Vendor import of llvm release_50 branch r309439:vendor/llvm/llvm-release_50-r309439

https://llvm.org/svn/llvm-project/llvm/branches/release_50@309439

51 files changed, 1082 insertions, 1383 deletions

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 61ecfdf970d0..6af2cba10093 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -29,7 +29,7 @@ if(NOT DEFINED LLVM_VERSION_PATCH)
   set(LLVM_VERSION_PATCH 0)
 endif()
 if(NOT DEFINED LLVM_VERSION_SUFFIX)
-  set(LLVM_VERSION_SUFFIX svn)
+  set(LLVM_VERSION_SUFFIX "")
 endif()
 
 if (POLICY CMP0048)
diff --git a/docs/ReleaseNotes.rst b/docs/ReleaseNotes.rst
index ddb31acfd029..800a8ec115b3 100644
--- a/docs/ReleaseNotes.rst
+++ b/docs/ReleaseNotes.rst
@@ -71,6 +71,12 @@ Non-comprehensive list of changes in this release
 Changes to the LLVM IR
 ----------------------
 
+* The datalayout string may now indicate an address space to use for
+  the pointer type of alloca rather than the default of 0.
+
+* Added speculatable attribute indicating a function which does has no
+  side-effects which could inhibit hoisting of calls.
+
 Changes to the ARM Backend
 --------------------------
 
@@ -91,12 +97,30 @@ Changes to the PowerPC Target
 Changes to the X86 Target
 -------------------------
 
- During this release ...
+* Added initial AMD Ryzen (znver1) scheduler support.
+
+* Added support for Intel Goldmont CPUs.
+
+* Add support for avx512vpopcntdq instructions.
+
+* Added heuristics to convert CMOV into branches when it may be profitable.
+
+* More aggressive inlining of memcmp calls.
+
+* Improve vXi64 shuffles on 32-bit targets.
+
+* Improved use of PMOVMSKB for any_of/all_of comparision reductions.
+
+* Improved Silvermont, Sandybridge, and Jaguar (btver2) schedulers.
+
+* Improved support for AVX512 vector rotations.
+
+* Added support for AMD Lightweight Profiling (LWP) instructions.
 
 Changes to the AMDGPU Target
 -----------------------------
 
- During this release ...
+* Initial gfx9 support
 
 Changes to the AVR Target
 -----------------------------
diff --git a/examples/ParallelJIT/ParallelJIT.cpp b/examples/ParallelJIT/ParallelJIT.cpp
index f1932d2471cb..ff44375e3921 100644
--- a/examples/ParallelJIT/ParallelJIT.cpp
+++ b/examples/ParallelJIT/ParallelJIT.cpp
@@ -145,6 +145,7 @@ public:
     waitFor = 0;
 
     int result = pthread_cond_init( &condition, nullptr );
+    (void)result;
     assert( result == 0 );
 
     result = pthread_mutex_init( &mutex, nullptr );
diff --git a/include/llvm/CodeGen/GlobalISel/InstructionSelector.h b/include/llvm/CodeGen/GlobalISel/InstructionSelector.h
index 59a4073646eb..1060d8fd667e 100644
--- a/include/llvm/CodeGen/GlobalISel/InstructionSelector.h
+++ b/include/llvm/CodeGen/GlobalISel/InstructionSelector.h
@@ -40,7 +40,8 @@ class TargetRegisterInfo;
 /// This is convenient because std::bitset does not have a constructor
 /// with an initializer list of set bits.
 ///
-/// Each InstructionSelector subclass should define a PredicateBitset class with:
+/// Each InstructionSelector subclass should define a PredicateBitset class
+/// with:
 ///   const unsigned MAX_SUBTARGET_PREDICATES = 192;
 ///   using PredicateBitset = PredicateBitsetImpl<MAX_SUBTARGET_PREDICATES>;
 /// and updating the constant to suit the target. Tablegen provides a suitable
@@ -102,7 +103,8 @@ enum {
   /// - OpIdx - Operand index
   /// - Expected integer
   GIM_CheckConstantInt,
-  /// Check the operand is a specific literal integer (i.e. MO.isImm() or MO.isCImm() is true).
+  /// Check the operand is a specific literal integer (i.e. MO.isImm() or
+  /// MO.isCImm() is true).
   /// - InsnID - Instruction ID
   /// - OpIdx - Operand index
   /// - Expected integer
diff --git a/include/llvm/Support/CommandLine.h b/include/llvm/Support/CommandLine.h
index 771b0a8c26a9..71d2f0293083 100644
--- a/include/llvm/Support/CommandLine.h
+++ b/include/llvm/Support/CommandLine.h
@@ -66,15 +66,12 @@ bool ParseCommandLineOptions(int argc, const char *const *argv,
 void ParseEnvironmentOptions(const char *progName, const char *envvar,
                              const char *Overview = "");
 
-// Function pointer type for printing version information.
-using VersionPrinterTy = std::function<void(raw_ostream &)>;
-
 ///===---------------------------------------------------------------------===//
 /// SetVersionPrinter - Override the default (LLVM specific) version printer
 ///                     used to print out the version when --version is given
 ///                     on the command line. This allows other systems using the
 ///                     CommandLine utilities to print their own version string.
-void SetVersionPrinter(VersionPrinterTy func);
+void SetVersionPrinter(void (*func)());
 
 ///===---------------------------------------------------------------------===//
 /// AddExtraVersionPrinter - Add an extra printer to use in addition to the
@@ -83,7 +80,7 @@ void SetVersionPrinter(VersionPrinterTy func);
 ///                          which will be called after the basic LLVM version
 ///                          printing is complete. Each can then add additional
 ///                          information specific to the tool.
-void AddExtraVersionPrinter(VersionPrinterTy func);
+void AddExtraVersionPrinter(void (*func)());
 
 // PrintOptionValues - Print option values.
 // With -print-options print the difference between option values and defaults.
diff --git a/include/llvm/Support/TargetRegistry.h b/include/llvm/Support/TargetRegistry.h
index 90d6c084ee95..8454b27b6f04 100644
--- a/include/llvm/Support/TargetRegistry.h
+++ b/include/llvm/Support/TargetRegistry.h
@@ -599,7 +599,7 @@ struct TargetRegistry {
 
   /// printRegisteredTargetsForVersion - Print the registered targets
   /// appropriately for inclusion in a tool's version output.
-  static void printRegisteredTargetsForVersion(raw_ostream &OS);
+  static void printRegisteredTargetsForVersion();
 
   /// @name Registry Access
   /// @{
diff --git a/include/llvm/Transforms/Utils/LoopUtils.h b/include/llvm/Transforms/Utils/LoopUtils.h
index 1344285917ba..94e20b83754e 100644
--- a/include/llvm/Transforms/Utils/LoopUtils.h
+++ b/include/llvm/Transforms/Utils/LoopUtils.h
@@ -531,8 +531,10 @@ Value *createTargetReduction(IRBuilder<> &B, const TargetTransformInfo *TTI,
 
 /// Get the intersection (logical and) of all of the potential IR flags
 /// of each scalar operation (VL) that will be converted into a vector (I).
+/// If OpValue is non-null, we only consider operations similar to OpValue
+/// when intersecting.
 /// Flag set: NSW, NUW, exact, and all of fast-math.
-void propagateIRFlags(Value *I, ArrayRef<Value *> VL);
+void propagateIRFlags(Value *I, ArrayRef<Value *> VL, Value *OpValue = nullptr);
 
 } // end namespace llvm
 
diff --git a/lib/CodeGen/CodeGenPrepare.cpp b/lib/CodeGen/CodeGenPrepare.cpp
index 45dc13d58de7..dc02a00e0fcc 100644
--- a/lib/CodeGen/CodeGenPrepare.cpp
+++ b/lib/CodeGen/CodeGenPrepare.cpp
@@ -4016,14 +4016,18 @@ static bool IsOperandAMemoryOperand(CallInst *CI, InlineAsm *IA, Value *OpVal,
   return true;
 }
 
+// Max number of memory uses to look at before aborting the search to conserve
+// compile time.
+static constexpr int MaxMemoryUsesToScan = 20;
+
 /// Recursively walk all the uses of I until we find a memory use.
 /// If we find an obviously non-foldable instruction, return true.
 /// Add the ultimately found memory instructions to MemoryUses.
 static bool FindAllMemoryUses(
     Instruction *I,
     SmallVectorImpl<std::pair<Instruction *, unsigned>> &MemoryUses,
-    SmallPtrSetImpl<Instruction *> &ConsideredInsts,
-    const TargetLowering &TLI, const TargetRegisterInfo &TRI) {
+    SmallPtrSetImpl<Instruction *> &ConsideredInsts, const TargetLowering &TLI,
+    const TargetRegisterInfo &TRI, int SeenInsts = 0) {
   // If we already considered this instruction, we're done.
   if (!ConsideredInsts.insert(I).second)
     return false;
@@ -4036,8 +4040,12 @@ static bool FindAllMemoryUses(
 
   // Loop over all the uses, recursively processing them.
   for (Use &U : I->uses()) {
-    Instruction *UserI = cast<Instruction>(U.getUser());
+    // Conservatively return true if we're seeing a large number or a deep chain
+    // of users. This avoids excessive compilation times in pathological cases.
+    if (SeenInsts++ >= MaxMemoryUsesToScan)
+      return true;
 
+    Instruction *UserI = cast<Instruction>(U.getUser());
     if (LoadInst *LI = dyn_cast<LoadInst>(UserI)) {
       MemoryUses.push_back(std::make_pair(LI, U.getOperandNo()));
       continue;
@@ -4082,7 +4090,8 @@ static bool FindAllMemoryUses(
       continue;
     }
 
-    if (FindAllMemoryUses(UserI, MemoryUses, ConsideredInsts, TLI, TRI))
+    if (FindAllMemoryUses(UserI, MemoryUses, ConsideredInsts, TLI, TRI,
+                          SeenInsts))
       return true;
   }
 
diff --git a/lib/CodeGen/InlineSpiller.cpp b/lib/CodeGen/InlineSpiller.cpp
index 4e6a3ec21866..eda4f74c7874 100644
--- a/lib/CodeGen/InlineSpiller.cpp
+++ b/lib/CodeGen/InlineSpiller.cpp
@@ -643,8 +643,11 @@ void InlineSpiller::reMaterializeAll() {
       Edit->eraseVirtReg(Reg);
       continue;
     }
-    assert((LIS.hasInterval(Reg) && !LIS.getInterval(Reg).empty()) &&
-           "Reg with empty interval has reference");
+
+    assert(LIS.hasInterval(Reg) &&
+           (!LIS.getInterval(Reg).empty() || !MRI.reg_nodbg_empty(Reg)) &&
+           "Empty and not used live-range?!");
+
     RegsToSpill[ResultPos++] = Reg;
   }
   RegsToSpill.erase(RegsToSpill.begin() + ResultPos, RegsToSpill.end());
diff --git a/lib/CodeGen/RegAllocBase.cpp b/lib/CodeGen/RegAllocBase.cpp
index a7b7a9f8ab15..7b4fbace2c1c 100644
--- a/lib/CodeGen/RegAllocBase.cpp
+++ b/lib/CodeGen/RegAllocBase.cpp
@@ -133,18 +133,19 @@ void RegAllocBase::allocatePhysRegs() {
     if (AvailablePhysReg)
       Matrix->assign(*VirtReg, AvailablePhysReg);
 
-    for (VirtRegVec::iterator I = SplitVRegs.begin(), E = SplitVRegs.end();
-         I != E; ++I) {
-      LiveInterval *SplitVirtReg = &LIS->getInterval(*I);
+    for (unsigned Reg : SplitVRegs) {
+      assert(LIS->hasInterval(Reg));
+
+      LiveInterval *SplitVirtReg = &LIS->getInterval(Reg);
       assert(!VRM->hasPhys(SplitVirtReg->reg) && "Register already assigned");
       if (MRI->reg_nodbg_empty(SplitVirtReg->reg)) {
+        assert(SplitVirtReg->empty() && "Non-empty but used interval");
         DEBUG(dbgs() << "not queueing unused  " << *SplitVirtReg << '\n');
         aboutToRemoveInterval(*SplitVirtReg);
         LIS->removeInterval(SplitVirtReg->reg);
         continue;
       }
       DEBUG(dbgs() << "queuing new interval: " << *SplitVirtReg << "\n");
-      assert(!SplitVirtReg->empty() && "expecting non-empty interval");
       assert(TargetRegisterInfo::isVirtualRegister(SplitVirtReg->reg) &&
              "expect split value in virtual register");
       enqueue(SplitVirtReg);
diff --git a/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp b/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp
index d41054b15bbc..0cad20db0964 100644
--- a/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp
+++ b/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp
@@ -2965,7 +2965,12 @@ static inline bool isSETCCorConvertedSETCC(SDValue N) {
   else if (N.getOpcode() == ISD::SIGN_EXTEND)
     N = N.getOperand(0);
 
-  return (N.getOpcode() == ISD::SETCC);
+  if (isLogicalMaskOp(N.getOpcode()))
+    return isSETCCorConvertedSETCC(N.getOperand(0)) &&
+           isSETCCorConvertedSETCC(N.getOperand(1));
+
+  return (N.getOpcode() == ISD::SETCC ||
+          ISD::isBuildVectorOfConstantSDNodes(N.getNode()));
 }
 #endif
 
@@ -2973,28 +2978,20 @@ static inline bool isSETCCorConvertedSETCC(SDValue N) {
 // to ToMaskVT if needed with vector extension or truncation.
 SDValue DAGTypeLegalizer::convertMask(SDValue InMask, EVT MaskVT,
                                       EVT ToMaskVT) {
-  LLVMContext &Ctx = *DAG.getContext();
-
   // Currently a SETCC or a AND/OR/XOR with two SETCCs are handled.
-  unsigned InMaskOpc = InMask->getOpcode();
-
   // FIXME: This code seems to be too restrictive, we might consider
   // generalizing it or dropping it.
-  assert((InMaskOpc == ISD::SETCC ||
-          ISD::isBuildVectorOfConstantSDNodes(InMask.getNode()) ||
-          (isLogicalMaskOp(InMaskOpc) &&
-           isSETCCorConvertedSETCC(InMask->getOperand(0)) &&
-           isSETCCorConvertedSETCC(InMask->getOperand(1)))) &&
-         "Unexpected mask argument.");
+  assert(isSETCCorConvertedSETCC(InMask) && "Unexpected mask argument.");
 
   // Make a new Mask node, with a legal result VT.
   SmallVector<SDValue, 4> Ops;
   for (unsigned i = 0; i < InMask->getNumOperands(); ++i)
     Ops.push_back(InMask->getOperand(i));
-  SDValue Mask = DAG.getNode(InMaskOpc, SDLoc(InMask), MaskVT, Ops);
+  SDValue Mask = DAG.getNode(InMask->getOpcode(), SDLoc(InMask), MaskVT, Ops);
 
   // If MaskVT has smaller or bigger elements than ToMaskVT, a vector sign
   // extend or truncate is needed.
+  LLVMContext &Ctx = *DAG.getContext();
   unsigned MaskScalarBits = MaskVT.getScalarSizeInBits();
   unsigned ToMaskScalBits = ToMaskVT.getScalarSizeInBits();
   if (MaskScalarBits < ToMaskScalBits) {
diff --git a/lib/Option/OptTable.cpp b/lib/Option/OptTable.cpp
index f3b438e829d6..51c62d33f8e1 100644
--- a/lib/Option/OptTable.cpp
+++ b/lib/Option/OptTable.cpp
@@ -235,7 +235,9 @@ OptTable::findByPrefix(StringRef Cur, unsigned short DisableFlags) const {
       continue;
 
     for (int I = 0; In.Prefixes[I]; I++) {
-      std::string S = std::string(In.Prefixes[I]) + std::string(In.Name);
+      std::string S = std::string(In.Prefixes[I]) + std::string(In.Name) + "\t";
+      if (In.HelpText)
+        S += In.HelpText;
       if (StringRef(S).startswith(Cur))
         Ret.push_back(S);
     }
diff --git a/lib/Support/CommandLine.cpp b/lib/Support/CommandLine.cpp
index 50173f5256bf..8eeb685a18a9 100644
--- a/lib/Support/CommandLine.cpp
+++ b/lib/Support/CommandLine.cpp
@@ -2039,9 +2039,9 @@ void CommandLineParser::printOptionValues() {
     Opts[i].second->printOptionValue(MaxArgLen, PrintAllOptions);
 }
 
-static VersionPrinterTy OverrideVersionPrinter = nullptr;
+static void (*OverrideVersionPrinter)() = nullptr;
 
-static std::vector<VersionPrinterTy> *ExtraVersionPrinters = nullptr;
+static std::vector<void (*)()> *ExtraVersionPrinters = nullptr;
 
 namespace {
 class VersionPrinter {
@@ -2081,7 +2081,7 @@ public:
       return;
 
     if (OverrideVersionPrinter != nullptr) {
-      OverrideVersionPrinter(outs());
+      (*OverrideVersionPrinter)();
       exit(0);
     }
     print();
@@ -2090,8 +2090,10 @@ public:
     // information.
     if (ExtraVersionPrinters != nullptr) {
       outs() << '\n';
-      for (auto I : *ExtraVersionPrinters)
-        I(outs());
+      for (std::vector<void (*)()>::iterator I = ExtraVersionPrinters->begin(),
+                                             E = ExtraVersionPrinters->end();
+           I != E; ++I)
+        (*I)();
     }
 
     exit(0);
@@ -2129,11 +2131,11 @@ void cl::PrintHelpMessage(bool Hidden, bool Categorized) {
 /// Utility function for printing version number.
 void cl::PrintVersionMessage() { VersionPrinterInstance.print(); }
 
-void cl::SetVersionPrinter(VersionPrinterTy func) { OverrideVersionPrinter = func; }
+void cl::SetVersionPrinter(void (*func)()) { OverrideVersionPrinter = func; }
 
-void cl::AddExtraVersionPrinter(VersionPrinterTy func) {
+void cl::AddExtraVersionPrinter(void (*func)()) {
   if (!ExtraVersionPrinters)
-    ExtraVersionPrinters = new std::vector<VersionPrinterTy>;
+    ExtraVersionPrinters = new std::vector<void (*)()>;
 
   ExtraVersionPrinters->push_back(func);
 }
diff --git a/lib/Support/ErrorHandling.cpp b/lib/Support/ErrorHandling.cpp
index 2fd4f3ea0d45..fb8ae4c1cd5e 100644
--- a/lib/Support/ErrorHandling.cpp
+++ b/lib/Support/ErrorHandling.cpp
@@ -169,7 +169,8 @@ void llvm::report_bad_alloc_error(const char *Reason, bool GenCrashDiag) {
   // Don't call the normal error handler. It may allocate memory. Directly write
   // an OOM to stderr and abort.
   char OOMMessage[] = "LLVM ERROR: out of memory\n";
-  (void)::write(2, OOMMessage, strlen(OOMMessage));
+  ssize_t written = ::write(2, OOMMessage, strlen(OOMMessage));
+  (void)written;
   abort();
 #endif
 }
diff --git a/lib/Support/TargetRegistry.cpp b/lib/Support/TargetRegistry.cpp
index b5c283253117..bed9ed64f802 100644
--- a/lib/Support/TargetRegistry.cpp
+++ b/lib/Support/TargetRegistry.cpp
@@ -114,7 +114,7 @@ static int TargetArraySortFn(const std::pair<StringRef, const Target *> *LHS,
   return LHS->first.compare(RHS->first);
 }
 
-void TargetRegistry::printRegisteredTargetsForVersion(raw_ostream &OS) {
+void TargetRegistry::printRegisteredTargetsForVersion() {
   std::vector<std::pair<StringRef, const Target*> > Targets;
   size_t Width = 0;
   for (const auto &T : TargetRegistry::targets()) {
@@ -123,6 +123,7 @@ void TargetRegistry::printRegisteredTargetsForVersion(raw_ostream &OS) {
   }
   array_pod_sort(Targets.begin(), Targets.end(), TargetArraySortFn);
 
+  raw_ostream &OS = outs();
   OS << "  Registered Targets:\n";
   for (unsigned i = 0, e = Targets.size(); i != e; ++i) {
     OS << "    " << Targets[i].first;
diff --git a/lib/Target/AArch64/AArch64ISelLowering.cpp b/lib/Target/AArch64/AArch64ISelLowering.cpp
index c6150f9e5d1d..8c30c4410c09 100644
--- a/lib/Target/AArch64/AArch64ISelLowering.cpp
+++ b/lib/Target/AArch64/AArch64ISelLowering.cpp
@@ -2889,9 +2889,12 @@ void AArch64TargetLowering::saveVarArgRegisters(CCState &CCInfo,
   unsigned GPRSaveSize = 8 * (NumGPRArgRegs - FirstVariadicGPR);
   int GPRIdx = 0;
   if (GPRSaveSize != 0) {
-    if (IsWin64)
+    if (IsWin64) {
       GPRIdx = MFI.CreateFixedObject(GPRSaveSize, -(int)GPRSaveSize, false);
-    else
+      if (GPRSaveSize & 15)
+        // The extra size here, if triggered, will always be 8.
+        MFI.CreateFixedObject(16 - (GPRSaveSize & 15), -(int)alignTo(GPRSaveSize, 16), false);
+    } else
       GPRIdx = MFI.CreateStackObject(GPRSaveSize, 8, false);
 
