[X86] Make memcmp vector lowering handle arbitrary expansions

[davezarzycki]

Summary

Teach combineVectorSizedSetCCEquality() to handle arbitrary memcmp expansions but do not change any default policy for now.

This also fixes a bug in the memcmp expansion itself when large displacements are needed.

[davezarzycki]

Updated the tests after running "check-all` with all targets enabled.

[davezarzycki]

Ping. This patch is essentially just bug fixes and a prerequisite change so that we might have a more focused conversation in D69044. Thanks!

[craig.topper]

LGTM

[davezarzycki]

Thanks @craig.topper for suggesting the GEP fix to the memcmp expansion code.

f68925d45083407e6bf322a7ee18d85c83c001ce

lib/CodeGen/ExpandMemCmp.cpp

 
 /// Return a pointer to an element of type `LoadSizeType` at offset
 /// `OffsetBytes`.
 Value *MemCmpExpansion::getPtrToElementAtOffset(Value *Source,
                                                 Type *LoadSizeType,
                                                 uint64_t OffsetBytes) {
   if (OffsetBytes > 0) {
     auto *ByteType = Type::getInt8Ty(CI->getContext());
-    Source = Builder.CreateGEP(
+    Source = Builder.CreateConstGEP1_64(
         ByteType, Builder.CreateBitCast(Source, ByteType->getPointerTo()),
-        ConstantInt::get(ByteType, OffsetBytes));
+        OffsetBytes);
   }
   return Builder.CreateBitCast(Source, LoadSizeType->getPointerTo());
 }
 
 // This function creates the IR instructions for loading and comparing 1 byte.
 // It loads 1 byte from each source of the memcmp parameters with the given
 // GEPIndex. It then subtracts the two loaded values and adds this result to the
 // final phi node for selecting the memcmp result.

lib/Target/X86/X86ISelLowering.cpp

if ((N00.isUndef() || DAG.MaskedValueIsZero(N00, ZeroMask)) &&
         (N01.isUndef() || DAG.MaskedValueIsZero(N01, ZeroMask))) {
       return concatSubVectors(N00, N01, DAG, dl);
     }
   }
 
   return SDValue();
 }
 
+/// Recursive helper for combineVectorSizedSetCCEquality() to see if we have a
+/// recognizable memcmp expansion.
+static bool isOrXorXorTree(SDValue X, bool Root = true) {
+  if (X.getOpcode() == ISD::OR)
+    return isOrXorXorTree(X.getOperand(0), false) &&
+           isOrXorXorTree(X.getOperand(1), false);
+  if (Root)
+    return false;
+  return X.getOpcode() == ISD::XOR;
+}
+
+/// Recursive helper for combineVectorSizedSetCCEquality() to emit the memcmp
+/// expansion.
+template<typename F>
+static SDValue emitOrXorXorTree(SDValue X, SDLoc &DL, SelectionDAG &DAG,
+                                EVT VecVT, EVT CmpVT, bool HasPT, F SToV) {
+  if (X.getOpcode() == ISD::OR) {
+    SDValue A = emitOrXorXorTree(X.getOperand(0), DL, DAG, VecVT, CmpVT, HasPT, SToV);
+    SDValue B = emitOrXorXorTree(X.getOperand(1), DL, DAG, VecVT, CmpVT, HasPT, SToV);
+    if (VecVT != CmpVT)
+      return DAG.getNode(ISD::OR, DL, CmpVT, A, B);
+    if (HasPT)
+      return DAG.getNode(ISD::OR, DL, VecVT, A, B);
+    return DAG.getNode(ISD::AND, DL, CmpVT, A, B);
+  } else if (X.getOpcode() == ISD::XOR) {
+    SDValue A = SToV(X.getOperand(0));
+    SDValue B = SToV(X.getOperand(1));
+    if (VecVT != CmpVT)
+      return DAG.getSetCC(DL, CmpVT, A, B, ISD::SETNE);
+    if (HasPT)
+      return DAG.getNode(ISD::XOR, DL, VecVT, A, B);
+    return DAG.getSetCC(DL, CmpVT, A, B, ISD::SETEQ);
+  }
+  llvm_unreachable("Impossible");
+}
+
 /// Try to map a 128-bit or larger integer comparison to vector instructions
 /// before type legalization splits it up into chunks.
 static SDValue combineVectorSizedSetCCEquality(SDNode *SetCC, SelectionDAG &DAG,
                                                const X86Subtarget &Subtarget) {
   ISD::CondCode CC = cast<CondCodeSDNode>(SetCC->getOperand(2))->get();
   assert((CC == ISD::SETNE || CC == ISD::SETEQ) && "Bad comparison predicate");
 
   // We're looking for an oversized integer equality comparison.
   SDValue X = SetCC->getOperand(0);
   SDValue Y = SetCC->getOperand(1);
   EVT OpVT = X.getValueType();
   unsigned OpSize = OpVT.getSizeInBits();
   if (!OpVT.isScalarInteger() || OpSize < 128)
     return SDValue();
 
   // Ignore a comparison with zero because that gets special treatment in
   // EmitTest(). But make an exception for the special case of a pair of
   // logically-combined vector-sized operands compared to zero. This pattern may
   // be generated by the memcmp expansion pass with oversized integer compares
   // (see PR33325).
-  bool IsOrXorXorCCZero = isNullConstant(Y) && X.getOpcode() == ISD::OR &&
-                          X.getOperand(0).getOpcode() == ISD::XOR &&
-                          X.getOperand(1).getOpcode() == ISD::XOR;
-  if (isNullConstant(Y) && !IsOrXorXorCCZero)
+  bool IsOrXorXorTreeCCZero = isNullConstant(Y) && isOrXorXorTree(X);
+  if (isNullConstant(Y) && !IsOrXorXorTreeCCZero)
     return SDValue();
 
   // Don't perform this combine if constructing the vector will be expensive.
   auto IsVectorBitCastCheap = [](SDValue X) {
     X = peekThroughBitcasts(X);
     return isa<ConstantSDNode>(X) || X.getValueType().isVector() ||
            X.getOpcode() == ISD::LOAD;
   };
   if ((!IsVectorBitCastCheap(X) || !IsVectorBitCastCheap(Y)) &&
-      !IsOrXorXorCCZero)
+      !IsOrXorXorTreeCCZero)
     return SDValue();
 
   EVT VT = SetCC->getValueType(0);
   SDLoc DL(SetCC);
   bool HasAVX = Subtarget.hasAVX();
 
   // Use XOR (plus OR) and PTEST after SSE4.1 for 128/256-bit operands.
   // Use PCMPNEQ (plus OR) and KORTEST for 512-bit operands.
auto ScalarToVector = [&](SDValue X) -> SDValue {
       const TargetLowering &TLI = DAG.getTargetLoweringInfo();
       MVT VecIdxVT = TLI.getVectorIdxTy(DAG.getDataLayout());
       return DAG.getNode(ISD::INSERT_SUBVECTOR, DL, VecVT,
                          DAG.getConstant(0, DL, VecVT), X,
                          DAG.getConstant(0, DL, VecIdxVT));
     };
 
     SDValue Cmp;
-    if (IsOrXorXorCCZero) {
+    if (IsOrXorXorTreeCCZero) {
       // This is a bitwise-combined equality comparison of 2 pairs of vectors:
       // setcc i128 (or (xor A, B), (xor C, D)), 0, eq|ne
       // Use 2 vector equality compares and 'and' the results before doing a
       // MOVMSK.
-      SDValue A = ScalarToVector(X.getOperand(0).getOperand(0));
-      SDValue B = ScalarToVector(X.getOperand(0).getOperand(1));
-      SDValue C = ScalarToVector(X.getOperand(1).getOperand(0));
-      SDValue D = ScalarToVector(X.getOperand(1).getOperand(1));
-      if (VecVT != CmpVT) {
-        SDValue Cmp1 = DAG.getSetCC(DL, CmpVT, A, B, ISD::SETNE);
-        SDValue Cmp2 = DAG.getSetCC(DL, CmpVT, C, D, ISD::SETNE);
-        Cmp = DAG.getNode(ISD::OR, DL, CmpVT, Cmp1, Cmp2);
-      } else if (HasPT) {
-        SDValue Cmp1 = DAG.getNode(ISD::XOR, DL, VecVT, A, B);
-        SDValue Cmp2 = DAG.getNode(ISD::XOR, DL, VecVT, C, D);
-        Cmp = DAG.getNode(ISD::OR, DL, VecVT, Cmp1, Cmp2);
-      } else {
-        SDValue Cmp1 = DAG.getSetCC(DL, CmpVT, A, B, ISD::SETEQ);
-        SDValue Cmp2 = DAG.getSetCC(DL, CmpVT, C, D, ISD::SETEQ);
-        Cmp = DAG.getNode(ISD::AND, DL, CmpVT, Cmp1, Cmp2);
-      }
+      Cmp = emitOrXorXorTree(X, DL, DAG, VecVT, CmpVT, HasPT, ScalarToVector);
     } else {
       SDValue VecX = ScalarToVector(X);
       SDValue VecY = ScalarToVector(Y);
       if (VecVT != CmpVT) {
         Cmp = DAG.getSetCC(DL, CmpVT, VecX, VecY, ISD::SETNE);
       } else if (HasPT) {
         Cmp = DAG.getNode(ISD::XOR, DL, VecVT, VecX, VecY);
       } else {

test/CodeGen/PowerPC/memcmpIR.ll

entry:
   ; CHECK-LABEL: res_block:{{.*}}
   ; CHECK: [[ICMP2:%[0-9]+]] = icmp ult i64
   ; CHECK-NEXT: [[SELECT:%[0-9]+]] = select i1 [[ICMP2]], i32 -1, i32 1
   ; CHECK-NEXT: br label %endblock
 
   ; CHECK-LABEL: loadbb1:{{.*}}
   ; CHECK: [[BCC1:%[0-9]+]] = bitcast i32* {{.*}} to i8*
   ; CHECK-NEXT: [[BCC2:%[0-9]+]] = bitcast i32* {{.*}} to i8*
-  ; CHECK-NEXT: [[GEP1:%[0-9]+]] = getelementptr i8, i8* [[BCC2]], i8 8
+  ; CHECK-NEXT: [[GEP1:%[0-9]+]] = getelementptr i8, i8* [[BCC2]], i64 8
   ; CHECK-NEXT: [[BCL1:%[0-9]+]] = bitcast i8* [[GEP1]] to i64*
-  ; CHECK-NEXT: [[GEP2:%[0-9]+]] = getelementptr i8, i8* [[BCC1]], i8 8
+  ; CHECK-NEXT: [[GEP2:%[0-9]+]] = getelementptr i8, i8* [[BCC1]], i64 8
   ; CHECK-NEXT: [[BCL2:%[0-9]+]] = bitcast i8* [[GEP2]] to i64*
   ; CHECK-NEXT: [[LOAD1:%[0-9]+]] = load i64, i64* [[BCL1]]
   ; CHECK-NEXT: [[LOAD2:%[0-9]+]] = load i64, i64* [[BCL2]]
   ; CHECK-NEXT: [[BSWAP1:%[0-9]+]] = call i64 @llvm.bswap.i64(i64 [[LOAD1]])
   ; CHECK-NEXT: [[BSWAP2:%[0-9]+]] = call i64 @llvm.bswap.i64(i64 [[LOAD2]])
   ; CHECK-NEXT: [[ICMP:%[0-9]+]] = icmp eq i64 [[BSWAP1]], [[BSWAP2]]
   ; CHECK-NEXT:  br i1 [[ICMP]], label %endblock, label %res_block
 
   ; CHECK-BE-LABEL: @test1(
   ; CHECK-BE: [[LOAD1:%[0-9]+]] = load i64, i64*
   ; CHECK-BE-NEXT: [[LOAD2:%[0-9]+]] = load i64, i64*
   ; CHECK-BE-NEXT: [[ICMP:%[0-9]+]] = icmp eq i64 [[LOAD1]], [[LOAD2]]
   ; CHECK-BE-NEXT:  br i1 [[ICMP]], label %loadbb1, label %res_block
 
   ; CHECK-BE-LABEL: res_block:{{.*}}
   ; CHECK-BE: [[ICMP2:%[0-9]+]] = icmp ult i64
   ; CHECK-BE-NEXT: [[SELECT:%[0-9]+]] = select i1 [[ICMP2]], i32 -1, i32 1
   ; CHECK-BE-NEXT: br label %endblock
 
   ; CHECK-BE-LABEL: loadbb1:{{.*}}
   ; CHECK-BE: [[BCC1:%[0-9]+]] = bitcast i32* {{.*}} to i8*
   ; CHECK-BE-NEXT: [[BCC2:%[0-9]+]] = bitcast i32* {{.*}} to i8*
-  ; CHECK-BE-NEXT: [[GEP1:%[0-9]+]] = getelementptr i8, i8* [[BCC2]], i8 8
+  ; CHECK-BE-NEXT: [[GEP1:%[0-9]+]] = getelementptr i8, i8* [[BCC2]], i64 8
   ; CHECK-BE-NEXT: [[BCL1:%[0-9]+]] = bitcast i8* [[GEP1]] to i64*
-  ; CHECK-BE-NEXT: [[GEP2:%[0-9]+]] = getelementptr i8, i8* [[BCC1]], i8 8
+  ; CHECK-BE-NEXT: [[GEP2:%[0-9]+]] = getelementptr i8, i8* [[BCC1]], i64 8
   ; CHECK-BE-NEXT: [[BCL2:%[0-9]+]] = bitcast i8* [[GEP2]] to i64*
   ; CHECK-BE-NEXT: [[LOAD1:%[0-9]+]] = load i64, i64* [[BCL1]]
   ; CHECK-BE-NEXT: [[LOAD2:%[0-9]+]] = load i64, i64* [[BCL2]]
   ; CHECK-BE-NEXT: [[ICMP:%[0-9]+]] = icmp eq i64 [[LOAD1]], [[LOAD2]]
   ; CHECK-BE-NEXT:  br i1 [[ICMP]], label %endblock, label %res_block
 
