_0_SRC_SROA_IDX]], i64 3
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_31_SROA_5]], ptr align 1 [[A_SROA_31_SROA_5_0_A_SROA_31_0_SRC_SROA_IDX_SROA_IDX]], i32 7, i1 false), !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_31_SROA_0_0_COPYLOAD:%.*]] = load <10 x i8>, ptr [[A_SROA_31_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_6_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 50
-; CHECK-NEXT:    [[A_SROA_6_SROA_0_0_COPYLOAD:%.*]] = load i16, ptr [[A_SROA_6_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
-; CHECK-NEXT:    [[A_SROA_6_SROA_3_0_A_SROA_6_0_SRC_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_6_0_SRC_SROA_IDX]], i64 2
-; CHECK-NEXT:    [[A_SROA_6_SROA_3_0_COPYLOAD:%.*]] = load i8, ptr [[A_SROA_6_SROA_3_0_A_SROA_6_0_SRC_SROA_IDX_SROA_IDX]], align 1, !tbaa [[TBAA0]]
-; CHECK-NEXT:    [[A_SROA_6_SROA_4_0_A_SROA_6_0_SRC_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_6_0_SRC_SROA_IDX]], i64 3
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_6_SROA_4]], ptr align 1 [[A_SROA_6_SROA_4_0_A_SROA_6_0_SRC_SROA_IDX_SROA_IDX]], i32 7, i1 false), !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_6_SROA_0_0_COPYLOAD:%.*]] = load <10 x i8>, ptr [[A_SROA_6_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_7_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 60
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_7]], ptr align 1 [[A_SROA_7_0_SRC_SROA_IDX]], i32 40, i1 false), !tbaa [[TBAA0]]
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_31_SROA_5]], ptr align 1 [[A_SROA_2_SROA_4]], i32 7, i1 false), !tbaa [[TBAA3]]
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_31_SROA_5]], ptr align 1 [[A_SROA_6_SROA_4]], i32 7, i1 false), !tbaa [[TBAA5]]
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[DST:%.*]], ptr align 1 [[A_SROA_0]], i32 20, i1 false), !tbaa [[TBAA7]]
+; CHECK-NEXT:    [[A_SROA_7_SROA_0_0_COPYLOAD:%.*]] = load <40 x i8>, ptr [[A_SROA_7_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
+; CHECK-NEXT:    [[A_SROA_31_SROA_0_2_VEC_INSERT:%.*]] = insertelement <10 x i8> [[A_SROA_2_SROA_0_0_COPYLOAD]], i8 0, i32 2
+; CHECK-NEXT:    store <20 x i8> [[A_SROA_0_SROA_0_0_COPYLOAD]], ptr [[DST:%.*]], align 1, !tbaa [[TBAA11:![0-9]+]]
 ; CHECK-NEXT:    [[A_SROA_2_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 20
-; CHECK-NEXT:    store i16 [[A_SROA_2_SROA_0_0_COPYLOAD]], ptr [[A_SROA_2_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA7]]
-; CHECK-NEXT:    [[A_SROA_2_SROA_3_0_A_SROA_2_0_DST_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_2_0_DST_SROA_IDX]], i64 2
-; CHECK-NEXT:    store i8 [[A_SROA_2_SROA_3_0_COPYLOAD]], ptr [[A_SROA_2_SROA_3_0_A_SROA_2_0_DST_SROA_IDX_SROA_IDX]], align 1, !tbaa [[TBAA7]]
-; CHECK-NEXT:    [[A_SROA_2_SROA_4_0_A_SROA_2_0_DST_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_2_0_DST_SROA_IDX]], i64 3
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_2_SROA_4_0_A_SROA_2_0_DST_SROA_IDX_SROA_IDX]], ptr align 1 [[A_SROA_2_SROA_4]], i32 7, i1 false), !tbaa [[TBAA7]]
+; CHECK-NEXT:    store <10 x i8> [[A_SROA_2_SROA_0_0_COPYLOAD]], ptr [[A_SROA_2_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA11]]
 ; CHECK-NEXT:    [[A_SROA_3_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 30
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_3_0_DST_SROA_IDX]], ptr align 1 [[A_SROA_3]], i32 10, i1 false), !tbaa [[TBAA7]]
+; CHECK-NEXT:    store <10 x i8> [[A_SROA_3_SROA_0_0_COPYLOAD]], ptr [[A_SROA_3_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA11]]
 ; CHECK-NEXT:    [[A_SROA_31_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 40
-; CHECK-NEXT:    store i16 [[A_SROA_6_SROA_0_0_COPYLOAD]], ptr [[A_SROA_31_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA7]]
-; CHECK-NEXT:    [[A_SROA_31_SROA_4_0_A_SROA_31_0_DST_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_31_0_DST_SROA_IDX]], i64 2
-; CHECK-NEXT:    store i8 [[A_SROA_6_SROA_3_0_COPYLOAD]], ptr [[A_SROA_31_SROA_4_0_A_SROA_31_0_DST_SROA_IDX_SROA_IDX]], align 1, !tbaa [[TBAA7]]
-; CHECK-NEXT:    [[A_SROA_31_SROA_5_0_A_SROA_31_0_DST_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_31_0_DST_SROA_IDX]], i64 3
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_31_SROA_5_0_A_SROA_31_0_DST_SROA_IDX_SROA_IDX]], ptr align 1 [[A_SROA_31_SROA_5]], i32 7, i1 false), !tbaa [[TBAA7]]
+; CHECK-NEXT:    store <10 x i8> [[A_SROA_6_SROA_0_0_COPYLOAD]], ptr [[A_SROA_31_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA11]]
 ; CHECK-NEXT:    [[A_SROA_6_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 50
-; CHECK-NEXT:    store i16 [[A_SROA_6_SROA_0_0_COPYLOAD]], ptr [[A_SROA_6_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA7]]
-; CHECK-NEXT:    [[A_SROA_6_SROA_3_0_A_SROA_6_0_DST_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_6_0_DST_SROA_IDX]], i64 2
-; CHECK-NEXT:    store i8 [[A_SROA_6_SROA_3_0_COPYLOAD]], ptr [[A_SROA_6_SROA_3_0_A_SROA_6_0_DST_SROA_IDX_SROA_IDX]], align 1, !tbaa [[TBAA7]]
-; CHECK-NEXT:    [[A_SROA_6_SROA_4_0_A_SROA_6_0_DST_SROA_IDX_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[A_SROA_6_0_DST_SROA_IDX]], i64 3
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_6_SROA_4_0_A_SROA_6_0_DST_SROA_IDX_SROA_IDX]], ptr align 1 [[A_SROA_6_SROA_4]], i32 7, i1 false), !tbaa [[TBAA7]]
+; CHECK-NEXT:    store <10 x i8> [[A_SROA_6_SROA_0_0_COPYLOAD]], ptr [[A_SROA_6_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA11]]
 ; CHECK-NEXT:    [[A_SROA_7_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 60
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_7_0_DST_SROA_IDX]], ptr align 1 [[A_SROA_7]], i32 40, i1 false), !tbaa [[TBAA7]]
+; CHECK-NEXT:    store <40 x i8> [[A_SROA_7_SROA_0_0_COPYLOAD]], ptr [[A_SROA_7_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA11]]
 ; CHECK-NEXT:    ret void
 ;
 