     SDValue FIN = DAG.getFrameIndex(GPRIdx, PtrVT);
diff --git a/lib/Target/AMDGPU/InstPrinter/AMDGPUInstPrinter.cpp b/lib/Target/AMDGPU/InstPrinter/AMDGPUInstPrinter.cpp
index 7c31c8e397ba..a844081db5b2 100644
--- a/lib/Target/AMDGPU/InstPrinter/AMDGPUInstPrinter.cpp
+++ b/lib/Target/AMDGPU/InstPrinter/AMDGPUInstPrinter.cpp
@@ -297,6 +297,11 @@ void AMDGPUInstPrinter::printRegOperand(unsigned RegNo, raw_ostream &O,
   case AMDGPU::FLAT_SCR_HI:
     O << "flat_scratch_hi";
     return;
+  case AMDGPU::FP_REG:
+  case AMDGPU::SP_REG:
+  case AMDGPU::SCRATCH_WAVE_OFFSET_REG:
+  case AMDGPU::PRIVATE_RSRC_REG:
+    llvm_unreachable("pseudo-register should not ever be emitted");
   default:
     break;
   }
diff --git a/lib/Target/AMDGPU/SIRegisterInfo.td b/lib/Target/AMDGPU/SIRegisterInfo.td
index 54ea7805e18d..d097b78890e3 100644
--- a/lib/Target/AMDGPU/SIRegisterInfo.td
+++ b/lib/Target/AMDGPU/SIRegisterInfo.td
@@ -274,8 +274,7 @@ def VGPR_512 : RegisterTuples<[sub0, sub1, sub2, sub3, sub4, sub5, sub6, sub7,
 def SReg_32_XM0_XEXEC : RegisterClass<"AMDGPU", [i32, f32, i16, f16, v2i16, v2f16], 32,
   (add SGPR_32, VCC_LO, VCC_HI, FLAT_SCR_LO, FLAT_SCR_HI,
    TTMP_32, TMA_LO, TMA_HI, TBA_LO, TBA_HI, SRC_SHARED_BASE, SRC_SHARED_LIMIT,
-   SRC_PRIVATE_BASE, SRC_PRIVATE_LIMIT,
-   FP_REG, SP_REG, SCRATCH_WAVE_OFFSET_REG)> {
+   SRC_PRIVATE_BASE, SRC_PRIVATE_LIMIT)> {
   let AllocationPriority = 7;
 }
 
diff --git a/lib/Target/Sparc/MCTargetDesc/SparcAsmBackend.cpp b/lib/Target/Sparc/MCTargetDesc/SparcAsmBackend.cpp
index 0d021d67033e..0a72a4438218 100644
--- a/lib/Target/Sparc/MCTargetDesc/SparcAsmBackend.cpp
+++ b/lib/Target/Sparc/MCTargetDesc/SparcAsmBackend.cpp
@@ -61,14 +61,6 @@ static unsigned adjustFixupValue(unsigned Kind, uint64_t Value) {
   case Sparc::fixup_sparc_lo10:
     return Value & 0x3ff;
 
-  case Sparc::fixup_sparc_tls_ldo_hix22:
-  case Sparc::fixup_sparc_tls_le_hix22:
-    return (~Value >> 10) & 0x3fffff;
-
-  case Sparc::fixup_sparc_tls_ldo_lox10:
-  case Sparc::fixup_sparc_tls_le_lox10:
-    return (~(~Value & 0x3ff)) & 0x1fff;
-
   case Sparc::fixup_sparc_h44:
     return (Value >> 22) & 0x3fffff;
 
@@ -84,6 +76,13 @@ static unsigned adjustFixupValue(unsigned Kind, uint64_t Value) {
   case Sparc::fixup_sparc_hm:
     return (Value >> 32) & 0x3ff;
 
+  case Sparc::fixup_sparc_tls_ldo_hix22:
+  case Sparc::fixup_sparc_tls_le_hix22:
+  case Sparc::fixup_sparc_tls_ldo_lox10:
+  case Sparc::fixup_sparc_tls_le_lox10:
+    assert(Value == 0 && "Sparc TLS relocs expect zero Value");
+    return 0;
+
   case Sparc::fixup_sparc_tls_gd_add:
   case Sparc::fixup_sparc_tls_gd_call:
   case Sparc::fixup_sparc_tls_ldm_add:
diff --git a/lib/Target/SystemZ/SystemZScheduleZ14.td b/lib/Target/SystemZ/SystemZScheduleZ14.td
index f11177af91a5..698eb5627d19 100644
--- a/lib/Target/SystemZ/SystemZScheduleZ14.td
+++ b/lib/Target/SystemZ/SystemZScheduleZ14.td
@@ -455,10 +455,10 @@ def : InstRW<[FXa, LSU, Lat8], (instregex "MH(Y)?$")>;
 def : InstRW<[FXa2, Lat6, GroupAlone], (instregex "M(L)?R$")>;
 def : InstRW<[FXa2, LSU, Lat10, GroupAlone], (instregex "M(FY|L)?$")>;
 def : InstRW<[FXa, LSU, Lat8], (instregex "MGH$")>;
-def : InstRW<[FXa, LSU, Lat12, GroupAlone], (instregex "MG$")>;
-def : InstRW<[FXa, Lat8, GroupAlone], (instregex "MGRK$")>;
-def : InstRW<[FXa, LSU, Lat9, GroupAlone], (instregex "MSC$")>;
-def : InstRW<[FXa, LSU, Lat11, GroupAlone], (instregex "MSGC$")>;
+def : InstRW<[FXa, FXa, LSU, Lat12, GroupAlone], (instregex "MG$")>;
+def : InstRW<[FXa, FXa, Lat8, GroupAlone], (instregex "MGRK$")>;
+def : InstRW<[FXa, LSU, Lat9], (instregex "MSC$")>;
+def : InstRW<[FXa, LSU, Lat11], (instregex "MSGC$")>;
 def : InstRW<[FXa, Lat5], (instregex "MSRKC$")>;
 def : InstRW<[FXa, Lat7], (instregex "MSGRKC$")>;
 
@@ -620,7 +620,7 @@ def : InstRW<[FXa, Lat30], (instregex "(PCC|PPNO|PRNO)$")>;
 
 def : InstRW<[LSU], (instregex "LGG$")>;
 def : InstRW<[LSU, Lat5], (instregex "LLGFSG$")>;
-def : InstRW<[LSU, Lat30, GroupAlone], (instregex "(L|ST)GSC$")>;
+def : InstRW<[LSU, Lat30], (instregex "(L|ST)GSC$")>;
 
 //===----------------------------------------------------------------------===//
 // Decimal arithmetic
@@ -708,7 +708,7 @@ def : InstRW<[FXb, LSU, Lat5], (instregex "NTSTG$")>;
 // Processor assist
 //===----------------------------------------------------------------------===//
 
-def : InstRW<[FXb], (instregex "PPA$")>;
+def : InstRW<[FXb, GroupAlone], (instregex "PPA$")>;
 
 //===----------------------------------------------------------------------===//
 // Miscellaneous Instructions.
@@ -1276,9 +1276,9 @@ def : InstRW<[VecXsPm], (instregex "VESRL(B|F|G|H)?$")>;
 def : InstRW<[VecXsPm], (instregex "VESRLV(B|F|G|H)?$")>;
 
 def : InstRW<[VecXsPm], (instregex "VSL(DB)?$")>;
-def : InstRW<[VecXsPm, VecXsPm, Lat8], (instregex "VSLB$")>;
+def : InstRW<[VecXsPm], (instregex "VSLB$")>;
 def : InstRW<[VecXsPm], (instregex "VSR(A|L)$")>;
-def : InstRW<[VecXsPm, VecXsPm, Lat8], (instregex "VSR(A|L)B$")>;
+def : InstRW<[VecXsPm], (instregex "VSR(A|L)B$")>;
 
 def : InstRW<[VecXsPm], (instregex "VSB(I|IQ|CBI|CBIQ)?$")>;
 def : InstRW<[VecXsPm], (instregex "VSCBI(B|F|G|H|Q)?$")>;
@@ -1435,9 +1435,9 @@ def : InstRW<[VecStr, Lat5], (instregex "VSTRCZ(B|F|H)S$")>;
 // Vector: Packed-decimal instructions
 //===----------------------------------------------------------------------===//
 
-def : InstRW<[VecDF, VecDF, Lat10, GroupAlone], (instregex "VLIP$")>;
-def : InstRW<[VecDFX, LSU, Lat12, GroupAlone], (instregex "VPKZ$")>;
-def : InstRW<[VecDFX, FXb, LSU, Lat12, GroupAlone], (instregex "VUPKZ$")>;
+def : InstRW<[VecDF, VecDF, Lat10], (instregex "VLIP$")>;
+def : InstRW<[VecDFX, LSU, GroupAlone], (instregex "VPKZ$")>;
+def : InstRW<[VecDFX, FXb, LSU, Lat12, BeginGroup], (instregex "VUPKZ$")>;
 def : InstRW<[VecDF, VecDF, FXb, Lat20, GroupAlone], (instregex "VCVB(G)?$")>;
 def : InstRW<[VecDF, VecDF, FXb, Lat20, GroupAlone], (instregex "VCVD(G)?$")>;
 def : InstRW<[VecDFX], (instregex "V(A|S)P$")>;
diff --git a/lib/Target/X86/X86ISelDAGToDAG.cpp b/lib/Target/X86/X86ISelDAGToDAG.cpp
index 3c4589ab18f6..8f24f98be681 100644
--- a/lib/Target/X86/X86ISelDAGToDAG.cpp
+++ b/lib/Target/X86/X86ISelDAGToDAG.cpp
@@ -1055,7 +1055,10 @@ static bool foldMaskAndShiftToScale(SelectionDAG &DAG, SDValue N,
 
   // Scale the leading zero count down based on the actual size of the value.
   // Also scale it down based on the size of the shift.
-  MaskLZ -= (64 - X.getSimpleValueType().getSizeInBits()) + ShiftAmt;
+  unsigned ScaleDown = (64 - X.getSimpleValueType().getSizeInBits()) + ShiftAmt;
+  if (MaskLZ < ScaleDown)
+    return true;
+  MaskLZ -= ScaleDown;
 
   // The final check is to ensure that any masked out high bits of X are
   // already known to be zero. Otherwise, the mask has a semantic impact
diff --git a/lib/Target/X86/X86ISelLowering.cpp b/lib/Target/X86/X86ISelLowering.cpp
index 44eecd664714..ba8eb8656585 100644
--- a/lib/Target/X86/X86ISelLowering.cpp
+++ b/lib/Target/X86/X86ISelLowering.cpp
@@ -1672,8 +1672,8 @@ X86TargetLowering::X86TargetLowering(const X86TargetMachine &TM,
 
   // TODO: These control memcmp expansion in CGP and could be raised higher, but
   // that needs to benchmarked and balanced with the potential use of vector
-  // load/store types (PR33329).
-  MaxLoadsPerMemcmp = 4;
+  // load/store types (PR33329, PR33914).
+  MaxLoadsPerMemcmp = 2;
   MaxLoadsPerMemcmpOptSize = 2;
 
   // Set loop alignment to 2^ExperimentalPrefLoopAlignment bytes (default: 2^4).
@@ -22022,8 +22022,9 @@ static SDValue LowerScalarImmediateShift(SDValue Op, SelectionDAG &DAG,
         return getTargetVShiftByConstNode(X86Opc, dl, VT, R, ShiftAmt, DAG);
 
       // i64 SRA needs to be performed as partial shifts.
-      if ((VT == MVT::v2i64 || (Subtarget.hasInt256() && VT == MVT::v4i64)) &&
-          Op.getOpcode() == ISD::SRA && !Subtarget.hasXOP())
+      if (((!Subtarget.hasXOP() && VT == MVT::v2i64) ||
+           (Subtarget.hasInt256() && VT == MVT::v4i64)) &&
+          Op.getOpcode() == ISD::SRA)
         return ArithmeticShiftRight64(ShiftAmt);
 
       if (VT == MVT::v16i8 ||
diff --git a/lib/Transforms/Scalar/JumpThreading.cpp b/lib/Transforms/Scalar/JumpThreading.cpp
index 4056cc5cb346..dc9143bebc45 100644
--- a/lib/Transforms/Scalar/JumpThreading.cpp
+++ b/lib/Transforms/Scalar/JumpThreading.cpp
@@ -64,6 +64,11 @@ ImplicationSearchThreshold(
            "condition to use to thread over a weaker condition"),
   cl::init(3), cl::Hidden);
 
+static cl::opt<bool> PrintLVIAfterJumpThreading(
+    "print-lvi-after-jump-threading",
+    cl::desc("Print the LazyValueInfo cache after JumpThreading"), cl::init(false),
+    cl::Hidden);
+
 namespace {
   /// This pass performs 'jump threading', which looks at blocks that have
   /// multiple predecessors and multiple successors.  If one or more of the
@@ -93,9 +98,10 @@ namespace {
     bool runOnFunction(Function &F) override;
 
     void getAnalysisUsage(AnalysisUsage &AU) const override {
+      if (PrintLVIAfterJumpThreading)
+        AU.addRequired<DominatorTreeWrapperPass>();
       AU.addRequired<AAResultsWrapperPass>();
       AU.addRequired<LazyValueInfoWrapperPass>();
-      AU.addPreserved<LazyValueInfoWrapperPass>();
       AU.addPreserved<GlobalsAAWrapperPass>();
       AU.addRequired<TargetLibraryInfoWrapperPass>();
     }
@@ -137,8 +143,14 @@ bool JumpThreading::runOnFunction(Function &F) {
     BFI.reset(new BlockFrequencyInfo(F, *BPI, LI));
   }
 
-  return Impl.runImpl(F, TLI, LVI, AA, HasProfileData, std::move(BFI),
-                      std::move(BPI));
+  bool Changed = Impl.runImpl(F, TLI, LVI, AA, HasProfileData, std::move(BFI),
+                              std::move(BPI));
+  if (PrintLVIAfterJumpThreading) {
+    dbgs() << "LVI for function '" << F.getName() << "':\n";
+    LVI->printLVI(F, getAnalysis<DominatorTreeWrapperPass>().getDomTree(),
+                  dbgs());
+  }
+  return Changed;
 }
 
 PreservedAnalyses JumpThreadingPass::run(Function &F,
@@ -231,13 +243,15 @@ bool JumpThreadingPass::runImpl(Function &F, TargetLibraryInfo *TLI_,
       // Can't thread an unconditional jump, but if the block is "almost
       // empty", we can replace uses of it with uses of the successor and make
       // this dead.
-      // We should not eliminate the loop header either, because eliminating
-      // a loop header might later prevent LoopSimplify from transforming nested
-      // loops into simplified form.
+      // We should not eliminate the loop header or latch either, because
+      // eliminating a loop header or latch might later prevent LoopSimplify
+      // from transforming nested loops into simplified form. We will rely on
+      // later passes in backend to clean up empty blocks.
       if (BI && BI->isUnconditional() &&
           BB != &BB->getParent()->getEntryBlock() &&
           // If the terminator is the only non-phi instruction, try to nuke it.
-          BB->getFirstNonPHIOrDbg()->isTerminator() && !LoopHeaders.count(BB)) {
+          BB->getFirstNonPHIOrDbg()->isTerminator() && !LoopHeaders.count(BB) &&
+          !LoopHeaders.count(BI->getSuccessor(0))) {
         // FIXME: It is always conservatively correct to drop the info
         // for a block even if it doesn't get erased.  This isn't totally
         // awesome, but it allows us to use AssertingVH to prevent nasty
diff --git a/lib/Transforms/Utils/LoopUtils.cpp b/lib/Transforms/Utils/LoopUtils.cpp
index 58b70be95d99..3c522786641a 100644
--- a/lib/Transforms/Utils/LoopUtils.cpp
+++ b/lib/Transforms/Utils/LoopUtils.cpp
@@ -1376,16 +1376,21 @@ Value *llvm::createTargetReduction(IRBuilder<> &Builder,
   }
 }
 
-void llvm::propagateIRFlags(Value *I, ArrayRef<Value *> VL) {
-  if (auto *VecOp = dyn_cast<Instruction>(I)) {
-    if (auto *I0 = dyn_cast<Instruction>(VL[0])) {
-      // VecOVp is initialized to the 0th scalar, so start counting from index
-      // '1'.
-      VecOp->copyIRFlags(I0);
-      for (int i = 1, e = VL.size(); i < e; ++i) {
-        if (auto *Scalar = dyn_cast<Instruction>(VL[i]))
-          VecOp->andIRFlags(Scalar);
-      }
-    }
+void llvm::propagateIRFlags(Value *I, ArrayRef<Value *> VL, Value *OpValue) {
+  auto *VecOp = dyn_cast<Instruction>(I);
+  if (!VecOp)
+    return;
+  auto *Intersection = (OpValue == nullptr) ? dyn_cast<Instruction>(VL[0])
+                                            : dyn_cast<Instruction>(OpValue);
+  if (!Intersection)
+    return;
+  const unsigned Opcode = Intersection->getOpcode();
+  VecOp->copyIRFlags(Intersection);
+  for (auto *V : VL) {
+    auto *Instr = dyn_cast<Instruction>(V);
+    if (!Instr)
+      continue;
+    if (OpValue == nullptr || Opcode == Instr->getOpcode())
+      VecOp->andIRFlags(V);
   }
 }
diff --git a/lib/Transforms/Utils/SimplifyCFG.cpp b/lib/Transforms/Utils/SimplifyCFG.cpp
index dee658f98393..8784b9702141 100644
--- a/lib/Transforms/Utils/SimplifyCFG.cpp
+++ b/lib/Transforms/Utils/SimplifyCFG.cpp
@@ -5656,20 +5656,22 @@ static bool TryToMergeLandingPad(LandingPadInst *LPad, BranchInst *BI,
 bool SimplifyCFGOpt::SimplifyUncondBranch(BranchInst *BI,
                                           IRBuilder<> &Builder) {
   BasicBlock *BB = BI->getParent();
+  BasicBlock *Succ = BI->getSuccessor(0);
 
   if (SinkCommon && SinkThenElseCodeToEnd(BI))
     return true;
 
   // If the Terminator is the only non-phi instruction, simplify the block.
-  // if LoopHeader is provided, check if the block is a loop header
-  // (This is for early invocations before loop simplify and vectorization
-  // to keep canonical loop forms for nested loops.
-  // These blocks can be eliminated when the pass is invoked later
-  // in the back-end.)
+  // if LoopHeader is provided, check if the block or its successor is a loop
+  // header (This is for early invocations before loop simplify and
+  // vectorization to keep canonical loop forms for nested loops. These blocks
+  // can be eliminated when the pass is invoked later in the back-end.)
+  bool NeedCanonicalLoop =
+      !LateSimplifyCFG &&
+      (LoopHeaders && (LoopHeaders->count(BB) || LoopHeaders->count(Succ)));
   BasicBlock::iterator I = BB->getFirstNonPHIOrDbg()->getIterator();
   if (I->isTerminator() && BB != &BB->getParent()->getEntryBlock() &&
-      (!LoopHeaders || !LoopHeaders->count(BB)) &&
-      TryToSimplifyUncondBranchFromEmptyBlock(BB))
+      !NeedCanonicalLoop && TryToSimplifyUncondBranchFromEmptyBlock(BB))
     return true;
 
   // If the only instruction in the block is a seteq/setne comparison
diff --git a/test/Analysis/LazyValueAnalysis/lvi-after-jumpthreading.ll b/test/Analysis/LazyValueAnalysis/lvi-after-jumpthreading.ll
index e797b377556e..41bb8c9c8201 100644
--- a/test/Analysis/LazyValueAnalysis/lvi-after-jumpthreading.ll
+++ b/test/Analysis/LazyValueAnalysis/lvi-after-jumpthreading.ll
@@ -1,4 +1,4 @@
-; RUN: opt < %s -jump-threading -print-lazy-value-info -disable-output 2>&1 | FileCheck %s
+; RUN: opt < %s -jump-threading -print-lvi-after-jump-threading -disable-output 2>&1 | FileCheck %s
 
 ; Testing LVI cache after jump-threading
 
@@ -19,13 +19,10 @@ entry:
 ; CHECK-NEXT:     ; LatticeVal for: 'i32 %a' is: overdefined
 ; CHECK-NEXT:     ; LatticeVal for: 'i32 %length' is: overdefined
 ; CHECK-NEXT:     ; LatticeVal for: '  %iv = phi i32 [ 0, %entry ], [ %iv.next, %backedge ]' in BB: '%backedge' is: constantrange<0, 400>
-; CHECK-NEXT:     ; LatticeVal for: '  %iv = phi i32 [ 0, %entry ], [ %iv.next, %backedge ]' in BB: '%exit' is: constantrange<399, 400>
 ; CHECK-NEXT:  %iv = phi i32 [ 0, %entry ], [ %iv.next, %backedge ]
 ; CHECK-NEXT:     ; LatticeVal for: '  %iv.next = add nsw i32 %iv, 1' in BB: '%backedge' is: constantrange<1, 401>
-; CHECK-NEXT:     ; LatticeVal for: '  %iv.next = add nsw i32 %iv, 1' in BB: '%exit' is: constantrange<400, 401>
 ; CHECK-NEXT:  %iv.next = add nsw i32 %iv, 1
 ; CHECK-NEXT:     ; LatticeVal for: '  %cont = icmp slt i32 %iv.next, 400' in BB: '%backedge' is: overdefined
-; CHECK-NEXT:     ; LatticeVal for: '  %cont = icmp slt i32 %iv.next, 400' in BB: '%exit' is: constantrange<0, -1>
 ; CHECK-NEXT:  %cont = icmp slt i32 %iv.next, 400
 ; CHECK-NOT: loop
 loop:
diff --git a/test/CodeGen/AArch64/aarch64-loop-gep-opt.ll b/test/CodeGen/AArch64/aarch64-loop-gep-opt.ll
index 1b2ed4b89521..2b4e438a13aa 100644
--- a/test/CodeGen/AArch64/aarch64-loop-gep-opt.ll
+++ b/test/CodeGen/AArch64/aarch64-loop-gep-opt.ll
@@ -19,9 +19,9 @@ entry:
 
 do.body.i:
 ; CHECK-LABEL: do.body.i:
-; CHECK:          %uglygep2 = getelementptr i8, i8* %uglygep, i64 %3
-; CHECK-NEXT:     %4 = bitcast i8* %uglygep2 to i32*
-; CHECK-NOT:      %uglygep2 = getelementptr i8, i8* %uglygep, i64 1032
+; CHECK:          %uglygep1 = getelementptr i8, i8* %uglygep, i64 %3
+; CHECK-NEXT:     %4 = bitcast i8* %uglygep1 to i32*
+; CHECK-NOT:      %uglygep1 = getelementptr i8, i8* %uglygep, i64 1032
 
 
   %0 = phi i32 [ 256, %entry ], [ %.be, %do.body.i.backedge ]
diff --git a/test/CodeGen/AArch64/aarch64_win64cc_vararg.ll b/test/CodeGen/AArch64/aarch64_win64cc_vararg.ll
index 2546e7c90ce5..89efe335e329 100644
--- a/test/CodeGen/AArch64/aarch64_win64cc_vararg.ll
+++ b/test/CodeGen/AArch64/aarch64_win64cc_vararg.ll
@@ -59,10 +59,12 @@ entry:
 }
 