   %0 = bitcast i32* %buffer1 to i8*
   %1 = bitcast i32* %buffer2 to i8*

test/CodeGen/X86/memcmp-more-load-pairs.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; NOTE: This is a copy of llvm/test/CodeGen/X86/memcmp.ll with more load pairs. Please keep it that way.
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=i686-unknown-unknown -mattr=cmov     | FileCheck %s --check-prefixes=X86,X86-NOSSE
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=i686-unknown-unknown -mattr=+sse     | FileCheck %s --check-prefixes=X86,X86-SSE,X86-SSE1
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=i686-unknown-unknown -mattr=+sse2    | FileCheck %s --check-prefixes=X86,X86-SSE,X86-SSE2
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=i686-unknown-unknown -mattr=+sse4.1  | FileCheck %s --check-prefixes=X86,X86-SSE,X86-SSE41
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=x86_64-unknown-unknown               | FileCheck %s --check-prefixes=X64,X64-SSE,X64-SSE2
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=x86_64-unknown-unknown -mattr=sse4.1 | FileCheck %s --check-prefixes=X64,X64-SSE,X64-SSE41
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=x86_64-unknown-unknown -mattr=avx    | FileCheck %s --check-prefixes=X64,X64-AVX,X64-AVX1
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=x86_64-unknown-unknown -mattr=avx2   | FileCheck %s --check-prefixes=X64,X64-AVX,X64-AVX2
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=x86_64-unknown-unknown -mattr=avx512bw,+prefer-256-bit | FileCheck %s --check-prefixes=X64,X64-AVX,X64-AVX2
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=x86_64-unknown-unknown -mattr=avx512bw,-prefer-256-bit | FileCheck %s --check-prefixes=X64,X64-AVX,X64-AVX512,X64-AVX512BW
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=x86_64-unknown-unknown -mattr=avx512f,+prefer-256-bit,-prefer-mask-registers | FileCheck %s --check-prefixes=X64,X64-AVX,X64-AVX2
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=x86_64-unknown-unknown -mattr=avx512f,-prefer-256-bit,-prefer-mask-registers | FileCheck %s --check-prefixes=X64,X64-AVX,X64-AVX512,X64-AVX512F
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=x86_64-unknown-unknown -mattr=avx512f,+prefer-256-bit,+prefer-mask-registers | FileCheck %s --check-prefixes=X64,X64-MIC-AVX,X64-MIC-AVX2
+; RUN: llc -max-loads-per-memcmp=4 -memcmp-num-loads-per-block=4 < %s -mtriple=x86_64-unknown-unknown -mattr=avx512f,-prefer-256-bit,+prefer-mask-registers | FileCheck %s --check-prefixes=X64,X64-MIC-AVX,X64-MIC-AVX512F
+
+; This tests codegen time inlining/optimization of memcmp
+; rdar://6480398
+
+@.str = private constant [513 x i8] c"01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901\00", align 1
+
+declare i32 @memcmp(i8*, i8*, i64)
+
+define i32 @length0(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length0:
+; X86:       # %bb.0:
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length0:
+; X64:       # %bb.0:
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    retq
+   %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 0) nounwind
+   ret i32 %m
+ }
+
+define i1 @length0_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length0_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    movb $1, %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length0_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movb $1, %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 0) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length0_lt(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length0_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length0_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 0) nounwind
+  %c = icmp slt i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length2(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length2:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movzwl (%ecx), %ecx
+; X86-NEXT:    movzwl (%eax), %edx
+; X86-NEXT:    rolw $8, %cx
+; X86-NEXT:    rolw $8, %dx
+; X86-NEXT:    movzwl %cx, %eax
+; X86-NEXT:    movzwl %dx, %ecx
+; X86-NEXT:    subl %ecx, %eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length2:
+; X64:       # %bb.0:
+; X64-NEXT:    movzwl (%rdi), %eax
+; X64-NEXT:    movzwl (%rsi), %ecx
+; X64-NEXT:    rolw $8, %ax
+; X64-NEXT:    rolw $8, %cx
+; X64-NEXT:    movzwl %ax, %eax
+; X64-NEXT:    movzwl %cx, %ecx
+; X64-NEXT:    subl %ecx, %eax
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 2) nounwind
+  ret i32 %m
+}
+
+define i1 @length2_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length2_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movzwl (%ecx), %ecx
+; X86-NEXT:    cmpw (%eax), %cx
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length2_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movzwl (%rdi), %eax
+; X64-NEXT:    cmpw (%rsi), %ax
+; X64-NEXT:    sete %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 2) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length2_lt(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length2_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movzwl (%ecx), %ecx
+; X86-NEXT:    movzwl (%eax), %edx
+; X86-NEXT:    rolw $8, %cx
+; X86-NEXT:    rolw $8, %dx
+; X86-NEXT:    movzwl %cx, %eax
+; X86-NEXT:    movzwl %dx, %ecx
+; X86-NEXT:    subl %ecx, %eax
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length2_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    movzwl (%rdi), %eax
+; X64-NEXT:    movzwl (%rsi), %ecx
+; X64-NEXT:    rolw $8, %ax
+; X64-NEXT:    rolw $8, %cx
+; X64-NEXT:    movzwl %ax, %eax
+; X64-NEXT:    movzwl %cx, %ecx
+; X64-NEXT:    subl %ecx, %eax
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 2) nounwind
+  %c = icmp slt i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length2_gt(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length2_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movzwl (%ecx), %ecx
+; X86-NEXT:    movzwl (%eax), %eax
+; X86-NEXT:    rolw $8, %cx
+; X86-NEXT:    rolw $8, %ax
+; X86-NEXT:    movzwl %cx, %ecx
+; X86-NEXT:    movzwl %ax, %eax
+; X86-NEXT:    subl %eax, %ecx
+; X86-NEXT:    testl %ecx, %ecx
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length2_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    movzwl (%rdi), %eax
+; X64-NEXT:    movzwl (%rsi), %ecx
+; X64-NEXT:    rolw $8, %ax
+; X64-NEXT:    rolw $8, %cx
+; X64-NEXT:    movzwl %ax, %eax
+; X64-NEXT:    movzwl %cx, %ecx
+; X64-NEXT:    subl %ecx, %eax
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 2) nounwind
+  %c = icmp sgt i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length2_eq_const(i8* %X) nounwind {
+; X86-LABEL: length2_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movzwl (%eax), %eax
+; X86-NEXT:    cmpl $12849, %eax # imm = 0x3231
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length2_eq_const:
+; X64:       # %bb.0:
+; X64-NEXT:    movzwl (%rdi), %eax
+; X64-NEXT:    cmpl $12849, %eax # imm = 0x3231
+; X64-NEXT:    setne %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 1), i64 2) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length2_eq_nobuiltin_attr(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length2_eq_nobuiltin_attr:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $2
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length2_eq_nobuiltin_attr:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $2, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    sete %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 2) nounwind nobuiltin
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length3(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length3:
+; X86:       # %bb.0: # %loadbb
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movzwl (%eax), %edx
+; X86-NEXT:    movzwl (%ecx), %esi
+; X86-NEXT:    rolw $8, %dx
+; X86-NEXT:    rolw $8, %si
+; X86-NEXT:    cmpw %si, %dx
+; X86-NEXT:    jne .LBB9_1
+; X86-NEXT:  # %bb.2: # %loadbb1
+; X86-NEXT:    movzbl 2(%eax), %eax
+; X86-NEXT:    movzbl 2(%ecx), %ecx
+; X86-NEXT:    subl %ecx, %eax
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+; X86-NEXT:  .LBB9_1: # %res_block
+; X86-NEXT:    setae %al
+; X86-NEXT:    movzbl %al, %eax
+; X86-NEXT:    leal -1(%eax,%eax), %eax
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length3:
+; X64:       # %bb.0: # %loadbb
+; X64-NEXT:    movzwl (%rdi), %eax
+; X64-NEXT:    movzwl (%rsi), %ecx
+; X64-NEXT:    rolw $8, %ax
+; X64-NEXT:    rolw $8, %cx
+; X64-NEXT:    cmpw %cx, %ax
+; X64-NEXT:    jne .LBB9_1
+; X64-NEXT:  # %bb.2: # %loadbb1
+; X64-NEXT:    movzbl 2(%rdi), %eax
+; X64-NEXT:    movzbl 2(%rsi), %ecx
+; X64-NEXT:    subl %ecx, %eax
+; X64-NEXT:    retq
+; X64-NEXT:  .LBB9_1: # %res_block
+; X64-NEXT:    setae %al
+; X64-NEXT:    movzbl %al, %eax
+; X64-NEXT:    leal -1(%rax,%rax), %eax
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 3) nounwind
+  ret i32 %m
+}
+
+define i1 @length3_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length3_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movzwl (%ecx), %edx
+; X86-NEXT:    xorw (%eax), %dx
+; X86-NEXT:    movb 2(%ecx), %cl
+; X86-NEXT:    xorb 2(%eax), %cl
+; X86-NEXT:    movzbl %cl, %eax
+; X86-NEXT:    orw %dx, %ax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length3_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movzwl (%rdi), %eax
+; X64-NEXT:    xorw (%rsi), %ax
+; X64-NEXT:    movb 2(%rdi), %cl
+; X64-NEXT:    xorb 2(%rsi), %cl
+; X64-NEXT:    movzbl %cl, %ecx
+; X64-NEXT:    orw %ax, %cx
+; X64-NEXT:    setne %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 3) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length4(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length4:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %ecx
+; X86-NEXT:    movl (%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    seta %al
+; X86-NEXT:    sbbl $0, %eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length4:
+; X64:       # %bb.0:
+; X64-NEXT:    movl (%rdi), %ecx
+; X64-NEXT:    movl (%rsi), %edx
+; X64-NEXT:    bswapl %ecx
+; X64-NEXT:    bswapl %edx
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpl %edx, %ecx
+; X64-NEXT:    seta %al
+; X64-NEXT:    sbbl $0, %eax
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 4) nounwind
+  ret i32 %m
+}
+
+define i1 @length4_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length4_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %ecx
+; X86-NEXT:    cmpl (%eax), %ecx
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length4_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movl (%rdi), %eax
+; X64-NEXT:    cmpl (%rsi), %eax
+; X64-NEXT:    setne %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 4) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length4_lt(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length4_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %ecx
+; X86-NEXT:    movl (%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    seta %al
+; X86-NEXT:    sbbl $0, %eax
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length4_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    movl (%rdi), %ecx
+; X64-NEXT:    movl (%rsi), %edx
+; X64-NEXT:    bswapl %ecx
+; X64-NEXT:    bswapl %edx
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpl %edx, %ecx
+; X64-NEXT:    seta %al
+; X64-NEXT:    sbbl $0, %eax
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 4) nounwind
+  %c = icmp slt i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length4_gt(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length4_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %ecx
+; X86-NEXT:    movl (%eax), %eax
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %eax
+; X86-NEXT:    xorl %edx, %edx
+; X86-NEXT:    cmpl %eax, %ecx
+; X86-NEXT:    seta %dl
+; X86-NEXT:    sbbl $0, %edx
+; X86-NEXT:    testl %edx, %edx
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length4_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    movl (%rdi), %eax
+; X64-NEXT:    movl (%rsi), %ecx
+; X64-NEXT:    bswapl %eax
+; X64-NEXT:    bswapl %ecx
+; X64-NEXT:    xorl %edx, %edx
+; X64-NEXT:    cmpl %ecx, %eax
+; X64-NEXT:    seta %dl
+; X64-NEXT:    sbbl $0, %edx
+; X64-NEXT:    testl %edx, %edx
+; X64-NEXT:    setg %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 4) nounwind
+  %c = icmp sgt i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length4_eq_const(i8* %X) nounwind {
+; X86-LABEL: length4_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    cmpl $875770417, (%eax) # imm = 0x34333231
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length4_eq_const:
+; X64:       # %bb.0:
+; X64-NEXT:    cmpl $875770417, (%rdi) # imm = 0x34333231
+; X64-NEXT:    sete %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 1), i64 4) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length5(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length5:
+; X86:       # %bb.0: # %loadbb
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl (%eax), %edx
+; X86-NEXT:    movl (%ecx), %esi
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    bswapl %esi
+; X86-NEXT:    cmpl %esi, %edx
+; X86-NEXT:    jne .LBB16_1
+; X86-NEXT:  # %bb.2: # %loadbb1
+; X86-NEXT:    movzbl 4(%eax), %eax
+; X86-NEXT:    movzbl 4(%ecx), %ecx
+; X86-NEXT:    subl %ecx, %eax
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+; X86-NEXT:  .LBB16_1: # %res_block
+; X86-NEXT:    setae %al
+; X86-NEXT:    movzbl %al, %eax
+; X86-NEXT:    leal -1(%eax,%eax), %eax
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length5:
+; X64:       # %bb.0: # %loadbb
+; X64-NEXT:    movl (%rdi), %eax
+; X64-NEXT:    movl (%rsi), %ecx
+; X64-NEXT:    bswapl %eax
+; X64-NEXT:    bswapl %ecx
+; X64-NEXT:    cmpl %ecx, %eax
+; X64-NEXT:    jne .LBB16_1
+; X64-NEXT:  # %bb.2: # %loadbb1
+; X64-NEXT:    movzbl 4(%rdi), %eax
+; X64-NEXT:    movzbl 4(%rsi), %ecx
+; X64-NEXT:    subl %ecx, %eax
+; X64-NEXT:    retq
+; X64-NEXT:  .LBB16_1: # %res_block
+; X64-NEXT:    setae %al
+; X64-NEXT:    movzbl %al, %eax
+; X64-NEXT:    leal -1(%rax,%rax), %eax
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 5) nounwind
+  ret i32 %m
+}
+
+define i1 @length5_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length5_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %edx
+; X86-NEXT:    xorl (%eax), %edx
+; X86-NEXT:    movb 4(%ecx), %cl
+; X86-NEXT:    xorb 4(%eax), %cl
+; X86-NEXT:    movzbl %cl, %eax
+; X86-NEXT:    orl %edx, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length5_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movl (%rdi), %eax
+; X64-NEXT:    xorl (%rsi), %eax
+; X64-NEXT:    movb 4(%rdi), %cl
+; X64-NEXT:    xorb 4(%rsi), %cl
+; X64-NEXT:    movzbl %cl, %ecx
+; X64-NEXT:    orl %eax, %ecx
+; X64-NEXT:    setne %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 5) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length5_lt(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length5_lt:
+; X86:       # %bb.0: # %loadbb
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl (%eax), %edx
+; X86-NEXT:    movl (%ecx), %esi
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    bswapl %esi
+; X86-NEXT:    cmpl %esi, %edx
+; X86-NEXT:    jne .LBB18_1
+; X86-NEXT:  # %bb.2: # %loadbb1
+; X86-NEXT:    movzbl 4(%eax), %eax
+; X86-NEXT:    movzbl 4(%ecx), %ecx
+; X86-NEXT:    subl %ecx, %eax
+; X86-NEXT:    jmp .LBB18_3
+; X86-NEXT:  .LBB18_1: # %res_block
+; X86-NEXT:    setae %al
+; X86-NEXT:    movzbl %al, %eax
+; X86-NEXT:    leal -1(%eax,%eax), %eax
+; X86-NEXT:  .LBB18_3: # %endblock
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length5_lt:
+; X64:       # %bb.0: # %loadbb
+; X64-NEXT:    movl (%rdi), %eax
+; X64-NEXT:    movl (%rsi), %ecx
+; X64-NEXT:    bswapl %eax
+; X64-NEXT:    bswapl %ecx
+; X64-NEXT:    cmpl %ecx, %eax
+; X64-NEXT:    jne .LBB18_1
+; X64-NEXT:  # %bb.2: # %loadbb1
+; X64-NEXT:    movzbl 4(%rdi), %eax
+; X64-NEXT:    movzbl 4(%rsi), %ecx
+; X64-NEXT:    subl %ecx, %eax
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    retq
+; X64-NEXT:  .LBB18_1: # %res_block
+; X64-NEXT:    setae %al
+; X64-NEXT:    movzbl %al, %eax
+; X64-NEXT:    leal -1(%rax,%rax), %eax
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 5) nounwind
+  %c = icmp slt i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length7_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length7_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %edx
+; X86-NEXT:    movl 3(%ecx), %ecx
+; X86-NEXT:    xorl (%eax), %edx
+; X86-NEXT:    xorl 3(%eax), %ecx
+; X86-NEXT:    orl %edx, %ecx
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length7_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movl (%rdi), %eax
+; X64-NEXT:    movl 3(%rdi), %ecx
+; X64-NEXT:    xorl (%rsi), %eax
+; X64-NEXT:    xorl 3(%rsi), %ecx
+; X64-NEXT:    orl %eax, %ecx
+; X64-NEXT:    setne %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 7) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length8(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length8:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %esi
+; X86-NEXT:    movl (%esi), %ecx
+; X86-NEXT:    movl (%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    jne .LBB20_2
+; X86-NEXT:  # %bb.1: # %loadbb1
+; X86-NEXT:    movl 4(%esi), %ecx
+; X86-NEXT:    movl 4(%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    je .LBB20_3
+; X86-NEXT:  .LBB20_2: # %res_block
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    setae %al
+; X86-NEXT:    leal -1(%eax,%eax), %eax
+; X86-NEXT:  .LBB20_3: # %endblock
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length8:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rcx
+; X64-NEXT:    movq (%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    seta %al
+; X64-NEXT:    sbbl $0, %eax
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 8) nounwind
+  ret i32 %m
+}
+
+define i1 @length8_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length8_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %edx
+; X86-NEXT:    movl 4(%ecx), %ecx
+; X86-NEXT:    xorl (%eax), %edx
+; X86-NEXT:    xorl 4(%eax), %ecx
+; X86-NEXT:    orl %edx, %ecx
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length8_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    cmpq (%rsi), %rax
+; X64-NEXT:    sete %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 8) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length8_eq_const(i8* %X) nounwind {
+; X86-LABEL: length8_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl $858927408, %ecx # imm = 0x33323130
+; X86-NEXT:    xorl (%eax), %ecx
+; X86-NEXT:    movl $926299444, %edx # imm = 0x37363534
+; X86-NEXT:    xorl 4(%eax), %edx
+; X86-NEXT:    orl %ecx, %edx
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length8_eq_const:
+; X64:       # %bb.0:
+; X64-NEXT:    movabsq $3978425819141910832, %rax # imm = 0x3736353433323130
+; X64-NEXT:    cmpq %rax, (%rdi)
+; X64-NEXT:    setne %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 8) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length9_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length9_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %edx
+; X86-NEXT:    movl 4(%ecx), %esi
+; X86-NEXT:    xorl (%eax), %edx
+; X86-NEXT:    xorl 4(%eax), %esi
+; X86-NEXT:    orl %edx, %esi
+; X86-NEXT:    movb 8(%ecx), %cl
+; X86-NEXT:    xorb 8(%eax), %cl
+; X86-NEXT:    movzbl %cl, %eax
+; X86-NEXT:    orl %esi, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length9_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    xorq (%rsi), %rax
+; X64-NEXT:    movb 8(%rdi), %cl
+; X64-NEXT:    xorb 8(%rsi), %cl
+; X64-NEXT:    movzbl %cl, %ecx
+; X64-NEXT:    orq %rax, %rcx
+; X64-NEXT:    sete %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 9) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length10_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length10_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %edx
+; X86-NEXT:    movl 4(%ecx), %esi
+; X86-NEXT:    xorl (%eax), %edx
+; X86-NEXT:    xorl 4(%eax), %esi
+; X86-NEXT:    orl %edx, %esi
+; X86-NEXT:    movzwl 8(%ecx), %ecx
+; X86-NEXT:    xorw 8(%eax), %cx
+; X86-NEXT:    movzwl %cx, %eax
+; X86-NEXT:    orl %esi, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length10_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    xorq (%rsi), %rax
+; X64-NEXT:    movzwl 8(%rdi), %ecx
+; X64-NEXT:    xorw 8(%rsi), %cx
+; X64-NEXT:    movzwl %cx, %ecx
+; X64-NEXT:    orq %rax, %rcx
+; X64-NEXT:    sete %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 10) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length11_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length11_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %edx
+; X86-NEXT:    movl 4(%ecx), %esi
+; X86-NEXT:    xorl (%eax), %edx
+; X86-NEXT:    xorl 4(%eax), %esi
+; X86-NEXT:    orl %edx, %esi
+; X86-NEXT:    movl 7(%ecx), %ecx
+; X86-NEXT:    xorl 7(%eax), %ecx
+; X86-NEXT:    orl %esi, %ecx
+; X86-NEXT:    sete %al
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length11_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    movq 3(%rdi), %rcx
+; X64-NEXT:    xorq (%rsi), %rax
+; X64-NEXT:    xorq 3(%rsi), %rcx
+; X64-NEXT:    orq %rax, %rcx
+; X64-NEXT:    sete %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 11) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length12_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length12_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %edx
+; X86-NEXT:    movl 4(%ecx), %esi
+; X86-NEXT:    xorl (%eax), %edx
+; X86-NEXT:    xorl 4(%eax), %esi
+; X86-NEXT:    orl %edx, %esi
+; X86-NEXT:    movl 8(%ecx), %ecx
+; X86-NEXT:    xorl 8(%eax), %ecx
+; X86-NEXT:    orl %esi, %ecx
+; X86-NEXT:    setne %al
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length12_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    xorq (%rsi), %rax
+; X64-NEXT:    movl 8(%rdi), %ecx
+; X64-NEXT:    xorl 8(%rsi), %ecx
+; X64-NEXT:    orq %rax, %rcx
+; X64-NEXT:    setne %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 12) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length12(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length12:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %esi
+; X86-NEXT:    movl (%esi), %ecx
+; X86-NEXT:    movl (%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    jne .LBB27_3
+; X86-NEXT:  # %bb.1: # %loadbb1
+; X86-NEXT:    movl 4(%esi), %ecx
+; X86-NEXT:    movl 4(%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    jne .LBB27_3
+; X86-NEXT:  # %bb.2: # %loadbb2
+; X86-NEXT:    movl 8(%esi), %ecx
+; X86-NEXT:    movl 8(%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    je .LBB27_4
+; X86-NEXT:  .LBB27_3: # %res_block
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    setae %al
+; X86-NEXT:    leal -1(%eax,%eax), %eax
+; X86-NEXT:  .LBB27_4: # %endblock
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length12:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rcx
+; X64-NEXT:    movq (%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB27_2
+; X64-NEXT:  # %bb.1: # %loadbb1
+; X64-NEXT:    movl 8(%rdi), %ecx
+; X64-NEXT:    movl 8(%rsi), %edx
+; X64-NEXT:    bswapl %ecx
+; X64-NEXT:    bswapl %edx
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    je .LBB27_3
+; X64-NEXT:  .LBB27_2: # %res_block
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    setae %al
+; X64-NEXT:    leal -1(%rax,%rax), %eax
+; X64-NEXT:  .LBB27_3: # %endblock
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 12) nounwind
+  ret i32 %m
+}
+
+define i1 @length13_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length13_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %edx
+; X86-NEXT:    movl 4(%ecx), %esi
+; X86-NEXT:    xorl (%eax), %edx
+; X86-NEXT:    xorl 4(%eax), %esi
+; X86-NEXT:    orl %edx, %esi
+; X86-NEXT:    movl 8(%ecx), %edx
+; X86-NEXT:    xorl 8(%eax), %edx
+; X86-NEXT:    movb 12(%ecx), %cl
+; X86-NEXT:    xorb 12(%eax), %cl
+; X86-NEXT:    movzbl %cl, %eax
+; X86-NEXT:    orl %edx, %eax
+; X86-NEXT:    orl %esi, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length13_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    movq 5(%rdi), %rcx
+; X64-NEXT:    xorq (%rsi), %rax
+; X64-NEXT:    xorq 5(%rsi), %rcx
+; X64-NEXT:    orq %rax, %rcx
+; X64-NEXT:    sete %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 13) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length14_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length14_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %edx
+; X86-NEXT:    movl 4(%ecx), %esi
+; X86-NEXT:    xorl (%eax), %edx
+; X86-NEXT:    xorl 4(%eax), %esi
+; X86-NEXT:    orl %edx, %esi
+; X86-NEXT:    movl 8(%ecx), %edx
+; X86-NEXT:    xorl 8(%eax), %edx
+; X86-NEXT:    movzwl 12(%ecx), %ecx
+; X86-NEXT:    xorw 12(%eax), %cx
+; X86-NEXT:    movzwl %cx, %eax
+; X86-NEXT:    orl %edx, %eax
+; X86-NEXT:    orl %esi, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length14_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    movq 6(%rdi), %rcx
+; X64-NEXT:    xorq (%rsi), %rax
+; X64-NEXT:    xorq 6(%rsi), %rcx
+; X64-NEXT:    orq %rax, %rcx
+; X64-NEXT:    sete %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 14) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i1 @length15_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length15_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl (%ecx), %edx
+; X86-NEXT:    movl 4(%ecx), %esi
+; X86-NEXT:    xorl (%eax), %edx
+; X86-NEXT:    xorl 4(%eax), %esi
+; X86-NEXT:    orl %edx, %esi
+; X86-NEXT:    movl 8(%ecx), %edx
+; X86-NEXT:    xorl 8(%eax), %edx
+; X86-NEXT:    movl 11(%ecx), %ecx
+; X86-NEXT:    xorl 11(%eax), %ecx
+; X86-NEXT:    orl %edx, %ecx
+; X86-NEXT:    orl %esi, %ecx
+; X86-NEXT:    sete %al
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length15_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    movq 7(%rdi), %rcx
+; X64-NEXT:    xorq (%rsi), %rax
+; X64-NEXT:    xorq 7(%rsi), %rcx
+; X64-NEXT:    orq %rax, %rcx
+; X64-NEXT:    sete %al
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 15) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+; PR33329 - https://bugs.llvm.org/show_bug.cgi?id=33329
+
+define i32 @length16(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length16:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %esi
+; X86-NEXT:    movl (%esi), %ecx
+; X86-NEXT:    movl (%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    jne .LBB31_4
+; X86-NEXT:  # %bb.1: # %loadbb1
+; X86-NEXT:    movl 4(%esi), %ecx
+; X86-NEXT:    movl 4(%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    jne .LBB31_4
+; X86-NEXT:  # %bb.2: # %loadbb2
+; X86-NEXT:    movl 8(%esi), %ecx
+; X86-NEXT:    movl 8(%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    jne .LBB31_4
+; X86-NEXT:  # %bb.3: # %loadbb3
+; X86-NEXT:    movl 12(%esi), %ecx
+; X86-NEXT:    movl 12(%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    je .LBB31_5
+; X86-NEXT:  .LBB31_4: # %res_block
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    setae %al
+; X86-NEXT:    leal -1(%eax,%eax), %eax
+; X86-NEXT:  .LBB31_5: # %endblock
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length16:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rcx
+; X64-NEXT:    movq (%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB31_2
+; X64-NEXT:  # %bb.1: # %loadbb1
+; X64-NEXT:    movq 8(%rdi), %rcx
+; X64-NEXT:    movq 8(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    je .LBB31_3
+; X64-NEXT:  .LBB31_2: # %res_block
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    setae %al
+; X64-NEXT:    leal -1(%rax,%rax), %eax
+; X64-NEXT:  .LBB31_3: # %endblock
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 16) nounwind
+  ret i32 %m
+}
+
+define i1 @length16_eq(i8* %x, i8* %y) nounwind {
+; X86-NOSSE-LABEL: length16_eq:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl %esi
+; X86-NOSSE-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NOSSE-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NOSSE-NEXT:    movl (%ecx), %edx
+; X86-NOSSE-NEXT:    movl 4(%ecx), %esi
+; X86-NOSSE-NEXT:    xorl (%eax), %edx
+; X86-NOSSE-NEXT:    xorl 4(%eax), %esi
+; X86-NOSSE-NEXT:    orl %edx, %esi
+; X86-NOSSE-NEXT:    movl 8(%ecx), %edx
+; X86-NOSSE-NEXT:    xorl 8(%eax), %edx
+; X86-NOSSE-NEXT:    movl 12(%ecx), %ecx
+; X86-NOSSE-NEXT:    xorl 12(%eax), %ecx
+; X86-NOSSE-NEXT:    orl %edx, %ecx
+; X86-NOSSE-NEXT:    orl %esi, %ecx
+; X86-NOSSE-NEXT:    setne %al
+; X86-NOSSE-NEXT:    popl %esi
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length16_eq:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl %esi
+; X86-SSE1-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE1-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE1-NEXT:    movl (%ecx), %edx
+; X86-SSE1-NEXT:    movl 4(%ecx), %esi
+; X86-SSE1-NEXT:    xorl (%eax), %edx
+; X86-SSE1-NEXT:    xorl 4(%eax), %esi
+; X86-SSE1-NEXT:    orl %edx, %esi
+; X86-SSE1-NEXT:    movl 8(%ecx), %edx
+; X86-SSE1-NEXT:    xorl 8(%eax), %edx
+; X86-SSE1-NEXT:    movl 12(%ecx), %ecx
+; X86-SSE1-NEXT:    xorl 12(%eax), %ecx
+; X86-SSE1-NEXT:    orl %edx, %ecx
+; X86-SSE1-NEXT:    orl %esi, %ecx
+; X86-SSE1-NEXT:    setne %al
+; X86-SSE1-NEXT:    popl %esi
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length16_eq:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE2-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm1
+; X86-SSE2-NEXT:    pcmpeqb %xmm0, %xmm1
+; X86-SSE2-NEXT:    pmovmskb %xmm1, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    setne %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length16_eq:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE41-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm1
+; X86-SSE41-NEXT:    pxor %xmm0, %xmm1
+; X86-SSE41-NEXT:    ptest %xmm1, %xmm1
+; X86-SSE41-NEXT:    setne %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length16_eq:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu (%rsi), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm1
+; X64-SSE2-NEXT:    pmovmskb %xmm1, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    setne %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length16_eq:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu (%rsi), %xmm1
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm1
+; X64-SSE41-NEXT:    ptest %xmm1, %xmm1
+; X64-SSE41-NEXT:    setne %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX-LABEL: length16_eq:
+; X64-AVX:       # %bb.0:
+; X64-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX-NEXT:    vpxor (%rsi), %xmm0, %xmm0
+; X64-AVX-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX-NEXT:    setne %al
+; X64-AVX-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length16_eq:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-MIC-AVX-NEXT:    vmovdqu (%rsi), %xmm1
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm1, %zmm0, %k0
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k0
+; X64-MIC-AVX-NEXT:    setne %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 16) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length16_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length16_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %esi
+; X86-NEXT:    movl (%esi), %ecx
+; X86-NEXT:    movl (%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    jne .LBB33_4
+; X86-NEXT:  # %bb.1: # %loadbb1
+; X86-NEXT:    movl 4(%esi), %ecx
+; X86-NEXT:    movl 4(%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    jne .LBB33_4
+; X86-NEXT:  # %bb.2: # %loadbb2
+; X86-NEXT:    movl 8(%esi), %ecx
+; X86-NEXT:    movl 8(%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    jne .LBB33_4