@@ -455,8 +407,8 @@ define void @test7(ptr %src, ptr %dst) {
 ; CHECK-NEXT:    [[A_SROA_0:%.*]] = alloca i32, align 4
 ; CHECK-NEXT:    [[A_SROA_0_0_COPYLOAD:%.*]] = load volatile i32, ptr [[SRC:%.*]], align 1, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    store volatile i32 [[A_SROA_0_0_COPYLOAD]], ptr [[A_SROA_0]], align 4, !tbaa [[TBAA0]]
-; CHECK-NEXT:    [[A_SROA_0_0_A_SROA_0_0_COPYLOAD1:%.*]] = load volatile i32, ptr [[A_SROA_0]], align 4, !tbaa [[TBAA3]]
-; CHECK-NEXT:    store volatile i32 [[A_SROA_0_0_A_SROA_0_0_COPYLOAD1]], ptr [[DST:%.*]], align 1, !tbaa [[TBAA3]]
+; CHECK-NEXT:    [[A_SROA_0_0_A_SROA_0_0_COPYLOAD1:%.*]] = load volatile i32, ptr [[A_SROA_0]], align 4, !tbaa [[TBAA13:![0-9]+]]
+; CHECK-NEXT:    store volatile i32 [[A_SROA_0_0_A_SROA_0_0_COPYLOAD1]], ptr [[DST:%.*]], align 1, !tbaa [[TBAA13]]
 ; CHECK-NEXT:    ret void
 ;
 
@@ -476,9 +428,9 @@ define %S2 @test8(ptr %arg) {
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    [[S2_NEXT_PTR:%.*]] = getelementptr [[S2:%.*]], ptr [[ARG:%.*]], i64 0, i32 1
 ; CHECK-NEXT:    [[S2_NEXT:%.*]] = load ptr, ptr [[S2_NEXT_PTR]], align 8, !tbaa [[TBAA0]]
-; CHECK-NEXT:    [[S2_NEXT_S1:%.*]] = load ptr, ptr [[S2_NEXT]], align 8, !tbaa [[TBAA3]]
+; CHECK-NEXT:    [[S2_NEXT_S1:%.*]] = load ptr, ptr [[S2_NEXT]], align 8, !tbaa [[TBAA13]]
 ; CHECK-NEXT:    [[S2_NEXT_NEXT_PTR:%.*]] = getelementptr [[S2]], ptr [[S2_NEXT]], i64 0, i32 1
-; CHECK-NEXT:    [[S2_NEXT_NEXT:%.*]] = load ptr, ptr [[S2_NEXT_NEXT_PTR]], align 8, !tbaa [[TBAA7]]
+; CHECK-NEXT:    [[S2_NEXT_NEXT:%.*]] = load ptr, ptr [[S2_NEXT_NEXT_PTR]], align 8, !tbaa [[TBAA11]]
 ; CHECK-NEXT:    [[RESULT1:%.*]] = insertvalue [[S2]] poison, ptr [[S2_NEXT_S1]], 0
 ; CHECK-NEXT:    [[RESULT2:%.*]] = insertvalue [[S2]] [[RESULT1]], ptr [[S2_NEXT_NEXT]], 1
 ; CHECK-NEXT:    ret [[S2]] [[RESULT2]]
@@ -725,7 +677,7 @@ define void @test16(ptr %src, ptr %dst) {
 ; CHECK-LABEL: @test16(
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    [[A_SROA_0_0_COPYLOAD:%.*]] = load i24, ptr [[SRC:%.*]], align 1, !tbaa [[TBAA0]]
-; CHECK-NEXT:    store i24 0, ptr [[DST:%.*]], align 1, !tbaa [[TBAA5]]
+; CHECK-NEXT:    store i24 0, ptr [[DST:%.*]], align 1, !tbaa [[TBAA15:![0-9]+]]
 ; CHECK-NEXT:    ret void
 ;
 