 ; CHECK-LABEL: f7:
-; CHECK: sub     sp, sp, #16
-; CHECK: add     x8, sp, #8
-; CHECK: add     x0, sp, #8
-; CHECK: stp     x8, x7, [sp], #16
+; CHECK: sub     sp, sp, #32
+; CHECK: add     x8, sp, #24
+; CHECK: str     x7, [sp, #24]
+; CHECK: add     x0, sp, #24
+; CHECK: str     x8, [sp, #8]
+; CHECK: add     sp, sp, #32
 ; CHECK: ret
 define win64cc i8* @f7(i64 %a0, i64 %a1, i64 %a2, i64 %a3, i64 %a4, i64 %a5, i64 %a6, ...) nounwind {
 entry:
diff --git a/test/CodeGen/AArch64/win64_vararg.ll b/test/CodeGen/AArch64/win64_vararg.ll
index b760e4acd16a..3acc7e520c96 100644
--- a/test/CodeGen/AArch64/win64_vararg.ll
+++ b/test/CodeGen/AArch64/win64_vararg.ll
@@ -59,10 +59,12 @@ entry:
 }
 
 ; CHECK-LABEL: f7:
-; CHECK: sub     sp, sp, #16
-; CHECK: add     x8, sp, #8
-; CHECK: add     x0, sp, #8
-; CHECK: stp     x8, x7, [sp], #16
+; CHECK: sub     sp, sp, #32
+; CHECK: add     x8, sp, #24
+; CHECK: str     x7, [sp, #24]
+; CHECK: add     x0, sp, #24
+; CHECK: str     x8, [sp, #8]
+; CHECK: add     sp, sp, #32
 ; CHECK: ret
 define i8* @f7(i64 %a0, i64 %a1, i64 %a2, i64 %a3, i64 %a4, i64 %a5, i64 %a6, ...) nounwind {
 entry:
@@ -79,9 +81,8 @@ entry:
 ; CHECK: stp     x6, x7, [sp, #64]
 ; CHECK: stp     x4, x5, [sp, #48]
 ; CHECK: stp     x2, x3, [sp, #32]
-; CHECK: stp     x8, x1, [sp, #16]
-; CHECK: str     x8, [sp, #8]
-; CHECK: add     sp, sp, #80
+; CHECK: str     x1, [sp, #24]
+; CHECK: stp     x8, x8, [sp], #80
 ; CHECK: ret
 define void @copy1(i64 %a0, ...) nounwind {
 entry:
@@ -93,3 +94,54 @@ entry:
   call void @llvm.va_copy(i8* %cp1, i8* %ap1)
   ret void
 }
+
+declare void @llvm.va_end(i8*)
+declare void @llvm.lifetime.start.p0i8(i64, i8* nocapture) #1
+declare void @llvm.lifetime.end.p0i8(i64, i8* nocapture) #1
+
+declare i32 @__stdio_common_vsprintf(i64, i8*, i64, i8*, i8*, i8*) local_unnamed_addr #3
+declare i64* @__local_stdio_printf_options() local_unnamed_addr #4
+
+; CHECK-LABEL: snprintf
+; CHECK: sub     sp,  sp, #96
+; CHECK: stp     x21, x20, [sp, #16]
+; CHECK: stp     x19, x30, [sp, #32]
+; CHECK: add     x8, sp, #56
+; CHECK: mov     x19, x2
+; CHECK: mov     x20, x1
+; CHECK: mov     x21, x0
+; CHECK: stp     x6, x7, [sp, #80]
+; CHECK: stp     x4, x5, [sp, #64]
+; CHECK: str     x3, [sp, #56]
+; CHECK: str     x8, [sp, #8]
+; CHECK: bl      __local_stdio_printf_options
+; CHECK: ldr     x8, [x0]
+; CHECK: add     x5, sp, #56
+; CHECK: mov     x1, x21
+; CHECK: mov     x2, x20
+; CHECK: orr     x0, x8, #0x2
+; CHECK: mov     x3, x19
+; CHECK: mov     x4, xzr
+; CHECK: bl      __stdio_common_vsprintf
+; CHECK: ldp     x19, x30, [sp, #32]
+; CHECK: ldp     x21, x20, [sp, #16]
+; CHECK: cmp     w0, #0
+; CHECK: csinv   w0, w0, wzr, ge
+; CHECK: add     sp, sp, #96
+; CHECK: ret
+define i32 @snprintf(i8*, i64, i8*, ...) local_unnamed_addr #5 {
+  %4 = alloca i8*, align 8
+  %5 = bitcast i8** %4 to i8*
+  call void @llvm.lifetime.start.p0i8(i64 8, i8* nonnull %5) #2
+  call void @llvm.va_start(i8* nonnull %5)
+  %6 = load i8*, i8** %4, align 8
+  %7 = call i64* @__local_stdio_printf_options() #2
+  %8 = load i64, i64* %7, align 8
+  %9 = or i64 %8, 2
+  %10 = call i32 @__stdio_common_vsprintf(i64 %9, i8* %0, i64 %1, i8* %2, i8* null, i8* %6) #2
+  %11 = icmp sgt i32 %10, -1
+  %12 = select i1 %11, i32 %10, i32 -1
+  call void @llvm.va_end(i8* nonnull %5)
+  call void @llvm.lifetime.end.p0i8(i64 8, i8* nonnull %5) #2
+  ret i32 %12
+}
diff --git a/test/CodeGen/AMDGPU/spill-empty-live-interval.mir b/test/CodeGen/AMDGPU/spill-empty-live-interval.mir
new file mode 100644
index 000000000000..bb58da7a59f6
--- /dev/null
+++ b/test/CodeGen/AMDGPU/spill-empty-live-interval.mir
@@ -0,0 +1,74 @@
+# RUN: llc -mtriple=amdgcn-amd-amdhsa-opencl -verify-machineinstrs -stress-regalloc=1 -start-before=simple-register-coalescing -stop-after=greedy -o - %s | FileCheck %s
+# https://bugs.llvm.org/show_bug.cgi?id=33620
+
+---
+# This would assert due to the empty live interval created for %vreg9
+# on the last S_NOP with an undef subreg use.
+
+# CHECK-LABEL: name: expecting_non_empty_interval
+
+# CHECK: undef %7.sub1 = V_MAC_F32_e32 0, undef %1, undef %7.sub1, implicit %exec
+# CHECK-NEXT: SI_SPILL_V64_SAVE %7, %stack.0, %sgpr0_sgpr1_sgpr2_sgpr3, %sgpr5, 0, implicit %exec :: (store 8 into %stack.0, align 4)
+# CHECK-NEXT: undef %5.sub1 = V_MOV_B32_e32 1786773504, implicit %exec
+# CHECK-NEXT: dead %2 = V_MUL_F32_e32 0, %5.sub1, implicit %exec
+
+# CHECK: S_NOP 0, implicit %6.sub1
+# CHECK-NEXT: %8 = SI_SPILL_V64_RESTORE %stack.0, %sgpr0_sgpr1_sgpr2_sgpr3, %sgpr5, 0, implicit %exec :: (load 8 from %stack.0, align 4)
+# CHECK-NEXT: S_NOP 0, implicit %8.sub1
+# CHECK-NEXT: S_NOP 0, implicit undef %9.sub0
+
+name: expecting_non_empty_interval
+tracksRegLiveness: true
+registers:
+  - { id: 0, class: vreg_64, preferred-register: '' }
+  - { id: 1, class: vgpr_32, preferred-register: '' }
+  - { id: 2, class: vgpr_32, preferred-register: '' }
+  - { id: 3, class: vreg_64, preferred-register: '' }
+body:             |
+  bb.0:
+    successors: %bb.1
+    undef %0.sub1 = V_MAC_F32_e32 0, undef %1, undef %0.sub1, implicit %exec
+    undef %3.sub1 = V_MOV_B32_e32 1786773504, implicit %exec
+    dead %2 = V_MUL_F32_e32 0, %3.sub1, implicit %exec
+
+  bb.1:
+    S_NOP 0, implicit %3.sub1
+    S_NOP 0, implicit %0.sub1
+    S_NOP 0, implicit undef %0.sub0
+    S_ENDPGM
+
+...
+
+# Similar assert which happens when trying to rematerialize.
+# https://bugs.llvm.org/show_bug.cgi?id=33884
+---
+# CHECK-LABEL: name: rematerialize_empty_interval_has_reference
+
+# CHECK-NOT: MOV
+# CHECK: undef %3.sub2 = V_MOV_B32_e32 1786773504, implicit %exec
+
+# CHECK: bb.1:
+# CHECK-NEXT: S_NOP 0, implicit %3.sub2
+# CHECK-NEXT: S_NOP 0, implicit undef %6.sub0
+# CHECK-NEXT: undef %4.sub2 = V_MOV_B32_e32 0, implicit %exec
+# CHECK-NEXT: S_NOP 0, implicit %4.sub2
+name: rematerialize_empty_interval_has_reference
+tracksRegLiveness: true
+registers:
+  - { id: 0, class: vreg_128, preferred-register: '' }
+  - { id: 1, class: vgpr_32, preferred-register: '' }
+  - { id: 2, class: vgpr_32, preferred-register: '' }
+  - { id: 3, class: vreg_128, preferred-register: '' }
+body:             |
+  bb.0:
+    successors: %bb.1
+
+    undef %0.sub2 = V_MOV_B32_e32 0, implicit %exec
+    undef %3.sub2 = V_MOV_B32_e32 1786773504, implicit %exec
+
+  bb.1:
+    S_NOP 0, implicit %3.sub2
+    S_NOP 0, implicit undef %0.sub0
+    S_NOP 0, implicit %0.sub2
+
+...
diff --git a/test/CodeGen/X86/memcmp-minsize.ll b/test/CodeGen/X86/memcmp-minsize.ll
index a7f42644ca2d..a55c40f5bda8 100644
--- a/test/CodeGen/X86/memcmp-minsize.ll
+++ b/test/CodeGen/X86/memcmp-minsize.ll
@@ -527,6 +527,93 @@ define i1 @length16_eq_const(i8* %X) nounwind minsize {
   ret i1 %c
 }
 
+; PR33914 - https://bugs.llvm.org/show_bug.cgi?id=33914
+
+define i32 @length24(i8* %X, i8* %Y) nounwind minsize {
+; X86-LABEL: length24:
+; X86:       # BB#0:
+; X86-NEXT:    subl $16, %esp
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl %ecx, {{[0-9]+}}(%esp)
+; X86-NEXT:    movl %eax, (%esp)
+; X86-NEXT:    andl $0, {{[0-9]+}}(%esp)
+; X86-NEXT:    movl $24, {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length24:
+; X64:       # BB#0:
+; X64-NEXT:    pushq $24
+; X64-NEXT:    popq %rdx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 24) nounwind
+  ret i32 %m
+}
+
+define i1 @length24_eq(i8* %x, i8* %y) nounwind minsize {
+; X86-LABEL: length24_eq:
+; X86:       # BB#0:
+; X86-NEXT:    subl $16, %esp
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    movl %ecx, {{[0-9]+}}(%esp)
+; X86-NEXT:    movl %eax, (%esp)
+; X86-NEXT:    andl $0, {{[0-9]+}}(%esp)
+; X86-NEXT:    movl $24, {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    sete %al
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length24_eq:
+; X64:       # BB#0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    pushq $24
+; X64-NEXT:    popq %rdx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    sete %al
+; X64-NEXT:    popq %rcx
+; X64-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_eq_const(i8* %X) nounwind minsize {
+; X86-LABEL: length24_eq_const:
+; X86:       # BB#0:
+; X86-NEXT:    subl $16, %esp
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-NEXT:    movl %eax, (%esp)
+; X86-NEXT:    andl $0, {{[0-9]+}}(%esp)
+; X86-NEXT:    movl $24, {{[0-9]+}}(%esp)
+; X86-NEXT:    movl $.L.str, {{[0-9]+}}(%esp)
+; X86-NEXT:    calll memcmp
+; X86-NEXT:    testl %eax, %eax
+; X86-NEXT:    setne %al
+; X86-NEXT:    addl $16, %esp
+; X86-NEXT:    retl
+;
+; X64-LABEL: length24_eq_const:
+; X64:       # BB#0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    pushq $24
+; X64-NEXT:    popq %rdx
+; X64-NEXT:    movl $.L.str, %esi
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setne %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([65 x i8], [65 x i8]* @.str, i32 0, i32 0), i64 24) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
 define i32 @length32(i8* %X, i8* %Y) nounwind minsize {
 ; X86-LABEL: length32:
 ; X86:       # BB#0:
diff --git a/test/CodeGen/X86/memcmp-optsize.ll b/test/CodeGen/X86/memcmp-optsize.ll
index 450205a966d2..4a5f30890513 100644
--- a/test/CodeGen/X86/memcmp-optsize.ll
+++ b/test/CodeGen/X86/memcmp-optsize.ll
@@ -699,6 +699,82 @@ define i1 @length16_eq_const(i8* %X) nounwind optsize {
   ret i1 %c
 }
 
+; PR33914 - https://bugs.llvm.org/show_bug.cgi?id=33914
+
+define i32 @length24(i8* %X, i8* %Y) nounwind optsize {
+; 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:    movl $24, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 24) nounwind
+  ret i32 %m
+}
+
+define i1 @length24_eq(i8* %x, i8* %y) nounwind optsize {
+; X86-LABEL: length24_eq:
+; 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:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length24_eq:
+; X64:       # BB#0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $24, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    sete %al
+; X64-NEXT:    popq %rcx
+; X64-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_eq_const(i8* %X) nounwind optsize {
+; X86-LABEL: length24_eq_const:
+; X86:       # BB#0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $24
+; 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:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length24_eq_const:
+; X64:       # BB#0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $.L.str, %esi
+; X64-NEXT:    movl $24, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setne %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([65 x i8], [65 x i8]* @.str, i32 0, i32 0), i64 24) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
 define i32 @length32(i8* %X, i8* %Y) nounwind optsize {
 ; X86-LABEL: length32:
 ; X86:       # BB#0:
diff --git a/test/CodeGen/X86/memcmp.ll b/test/CodeGen/X86/memcmp.ll
index 2e6782765462..889f6a74bf7f 100644
--- a/test/CodeGen/X86/memcmp.ll
+++ b/test/CodeGen/X86/memcmp.ll
@@ -475,25 +475,14 @@ define i1 @length8_eq_const(i8* %X) nounwind {
 
 define i1 @length12_eq(i8* %X, i8* %Y) nounwind {
 ; X86-LABEL: length12_eq:
-; X86:       # BB#0: # %loadbb
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
-; X86-NEXT:    movl (%ecx), %edx
-; X86-NEXT:    cmpl (%eax), %edx
-; X86-NEXT:    jne .LBB14_1
-; X86-NEXT:  # BB#2: # %loadbb1
-; X86-NEXT:    movl 4(%ecx), %edx
-; X86-NEXT:    cmpl 4(%eax), %edx
-; X86-NEXT:    jne .LBB14_1
-; X86-NEXT:  # BB#3: # %loadbb2
-; X86-NEXT:    movl 8(%ecx), %edx
-; X86-NEXT:    xorl %ecx, %ecx
-; X86-NEXT:    cmpl 8(%eax), %edx
-; X86-NEXT:    je .LBB14_4
-; X86-NEXT:  .LBB14_1: # %res_block
-; X86-NEXT:    movl $1, %ecx
-; X86-NEXT:  .LBB14_4: # %endblock
-; X86-NEXT:    testl %ecx, %ecx
+; X86:       # BB#0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $12
+; 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
 ;
@@ -520,40 +509,13 @@ define i1 @length12_eq(i8* %X, i8* %Y) nounwind {
 
 define i32 @length12(i8* %X, i8* %Y) nounwind {
 ; X86-LABEL: length12:
-; X86:       # BB#0: # %loadbb
-; 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 .LBB15_1
-; X86-NEXT:  # BB#2: # %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 .LBB15_1
-; X86-NEXT:  # BB#3: # %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:    jne .LBB15_1
-; X86-NEXT:  # BB#4: # %endblock
-; X86-NEXT:    popl %esi
-; X86-NEXT:    retl
-; X86-NEXT:  .LBB15_1: # %res_block
-; X86-NEXT:    cmpl %edx, %ecx
-; X86-NEXT:    movl $-1, %ecx
-; X86-NEXT:    movl $1, %eax
-; X86-NEXT:    cmovbl %ecx, %eax
-; X86-NEXT:    popl %esi
+; X86:       # BB#0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $12
+; 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: length12:
@@ -588,47 +550,13 @@ define i32 @length12(i8* %X, i8* %Y) nounwind {
 
 define i32 @length16(i8* %X, i8* %Y) nounwind {
 ; X86-LABEL: length16:
-; X86:       # BB#0: # %loadbb
-; 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 .LBB16_1
-; X86-NEXT:  # BB#2: # %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 .LBB16_1
-; X86-NEXT:  # BB#3: # %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 .LBB16_1
-; X86-NEXT:  # BB#4: # %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:    jne .LBB16_1
-; X86-NEXT:  # BB#5: # %endblock
-; X86-NEXT:    popl %esi
-; X86-NEXT:    retl
-; X86-NEXT:  .LBB16_1: # %res_block
-; X86-NEXT:    cmpl %edx, %ecx
-; X86-NEXT:    movl $-1, %ecx
-; X86-NEXT:    movl $1, %eax
-; X86-NEXT:    cmovbl %ecx, %eax
-; X86-NEXT:    popl %esi
+; X86:       # BB#0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $16
+; 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: length16:
@@ -660,32 +588,29 @@ define i32 @length16(i8* %X, i8* %Y) nounwind {
 }
 
 define i1 @length16_eq(i8* %x, i8* %y) nounwind {
-; X86-LABEL: length16_eq:
-; X86:       # BB#0: # %loadbb
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
-; X86-NEXT:    movl (%ecx), %edx
-; X86-NEXT:    cmpl (%eax), %edx
-; X86-NEXT:    jne .LBB17_1
-; X86-NEXT:  # BB#2: # %loadbb1
-; X86-NEXT:    movl 4(%ecx), %edx
-; X86-NEXT:    cmpl 4(%eax), %edx
-; X86-NEXT:    jne .LBB17_1
-; X86-NEXT:  # BB#3: # %loadbb2
-; X86-NEXT:    movl 8(%ecx), %edx
-; X86-NEXT:    cmpl 8(%eax), %edx
-; X86-NEXT:    jne .LBB17_1
-; X86-NEXT:  # BB#4: # %loadbb3
-; X86-NEXT:    movl 12(%ecx), %edx
-; X86-NEXT:    xorl %ecx, %ecx
-; X86-NEXT:    cmpl 12(%eax), %edx
-; X86-NEXT:    je .LBB17_5
-; X86-NEXT:  .LBB17_1: # %res_block
-; X86-NEXT:    movl $1, %ecx
-; X86-NEXT:  .LBB17_5: # %endblock
-; X86-NEXT:    testl %ecx, %ecx
-; X86-NEXT:    setne %al
-; X86-NEXT:    retl
+; X86-NOSSE-LABEL: length16_eq:
+; X86-NOSSE:       # BB#0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $16
+; 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-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
 ;
 ; X64-LABEL: length16_eq:
 ; X64:       # BB#0: # %loadbb
@@ -709,27 +634,27 @@ define i1 @length16_eq(i8* %x, i8* %y) nounwind {
 }
 