+; X86-NEXT:  # %bb.3: # %loadbb3
+; X86-NEXT:    movl 12(%esi), %ecx
+; X86-NEXT:    movl 12(%eax), %edx
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    bswapl %edx
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    je .LBB33_5
+; X86-NEXT:  .LBB33_4: # %res_block
+; X86-NEXT:    xorl %eax, %eax
+; X86-NEXT:    cmpl %edx, %ecx
+; X86-NEXT:    setae %al
+; X86-NEXT:    leal -1(%eax,%eax), %eax
+; X86-NEXT:  .LBB33_5: # %endblock
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length16_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rcx
+; X64-NEXT:    movq (%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB33_2
+; X64-NEXT:  # %bb.1: # %loadbb1
+; X64-NEXT:    movq 8(%rdi), %rcx
+; X64-NEXT:    movq 8(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    je .LBB33_3
+; X64-NEXT:  .LBB33_2: # %res_block
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    setae %al
+; X64-NEXT:    leal -1(%rax,%rax), %eax
+; X64-NEXT:  .LBB33_3: # %endblock
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 16) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length16_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length16_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %esi
+; X86-NEXT:    movl (%esi), %eax
+; X86-NEXT:    movl (%edx), %ecx
+; X86-NEXT:    bswapl %eax
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    cmpl %ecx, %eax
+; X86-NEXT:    jne .LBB34_4
+; X86-NEXT:  # %bb.1: # %loadbb1
+; X86-NEXT:    movl 4(%esi), %eax
+; X86-NEXT:    movl 4(%edx), %ecx
+; X86-NEXT:    bswapl %eax
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    cmpl %ecx, %eax
+; X86-NEXT:    jne .LBB34_4
+; X86-NEXT:  # %bb.2: # %loadbb2
+; X86-NEXT:    movl 8(%esi), %eax
+; X86-NEXT:    movl 8(%edx), %ecx
+; X86-NEXT:    bswapl %eax
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    cmpl %ecx, %eax
+; X86-NEXT:    jne .LBB34_4
+; X86-NEXT:  # %bb.3: # %loadbb3
+; X86-NEXT:    movl 12(%esi), %eax
+; X86-NEXT:    movl 12(%edx), %ecx
+; X86-NEXT:    bswapl %eax
+; X86-NEXT:    bswapl %ecx
+; X86-NEXT:    xorl %edx, %edx
+; X86-NEXT:    cmpl %ecx, %eax
+; X86-NEXT:    je .LBB34_5
+; X86-NEXT:  .LBB34_4: # %res_block
+; X86-NEXT:    xorl %edx, %edx
+; X86-NEXT:    cmpl %ecx, %eax
+; X86-NEXT:    setae %dl
+; X86-NEXT:    leal -1(%edx,%edx), %edx
+; X86-NEXT:  .LBB34_5: # %endblock
+; X86-NEXT:    testl %edx, %edx
+; X86-NEXT:    setg %al
+; X86-NEXT:    popl %esi
+; X86-NEXT:    retl
+;
+; X64-LABEL: length16_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    movq (%rsi), %rcx
+; X64-NEXT:    bswapq %rax
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    jne .LBB34_2
+; X64-NEXT:  # %bb.1: # %loadbb1
+; X64-NEXT:    movq 8(%rdi), %rax
+; X64-NEXT:    movq 8(%rsi), %rcx
+; X64-NEXT:    bswapq %rax
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    xorl %edx, %edx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    je .LBB34_3
+; X64-NEXT:  .LBB34_2: # %res_block
+; X64-NEXT:    xorl %edx, %edx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    setae %dl
+; X64-NEXT:    leal -1(%rdx,%rdx), %edx
+; X64-NEXT:  .LBB34_3: # %endblock
+; X64-NEXT:    testl %edx, %edx
+; X64-NEXT:    setg %al
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 16) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length16_eq_const(i8* %X) nounwind {
+; X86-NOSSE-LABEL: length16_eq_const:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl %esi
+; X86-NOSSE-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NOSSE-NEXT:    movl $858927408, %ecx # imm = 0x33323130
+; X86-NOSSE-NEXT:    xorl (%eax), %ecx
+; X86-NOSSE-NEXT:    movl $926299444, %edx # imm = 0x37363534
+; X86-NOSSE-NEXT:    xorl 4(%eax), %edx
+; X86-NOSSE-NEXT:    orl %ecx, %edx
+; X86-NOSSE-NEXT:    movl $825243960, %ecx # imm = 0x31303938
+; X86-NOSSE-NEXT:    xorl 8(%eax), %ecx
+; X86-NOSSE-NEXT:    movl $892613426, %esi # imm = 0x35343332
+; X86-NOSSE-NEXT:    xorl 12(%eax), %esi
+; X86-NOSSE-NEXT:    orl %ecx, %esi
+; X86-NOSSE-NEXT:    orl %edx, %esi
+; X86-NOSSE-NEXT:    sete %al
+; X86-NOSSE-NEXT:    popl %esi
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length16_eq_const:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl %esi
+; X86-SSE1-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE1-NEXT:    movl $858927408, %ecx # imm = 0x33323130
+; X86-SSE1-NEXT:    xorl (%eax), %ecx
+; X86-SSE1-NEXT:    movl $926299444, %edx # imm = 0x37363534
+; X86-SSE1-NEXT:    xorl 4(%eax), %edx
+; X86-SSE1-NEXT:    orl %ecx, %edx
+; X86-SSE1-NEXT:    movl $825243960, %ecx # imm = 0x31303938
+; X86-SSE1-NEXT:    xorl 8(%eax), %ecx
+; X86-SSE1-NEXT:    movl $892613426, %esi # imm = 0x35343332
+; X86-SSE1-NEXT:    xorl 12(%eax), %esi
+; X86-SSE1-NEXT:    orl %ecx, %esi
+; X86-SSE1-NEXT:    orl %edx, %esi
+; X86-SSE1-NEXT:    sete %al
+; X86-SSE1-NEXT:    popl %esi
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length16_eq_const:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    sete %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length16_eq_const:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    sete %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length16_eq_const:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm0
+; X64-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length16_eq_const:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm0
+; X64-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X64-SSE41-NEXT:    sete %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX-LABEL: length16_eq_const:
+; X64-AVX:       # %bb.0:
+; X64-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX-NEXT:    vpxor {{.*}}(%rip), %xmm0, %xmm0
+; X64-AVX-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX-NEXT:    sete %al
+; X64-AVX-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length16_eq_const:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-MIC-AVX-NEXT:    vmovdqa {{.*#+}} xmm1 = [858927408,926299444,825243960,892613426]
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm1, %zmm0, %k0
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k0
+; X64-MIC-AVX-NEXT:    sete %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 16) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+; PR33914 - https://bugs.llvm.org/show_bug.cgi?id=33914
+
+define i32 @length24(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length24:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $24
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length24:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rcx
+; X64-NEXT:    movq (%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB36_3
+; X64-NEXT:  # %bb.1: # %loadbb1
+; X64-NEXT:    movq 8(%rdi), %rcx
+; X64-NEXT:    movq 8(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB36_3
+; X64-NEXT:  # %bb.2: # %loadbb2
+; X64-NEXT:    movq 16(%rdi), %rcx
+; X64-NEXT:    movq 16(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    je .LBB36_4
+; X64-NEXT:  .LBB36_3: # %res_block
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    setae %al
+; X64-NEXT:    leal -1(%rax,%rax), %eax
+; X64-NEXT:  .LBB36_4: # %endblock
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 24) nounwind
+  ret i32 %m
+}
+
+define i1 @length24_eq(i8* %x, i8* %y) nounwind {
+; X86-NOSSE-LABEL: length24_eq:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $24
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    sete %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length24_eq:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $24
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    sete %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length24_eq:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE2-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE2-NEXT:    movdqu 8(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X86-SSE2-NEXT:    movdqu 8(%eax), %xmm0
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X86-SSE2-NEXT:    pand %xmm2, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    sete %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length24_eq:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE41-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE41-NEXT:    movdqu 8(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X86-SSE41-NEXT:    movdqu 8(%eax), %xmm0
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X86-SSE41-NEXT:    por %xmm2, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    sete %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length24_eq:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu (%rsi), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm1
+; X64-SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; X64-SSE2-NEXT:    movq {{.*#+}} xmm2 = mem[0],zero
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X64-SSE2-NEXT:    pand %xmm1, %xmm2
+; X64-SSE2-NEXT:    pmovmskb %xmm2, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length24_eq:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu (%rsi), %xmm1
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm1
+; X64-SSE41-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; X64-SSE41-NEXT:    movq {{.*#+}} xmm2 = mem[0],zero
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X64-SSE41-NEXT:    por %xmm1, %xmm2
+; X64-SSE41-NEXT:    ptest %xmm2, %xmm2
+; X64-SSE41-NEXT:    sete %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX-LABEL: length24_eq:
+; X64-AVX:       # %bb.0:
+; X64-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
+; X64-AVX-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
+; X64-AVX-NEXT:    vpxor %xmm2, %xmm1, %xmm1
+; X64-AVX-NEXT:    vpxor (%rsi), %xmm0, %xmm0
+; X64-AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX-NEXT:    sete %al
+; X64-AVX-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length24_eq:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-MIC-AVX-NEXT:    vmovdqu (%rsi), %xmm1
+; X64-MIC-AVX-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
+; X64-MIC-AVX-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm3, %zmm2, %k0
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm1, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX-NEXT:    sete %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 24) nounwind
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length24_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length24_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $24
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length24_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rcx
+; X64-NEXT:    movq (%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB38_3
+; X64-NEXT:  # %bb.1: # %loadbb1
+; X64-NEXT:    movq 8(%rdi), %rcx
+; X64-NEXT:    movq 8(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB38_3
+; X64-NEXT:  # %bb.2: # %loadbb2
+; X64-NEXT:    movq 16(%rdi), %rcx
+; X64-NEXT:    movq 16(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    je .LBB38_4
+; X64-NEXT:  .LBB38_3: # %res_block
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    setae %al
+; X64-NEXT:    leal -1(%rax,%rax), %eax
+; X64-NEXT:  .LBB38_4: # %endblock
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 24) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length24_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length24_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $24
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length24_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    movq (%rsi), %rcx
+; X64-NEXT:    bswapq %rax
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    jne .LBB39_3
+; X64-NEXT:  # %bb.1: # %loadbb1
+; X64-NEXT:    movq 8(%rdi), %rax
+; X64-NEXT:    movq 8(%rsi), %rcx
+; X64-NEXT:    bswapq %rax
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    jne .LBB39_3
+; X64-NEXT:  # %bb.2: # %loadbb2
+; X64-NEXT:    movq 16(%rdi), %rax
+; X64-NEXT:    movq 16(%rsi), %rcx
+; X64-NEXT:    bswapq %rax
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    xorl %edx, %edx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    je .LBB39_4
+; X64-NEXT:  .LBB39_3: # %res_block
+; X64-NEXT:    xorl %edx, %edx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    setae %dl
+; X64-NEXT:    leal -1(%rdx,%rdx), %edx
+; X64-NEXT:  .LBB39_4: # %endblock
+; X64-NEXT:    testl %edx, %edx
+; X64-NEXT:    setg %al
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 24) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length24_eq_const(i8* %X) nounwind {
+; X86-NOSSE-LABEL: length24_eq_const:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $24
+; X86-NOSSE-NEXT:    pushl $.L.str
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    setne %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length24_eq_const:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $24
+; X86-SSE1-NEXT:    pushl $.L.str
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    setne %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length24_eq_const:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE2-NEXT:    movdqu 8(%eax), %xmm1
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm1
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm0
+; X86-SSE2-NEXT:    pand %xmm1, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    setne %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length24_eq_const:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE41-NEXT:    movdqu 8(%eax), %xmm1
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm1
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm0
+; X86-SSE41-NEXT:    por %xmm1, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    setne %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length24_eq_const:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm0
+; X64-SSE2-NEXT:    pand %xmm1, %xmm0
+; X64-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    setne %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length24_eq_const:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm1
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm0
+; X64-SSE41-NEXT:    por %xmm1, %xmm0
+; X64-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X64-SSE41-NEXT:    setne %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX-LABEL: length24_eq_const:
+; X64-AVX:       # %bb.0:
+; X64-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
+; X64-AVX-NEXT:    vpxor {{.*}}(%rip), %xmm1, %xmm1
+; X64-AVX-NEXT:    vpxor {{.*}}(%rip), %xmm0, %xmm0
+; X64-AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX-NEXT:    setne %al
+; X64-AVX-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length24_eq_const:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-MIC-AVX-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
+; X64-MIC-AVX-NEXT:    vmovdqa {{.*#+}} xmm2 = [959985462,858927408,0,0]
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm2, %zmm1, %k0
+; X64-MIC-AVX-NEXT:    vmovdqa {{.*#+}} xmm1 = [858927408,926299444,825243960,892613426]
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm1, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX-NEXT:    setne %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 24) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length31(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length31:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $31
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length31:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $31, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 31) nounwind
+  ret i32 %m
+}
+
+define i1 @length31_eq(i8* %x, i8* %y) nounwind {
+; X86-NOSSE-LABEL: length31_eq:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $31
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    sete %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length31_eq:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $31
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    sete %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length31_eq:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE2-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE2-NEXT:    movdqu 15(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X86-SSE2-NEXT:    movdqu 15(%eax), %xmm0
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X86-SSE2-NEXT:    pand %xmm2, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    sete %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length31_eq:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE41-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE41-NEXT:    movdqu 15(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X86-SSE41-NEXT:    movdqu 15(%eax), %xmm0
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X86-SSE41-NEXT:    por %xmm2, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    sete %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length31_eq:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 15(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu (%rsi), %xmm2
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X64-SSE2-NEXT:    movdqu 15(%rsi), %xmm0
+; X64-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X64-SSE2-NEXT:    pand %xmm2, %xmm0
+; X64-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length31_eq:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 15(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu (%rsi), %xmm2
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X64-SSE41-NEXT:    movdqu 15(%rsi), %xmm0
+; X64-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X64-SSE41-NEXT:    por %xmm2, %xmm0
+; X64-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X64-SSE41-NEXT:    sete %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX-LABEL: length31_eq:
+; X64-AVX:       # %bb.0:
+; X64-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX-NEXT:    vmovdqu 15(%rdi), %xmm1
+; X64-AVX-NEXT:    vpxor 15(%rsi), %xmm1, %xmm1
+; X64-AVX-NEXT:    vpxor (%rsi), %xmm0, %xmm0
+; X64-AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX-NEXT:    sete %al
+; X64-AVX-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length31_eq:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-MIC-AVX-NEXT:    vmovdqu 15(%rdi), %xmm1
+; X64-MIC-AVX-NEXT:    vmovdqu (%rsi), %xmm2
+; X64-MIC-AVX-NEXT:    vmovdqu 15(%rsi), %xmm3
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm3, %zmm1, %k0
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm2, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX-NEXT:    sete %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 31) nounwind
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length31_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length31_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $31
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length31_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $31, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 31) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length31_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length31_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $31
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length31_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $31, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 31) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length31_eq_prefer128(i8* %x, i8* %y) nounwind "prefer-vector-width"="128" {
+; X86-NOSSE-LABEL: length31_eq_prefer128:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $31
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    sete %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length31_eq_prefer128:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $31
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    sete %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length31_eq_prefer128:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE2-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE2-NEXT:    movdqu 15(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X86-SSE2-NEXT:    movdqu 15(%eax), %xmm0
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X86-SSE2-NEXT:    pand %xmm2, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    sete %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length31_eq_prefer128:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE41-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE41-NEXT:    movdqu 15(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X86-SSE41-NEXT:    movdqu 15(%eax), %xmm0
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X86-SSE41-NEXT:    por %xmm2, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    sete %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length31_eq_prefer128:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 15(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu (%rsi), %xmm2
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X64-SSE2-NEXT:    movdqu 15(%rsi), %xmm0
+; X64-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X64-SSE2-NEXT:    pand %xmm2, %xmm0
+; X64-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length31_eq_prefer128:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 15(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu (%rsi), %xmm2
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X64-SSE41-NEXT:    movdqu 15(%rsi), %xmm0
+; X64-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X64-SSE41-NEXT:    por %xmm2, %xmm0
+; X64-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X64-SSE41-NEXT:    sete %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX-LABEL: length31_eq_prefer128:
+; X64-AVX:       # %bb.0:
+; X64-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX-NEXT:    vmovdqu 15(%rdi), %xmm1
+; X64-AVX-NEXT:    vpxor 15(%rsi), %xmm1, %xmm1
+; X64-AVX-NEXT:    vpxor (%rsi), %xmm0, %xmm0
+; X64-AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX-NEXT:    sete %al
+; X64-AVX-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length31_eq_prefer128:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-MIC-AVX-NEXT:    vmovdqu 15(%rdi), %xmm1
+; X64-MIC-AVX-NEXT:    vmovdqu (%rsi), %xmm2
+; X64-MIC-AVX-NEXT:    vmovdqu 15(%rsi), %xmm3
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm3, %zmm1, %k0
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm2, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX-NEXT:    sete %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 31) nounwind
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length31_eq_const(i8* %X) nounwind {
+; X86-NOSSE-LABEL: length31_eq_const:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $31
+; X86-NOSSE-NEXT:    pushl $.L.str
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    setne %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length31_eq_const:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $31
+; X86-SSE1-NEXT:    pushl $.L.str
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    setne %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length31_eq_const:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE2-NEXT:    movdqu 15(%eax), %xmm1
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm1
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm0
+; X86-SSE2-NEXT:    pand %xmm1, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    setne %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length31_eq_const:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE41-NEXT:    movdqu 15(%eax), %xmm1
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm1
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm0
+; X86-SSE41-NEXT:    por %xmm1, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    setne %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length31_eq_const:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 15(%rdi), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm0
+; X64-SSE2-NEXT:    pand %xmm1, %xmm0
+; X64-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    setne %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length31_eq_const:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 15(%rdi), %xmm1
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm1
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm0
+; X64-SSE41-NEXT:    por %xmm1, %xmm0
+; X64-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X64-SSE41-NEXT:    setne %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX-LABEL: length31_eq_const:
+; X64-AVX:       # %bb.0:
+; X64-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX-NEXT:    vmovdqu 15(%rdi), %xmm1
+; X64-AVX-NEXT:    vpxor {{.*}}(%rip), %xmm1, %xmm1
+; X64-AVX-NEXT:    vpxor {{.*}}(%rip), %xmm0, %xmm0
+; X64-AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX-NEXT:    setne %al
+; X64-AVX-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length31_eq_const:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-MIC-AVX-NEXT:    vmovdqu 15(%rdi), %xmm1
+; X64-MIC-AVX-NEXT:    vmovdqa {{.*#+}} xmm2 = [943142453,842084409,909456435,809056311]
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm2, %zmm1, %k0
+; X64-MIC-AVX-NEXT:    vmovdqa {{.*#+}} xmm1 = [858927408,926299444,825243960,892613426]
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm1, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX-NEXT:    setne %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 31) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length32(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length32:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $32
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length32:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rcx
+; X64-NEXT:    movq (%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB47_4
+; X64-NEXT:  # %bb.1: # %loadbb1
+; X64-NEXT:    movq 8(%rdi), %rcx
+; X64-NEXT:    movq 8(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB47_4
+; X64-NEXT:  # %bb.2: # %loadbb2
+; X64-NEXT:    movq 16(%rdi), %rcx
+; X64-NEXT:    movq 16(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB47_4
+; X64-NEXT:  # %bb.3: # %loadbb3
+; X64-NEXT:    movq 24(%rdi), %rcx
+; X64-NEXT:    movq 24(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    je .LBB47_5
+; X64-NEXT:  .LBB47_4: # %res_block
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    setae %al
+; X64-NEXT:    leal -1(%rax,%rax), %eax
+; X64-NEXT:  .LBB47_5: # %endblock
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 32) nounwind
+  ret i32 %m
+}
+
+; PR33325 - https://bugs.llvm.org/show_bug.cgi?id=33325
+
+define i1 @length32_eq(i8* %x, i8* %y) nounwind {
+; X86-NOSSE-LABEL: length32_eq:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $32
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    sete %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length32_eq:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $32
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    sete %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length32_eq:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE2-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE2-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X86-SSE2-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X86-SSE2-NEXT:    pand %xmm2, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    sete %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length32_eq:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE41-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE41-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X86-SSE41-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X86-SSE41-NEXT:    por %xmm2, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    sete %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length32_eq:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu (%rsi), %xmm2
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X64-SSE2-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X64-SSE2-NEXT:    pand %xmm2, %xmm0
+; X64-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length32_eq:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu (%rsi), %xmm2
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X64-SSE41-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X64-SSE41-NEXT:    por %xmm2, %xmm0
+; X64-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X64-SSE41-NEXT:    sete %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX1-LABEL: length32_eq:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX1-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-AVX1-NEXT:    vpxor 16(%rsi), %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpxor (%rsi), %xmm0, %xmm0
+; X64-AVX1-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX1-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX1-NEXT:    sete %al
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length32_eq:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vpxor (%rsi), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512-LABEL: length32_eq:
+; X64-AVX512:       # %bb.0:
+; X64-AVX512-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX512-NEXT:    vpxor (%rsi), %ymm0, %ymm0
+; X64-AVX512-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX512-NEXT:    sete %al
+; X64-AVX512-NEXT:    vzeroupper
+; X64-AVX512-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length32_eq:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX-NEXT:    vmovdqu (%rsi), %ymm1
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm1, %zmm0, %k0
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k0
+; X64-MIC-AVX-NEXT:    sete %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 32) nounwind
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length32_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length32_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $32
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length32_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rcx
+; X64-NEXT:    movq (%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB49_4
+; X64-NEXT:  # %bb.1: # %loadbb1
+; X64-NEXT:    movq 8(%rdi), %rcx
+; X64-NEXT:    movq 8(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB49_4
+; X64-NEXT:  # %bb.2: # %loadbb2
+; X64-NEXT:    movq 16(%rdi), %rcx
+; X64-NEXT:    movq 16(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    jne .LBB49_4
+; X64-NEXT:  # %bb.3: # %loadbb3
+; X64-NEXT:    movq 24(%rdi), %rcx
+; X64-NEXT:    movq 24(%rsi), %rdx
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    bswapq %rdx
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    je .LBB49_5
+; X64-NEXT:  .LBB49_4: # %res_block
+; X64-NEXT:    xorl %eax, %eax
+; X64-NEXT:    cmpq %rdx, %rcx
+; X64-NEXT:    setae %al
+; X64-NEXT:    leal -1(%rax,%rax), %eax
+; X64-NEXT:  .LBB49_5: # %endblock
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 32) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length32_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length32_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $32
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length32_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    movq (%rdi), %rax
+; X64-NEXT:    movq (%rsi), %rcx
+; X64-NEXT:    bswapq %rax
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    jne .LBB50_4
+; X64-NEXT:  # %bb.1: # %loadbb1
+; X64-NEXT:    movq 8(%rdi), %rax
+; X64-NEXT:    movq 8(%rsi), %rcx
+; X64-NEXT:    bswapq %rax
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    jne .LBB50_4
+; X64-NEXT:  # %bb.2: # %loadbb2
+; X64-NEXT:    movq 16(%rdi), %rax
+; X64-NEXT:    movq 16(%rsi), %rcx
+; X64-NEXT:    bswapq %rax
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    jne .LBB50_4
+; X64-NEXT:  # %bb.3: # %loadbb3
+; X64-NEXT:    movq 24(%rdi), %rax
+; X64-NEXT:    movq 24(%rsi), %rcx