@@ -744,7 +696,7 @@ define void @test17(ptr %src, ptr %dst) {
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    [[A:%.*]] = alloca [3 x i8], align 1
 ; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr [[A]], ptr [[SRC:%.*]], i32 4, i1 true), !tbaa [[TBAA0]]
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr [[DST:%.*]], ptr [[A]], i32 4, i1 true), !tbaa [[TBAA3]]
+; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr [[DST:%.*]], ptr [[A]], i32 4, i1 true), !tbaa [[TBAA13]]
 ; CHECK-NEXT:    ret void
 ;
 
@@ -765,12 +717,12 @@ define void @test18(ptr %src, ptr %dst, i32 %size) {
 ; CHECK-NEXT:    [[A_SROA_0_0_COPYLOAD:%.*]] = load i32, ptr [[SRC:%.*]], align 1, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    [[A_SROA_3_0_SRC_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i64 4
 ; CHECK-NEXT:    [[A_SROA_3_0_COPYLOAD:%.*]] = load i32, ptr [[A_SROA_3_0_SRC_SROA_IDX]], align 1, !tbaa [[TBAA0]]
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_33]], ptr [[SRC]], i32 [[SIZE:%.*]], i1 false), !tbaa [[TBAA3]]
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 1 [[A_SROA_33]], i8 42, i32 [[SIZE]], i1 false), !tbaa [[TBAA5]]
-; CHECK-NEXT:    store i32 42, ptr [[DST:%.*]], align 1, !tbaa [[TBAA9]]
+; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[A_SROA_33]], ptr [[SRC]], i32 [[SIZE:%.*]], i1 false), !tbaa [[TBAA13]]
+; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 1 [[A_SROA_33]], i8 42, i32 [[SIZE]], i1 false), !tbaa [[TBAA15]]
+; CHECK-NEXT:    store i32 42, ptr [[DST:%.*]], align 1, !tbaa [[TBAA17:![0-9]+]]
 ; CHECK-NEXT:    [[A_SROA_3_0_DST_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[DST]], i64 4
-; CHECK-NEXT:    store i32 [[A_SROA_3_0_COPYLOAD]], ptr [[A_SROA_3_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA9]]
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr [[DST]], ptr align 1 [[A_SROA_33]], i32 [[SIZE]], i1 false), !tbaa [[TBAA11]]
+; CHECK-NEXT:    store i32 [[A_SROA_3_0_COPYLOAD]], ptr [[A_SROA_3_0_DST_SROA_IDX]], align 1, !tbaa [[TBAA17]]
+; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr [[DST]], ptr align 1 [[A_SROA_33]], i32 [[SIZE]], i1 false), !tbaa [[TBAA19:![0-9]+]]
 ; CHECK-NEXT:    ret void
 ;
 
@@ -1005,8 +957,7 @@ define void @PR14034(ptr %ptr, ptr %ptr2) {
 ; thing is to handle empty structs gracefully.
 ; CHECK-LABEL: @PR14034(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[A_SROA_0:%.*]] = alloca [12 x i8], align 8
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[PTR2:%.*]], ptr align 8 [[A_SROA_0]], i32 12, i1 false)
+; CHECK-NEXT:    store <12 x i8> undef, ptr [[PTR2:%.*]], align 1
 ; CHECK-NEXT:    ret void
 ;
 
@@ -1227,8 +1178,6 @@ define void @PR14465() {
 ; Ensure that we don't crash when analyzing a alloca larger than the maximum
 ; integer type width (MAX_INT_BITS) supported by llvm (1048576*32 > (1<<23)-1).
 ; CHECK-LABEL: @PR14465(
-; CHECK-NEXT:    [[STACK:%.*]] = alloca [1048576 x i32], align 16
-; CHECK-NEXT:    call void @llvm.memset.p0.i64(ptr align 16 [[STACK]], i8 -2, i64 4194304, i1 false)
 ; CHECK-NEXT:    ret void
 ;
 
@@ -1547,8 +1496,8 @@ define void @test24(ptr %src, ptr %dst) {
 ; CHECK-NEXT:    [[A:%.*]] = alloca i64, align 16
 ; CHECK-NEXT:    [[A_0_COPYLOAD:%.*]] = load volatile i64, ptr [[SRC:%.*]], align 1, !tbaa [[TBAA0]]
 ; CHECK-NEXT:    store volatile i64 [[A_0_COPYLOAD]], ptr [[A]], align 16, !tbaa [[TBAA0]]
-; CHECK-NEXT:    [[A_0_COPYLOAD1:%.*]] = load volatile i64, ptr [[A]], align 16, !tbaa [[TBAA3]]
-; CHECK-NEXT:    store volatile i64 [[A_0_COPYLOAD1]], ptr [[DST:%.*]], align 1, !tbaa [[TBAA3]]
+; CHECK-NEXT:    [[A_0_COPYLOAD1:%.*]] = load volatile i64, ptr [[A]], align 16, !tbaa [[TBAA13]]
+; CHECK-NEXT:    store volatile i64 [[A_0_COPYLOAD1]], ptr [[DST:%.*]], align 1, !tbaa [[TBAA13]]
 ; CHECK-NEXT:    ret void
 ;
 

