Vendor import of clang trunk r300422:vendor/clang/clang-trunk-r300422

https://llvm.org/svn/llvm-project/cfe/trunk@300422

392 files changed, 51438 insertions, 31344 deletions

diff --git a/lib/AST/ASTContext.cpp b/lib/AST/ASTContext.cpp
index b531a66dbab3..7b337b061a03 100644
--- a/lib/AST/ASTContext.cpp
+++ b/lib/AST/ASTContext.cpp
@@ -703,6 +703,7 @@ static const LangAS::Map *getAddressSpaceMap(const TargetInfo &T,
     // The fake address space map must have a distinct entry for each
     // language-specific address space.
     static const unsigned FakeAddrSpaceMap[] = {
+      0, // Default
       1, // opencl_global
       3, // opencl_local
       2, // opencl_constant
@@ -748,6 +749,8 @@ ASTContext::ASTContext(LangOptions &LOpts, SourceManager &SM,
       ExternCContext(nullptr), MakeIntegerSeqDecl(nullptr),
       TypePackElementDecl(nullptr), SourceMgr(SM), LangOpts(LOpts),
       SanitizerBL(new SanitizerBlacklist(LangOpts.SanitizerBlacklistFiles, SM)),
+      XRayFilter(new XRayFunctionFilter(LangOpts.XRayAlwaysInstrumentFiles,
+                                        LangOpts.XRayNeverInstrumentFiles, SM)),
       AddrSpaceMap(nullptr), Target(nullptr), AuxTarget(nullptr),
       PrintingPolicy(LOpts), Idents(idents), Selectors(sels),
       BuiltinInfo(builtins), DeclarationNames(*this), ExternalSource(nullptr),
@@ -1167,7 +1170,6 @@ void ASTContext::InitBuiltinTypes(const TargetInfo &Target,
     InitBuiltinType(OCLEventTy, BuiltinType::OCLEvent);
     InitBuiltinType(OCLClkEventTy, BuiltinType::OCLClkEvent);
     InitBuiltinType(OCLQueueTy, BuiltinType::OCLQueue);
-    InitBuiltinType(OCLNDRangeTy, BuiltinType::OCLNDRange);
     InitBuiltinType(OCLReserveIDTy, BuiltinType::OCLReserveID);
   }
   
@@ -1474,6 +1476,8 @@ CharUnits ASTContext::getDeclAlign(const Decl *D, bool ForAlignof) const {
         }
       }
       Align = std::max(Align, getPreferredTypeAlign(T.getTypePtr()));
+      if (BaseT.getQualifiers().hasUnaligned())
+        Align = Target->getCharWidth();
       if (const VarDecl *VD = dyn_cast<VarDecl>(D)) {
         if (VD->hasGlobalStorage() && !ForAlignof)
           Align = std::max(Align, getTargetInfo().getMinGlobalAlign());
@@ -1775,7 +1779,6 @@ TypeInfo ASTContext::getTypeInfoImpl(const Type *T) const {
     case BuiltinType::OCLEvent:
     case BuiltinType::OCLClkEvent:
     case BuiltinType::OCLQueue:
-    case BuiltinType::OCLNDRange:
     case BuiltinType::OCLReserveID:
       // Currently these types are pointers to opaque types.
       Width = Target->getPointerWidth(0);
@@ -1877,8 +1880,9 @@ TypeInfo ASTContext::getTypeInfoImpl(const Type *T) const {
     return getTypeInfo(cast<SubstTemplateTypeParmType>(T)->
                        getReplacementType().getTypePtr());
 
-  case Type::Auto: {
-    const AutoType *A = cast<AutoType>(T);
+  case Type::Auto:
+  case Type::DeducedTemplateSpecialization: {
+    const DeducedType *A = cast<DeducedType>(T);
     assert(!A->getDeducedType().isNull() &&
            "cannot request the size of an undeduced or dependent auto type");
     return getTypeInfo(A->getDeducedType().getTypePtr());
@@ -2691,8 +2695,7 @@ QualType ASTContext::getConstantArrayType(QualType EltTy,
   // Convert the array size into a canonical width matching the pointer size for
   // the target.
   llvm::APInt ArySize(ArySizeIn);
-  ArySize =
-    ArySize.zextOrTrunc(Target->getPointerWidth(getTargetAddressSpace(EltTy)));
+  ArySize = ArySize.zextOrTrunc(Target->getMaxPointerWidth());
 
   llvm::FoldingSetNodeID ID;
   ConstantArrayType::Profile(ID, EltTy, ArySize, ASM, IndexTypeQuals);
@@ -2765,6 +2768,7 @@ QualType ASTContext::getVariableArrayDecayedType(QualType type) const {
   case Type::TemplateTypeParm:
   case Type::SubstTemplateTypeParmPack:
   case Type::Auto:
+  case Type::DeducedTemplateSpecialization:
   case Type::PackExpansion:
     llvm_unreachable("type should never be variably-modified");
 