+; X64-NEXT:    bswapq %rax
+; X64-NEXT:    bswapq %rcx
+; X64-NEXT:    xorl %edx, %edx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    je .LBB50_5
+; X64-NEXT:  .LBB50_4: # %res_block
+; X64-NEXT:    xorl %edx, %edx
+; X64-NEXT:    cmpq %rcx, %rax
+; X64-NEXT:    setae %dl
+; X64-NEXT:    leal -1(%rdx,%rdx), %edx
+; X64-NEXT:  .LBB50_5: # %endblock
+; X64-NEXT:    testl %edx, %edx
+; X64-NEXT:    setg %al
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 32) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length32_eq_prefer128(i8* %x, i8* %y) nounwind "prefer-vector-width"="128" {
+; X86-NOSSE-LABEL: length32_eq_prefer128:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $32
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    sete %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length32_eq_prefer128:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $32
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    sete %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length32_eq_prefer128:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE2-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE2-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X86-SSE2-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X86-SSE2-NEXT:    pand %xmm2, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    sete %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length32_eq_prefer128:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE41-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE41-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X86-SSE41-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X86-SSE41-NEXT:    por %xmm2, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    sete %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length32_eq_prefer128:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu (%rsi), %xmm2
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X64-SSE2-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X64-SSE2-NEXT:    pand %xmm2, %xmm0
+; X64-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length32_eq_prefer128:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu (%rsi), %xmm2
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X64-SSE41-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X64-SSE41-NEXT:    por %xmm2, %xmm0
+; X64-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X64-SSE41-NEXT:    sete %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX-LABEL: length32_eq_prefer128:
+; X64-AVX:       # %bb.0:
+; X64-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-AVX-NEXT:    vpxor 16(%rsi), %xmm1, %xmm1
+; X64-AVX-NEXT:    vpxor (%rsi), %xmm0, %xmm0
+; X64-AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX-NEXT:    sete %al
+; X64-AVX-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length32_eq_prefer128:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-MIC-AVX-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-MIC-AVX-NEXT:    vmovdqu (%rsi), %xmm2
+; X64-MIC-AVX-NEXT:    vmovdqu 16(%rsi), %xmm3
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm3, %zmm1, %k0
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm2, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX-NEXT:    sete %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 32) nounwind
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length32_eq_const(i8* %X) nounwind {
+; X86-NOSSE-LABEL: length32_eq_const:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $32
+; X86-NOSSE-NEXT:    pushl $.L.str
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    setne %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length32_eq_const:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $32
+; X86-SSE1-NEXT:    pushl $.L.str
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    setne %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length32_eq_const:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE2-NEXT:    movdqu 16(%eax), %xmm1
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm1
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm0
+; X86-SSE2-NEXT:    pand %xmm1, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    setne %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length32_eq_const:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE41-NEXT:    movdqu 16(%eax), %xmm1
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm1
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm0
+; X86-SSE41-NEXT:    por %xmm1, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    setne %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length32_eq_const:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm0
+; X64-SSE2-NEXT:    pand %xmm1, %xmm0
+; X64-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    setne %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length32_eq_const:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm1
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm0
+; X64-SSE41-NEXT:    por %xmm1, %xmm0
+; X64-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X64-SSE41-NEXT:    setne %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX1-LABEL: length32_eq_const:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX1-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm0, %xmm0
+; X64-AVX1-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX1-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX1-NEXT:    setne %al
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length32_eq_const:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512-LABEL: length32_eq_const:
+; X64-AVX512:       # %bb.0:
+; X64-AVX512-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX512-NEXT:    vpxor {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX512-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX512-NEXT:    setne %al
+; X64-AVX512-NEXT:    vzeroupper
+; X64-AVX512-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length32_eq_const:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [858927408,926299444,825243960,892613426,959985462,858927408,926299444,825243960]
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm1, %zmm0, %k0
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k0
+; X64-MIC-AVX-NEXT:    setne %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 32) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length48(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length48:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $48
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length48:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $48, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 48) nounwind
+  ret i32 %m
+}
+
+define i1 @length48_eq(i8* %x, i8* %y) nounwind {
+; X86-NOSSE-LABEL: length48_eq:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $48
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    sete %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length48_eq:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $48
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    sete %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length48_eq:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE2-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE2-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X86-SSE2-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X86-SSE2-NEXT:    movdqu 32(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu 32(%eax), %xmm3
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm3
+; X86-SSE2-NEXT:    pand %xmm0, %xmm3
+; X86-SSE2-NEXT:    pand %xmm2, %xmm3
+; X86-SSE2-NEXT:    pmovmskb %xmm3, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    sete %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length48_eq:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE41-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE41-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X86-SSE41-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X86-SSE41-NEXT:    movdqu 32(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu 32(%eax), %xmm3
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm3
+; X86-SSE41-NEXT:    por %xmm0, %xmm3
+; X86-SSE41-NEXT:    por %xmm2, %xmm3
+; X86-SSE41-NEXT:    ptest %xmm3, %xmm3
+; X86-SSE41-NEXT:    sete %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length48_eq:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE2-NEXT:    movdqu (%rsi), %xmm3
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm3
+; X64-SSE2-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X64-SSE2-NEXT:    movdqu 32(%rsi), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb %xmm2, %xmm1
+; X64-SSE2-NEXT:    pand %xmm0, %xmm1
+; X64-SSE2-NEXT:    pand %xmm3, %xmm1
+; X64-SSE2-NEXT:    pmovmskb %xmm1, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length48_eq:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE41-NEXT:    movdqu (%rsi), %xmm3
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm3
+; X64-SSE41-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X64-SSE41-NEXT:    movdqu 32(%rsi), %xmm1
+; X64-SSE41-NEXT:    pxor %xmm2, %xmm1
+; X64-SSE41-NEXT:    por %xmm0, %xmm1
+; X64-SSE41-NEXT:    por %xmm3, %xmm1
+; X64-SSE41-NEXT:    ptest %xmm1, %xmm1
+; X64-SSE41-NEXT:    sete %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX1-LABEL: length48_eq:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX1-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-AVX1-NEXT:    vmovdqu 32(%rdi), %xmm2
+; X64-AVX1-NEXT:    vpxor 16(%rsi), %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpxor (%rsi), %xmm0, %xmm0
+; X64-AVX1-NEXT:    vpxor 32(%rsi), %xmm2, %xmm2
+; X64-AVX1-NEXT:    vpor %xmm2, %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX1-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX1-NEXT:    sete %al
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length48_eq:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 32(%rdi), %xmm1
+; X64-AVX2-NEXT:    vmovdqu 32(%rsi), %xmm2
+; X64-AVX2-NEXT:    vpxor (%rsi), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpxor %ymm2, %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512-LABEL: length48_eq:
+; X64-AVX512:       # %bb.0:
+; X64-AVX512-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX512-NEXT:    vmovdqu 32(%rdi), %xmm1
+; X64-AVX512-NEXT:    vmovdqu 32(%rsi), %xmm2
+; X64-AVX512-NEXT:    vpxor (%rsi), %ymm0, %ymm0
+; X64-AVX512-NEXT:    vpxor %ymm2, %ymm1, %ymm1
+; X64-AVX512-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX512-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX512-NEXT:    sete %al
+; X64-AVX512-NEXT:    vzeroupper
+; X64-AVX512-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length48_eq:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX-NEXT:    vmovdqu (%rsi), %ymm1
+; X64-MIC-AVX-NEXT:    vmovdqu 32(%rdi), %xmm2
+; X64-MIC-AVX-NEXT:    vmovdqu 32(%rsi), %xmm3
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm3, %zmm2, %k0
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm1, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX-NEXT:    sete %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 48) nounwind
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length48_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length48_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $48
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length48_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $48, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 48) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length48_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length48_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $48
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length48_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $48, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 48) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length48_eq_prefer128(i8* %x, i8* %y) nounwind "prefer-vector-width"="128" {
+; X86-NOSSE-LABEL: length48_eq_prefer128:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $48
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    sete %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length48_eq_prefer128:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $48
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    sete %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length48_eq_prefer128:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE2-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE2-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X86-SSE2-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X86-SSE2-NEXT:    movdqu 32(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu 32(%eax), %xmm3
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm3
+; X86-SSE2-NEXT:    pand %xmm0, %xmm3
+; X86-SSE2-NEXT:    pand %xmm2, %xmm3
+; X86-SSE2-NEXT:    pmovmskb %xmm3, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    sete %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length48_eq_prefer128:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE41-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE41-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X86-SSE41-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X86-SSE41-NEXT:    movdqu 32(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu 32(%eax), %xmm3
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm3
+; X86-SSE41-NEXT:    por %xmm0, %xmm3
+; X86-SSE41-NEXT:    por %xmm2, %xmm3
+; X86-SSE41-NEXT:    ptest %xmm3, %xmm3
+; X86-SSE41-NEXT:    sete %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length48_eq_prefer128:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE2-NEXT:    movdqu (%rsi), %xmm3
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm3
+; X64-SSE2-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X64-SSE2-NEXT:    movdqu 32(%rsi), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb %xmm2, %xmm1
+; X64-SSE2-NEXT:    pand %xmm0, %xmm1
+; X64-SSE2-NEXT:    pand %xmm3, %xmm1
+; X64-SSE2-NEXT:    pmovmskb %xmm1, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length48_eq_prefer128:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE41-NEXT:    movdqu (%rsi), %xmm3
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm3
+; X64-SSE41-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X64-SSE41-NEXT:    movdqu 32(%rsi), %xmm1
+; X64-SSE41-NEXT:    pxor %xmm2, %xmm1
+; X64-SSE41-NEXT:    por %xmm0, %xmm1
+; X64-SSE41-NEXT:    por %xmm3, %xmm1
+; X64-SSE41-NEXT:    ptest %xmm1, %xmm1
+; X64-SSE41-NEXT:    sete %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX-LABEL: length48_eq_prefer128:
+; X64-AVX:       # %bb.0:
+; X64-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-AVX-NEXT:    vmovdqu 32(%rdi), %xmm2
+; X64-AVX-NEXT:    vpxor 16(%rsi), %xmm1, %xmm1
+; X64-AVX-NEXT:    vpxor (%rsi), %xmm0, %xmm0
+; X64-AVX-NEXT:    vpxor 32(%rsi), %xmm2, %xmm2
+; X64-AVX-NEXT:    vpor %xmm2, %xmm1, %xmm1
+; X64-AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX-NEXT:    sete %al
+; X64-AVX-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length48_eq_prefer128:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-MIC-AVX-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-MIC-AVX-NEXT:    vmovdqu 32(%rdi), %xmm2
+; X64-MIC-AVX-NEXT:    vmovdqu (%rsi), %xmm3
+; X64-MIC-AVX-NEXT:    vmovdqu 16(%rsi), %xmm4
+; X64-MIC-AVX-NEXT:    vmovdqu 32(%rsi), %xmm5
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm4, %zmm1, %k0
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm3, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm5, %zmm2, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k1, %k0
+; X64-MIC-AVX-NEXT:    sete %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 48) nounwind
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length48_eq_const(i8* %X) nounwind {
+; X86-NOSSE-LABEL: length48_eq_const:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $48
+; X86-NOSSE-NEXT:    pushl $.L.str
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    setne %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length48_eq_const:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $48
+; X86-SSE1-NEXT:    pushl $.L.str
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    setne %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length48_eq_const:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE2-NEXT:    movdqu 16(%eax), %xmm1
+; X86-SSE2-NEXT:    movdqu 32(%eax), %xmm2
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm1
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm0
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm2
+; X86-SSE2-NEXT:    pand %xmm1, %xmm2
+; X86-SSE2-NEXT:    pand %xmm0, %xmm2
+; X86-SSE2-NEXT:    pmovmskb %xmm2, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    setne %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length48_eq_const:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE41-NEXT:    movdqu 16(%eax), %xmm1
+; X86-SSE41-NEXT:    movdqu 32(%eax), %xmm2
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm1
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm0
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm2
+; X86-SSE41-NEXT:    por %xmm1, %xmm2
+; X86-SSE41-NEXT:    por %xmm0, %xmm2
+; X86-SSE41-NEXT:    ptest %xmm2, %xmm2
+; X86-SSE41-NEXT:    setne %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length48_eq_const:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm0
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm2
+; X64-SSE2-NEXT:    pand %xmm1, %xmm2
+; X64-SSE2-NEXT:    pand %xmm0, %xmm2
+; X64-SSE2-NEXT:    pmovmskb %xmm2, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    setne %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length48_eq_const:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm1
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm0
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm2
+; X64-SSE41-NEXT:    por %xmm1, %xmm2
+; X64-SSE41-NEXT:    por %xmm0, %xmm2
+; X64-SSE41-NEXT:    ptest %xmm2, %xmm2
+; X64-SSE41-NEXT:    setne %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX1-LABEL: length48_eq_const:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX1-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-AVX1-NEXT:    vmovdqu 32(%rdi), %xmm2
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm0, %xmm0
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm2, %xmm2
+; X64-AVX1-NEXT:    vpor %xmm2, %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX1-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX1-NEXT:    setne %al
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length48_eq_const:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 32(%rdi), %xmm1
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512-LABEL: length48_eq_const:
+; X64-AVX512:       # %bb.0:
+; X64-AVX512-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX512-NEXT:    vmovdqu 32(%rdi), %xmm1
+; X64-AVX512-NEXT:    vpxor {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX512-NEXT:    vpxor {{.*}}(%rip), %ymm1, %ymm1
+; X64-AVX512-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX512-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX512-NEXT:    setne %al
+; X64-AVX512-NEXT:    vzeroupper
+; X64-AVX512-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length48_eq_const:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX-NEXT:    vmovdqu 32(%rdi), %xmm1
+; X64-MIC-AVX-NEXT:    vmovdqa {{.*#+}} ymm2 = [892613426,959985462,858927408,926299444,0,0,0,0]
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm2, %zmm1, %k0
+; X64-MIC-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [858927408,926299444,825243960,892613426,959985462,858927408,926299444,825243960]
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm1, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX-NEXT:    setne %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 48) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length63(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length63:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $63
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length63:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $63, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 63) nounwind
+  ret i32 %m
+}
+
+define i1 @length63_eq(i8* %x, i8* %y) nounwind {
+; X86-NOSSE-LABEL: length63_eq:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $63
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    setne %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length63_eq:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $63
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    setne %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length63_eq:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE2-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE2-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X86-SSE2-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X86-SSE2-NEXT:    movdqu 32(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu 32(%eax), %xmm3
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm3
+; X86-SSE2-NEXT:    movdqu 47(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu 47(%eax), %xmm4
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm4
+; X86-SSE2-NEXT:    pand %xmm3, %xmm4
+; X86-SSE2-NEXT:    pand %xmm0, %xmm4
+; X86-SSE2-NEXT:    pand %xmm2, %xmm4
+; X86-SSE2-NEXT:    pmovmskb %xmm4, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    setne %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length63_eq:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE41-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE41-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X86-SSE41-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X86-SSE41-NEXT:    movdqu 32(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu 32(%eax), %xmm3
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm3
+; X86-SSE41-NEXT:    movdqu 47(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu 47(%eax), %xmm4
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm4
+; X86-SSE41-NEXT:    por %xmm3, %xmm4
+; X86-SSE41-NEXT:    por %xmm0, %xmm4
+; X86-SSE41-NEXT:    por %xmm2, %xmm4
+; X86-SSE41-NEXT:    ptest %xmm4, %xmm4
+; X86-SSE41-NEXT:    setne %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length63_eq:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE2-NEXT:    movdqu 47(%rdi), %xmm3
+; X64-SSE2-NEXT:    movdqu (%rsi), %xmm4
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm4
+; X64-SSE2-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X64-SSE2-NEXT:    movdqu 32(%rsi), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb %xmm2, %xmm1
+; X64-SSE2-NEXT:    movdqu 47(%rsi), %xmm2
+; X64-SSE2-NEXT:    pcmpeqb %xmm3, %xmm2
+; X64-SSE2-NEXT:    pand %xmm1, %xmm2
+; X64-SSE2-NEXT:    pand %xmm0, %xmm2
+; X64-SSE2-NEXT:    pand %xmm4, %xmm2
+; X64-SSE2-NEXT:    pmovmskb %xmm2, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    setne %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length63_eq:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE41-NEXT:    movdqu 47(%rdi), %xmm3
+; X64-SSE41-NEXT:    movdqu (%rsi), %xmm4
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm4
+; X64-SSE41-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X64-SSE41-NEXT:    movdqu 32(%rsi), %xmm1
+; X64-SSE41-NEXT:    pxor %xmm2, %xmm1
+; X64-SSE41-NEXT:    movdqu 47(%rsi), %xmm2
+; X64-SSE41-NEXT:    pxor %xmm3, %xmm2
+; X64-SSE41-NEXT:    por %xmm1, %xmm2
+; X64-SSE41-NEXT:    por %xmm0, %xmm2
+; X64-SSE41-NEXT:    por %xmm4, %xmm2
+; X64-SSE41-NEXT:    ptest %xmm2, %xmm2
+; X64-SSE41-NEXT:    setne %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX1-LABEL: length63_eq:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX1-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-AVX1-NEXT:    vmovdqu 32(%rdi), %xmm2
+; X64-AVX1-NEXT:    vmovdqu 47(%rdi), %xmm3
+; X64-AVX1-NEXT:    vpxor 47(%rsi), %xmm3, %xmm3
+; X64-AVX1-NEXT:    vpxor 32(%rsi), %xmm2, %xmm2
+; X64-AVX1-NEXT:    vpor %xmm3, %xmm2, %xmm2
+; X64-AVX1-NEXT:    vpxor 16(%rsi), %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpor %xmm2, %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpxor (%rsi), %xmm0, %xmm0
+; X64-AVX1-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX1-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX1-NEXT:    setne %al
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length63_eq:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 31(%rdi), %ymm1
+; X64-AVX2-NEXT:    vpxor 31(%rsi), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpxor (%rsi), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512-LABEL: length63_eq:
+; X64-AVX512:       # %bb.0:
+; X64-AVX512-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX512-NEXT:    vmovdqu 31(%rdi), %ymm1
+; X64-AVX512-NEXT:    vpxor 31(%rsi), %ymm1, %ymm1
+; X64-AVX512-NEXT:    vpxor (%rsi), %ymm0, %ymm0
+; X64-AVX512-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX512-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX512-NEXT:    setne %al
+; X64-AVX512-NEXT:    vzeroupper
+; X64-AVX512-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length63_eq:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX-NEXT:    vmovdqu 31(%rdi), %ymm1
+; X64-MIC-AVX-NEXT:    vmovdqu (%rsi), %ymm2
+; X64-MIC-AVX-NEXT:    vmovdqu 31(%rsi), %ymm3
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm3, %zmm1, %k0
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm2, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX-NEXT:    setne %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 63) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length63_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length63_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $63
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length63_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $63, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 63) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length63_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length63_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $63
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length63_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $63, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 63) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length63_eq_const(i8* %X) nounwind {
+; X86-NOSSE-LABEL: length63_eq_const:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $63
+; X86-NOSSE-NEXT:    pushl $.L.str
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    sete %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length63_eq_const:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $63
+; X86-SSE1-NEXT:    pushl $.L.str
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    sete %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length63_eq_const:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE2-NEXT:    movdqu 16(%eax), %xmm1
+; X86-SSE2-NEXT:    movdqu 32(%eax), %xmm2
+; X86-SSE2-NEXT:    movdqu 47(%eax), %xmm3
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm3
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm2
+; X86-SSE2-NEXT:    pand %xmm3, %xmm2
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm1
+; X86-SSE2-NEXT:    pand %xmm2, %xmm1
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm0
+; X86-SSE2-NEXT:    pand %xmm1, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    sete %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length63_eq_const:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE41-NEXT:    movdqu 16(%eax), %xmm1
+; X86-SSE41-NEXT:    movdqu 32(%eax), %xmm2
+; X86-SSE41-NEXT:    movdqu 47(%eax), %xmm3
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm3
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm2
+; X86-SSE41-NEXT:    por %xmm3, %xmm2
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm1
+; X86-SSE41-NEXT:    por %xmm2, %xmm1
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm0
+; X86-SSE41-NEXT:    por %xmm1, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    sete %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length63_eq_const:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE2-NEXT:    movdqu 47(%rdi), %xmm3
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm3
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm2
+; X64-SSE2-NEXT:    pand %xmm3, %xmm2
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm1
+; X64-SSE2-NEXT:    pand %xmm2, %xmm1
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm0
+; X64-SSE2-NEXT:    pand %xmm1, %xmm0
+; X64-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length63_eq_const:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE41-NEXT:    movdqu 47(%rdi), %xmm3
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm3
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm2
+; X64-SSE41-NEXT:    por %xmm3, %xmm2
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm1
+; X64-SSE41-NEXT:    por %xmm2, %xmm1
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm0
+; X64-SSE41-NEXT:    por %xmm1, %xmm0
+; X64-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X64-SSE41-NEXT:    sete %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX1-LABEL: length63_eq_const:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX1-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-AVX1-NEXT:    vmovdqu 32(%rdi), %xmm2
+; X64-AVX1-NEXT:    vmovdqu 47(%rdi), %xmm3
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm3, %xmm3
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm2, %xmm2
+; X64-AVX1-NEXT:    vpor %xmm3, %xmm2, %xmm2
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpor %xmm2, %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm0, %xmm0
+; X64-AVX1-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX1-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX1-NEXT:    sete %al
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length63_eq_const:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 31(%rdi), %ymm1
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512-LABEL: length63_eq_const:
+; X64-AVX512:       # %bb.0:
+; X64-AVX512-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX512-NEXT:    vmovdqu 31(%rdi), %ymm1
+; X64-AVX512-NEXT:    vpxor {{.*}}(%rip), %ymm1, %ymm1
+; X64-AVX512-NEXT:    vpxor {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX512-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX512-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX512-NEXT:    sete %al
+; X64-AVX512-NEXT:    vzeroupper
+; X64-AVX512-NEXT:    retq
+;
+; X64-MIC-AVX-LABEL: length63_eq_const:
+; X64-MIC-AVX:       # %bb.0:
+; X64-MIC-AVX-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX-NEXT:    vmovdqu 31(%rdi), %ymm1
+; X64-MIC-AVX-NEXT:    vmovdqa {{.*#+}} ymm2 = [875770417,943142453,842084409,909456435,809056311,875770417,943142453,842084409]
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm2, %zmm1, %k0
+; X64-MIC-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [858927408,926299444,825243960,892613426,959985462,858927408,926299444,825243960]
+; X64-MIC-AVX-NEXT:    vpcmpneqd %zmm1, %zmm0, %k1
+; X64-MIC-AVX-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX-NEXT:    sete %al
+; X64-MIC-AVX-NEXT:    vzeroupper
+; X64-MIC-AVX-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 63) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length64(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length64:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $64
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length64:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $64, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 64) nounwind