diff  --git a/llvm/test/Transforms/SROA/pointer-offset-size.ll b/llvm/test/Transforms/SROA/pointer-offset-size.ll
index 76b52098a7e62..bf3c63c1ae7a3 100644
--- a/llvm/test/Transforms/SROA/pointer-offset-size.ll
+++ b/llvm/test/Transforms/SROA/pointer-offset-size.ll
@@ -8,8 +8,7 @@ target datalayout = "e-p:64:64:64:32"
 define i16 @test(ptr %ts2.i) {
 ; CHECK-LABEL: @test(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[S_SROA_0:%.*]] = alloca [3 x i8], align 8
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[TS2_I:%.*]], ptr align 8 [[S_SROA_0]], i32 3, i1 false)
+; CHECK-NEXT:    store <3 x i8> undef, ptr [[TS2_I:%.*]], align 1
 ; CHECK-NEXT:    [[TMP0:%.*]] = load i16, ptr [[TS2_I]], align 2
 ; CHECK-NEXT:    ret i16 [[TMP0]]
 ;

diff  --git a/llvm/test/Transforms/SROA/scalable-vectors.ll b/llvm/test/Transforms/SROA/scalable-vectors.ll
index d96f4dba868b3..bf9742e570084 100644
--- a/llvm/test/Transforms/SROA/scalable-vectors.ll
+++ b/llvm/test/Transforms/SROA/scalable-vectors.ll
@@ -67,7 +67,7 @@ define <vscale x 4 x i32> @cast_alloca_to_svint32_t(<vscale x 4 x i32> %type.coe
 define <vscale x 4 x i32> @cast_alloca_from_svint32_t() {
 ; CHECK-LABEL: @cast_alloca_from_svint32_t(
 ; CHECK-NEXT:    [[RETVAL_COERCE:%.*]] = alloca <vscale x 4 x i32>, align 16
-; CHECK-NEXT:    store <16 x i32> undef, ptr [[RETVAL_COERCE]], align 16
+; CHECK-NEXT:    store <64 x i8> undef, ptr [[RETVAL_COERCE]], align 16
 ; CHECK-NEXT:    [[TMP1:%.*]] = load <vscale x 4 x i32>, ptr [[RETVAL_COERCE]], align 16
 ; CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 ;

diff  --git a/llvm/test/Transforms/SROA/slice-width.ll b/llvm/test/Transforms/SROA/slice-width.ll
index 7d2aeaaff57bc..2aa54a104153f 100644
--- a/llvm/test/Transforms/SROA/slice-width.ll
+++ b/llvm/test/Transforms/SROA/slice-width.ll
@@ -46,8 +46,7 @@ load_i1:
 
 define void @memcpy_fp80_padding() {
 ; CHECK-LABEL: @memcpy_fp80_padding(
-; CHECK-NEXT:    [[X_SROA_0:%.*]] = alloca x86_fp80, align 16
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 16 [[X_SROA_0]], ptr align 16 @foo_copy_source, i32 16, i1 false)
+; CHECK-NEXT:    [[X_SROA_0_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr @foo_copy_source, align 16
 ; CHECK-NEXT:    [[X_SROA_1_0_COPYLOAD:%.*]] = load i64, ptr getelementptr inbounds (i8, ptr @foo_copy_source, i64 16), align 16
 ; CHECK-NEXT:    [[X_SROA_2_0_COPYLOAD:%.*]] = load i64, ptr getelementptr inbounds (i8, ptr @foo_copy_source, i64 24), align 8
 ; CHECK-NEXT:    store i64 [[X_SROA_1_0_COPYLOAD]], ptr @i64_sink, align 4
@@ -67,8 +66,6 @@ define void @memcpy_fp80_padding() {
 
 define void @memset_fp80_padding() {
 ; CHECK-LABEL: @memset_fp80_padding(
-; CHECK-NEXT:    [[X_SROA_0:%.*]] = alloca x86_fp80, align 16
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 16 [[X_SROA_0]], i8 -1, i32 16, i1 false)
 ; CHECK-NEXT:    store i64 -1, ptr @i64_sink, align 4
 ; CHECK-NEXT:    ret void
 ;
@@ -136,8 +133,7 @@ define void @PR50888() {
 
 define void @PR50910() {
 ; CHECK-LABEL: @PR50910(
-; CHECK-NEXT:    [[T1:%.*]] = alloca i8, i64 1, align 8
-; CHECK-NEXT:    call void @llvm.memset.p0.i64(ptr align 8 [[T1]], i8 0, i64 1, i1 false)
+; CHECK-NEXT:    [[T1_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i8> undef, i8 0, i32 0
 ; CHECK-NEXT:    ret void
 ;
   %t1 = alloca i8, i64 1, align 8