 define i1 @length16_eq_const(i8* %X) nounwind {
-; X86-LABEL: length16_eq_const:
-; X86:       # BB#0: # %loadbb
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X86-NEXT:    cmpl $858927408, (%eax) # imm = 0x33323130
-; X86-NEXT:    jne .LBB18_1
-; X86-NEXT:  # BB#2: # %loadbb1
-; X86-NEXT:    cmpl $926299444, 4(%eax) # imm = 0x37363534
-; X86-NEXT:    jne .LBB18_1
-; X86-NEXT:  # BB#3: # %loadbb2
-; X86-NEXT:    cmpl $825243960, 8(%eax) # imm = 0x31303938
-; X86-NEXT:    jne .LBB18_1
-; X86-NEXT:  # BB#4: # %loadbb3
-; X86-NEXT:    xorl %ecx, %ecx
-; X86-NEXT:    cmpl $892613426, 12(%eax) # imm = 0x35343332
-; X86-NEXT:    je .LBB18_5
-; X86-NEXT:  .LBB18_1: # %res_block
-; X86-NEXT:    movl $1, %ecx
-; X86-NEXT:  .LBB18_5: # %endblock
-; X86-NEXT:    testl %ecx, %ecx
-; X86-NEXT:    sete %al
-; X86-NEXT:    retl
+; X86-NOSSE-LABEL: length16_eq_const:
+; X86-NOSSE:       # BB#0:
+; X86-NOSSE-NEXT:    pushl $0
+; X86-NOSSE-NEXT:    pushl $16
+; 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-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
 ;
 ; X64-LABEL: length16_eq_const:
 ; X64:       # BB#0: # %loadbb
@@ -752,6 +677,82 @@ define i1 @length16_eq_const(i8* %X) nounwind {
   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:    movl $24, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
+  %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-LABEL: length24_eq:
+; 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:    sete %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length24_eq:
+; X64:       # BB#0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $24, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    sete %al
+; X64-NEXT:    popq %rcx
+; X64-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_eq_const(i8* %X) nounwind {
+; X86-LABEL: length24_eq_const:
+; X86:       # BB#0:
+; X86-NEXT:    pushl $0
+; X86-NEXT:    pushl $24
+; 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:    setne %al
+; X86-NEXT:    retl
+;
+; X64-LABEL: length24_eq_const:
+; X64:       # BB#0:
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    movl $.L.str, %esi
+; X64-NEXT:    movl $24, %edx
+; X64-NEXT:    callq memcmp
+; X64-NEXT:    testl %eax, %eax
+; X64-NEXT:    setne %al
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+  %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([65 x i8], [65 x i8]* @.str, i32 0, i32 0), i64 24) nounwind
+  %c = icmp ne i32 %m, 0
+  ret i1 %c
+}
+
 define i32 @length32(i8* %X, i8* %Y) nounwind {
 ; X86-LABEL: length32:
 ; X86:       # BB#0:
@@ -764,43 +765,9 @@ define i32 @length32(i8* %X, i8* %Y) nounwind {
 ; X86-NEXT:    retl
 ;
 ; X64-LABEL: length32:
-; X64:       # BB#0: # %loadbb
-; 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 .LBB19_1
-; X64-NEXT:  # BB#2: # %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 .LBB19_1
-; X64-NEXT:  # BB#3: # %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 .LBB19_1
-; X64-NEXT:  # BB#4: # %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:    jne .LBB19_1
-; X64-NEXT:  # BB#5: # %endblock
-; X64-NEXT:    retq
-; X64-NEXT:  .LBB19_1: # %res_block
-; X64-NEXT:    cmpq %rdx, %rcx
-; X64-NEXT:    movl $-1, %ecx
-; X64-NEXT:    movl $1, %eax
-; X64-NEXT:    cmovbl %ecx, %eax
-; X64-NEXT:    retq
+; X64:       # BB#0:
+; X64-NEXT:    movl $32, %edx
+; X64-NEXT:    jmp memcmp # TAILCALL
   %m = tail call i32 @memcmp(i8* %X, i8* %Y, i64 32) nounwind
   ret i32 %m
 }
@@ -820,30 +787,25 @@ define i1 @length32_eq(i8* %x, i8* %y) nounwind {
 ; X86-NEXT:    sete %al
 ; X86-NEXT:    retl
 ;
-; X64-LABEL: length32_eq:
-; X64:       # BB#0: # %loadbb
-; X64-NEXT:    movq (%rdi), %rax
-; X64-NEXT:    cmpq (%rsi), %rax
-; X64-NEXT:    jne .LBB20_1
-; X64-NEXT:  # BB#2: # %loadbb1
-; X64-NEXT:    movq 8(%rdi), %rax
-; X64-NEXT:    cmpq 8(%rsi), %rax
-; X64-NEXT:    jne .LBB20_1
-; X64-NEXT:  # BB#3: # %loadbb2
-; X64-NEXT:    movq 16(%rdi), %rax
-; X64-NEXT:    cmpq 16(%rsi), %rax
-; X64-NEXT:    jne .LBB20_1
-; X64-NEXT:  # BB#4: # %loadbb3
-; X64-NEXT:    movq 24(%rdi), %rcx
-; X64-NEXT:    xorl %eax, %eax
-; X64-NEXT:    cmpq 24(%rsi), %rcx
-; X64-NEXT:    je .LBB20_5
-; X64-NEXT:  .LBB20_1: # %res_block
-; X64-NEXT:    movl $1, %eax
-; X64-NEXT:  .LBB20_5: # %endblock
-; X64-NEXT:    testl %eax, %eax
-; X64-NEXT:    sete %al
-; X64-NEXT:    retq
+; X64-SSE2-LABEL: length32_eq:
+; X64-SSE2:       # BB#0:
+; X64-SSE2-NEXT:    pushq %rax
+; X64-SSE2-NEXT:    movl $32, %edx
+; X64-SSE2-NEXT:    callq memcmp
+; X64-SSE2-NEXT:    testl %eax, %eax
+; X64-SSE2-NEXT:    sete %al
+; X64-SSE2-NEXT:    popq %rcx
+; X64-SSE2-NEXT:    retq
+;
+; X64-AVX2-LABEL: length32_eq:
+; X64-AVX2:       # BB#0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vpcmpeqb (%rsi), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpmovmskb %ymm0, %eax
+; X64-AVX2-NEXT:    cmpl $-1, %eax
+; X64-AVX2-NEXT:    sete %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 32) nounwind
   %cmp = icmp eq i32 %call, 0
   ret i1 %cmp
@@ -862,30 +824,26 @@ define i1 @length32_eq_const(i8* %X) nounwind {
 ; X86-NEXT:    setne %al
 ; X86-NEXT:    retl
 ;
-; X64-LABEL: length32_eq_const:
-; X64:       # BB#0: # %loadbb
-; X64-NEXT:    movabsq $3978425819141910832, %rax # imm = 0x3736353433323130
-; X64-NEXT:    cmpq %rax, (%rdi)
-; X64-NEXT:    jne .LBB21_1
-; X64-NEXT:  # BB#2: # %loadbb1
-; X64-NEXT:    movabsq $3833745473465760056, %rax # imm = 0x3534333231303938
-; X64-NEXT:    cmpq %rax, 8(%rdi)
-; X64-NEXT:    jne .LBB21_1
-; X64-NEXT:  # BB#3: # %loadbb2
-; X64-NEXT:    movabsq $3689065127958034230, %rax # imm = 0x3332313039383736
-; X64-NEXT:    cmpq %rax, 16(%rdi)
-; X64-NEXT:    jne .LBB21_1
-; X64-NEXT:  # BB#4: # %loadbb3
-; X64-NEXT:    xorl %eax, %eax
-; X64-NEXT:    movabsq $3544395820347831604, %rcx # imm = 0x3130393837363534
-; X64-NEXT:    cmpq %rcx, 24(%rdi)
-; X64-NEXT:    je .LBB21_5
-; X64-NEXT:  .LBB21_1: # %res_block
-; X64-NEXT:    movl $1, %eax
-; X64-NEXT:  .LBB21_5: # %endblock
-; X64-NEXT:    testl %eax, %eax
-; X64-NEXT:    setne %al
-; X64-NEXT:    retq
+; X64-SSE2-LABEL: length32_eq_const:
+; X64-SSE2:       # BB#0:
+; X64-SSE2-NEXT:    pushq %rax
+; X64-SSE2-NEXT:    movl $.L.str, %esi
+; X64-SSE2-NEXT:    movl $32, %edx
+; X64-SSE2-NEXT:    callq memcmp
+; X64-SSE2-NEXT:    testl %eax, %eax
+; X64-SSE2-NEXT:    setne %al
+; X64-SSE2-NEXT:    popq %rcx
+; X64-SSE2-NEXT:    retq
+;
+; X64-AVX2-LABEL: length32_eq_const:
+; X64-AVX2:       # BB#0:
+; X64-AVX2-NEXT:    vmovdqu (%rdi), %ymm0
+; X64-AVX2-NEXT:    vpcmpeqb {{.*}}(%rip), %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpmovmskb %ymm0, %eax
+; X64-AVX2-NEXT:    cmpl $-1, %eax
+; X64-AVX2-NEXT:    setne %al
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
   %m = tail call i32 @memcmp(i8* %X, i8* getelementptr inbounds ([65 x i8], [65 x i8]* @.str, i32 0, i32 0), i64 32) nounwind
   %c = icmp ne i32 %m, 0
   ret i1 %c
diff --git a/test/CodeGen/X86/pr33844.ll b/test/CodeGen/X86/pr33844.ll
new file mode 100644
index 000000000000..2585945aa109
--- /dev/null
+++ b/test/CodeGen/X86/pr33844.ll
@@ -0,0 +1,38 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -o - %s | FileCheck %s
+
+target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+@global = external global i32
+@global.1 = external global i64
+
+define void @patatino() {
+; CHECK-LABEL: patatino:
+; CHECK:       # BB#0: # %bb
+; CHECK-NEXT:    movl {{.*}}(%rip), %eax
+; CHECK-NEXT:    movl %eax, %ecx
+; CHECK-NEXT:    shrl $31, %ecx
+; CHECK-NEXT:    addl $2147483647, %ecx # imm = 0x7FFFFFFF
+; CHECK-NEXT:    shrl $31, %ecx
+; CHECK-NEXT:    andl $62, %ecx
+; CHECK-NEXT:    andl $-536870912, %eax # imm = 0xE0000000
+; CHECK-NEXT:    orl %ecx, %eax
+; CHECK-NEXT:    movl %eax, {{.*}}(%rip)
+; CHECK-NEXT:    retq
+bb:
+  %tmp = load i32, i32* @global
+  %tmp1 = lshr i32 %tmp, 31
+  %tmp2 = add nuw nsw i32 %tmp1, 2147483647
+  %tmp3 = load i64, i64* @global.1
+  %tmp4 = shl i64 %tmp3, 23
+  %tmp5 = add nsw i64 %tmp4, 8388639
+  %tmp6 = trunc i64 %tmp5 to i32
+  %tmp7 = lshr i32 %tmp2, %tmp6
+  %tmp8 = load i32, i32* @global
+  %tmp9 = and i32 %tmp7, 62
+  %tmp10 = and i32 %tmp8, -536870912
+  %tmp11 = or i32 %tmp9, %tmp10
+  store i32 %tmp11, i32* @global
+  ret void
+}
diff --git a/test/CodeGen/X86/pr33960.ll b/test/CodeGen/X86/pr33960.ll
new file mode 100644
index 000000000000..fb9236d3ffa2
--- /dev/null
+++ b/test/CodeGen/X86/pr33960.ll
@@ -0,0 +1,39 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc < %s -mtriple=i686-unknown -mattr=+avx | FileCheck %s --check-prefix=X86
+; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+avx | FileCheck %s --check-prefix=X64
+
+@b = external local_unnamed_addr global i32, align 4
+
+define void @PR33960() {
+; X86-LABEL: PR33960:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    movl $0, b
+; X86-NEXT:    retl
+;
+; X64-LABEL: PR33960:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    movl $0, {{.*}}(%rip)
+; X64-NEXT:    retq
+entry:
+  %tmp = insertelement <4 x i32> <i32 undef, i32 -7, i32 -3, i32 undef>, i32 -2, i32 3
+  %predphi26 = insertelement <4 x i32> %tmp, i32 -7, i32 0
+  %tmp1 = trunc <4 x i32> %predphi26 to <4 x i16>
+  %tmp2 = icmp eq <4 x i16> %tmp1, zeroinitializer
+  %tmp3 = icmp eq <4 x i32> undef, zeroinitializer
+  %tmp4 = and <4 x i1> %tmp2, %tmp3
+  %predphi17 = select <4 x i1> %tmp4, <4 x i32> undef, <4 x i32> zeroinitializer
+  %tmp5 = shl <4 x i32> %predphi17, <i32 16, i32 16, i32 16, i32 16>
+  %tmp6 = ashr exact <4 x i32> %tmp5, <i32 16, i32 16, i32 16, i32 16>
+  %tmp7 = or <4 x i32> %tmp6, undef
+  %tmp8 = or <4 x i32> undef, %tmp7
+  %tmp9 = or <4 x i32> undef, %tmp8
+  %tmp10 = or <4 x i32> undef, %tmp9
+  %tmp11 = or <4 x i32> undef, %tmp10
+  %tmp12 = or <4 x i32> undef, %tmp11
+  %bin.rdx = or <4 x i32> %tmp12, undef
+  %bin.rdx19 = or <4 x i32> %bin.rdx, undef
+  %tmp13 = extractelement <4 x i32> %bin.rdx19, i32 0
+  %or = or i32 0, %tmp13
+  store i32 %or, i32* @b, align 4
+  ret void
+}
diff --git a/test/CodeGen/X86/vector-shift-ashr-256.ll b/test/CodeGen/X86/vector-shift-ashr-256.ll
index 5f2b18fc9c03..6bb57d8f5f71 100644
--- a/test/CodeGen/X86/vector-shift-ashr-256.ll
+++ b/test/CodeGen/X86/vector-shift-ashr-256.ll
@@ -1699,10 +1699,9 @@ define <4 x i64> @splatconstant_shift_v4i64(<4 x i64> %a) nounwind {
 ;
 ; XOPAVX2-LABEL: splatconstant_shift_v4i64:
 ; XOPAVX2:       # BB#0:
+; XOPAVX2-NEXT:    vpsrad $7, %ymm0, %ymm1
 ; XOPAVX2-NEXT:    vpsrlq $7, %ymm0, %ymm0
-; XOPAVX2-NEXT:    vpbroadcastq {{.*#+}} ymm1 = [72057594037927936,72057594037927936,72057594037927936,72057594037927936]
-; XOPAVX2-NEXT:    vpxor %ymm1, %ymm0, %ymm0
-; XOPAVX2-NEXT:    vpsubq %ymm1, %ymm0, %ymm0
+; XOPAVX2-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7]
 ; XOPAVX2-NEXT:    retq
 ;
 ; AVX512-LABEL: splatconstant_shift_v4i64:
diff --git a/test/MC/Sparc/sparc-tls-relocations.s b/test/MC/Sparc/sparc-tls-relocations.s
new file mode 100644
index 000000000000..3d1b80b5f191
--- /dev/null
+++ b/test/MC/Sparc/sparc-tls-relocations.s
@@ -0,0 +1,83 @@
+! Testing Sparc TLS relocations emission
+! (for now a couple local ones).
+!
+! RUN: llvm-mc %s -arch=sparc -show-encoding | FileCheck %s --check-prefix=ASM
+! RUN: llvm-mc %s -arch=sparcv9 -show-encoding | FileCheck %s --check-prefix=ASM
+! RUN: llvm-mc %s -arch=sparc -filetype=obj | llvm-readobj -r | FileCheck %s --check-prefix=REL
+! RUN: llvm-mc %s -arch=sparcv9 -filetype=obj | llvm-readobj -r | FileCheck %s --check-prefix=REL
+! RUN: llvm-mc %s -arch=sparc -filetype=obj | llvm-objdump -r -d - | FileCheck %s --check-prefix=OBJDUMP
+! RUN: llvm-mc %s -arch=sparcv9 -filetype=obj | llvm-objdump -r -d - | FileCheck %s --check-prefix=OBJDUMP
+
+! REL: Arch: sparc
+! REL: Relocations [
+! REL: 0x{{[0-9,A-F]+}} R_SPARC_TLS_LE_HIX22 Head 0x0
+! REL: 0x{{[0-9,A-F]+}} R_SPARC_TLS_LE_LOX10 Head 0x0
+! REL: 0x{{[0-9,A-F]+}} R_SPARC_TLS_LDO_HIX22 Head 0x0
+! REL: 0x{{[0-9,A-F]+}} R_SPARC_TLS_LDM_HI22  Head 0x0
+! REL: 0x{{[0-9,A-F]+}} R_SPARC_TLS_LDM_LO10  Head 0x0
+! REL: 0x{{[0-9,A-F]+}} R_SPARC_TLS_LDO_LOX10 Head 0x0
+! REL: ]
+
+
+! OBJDUMP: foo:
+foo:
+! Here we use two different sequences to get the address of a static TLS variable 'Head'
+! (note - there is no intent to have valid assembler function here,
+!  we just check how TLS relocations are emitted)
+!
+! First sequence uses LE_HIX22/LE_LOX10
+
+! OBJDUMP: {{[0-9,a-f]+}}:  31 00 00 00  sethi 0, %i0
+! OBJDUMP: {{[0-9,a-f]+}}:     R_SPARC_TLS_LE_HIX22 Unknown
+! ASM: sethi %tle_hix22(Head), %i0 ! encoding: [0x31,0x00,0x00,0x00]
+! ASM:                                 !   fixup A - offset: 0, value: %tle_hix22(Head), kind: fixup_sparc_tls_le_hix22
+        sethi %tle_hix22(Head), %i0
+
+! OBJDUMP: {{[0-9,a-f]+}}:  b0 1e 20 00  xor %i0, 0, %i0
+! OBJDUMP: {{[0-9,a-f]+}}:     R_SPARC_TLS_LE_LOX10 Unknown
+! ASM: xor %i0, %tle_lox10(Head), %i0 ! encoding: [0xb0,0x1e,0x20,0x00]
+! ASM:                                    !   fixup A - offset: 0, value: %tle_lox10(Head), kind: fixup_sparc_tls_le_lox10
+        xor %i0, %tle_lox10(Head), %i0
+
+
+! Second sequence is for PIC, so it is more complicated.
+! It uses LDO_HIX22/LDO_LOX10/LDO_ADD/LDM_HI22/LDM_LO10/LDM_ADD/LDM_CALL
+
+! OBJDUMP: {{[0-9,a-f]+}}:  33 00 00 00  sethi 0, %i1
+! OBJDUMP: {{[0-9,a-f]+}}:     R_SPARC_TLS_LDO_HIX22 Unknown
+! ASM: sethi %tldo_hix22(Head), %i1 ! encoding: [0x33,0b00AAAAAA,A,A]
+! ASM:                                  !   fixup A - offset: 0, value: %tldo_hix22(Head), kind: fixup_sparc_tls_ldo_hix22
+        sethi %tldo_hix22(Head), %i1
+
+! OBJDUMP: {{[0-9,a-f]+}}:  35 00 00 00  sethi 0, %i2
+! OBJDUMP: {{[0-9,a-f]+}}:     R_SPARC_TLS_LDM_HI22 Unknown
+! ASM: sethi %tldm_hi22(Head), %i2 ! encoding: [0x35,0b00AAAAAA,A,A]
+! ASM:                                 !   fixup A - offset: 0, value: %tldm_hi22(Head), kind: fixup_sparc_tls_ldm_hi22
+        sethi %tldm_hi22(Head), %i2
+
+! OBJDUMP: {{[0-9,a-f]+}}:  b4 06 a0 00  add %i2, 0, %i2
+! OBJDUMP: {{[0-9,a-f]+}}:     R_SPARC_TLS_LDM_LO10 Unknown
+! ASM: add %i2, %tldm_lo10(Head), %i2 ! encoding: [0xb4,0x06,0b101000AA,A]
+! ASM:                                    !   fixup A - offset: 0, value: %tldm_lo10(Head), kind: fixup_sparc_tls_ldm_lo10
+        add %i2, %tldm_lo10(Head), %i2
+
+	! ???error from llvm-mc on the next asm line???
+	! add %i0, %i2, %o0, %tldm_add(Head)
+
+! OBJDUMP: {{[0-9,a-f]+}}:  b0 1e 60 00  xor %i1, 0, %i0
+! OBJDUMP: {{[0-9,a-f]+}}:     R_SPARC_TLS_LDO_LOX10 Unknown
+! ASM: xor %i1, %tldo_lox10(Head), %i0 ! encoding: [0xb0,0x1e,0b011000AA,A]
+! ASM:                                     !   fixup A - offset: 0, value: %tldo_lox10(Head), kind: fixup_sparc_tls_ldo_lox10
+        xor %i1, %tldo_lox10(Head), %i0
+
+        ! ???error from llvm-mc on the next asm line???
+        ! call __tls_get_addr, %tldm_call(Head)
+        ! nop
+        ! ???error from llvm-mc on the next asm line???
+        ! add %o0, %i0, %i0, %tldo_add(Head)
+
+        .type  Head,@object
+        .section      .tbss,#alloc,#write,#tls
+Head:
+        .word  0
+        .size  Head, 4
diff --git a/test/Transforms/CodeGenPrepare/X86/memcmp.ll b/test/Transforms/CodeGenPrepare/X86/memcmp.ll
index 1dfc08761965..c5281a9e5733 100644
--- a/test/Transforms/CodeGenPrepare/X86/memcmp.ll
+++ b/test/Transforms/CodeGenPrepare/X86/memcmp.ll
@@ -238,91 +238,9 @@ define i32 @cmp6(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 }
 