+  ret i32 %m
+}
+
+define i1 @length64_eq(i8* %x, i8* %y) nounwind {
+; X86-NOSSE-LABEL: length64_eq:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $64
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    setne %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length64_eq:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $64
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    setne %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length64_eq:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE2-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE2-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE2-NEXT:    pcmpeqb %xmm0, %xmm2
+; X86-SSE2-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X86-SSE2-NEXT:    movdqu 32(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu 32(%eax), %xmm3
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm3
+; X86-SSE2-NEXT:    movdqu 48(%ecx), %xmm1
+; X86-SSE2-NEXT:    movdqu 48(%eax), %xmm4
+; X86-SSE2-NEXT:    pcmpeqb %xmm1, %xmm4
+; X86-SSE2-NEXT:    pand %xmm3, %xmm4
+; X86-SSE2-NEXT:    pand %xmm0, %xmm4
+; X86-SSE2-NEXT:    pand %xmm2, %xmm4
+; X86-SSE2-NEXT:    pmovmskb %xmm4, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    setne %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length64_eq:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-SSE41-NEXT:    movdqu (%ecx), %xmm0
+; X86-SSE41-NEXT:    movdqu 16(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm2
+; X86-SSE41-NEXT:    pxor %xmm0, %xmm2
+; X86-SSE41-NEXT:    movdqu 16(%eax), %xmm0
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X86-SSE41-NEXT:    movdqu 32(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu 32(%eax), %xmm3
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm3
+; X86-SSE41-NEXT:    movdqu 48(%ecx), %xmm1
+; X86-SSE41-NEXT:    movdqu 48(%eax), %xmm4
+; X86-SSE41-NEXT:    pxor %xmm1, %xmm4
+; X86-SSE41-NEXT:    por %xmm3, %xmm4
+; X86-SSE41-NEXT:    por %xmm0, %xmm4
+; X86-SSE41-NEXT:    por %xmm2, %xmm4
+; X86-SSE41-NEXT:    ptest %xmm4, %xmm4
+; X86-SSE41-NEXT:    setne %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length64_eq:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE2-NEXT:    movdqu 48(%rdi), %xmm3
+; X64-SSE2-NEXT:    movdqu (%rsi), %xmm4
+; X64-SSE2-NEXT:    pcmpeqb %xmm0, %xmm4
+; X64-SSE2-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE2-NEXT:    pcmpeqb %xmm1, %xmm0
+; X64-SSE2-NEXT:    movdqu 32(%rsi), %xmm1
+; X64-SSE2-NEXT:    pcmpeqb %xmm2, %xmm1
+; X64-SSE2-NEXT:    movdqu 48(%rsi), %xmm2
+; X64-SSE2-NEXT:    pcmpeqb %xmm3, %xmm2
+; X64-SSE2-NEXT:    pand %xmm1, %xmm2
+; X64-SSE2-NEXT:    pand %xmm0, %xmm2
+; X64-SSE2-NEXT:    pand %xmm4, %xmm2
+; X64-SSE2-NEXT:    pmovmskb %xmm2, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    setne %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length64_eq:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE41-NEXT:    movdqu 48(%rdi), %xmm3
+; X64-SSE41-NEXT:    movdqu (%rsi), %xmm4
+; X64-SSE41-NEXT:    pxor %xmm0, %xmm4
+; X64-SSE41-NEXT:    movdqu 16(%rsi), %xmm0
+; X64-SSE41-NEXT:    pxor %xmm1, %xmm0
+; X64-SSE41-NEXT:    movdqu 32(%rsi), %xmm1
+; X64-SSE41-NEXT:    pxor %xmm2, %xmm1
+; X64-SSE41-NEXT:    movdqu 48(%rsi), %xmm2
+; X64-SSE41-NEXT:    pxor %xmm3, %xmm2
+; X64-SSE41-NEXT:    por %xmm1, %xmm2
+; X64-SSE41-NEXT:    por %xmm0, %xmm2
+; X64-SSE41-NEXT:    por %xmm4, %xmm2
+; X64-SSE41-NEXT:    ptest %xmm2, %xmm2
+; X64-SSE41-NEXT:    setne %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX1-LABEL: length64_eq:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX1-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-AVX1-NEXT:    vmovdqu 32(%rdi), %xmm2
+; X64-AVX1-NEXT:    vmovdqu 48(%rdi), %xmm3
+; X64-AVX1-NEXT:    vpxor 48(%rsi), %xmm3, %xmm3
+; X64-AVX1-NEXT:    vpxor 32(%rsi), %xmm2, %xmm2
+; X64-AVX1-NEXT:    vpor %xmm3, %xmm2, %xmm2
+; X64-AVX1-NEXT:    vpxor 16(%rsi), %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpor %xmm2, %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpxor (%rsi), %xmm0, %xmm0
+; X64-AVX1-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX1-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX1-NEXT:    setne %al
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length64_eq:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-AVX2-NEXT:    vpxor 32(%rsi), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpxor (%rsi), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length64_eq:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vpcmpneqb (%rsi), %zmm0, %k0
+; X64-AVX512BW-NEXT:    kortestq %k0, %k0
+; X64-AVX512BW-NEXT:    setne %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length64_eq:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k0
+; X64-AVX512F-NEXT:    kortestw %k0, %k0
+; X64-AVX512F-NEXT:    setne %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length64_eq:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rsi), %ymm2
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rsi), %ymm3
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm3, %zmm1, %k0
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm2, %zmm0, %k1
+; X64-MIC-AVX2-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX2-NEXT:    setne %al
+; X64-MIC-AVX2-NEXT:    vzeroupper
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length64_eq:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k0
+; X64-MIC-AVX512F-NEXT:    kortestw %k0, %k0
+; X64-MIC-AVX512F-NEXT:    setne %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 64) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length64_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length64_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $64
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length64_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $64, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 64) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length64_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length64_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $64
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length64_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $64, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 64) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length64_eq_const(i8* %X) nounwind {
+; X86-NOSSE-LABEL: length64_eq_const:
+; X86-NOSSE:       # %bb.0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $64
+; X86-NOSSE-NEXT:    pushl $.L.str
+; X86-NOSSE-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NOSSE-NEXT:    calll memcmp
+; X86-NOSSE-NEXT:    addl $16, %esp
+; X86-NOSSE-NEXT:    testl %eax, %eax
+; X86-NOSSE-NEXT:    sete %al
+; X86-NOSSE-NEXT:    retl
+;
+; X86-SSE1-LABEL: length64_eq_const:
+; X86-SSE1:       # %bb.0:
+; X86-SSE1-NEXT:    pushl $0
+; X86-SSE1-NEXT:    pushl $64
+; X86-SSE1-NEXT:    pushl $.L.str
+; X86-SSE1-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-SSE1-NEXT:    calll memcmp
+; X86-SSE1-NEXT:    addl $16, %esp
+; X86-SSE1-NEXT:    testl %eax, %eax
+; X86-SSE1-NEXT:    sete %al
+; X86-SSE1-NEXT:    retl
+;
+; X86-SSE2-LABEL: length64_eq_const:
+; X86-SSE2:       # %bb.0:
+; X86-SSE2-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE2-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE2-NEXT:    movdqu 16(%eax), %xmm1
+; X86-SSE2-NEXT:    movdqu 32(%eax), %xmm2
+; X86-SSE2-NEXT:    movdqu 48(%eax), %xmm3
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm3
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm2
+; X86-SSE2-NEXT:    pand %xmm3, %xmm2
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm1
+; X86-SSE2-NEXT:    pand %xmm2, %xmm1
+; X86-SSE2-NEXT:    pcmpeqb {{\.LCPI.*}}, %xmm0
+; X86-SSE2-NEXT:    pand %xmm1, %xmm0
+; X86-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X86-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X86-SSE2-NEXT:    sete %al
+; X86-SSE2-NEXT:    retl
+;
+; X86-SSE41-LABEL: length64_eq_const:
+; X86-SSE41:       # %bb.0:
+; X86-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-SSE41-NEXT:    movdqu (%eax), %xmm0
+; X86-SSE41-NEXT:    movdqu 16(%eax), %xmm1
+; X86-SSE41-NEXT:    movdqu 32(%eax), %xmm2
+; X86-SSE41-NEXT:    movdqu 48(%eax), %xmm3
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm3
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm2
+; X86-SSE41-NEXT:    por %xmm3, %xmm2
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm1
+; X86-SSE41-NEXT:    por %xmm2, %xmm1
+; X86-SSE41-NEXT:    pxor {{\.LCPI.*}}, %xmm0
+; X86-SSE41-NEXT:    por %xmm1, %xmm0
+; X86-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X86-SSE41-NEXT:    sete %al
+; X86-SSE41-NEXT:    retl
+;
+; X64-SSE2-LABEL: length64_eq_const:
+; X64-SSE2:       # %bb.0:
+; X64-SSE2-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE2-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE2-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE2-NEXT:    movdqu 48(%rdi), %xmm3
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm3
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm2
+; X64-SSE2-NEXT:    pand %xmm3, %xmm2
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm1
+; X64-SSE2-NEXT:    pand %xmm2, %xmm1
+; X64-SSE2-NEXT:    pcmpeqb {{.*}}(%rip), %xmm0
+; X64-SSE2-NEXT:    pand %xmm1, %xmm0
+; X64-SSE2-NEXT:    pmovmskb %xmm0, %eax
+; X64-SSE2-NEXT:    cmpl $65535, %eax # imm = 0xFFFF
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    retq
+;
+; X64-SSE41-LABEL: length64_eq_const:
+; X64-SSE41:       # %bb.0:
+; X64-SSE41-NEXT:    movdqu (%rdi), %xmm0
+; X64-SSE41-NEXT:    movdqu 16(%rdi), %xmm1
+; X64-SSE41-NEXT:    movdqu 32(%rdi), %xmm2
+; X64-SSE41-NEXT:    movdqu 48(%rdi), %xmm3
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm3
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm2
+; X64-SSE41-NEXT:    por %xmm3, %xmm2
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm1
+; X64-SSE41-NEXT:    por %xmm2, %xmm1
+; X64-SSE41-NEXT:    pxor {{.*}}(%rip), %xmm0
+; X64-SSE41-NEXT:    por %xmm1, %xmm0
+; X64-SSE41-NEXT:    ptest %xmm0, %xmm0
+; X64-SSE41-NEXT:    sete %al
+; X64-SSE41-NEXT:    retq
+;
+; X64-AVX1-LABEL: length64_eq_const:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    vmovdqu (%rdi), %xmm0
+; X64-AVX1-NEXT:    vmovdqu 16(%rdi), %xmm1
+; X64-AVX1-NEXT:    vmovdqu 32(%rdi), %xmm2
+; X64-AVX1-NEXT:    vmovdqu 48(%rdi), %xmm3
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm3, %xmm3
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm2, %xmm2
+; X64-AVX1-NEXT:    vpor %xmm3, %xmm2, %xmm2
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpor %xmm2, %xmm1, %xmm1
+; X64-AVX1-NEXT:    vpxor {{.*}}(%rip), %xmm0, %xmm0
+; X64-AVX1-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; X64-AVX1-NEXT:    vptest %xmm0, %xmm0
+; X64-AVX1-NEXT:    sete %al
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length64_eq_const:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length64_eq_const:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vpcmpneqb {{.*}}(%rip), %zmm0, %k0
+; X64-AVX512BW-NEXT:    kortestq %k0, %k0
+; X64-AVX512BW-NEXT:    sete %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length64_eq_const:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k0
+; X64-AVX512F-NEXT:    kortestw %k0, %k0
+; X64-AVX512F-NEXT:    sete %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length64_eq_const:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [892613426,959985462,858927408,926299444,825243960,892613426,959985462,858927408]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm2, %zmm1, %k0
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [858927408,926299444,825243960,892613426,959985462,858927408,926299444,825243960]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm1, %zmm0, %k1
+; X64-MIC-AVX2-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX2-NEXT:    sete %al
+; X64-MIC-AVX2-NEXT:    vzeroupper
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length64_eq_const:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k0
+; X64-MIC-AVX512F-NEXT:    kortestw %k0, %k0
+; X64-MIC-AVX512F-NEXT:    sete %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 64) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length96(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length96:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $96
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length96:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $96, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 96) nounwind
+  ret i32 %m
+}
+
+define i1 @length96_eq(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length96_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $96
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length96_eq:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $96, %edx
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    setne %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length96_eq:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $96, %edx
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    setne %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length96_eq:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-AVX2-NEXT:    vpxor 32(%rsi), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpxor (%rsi), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpxor 64(%rsi), %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpor %ymm2, %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length96_eq:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu 64(%rdi), %ymm1
+; X64-AVX512BW-NEXT:    vmovdqu 64(%rsi), %ymm2
+; X64-AVX512BW-NEXT:    vpcmpneqb (%rsi), %zmm0, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb %zmm2, %zmm1, %k1
+; X64-AVX512BW-NEXT:    kortestq %k1, %k0
+; X64-AVX512BW-NEXT:    setne %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length96_eq:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu 64(%rdi), %ymm1
+; X64-AVX512F-NEXT:    vmovdqu 64(%rsi), %ymm2
+; X64-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd %zmm2, %zmm1, %k1
+; X64-AVX512F-NEXT:    kortestw %k1, %k0
+; X64-AVX512F-NEXT:    setne %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length96_eq:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-MIC-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rsi), %ymm3
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rsi), %ymm4
+; X64-MIC-AVX2-NEXT:    vmovdqu 64(%rsi), %ymm5
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm4, %zmm1, %k0
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm3, %zmm0, %k1
+; X64-MIC-AVX2-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm5, %zmm2, %k1
+; X64-MIC-AVX2-NEXT:    kortestw %k1, %k0
+; X64-MIC-AVX2-NEXT:    setne %al
+; X64-MIC-AVX2-NEXT:    vzeroupper
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length96_eq:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu 64(%rdi), %ymm1
+; X64-MIC-AVX512F-NEXT:    vmovdqu 64(%rsi), %ymm2
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd %zmm2, %zmm1, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k1, %k0
+; X64-MIC-AVX512F-NEXT:    setne %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 96) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length96_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length96_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $96
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length96_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $96, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 96) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length96_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length96_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $96
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length96_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $96, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 96) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length96_eq_const(i8* %X) nounwind {
+; X86-LABEL: length96_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $96
+; X86-NEXT:    pushl $.L.str
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length96_eq_const:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $.L.str, %esi
+; X64-SSE-NEXT:    movl $96, %edx
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    sete %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length96_eq_const:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $.L.str, %esi
+; X64-AVX1-NEXT:    movl $96, %edx
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    sete %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length96_eq_const:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpor %ymm2, %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length96_eq_const:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu 64(%rdi), %ymm1
+; X64-AVX512BW-NEXT:    vpcmpneqb {{.*}}(%rip), %zmm0, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb {{.*}}(%rip), %zmm1, %k1
+; X64-AVX512BW-NEXT:    kortestq %k1, %k0
+; X64-AVX512BW-NEXT:    sete %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length96_eq_const:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu 64(%rdi), %ymm1
+; X64-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm1, %k1
+; X64-AVX512F-NEXT:    kortestw %k1, %k0
+; X64-AVX512F-NEXT:    sete %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length96_eq_const:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-MIC-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm3 = [892613426,959985462,858927408,926299444,825243960,892613426,959985462,858927408]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm3, %zmm1, %k0
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [858927408,926299444,825243960,892613426,959985462,858927408,926299444,825243960]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm1, %zmm0, %k1
+; X64-MIC-AVX2-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm0 = [926299444,825243960,892613426,959985462,858927408,926299444,825243960,892613426]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm0, %zmm2, %k1
+; X64-MIC-AVX2-NEXT:    kortestw %k1, %k0
+; X64-MIC-AVX2-NEXT:    sete %al
+; X64-MIC-AVX2-NEXT:    vzeroupper
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length96_eq_const:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu 64(%rdi), %ymm1
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm1, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k1, %k0
+; X64-MIC-AVX512F-NEXT:    sete %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 96) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length127(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length127:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $127
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length127:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $127, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 127) nounwind
+  ret i32 %m
+}
+
+define i1 @length127_eq(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length127_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $127
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length127_eq:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $127, %edx
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    setne %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length127_eq:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $127, %edx
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    setne %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length127_eq:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-AVX2-NEXT:    vmovdqu 95(%rdi), %ymm3
+; X64-AVX2-NEXT:    vpxor 95(%rsi), %ymm3, %ymm3
+; X64-AVX2-NEXT:    vpxor 64(%rsi), %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpor %ymm3, %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpxor 32(%rsi), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpor %ymm2, %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpxor (%rsi), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length127_eq:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu64 63(%rdi), %zmm1
+; X64-AVX512BW-NEXT:    vpcmpneqb 63(%rsi), %zmm1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb (%rsi), %zmm0, %k1
+; X64-AVX512BW-NEXT:    kortestq %k0, %k1
+; X64-AVX512BW-NEXT:    setne %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length127_eq:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu64 63(%rdi), %zmm1
+; X64-AVX512F-NEXT:    vpcmpneqd 63(%rsi), %zmm1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k1
+; X64-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-AVX512F-NEXT:    setne %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length127_eq:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-MIC-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-MIC-AVX2-NEXT:    vmovdqu 95(%rdi), %ymm3
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rsi), %ymm4
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rsi), %ymm5
+; X64-MIC-AVX2-NEXT:    vmovdqu 64(%rsi), %ymm6
+; X64-MIC-AVX2-NEXT:    vmovdqu 95(%rsi), %ymm7
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm7, %zmm3, %k0
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm6, %zmm2, %k1
+; X64-MIC-AVX2-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm5, %zmm1, %k1
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm4, %zmm0, %k2
+; X64-MIC-AVX2-NEXT:    korw %k1, %k2, %k1
+; X64-MIC-AVX2-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX2-NEXT:    setne %al
+; X64-MIC-AVX2-NEXT:    vzeroupper
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length127_eq:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 63(%rdi), %zmm1
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd 63(%rsi), %zmm1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX512F-NEXT:    setne %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 127) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length127_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length127_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $127
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length127_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $127, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 127) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length127_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length127_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $127
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length127_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $127, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 127) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length127_eq_const(i8* %X) nounwind {
+; X86-LABEL: length127_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $127
+; X86-NEXT:    pushl $.L.str
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length127_eq_const:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $.L.str, %esi
+; X64-SSE-NEXT:    movl $127, %edx
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    sete %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length127_eq_const:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $.L.str, %esi
+; X64-AVX1-NEXT:    movl $127, %edx
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    sete %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length127_eq_const:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-AVX2-NEXT:    vmovdqu 95(%rdi), %ymm3
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm3, %ymm3
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpor %ymm3, %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpor %ymm2, %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length127_eq_const:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu64 63(%rdi), %zmm1
+; X64-AVX512BW-NEXT:    vpcmpneqb .L.str+{{.*}}(%rip), %zmm1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb {{.*}}(%rip), %zmm0, %k1
+; X64-AVX512BW-NEXT:    kortestq %k0, %k1
+; X64-AVX512BW-NEXT:    sete %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length127_eq_const:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu64 63(%rdi), %zmm1
+; X64-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k1
+; X64-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-AVX512F-NEXT:    sete %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length127_eq_const:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-MIC-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-MIC-AVX2-NEXT:    vmovdqu 95(%rdi), %ymm3
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm4 = [943142453,842084409,909456435,809056311,875770417,943142453,842084409,909456435]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm4, %zmm3, %k0
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm3 = [926299444,825243960,892613426,959985462,858927408,926299444,825243960,892613426]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm3, %zmm2, %k1
+; X64-MIC-AVX2-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [892613426,959985462,858927408,926299444,825243960,892613426,959985462,858927408]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm2, %zmm1, %k1
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [858927408,926299444,825243960,892613426,959985462,858927408,926299444,825243960]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm1, %zmm0, %k2
+; X64-MIC-AVX2-NEXT:    korw %k1, %k2, %k1
+; X64-MIC-AVX2-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX2-NEXT:    sete %al
+; X64-MIC-AVX2-NEXT:    vzeroupper
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length127_eq_const:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 63(%rdi), %zmm1
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX512F-NEXT:    sete %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 127) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length128(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length128:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $128
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length128:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $128, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 128) nounwind
+  ret i32 %m
+}
+
+define i1 @length128_eq(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length128_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $128
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length128_eq:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $128, %edx
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    setne %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length128_eq:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $128, %edx
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    setne %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length128_eq:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-AVX2-NEXT:    vmovdqu 96(%rdi), %ymm3
+; X64-AVX2-NEXT:    vpxor 96(%rsi), %ymm3, %ymm3
+; X64-AVX2-NEXT:    vpxor 64(%rsi), %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpor %ymm3, %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpxor 32(%rsi), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpor %ymm2, %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpxor (%rsi), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length128_eq:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512BW-NEXT:    vpcmpneqb 64(%rsi), %zmm1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb (%rsi), %zmm0, %k1
+; X64-AVX512BW-NEXT:    kortestq %k0, %k1
+; X64-AVX512BW-NEXT:    setne %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length128_eq:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512F-NEXT:    vpcmpneqd 64(%rsi), %zmm1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k1
+; X64-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-AVX512F-NEXT:    setne %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length128_eq:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-MIC-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-MIC-AVX2-NEXT:    vmovdqu 96(%rdi), %ymm3
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rsi), %ymm4
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rsi), %ymm5
+; X64-MIC-AVX2-NEXT:    vmovdqu 64(%rsi), %ymm6
+; X64-MIC-AVX2-NEXT:    vmovdqu 96(%rsi), %ymm7
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm7, %zmm3, %k0
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm6, %zmm2, %k1
+; X64-MIC-AVX2-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm5, %zmm1, %k1
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm4, %zmm0, %k2
+; X64-MIC-AVX2-NEXT:    korw %k1, %k2, %k1
+; X64-MIC-AVX2-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX2-NEXT:    setne %al
+; X64-MIC-AVX2-NEXT:    vzeroupper
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length128_eq:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd 64(%rsi), %zmm1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX512F-NEXT:    setne %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 128) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length128_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length128_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $128
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length128_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $128, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 128) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length128_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length128_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $128
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length128_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $128, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 128) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length128_eq_const(i8* %X) nounwind {
+; X86-LABEL: length128_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $128
+; X86-NEXT:    pushl $.L.str
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length128_eq_const:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $.L.str, %esi
+; X64-SSE-NEXT:    movl $128, %edx
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    sete %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length128_eq_const:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $.L.str, %esi
+; X64-AVX1-NEXT:    movl $128, %edx
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    sete %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length128_eq_const:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-AVX2-NEXT:    vmovdqu 96(%rdi), %ymm3
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm3, %ymm3
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpor %ymm3, %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpor %ymm2, %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpxor {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vptest %ymm0, %ymm0
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length128_eq_const:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512BW-NEXT:    vpcmpneqb .L.str+{{.*}}(%rip), %zmm1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb {{.*}}(%rip), %zmm0, %k1
+; X64-AVX512BW-NEXT:    kortestq %k0, %k1
+; X64-AVX512BW-NEXT:    sete %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length128_eq_const:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k1
+; X64-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-AVX512F-NEXT:    sete %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length128_eq_const:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-MIC-AVX2-NEXT:    vmovdqu 32(%rdi), %ymm1
+; X64-MIC-AVX2-NEXT:    vmovdqu 64(%rdi), %ymm2
+; X64-MIC-AVX2-NEXT:    vmovdqu 96(%rdi), %ymm3