diff  --git a/llvm/test/Transforms/SROA/sroa-common-type-fail-promotion.ll b/llvm/test/Transforms/SROA/sroa-common-type-fail-promotion.ll
index 00cbe56929c02..c4e68ff0799ea 100644
--- a/llvm/test/Transforms/SROA/sroa-common-type-fail-promotion.ll
+++ b/llvm/test/Transforms/SROA/sroa-common-type-fail-promotion.ll
@@ -43,11 +43,15 @@ define amdgpu_kernel void @test_memset() #0 {
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    [[DATA:%.*]] = load <4 x float>, ptr undef, align 16
 ; CHECK-NEXT:    [[TMP0:%.*]] = bitcast <4 x float> [[DATA]] to <8 x half>
+; CHECK-NEXT:    [[TMP1:%.*]] = bitcast <8 x half> [[TMP0]] to <8 x i16>
 ; CHECK-NEXT:    br label [[BB:%.*]]
 ; CHECK:       bb:
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_0_VEC_EXTRACT:%.*]] = extractelement <8 x half> [[TMP0]], i32 0
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_2_VEC_EXTRACT:%.*]] = extractelement <8 x half> [[TMP0]], i32 1
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_4_VEC_EXTRACT:%.*]] = extractelement <8 x half> [[TMP0]], i32 2
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_0_VEC_EXTRACT:%.*]] = extractelement <8 x i16> [[TMP1]], i32 0
+; CHECK-NEXT:    [[TMP2:%.*]] = bitcast i16 [[B_BLOCKWISE_COPY_SROA_0_SROA_0_0_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_2_VEC_EXTRACT:%.*]] = extractelement <8 x i16> [[TMP1]], i32 1
+; CHECK-NEXT:    [[TMP3:%.*]] = bitcast i16 [[B_BLOCKWISE_COPY_SROA_0_SROA_0_2_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_4_VEC_EXTRACT:%.*]] = extractelement <8 x i16> [[TMP1]], i32 2
+; CHECK-NEXT:    [[TMP4:%.*]] = bitcast i16 [[B_BLOCKWISE_COPY_SROA_0_SROA_0_4_VEC_EXTRACT]] to half
 ; CHECK-NEXT:    ret void
 ;
 entry:
@@ -235,25 +239,23 @@ bb:
 define amdgpu_kernel void @test_half_array() #0 {
 ; CHECK-LABEL: @test_half_array(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0:%.*]] = alloca float, align 16
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_4:%.*]] = alloca float, align 4
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 16 [[B_BLOCKWISE_COPY_SROA_0]], i8 0, i32 4, i1 false)
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 4 [[B_BLOCKWISE_COPY_SROA_4]], i8 0, i32 4, i1 false)
 ; CHECK-NEXT:    [[TMP0:%.*]] = bitcast float undef to i32
 ; CHECK-NEXT:    [[TMP1:%.*]] = bitcast float undef to i32
 ; CHECK-NEXT:    [[DATA:%.*]] = load [4 x float], ptr undef, align 4
 ; CHECK-NEXT:    [[DATA_FCA_0_EXTRACT:%.*]] = extractvalue [4 x float] [[DATA]], 0
-; CHECK-NEXT:    store float [[DATA_FCA_0_EXTRACT]], ptr [[B_BLOCKWISE_COPY_SROA_0]], align 16
+; CHECK-NEXT:    [[TMP2:%.*]] = bitcast float [[DATA_FCA_0_EXTRACT]] to <4 x i8>
 ; CHECK-NEXT:    [[DATA_FCA_1_EXTRACT:%.*]] = extractvalue [4 x float] [[DATA]], 1
-; CHECK-NEXT:    store float [[DATA_FCA_1_EXTRACT]], ptr [[B_BLOCKWISE_COPY_SROA_4]], align 4
+; CHECK-NEXT:    [[TMP3:%.*]] = bitcast float [[DATA_FCA_1_EXTRACT]] to <4 x i8>
 ; CHECK-NEXT:    [[DATA_FCA_2_EXTRACT:%.*]] = extractvalue [4 x float] [[DATA]], 2
 ; CHECK-NEXT:    [[DATA_FCA_3_EXTRACT:%.*]] = extractvalue [4 x float] [[DATA]], 3
 ; CHECK-NEXT:    br label [[BB:%.*]]
 ; CHECK:       bb:
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_0_B_BLOCKWISE_COPY_SROA_0_0_LOAD1:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_0]], align 16
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_2_PTR2_SROA_IDX1:%.*]] = getelementptr inbounds i8, ptr [[B_BLOCKWISE_COPY_SROA_0]], i64 2
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_2_B_BLOCKWISE_COPY_SROA_0_2_LOAD2:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_0_2_PTR2_SROA_IDX1]], align 2
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_4_0_B_BLOCKWISE_COPY_SROA_4_4_LOAD3:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_4]], align 4
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_0_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP2]], <4 x i8> poison, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[TMP4:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_0_SROA_0_0_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_2_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP2]], <4 x i8> poison, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[TMP5:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_0_SROA_0_2_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_4_SROA_0_0_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP3]], <4 x i8> poison, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[TMP6:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_4_SROA_0_0_VEC_EXTRACT]] to half
 ; CHECK-NEXT:    ret void
 ;
 entry:
@@ -275,15 +277,17 @@ bb:
 define amdgpu_kernel void @test_array_vector() #0 {
 ; CHECK-LABEL: @test_array_vector(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_5:%.*]] = alloca <8 x half>, align 16
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 16 [[B_BLOCKWISE_COPY_SROA_5]], i8 0, i32 16, i1 false)
 ; CHECK-NEXT:    [[DATA:%.*]] = load <4 x float>, ptr undef, align 16
 ; CHECK-NEXT:    [[TMP0:%.*]] = bitcast <4 x float> [[DATA]] to <8 x half>
+; CHECK-NEXT:    [[TMP1:%.*]] = bitcast <8 x half> [[TMP0]] to <8 x i16>
 ; CHECK-NEXT:    br label [[BB:%.*]]
 ; CHECK:       bb:
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_0_VEC_EXTRACT:%.*]] = extractelement <8 x half> [[TMP0]], i32 0
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_2_VEC_EXTRACT:%.*]] = extractelement <8 x half> [[TMP0]], i32 1
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_4_VEC_EXTRACT:%.*]] = extractelement <8 x half> [[TMP0]], i32 2
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_0_VEC_EXTRACT:%.*]] = extractelement <8 x i16> [[TMP1]], i32 0
+; CHECK-NEXT:    [[TMP2:%.*]] = bitcast i16 [[B_BLOCKWISE_COPY_SROA_0_SROA_0_0_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_2_VEC_EXTRACT:%.*]] = extractelement <8 x i16> [[TMP1]], i32 1
+; CHECK-NEXT:    [[TMP3:%.*]] = bitcast i16 [[B_BLOCKWISE_COPY_SROA_0_SROA_0_2_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_4_VEC_EXTRACT:%.*]] = extractelement <8 x i16> [[TMP1]], i32 2
+; CHECK-NEXT:    [[TMP4:%.*]] = bitcast i16 [[B_BLOCKWISE_COPY_SROA_0_SROA_0_4_VEC_EXTRACT]] to half
 ; CHECK-NEXT:    ret void
 ;
 entry:
@@ -305,15 +309,17 @@ bb:
 define amdgpu_kernel void @test_array_vector2() #0 {
 ; CHECK-LABEL: @test_array_vector2(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_5:%.*]] = alloca <8 x half>, align 16
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 16 [[B_BLOCKWISE_COPY_SROA_5]], i8 0, i32 16, i1 false)
 ; CHECK-NEXT:    [[DATA:%.*]] = load <4 x float>, ptr undef, align 16
 ; CHECK-NEXT:    [[TMP0:%.*]] = bitcast <4 x float> [[DATA]] to <8 x half>
+; CHECK-NEXT:    [[TMP1:%.*]] = bitcast <8 x half> [[TMP0]] to <8 x i16>
 ; CHECK-NEXT:    br label [[BB:%.*]]
 ; CHECK:       bb:
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_0_VEC_EXTRACT:%.*]] = extractelement <8 x half> [[TMP0]], i32 0
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_2_VEC_EXTRACT:%.*]] = extractelement <8 x half> [[TMP0]], i32 1