 define i32 @cmp7(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
-; X32-LABEL: @cmp7(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP2]])
-; X32-NEXT:    [[TMP5:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP3]])
-; X32-NEXT:    [[TMP6:%.*]] = icmp eq i32 [[TMP4]], [[TMP5]]
-; X32-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X32:       res_block:
-; X32-NEXT:    [[PHI_SRC1:%.*]] = phi i32 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP17:%.*]], [[LOADBB1]] ]
-; X32-NEXT:    [[PHI_SRC2:%.*]] = phi i32 [ [[TMP5]], [[LOADBB]] ], [ [[TMP18:%.*]], [[LOADBB1]] ]
-; X32-NEXT:    [[TMP7:%.*]] = icmp ult i32 [[PHI_SRC1]], [[PHI_SRC2]]
-; X32-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i16*
-; X32-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i16*
-; X32-NEXT:    [[TMP11:%.*]] = getelementptr i16, i16* [[TMP9]], i16 2
-; X32-NEXT:    [[TMP12:%.*]] = getelementptr i16, i16* [[TMP10]], i16 2
-; X32-NEXT:    [[TMP13:%.*]] = load i16, i16* [[TMP11]]
-; X32-NEXT:    [[TMP14:%.*]] = load i16, i16* [[TMP12]]
-; X32-NEXT:    [[TMP15:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP13]])
-; X32-NEXT:    [[TMP16:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP14]])
-; X32-NEXT:    [[TMP17]] = zext i16 [[TMP15]] to i32
-; X32-NEXT:    [[TMP18]] = zext i16 [[TMP16]] to i32
-; X32-NEXT:    [[TMP19:%.*]] = icmp eq i32 [[TMP17]], [[TMP18]]
-; X32-NEXT:    br i1 [[TMP19]], label [[LOADBB2:%.*]], label [[RES_BLOCK]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP20:%.*]] = getelementptr i8, i8* [[X]], i8 6
-; X32-NEXT:    [[TMP21:%.*]] = getelementptr i8, i8* [[Y]], i8 6
-; X32-NEXT:    [[TMP22:%.*]] = load i8, i8* [[TMP20]]
-; X32-NEXT:    [[TMP23:%.*]] = load i8, i8* [[TMP21]]
-; X32-NEXT:    [[TMP24:%.*]] = zext i8 [[TMP22]] to i32
-; X32-NEXT:    [[TMP25:%.*]] = zext i8 [[TMP23]] to i32
-; X32-NEXT:    [[TMP26:%.*]] = sub i32 [[TMP24]], [[TMP25]]
-; X32-NEXT:    br label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP26]], [[LOADBB2]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X32-NEXT:    ret i32 [[PHI_RES]]
-;
-; X64-LABEL: @cmp7(
-; X64-NEXT:  loadbb:
-; X64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X64-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X64-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X64-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X64-NEXT:    [[TMP4:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP2]])
-; X64-NEXT:    [[TMP5:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP3]])
-; X64-NEXT:    [[TMP6:%.*]] = zext i32 [[TMP4]] to i64
-; X64-NEXT:    [[TMP7:%.*]] = zext i32 [[TMP5]] to i64
-; X64-NEXT:    [[TMP8:%.*]] = icmp eq i64 [[TMP6]], [[TMP7]]
-; X64-NEXT:    br i1 [[TMP8]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X64:       res_block:
-; X64-NEXT:    [[PHI_SRC1:%.*]] = phi i64 [ [[TMP6]], [[LOADBB:%.*]] ], [ [[TMP19:%.*]], [[LOADBB1]] ]
-; X64-NEXT:    [[PHI_SRC2:%.*]] = phi i64 [ [[TMP7]], [[LOADBB]] ], [ [[TMP20:%.*]], [[LOADBB1]] ]
-; X64-NEXT:    [[TMP9:%.*]] = icmp ult i64 [[PHI_SRC1]], [[PHI_SRC2]]
-; X64-NEXT:    [[TMP10:%.*]] = select i1 [[TMP9]], i32 -1, i32 1
-; X64-NEXT:    br label [[ENDBLOCK:%.*]]
-; X64:       loadbb1:
-; X64-NEXT:    [[TMP11:%.*]] = bitcast i8* [[X]] to i16*
-; X64-NEXT:    [[TMP12:%.*]] = bitcast i8* [[Y]] to i16*
-; X64-NEXT:    [[TMP13:%.*]] = getelementptr i16, i16* [[TMP11]], i16 2
-; X64-NEXT:    [[TMP14:%.*]] = getelementptr i16, i16* [[TMP12]], i16 2
-; X64-NEXT:    [[TMP15:%.*]] = load i16, i16* [[TMP13]]
-; X64-NEXT:    [[TMP16:%.*]] = load i16, i16* [[TMP14]]
-; X64-NEXT:    [[TMP17:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP15]])
-; X64-NEXT:    [[TMP18:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP16]])
-; X64-NEXT:    [[TMP19]] = zext i16 [[TMP17]] to i64
-; X64-NEXT:    [[TMP20]] = zext i16 [[TMP18]] to i64
-; X64-NEXT:    [[TMP21:%.*]] = icmp eq i64 [[TMP19]], [[TMP20]]
-; X64-NEXT:    br i1 [[TMP21]], label [[LOADBB2:%.*]], label [[RES_BLOCK]]
-; X64:       loadbb2:
-; X64-NEXT:    [[TMP22:%.*]] = getelementptr i8, i8* [[X]], i8 6
-; X64-NEXT:    [[TMP23:%.*]] = getelementptr i8, i8* [[Y]], i8 6
-; X64-NEXT:    [[TMP24:%.*]] = load i8, i8* [[TMP22]]
-; X64-NEXT:    [[TMP25:%.*]] = load i8, i8* [[TMP23]]
-; X64-NEXT:    [[TMP26:%.*]] = zext i8 [[TMP24]] to i32
-; X64-NEXT:    [[TMP27:%.*]] = zext i8 [[TMP25]] to i32
-; X64-NEXT:    [[TMP28:%.*]] = sub i32 [[TMP26]], [[TMP27]]
-; X64-NEXT:    br label [[ENDBLOCK]]
-; X64:       endblock:
-; X64-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP28]], [[LOADBB2]] ], [ [[TMP10]], [[RES_BLOCK]] ]
-; X64-NEXT:    ret i32 [[PHI_RES]]
+; ALL-LABEL: @cmp7(
+; ALL-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 7)
+; ALL-NEXT:    ret i32 [[CALL]]
 ;
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 7)
   ret i32 %call
@@ -379,44 +297,8 @@ define i32 @cmp8(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 
 define i32 @cmp9(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp9(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP2]])
-; X32-NEXT:    [[TMP5:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP3]])
-; X32-NEXT:    [[TMP6:%.*]] = icmp eq i32 [[TMP4]], [[TMP5]]
-; X32-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X32:       res_block:
-; X32-NEXT:    [[PHI_SRC1:%.*]] = phi i32 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP15:%.*]], [[LOADBB1]] ]
-; X32-NEXT:    [[PHI_SRC2:%.*]] = phi i32 [ [[TMP5]], [[LOADBB]] ], [ [[TMP16:%.*]], [[LOADBB1]] ]
-; X32-NEXT:    [[TMP7:%.*]] = icmp ult i32 [[PHI_SRC1]], [[PHI_SRC2]]
-; X32-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP11:%.*]] = getelementptr i32, i32* [[TMP9]], i32 1
-; X32-NEXT:    [[TMP12:%.*]] = getelementptr i32, i32* [[TMP10]], i32 1
-; X32-NEXT:    [[TMP13:%.*]] = load i32, i32* [[TMP11]]
-; X32-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP12]]
-; X32-NEXT:    [[TMP15]] = call i32 @llvm.bswap.i32(i32 [[TMP13]])
-; X32-NEXT:    [[TMP16]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
-; X32-NEXT:    [[TMP17:%.*]] = icmp eq i32 [[TMP15]], [[TMP16]]
-; X32-NEXT:    br i1 [[TMP17]], label [[LOADBB2:%.*]], label [[RES_BLOCK]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP18:%.*]] = getelementptr i8, i8* [[X]], i8 8
-; X32-NEXT:    [[TMP19:%.*]] = getelementptr i8, i8* [[Y]], i8 8
-; X32-NEXT:    [[TMP20:%.*]] = load i8, i8* [[TMP18]]
-; X32-NEXT:    [[TMP21:%.*]] = load i8, i8* [[TMP19]]
-; X32-NEXT:    [[TMP22:%.*]] = zext i8 [[TMP20]] to i32
-; X32-NEXT:    [[TMP23:%.*]] = zext i8 [[TMP21]] to i32
-; X32-NEXT:    [[TMP24:%.*]] = sub i32 [[TMP22]], [[TMP23]]
-; X32-NEXT:    br label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP24]], [[LOADBB2]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X32-NEXT:    ret i32 [[PHI_RES]]
+; X32-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 9)
+; X32-NEXT:    ret i32 [[CALL]]
 ;
 ; X64-LABEL: @cmp9(
 ; X64-NEXT:  loadbb:
@@ -451,48 +333,8 @@ define i32 @cmp9(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 
 define i32 @cmp10(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp10(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP2]])
-; X32-NEXT:    [[TMP5:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP3]])
-; X32-NEXT:    [[TMP6:%.*]] = icmp eq i32 [[TMP4]], [[TMP5]]
-; X32-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X32:       res_block:
-; X32-NEXT:    [[PHI_SRC1:%.*]] = phi i32 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP15:%.*]], [[LOADBB1]] ], [ [[TMP26:%.*]], [[LOADBB2:%.*]] ]
-; X32-NEXT:    [[PHI_SRC2:%.*]] = phi i32 [ [[TMP5]], [[LOADBB]] ], [ [[TMP16:%.*]], [[LOADBB1]] ], [ [[TMP27:%.*]], [[LOADBB2]] ]
-; X32-NEXT:    [[TMP7:%.*]] = icmp ult i32 [[PHI_SRC1]], [[PHI_SRC2]]
-; X32-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP11:%.*]] = getelementptr i32, i32* [[TMP9]], i32 1
-; X32-NEXT:    [[TMP12:%.*]] = getelementptr i32, i32* [[TMP10]], i32 1
-; X32-NEXT:    [[TMP13:%.*]] = load i32, i32* [[TMP11]]
-; X32-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP12]]
-; X32-NEXT:    [[TMP15]] = call i32 @llvm.bswap.i32(i32 [[TMP13]])
-; X32-NEXT:    [[TMP16]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
-; X32-NEXT:    [[TMP17:%.*]] = icmp eq i32 [[TMP15]], [[TMP16]]
-; X32-NEXT:    br i1 [[TMP17]], label [[LOADBB2]], label [[RES_BLOCK]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP18:%.*]] = bitcast i8* [[X]] to i16*
-; X32-NEXT:    [[TMP19:%.*]] = bitcast i8* [[Y]] to i16*
-; X32-NEXT:    [[TMP20:%.*]] = getelementptr i16, i16* [[TMP18]], i16 4
-; X32-NEXT:    [[TMP21:%.*]] = getelementptr i16, i16* [[TMP19]], i16 4
-; X32-NEXT:    [[TMP22:%.*]] = load i16, i16* [[TMP20]]
-; X32-NEXT:    [[TMP23:%.*]] = load i16, i16* [[TMP21]]
-; X32-NEXT:    [[TMP24:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP22]])
-; X32-NEXT:    [[TMP25:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP23]])
-; X32-NEXT:    [[TMP26]] = zext i16 [[TMP24]] to i32
-; X32-NEXT:    [[TMP27]] = zext i16 [[TMP25]] to i32
-; X32-NEXT:    [[TMP28:%.*]] = icmp eq i32 [[TMP26]], [[TMP27]]
-; X32-NEXT:    br i1 [[TMP28]], label [[ENDBLOCK]], label [[RES_BLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB2]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X32-NEXT:    ret i32 [[PHI_RES]]
+; X32-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 10)
+; X32-NEXT:    ret i32 [[CALL]]
 ;
 ; X64-LABEL: @cmp10(
 ; X64-NEXT:  loadbb:
@@ -532,100 +374,9 @@ define i32 @cmp10(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 }
 
 define i32 @cmp11(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
-; X32-LABEL: @cmp11(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP2]])
-; X32-NEXT:    [[TMP5:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP3]])
-; X32-NEXT:    [[TMP6:%.*]] = icmp eq i32 [[TMP4]], [[TMP5]]
-; X32-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X32:       res_block:
-; X32-NEXT:    [[PHI_SRC1:%.*]] = phi i32 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP15:%.*]], [[LOADBB1]] ], [ [[TMP26:%.*]], [[LOADBB2:%.*]] ]
-; X32-NEXT:    [[PHI_SRC2:%.*]] = phi i32 [ [[TMP5]], [[LOADBB]] ], [ [[TMP16:%.*]], [[LOADBB1]] ], [ [[TMP27:%.*]], [[LOADBB2]] ]
-; X32-NEXT:    [[TMP7:%.*]] = icmp ult i32 [[PHI_SRC1]], [[PHI_SRC2]]
-; X32-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP11:%.*]] = getelementptr i32, i32* [[TMP9]], i32 1
-; X32-NEXT:    [[TMP12:%.*]] = getelementptr i32, i32* [[TMP10]], i32 1
-; X32-NEXT:    [[TMP13:%.*]] = load i32, i32* [[TMP11]]
-; X32-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP12]]
-; X32-NEXT:    [[TMP15]] = call i32 @llvm.bswap.i32(i32 [[TMP13]])
-; X32-NEXT:    [[TMP16]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
-; X32-NEXT:    [[TMP17:%.*]] = icmp eq i32 [[TMP15]], [[TMP16]]
-; X32-NEXT:    br i1 [[TMP17]], label [[LOADBB2]], label [[RES_BLOCK]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP18:%.*]] = bitcast i8* [[X]] to i16*
-; X32-NEXT:    [[TMP19:%.*]] = bitcast i8* [[Y]] to i16*
-; X32-NEXT:    [[TMP20:%.*]] = getelementptr i16, i16* [[TMP18]], i16 4
-; X32-NEXT:    [[TMP21:%.*]] = getelementptr i16, i16* [[TMP19]], i16 4
-; X32-NEXT:    [[TMP22:%.*]] = load i16, i16* [[TMP20]]
-; X32-NEXT:    [[TMP23:%.*]] = load i16, i16* [[TMP21]]
-; X32-NEXT:    [[TMP24:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP22]])
-; X32-NEXT:    [[TMP25:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP23]])
-; X32-NEXT:    [[TMP26]] = zext i16 [[TMP24]] to i32
-; X32-NEXT:    [[TMP27]] = zext i16 [[TMP25]] to i32
-; X32-NEXT:    [[TMP28:%.*]] = icmp eq i32 [[TMP26]], [[TMP27]]
-; X32-NEXT:    br i1 [[TMP28]], label [[LOADBB3:%.*]], label [[RES_BLOCK]]
-; X32:       loadbb3:
-; X32-NEXT:    [[TMP29:%.*]] = getelementptr i8, i8* [[X]], i8 10
-; X32-NEXT:    [[TMP30:%.*]] = getelementptr i8, i8* [[Y]], i8 10
-; X32-NEXT:    [[TMP31:%.*]] = load i8, i8* [[TMP29]]
-; X32-NEXT:    [[TMP32:%.*]] = load i8, i8* [[TMP30]]
-; X32-NEXT:    [[TMP33:%.*]] = zext i8 [[TMP31]] to i32
-; X32-NEXT:    [[TMP34:%.*]] = zext i8 [[TMP32]] to i32
-; X32-NEXT:    [[TMP35:%.*]] = sub i32 [[TMP33]], [[TMP34]]
-; X32-NEXT:    br label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP35]], [[LOADBB3]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X32-NEXT:    ret i32 [[PHI_RES]]
-;
-; X64-LABEL: @cmp11(
-; X64-NEXT:  loadbb:
-; X64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i64*
-; X64-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i64*
-; X64-NEXT:    [[TMP2:%.*]] = load i64, i64* [[TMP0]]
-; X64-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
-; X64-NEXT:    [[TMP4:%.*]] = call i64 @llvm.bswap.i64(i64 [[TMP2]])
-; X64-NEXT:    [[TMP5:%.*]] = call i64 @llvm.bswap.i64(i64 [[TMP3]])
-; X64-NEXT:    [[TMP6:%.*]] = icmp eq i64 [[TMP4]], [[TMP5]]
-; X64-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X64:       res_block:
-; X64-NEXT:    [[PHI_SRC1:%.*]] = phi i64 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP17:%.*]], [[LOADBB1]] ]
-; X64-NEXT:    [[PHI_SRC2:%.*]] = phi i64 [ [[TMP5]], [[LOADBB]] ], [ [[TMP18:%.*]], [[LOADBB1]] ]
-; X64-NEXT:    [[TMP7:%.*]] = icmp ult i64 [[PHI_SRC1]], [[PHI_SRC2]]
-; X64-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X64-NEXT:    br label [[ENDBLOCK:%.*]]
-; X64:       loadbb1:
-; X64-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i16*
-; X64-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i16*
-; X64-NEXT:    [[TMP11:%.*]] = getelementptr i16, i16* [[TMP9]], i16 4
-; X64-NEXT:    [[TMP12:%.*]] = getelementptr i16, i16* [[TMP10]], i16 4
-; X64-NEXT:    [[TMP13:%.*]] = load i16, i16* [[TMP11]]
-; X64-NEXT:    [[TMP14:%.*]] = load i16, i16* [[TMP12]]
-; X64-NEXT:    [[TMP15:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP13]])
-; X64-NEXT:    [[TMP16:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP14]])
-; X64-NEXT:    [[TMP17]] = zext i16 [[TMP15]] to i64
-; X64-NEXT:    [[TMP18]] = zext i16 [[TMP16]] to i64
-; X64-NEXT:    [[TMP19:%.*]] = icmp eq i64 [[TMP17]], [[TMP18]]
-; X64-NEXT:    br i1 [[TMP19]], label [[LOADBB2:%.*]], label [[RES_BLOCK]]
-; X64:       loadbb2:
-; X64-NEXT:    [[TMP20:%.*]] = getelementptr i8, i8* [[X]], i8 10
-; X64-NEXT:    [[TMP21:%.*]] = getelementptr i8, i8* [[Y]], i8 10
-; X64-NEXT:    [[TMP22:%.*]] = load i8, i8* [[TMP20]]
-; X64-NEXT:    [[TMP23:%.*]] = load i8, i8* [[TMP21]]
-; X64-NEXT:    [[TMP24:%.*]] = zext i8 [[TMP22]] to i32
-; X64-NEXT:    [[TMP25:%.*]] = zext i8 [[TMP23]] to i32
-; X64-NEXT:    [[TMP26:%.*]] = sub i32 [[TMP24]], [[TMP25]]
-; X64-NEXT:    br label [[ENDBLOCK]]
-; X64:       endblock:
-; X64-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP26]], [[LOADBB2]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X64-NEXT:    ret i32 [[PHI_RES]]
+; ALL-LABEL: @cmp11(
+; ALL-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 11)
+; ALL-NEXT:    ret i32 [[CALL]]
 ;
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 11)
   ret i32 %call
@@ -633,46 +384,8 @@ define i32 @cmp11(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 
 define i32 @cmp12(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp12(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP2]])
-; X32-NEXT:    [[TMP5:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP3]])
-; X32-NEXT:    [[TMP6:%.*]] = icmp eq i32 [[TMP4]], [[TMP5]]
-; X32-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X32:       res_block:
-; X32-NEXT:    [[PHI_SRC1:%.*]] = phi i32 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP15:%.*]], [[LOADBB1]] ], [ [[TMP24:%.*]], [[LOADBB2:%.*]] ]
-; X32-NEXT:    [[PHI_SRC2:%.*]] = phi i32 [ [[TMP5]], [[LOADBB]] ], [ [[TMP16:%.*]], [[LOADBB1]] ], [ [[TMP25:%.*]], [[LOADBB2]] ]
-; X32-NEXT:    [[TMP7:%.*]] = icmp ult i32 [[PHI_SRC1]], [[PHI_SRC2]]
-; X32-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP11:%.*]] = getelementptr i32, i32* [[TMP9]], i32 1
-; X32-NEXT:    [[TMP12:%.*]] = getelementptr i32, i32* [[TMP10]], i32 1
-; X32-NEXT:    [[TMP13:%.*]] = load i32, i32* [[TMP11]]
-; X32-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP12]]
-; X32-NEXT:    [[TMP15]] = call i32 @llvm.bswap.i32(i32 [[TMP13]])
-; X32-NEXT:    [[TMP16]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
-; X32-NEXT:    [[TMP17:%.*]] = icmp eq i32 [[TMP15]], [[TMP16]]
-; X32-NEXT:    br i1 [[TMP17]], label [[LOADBB2]], label [[RES_BLOCK]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP18:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP19:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP20:%.*]] = getelementptr i32, i32* [[TMP18]], i32 2
-; X32-NEXT:    [[TMP21:%.*]] = getelementptr i32, i32* [[TMP19]], i32 2
-; X32-NEXT:    [[TMP22:%.*]] = load i32, i32* [[TMP20]]
-; X32-NEXT:    [[TMP23:%.*]] = load i32, i32* [[TMP21]]
-; X32-NEXT:    [[TMP24]] = call i32 @llvm.bswap.i32(i32 [[TMP22]])
-; X32-NEXT:    [[TMP25]] = call i32 @llvm.bswap.i32(i32 [[TMP23]])
-; X32-NEXT:    [[TMP26:%.*]] = icmp eq i32 [[TMP24]], [[TMP25]]
-; X32-NEXT:    br i1 [[TMP26]], label [[ENDBLOCK]], label [[RES_BLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB2]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X32-NEXT:    ret i32 [[PHI_RES]]
+; X32-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 12)
+; X32-NEXT:    ret i32 [[CALL]]
 ;
 ; X64-LABEL: @cmp12(
 ; X64-NEXT:  loadbb:
@@ -712,268 +425,27 @@ define i32 @cmp12(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 }
 
 define i32 @cmp13(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
-; X32-LABEL: @cmp13(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP2]])
-; X32-NEXT:    [[TMP5:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP3]])
-; X32-NEXT:    [[TMP6:%.*]] = icmp eq i32 [[TMP4]], [[TMP5]]
-; X32-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X32:       res_block:
-; X32-NEXT:    [[PHI_SRC1:%.*]] = phi i32 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP15:%.*]], [[LOADBB1]] ], [ [[TMP24:%.*]], [[LOADBB2:%.*]] ]
-; X32-NEXT:    [[PHI_SRC2:%.*]] = phi i32 [ [[TMP5]], [[LOADBB]] ], [ [[TMP16:%.*]], [[LOADBB1]] ], [ [[TMP25:%.*]], [[LOADBB2]] ]
-; X32-NEXT:    [[TMP7:%.*]] = icmp ult i32 [[PHI_SRC1]], [[PHI_SRC2]]
-; X32-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP11:%.*]] = getelementptr i32, i32* [[TMP9]], i32 1
-; X32-NEXT:    [[TMP12:%.*]] = getelementptr i32, i32* [[TMP10]], i32 1
-; X32-NEXT:    [[TMP13:%.*]] = load i32, i32* [[TMP11]]
-; X32-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP12]]
-; X32-NEXT:    [[TMP15]] = call i32 @llvm.bswap.i32(i32 [[TMP13]])
-; X32-NEXT:    [[TMP16]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
-; X32-NEXT:    [[TMP17:%.*]] = icmp eq i32 [[TMP15]], [[TMP16]]
-; X32-NEXT:    br i1 [[TMP17]], label [[LOADBB2]], label [[RES_BLOCK]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP18:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP19:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP20:%.*]] = getelementptr i32, i32* [[TMP18]], i32 2
-; X32-NEXT:    [[TMP21:%.*]] = getelementptr i32, i32* [[TMP19]], i32 2
-; X32-NEXT:    [[TMP22:%.*]] = load i32, i32* [[TMP20]]
-; X32-NEXT:    [[TMP23:%.*]] = load i32, i32* [[TMP21]]
-; X32-NEXT:    [[TMP24]] = call i32 @llvm.bswap.i32(i32 [[TMP22]])
-; X32-NEXT:    [[TMP25]] = call i32 @llvm.bswap.i32(i32 [[TMP23]])
-; X32-NEXT:    [[TMP26:%.*]] = icmp eq i32 [[TMP24]], [[TMP25]]
-; X32-NEXT:    br i1 [[TMP26]], label [[LOADBB3:%.*]], label [[RES_BLOCK]]
-; X32:       loadbb3:
-; X32-NEXT:    [[TMP27:%.*]] = getelementptr i8, i8* [[X]], i8 12
-; X32-NEXT:    [[TMP28:%.*]] = getelementptr i8, i8* [[Y]], i8 12
-; X32-NEXT:    [[TMP29:%.*]] = load i8, i8* [[TMP27]]
-; X32-NEXT:    [[TMP30:%.*]] = load i8, i8* [[TMP28]]
-; X32-NEXT:    [[TMP31:%.*]] = zext i8 [[TMP29]] to i32
-; X32-NEXT:    [[TMP32:%.*]] = zext i8 [[TMP30]] to i32
-; X32-NEXT:    [[TMP33:%.*]] = sub i32 [[TMP31]], [[TMP32]]
-; X32-NEXT:    br label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP33]], [[LOADBB3]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X32-NEXT:    ret i32 [[PHI_RES]]
-;
-; X64-LABEL: @cmp13(
-; X64-NEXT:  loadbb:
-; X64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i64*
-; X64-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i64*
-; X64-NEXT:    [[TMP2:%.*]] = load i64, i64* [[TMP0]]
-; X64-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
-; X64-NEXT:    [[TMP4:%.*]] = call i64 @llvm.bswap.i64(i64 [[TMP2]])
-; X64-NEXT:    [[TMP5:%.*]] = call i64 @llvm.bswap.i64(i64 [[TMP3]])
-; X64-NEXT:    [[TMP6:%.*]] = icmp eq i64 [[TMP4]], [[TMP5]]
-; X64-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X64:       res_block:
-; X64-NEXT:    [[PHI_SRC1:%.*]] = phi i64 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP17:%.*]], [[LOADBB1]] ]
-; X64-NEXT:    [[PHI_SRC2:%.*]] = phi i64 [ [[TMP5]], [[LOADBB]] ], [ [[TMP18:%.*]], [[LOADBB1]] ]
-; X64-NEXT:    [[TMP7:%.*]] = icmp ult i64 [[PHI_SRC1]], [[PHI_SRC2]]
-; X64-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X64-NEXT:    br label [[ENDBLOCK:%.*]]
-; X64:       loadbb1:
-; X64-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i32*
-; X64-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i32*
-; X64-NEXT:    [[TMP11:%.*]] = getelementptr i32, i32* [[TMP9]], i32 2
-; X64-NEXT:    [[TMP12:%.*]] = getelementptr i32, i32* [[TMP10]], i32 2
-; X64-NEXT:    [[TMP13:%.*]] = load i32, i32* [[TMP11]]
-; X64-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP12]]
-; X64-NEXT:    [[TMP15:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP13]])
-; X64-NEXT:    [[TMP16:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
-; X64-NEXT:    [[TMP17]] = zext i32 [[TMP15]] to i64
-; X64-NEXT:    [[TMP18]] = zext i32 [[TMP16]] to i64
-; X64-NEXT:    [[TMP19:%.*]] = icmp eq i64 [[TMP17]], [[TMP18]]
-; X64-NEXT:    br i1 [[TMP19]], label [[LOADBB2:%.*]], label [[RES_BLOCK]]
-; X64:       loadbb2:
-; X64-NEXT:    [[TMP20:%.*]] = getelementptr i8, i8* [[X]], i8 12
-; X64-NEXT:    [[TMP21:%.*]] = getelementptr i8, i8* [[Y]], i8 12
-; X64-NEXT:    [[TMP22:%.*]] = load i8, i8* [[TMP20]]
-; X64-NEXT:    [[TMP23:%.*]] = load i8, i8* [[TMP21]]
-; X64-NEXT:    [[TMP24:%.*]] = zext i8 [[TMP22]] to i32
-; X64-NEXT:    [[TMP25:%.*]] = zext i8 [[TMP23]] to i32
-; X64-NEXT:    [[TMP26:%.*]] = sub i32 [[TMP24]], [[TMP25]]
-; X64-NEXT:    br label [[ENDBLOCK]]
-; X64:       endblock:
-; X64-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP26]], [[LOADBB2]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X64-NEXT:    ret i32 [[PHI_RES]]
+; ALL-LABEL: @cmp13(
+; ALL-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 13)
+; ALL-NEXT:    ret i32 [[CALL]]
 ;
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 13)
   ret i32 %call
 }
 