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm4 = [959985462,858927408,926299444,825243960,892613426,959985462,858927408,926299444]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm4, %zmm3, %k0
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm3 = [926299444,825243960,892613426,959985462,858927408,926299444,825243960,892613426]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm3, %zmm2, %k1
+; X64-MIC-AVX2-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [892613426,959985462,858927408,926299444,825243960,892613426,959985462,858927408]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm2, %zmm1, %k1
+; X64-MIC-AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [858927408,926299444,825243960,892613426,959985462,858927408,926299444,825243960]
+; X64-MIC-AVX2-NEXT:    vpcmpneqd %zmm1, %zmm0, %k2
+; X64-MIC-AVX2-NEXT:    korw %k1, %k2, %k1
+; X64-MIC-AVX2-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX2-NEXT:    sete %al
+; X64-MIC-AVX2-NEXT:    vzeroupper
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length128_eq_const:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX512F-NEXT:    sete %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 128) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length192(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length192:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $192
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length192:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $192, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 192) nounwind
+  ret i32 %m
+}
+
+define i1 @length192_eq(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length192_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $192
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length192_eq:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $192, %edx
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    setne %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length192_eq:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $192, %edx
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    setne %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length192_eq:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    pushq %rax
+; X64-AVX2-NEXT:    movl $192, %edx
+; X64-AVX2-NEXT:    callq memcmp
+; X64-AVX2-NEXT:    testl %eax, %eax
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    popq %rcx
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length192_eq:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512BW-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512BW-NEXT:    vpcmpneqb 64(%rsi), %zmm1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb (%rsi), %zmm0, %k1
+; X64-AVX512BW-NEXT:    korq %k0, %k1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb 128(%rsi), %zmm2, %k1
+; X64-AVX512BW-NEXT:    kortestq %k1, %k0
+; X64-AVX512BW-NEXT:    setne %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length192_eq:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512F-NEXT:    vpcmpneqd 64(%rsi), %zmm1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k1
+; X64-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd 128(%rsi), %zmm2, %k1
+; X64-AVX512F-NEXT:    kortestw %k1, %k0
+; X64-AVX512F-NEXT:    setne %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length192_eq:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    pushq %rax
+; X64-MIC-AVX2-NEXT:    movl $192, %edx
+; X64-MIC-AVX2-NEXT:    callq memcmp
+; X64-MIC-AVX2-NEXT:    testl %eax, %eax
+; X64-MIC-AVX2-NEXT:    setne %al
+; X64-MIC-AVX2-NEXT:    popq %rcx
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length192_eq:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd 64(%rsi), %zmm1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k1
+; X64-MIC-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd 128(%rsi), %zmm2, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k1, %k0
+; X64-MIC-AVX512F-NEXT:    setne %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 192) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length192_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length192_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $192
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length192_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $192, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 192) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length192_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length192_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $192
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length192_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $192, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 192) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length192_eq_const(i8* %X) nounwind {
+; X86-LABEL: length192_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $192
+; X86-NEXT:    pushl $.L.str
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length192_eq_const:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $.L.str, %esi
+; X64-SSE-NEXT:    movl $192, %edx
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    sete %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length192_eq_const:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $.L.str, %esi
+; X64-AVX1-NEXT:    movl $192, %edx
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    sete %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length192_eq_const:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    pushq %rax
+; X64-AVX2-NEXT:    movl $.L.str, %esi
+; X64-AVX2-NEXT:    movl $192, %edx
+; X64-AVX2-NEXT:    callq memcmp
+; X64-AVX2-NEXT:    testl %eax, %eax
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    popq %rcx
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length192_eq_const:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512BW-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512BW-NEXT:    vpcmpneqb .L.str+{{.*}}(%rip), %zmm1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb {{.*}}(%rip), %zmm0, %k1
+; X64-AVX512BW-NEXT:    korq %k0, %k1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb .L.str+{{.*}}(%rip), %zmm2, %k1
+; X64-AVX512BW-NEXT:    kortestq %k1, %k0
+; X64-AVX512BW-NEXT:    sete %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length192_eq_const:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k1
+; X64-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm2, %k1
+; X64-AVX512F-NEXT:    kortestw %k1, %k0
+; X64-AVX512F-NEXT:    sete %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length192_eq_const:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    pushq %rax
+; X64-MIC-AVX2-NEXT:    movl $.L.str, %esi
+; X64-MIC-AVX2-NEXT:    movl $192, %edx
+; X64-MIC-AVX2-NEXT:    callq memcmp
+; X64-MIC-AVX2-NEXT:    testl %eax, %eax
+; X64-MIC-AVX2-NEXT:    sete %al
+; X64-MIC-AVX2-NEXT:    popq %rcx
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length192_eq_const:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k1
+; X64-MIC-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm2, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k1, %k0
+; X64-MIC-AVX512F-NEXT:    sete %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 192) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length255(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length255:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $255
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length255:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $255, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 255) nounwind
+  ret i32 %m
+}
+
+define i1 @length255_eq(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length255_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $255
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length255_eq:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $255, %edx
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    setne %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length255_eq:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $255, %edx
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    setne %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length255_eq:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    pushq %rax
+; X64-AVX2-NEXT:    movl $255, %edx
+; X64-AVX2-NEXT:    callq memcmp
+; X64-AVX2-NEXT:    testl %eax, %eax
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    popq %rcx
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length255_eq:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512BW-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512BW-NEXT:    vmovdqu64 191(%rdi), %zmm3
+; X64-AVX512BW-NEXT:    vpcmpneqb 191(%rsi), %zmm3, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb 128(%rsi), %zmm2, %k1
+; X64-AVX512BW-NEXT:    korq %k0, %k1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb 64(%rsi), %zmm1, %k1
+; X64-AVX512BW-NEXT:    vpcmpneqb (%rsi), %zmm0, %k2
+; X64-AVX512BW-NEXT:    korq %k1, %k2, %k1
+; X64-AVX512BW-NEXT:    kortestq %k0, %k1
+; X64-AVX512BW-NEXT:    setne %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length255_eq:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512F-NEXT:    vmovdqu64 191(%rdi), %zmm3
+; X64-AVX512F-NEXT:    vpcmpneqd 191(%rsi), %zmm3, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd 128(%rsi), %zmm2, %k1
+; X64-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd 64(%rsi), %zmm1, %k1
+; X64-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k2
+; X64-AVX512F-NEXT:    korw %k1, %k2, %k1
+; X64-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-AVX512F-NEXT:    setne %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length255_eq:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    pushq %rax
+; X64-MIC-AVX2-NEXT:    movl $255, %edx
+; X64-MIC-AVX2-NEXT:    callq memcmp
+; X64-MIC-AVX2-NEXT:    testl %eax, %eax
+; X64-MIC-AVX2-NEXT:    setne %al
+; X64-MIC-AVX2-NEXT:    popq %rcx
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length255_eq:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 191(%rdi), %zmm3
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd 191(%rsi), %zmm3, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd 128(%rsi), %zmm2, %k1
+; X64-MIC-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd 64(%rsi), %zmm1, %k1
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k2
+; X64-MIC-AVX512F-NEXT:    korw %k1, %k2, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX512F-NEXT:    setne %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 255) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length255_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length255_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $255
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length255_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $255, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 255) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length255_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length255_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $255
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length255_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $255, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 255) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length255_eq_const(i8* %X) nounwind {
+; X86-LABEL: length255_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $255
+; X86-NEXT:    pushl $.L.str
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length255_eq_const:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $.L.str, %esi
+; X64-SSE-NEXT:    movl $255, %edx
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    sete %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length255_eq_const:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $.L.str, %esi
+; X64-AVX1-NEXT:    movl $255, %edx
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    sete %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length255_eq_const:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    pushq %rax
+; X64-AVX2-NEXT:    movl $.L.str, %esi
+; X64-AVX2-NEXT:    movl $255, %edx
+; X64-AVX2-NEXT:    callq memcmp
+; X64-AVX2-NEXT:    testl %eax, %eax
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    popq %rcx
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length255_eq_const:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512BW-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512BW-NEXT:    vmovdqu64 191(%rdi), %zmm3
+; X64-AVX512BW-NEXT:    vpcmpneqb .L.str+{{.*}}(%rip), %zmm3, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb .L.str+{{.*}}(%rip), %zmm2, %k1
+; X64-AVX512BW-NEXT:    korq %k0, %k1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb .L.str+{{.*}}(%rip), %zmm1, %k1
+; X64-AVX512BW-NEXT:    vpcmpneqb {{.*}}(%rip), %zmm0, %k2
+; X64-AVX512BW-NEXT:    korq %k1, %k2, %k1
+; X64-AVX512BW-NEXT:    kortestq %k0, %k1
+; X64-AVX512BW-NEXT:    sete %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length255_eq_const:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512F-NEXT:    vmovdqu64 191(%rdi), %zmm3
+; X64-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm3, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm2, %k1
+; X64-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm1, %k1
+; X64-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k2
+; X64-AVX512F-NEXT:    korw %k1, %k2, %k1
+; X64-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-AVX512F-NEXT:    sete %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length255_eq_const:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    pushq %rax
+; X64-MIC-AVX2-NEXT:    movl $.L.str, %esi
+; X64-MIC-AVX2-NEXT:    movl $255, %edx
+; X64-MIC-AVX2-NEXT:    callq memcmp
+; X64-MIC-AVX2-NEXT:    testl %eax, %eax
+; X64-MIC-AVX2-NEXT:    sete %al
+; X64-MIC-AVX2-NEXT:    popq %rcx
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length255_eq_const:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 191(%rdi), %zmm3
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm3, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm2, %k1
+; X64-MIC-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm1, %k1
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k2
+; X64-MIC-AVX512F-NEXT:    korw %k1, %k2, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX512F-NEXT:    sete %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 255) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length256(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length256:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $256 # imm = 0x100
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length256:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $256, %edx # imm = 0x100
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 256) nounwind
+  ret i32 %m
+}
+
+define i1 @length256_eq(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length256_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $256 # imm = 0x100
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length256_eq:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $256, %edx # imm = 0x100
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    setne %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length256_eq:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $256, %edx # imm = 0x100
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    setne %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length256_eq:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    pushq %rax
+; X64-AVX2-NEXT:    movl $256, %edx # imm = 0x100
+; X64-AVX2-NEXT:    callq memcmp
+; X64-AVX2-NEXT:    testl %eax, %eax
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    popq %rcx
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length256_eq:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512BW-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512BW-NEXT:    vmovdqu64 192(%rdi), %zmm3
+; X64-AVX512BW-NEXT:    vpcmpneqb 192(%rsi), %zmm3, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb 128(%rsi), %zmm2, %k1
+; X64-AVX512BW-NEXT:    korq %k0, %k1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb 64(%rsi), %zmm1, %k1
+; X64-AVX512BW-NEXT:    vpcmpneqb (%rsi), %zmm0, %k2
+; X64-AVX512BW-NEXT:    korq %k1, %k2, %k1
+; X64-AVX512BW-NEXT:    kortestq %k0, %k1
+; X64-AVX512BW-NEXT:    setne %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length256_eq:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512F-NEXT:    vmovdqu64 192(%rdi), %zmm3
+; X64-AVX512F-NEXT:    vpcmpneqd 192(%rsi), %zmm3, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd 128(%rsi), %zmm2, %k1
+; X64-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd 64(%rsi), %zmm1, %k1
+; X64-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k2
+; X64-AVX512F-NEXT:    korw %k1, %k2, %k1
+; X64-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-AVX512F-NEXT:    setne %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length256_eq:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    pushq %rax
+; X64-MIC-AVX2-NEXT:    movl $256, %edx # imm = 0x100
+; X64-MIC-AVX2-NEXT:    callq memcmp
+; X64-MIC-AVX2-NEXT:    testl %eax, %eax
+; X64-MIC-AVX2-NEXT:    setne %al
+; X64-MIC-AVX2-NEXT:    popq %rcx
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length256_eq:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 192(%rdi), %zmm3
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd 192(%rsi), %zmm3, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd 128(%rsi), %zmm2, %k1
+; X64-MIC-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd 64(%rsi), %zmm1, %k1
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd (%rsi), %zmm0, %k2
+; X64-MIC-AVX512F-NEXT:    korw %k1, %k2, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX512F-NEXT:    setne %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 256) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length256_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length256_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $256 # imm = 0x100
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length256_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $256, %edx # imm = 0x100
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 256) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length256_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length256_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $256 # imm = 0x100
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length256_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $256, %edx # imm = 0x100
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 256) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length256_eq_const(i8* %X) nounwind {
+; X86-LABEL: length256_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $256 # imm = 0x100
+; X86-NEXT:    pushl $.L.str
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-SSE-LABEL: length256_eq_const:
+; X64-SSE:       # %bb.0:
+; X64-SSE-NEXT:    pushq %rax
+; X64-SSE-NEXT:    movl $.L.str, %esi
+; X64-SSE-NEXT:    movl $256, %edx # imm = 0x100
+; X64-SSE-NEXT:    callq memcmp
+; X64-SSE-NEXT:    testl %eax, %eax
+; X64-SSE-NEXT:    sete %al
+; X64-SSE-NEXT:    popq %rcx
+; X64-SSE-NEXT:    retq
+;
+; X64-AVX1-LABEL: length256_eq_const:
+; X64-AVX1:       # %bb.0:
+; X64-AVX1-NEXT:    pushq %rax
+; X64-AVX1-NEXT:    movl $.L.str, %esi
+; X64-AVX1-NEXT:    movl $256, %edx # imm = 0x100
+; X64-AVX1-NEXT:    callq memcmp
+; X64-AVX1-NEXT:    testl %eax, %eax
+; X64-AVX1-NEXT:    sete %al
+; X64-AVX1-NEXT:    popq %rcx
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: length256_eq_const:
+; X64-AVX2:       # %bb.0:
+; X64-AVX2-NEXT:    pushq %rax
+; X64-AVX2-NEXT:    movl $.L.str, %esi
+; X64-AVX2-NEXT:    movl $256, %edx # imm = 0x100
+; X64-AVX2-NEXT:    callq memcmp
+; X64-AVX2-NEXT:    testl %eax, %eax
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    popq %rcx
+; X64-AVX2-NEXT:    retq
+;
+; X64-AVX512BW-LABEL: length256_eq_const:
+; X64-AVX512BW:       # %bb.0:
+; X64-AVX512BW-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512BW-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512BW-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512BW-NEXT:    vmovdqu64 192(%rdi), %zmm3
+; X64-AVX512BW-NEXT:    vpcmpneqb .L.str+{{.*}}(%rip), %zmm3, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb .L.str+{{.*}}(%rip), %zmm2, %k1
+; X64-AVX512BW-NEXT:    korq %k0, %k1, %k0
+; X64-AVX512BW-NEXT:    vpcmpneqb .L.str+{{.*}}(%rip), %zmm1, %k1
+; X64-AVX512BW-NEXT:    vpcmpneqb {{.*}}(%rip), %zmm0, %k2
+; X64-AVX512BW-NEXT:    korq %k1, %k2, %k1
+; X64-AVX512BW-NEXT:    kortestq %k0, %k1
+; X64-AVX512BW-NEXT:    sete %al
+; X64-AVX512BW-NEXT:    vzeroupper
+; X64-AVX512BW-NEXT:    retq
+;
+; X64-AVX512F-LABEL: length256_eq_const:
+; X64-AVX512F:       # %bb.0:
+; X64-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-AVX512F-NEXT:    vmovdqu64 192(%rdi), %zmm3
+; X64-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm3, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm2, %k1
+; X64-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm1, %k1
+; X64-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k2
+; X64-AVX512F-NEXT:    korw %k1, %k2, %k1
+; X64-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-AVX512F-NEXT:    sete %al
+; X64-AVX512F-NEXT:    vzeroupper
+; X64-AVX512F-NEXT:    retq
+;
+; X64-MIC-AVX2-LABEL: length256_eq_const:
+; X64-MIC-AVX2:       # %bb.0:
+; X64-MIC-AVX2-NEXT:    pushq %rax
+; X64-MIC-AVX2-NEXT:    movl $.L.str, %esi
+; X64-MIC-AVX2-NEXT:    movl $256, %edx # imm = 0x100
+; X64-MIC-AVX2-NEXT:    callq memcmp
+; X64-MIC-AVX2-NEXT:    testl %eax, %eax
+; X64-MIC-AVX2-NEXT:    sete %al
+; X64-MIC-AVX2-NEXT:    popq %rcx
+; X64-MIC-AVX2-NEXT:    retq
+;
+; X64-MIC-AVX512F-LABEL: length256_eq_const:
+; X64-MIC-AVX512F:       # %bb.0:
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 (%rdi), %zmm0
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 64(%rdi), %zmm1
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 128(%rdi), %zmm2
+; X64-MIC-AVX512F-NEXT:    vmovdqu64 192(%rdi), %zmm3
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm3, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm2, %k1
+; X64-MIC-AVX512F-NEXT:    korw %k0, %k1, %k0
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd .L.str+{{.*}}(%rip), %zmm1, %k1
+; X64-MIC-AVX512F-NEXT:    vpcmpneqd {{.*}}(%rip), %zmm0, %k2
+; X64-MIC-AVX512F-NEXT:    korw %k1, %k2, %k1
+; X64-MIC-AVX512F-NEXT:    kortestw %k0, %k1
+; X64-MIC-AVX512F-NEXT:    sete %al
+; X64-MIC-AVX512F-NEXT:    vzeroupper
+; X64-MIC-AVX512F-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 256) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length384(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length384:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $384 # imm = 0x180
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length384:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $384, %edx # imm = 0x180
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 384) nounwind
+  ret i32 %m
+}
+
+define i1 @length384_eq(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length384_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $384 # imm = 0x180
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length384_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $384, %edx # imm = 0x180
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setne %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 384) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length384_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length384_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $384 # imm = 0x180
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length384_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $384, %edx # imm = 0x180
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 384) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length384_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length384_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $384 # imm = 0x180
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length384_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $384, %edx # imm = 0x180
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 384) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length384_eq_const(i8* %X) nounwind {
+; X86-LABEL: length384_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $384 # imm = 0x180
+; X86-NEXT:    pushl $.L.str
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length384_eq_const:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $.L.str, %esi
+; X64-NEXT:    movl $384, %edx # imm = 0x180
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    sete %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 384) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length511(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length511:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $511 # imm = 0x1FF
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length511:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $511, %edx # imm = 0x1FF
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 511) nounwind
+  ret i32 %m
+}
+
+define i1 @length511_eq(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length511_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $511 # imm = 0x1FF
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length511_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $511, %edx # imm = 0x1FF
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setne %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 511) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length511_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length511_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $511 # imm = 0x1FF
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length511_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $511, %edx # imm = 0x1FF
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 511) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length511_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length511_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $511 # imm = 0x1FF
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length511_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $511, %edx # imm = 0x1FF
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 511) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length511_eq_const(i8* %X) nounwind {
+; X86-LABEL: length511_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $511 # imm = 0x1FF
+; X86-NEXT:    pushl $.L.str
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length511_eq_const:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $.L.str, %esi
+; X64-NEXT:    movl $511, %edx # imm = 0x1FF
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    sete %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 511) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+define i32 @length512(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: length512:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $512 # imm = 0x200
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length512:
+; X64:       # %bb.0:
+; X64-NEXT:    movl $512, %edx # imm = 0x200
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 512) nounwind
+  ret i32 %m
+}
+
+define i1 @length512_eq(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length512_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $512 # imm = 0x200
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length512_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $512, %edx # imm = 0x200
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setne %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 512) nounwind
+  %cmp = icmp ne i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length512_lt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length512_lt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $512 # imm = 0x200
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    shrl $31, %eax
+; X86-NEXT:    # kill: def $al killed $al killed $eax
+; X86-NEXT:    retl
+;
+; X64-LABEL: length512_lt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $512, %edx # imm = 0x200
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    shrl $31, %eax
+; X64-NEXT:    # kill: def $al killed $al killed $eax
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 512) nounwind
+  %cmp = icmp slt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length512_gt(i8* %x, i8* %y) nounwind {
+; X86-LABEL: length512_gt:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $512 # imm = 0x200
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setg %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length512_gt:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $512, %edx # imm = 0x200
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setg %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 512) nounwind
+  %cmp = icmp sgt i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @length512_eq_const(i8* %X) nounwind {
+; X86-LABEL: length512_eq_const:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $512 # imm = 0x200
+; X86-NEXT:    pushl $.L.str
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length512_eq_const:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $.L.str, %esi
+; X64-NEXT:    movl $512, %edx # imm = 0x200
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    sete %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([513 x i8], [513 x i8]* @.str, i32 0, i32 0), i64 512) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+; This checks that we do not do stupid things with huge sizes.
+define i32 @huge_length(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: huge_length:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $2147483647 # imm = 0x7FFFFFFF
+; X86-NEXT:    pushl $-1
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: huge_length:
+; X64:       # %bb.0:
+; X64-NEXT:    movabsq $9223372036854775807, %rdx # imm = 0x7FFFFFFFFFFFFFFF
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 9223372036854775807) nounwind
+  ret i32 %m
+}
+
+define i1 @huge_length_eq(i8* %X, i8* %Y) nounwind {
+; X86-LABEL: huge_length_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl $2147483647 # imm = 0x7FFFFFFF
+; X86-NEXT:    pushl $-1
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: huge_length_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movabsq $9223372036854775807, %rdx # imm = 0x7FFFFFFFFFFFFFFF
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    sete %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 9223372036854775807) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}
+
+; This checks non-constant sizes.
+define i32 @nonconst_length(i8* %X, i8* %Y, i64 %size) nounwind {
+; X86-LABEL: nonconst_length:
+; X86:       # %bb.0:
+; X86-NEXT:    jmp memcmp # TAILCALL
+;
+; X64-LABEL: nonconst_length:
+; X64:       # %bb.0:
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 %size) nounwind
+  ret i32 %m
+}
+
+define i1 @nonconst_length_eq(i8* %X, i8* %Y, i64 %size) nounwind {
+; X86-LABEL: nonconst_length_eq:
+; X86:       # %bb.0:
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    pushl {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: nonconst_length_eq:
+; X64:       # %bb.0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    sete %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 %size) nounwind
+  %c = icmp eq i32 %m, 0
+  ret i1 %c
+}