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_4_VEC_EXTRACT:%.*]] = extractelement <8 x half> [[TMP0]], i32 2
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_0_VEC_EXTRACT:%.*]] = extractelement <8 x i16> [[TMP1]], i32 0
+; CHECK-NEXT:    [[TMP2:%.*]] = bitcast i16 [[B_BLOCKWISE_COPY_SROA_0_SROA_0_0_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_2_VEC_EXTRACT:%.*]] = extractelement <8 x i16> [[TMP1]], i32 1
+; CHECK-NEXT:    [[TMP3:%.*]] = bitcast i16 [[B_BLOCKWISE_COPY_SROA_0_SROA_0_2_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_4_VEC_EXTRACT:%.*]] = extractelement <8 x i16> [[TMP1]], i32 2
+; CHECK-NEXT:    [[TMP4:%.*]] = bitcast i16 [[B_BLOCKWISE_COPY_SROA_0_SROA_0_4_VEC_EXTRACT]] to half
 ; CHECK-NEXT:    ret void
 ;
 entry:
@@ -335,38 +341,32 @@ bb:
 define amdgpu_kernel void @test_array_vector_no_vector_common_type() #0 {
 ; CHECK-LABEL: @test_array_vector_no_vector_common_type(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0:%.*]] = alloca float, align 16
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_4:%.*]] = alloca float, align 4
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_7:%.*]] = alloca float, align 8
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_10:%.*]] = alloca float, align 4
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_13:%.*]] = alloca <8 x half>, align 16
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 16 [[B_BLOCKWISE_COPY_SROA_0]], i8 0, i32 4, i1 false)
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 4 [[B_BLOCKWISE_COPY_SROA_4]], i8 0, i32 4, i1 false)
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 8 [[B_BLOCKWISE_COPY_SROA_7]], i8 0, i32 4, i1 false)
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 4 [[B_BLOCKWISE_COPY_SROA_10]], i8 0, i32 4, i1 false)
-; CHECK-NEXT:    call void @llvm.memset.p0.i32(ptr align 16 [[B_BLOCKWISE_COPY_SROA_13]], i8 0, i32 16, i1 false)
 ; CHECK-NEXT:    [[DATA1:%.*]] = load float, ptr undef, align 4
 ; CHECK-NEXT:    [[DATA2:%.*]] = load float, ptr undef, align 4
 ; CHECK-NEXT:    [[DATA3:%.*]] = load float, ptr undef, align 4
 ; CHECK-NEXT:    [[DATA4:%.*]] = load float, ptr undef, align 4
-; CHECK-NEXT:    store float [[DATA1]], ptr [[B_BLOCKWISE_COPY_SROA_0]], align 16
-; CHECK-NEXT:    store float [[DATA2]], ptr [[B_BLOCKWISE_COPY_SROA_4]], align 4
-; CHECK-NEXT:    store float [[DATA3]], ptr [[B_BLOCKWISE_COPY_SROA_7]], align 8
-; CHECK-NEXT:    store float [[DATA4]], ptr [[B_BLOCKWISE_COPY_SROA_10]], align 4
+; CHECK-NEXT:    [[TMP0:%.*]] = bitcast float [[DATA1]] to <4 x i8>
+; CHECK-NEXT:    [[TMP1:%.*]] = bitcast float [[DATA2]] to <4 x i8>
+; CHECK-NEXT:    [[TMP2:%.*]] = bitcast float [[DATA3]] to <4 x i8>
+; CHECK-NEXT:    [[TMP3:%.*]] = bitcast float [[DATA4]] to <4 x i8>
 ; CHECK-NEXT:    br label [[BB:%.*]]
 ; CHECK:       bb:
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_0_B_BLOCKWISE_COPY_SROA_0_0_LOAD1:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_0]], align 16
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_2_PTR2_SROA_IDX1:%.*]] = getelementptr inbounds i8, ptr [[B_BLOCKWISE_COPY_SROA_0]], i64 2
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_2_B_BLOCKWISE_COPY_SROA_0_2_LOAD2:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_0_2_PTR2_SROA_IDX1]], align 2
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_4_0_B_BLOCKWISE_COPY_SROA_4_4_LOAD3:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_4]], align 4
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_4_2_PTR4_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[B_BLOCKWISE_COPY_SROA_4]], i64 2
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_4_2_B_BLOCKWISE_COPY_SROA_4_6_LOAD4:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_4_2_PTR4_SROA_IDX]], align 2
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_7_0_B_BLOCKWISE_COPY_SROA_7_8_LOAD5:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_7]], align 8
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_7_2_PTR6_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[B_BLOCKWISE_COPY_SROA_7]], i64 2
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_7_2_B_BLOCKWISE_COPY_SROA_7_10_LOAD6:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_7_2_PTR6_SROA_IDX]], align 2
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_10_0_B_BLOCKWISE_COPY_SROA_10_12_LOAD7:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_10]], align 4
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_10_2_PTR8_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[B_BLOCKWISE_COPY_SROA_10]], i64 2
-; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_10_2_B_BLOCKWISE_COPY_SROA_10_14_LOAD8:%.*]] = load half, ptr [[B_BLOCKWISE_COPY_SROA_10_2_PTR8_SROA_IDX]], align 2
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_0_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP0]], <4 x i8> poison, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[TMP4:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_0_SROA_0_0_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_0_SROA_0_2_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP0]], <4 x i8> poison, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[TMP5:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_0_SROA_0_2_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_4_SROA_0_0_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP1]], <4 x i8> poison, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[TMP6:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_4_SROA_0_0_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_4_SROA_0_2_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP1]], <4 x i8> poison, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[TMP7:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_4_SROA_0_2_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_7_SROA_0_0_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP2]], <4 x i8> poison, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[TMP8:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_7_SROA_0_0_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_7_SROA_0_2_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP2]], <4 x i8> poison, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[TMP9:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_7_SROA_0_2_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_10_SROA_0_0_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP3]], <4 x i8> poison, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[TMP10:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_10_SROA_0_0_VEC_EXTRACT]] to half
+; CHECK-NEXT:    [[B_BLOCKWISE_COPY_SROA_10_SROA_0_2_VEC_EXTRACT:%.*]] = shufflevector <4 x i8> [[TMP3]], <4 x i8> poison, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[TMP11:%.*]] = bitcast <2 x i8> [[B_BLOCKWISE_COPY_SROA_10_SROA_0_2_VEC_EXTRACT]] to half
 ; CHECK-NEXT:    ret void
 ;
 entry:

diff  --git a/llvm/test/Transforms/SROA/tbaa-struct.ll b/llvm/test/Transforms/SROA/tbaa-struct.ll
index 3e9332c5b11c0..3d55b72e91a60 100644
--- a/llvm/test/Transforms/SROA/tbaa-struct.ll
+++ b/llvm/test/Transforms/SROA/tbaa-struct.ll
@@ -10,7 +10,8 @@ declare <2 x float> @foo(ptr %0)
 define void @bar(ptr %y2) {
 ; CHECK-LABEL: @bar(
 ; CHECK-NEXT:    [[X14:%.*]] = call <2 x float> @foo(ptr [[Y2:%.*]])
-; CHECK-NEXT:    store <2 x float> [[X14]], ptr [[Y2]], align 4, !tbaa.struct !0
+; CHECK-NEXT:    [[TMP1:%.*]] = bitcast <2 x float> [[X14]] to <2 x i32>
+; CHECK-NEXT:    store <2 x i32> [[TMP1]], ptr [[Y2]], align 4, !tbaa.struct !0
 ; CHECK-NEXT:    ret void
 ;
   %x7 = alloca %vector

diff  --git a/llvm/test/Transforms/SROA/tbaa-struct2.ll b/llvm/test/Transforms/SROA/tbaa-struct2.ll
index 1c81fc6163bbc..e7d5f4e74de52 100644
--- a/llvm/test/Transforms/SROA/tbaa-struct2.ll
+++ b/llvm/test/Transforms/SROA/tbaa-struct2.ll
@@ -9,12 +9,11 @@ declare double @subcall(double %g, i32 %m)
 
 define double @bar(ptr %wishart) {
 ; CHECK-LABEL: @bar(
-; CHECK-NEXT:    [[TMP_SROA_3:%.*]] = alloca [4 x i8], align 4
 ; CHECK-NEXT:    [[TMP_SROA_0_0_COPYLOAD:%.*]] = load double, ptr [[WISHART:%.*]], align 8, !tbaa.struct !0
 ; CHECK-NEXT:    [[TMP_SROA_2_0_WISHART_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[WISHART]], i64 8
 ; CHECK-NEXT:    [[TMP_SROA_2_0_COPYLOAD:%.*]] = load i32, ptr [[TMP_SROA_2_0_WISHART_SROA_IDX]], align 8, !tbaa.struct !7
 ; CHECK-NEXT:    [[TMP_SROA_3_0_WISHART_SROA_IDX:%.*]] = getelementptr inbounds i8, ptr [[WISHART]], i64 12
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i64(ptr align 4 [[TMP_SROA_3]], ptr align 4 [[TMP_SROA_3_0_WISHART_SROA_IDX]], i64 4, i1 false), !tbaa.struct !8
+; CHECK-NEXT:    [[TMP_SROA_3_SROA_0_0_COPYLOAD:%.*]] = load <4 x i8>, ptr [[TMP_SROA_3_0_WISHART_SROA_IDX]], align 4, !tbaa.struct !8
 ; CHECK-NEXT:    [[CALL:%.*]] = call double @subcall(double [[TMP_SROA_0_0_COPYLOAD]], i32 [[TMP_SROA_2_0_COPYLOAD]])
 ; CHECK-NEXT:    ret double [[CALL]]
 ;