 define i32 @cmp14(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
-; X32-LABEL: @cmp14(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP2]])
-; X32-NEXT:    [[TMP5:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP3]])
-; X32-NEXT:    [[TMP6:%.*]] = icmp eq i32 [[TMP4]], [[TMP5]]
-; X32-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X32:       res_block:
-; X32-NEXT:    [[PHI_SRC1:%.*]] = phi i32 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP15:%.*]], [[LOADBB1]] ], [ [[TMP24:%.*]], [[LOADBB2:%.*]] ], [ [[TMP35:%.*]], [[LOADBB3:%.*]] ]
-; X32-NEXT:    [[PHI_SRC2:%.*]] = phi i32 [ [[TMP5]], [[LOADBB]] ], [ [[TMP16:%.*]], [[LOADBB1]] ], [ [[TMP25:%.*]], [[LOADBB2]] ], [ [[TMP36:%.*]], [[LOADBB3]] ]
-; X32-NEXT:    [[TMP7:%.*]] = icmp ult i32 [[PHI_SRC1]], [[PHI_SRC2]]
-; X32-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP11:%.*]] = getelementptr i32, i32* [[TMP9]], i32 1
-; X32-NEXT:    [[TMP12:%.*]] = getelementptr i32, i32* [[TMP10]], i32 1
-; X32-NEXT:    [[TMP13:%.*]] = load i32, i32* [[TMP11]]
-; X32-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP12]]
-; X32-NEXT:    [[TMP15]] = call i32 @llvm.bswap.i32(i32 [[TMP13]])
-; X32-NEXT:    [[TMP16]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
-; X32-NEXT:    [[TMP17:%.*]] = icmp eq i32 [[TMP15]], [[TMP16]]
-; X32-NEXT:    br i1 [[TMP17]], label [[LOADBB2]], label [[RES_BLOCK]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP18:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP19:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP20:%.*]] = getelementptr i32, i32* [[TMP18]], i32 2
-; X32-NEXT:    [[TMP21:%.*]] = getelementptr i32, i32* [[TMP19]], i32 2
-; X32-NEXT:    [[TMP22:%.*]] = load i32, i32* [[TMP20]]
-; X32-NEXT:    [[TMP23:%.*]] = load i32, i32* [[TMP21]]
-; X32-NEXT:    [[TMP24]] = call i32 @llvm.bswap.i32(i32 [[TMP22]])
-; X32-NEXT:    [[TMP25]] = call i32 @llvm.bswap.i32(i32 [[TMP23]])
-; X32-NEXT:    [[TMP26:%.*]] = icmp eq i32 [[TMP24]], [[TMP25]]
-; X32-NEXT:    br i1 [[TMP26]], label [[LOADBB3]], label [[RES_BLOCK]]
-; X32:       loadbb3:
-; X32-NEXT:    [[TMP27:%.*]] = bitcast i8* [[X]] to i16*
-; X32-NEXT:    [[TMP28:%.*]] = bitcast i8* [[Y]] to i16*
-; X32-NEXT:    [[TMP29:%.*]] = getelementptr i16, i16* [[TMP27]], i16 6
-; X32-NEXT:    [[TMP30:%.*]] = getelementptr i16, i16* [[TMP28]], i16 6
-; X32-NEXT:    [[TMP31:%.*]] = load i16, i16* [[TMP29]]
-; X32-NEXT:    [[TMP32:%.*]] = load i16, i16* [[TMP30]]
-; X32-NEXT:    [[TMP33:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP31]])
-; X32-NEXT:    [[TMP34:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP32]])
-; X32-NEXT:    [[TMP35]] = zext i16 [[TMP33]] to i32
-; X32-NEXT:    [[TMP36]] = zext i16 [[TMP34]] to i32
-; X32-NEXT:    [[TMP37:%.*]] = icmp eq i32 [[TMP35]], [[TMP36]]
-; X32-NEXT:    br i1 [[TMP37]], label [[ENDBLOCK]], label [[RES_BLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB3]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X32-NEXT:    ret i32 [[PHI_RES]]
-;
-; X64-LABEL: @cmp14(
-; X64-NEXT:  loadbb:
-; X64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i64*
-; X64-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i64*
-; X64-NEXT:    [[TMP2:%.*]] = load i64, i64* [[TMP0]]
-; X64-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
-; X64-NEXT:    [[TMP4:%.*]] = call i64 @llvm.bswap.i64(i64 [[TMP2]])
-; X64-NEXT:    [[TMP5:%.*]] = call i64 @llvm.bswap.i64(i64 [[TMP3]])
-; X64-NEXT:    [[TMP6:%.*]] = icmp eq i64 [[TMP4]], [[TMP5]]
-; X64-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X64:       res_block:
-; X64-NEXT:    [[PHI_SRC1:%.*]] = phi i64 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP17:%.*]], [[LOADBB1]] ], [ [[TMP28:%.*]], [[LOADBB2:%.*]] ]
-; X64-NEXT:    [[PHI_SRC2:%.*]] = phi i64 [ [[TMP5]], [[LOADBB]] ], [ [[TMP18:%.*]], [[LOADBB1]] ], [ [[TMP29:%.*]], [[LOADBB2]] ]
-; X64-NEXT:    [[TMP7:%.*]] = icmp ult i64 [[PHI_SRC1]], [[PHI_SRC2]]
-; X64-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X64-NEXT:    br label [[ENDBLOCK:%.*]]
-; X64:       loadbb1:
-; X64-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i32*
-; X64-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i32*
-; X64-NEXT:    [[TMP11:%.*]] = getelementptr i32, i32* [[TMP9]], i32 2
-; X64-NEXT:    [[TMP12:%.*]] = getelementptr i32, i32* [[TMP10]], i32 2
-; X64-NEXT:    [[TMP13:%.*]] = load i32, i32* [[TMP11]]
-; X64-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP12]]
-; X64-NEXT:    [[TMP15:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP13]])
-; X64-NEXT:    [[TMP16:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
-; X64-NEXT:    [[TMP17]] = zext i32 [[TMP15]] to i64
-; X64-NEXT:    [[TMP18]] = zext i32 [[TMP16]] to i64
-; X64-NEXT:    [[TMP19:%.*]] = icmp eq i64 [[TMP17]], [[TMP18]]
-; X64-NEXT:    br i1 [[TMP19]], label [[LOADBB2]], label [[RES_BLOCK]]
-; X64:       loadbb2:
-; X64-NEXT:    [[TMP20:%.*]] = bitcast i8* [[X]] to i16*
-; X64-NEXT:    [[TMP21:%.*]] = bitcast i8* [[Y]] to i16*
-; X64-NEXT:    [[TMP22:%.*]] = getelementptr i16, i16* [[TMP20]], i16 6
-; X64-NEXT:    [[TMP23:%.*]] = getelementptr i16, i16* [[TMP21]], i16 6
-; X64-NEXT:    [[TMP24:%.*]] = load i16, i16* [[TMP22]]
-; X64-NEXT:    [[TMP25:%.*]] = load i16, i16* [[TMP23]]
-; X64-NEXT:    [[TMP26:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP24]])
-; X64-NEXT:    [[TMP27:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP25]])
-; X64-NEXT:    [[TMP28]] = zext i16 [[TMP26]] to i64
-; X64-NEXT:    [[TMP29]] = zext i16 [[TMP27]] to i64
-; X64-NEXT:    [[TMP30:%.*]] = icmp eq i64 [[TMP28]], [[TMP29]]
-; X64-NEXT:    br i1 [[TMP30]], label [[ENDBLOCK]], label [[RES_BLOCK]]
-; X64:       endblock:
-; X64-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB2]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X64-NEXT:    ret i32 [[PHI_RES]]
+; ALL-LABEL: @cmp14(
+; ALL-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 14)
+; ALL-NEXT:    ret i32 [[CALL]]
 ;
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 14)
   ret i32 %call
 }
 
 define i32 @cmp15(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
-; X32-LABEL: @cmp15(
-; X32-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 15)
-; X32-NEXT:    ret i32 [[CALL]]
-;
-; X64-LABEL: @cmp15(
-; X64-NEXT:  loadbb:
-; X64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i64*
-; X64-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i64*
-; X64-NEXT:    [[TMP2:%.*]] = load i64, i64* [[TMP0]]
-; X64-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
-; X64-NEXT:    [[TMP4:%.*]] = call i64 @llvm.bswap.i64(i64 [[TMP2]])
-; X64-NEXT:    [[TMP5:%.*]] = call i64 @llvm.bswap.i64(i64 [[TMP3]])
-; X64-NEXT:    [[TMP6:%.*]] = icmp eq i64 [[TMP4]], [[TMP5]]
-; X64-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X64:       res_block:
-; X64-NEXT:    [[PHI_SRC1:%.*]] = phi i64 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP17:%.*]], [[LOADBB1]] ], [ [[TMP28:%.*]], [[LOADBB2:%.*]] ]
-; X64-NEXT:    [[PHI_SRC2:%.*]] = phi i64 [ [[TMP5]], [[LOADBB]] ], [ [[TMP18:%.*]], [[LOADBB1]] ], [ [[TMP29:%.*]], [[LOADBB2]] ]
-; X64-NEXT:    [[TMP7:%.*]] = icmp ult i64 [[PHI_SRC1]], [[PHI_SRC2]]
-; X64-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X64-NEXT:    br label [[ENDBLOCK:%.*]]
-; X64:       loadbb1:
-; X64-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i32*
-; X64-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i32*
-; X64-NEXT:    [[TMP11:%.*]] = getelementptr i32, i32* [[TMP9]], i32 2
-; X64-NEXT:    [[TMP12:%.*]] = getelementptr i32, i32* [[TMP10]], i32 2
-; X64-NEXT:    [[TMP13:%.*]] = load i32, i32* [[TMP11]]
-; X64-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP12]]
-; X64-NEXT:    [[TMP15:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP13]])
-; X64-NEXT:    [[TMP16:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
-; X64-NEXT:    [[TMP17]] = zext i32 [[TMP15]] to i64
-; X64-NEXT:    [[TMP18]] = zext i32 [[TMP16]] to i64
-; X64-NEXT:    [[TMP19:%.*]] = icmp eq i64 [[TMP17]], [[TMP18]]
-; X64-NEXT:    br i1 [[TMP19]], label [[LOADBB2]], label [[RES_BLOCK]]
-; X64:       loadbb2:
-; X64-NEXT:    [[TMP20:%.*]] = bitcast i8* [[X]] to i16*
-; X64-NEXT:    [[TMP21:%.*]] = bitcast i8* [[Y]] to i16*
-; X64-NEXT:    [[TMP22:%.*]] = getelementptr i16, i16* [[TMP20]], i16 6
-; X64-NEXT:    [[TMP23:%.*]] = getelementptr i16, i16* [[TMP21]], i16 6
-; X64-NEXT:    [[TMP24:%.*]] = load i16, i16* [[TMP22]]
-; X64-NEXT:    [[TMP25:%.*]] = load i16, i16* [[TMP23]]
-; X64-NEXT:    [[TMP26:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP24]])
-; X64-NEXT:    [[TMP27:%.*]] = call i16 @llvm.bswap.i16(i16 [[TMP25]])
-; X64-NEXT:    [[TMP28]] = zext i16 [[TMP26]] to i64
-; X64-NEXT:    [[TMP29]] = zext i16 [[TMP27]] to i64
-; X64-NEXT:    [[TMP30:%.*]] = icmp eq i64 [[TMP28]], [[TMP29]]
-; X64-NEXT:    br i1 [[TMP30]], label [[LOADBB3:%.*]], label [[RES_BLOCK]]
-; X64:       loadbb3:
-; X64-NEXT:    [[TMP31:%.*]] = getelementptr i8, i8* [[X]], i8 14
-; X64-NEXT:    [[TMP32:%.*]] = getelementptr i8, i8* [[Y]], i8 14
-; X64-NEXT:    [[TMP33:%.*]] = load i8, i8* [[TMP31]]
-; X64-NEXT:    [[TMP34:%.*]] = load i8, i8* [[TMP32]]
-; X64-NEXT:    [[TMP35:%.*]] = zext i8 [[TMP33]] to i32
-; X64-NEXT:    [[TMP36:%.*]] = zext i8 [[TMP34]] to i32
-; X64-NEXT:    [[TMP37:%.*]] = sub i32 [[TMP35]], [[TMP36]]
-; X64-NEXT:    br label [[ENDBLOCK]]
-; X64:       endblock:
-; X64-NEXT:    [[PHI_RES:%.*]] = phi i32 [ [[TMP37]], [[LOADBB3]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X64-NEXT:    ret i32 [[PHI_RES]]
+; ALL-LABEL: @cmp15(
+; ALL-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 15)
+; ALL-NEXT:    ret i32 [[CALL]]
 ;
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 15)
   ret i32 %call
@@ -981,57 +453,8 @@ define i32 @cmp15(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 
 define i32 @cmp16(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp16(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP2]])
-; X32-NEXT:    [[TMP5:%.*]] = call i32 @llvm.bswap.i32(i32 [[TMP3]])
-; X32-NEXT:    [[TMP6:%.*]] = icmp eq i32 [[TMP4]], [[TMP5]]
-; X32-NEXT:    br i1 [[TMP6]], label [[LOADBB1:%.*]], label [[RES_BLOCK:%.*]]
-; X32:       res_block:
-; X32-NEXT:    [[PHI_SRC1:%.*]] = phi i32 [ [[TMP4]], [[LOADBB:%.*]] ], [ [[TMP15:%.*]], [[LOADBB1]] ], [ [[TMP24:%.*]], [[LOADBB2:%.*]] ], [ [[TMP33:%.*]], [[LOADBB3:%.*]] ]
-; X32-NEXT:    [[PHI_SRC2:%.*]] = phi i32 [ [[TMP5]], [[LOADBB]] ], [ [[TMP16:%.*]], [[LOADBB1]] ], [ [[TMP25:%.*]], [[LOADBB2]] ], [ [[TMP34:%.*]], [[LOADBB3]] ]
-; X32-NEXT:    [[TMP7:%.*]] = icmp ult i32 [[PHI_SRC1]], [[PHI_SRC2]]
-; X32-NEXT:    [[TMP8:%.*]] = select i1 [[TMP7]], i32 -1, i32 1
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP9:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP10:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP11:%.*]] = getelementptr i32, i32* [[TMP9]], i32 1
-; X32-NEXT:    [[TMP12:%.*]] = getelementptr i32, i32* [[TMP10]], i32 1
-; X32-NEXT:    [[TMP13:%.*]] = load i32, i32* [[TMP11]]
-; X32-NEXT:    [[TMP14:%.*]] = load i32, i32* [[TMP12]]
-; X32-NEXT:    [[TMP15]] = call i32 @llvm.bswap.i32(i32 [[TMP13]])
-; X32-NEXT:    [[TMP16]] = call i32 @llvm.bswap.i32(i32 [[TMP14]])
-; X32-NEXT:    [[TMP17:%.*]] = icmp eq i32 [[TMP15]], [[TMP16]]
-; X32-NEXT:    br i1 [[TMP17]], label [[LOADBB2]], label [[RES_BLOCK]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP18:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP19:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP20:%.*]] = getelementptr i32, i32* [[TMP18]], i32 2
-; X32-NEXT:    [[TMP21:%.*]] = getelementptr i32, i32* [[TMP19]], i32 2
-; X32-NEXT:    [[TMP22:%.*]] = load i32, i32* [[TMP20]]
-; X32-NEXT:    [[TMP23:%.*]] = load i32, i32* [[TMP21]]
-; X32-NEXT:    [[TMP24]] = call i32 @llvm.bswap.i32(i32 [[TMP22]])
-; X32-NEXT:    [[TMP25]] = call i32 @llvm.bswap.i32(i32 [[TMP23]])
-; X32-NEXT:    [[TMP26:%.*]] = icmp eq i32 [[TMP24]], [[TMP25]]
-; X32-NEXT:    br i1 [[TMP26]], label [[LOADBB3]], label [[RES_BLOCK]]
-; X32:       loadbb3:
-; X32-NEXT:    [[TMP27:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP28:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP29:%.*]] = getelementptr i32, i32* [[TMP27]], i32 3
-; X32-NEXT:    [[TMP30:%.*]] = getelementptr i32, i32* [[TMP28]], i32 3
-; X32-NEXT:    [[TMP31:%.*]] = load i32, i32* [[TMP29]]
-; X32-NEXT:    [[TMP32:%.*]] = load i32, i32* [[TMP30]]
-; X32-NEXT:    [[TMP33]] = call i32 @llvm.bswap.i32(i32 [[TMP31]])
-; X32-NEXT:    [[TMP34]] = call i32 @llvm.bswap.i32(i32 [[TMP32]])
-; X32-NEXT:    [[TMP35:%.*]] = icmp eq i32 [[TMP33]], [[TMP34]]
-; X32-NEXT:    br i1 [[TMP35]], label [[ENDBLOCK]], label [[RES_BLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB3]] ], [ [[TMP8]], [[RES_BLOCK]] ]
-; X32-NEXT:    ret i32 [[PHI_RES]]
+; X32-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 16)
+; X32-NEXT:    ret i32 [[CALL]]
 ;
 ; X64-LABEL: @cmp16(
 ; X64-NEXT:  loadbb:
@@ -1198,34 +621,8 @@ define i32 @cmp_eq6(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 
 define i32 @cmp_eq7(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; ALL-LABEL: @cmp_eq7(
-; ALL-NEXT:  loadbb:
-; ALL-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; ALL-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; ALL-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; ALL-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; ALL-NEXT:    [[TMP4:%.*]] = icmp ne i32 [[TMP2]], [[TMP3]]
-; ALL-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; ALL:       res_block:
-; ALL-NEXT:    br label [[ENDBLOCK:%.*]]
-; ALL:       loadbb1:
-; ALL-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i16*
-; ALL-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i16*
-; ALL-NEXT:    [[TMP7:%.*]] = getelementptr i16, i16* [[TMP5]], i16 2
-; ALL-NEXT:    [[TMP8:%.*]] = getelementptr i16, i16* [[TMP6]], i16 2
-; ALL-NEXT:    [[TMP9:%.*]] = load i16, i16* [[TMP7]]
-; ALL-NEXT:    [[TMP10:%.*]] = load i16, i16* [[TMP8]]
-; ALL-NEXT:    [[TMP11:%.*]] = icmp ne i16 [[TMP9]], [[TMP10]]
-; ALL-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; ALL:       loadbb2:
-; ALL-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[X]], i8 6
-; ALL-NEXT:    [[TMP13:%.*]] = getelementptr i8, i8* [[Y]], i8 6
-; ALL-NEXT:    [[TMP14:%.*]] = load i8, i8* [[TMP12]]
-; ALL-NEXT:    [[TMP15:%.*]] = load i8, i8* [[TMP13]]
-; ALL-NEXT:    [[TMP16:%.*]] = icmp ne i8 [[TMP14]], [[TMP15]]
-; ALL-NEXT:    br i1 [[TMP16]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; ALL:       endblock:
-; ALL-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB2]] ], [ 1, [[RES_BLOCK]] ]
-; ALL-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
+; ALL-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 7)
+; ALL-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
 ; ALL-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; ALL-NEXT:    ret i32 [[CONV]]
 ;
@@ -1280,34 +677,8 @@ define i32 @cmp_eq8(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 
 define i32 @cmp_eq9(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp_eq9(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = icmp ne i32 [[TMP2]], [[TMP3]]
-; X32-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X32:       res_block:
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP7:%.*]] = getelementptr i32, i32* [[TMP5]], i32 1
-; X32-NEXT:    [[TMP8:%.*]] = getelementptr i32, i32* [[TMP6]], i32 1
-; X32-NEXT:    [[TMP9:%.*]] = load i32, i32* [[TMP7]]
-; X32-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP8]]
-; X32-NEXT:    [[TMP11:%.*]] = icmp ne i32 [[TMP9]], [[TMP10]]
-; X32-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[X]], i8 8
-; X32-NEXT:    [[TMP13:%.*]] = getelementptr i8, i8* [[Y]], i8 8
-; X32-NEXT:    [[TMP14:%.*]] = load i8, i8* [[TMP12]]
-; X32-NEXT:    [[TMP15:%.*]] = load i8, i8* [[TMP13]]
-; X32-NEXT:    [[TMP16:%.*]] = icmp ne i8 [[TMP14]], [[TMP15]]
-; X32-NEXT:    br i1 [[TMP16]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB2]] ], [ 1, [[RES_BLOCK]] ]
-; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
+; X32-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 9)
+; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
 ; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X32-NEXT:    ret i32 [[CONV]]
 ;
@@ -1342,36 +713,8 @@ define i32 @cmp_eq9(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 
 define i32 @cmp_eq10(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp_eq10(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = icmp ne i32 [[TMP2]], [[TMP3]]
-; X32-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X32:       res_block:
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP7:%.*]] = getelementptr i32, i32* [[TMP5]], i32 1
-; X32-NEXT:    [[TMP8:%.*]] = getelementptr i32, i32* [[TMP6]], i32 1
-; X32-NEXT:    [[TMP9:%.*]] = load i32, i32* [[TMP7]]
-; X32-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP8]]
-; X32-NEXT:    [[TMP11:%.*]] = icmp ne i32 [[TMP9]], [[TMP10]]
-; X32-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP12:%.*]] = bitcast i8* [[X]] to i16*
-; X32-NEXT:    [[TMP13:%.*]] = bitcast i8* [[Y]] to i16*
-; X32-NEXT:    [[TMP14:%.*]] = getelementptr i16, i16* [[TMP12]], i16 4
-; X32-NEXT:    [[TMP15:%.*]] = getelementptr i16, i16* [[TMP13]], i16 4
-; X32-NEXT:    [[TMP16:%.*]] = load i16, i16* [[TMP14]]
-; X32-NEXT:    [[TMP17:%.*]] = load i16, i16* [[TMP15]]
-; X32-NEXT:    [[TMP18:%.*]] = icmp ne i16 [[TMP16]], [[TMP17]]
-; X32-NEXT:    br i1 [[TMP18]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB2]] ], [ 1, [[RES_BLOCK]] ]
-; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
+; X32-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 10)
+; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
 ; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X32-NEXT:    ret i32 [[CONV]]
 ;
@@ -1407,78 +750,11 @@ define i32 @cmp_eq10(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 }
 