test/Transforms/ExpandMemCmp/X86/memcmp.ll

 ; ALL-NEXT:    [[TMP4:%.*]] = bitcast i8* [[Y:%.*]] to i16*
 ; ALL-NEXT:    [[TMP5:%.*]] = load i16, i16* [[TMP3]]
 ; ALL-NEXT:    [[TMP6:%.*]] = load i16, i16* [[TMP4]]
 ; ALL-NEXT:    [[TMP7]] = call i16 @llvm.bswap.i16(i16 [[TMP5]])
 ; ALL-NEXT:    [[TMP8]] = call i16 @llvm.bswap.i16(i16 [[TMP6]])
 ; ALL-NEXT:    [[TMP9:%.*]] = icmp eq i16 [[TMP7]], [[TMP8]]
 ; ALL-NEXT:    br i1 [[TMP9]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
 ; ALL:       loadbb1:
-; ALL-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i8 2
-; ALL-NEXT:    [[TMP11:%.*]] = getelementptr i8, i8* [[Y]], i8 2
+; ALL-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i64 2
+; ALL-NEXT:    [[TMP11:%.*]] = getelementptr i8, i8* [[Y]], i64 2
 ; ALL-NEXT:    [[TMP12:%.*]] = load i8, i8* [[TMP10]]
 ; ALL-NEXT:    [[TMP13:%.*]] = load i8, i8* [[TMP11]]
 ; ALL-NEXT:    [[TMP14:%.*]] = zext i8 [[TMP12]] to i32
 ; ALL-NEXT:    [[TMP15:%.*]] = zext i8 [[TMP13]] to i32
 ; ALL-NEXT:    [[TMP16:%.*]] = sub i32 [[TMP14]], [[TMP15]]
 ; ALL-NEXT:    br label [[ENDBLOCK]]
 ; ALL:       endblock:
 ; ALL-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP16]], [[LOADBB1]] ], [ [[TMP2]], [[RES_BLOCK]] ]

 ; ALL-NEXT:    [[TMP4:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; ALL-NEXT:    [[TMP5:%.*]] = load i32, i32* [[TMP3]]
 ; ALL-NEXT:    [[TMP6:%.*]] = load i32, i32* [[TMP4]]
 ; ALL-NEXT:    [[TMP7]] = call i32 @llvm.bswap.i32(i32 [[TMP5]])
 ; ALL-NEXT:    [[TMP8]] = call i32 @llvm.bswap.i32(i32 [[TMP6]])
 ; ALL-NEXT:    [[TMP9:%.*]] = icmp eq i32 [[TMP7]], [[TMP8]]
 ; ALL-NEXT:    br i1 [[TMP9]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
 ; ALL:       loadbb1:
-; ALL-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i8 4
-; ALL-NEXT:    [[TMP11:%.*]] = getelementptr i8, i8* [[Y]], i8 4
+; ALL-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i64 4
+; ALL-NEXT:    [[TMP11:%.*]] = getelementptr i8, i8* [[Y]], i64 4
 ; ALL-NEXT:    [[TMP12:%.*]] = load i8, i8* [[TMP10]]
 ; ALL-NEXT:    [[TMP13:%.*]] = load i8, i8* [[TMP11]]
 ; ALL-NEXT:    [[TMP14:%.*]] = zext i8 [[TMP12]] to i32
 ; ALL-NEXT:    [[TMP15:%.*]] = zext i8 [[TMP13]] to i32
 ; ALL-NEXT:    [[TMP16:%.*]] = sub i32 [[TMP14]], [[TMP15]]
 ; ALL-NEXT:    br label [[ENDBLOCK]]
 ; ALL:       endblock:
 ; ALL-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP16]], [[LOADBB1]] ], [ [[TMP2]], [[RES_BLOCK]] ]

 ; ALL-NEXT:    [[TMP4:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; ALL-NEXT:    [[TMP5:%.*]] = load i32, i32* [[TMP3]]
 ; ALL-NEXT:    [[TMP6:%.*]] = load i32, i32* [[TMP4]]
 ; ALL-NEXT:    [[TMP7]] = call i32 @llvm.bswap.i32(i32 [[TMP5]])
 ; ALL-NEXT:    [[TMP8]] = call i32 @llvm.bswap.i32(i32 [[TMP6]])
 ; ALL-NEXT:    [[TMP9:%.*]] = icmp eq i32 [[TMP7]], [[TMP8]]
 ; ALL-NEXT:    br i1 [[TMP9]], label [[LOADBB1]], label [[RES_BLOCK:%.*]]
 ; ALL:       loadbb1:
-; ALL-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i8 4
+; ALL-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i64 4
 ; ALL-NEXT:    [[TMP11:%.*]] = bitcast i8* [[TMP10]] to i16*
-; ALL-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[Y]], i8 4
+; ALL-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[Y]], i64 4
 ; ALL-NEXT:    [[TMP13:%.*]] = bitcast i8* [[TMP12]] to i16*
 ; ALL-NEXT:    [[TMP14:%.*]] = load i16, i16* [[TMP11]]
 ; ALL-NEXT:    [[TMP15:%.*]] = load i16, i16* [[TMP13]]
 ; ALL-NEXT:    [[TMP16:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP14]])
 ; ALL-NEXT:    [[TMP17:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP15]])
 ; ALL-NEXT:    [[TMP18]] = zext i16 [[TMP16]] to i32
 ; ALL-NEXT:    [[TMP19]] = zext i16 [[TMP17]] to i32
 ; ALL-NEXT:    [[TMP20:%.*]] = icmp eq i32 [[TMP18]], [[TMP19]]

 ; X32-NEXT:    [[TMP4:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X32-NEXT:    [[TMP5:%.*]] = load i32, i32* [[TMP3]]
 ; X32-NEXT:    [[TMP6:%.*]] = load i32, i32* [[TMP4]]
 ; X32-NEXT:    [[TMP7]] = call i32 @llvm.bswap.i32(i32 [[TMP5]])
 ; X32-NEXT:    [[TMP8]] = call i32 @llvm.bswap.i32(i32 [[TMP6]])
 ; X32-NEXT:    [[TMP9:%.*]] = icmp eq i32 [[TMP7]], [[TMP8]]
 ; X32-NEXT:    br i1 [[TMP9]], label [[LOADBB1]], label [[RES_BLOCK:%.*]]
 ; X32:       loadbb1:
-; X32-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i8 4
+; X32-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i64 4
 ; X32-NEXT:    [[TMP11:%.*]] = bitcast i8* [[TMP10]] to i32*
-; X32-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[Y]], i8 4
+; X32-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[Y]], i64 4
 ; X32-NEXT:    [[TMP13:%.*]] = bitcast i8* [[TMP12]] to i32*
 ; X32-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP11]]
 ; X32-NEXT:    [[TMP15:%.*]] = load i32, i32* [[TMP13]]
 ; X32-NEXT:    [[TMP16]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
 ; X32-NEXT:    [[TMP17]] = call i32 @llvm.bswap.i32(i32 [[TMP15]])
 ; X32-NEXT:    [[TMP18:%.*]] = icmp eq i32 [[TMP16]], [[TMP17]]
 ; X32-NEXT:    br i1 [[TMP18]], label [[ENDBLOCK]], label [[RES_BLOCK]]
 ; X32:       endblock:

 ; X64-NEXT:    [[TMP4:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64-NEXT:    [[TMP5:%.*]] = load i64, i64* [[TMP3]]
 ; X64-NEXT:    [[TMP6:%.*]] = load i64, i64* [[TMP4]]
 ; X64-NEXT:    [[TMP7]] = call i64 @llvm.bswap.i64(i64 [[TMP5]])
 ; X64-NEXT:    [[TMP8]] = call i64 @llvm.bswap.i64(i64 [[TMP6]])
 ; X64-NEXT:    [[TMP9:%.*]] = icmp eq i64 [[TMP7]], [[TMP8]]
 ; X64-NEXT:    br i1 [[TMP9]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
 ; X64:       loadbb1:
-; X64-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i8 8
-; X64-NEXT:    [[TMP11:%.*]] = getelementptr i8, i8* [[Y]], i8 8
+; X64-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i64 8
+; X64-NEXT:    [[TMP11:%.*]] = getelementptr i8, i8* [[Y]], i64 8
 ; X64-NEXT:    [[TMP12:%.*]] = load i8, i8* [[TMP10]]
 ; X64-NEXT:    [[TMP13:%.*]] = load i8, i8* [[TMP11]]
 ; X64-NEXT:    [[TMP14:%.*]] = zext i8 [[TMP12]] to i32
 ; X64-NEXT:    [[TMP15:%.*]] = zext i8 [[TMP13]] to i32
 ; X64-NEXT:    [[TMP16:%.*]] = sub i32 [[TMP14]], [[TMP15]]
 ; X64-NEXT:    br label [[ENDBLOCK]]
 ; X64:       endblock:
 ; X64-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP16]], [[LOADBB1]] ], [ [[TMP2]], [[RES_BLOCK]] ]

 ; X64-NEXT:    [[TMP4:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64-NEXT:    [[TMP5:%.*]] = load i64, i64* [[TMP3]]
 ; X64-NEXT:    [[TMP6:%.*]] = load i64, i64* [[TMP4]]
 ; X64-NEXT:    [[TMP7]] = call i64 @llvm.bswap.i64(i64 [[TMP5]])
 ; X64-NEXT:    [[TMP8]] = call i64 @llvm.bswap.i64(i64 [[TMP6]])
 ; X64-NEXT:    [[TMP9:%.*]] = icmp eq i64 [[TMP7]], [[TMP8]]
 ; X64-NEXT:    br i1 [[TMP9]], label [[LOADBB1]], label [[RES_BLOCK:%.*]]
 ; X64:       loadbb1:
-; X64-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i8 8
+; X64-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i64 8
 ; X64-NEXT:    [[TMP11:%.*]] = bitcast i8* [[TMP10]] to i16*
-; X64-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[Y]], i8 8
+; X64-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[Y]], i64 8
 ; X64-NEXT:    [[TMP13:%.*]] = bitcast i8* [[TMP12]] to i16*
 ; X64-NEXT:    [[TMP14:%.*]] = load i16, i16* [[TMP11]]
 ; X64-NEXT:    [[TMP15:%.*]] = load i16, i16* [[TMP13]]
 ; X64-NEXT:    [[TMP16:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP14]])
 ; X64-NEXT:    [[TMP17:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP15]])
 ; X64-NEXT:    [[TMP18]] = zext i16 [[TMP16]] to i64
 ; X64-NEXT:    [[TMP19]] = zext i16 [[TMP17]] to i64
 ; X64-NEXT:    [[TMP20:%.*]] = icmp eq i64 [[TMP18]], [[TMP19]]