diff  --git a/llvm/test/Transforms/SROA/vector-promotion.ll b/llvm/test/Transforms/SROA/vector-promotion.ll
index bdf50ea39af33..569dd05918332 100644
--- a/llvm/test/Transforms/SROA/vector-promotion.ll
+++ b/llvm/test/Transforms/SROA/vector-promotion.ll
@@ -567,9 +567,9 @@ define <4 x float> @test12(<4 x i32> %val) {
 
 define void @swap-8bytes(ptr %x, ptr %y) {
 ; CHECK-LABEL: @swap-8bytes(
-; CHECK-NEXT:    [[TMP_SROA_0_0_COPYLOAD:%.*]] = load i64, ptr [[X:%.*]], align 1
+; CHECK-NEXT:    [[TMP_SROA_0_0_COPYLOAD:%.*]] = load <8 x i8>, ptr [[X:%.*]], align 1
 ; CHECK-NEXT:    tail call void @llvm.memcpy.p0.p0.i64(ptr [[X]], ptr [[Y:%.*]], i64 8, i1 false)
-; CHECK-NEXT:    store i64 [[TMP_SROA_0_0_COPYLOAD]], ptr [[Y]], align 1
+; CHECK-NEXT:    store <8 x i8> [[TMP_SROA_0_0_COPYLOAD]], ptr [[Y]], align 1
 ; CHECK-NEXT:    ret void
 ;
   %tmp = alloca [2 x i32]
@@ -581,10 +581,9 @@ define void @swap-8bytes(ptr %x, ptr %y) {
 
 define void @swap-7bytes(ptr %x, ptr %y) {
 ; CHECK-LABEL: @swap-7bytes(
-; CHECK-NEXT:    [[TMP:%.*]] = alloca [7 x i8], align 1
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i64(ptr [[TMP]], ptr [[X:%.*]], i64 7, i1 false)
+; CHECK-NEXT:    [[TMP_SROA_0_0_COPYLOAD:%.*]] = load <7 x i8>, ptr [[X:%.*]], align 1
 ; CHECK-NEXT:    tail call void @llvm.memcpy.p0.p0.i64(ptr [[X]], ptr [[Y:%.*]], i64 7, i1 false)
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i64(ptr [[Y]], ptr [[TMP]], i64 7, i1 false)
+; CHECK-NEXT:    store <7 x i8> [[TMP_SROA_0_0_COPYLOAD]], ptr [[Y]], align 1
 ; CHECK-NEXT:    ret void
 ;
   %tmp = alloca [7 x i8]
@@ -596,10 +595,9 @@ define void @swap-7bytes(ptr %x, ptr %y) {
 
 define void @swap-16bytes(ptr %x, ptr %y) {
 ; CHECK-LABEL: @swap-16bytes(
-; CHECK-NEXT:    [[TMP:%.*]] = alloca [2 x i64], align 8
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i64(ptr [[TMP]], ptr [[X:%.*]], i64 16, i1 false)
+; CHECK-NEXT:    [[TMP_SROA_0_0_COPYLOAD:%.*]] = load <16 x i8>, ptr [[X:%.*]], align 1
 ; CHECK-NEXT:    tail call void @llvm.memcpy.p0.p0.i64(ptr [[X]], ptr [[Y:%.*]], i64 16, i1 false)
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i64(ptr [[Y]], ptr [[TMP]], i64 16, i1 false)
+; CHECK-NEXT:    store <16 x i8> [[TMP_SROA_0_0_COPYLOAD]], ptr [[Y]], align 1
 ; CHECK-NEXT:    ret void
 ;
   %tmp = alloca [2 x i64]
@@ -611,10 +609,9 @@ define void @swap-16bytes(ptr %x, ptr %y) {
 
 define void @swap-15bytes(ptr %x, ptr %y) {
 ; CHECK-LABEL: @swap-15bytes(
-; CHECK-NEXT:    [[TMP:%.*]] = alloca [15 x i8], align 1
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i64(ptr [[TMP]], ptr [[X:%.*]], i64 15, i1 false)
+; CHECK-NEXT:    [[TMP_SROA_0_0_COPYLOAD:%.*]] = load <15 x i8>, ptr [[X:%.*]], align 1
 ; CHECK-NEXT:    tail call void @llvm.memcpy.p0.p0.i64(ptr [[X]], ptr [[Y:%.*]], i64 15, i1 false)
-; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i64(ptr [[Y]], ptr [[TMP]], i64 15, i1 false)
+; CHECK-NEXT:    store <15 x i8> [[TMP_SROA_0_0_COPYLOAD]], ptr [[Y]], align 1
 ; CHECK-NEXT:    ret void
 ;
   %tmp = alloca [15 x i8]