@@ -3785,12 +3789,8 @@ QualType ASTContext::getDependentNameType(ElaboratedTypeKeyword Keyword,
                                           QualType Canon) const {
   if (Canon.isNull()) {
     NestedNameSpecifier *CanonNNS = getCanonicalNestedNameSpecifier(NNS);
-    ElaboratedTypeKeyword CanonKeyword = Keyword;
-    if (Keyword == ETK_None)
-      CanonKeyword = ETK_Typename;
-    
-    if (CanonNNS != NNS || CanonKeyword != Keyword)
-      Canon = getDependentNameType(CanonKeyword, CanonNNS, Name);
+    if (CanonNNS != NNS)
+      Canon = getDependentNameType(Keyword, CanonNNS, Name);
   }
 
   llvm::FoldingSetNodeID ID;
@@ -3874,42 +3874,45 @@ ASTContext::getDependentTemplateSpecializationType(
   return QualType(T, 0);
 }
 
+TemplateArgument ASTContext::getInjectedTemplateArg(NamedDecl *Param) {
+  TemplateArgument Arg;
+  if (auto *TTP = dyn_cast<TemplateTypeParmDecl>(Param)) {
+    QualType ArgType = getTypeDeclType(TTP);
+    if (TTP->isParameterPack())
+      ArgType = getPackExpansionType(ArgType, None);
+
+    Arg = TemplateArgument(ArgType);
+  } else if (auto *NTTP = dyn_cast<NonTypeTemplateParmDecl>(Param)) {
+    Expr *E = new (*this) DeclRefExpr(
+        NTTP, /*enclosing*/false,
+        NTTP->getType().getNonLValueExprType(*this),
+        Expr::getValueKindForType(NTTP->getType()), NTTP->getLocation());
+
+    if (NTTP->isParameterPack())
+      E = new (*this) PackExpansionExpr(DependentTy, E, NTTP->getLocation(),
+                                        None);
+    Arg = TemplateArgument(E);
+  } else {
+    auto *TTP = cast<TemplateTemplateParmDecl>(Param);
+    if (TTP->isParameterPack())
+      Arg = TemplateArgument(TemplateName(TTP), Optional<unsigned>());
+    else
+      Arg = TemplateArgument(TemplateName(TTP));
+  }
+
+  if (Param->isTemplateParameterPack())
+    Arg = TemplateArgument::CreatePackCopy(*this, Arg);
+
+  return Arg;
+}
+
 void
 ASTContext::getInjectedTemplateArgs(const TemplateParameterList *Params,
                                     SmallVectorImpl<TemplateArgument> &Args) {
   Args.reserve(Args.size() + Params->size());
 
-  for (NamedDecl *Param : *Params) {
-    TemplateArgument Arg;
-    if (auto *TTP = dyn_cast<TemplateTypeParmDecl>(Param)) {
-      QualType ArgType = getTypeDeclType(TTP);
-      if (TTP->isParameterPack())
-        ArgType = getPackExpansionType(ArgType, None);
-
-      Arg = TemplateArgument(ArgType);
-    } else if (auto *NTTP = dyn_cast<NonTypeTemplateParmDecl>(Param)) {
-      Expr *E = new (*this) DeclRefExpr(
-          NTTP, /*enclosing*/false,
-          NTTP->getType().getNonLValueExprType(*this),
-          Expr::getValueKindForType(NTTP->getType()), NTTP->getLocation());
-
-      if (NTTP->isParameterPack())
-        E = new (*this) PackExpansionExpr(DependentTy, E, NTTP->getLocation(),
-                                          None);
-      Arg = TemplateArgument(E);
-    } else {
-      auto *TTP = cast<TemplateTemplateParmDecl>(Param);
-      if (TTP->isParameterPack())
-        Arg = TemplateArgument(TemplateName(TTP), Optional<unsigned>());
-      else
-        Arg = TemplateArgument(TemplateName(TTP));
-    }
-
-    if (Param->isTemplateParameterPack())
-      Arg = TemplateArgument::CreatePackCopy(*this, Arg);
-
-    Args.push_back(Arg);
-  }
+  for (NamedDecl *Param : *Params)
+    Args.push_back(getInjectedTemplateArg(Param));
 }
 
 QualType ASTContext::getPackExpansionType(QualType Pattern,
@@ -4439,6 +4442,28 @@ QualType ASTContext::getAutoType(QualType DeducedType, AutoTypeKeyword Keyword,
   return QualType(AT, 0);
 }
 