 ; X64-NEXT:    [[TMP4:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64-NEXT:    [[TMP5:%.*]] = load i64, i64* [[TMP3]]
 ; X64-NEXT:    [[TMP6:%.*]] = load i64, i64* [[TMP4]]
 ; X64-NEXT:    [[TMP7]] = call i64 @llvm.bswap.i64(i64 [[TMP5]])
 ; X64-NEXT:    [[TMP8]] = call i64 @llvm.bswap.i64(i64 [[TMP6]])
 ; X64-NEXT:    [[TMP9:%.*]] = icmp eq i64 [[TMP7]], [[TMP8]]
 ; X64-NEXT:    br i1 [[TMP9]], label [[LOADBB1]], label [[RES_BLOCK:%.*]]
 ; X64:       loadbb1:
-; X64-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i8 8
+; X64-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i64 8
 ; X64-NEXT:    [[TMP11:%.*]] = bitcast i8* [[TMP10]] to i32*
-; X64-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[Y]], i8 8
+; X64-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[Y]], i64 8
 ; X64-NEXT:    [[TMP13:%.*]] = bitcast i8* [[TMP12]] to i32*
 ; X64-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP11]]
 ; X64-NEXT:    [[TMP15:%.*]] = load i32, i32* [[TMP13]]
 ; X64-NEXT:    [[TMP16:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
 ; X64-NEXT:    [[TMP17:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP15]])
 ; X64-NEXT:    [[TMP18]] = zext i32 [[TMP16]] to i64
 ; X64-NEXT:    [[TMP19]] = zext i32 [[TMP17]] to i64
 ; X64-NEXT:    [[TMP20:%.*]] = icmp eq i64 [[TMP18]], [[TMP19]]

 ; X64-NEXT:    [[TMP4:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64-NEXT:    [[TMP5:%.*]] = load i64, i64* [[TMP3]]
 ; X64-NEXT:    [[TMP6:%.*]] = load i64, i64* [[TMP4]]
 ; X64-NEXT:    [[TMP7]] = call i64 @llvm.bswap.i64(i64 [[TMP5]])
 ; X64-NEXT:    [[TMP8]] = call i64 @llvm.bswap.i64(i64 [[TMP6]])
 ; X64-NEXT:    [[TMP9:%.*]] = icmp eq i64 [[TMP7]], [[TMP8]]
 ; X64-NEXT:    br i1 [[TMP9]], label [[LOADBB1]], label [[RES_BLOCK:%.*]]
 ; X64:       loadbb1:
-; X64-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i8 8
+; X64-NEXT:    [[TMP10:%.*]] = getelementptr i8, i8* [[X]], i64 8
 ; X64-NEXT:    [[TMP11:%.*]] = bitcast i8* [[TMP10]] to i64*
-; X64-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[Y]], i8 8
+; X64-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[Y]], i64 8
 ; X64-NEXT:    [[TMP13:%.*]] = bitcast i8* [[TMP12]] to i64*
 ; X64-NEXT:    [[TMP14:%.*]] = load i64, i64* [[TMP11]]
 ; X64-NEXT:    [[TMP15:%.*]] = load i64, i64* [[TMP13]]
 ; X64-NEXT:    [[TMP16]] = call i64 @llvm.bswap.i64(i64 [[TMP14]])
 ; X64-NEXT:    [[TMP17]] = call i64 @llvm.bswap.i64(i64 [[TMP15]])
 ; X64-NEXT:    [[TMP18:%.*]] = icmp eq i64 [[TMP16]], [[TMP17]]
 ; X64-NEXT:    br i1 [[TMP18]], label [[ENDBLOCK]], label [[RES_BLOCK]]
 ; X64:       endblock:

 
 define i32 @cmp_eq3(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp_eq3(
 ; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i16*
 ; X32-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i16*
 ; X32-NEXT:    [[TMP3:%.*]] = load i16, i16* [[TMP1]]
 ; X32-NEXT:    [[TMP4:%.*]] = load i16, i16* [[TMP2]]
 ; X32-NEXT:    [[TMP5:%.*]] = xor i16 [[TMP3]], [[TMP4]]
-; X32-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 2
-; X32-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i8 2
+; X32-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 2
+; X32-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i64 2
 ; X32-NEXT:    [[TMP8:%.*]] = load i8, i8* [[TMP6]]
 ; X32-NEXT:    [[TMP9:%.*]] = load i8, i8* [[TMP7]]
 ; X32-NEXT:    [[TMP10:%.*]] = zext i8 [[TMP8]] to i16
 ; X32-NEXT:    [[TMP11:%.*]] = zext i8 [[TMP9]] to i16
 ; X32-NEXT:    [[TMP12:%.*]] = xor i16 [[TMP10]], [[TMP11]]
 ; X32-NEXT:    [[TMP13:%.*]] = or i16 [[TMP5]], [[TMP12]]
 ; X32-NEXT:    [[TMP14:%.*]] = icmp ne i16 [[TMP13]], 0
 ; X32-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32
 ; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[TMP15]], 0
 ; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X32-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_1LD-LABEL: @cmp_eq3(
 ; X64_1LD-NEXT:    br label [[LOADBB:%.*]]
 ; X64_1LD:       res_block:
 ; X64_1LD-NEXT:    br label [[ENDBLOCK:%.*]]
 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i16*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i16*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i16, i16* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i16, i16* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i16 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 2
-; X64_1LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i8 2
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 2
+; X64_1LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i64 2
 ; X64_1LD-NEXT:    [[TMP8:%.*]] = load i8, i8* [[TMP6]]
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = load i8, i8* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = icmp ne i8 [[TMP8]], [[TMP9]]
 ; X64_1LD-NEXT:    br i1 [[TMP10]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq3(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i16*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i16*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i16, i16* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i16, i16* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i16 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 2
-; X64_2LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i8 2
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 2
+; X64_2LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i64 2
 ; X64_2LD-NEXT:    [[TMP8:%.*]] = load i8, i8* [[TMP6]]
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = load i8, i8* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = zext i8 [[TMP8]] to i16
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = zext i8 [[TMP9]] to i16
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = xor i16 [[TMP10]], [[TMP11]]
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = or i16 [[TMP5]], [[TMP12]]
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = icmp ne i16 [[TMP13]], 0
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32

 
 define i32 @cmp_eq5(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp_eq5(
 ; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i32*
 ; X32-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
 ; X32-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP2]]
 ; X32-NEXT:    [[TMP5:%.*]] = xor i32 [[TMP3]], [[TMP4]]
-; X32-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 4
-; X32-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i8 4
+; X32-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 4
+; X32-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i64 4
 ; X32-NEXT:    [[TMP8:%.*]] = load i8, i8* [[TMP6]]
 ; X32-NEXT:    [[TMP9:%.*]] = load i8, i8* [[TMP7]]
 ; X32-NEXT:    [[TMP10:%.*]] = zext i8 [[TMP8]] to i32
 ; X32-NEXT:    [[TMP11:%.*]] = zext i8 [[TMP9]] to i32
 ; X32-NEXT:    [[TMP12:%.*]] = xor i32 [[TMP10]], [[TMP11]]
 ; X32-NEXT:    [[TMP13:%.*]] = or i32 [[TMP5]], [[TMP12]]
 ; X32-NEXT:    [[TMP14:%.*]] = icmp ne i32 [[TMP13]], 0
 ; X32-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32
 ; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[TMP15]], 0
 ; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X32-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_1LD-LABEL: @cmp_eq5(
 ; X64_1LD-NEXT:    br label [[LOADBB:%.*]]
 ; X64_1LD:       res_block:
 ; X64_1LD-NEXT:    br label [[ENDBLOCK:%.*]]
 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i32*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i32 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 4
-; X64_1LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i8 4
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 4
+; X64_1LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i64 4
 ; X64_1LD-NEXT:    [[TMP8:%.*]] = load i8, i8* [[TMP6]]
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = load i8, i8* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = icmp ne i8 [[TMP8]], [[TMP9]]
 ; X64_1LD-NEXT:    br i1 [[TMP10]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq5(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i32*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i32 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 4
-; X64_2LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i8 4
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 4
+; X64_2LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i64 4
 ; X64_2LD-NEXT:    [[TMP8:%.*]] = load i8, i8* [[TMP6]]
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = load i8, i8* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = zext i8 [[TMP8]] to i32
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = zext i8 [[TMP9]] to i32
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = xor i32 [[TMP10]], [[TMP11]]
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = or i32 [[TMP5]], [[TMP12]]
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = icmp ne i32 [[TMP13]], 0
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32

 
 define i32 @cmp_eq6(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp_eq6(
 ; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i32*
 ; X32-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
 ; X32-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP2]]
 ; X32-NEXT:    [[TMP5:%.*]] = xor i32 [[TMP3]], [[TMP4]]
-; X32-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 4
+; X32-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 4
 ; X32-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i16*
-; X32-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 4
+; X32-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 4
 ; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i16*
 ; X32-NEXT:    [[TMP10:%.*]] = load i16, i16* [[TMP7]]
 ; X32-NEXT:    [[TMP11:%.*]] = load i16, i16* [[TMP9]]
 ; X32-NEXT:    [[TMP12:%.*]] = zext i16 [[TMP10]] to i32
 ; X32-NEXT:    [[TMP13:%.*]] = zext i16 [[TMP11]] to i32
 ; X32-NEXT:    [[TMP14:%.*]] = xor i32 [[TMP12]], [[TMP13]]
 ; X32-NEXT:    [[TMP15:%.*]] = or i32 [[TMP5]], [[TMP14]]
 ; X32-NEXT:    [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0

 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i32*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i32 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 4
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 4
 ; X64_1LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i16*
-; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 4
+; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 4
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i16*
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = load i16, i16* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP11:%.*]] = load i16, i16* [[TMP9]]
 ; X64_1LD-NEXT:    [[TMP12:%.*]] = icmp ne i16 [[TMP10]], [[TMP11]]
 ; X64_1LD-NEXT:    br i1 [[TMP12]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq6(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i32*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i32 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 4
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 4
 ; X64_2LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i16*
-; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 4
+; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 4
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i16*
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = load i16, i16* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = load i16, i16* [[TMP9]]
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = zext i16 [[TMP10]] to i32
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = zext i16 [[TMP11]] to i32
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = xor i32 [[TMP12]], [[TMP13]]
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = or i32 [[TMP5]], [[TMP14]]
 ; X64_2LD-NEXT:    [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0

 
 define i32 @cmp_eq7(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp_eq7(
 ; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i32*
 ; X32-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
 ; X32-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP2]]
 ; X32-NEXT:    [[TMP5:%.*]] = xor i32 [[TMP3]], [[TMP4]]
-; X32-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 3
+; X32-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 3
 ; X32-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i32*
-; X32-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 3
+; X32-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 3
 ; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i32*
 ; X32-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP7]]
 ; X32-NEXT:    [[TMP11:%.*]] = load i32, i32* [[TMP9]]
 ; X32-NEXT:    [[TMP12:%.*]] = xor i32 [[TMP10]], [[TMP11]]
 ; X32-NEXT:    [[TMP13:%.*]] = or i32 [[TMP5]], [[TMP12]]
 ; X32-NEXT:    [[TMP14:%.*]] = icmp ne i32 [[TMP13]], 0
 ; X32-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32
 ; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[TMP15]], 0
 ; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X32-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_1LD-LABEL: @cmp_eq7(
 ; X64_1LD-NEXT:    br label [[LOADBB:%.*]]
 ; X64_1LD:       res_block:
 ; X64_1LD-NEXT:    br label [[ENDBLOCK:%.*]]
 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i32*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i32 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 3
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 3
 ; X64_1LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i32*
-; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 3
+; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 3
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i32*
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP11:%.*]] = load i32, i32* [[TMP9]]
 ; X64_1LD-NEXT:    [[TMP12:%.*]] = icmp ne i32 [[TMP10]], [[TMP11]]
 ; X64_1LD-NEXT:    br i1 [[TMP12]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq7(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i32*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i32 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 3
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 3
 ; X64_2LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i32*
-; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 3
+; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 3
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i32*
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = load i32, i32* [[TMP9]]
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = xor i32 [[TMP10]], [[TMP11]]
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = or i32 [[TMP5]], [[TMP12]]
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = icmp ne i32 [[TMP13]], 0
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32
 ; X64_2LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[TMP15]], 0
 ; X64_2LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_2LD-NEXT:    ret i32 [[CONV]]
 ;
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 7)
   %cmp = icmp eq i32 %call, 0
   %conv = zext i1 %cmp to i32
   ret i32 %conv
 }
 
 define i32 @cmp_eq8(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp_eq8(
 ; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i32*
 ; X32-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i32*
 ; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
 ; X32-NEXT:    [[TMP4:%.*]] = load i32, i32* [[TMP2]]
 ; X32-NEXT:    [[TMP5:%.*]] = xor i32 [[TMP3]], [[TMP4]]
-; X32-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 4
+; X32-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 4
 ; X32-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i32*
-; X32-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 4
+; X32-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 4
 ; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i32*
 ; X32-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP7]]
 ; X32-NEXT:    [[TMP11:%.*]] = load i32, i32* [[TMP9]]
 ; X32-NEXT:    [[TMP12:%.*]] = xor i32 [[TMP10]], [[TMP11]]
 ; X32-NEXT:    [[TMP13:%.*]] = or i32 [[TMP5]], [[TMP12]]
 ; X32-NEXT:    [[TMP14:%.*]] = icmp ne i32 [[TMP13]], 0
 ; X32-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32
 ; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[TMP15]], 0

 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i64 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 8
-; X64_1LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i8 8
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 8
+; X64_1LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i64 8
 ; X64_1LD-NEXT:    [[TMP8:%.*]] = load i8, i8* [[TMP6]]
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = load i8, i8* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = icmp ne i8 [[TMP8]], [[TMP9]]
 ; X64_1LD-NEXT:    br i1 [[TMP10]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq9(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i64 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 8
-; X64_2LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i8 8
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 8
+; X64_2LD-NEXT:    [[TMP7:%.*]] = getelementptr i8, i8* [[Y]], i64 8
 ; X64_2LD-NEXT:    [[TMP8:%.*]] = load i8, i8* [[TMP6]]
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = load i8, i8* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = zext i8 [[TMP8]] to i64
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = zext i8 [[TMP9]] to i64
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = xor i64 [[TMP10]], [[TMP11]]
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = or i64 [[TMP5]], [[TMP12]]
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = icmp ne i64 [[TMP13]], 0
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32

 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i64 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 8
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 8
 ; X64_1LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i16*
-; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 8
+; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 8
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i16*
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = load i16, i16* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP11:%.*]] = load i16, i16* [[TMP9]]
 ; X64_1LD-NEXT:    [[TMP12:%.*]] = icmp ne i16 [[TMP10]], [[TMP11]]
 ; X64_1LD-NEXT:    br i1 [[TMP12]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq10(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i64 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 8
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 8
 ; X64_2LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i16*
-; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 8
+; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 8
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i16*
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = load i16, i16* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = load i16, i16* [[TMP9]]
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = zext i16 [[TMP10]] to i64
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = zext i16 [[TMP11]] to i64
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = xor i64 [[TMP12]], [[TMP13]]
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = or i64 [[TMP5]], [[TMP14]]
 ; X64_2LD-NEXT:    [[TMP16:%.*]] = icmp ne i64 [[TMP15]], 0

 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i64 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 3
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 3
 ; X64_1LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i64*
-; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 3
+; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 3
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i64*
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = load i64, i64* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP11:%.*]] = load i64, i64* [[TMP9]]
 ; X64_1LD-NEXT:    [[TMP12:%.*]] = icmp ne i64 [[TMP10]], [[TMP11]]
 ; X64_1LD-NEXT:    br i1 [[TMP12]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq11(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i64 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 3
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 3
 ; X64_2LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i64*
-; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 3
+; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 3
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i64*
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = load i64, i64* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = load i64, i64* [[TMP9]]
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = xor i64 [[TMP10]], [[TMP11]]
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = or i64 [[TMP5]], [[TMP12]]
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = icmp ne i64 [[TMP13]], 0
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32
 ; X64_2LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[TMP15]], 0

 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i64 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 8
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 8
 ; X64_1LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i32*
-; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 8
+; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 8
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i32*
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP11:%.*]] = load i32, i32* [[TMP9]]
 ; X64_1LD-NEXT:    [[TMP12:%.*]] = icmp ne i32 [[TMP10]], [[TMP11]]
 ; X64_1LD-NEXT:    br i1 [[TMP12]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq12(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i64 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 8
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 8
 ; X64_2LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i32*
-; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 8
+; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 8
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i32*
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = load i32, i32* [[TMP9]]
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = zext i32 [[TMP10]] to i64
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = zext i32 [[TMP11]] to i64
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = xor i64 [[TMP12]], [[TMP13]]
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = or i64 [[TMP5]], [[TMP14]]
 ; X64_2LD-NEXT:    [[TMP16:%.*]] = icmp ne i64 [[TMP15]], 0

 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i64 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 5
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 5
 ; X64_1LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i64*
-; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 5
+; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 5
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i64*
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = load i64, i64* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP11:%.*]] = load i64, i64* [[TMP9]]
 ; X64_1LD-NEXT:    [[TMP12:%.*]] = icmp ne i64 [[TMP10]], [[TMP11]]
 ; X64_1LD-NEXT:    br i1 [[TMP12]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq13(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i64 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 5
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 5
 ; X64_2LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i64*
-; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 5
+; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 5
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i64*
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = load i64, i64* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = load i64, i64* [[TMP9]]
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = xor i64 [[TMP10]], [[TMP11]]
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = or i64 [[TMP5]], [[TMP12]]
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = icmp ne i64 [[TMP13]], 0
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32
 ; X64_2LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[TMP15]], 0

 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i64 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 6
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 6
 ; X64_1LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i64*
-; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 6
+; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 6
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i64*
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = load i64, i64* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP11:%.*]] = load i64, i64* [[TMP9]]
 ; X64_1LD-NEXT:    [[TMP12:%.*]] = icmp ne i64 [[TMP10]], [[TMP11]]
 ; X64_1LD-NEXT:    br i1 [[TMP12]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq14(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i64 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 6
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 6
 ; X64_2LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i64*
-; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 6
+; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 6
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i64*
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = load i64, i64* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = load i64, i64* [[TMP9]]
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = xor i64 [[TMP10]], [[TMP11]]
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = or i64 [[TMP5]], [[TMP12]]
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = icmp ne i64 [[TMP13]], 0
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32
 ; X64_2LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[TMP15]], 0

 ; X64_1LD:       loadbb:
 ; X64_1LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_1LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_1LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_1LD-NEXT:    [[TMP5:%.*]] = icmp ne i64 [[TMP3]], [[TMP4]]
 ; X64_1LD-NEXT:    br i1 [[TMP5]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
 ; X64_1LD:       loadbb1:
-; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 7
+; X64_1LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 7
 ; X64_1LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i64*
-; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 7
+; X64_1LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 7
 ; X64_1LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i64*
 ; X64_1LD-NEXT:    [[TMP10:%.*]] = load i64, i64* [[TMP7]]
 ; X64_1LD-NEXT:    [[TMP11:%.*]] = load i64, i64* [[TMP9]]
 ; X64_1LD-NEXT:    [[TMP12:%.*]] = icmp ne i64 [[TMP10]], [[TMP11]]
 ; X64_1LD-NEXT:    br i1 [[TMP12]], label [[RES_BLOCK]], label [[ENDBLOCK]]
 ; X64_1LD:       endblock:
 ; X64_1LD-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB1]] ], [ 1, [[RES_BLOCK]] ]
 ; X64_1LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
 ; X64_1LD-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X64_1LD-NEXT:    ret i32 [[CONV]]
 ;
 ; X64_2LD-LABEL: @cmp_eq15(
 ; X64_2LD-NEXT:    [[TMP1:%.*]] = bitcast i8* [[X:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP2:%.*]] = bitcast i8* [[Y:%.*]] to i64*
 ; X64_2LD-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
 ; X64_2LD-NEXT:    [[TMP4:%.*]] = load i64, i64* [[TMP2]]
 ; X64_2LD-NEXT:    [[TMP5:%.*]] = xor i64 [[TMP3]], [[TMP4]]
-; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i8 7
+; X64_2LD-NEXT:    [[TMP6:%.*]] = getelementptr i8, i8* [[X]], i64 7
 ; X64_2LD-NEXT:    [[TMP7:%.*]] = bitcast i8* [[TMP6]] to i64*
-; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i8 7
+; X64_2LD-NEXT:    [[TMP8:%.*]] = getelementptr i8, i8* [[Y]], i64 7
 ; X64_2LD-NEXT:    [[TMP9:%.*]] = bitcast i8* [[TMP8]] to i64*
 ; X64_2LD-NEXT:    [[TMP10:%.*]] = load i64, i64* [[TMP7]]
 ; X64_2LD-NEXT:    [[TMP11:%.*]] = load i64, i64* [[TMP9]]
 ; X64_2LD-NEXT:    [[TMP12:%.*]] = xor i64 [[TMP10]], [[TMP11]]
 ; X64_2LD-NEXT:    [[TMP13:%.*]] = or i64 [[TMP5]], [[TMP12]]
 ; X64_2LD-NEXT:    [[TMP14:%.*]] = icmp ne i64 [[TMP13]], 0
 ; X64_2LD-NEXT:    [[TMP15:%.*]] = zext i1 [[TMP14]] to i32
 ; X64_2LD-NEXT:    [[CMP:%.*]] = icmp eq i32 [[TMP15]], 0