 define i32 @cmp_eq11(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
-; X32-LABEL: @cmp_eq11(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = icmp ne i32 [[TMP2]], [[TMP3]]
-; X32-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X32:       res_block:
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP7:%.*]] = getelementptr i32, i32* [[TMP5]], i32 1
-; X32-NEXT:    [[TMP8:%.*]] = getelementptr i32, i32* [[TMP6]], i32 1
-; X32-NEXT:    [[TMP9:%.*]] = load i32, i32* [[TMP7]]
-; X32-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP8]]
-; X32-NEXT:    [[TMP11:%.*]] = icmp ne i32 [[TMP9]], [[TMP10]]
-; X32-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP12:%.*]] = bitcast i8* [[X]] to i16*
-; X32-NEXT:    [[TMP13:%.*]] = bitcast i8* [[Y]] to i16*
-; X32-NEXT:    [[TMP14:%.*]] = getelementptr i16, i16* [[TMP12]], i16 4
-; X32-NEXT:    [[TMP15:%.*]] = getelementptr i16, i16* [[TMP13]], i16 4
-; X32-NEXT:    [[TMP16:%.*]] = load i16, i16* [[TMP14]]
-; X32-NEXT:    [[TMP17:%.*]] = load i16, i16* [[TMP15]]
-; X32-NEXT:    [[TMP18:%.*]] = icmp ne i16 [[TMP16]], [[TMP17]]
-; X32-NEXT:    br i1 [[TMP18]], label [[RES_BLOCK]], label [[LOADBB3:%.*]]
-; X32:       loadbb3:
-; X32-NEXT:    [[TMP19:%.*]] = getelementptr i8, i8* [[X]], i8 10
-; X32-NEXT:    [[TMP20:%.*]] = getelementptr i8, i8* [[Y]], i8 10
-; X32-NEXT:    [[TMP21:%.*]] = load i8, i8* [[TMP19]]
-; X32-NEXT:    [[TMP22:%.*]] = load i8, i8* [[TMP20]]
-; X32-NEXT:    [[TMP23:%.*]] = icmp ne i8 [[TMP21]], [[TMP22]]
-; X32-NEXT:    br i1 [[TMP23]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB3]] ], [ 1, [[RES_BLOCK]] ]
-; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
-; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
-; X32-NEXT:    ret i32 [[CONV]]
-;
-; X64-LABEL: @cmp_eq11(
-; X64-NEXT:  loadbb:
-; X64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i64*
-; X64-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i64*
-; X64-NEXT:    [[TMP2:%.*]] = load i64, i64* [[TMP0]]
-; X64-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
-; X64-NEXT:    [[TMP4:%.*]] = icmp ne i64 [[TMP2]], [[TMP3]]
-; X64-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X64:       res_block:
-; X64-NEXT:    br label [[ENDBLOCK:%.*]]
-; X64:       loadbb1:
-; X64-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i16*
-; X64-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i16*
-; X64-NEXT:    [[TMP7:%.*]] = getelementptr i16, i16* [[TMP5]], i16 4
-; X64-NEXT:    [[TMP8:%.*]] = getelementptr i16, i16* [[TMP6]], i16 4
-; X64-NEXT:    [[TMP9:%.*]] = load i16, i16* [[TMP7]]
-; X64-NEXT:    [[TMP10:%.*]] = load i16, i16* [[TMP8]]
-; X64-NEXT:    [[TMP11:%.*]] = icmp ne i16 [[TMP9]], [[TMP10]]
-; X64-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X64:       loadbb2:
-; X64-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[X]], i8 10
-; X64-NEXT:    [[TMP13:%.*]] = getelementptr i8, i8* [[Y]], i8 10
-; X64-NEXT:    [[TMP14:%.*]] = load i8, i8* [[TMP12]]
-; X64-NEXT:    [[TMP15:%.*]] = load i8, i8* [[TMP13]]
-; X64-NEXT:    [[TMP16:%.*]] = icmp ne i8 [[TMP14]], [[TMP15]]
-; X64-NEXT:    br i1 [[TMP16]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X64:       endblock:
-; X64-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB2]] ], [ 1, [[RES_BLOCK]] ]
-; X64-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
-; X64-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
-; X64-NEXT:    ret i32 [[CONV]]
+; ALL-LABEL: @cmp_eq11(
+; ALL-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 11)
+; ALL-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
+; ALL-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
+; ALL-NEXT:    ret i32 [[CONV]]
 ;
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 11)
   %cmp = icmp eq i32 %call, 0
@@ -1488,36 +764,8 @@ define i32 @cmp_eq11(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 
 define i32 @cmp_eq12(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp_eq12(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = icmp ne i32 [[TMP2]], [[TMP3]]
-; X32-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X32:       res_block:
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP7:%.*]] = getelementptr i32, i32* [[TMP5]], i32 1
-; X32-NEXT:    [[TMP8:%.*]] = getelementptr i32, i32* [[TMP6]], i32 1
-; X32-NEXT:    [[TMP9:%.*]] = load i32, i32* [[TMP7]]
-; X32-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP8]]
-; X32-NEXT:    [[TMP11:%.*]] = icmp ne i32 [[TMP9]], [[TMP10]]
-; X32-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP12:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP13:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP14:%.*]] = getelementptr i32, i32* [[TMP12]], i32 2
-; X32-NEXT:    [[TMP15:%.*]] = getelementptr i32, i32* [[TMP13]], i32 2
-; X32-NEXT:    [[TMP16:%.*]] = load i32, i32* [[TMP14]]
-; X32-NEXT:    [[TMP17:%.*]] = load i32, i32* [[TMP15]]
-; X32-NEXT:    [[TMP18:%.*]] = icmp ne i32 [[TMP16]], [[TMP17]]
-; X32-NEXT:    br i1 [[TMP18]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB2]] ], [ 1, [[RES_BLOCK]] ]
-; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
+; X32-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 12)
+; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
 ; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X32-NEXT:    ret i32 [[CONV]]
 ;
@@ -1553,78 +801,11 @@ define i32 @cmp_eq12(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 }
 
 define i32 @cmp_eq13(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
-; X32-LABEL: @cmp_eq13(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = icmp ne i32 [[TMP2]], [[TMP3]]
-; X32-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X32:       res_block:
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP7:%.*]] = getelementptr i32, i32* [[TMP5]], i32 1
-; X32-NEXT:    [[TMP8:%.*]] = getelementptr i32, i32* [[TMP6]], i32 1
-; X32-NEXT:    [[TMP9:%.*]] = load i32, i32* [[TMP7]]
-; X32-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP8]]
-; X32-NEXT:    [[TMP11:%.*]] = icmp ne i32 [[TMP9]], [[TMP10]]
-; X32-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP12:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP13:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP14:%.*]] = getelementptr i32, i32* [[TMP12]], i32 2
-; X32-NEXT:    [[TMP15:%.*]] = getelementptr i32, i32* [[TMP13]], i32 2
-; X32-NEXT:    [[TMP16:%.*]] = load i32, i32* [[TMP14]]
-; X32-NEXT:    [[TMP17:%.*]] = load i32, i32* [[TMP15]]
-; X32-NEXT:    [[TMP18:%.*]] = icmp ne i32 [[TMP16]], [[TMP17]]
-; X32-NEXT:    br i1 [[TMP18]], label [[RES_BLOCK]], label [[LOADBB3:%.*]]
-; X32:       loadbb3:
-; X32-NEXT:    [[TMP19:%.*]] = getelementptr i8, i8* [[X]], i8 12
-; X32-NEXT:    [[TMP20:%.*]] = getelementptr i8, i8* [[Y]], i8 12
-; X32-NEXT:    [[TMP21:%.*]] = load i8, i8* [[TMP19]]
-; X32-NEXT:    [[TMP22:%.*]] = load i8, i8* [[TMP20]]
-; X32-NEXT:    [[TMP23:%.*]] = icmp ne i8 [[TMP21]], [[TMP22]]
-; X32-NEXT:    br i1 [[TMP23]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB3]] ], [ 1, [[RES_BLOCK]] ]
-; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
-; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
-; X32-NEXT:    ret i32 [[CONV]]
-;
-; X64-LABEL: @cmp_eq13(
-; X64-NEXT:  loadbb:
-; X64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i64*
-; X64-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i64*
-; X64-NEXT:    [[TMP2:%.*]] = load i64, i64* [[TMP0]]
-; X64-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
-; X64-NEXT:    [[TMP4:%.*]] = icmp ne i64 [[TMP2]], [[TMP3]]
-; X64-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X64:       res_block:
-; X64-NEXT:    br label [[ENDBLOCK:%.*]]
-; X64:       loadbb1:
-; X64-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i32*
-; X64-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i32*
-; X64-NEXT:    [[TMP7:%.*]] = getelementptr i32, i32* [[TMP5]], i32 2
-; X64-NEXT:    [[TMP8:%.*]] = getelementptr i32, i32* [[TMP6]], i32 2
-; X64-NEXT:    [[TMP9:%.*]] = load i32, i32* [[TMP7]]
-; X64-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP8]]
-; X64-NEXT:    [[TMP11:%.*]] = icmp ne i32 [[TMP9]], [[TMP10]]
-; X64-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X64:       loadbb2:
-; X64-NEXT:    [[TMP12:%.*]] = getelementptr i8, i8* [[X]], i8 12
-; X64-NEXT:    [[TMP13:%.*]] = getelementptr i8, i8* [[Y]], i8 12
-; X64-NEXT:    [[TMP14:%.*]] = load i8, i8* [[TMP12]]
-; X64-NEXT:    [[TMP15:%.*]] = load i8, i8* [[TMP13]]
-; X64-NEXT:    [[TMP16:%.*]] = icmp ne i8 [[TMP14]], [[TMP15]]
-; X64-NEXT:    br i1 [[TMP16]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X64:       endblock:
-; X64-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB2]] ], [ 1, [[RES_BLOCK]] ]
-; X64-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
-; X64-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
-; X64-NEXT:    ret i32 [[CONV]]
+; ALL-LABEL: @cmp_eq13(
+; ALL-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 13)
+; ALL-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
+; ALL-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
+; ALL-NEXT:    ret i32 [[CONV]]
 ;
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 13)
   %cmp = icmp eq i32 %call, 0
@@ -1633,82 +814,11 @@ define i32 @cmp_eq13(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 }
 
 define i32 @cmp_eq14(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
-; X32-LABEL: @cmp_eq14(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = icmp ne i32 [[TMP2]], [[TMP3]]
-; X32-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X32:       res_block:
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP7:%.*]] = getelementptr i32, i32* [[TMP5]], i32 1
-; X32-NEXT:    [[TMP8:%.*]] = getelementptr i32, i32* [[TMP6]], i32 1
-; X32-NEXT:    [[TMP9:%.*]] = load i32, i32* [[TMP7]]
-; X32-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP8]]
-; X32-NEXT:    [[TMP11:%.*]] = icmp ne i32 [[TMP9]], [[TMP10]]
-; X32-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP12:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP13:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP14:%.*]] = getelementptr i32, i32* [[TMP12]], i32 2
-; X32-NEXT:    [[TMP15:%.*]] = getelementptr i32, i32* [[TMP13]], i32 2
-; X32-NEXT:    [[TMP16:%.*]] = load i32, i32* [[TMP14]]
-; X32-NEXT:    [[TMP17:%.*]] = load i32, i32* [[TMP15]]
-; X32-NEXT:    [[TMP18:%.*]] = icmp ne i32 [[TMP16]], [[TMP17]]
-; X32-NEXT:    br i1 [[TMP18]], label [[RES_BLOCK]], label [[LOADBB3:%.*]]
-; X32:       loadbb3:
-; X32-NEXT:    [[TMP19:%.*]] = bitcast i8* [[X]] to i16*
-; X32-NEXT:    [[TMP20:%.*]] = bitcast i8* [[Y]] to i16*
-; X32-NEXT:    [[TMP21:%.*]] = getelementptr i16, i16* [[TMP19]], i16 6
-; X32-NEXT:    [[TMP22:%.*]] = getelementptr i16, i16* [[TMP20]], i16 6
-; X32-NEXT:    [[TMP23:%.*]] = load i16, i16* [[TMP21]]
-; X32-NEXT:    [[TMP24:%.*]] = load i16, i16* [[TMP22]]
-; X32-NEXT:    [[TMP25:%.*]] = icmp ne i16 [[TMP23]], [[TMP24]]
-; X32-NEXT:    br i1 [[TMP25]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB3]] ], [ 1, [[RES_BLOCK]] ]
-; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
-; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
-; X32-NEXT:    ret i32 [[CONV]]
-;
-; X64-LABEL: @cmp_eq14(
-; X64-NEXT:  loadbb:
-; X64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i64*
-; X64-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i64*
-; X64-NEXT:    [[TMP2:%.*]] = load i64, i64* [[TMP0]]
-; X64-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
-; X64-NEXT:    [[TMP4:%.*]] = icmp ne i64 [[TMP2]], [[TMP3]]
-; X64-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X64:       res_block:
-; X64-NEXT:    br label [[ENDBLOCK:%.*]]
-; X64:       loadbb1:
-; X64-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i32*
-; X64-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i32*
-; X64-NEXT:    [[TMP7:%.*]] = getelementptr i32, i32* [[TMP5]], i32 2
-; X64-NEXT:    [[TMP8:%.*]] = getelementptr i32, i32* [[TMP6]], i32 2
-; X64-NEXT:    [[TMP9:%.*]] = load i32, i32* [[TMP7]]
-; X64-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP8]]
-; X64-NEXT:    [[TMP11:%.*]] = icmp ne i32 [[TMP9]], [[TMP10]]
-; X64-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X64:       loadbb2:
-; X64-NEXT:    [[TMP12:%.*]] = bitcast i8* [[X]] to i16*
-; X64-NEXT:    [[TMP13:%.*]] = bitcast i8* [[Y]] to i16*
-; X64-NEXT:    [[TMP14:%.*]] = getelementptr i16, i16* [[TMP12]], i16 6
-; X64-NEXT:    [[TMP15:%.*]] = getelementptr i16, i16* [[TMP13]], i16 6
-; X64-NEXT:    [[TMP16:%.*]] = load i16, i16* [[TMP14]]
-; X64-NEXT:    [[TMP17:%.*]] = load i16, i16* [[TMP15]]
-; X64-NEXT:    [[TMP18:%.*]] = icmp ne i16 [[TMP16]], [[TMP17]]
-; X64-NEXT:    br i1 [[TMP18]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X64:       endblock:
-; X64-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB2]] ], [ 1, [[RES_BLOCK]] ]
-; X64-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
-; X64-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
-; X64-NEXT:    ret i32 [[CONV]]
+; ALL-LABEL: @cmp_eq14(
+; ALL-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 14)
+; ALL-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
+; ALL-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
+; ALL-NEXT:    ret i32 [[CONV]]
 ;
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 14)
   %cmp = icmp eq i32 %call, 0
@@ -1717,52 +827,11 @@ define i32 @cmp_eq14(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 }
 
 define i32 @cmp_eq15(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
-; X32-LABEL: @cmp_eq15(
-; X32-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 15)
-; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
-; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
-; X32-NEXT:    ret i32 [[CONV]]
-;
-; X64-LABEL: @cmp_eq15(
-; X64-NEXT:  loadbb:
-; X64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i64*
-; X64-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i64*
-; X64-NEXT:    [[TMP2:%.*]] = load i64, i64* [[TMP0]]
-; X64-NEXT:    [[TMP3:%.*]] = load i64, i64* [[TMP1]]
-; X64-NEXT:    [[TMP4:%.*]] = icmp ne i64 [[TMP2]], [[TMP3]]
-; X64-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X64:       res_block:
-; X64-NEXT:    br label [[ENDBLOCK:%.*]]
-; X64:       loadbb1:
-; X64-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i32*
-; X64-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i32*
-; X64-NEXT:    [[TMP7:%.*]] = getelementptr i32, i32* [[TMP5]], i32 2
-; X64-NEXT:    [[TMP8:%.*]] = getelementptr i32, i32* [[TMP6]], i32 2
-; X64-NEXT:    [[TMP9:%.*]] = load i32, i32* [[TMP7]]
-; X64-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP8]]
-; X64-NEXT:    [[TMP11:%.*]] = icmp ne i32 [[TMP9]], [[TMP10]]
-; X64-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X64:       loadbb2:
-; X64-NEXT:    [[TMP12:%.*]] = bitcast i8* [[X]] to i16*
-; X64-NEXT:    [[TMP13:%.*]] = bitcast i8* [[Y]] to i16*
-; X64-NEXT:    [[TMP14:%.*]] = getelementptr i16, i16* [[TMP12]], i16 6
-; X64-NEXT:    [[TMP15:%.*]] = getelementptr i16, i16* [[TMP13]], i16 6
-; X64-NEXT:    [[TMP16:%.*]] = load i16, i16* [[TMP14]]
-; X64-NEXT:    [[TMP17:%.*]] = load i16, i16* [[TMP15]]
-; X64-NEXT:    [[TMP18:%.*]] = icmp ne i16 [[TMP16]], [[TMP17]]
-; X64-NEXT:    br i1 [[TMP18]], label [[RES_BLOCK]], label [[LOADBB3:%.*]]
-; X64:       loadbb3:
-; X64-NEXT:    [[TMP19:%.*]] = getelementptr i8, i8* [[X]], i8 14
-; X64-NEXT:    [[TMP20:%.*]] = getelementptr i8, i8* [[Y]], i8 14
-; X64-NEXT:    [[TMP21:%.*]] = load i8, i8* [[TMP19]]
-; X64-NEXT:    [[TMP22:%.*]] = load i8, i8* [[TMP20]]
-; X64-NEXT:    [[TMP23:%.*]] = icmp ne i8 [[TMP21]], [[TMP22]]
-; X64-NEXT:    br i1 [[TMP23]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X64:       endblock:
-; X64-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB3]] ], [ 1, [[RES_BLOCK]] ]
-; X64-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
-; X64-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
-; X64-NEXT:    ret i32 [[CONV]]
+; ALL-LABEL: @cmp_eq15(
+; ALL-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 15)
+; ALL-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
+; ALL-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
+; ALL-NEXT:    ret i32 [[CONV]]
 ;
   %call = tail call i32 @memcmp(i8* %x, i8* %y, i64 15)
   %cmp = icmp eq i32 %call, 0
@@ -1772,45 +841,8 @@ define i32 @cmp_eq15(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 
 define i32 @cmp_eq16(i8* nocapture readonly %x, i8* nocapture readonly %y)  {
 ; X32-LABEL: @cmp_eq16(
-; X32-NEXT:  loadbb:
-; X32-NEXT:    [[TMP0:%.*]] = bitcast i8* [[X:%.*]] to i32*
-; X32-NEXT:    [[TMP1:%.*]] = bitcast i8* [[Y:%.*]] to i32*
-; X32-NEXT:    [[TMP2:%.*]] = load i32, i32* [[TMP0]]
-; X32-NEXT:    [[TMP3:%.*]] = load i32, i32* [[TMP1]]
-; X32-NEXT:    [[TMP4:%.*]] = icmp ne i32 [[TMP2]], [[TMP3]]
-; X32-NEXT:    br i1 [[TMP4]], label [[RES_BLOCK:%.*]], label [[LOADBB1:%.*]]
-; X32:       res_block:
-; X32-NEXT:    br label [[ENDBLOCK:%.*]]
-; X32:       loadbb1:
-; X32-NEXT:    [[TMP5:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP6:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP7:%.*]] = getelementptr i32, i32* [[TMP5]], i32 1
-; X32-NEXT:    [[TMP8:%.*]] = getelementptr i32, i32* [[TMP6]], i32 1
-; X32-NEXT:    [[TMP9:%.*]] = load i32, i32* [[TMP7]]
-; X32-NEXT:    [[TMP10:%.*]] = load i32, i32* [[TMP8]]
-; X32-NEXT:    [[TMP11:%.*]] = icmp ne i32 [[TMP9]], [[TMP10]]
-; X32-NEXT:    br i1 [[TMP11]], label [[RES_BLOCK]], label [[LOADBB2:%.*]]
-; X32:       loadbb2:
-; X32-NEXT:    [[TMP12:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP13:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP14:%.*]] = getelementptr i32, i32* [[TMP12]], i32 2
-; X32-NEXT:    [[TMP15:%.*]] = getelementptr i32, i32* [[TMP13]], i32 2
-; X32-NEXT:    [[TMP16:%.*]] = load i32, i32* [[TMP14]]
-; X32-NEXT:    [[TMP17:%.*]] = load i32, i32* [[TMP15]]
-; X32-NEXT:    [[TMP18:%.*]] = icmp ne i32 [[TMP16]], [[TMP17]]
-; X32-NEXT:    br i1 [[TMP18]], label [[RES_BLOCK]], label [[LOADBB3:%.*]]
-; X32:       loadbb3:
-; X32-NEXT:    [[TMP19:%.*]] = bitcast i8* [[X]] to i32*
-; X32-NEXT:    [[TMP20:%.*]] = bitcast i8* [[Y]] to i32*
-; X32-NEXT:    [[TMP21:%.*]] = getelementptr i32, i32* [[TMP19]], i32 3
-; X32-NEXT:    [[TMP22:%.*]] = getelementptr i32, i32* [[TMP20]], i32 3
-; X32-NEXT:    [[TMP23:%.*]] = load i32, i32* [[TMP21]]
-; X32-NEXT:    [[TMP24:%.*]] = load i32, i32* [[TMP22]]
-; X32-NEXT:    [[TMP25:%.*]] = icmp ne i32 [[TMP23]], [[TMP24]]
-; X32-NEXT:    br i1 [[TMP25]], label [[RES_BLOCK]], label [[ENDBLOCK]]
-; X32:       endblock:
-; X32-NEXT:    [[PHI_RES:%.*]] = phi i32 [ 0, [[LOADBB3]] ], [ 1, [[RES_BLOCK]] ]
-; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[PHI_RES]], 0
+; X32-NEXT:    [[CALL:%.*]] = tail call i32 @memcmp(i8* [[X:%.*]], i8* [[Y:%.*]], i64 16)
+; X32-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
 ; X32-NEXT:    [[CONV:%.*]] = zext i1 [[CMP]] to i32
 ; X32-NEXT:    ret i32 [[CONV]]
 ;
diff --git a/test/Transforms/JumpThreading/pr33605.ll b/test/Transforms/JumpThreading/pr33605.ll
new file mode 100644
index 000000000000..eb8cab90fa50
--- /dev/null
+++ b/test/Transforms/JumpThreading/pr33605.ll
@@ -0,0 +1,64 @@
+; RUN: opt < %s -jump-threading -S | FileCheck %s
+
+; Skip simplifying unconditional branches from empty blocks in simplifyCFG,
+; when it can destroy canonical loop structure.
+
+; void foo();
+; bool test(int a, int b, int *c) {
+;   bool changed = false;
+;   for (unsigned int i = 2; i--;) {
+;     int r = a | b;
+;     if ( r != c[i]) {
+;       c[i] = r;
+;       foo();
+;       changed = true;
+;     }
+;   }
+;   return changed;
+; }
+
+; CHECK-LABEL: @test(
+; CHECK: for.cond:
+; CHECK-NEXT: %i.0 = phi i32 [ 2, %entry ], [ %dec, %if.end ]
+; CHECK: for.body:
+; CHECK: br i1 %cmp, label %if.end, label %if.then
+; CHECK-NOT: br i1 %cmp, label %for.cond, label %if.then
+; CHECK: if.then:
+; CHECK: br label %if.end
+; CHECK-NOT: br label %for.cond
+; CHECK: if.end:
+; CHECK br label %for.cond
+define i1 @test(i32 %a, i32 %b, i32* %c) {
+entry:
+  br label %for.cond
+
+for.cond:                                         ; preds = %if.end, %entry
+  %i.0 = phi i32 [ 2, %entry ], [ %dec, %if.end ]
+  %changed.0.off0 = phi i1 [ false, %entry ], [ %changed.1.off0, %if.end ]
+  %dec = add nsw i32 %i.0, -1
+  %tobool = icmp eq i32 %i.0, 0
+  br i1 %tobool, label %for.cond.cleanup, label %for.body
+
+for.cond.cleanup:                                 ; preds = %for.cond
+  %changed.0.off0.lcssa = phi i1 [ %changed.0.off0, %for.cond ]
+  ret i1 %changed.0.off0.lcssa
+
+for.body:                                         ; preds = %for.cond
+  %or = or i32 %a, %b
+  %idxprom = sext i32 %dec to i64
+  %arrayidx = getelementptr inbounds i32, i32* %c, i64 %idxprom
+  %0 = load i32, i32* %arrayidx, align 4
+  %cmp = icmp eq i32 %or, %0
+  br i1 %cmp, label %if.end, label %if.then
+
+if.then:                                          ; preds = %for.body
+  store i32 %or, i32* %arrayidx, align 4
+  call void @foo()
+  br label %if.end
+
+if.end:                                           ; preds = %for.body, %if.then
+  %changed.1.off0 = phi i1 [ true, %if.then ], [ %changed.0.off0, %for.body ]
+  br label %for.cond
+}
+
+declare void @foo()
diff --git a/test/Transforms/JumpThreading/pr33917.ll b/test/Transforms/JumpThreading/pr33917.ll
new file mode 100644
index 000000000000..30652279a0e1
--- /dev/null
+++ b/test/Transforms/JumpThreading/pr33917.ll
@@ -0,0 +1,57 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -jump-threading -correlated-propagation %s -S | FileCheck %s
+
+target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+declare i8* @foo()
+
+declare i32 @rust_eh_personality() unnamed_addr
+
+; Function Attrs: nounwind
+declare void @llvm.assume(i1) #0
+
+define void @patatino() personality i32 ()* @rust_eh_personality {
+; CHECK-LABEL: @patatino(
+; CHECK-NEXT:  bb9:
+; CHECK-NEXT:    [[T9:%.*]] = invoke i8* @foo()
+; CHECK-NEXT:    to label [[GOOD:%.*]] unwind label [[BAD:%.*]]
+; CHECK:       bad:
+; CHECK-NEXT:    [[T10:%.*]] = landingpad { i8*, i32 }
+; CHECK-NEXT:    cleanup
+; CHECK-NEXT:    resume { i8*, i32 } [[T10]]
+; CHECK:       good:
+; CHECK-NEXT:    [[T11:%.*]] = icmp ne i8* [[T9]], null
+; CHECK-NEXT:    [[T12:%.*]] = zext i1 [[T11]] to i64
+; CHECK-NEXT:    [[COND:%.*]] = icmp eq i64 [[T12]], 1
+; CHECK-NEXT:    br i1 [[COND]], label [[IF_TRUE:%.*]], label [[DONE:%.*]]
+; CHECK:       if_true:
+; CHECK-NEXT:    call void @llvm.assume(i1 [[T11]])
+; CHECK-NEXT:    br label [[DONE]]
+; CHECK:       done:
+; CHECK-NEXT:    ret void
+;
+bb9:
+  %t9 = invoke i8* @foo()
+  to label %good unwind label %bad
+
+bad:
+  %t10 = landingpad { i8*, i32 }
+  cleanup
+  resume { i8*, i32 } %t10
+
+good:
+  %t11 = icmp ne i8* %t9, null
+  %t12 = zext i1 %t11 to i64
+  %cond = icmp eq i64 %t12, 1
+  br i1 %cond, label %if_true, label %done
+
+if_true:
+  call void @llvm.assume(i1 %t11)
+  br label %done
+
+done:
+  ret void
+}
+
+attributes #0 = { nounwind }
diff --git a/test/Transforms/JumpThreading/static-profile.ll b/test/Transforms/JumpThreading/static-profile.ll
index d634a607eabf..505e849f4806 100644
--- a/test/Transforms/JumpThreading/static-profile.ll
+++ b/test/Transforms/JumpThreading/static-profile.ll
@@ -86,7 +86,7 @@ eq_1:
 ; Verify the new backedge:
 ; CHECK: check_2.thread:
 ; CHECK-NEXT: call void @bar()
-; CHECK-NEXT: br label %check_1
+; CHECK-NEXT: br label %check_3.thread
 
 check_2:
   %cond2 = icmp eq i32 %v, 2
@@ -100,7 +100,7 @@ eq_2:
 ; Verify the new backedge:
 ; CHECK: eq_2:
 ; CHECK-NEXT: call void @bar()
-; CHECK-NEXT: br label %check_1
+; CHECK-NEXT: br label %check_3.thread
 
 check_3:
   %condE = icmp eq i32 %v, 3
diff --git a/test/Transforms/LoopUnroll/peel-loop.ll b/test/Transforms/LoopUnroll/peel-loop.ll
index bf0801fc760a..3f7c64d8154c 100644
--- a/test/Transforms/LoopUnroll/peel-loop.ll
+++ b/test/Transforms/LoopUnroll/peel-loop.ll
@@ -18,9 +18,11 @@
 ; CHECK: %[[INC2:.*]] = getelementptr inbounds i32, i32* %p, i64 2
 ; CHECK: store i32 2, i32* %[[INC2]], align 4
 ; CHECK: %[[CMP3:.*]] = icmp eq i32 %k, 3
-; CHECK: br i1 %[[CMP3]], label %for.end, label %[[LOOP:.*]]
+; CHECK: br i1 %[[CMP3]], label %for.end, label %[[LOOP_PH:.*]]
+; CHECK: [[LOOP_PH]]:
+; CHECK: br label %[[LOOP:.*]]
 ; CHECK: [[LOOP]]:
-; CHECK: %[[IV:.*]] = phi i32 [ {{.*}}, %[[LOOP]] ], [ 3, %[[NEXT2]] ]
+; CHECK: %[[IV:.*]] = phi i32 [ 3, %[[LOOP_PH]] ], [ {{.*}}, %[[LOOP]] ]
 
 define void @basic(i32* %p, i32 %k) #0 {
 entry:
@@ -65,9 +67,11 @@ for.end:                                          ; preds = %for.cond.for.end_cr
 ; CHECK: %[[INC2:.*]] = getelementptr inbounds i32, i32* %p, i64 2
 ; CHECK: store i32 2, i32* %[[INC2]], align 4
 ; CHECK: %[[CMP3:.*]] = icmp eq i32 %k, 3
-; CHECK: br i1 %[[CMP3]], label %for.end, label %[[LOOP:.*]]
+; CHECK: br i1 %[[CMP3]], label %for.end, label %[[LOOP_PH:.*]]
+; CHECK: [[LOOP_PH]]:
+; CHECK: br label %[[LOOP:.*]]
 ; CHECK: [[LOOP]]:
-; CHECK: %[[IV:.*]] = phi i32 [ %[[IV:.*]], %[[LOOP]] ], [ 3, %[[NEXT2]] ]
+; CHECK: %[[IV:.*]] = phi i32 [ 3, %[[LOOP_PH]] ], [ %[[IV:.*]], %[[LOOP]] ]
 ; CHECK: %ret = phi i32 [ 0, %entry ], [ 1, %[[NEXT0]] ], [ 2, %[[NEXT1]] ], [ 3, %[[NEXT2]] ], [ %[[IV]], %[[LOOP]] ]
 ; CHECK: ret i32 %ret
 define i32 @output(i32* %p, i32 %k) #0 {
diff --git a/test/Transforms/LoopUnswitch/2015-06-17-Metadata.ll b/test/Transforms/LoopUnswitch/2015-06-17-Metadata.ll
index a215be9d4877..8652829bc511 100644
--- a/test/Transforms/LoopUnswitch/2015-06-17-Metadata.ll
+++ b/test/Transforms/LoopUnswitch/2015-06-17-Metadata.ll
@@ -16,7 +16,7 @@ for.body:                                         ; preds = %for.inc, %for.body.
   %cmp1 = icmp eq i32 %a, 12345
   br i1 %cmp1, label %if.then, label %if.else, !prof !0
 ; CHECK: %cmp1 = icmp eq i32 %a, 12345
-; CHECK-NEXT: br i1 %cmp1, label %for.body.us, label %for.body, !prof !0
+; CHECK-NEXT: br i1 %cmp1, label %for.body.preheader.split.us, label %for.body.preheader.split, !prof !0
 if.then:                                          ; preds = %for.body
 ; CHECK: for.body.us:
 ; CHECK: add nsw i32 %{{.*}}, 123
@@ -53,7 +53,7 @@ entry:
   br label %for.body
 ;CHECK: entry:
 ;CHECK-NEXT: %cmp1 = icmp eq i32 1, 2
-;CHECK-NEXT: br i1 %cmp1, label %for.body, label %for.cond.cleanup.split, !prof !1
+;CHECK-NEXT: br i1 %cmp1, label %entry.split, label %for.cond.cleanup.split, !prof !1
 ;CHECK: for.body:
 for.body:                                         ; preds = %for.inc, %entry
   %inc.i = phi i32 [ 0, %entry ], [ %inc, %if.then ]
diff --git a/test/Transforms/LoopUnswitch/infinite-loop.ll b/test/Transforms/LoopUnswitch/infinite-loop.ll
index 0aef9092a1fe..af8725b02a14 100644
--- a/test/Transforms/LoopUnswitch/infinite-loop.ll
+++ b/test/Transforms/LoopUnswitch/infinite-loop.ll
@@ -6,7 +6,7 @@
 ; Loop unswitching shouldn't trivially unswitch the true case of condition %a
 ; in the code here because it leads to an infinite loop. While this doesn't
 ; contain any instructions with side effects, it's still a kind of side effect.
-; It can trivially unswitch on the false cas of condition %a though.
+; It can trivially unswitch on the false case of condition %a though.
 
 ; STATS: 2 loop-unswitch - Number of branches unswitched
 ; STATS: 2 loop-unswitch - Number of unswitches that are trivial
@@ -16,7 +16,7 @@
 ; CHECK-NEXT: br i1 %a, label %entry.split, label %abort0.split
 
 ; CHECK: entry.split:
-; CHECK-NEXT: br i1 %b, label %for.body, label %abort1.split
+; CHECK-NEXT: br i1 %b, label %entry.split.split, label %abort1.split
 
 ; CHECK: for.body:
 ; CHECK-NEXT: br label %for.body
diff --git a/test/Transforms/LoopVectorize/X86/float-induction-x86.ll b/test/Transforms/LoopVectorize/X86/float-induction-x86.ll
index 31c564779fb2..bf455807c586 100644
--- a/test/Transforms/LoopVectorize/X86/float-induction-x86.ll
+++ b/test/Transforms/LoopVectorize/X86/float-induction-x86.ll
@@ -1,4 +1,4 @@
-; RUN: opt < %s  -O3 -mcpu=core-avx2 -mtriple=x86_64-unknown-linux-gnu -S | FileCheck --check-prefix AUTO_VEC %s
+; RUN: opt < %s  -O3 -latesimplifycfg -mcpu=core-avx2 -mtriple=x86_64-unknown-linux-gnu -S | FileCheck --check-prefix AUTO_VEC %s
 
 ; This test checks auto-vectorization with FP induction variable.
 ; The FP operation is not "fast" and requires "fast-math" function attribute.
diff --git a/test/Transforms/LoopVectorize/float-induction.ll b/test/Transforms/LoopVectorize/float-induction.ll
index a7cc4530ceb3..cf6ec88478be 100644
--- a/test/Transforms/LoopVectorize/float-induction.ll
+++ b/test/Transforms/LoopVectorize/float-induction.ll
@@ -1,7 +1,7 @@
 ; RUN: opt < %s  -loop-vectorize -force-vector-interleave=1 -force-vector-width=4 -dce -instcombine -S | FileCheck --check-prefix VEC4_INTERL1 %s
 ; RUN: opt < %s  -loop-vectorize -force-vector-interleave=2 -force-vector-width=4 -dce -instcombine -S | FileCheck --check-prefix VEC4_INTERL2 %s
 ; RUN: opt < %s  -loop-vectorize -force-vector-interleave=2 -force-vector-width=1 -dce -instcombine -S | FileCheck --check-prefix VEC1_INTERL2 %s
-; RUN: opt < %s  -loop-vectorize -force-vector-interleave=1 -force-vector-width=2 -dce -simplifycfg -instcombine -S | FileCheck --check-prefix VEC2_INTERL1_PRED_STORE %s
+; RUN: opt < %s  -loop-vectorize -force-vector-interleave=1 -force-vector-width=2 -dce -simplifycfg -instcombine -latesimplifycfg -S | FileCheck --check-prefix VEC2_INTERL1_PRED_STORE %s
 
 @fp_inc = common global float 0.000000e+00, align 4
 
diff --git a/test/Transforms/SimplifyCFG/X86/switch_to_lookup_table.ll b/test/Transforms/SimplifyCFG/X86/switch_to_lookup_table.ll
index 656a276969f3..536d1d85c3d8 100644
--- a/test/Transforms/SimplifyCFG/X86/switch_to_lookup_table.ll
+++ b/test/Transforms/SimplifyCFG/X86/switch_to_lookup_table.ll
@@ -1322,8 +1322,8 @@ l6:
 ; Speculation depth must be limited to avoid a zero-cost instruction cycle.
 
 ; CHECK-LABEL: @PR26308(
-; CHECK:       while.body:
-; CHECK-NEXT:  br label %while.body
+; CHECK:       cleanup4:
+; CHECK-NEXT:  br label %cleanup4
 
 define i32 @PR26308(i1 %B, i64 %load) {
 entry:
diff --git a/test/Transforms/SimplifyCFG/multiple-phis.ll b/test/Transforms/SimplifyCFG/multiple-phis.ll
index a6eef09ae646..823e2500eac1 100644
--- a/test/Transforms/SimplifyCFG/multiple-phis.ll
+++ b/test/Transforms/SimplifyCFG/multiple-phis.ll
@@ -1,4 +1,4 @@
-; RUN: opt -simplifycfg -S < %s | FileCheck %s
+; RUN: opt -latesimplifycfg -S < %s | FileCheck %s
 
 ; It's not worthwhile to if-convert one of the phi nodes and leave
 ; the other behind, because that still requires a branch. If
diff --git a/test/Transforms/SimplifyCFG/pr33605.ll b/test/Transforms/SimplifyCFG/pr33605.ll
new file mode 100644
index 000000000000..963b15991263
--- /dev/null
+++ b/test/Transforms/SimplifyCFG/pr33605.ll
@@ -0,0 +1,64 @@
+; RUN: opt < %s -simplifycfg -S | FileCheck %s
+
+; Skip simplifying unconditional branches from empty blocks in simplifyCFG,
+; when it can destroy canonical loop structure.
+
+; void foo();
+; bool test(int a, int b, int *c) {
+;   bool changed = false;
+;   for (unsigned int i = 2; i--;) {
+;     int r = a | b;
+;     if ( r != c[i]) {
+;       c[i] = r;
+;       foo();
+;       changed = true;
+;     }
+;   }
+;   return changed;
+; }
+
+; CHECK-LABEL: @test(
+; CHECK: for.cond:
+; CHECK-NEXT: %i.0 = phi i32 [ 2, %entry ], [ %dec, %if.end ]
+; CHECK: for.body:
+; CHECK: br i1 %cmp, label %if.end, label %if.then
+; CHECK-NOT: br i1 %cmp, label %for.cond, label %if.then
+; CHECK: if.then:
+; CHECK: br label %if.end
+; CHECK-NOT: br label %for.cond
+; CHECK: if.end:
+; CHECK br label %for.cond
+define i1 @test(i32 %a, i32 %b, i32* %c) {
+entry:
+  br label %for.cond
+
+for.cond:                                         ; preds = %if.end, %entry
+  %i.0 = phi i32 [ 2, %entry ], [ %dec, %if.end ]
+  %changed.0.off0 = phi i1 [ false, %entry ], [ %changed.1.off0, %if.end ]
+  %dec = add nsw i32 %i.0, -1
+  %tobool = icmp eq i32 %i.0, 0
+  br i1 %tobool, label %for.cond.cleanup, label %for.body
+
+for.cond.cleanup:                                 ; preds = %for.cond
+  %changed.0.off0.lcssa = phi i1 [ %changed.0.off0, %for.cond ]
+  ret i1 %changed.0.off0.lcssa
+
+for.body:                                         ; preds = %for.cond
+  %or = or i32 %a, %b
+  %idxprom = sext i32 %dec to i64
+  %arrayidx = getelementptr inbounds i32, i32* %c, i64 %idxprom
+  %0 = load i32, i32* %arrayidx, align 4
+  %cmp = icmp eq i32 %or, %0
+  br i1 %cmp, label %if.end, label %if.then
+
+if.then:                                          ; preds = %for.body
+  store i32 %or, i32* %arrayidx, align 4
+  call void @foo()
+  br label %if.end
+
+if.end:                                           ; preds = %for.body, %if.then
+  %changed.1.off0 = phi i1 [ true, %if.then ], [ %changed.0.off0, %for.body ]
+  br label %for.cond
+}
+
+declare void @foo()
diff --git a/test/Transforms/SimplifyCFG/preserve-llvm-loop-metadata.ll b/test/Transforms/SimplifyCFG/preserve-llvm-loop-metadata.ll
index 12a908b20f9e..e357104f2a66 100644
--- a/test/Transforms/SimplifyCFG/preserve-llvm-loop-metadata.ll
+++ b/test/Transforms/SimplifyCFG/preserve-llvm-loop-metadata.ll
@@ -1,4 +1,4 @@
-; RUN: opt -simplifycfg -S < %s | FileCheck %s
+; RUN: opt -latesimplifycfg -S < %s | FileCheck %s
 
 define void @test1(i32 %n) #0 {
 entry:
diff --git a/utils/release/test-release.sh b/utils/release/test-release.sh
index 8ec3abb17551..02d8e7925f6e 100755
--- a/utils/release/test-release.sh
+++ b/utils/release/test-release.sh
@@ -562,7 +562,7 @@ for Flavor in $Flavors ; do
             # case there are build paths in the debug info. On some systems,
             # sed adds a newline to the output, so pass $p3 through sed too.
             if ! cmp -s \
-                <(env LC_CTYPE=C sed -e 's,Phase2,Phase3,g' $p2) \
+                <(env LC_CTYPE=C sed -e 's,Phase2,Phase3,g' -e 's,Phase1,Phase2,g' $p2) \
                 <(env LC_CTYPE=C sed -e '' $p3) 16 16; then
                 echo "file `basename $p2` differs between phase 2 and phase 3"
             fi