+/// Return the uniqued reference to the deduced template specialization type
+/// which has been deduced to the given type, or to the canonical undeduced
+/// such type, or the canonical deduced-but-dependent such type.
+QualType ASTContext::getDeducedTemplateSpecializationType(
+    TemplateName Template, QualType DeducedType, bool IsDependent) const {
+  // Look in the folding set for an existing type.
+  void *InsertPos = nullptr;
+  llvm::FoldingSetNodeID ID;
+  DeducedTemplateSpecializationType::Profile(ID, Template, DeducedType,
+                                             IsDependent);
+  if (DeducedTemplateSpecializationType *DTST =
+          DeducedTemplateSpecializationTypes.FindNodeOrInsertPos(ID, InsertPos))
+    return QualType(DTST, 0);
+
+  DeducedTemplateSpecializationType *DTST = new (*this, TypeAlignment)
+      DeducedTemplateSpecializationType(Template, DeducedType, IsDependent);
+  Types.push_back(DTST);
+  if (InsertPos)
+    DeducedTemplateSpecializationTypes.InsertNode(DTST, InsertPos);
+  return QualType(DTST, 0);
+}
+
 /// getAtomicType - Return the uniqued reference to the atomic type for
 /// the given value type.
 QualType ASTContext::getAtomicType(QualType T) const {
@@ -5922,7 +5947,6 @@ static char getObjCEncodingForPrimitiveKind(const ASTContext *C,
     case BuiltinType::OCLEvent:
     case BuiltinType::OCLClkEvent:
     case BuiltinType::OCLQueue:
-    case BuiltinType::OCLNDRange:
     case BuiltinType::OCLReserveID:
     case BuiltinType::OCLSampler:
     case BuiltinType::Dependent:
@@ -6337,6 +6361,7 @@ void ASTContext::getObjCEncodingForTypeImpl(QualType T, std::string& S,
   // We could see an undeduced auto type here during error recovery.
   // Just ignore it.
   case Type::Auto:
+  case Type::DeducedTemplateSpecialization:
     return;
 
   case Type::Pipe:
@@ -8043,20 +8068,12 @@ QualType ASTContext::mergeTypes(QualType LHS, QualType RHS,
   Qualifiers LQuals = LHSCan.getLocalQualifiers();
   Qualifiers RQuals = RHSCan.getLocalQualifiers();
   if (LQuals != RQuals) {
-    if (getLangOpts().OpenCL) {
-      if (LHSCan.getUnqualifiedType() != RHSCan.getUnqualifiedType() ||
-          LQuals.getCVRQualifiers() != RQuals.getCVRQualifiers())
-        return QualType();
-      if (LQuals.isAddressSpaceSupersetOf(RQuals))
-        return LHS;
-      if (RQuals.isAddressSpaceSupersetOf(LQuals))
-        return RHS;
-    }
     // If any of these qualifiers are different, we have a type
     // mismatch.
     if (LQuals.getCVRQualifiers() != RQuals.getCVRQualifiers() ||
         LQuals.getAddressSpace() != RQuals.getAddressSpace() ||
-        LQuals.getObjCLifetime() != RQuals.getObjCLifetime())
+        LQuals.getObjCLifetime() != RQuals.getObjCLifetime() ||
+        LQuals.hasUnaligned() != RQuals.hasUnaligned())
       return QualType();
 
     // Exactly one GC qualifier difference is allowed: __strong is
@@ -8136,6 +8153,7 @@ QualType ASTContext::mergeTypes(QualType LHS, QualType RHS,
     llvm_unreachable("Non-canonical and dependent types shouldn't get here");
 
   case Type::Auto:
+  case Type::DeducedTemplateSpecialization:
   case Type::LValueReference:
   case Type::RValueReference:
   case Type::MemberPointer:
@@ -8175,6 +8193,20 @@ QualType ASTContext::mergeTypes(QualType LHS, QualType RHS,
       LHSPointee = LHSPointee.getUnqualifiedType();
       RHSPointee = RHSPointee.getUnqualifiedType();
     }
+    if (getLangOpts().OpenCL) {
+      Qualifiers LHSPteeQual = LHSPointee.getQualifiers();
+      Qualifiers RHSPteeQual = RHSPointee.getQualifiers();
+      // Blocks can't be an expression in a ternary operator (OpenCL v2.0
+      // 6.12.5) thus the following check is asymmetric.
+      if (!LHSPteeQual.isAddressSpaceSupersetOf(RHSPteeQual))
+        return QualType();
+      LHSPteeQual.removeAddressSpace();
+      RHSPteeQual.removeAddressSpace();
+      LHSPointee =
+          QualType(LHSPointee.getTypePtr(), LHSPteeQual.getAsOpaqueValue());
+      RHSPointee =
+          QualType(RHSPointee.getTypePtr(), RHSPteeQual.getAsOpaqueValue());
+    }
     QualType ResultType = mergeTypes(LHSPointee, RHSPointee, OfBlockPointer,
                                      Unqualified);
     if (ResultType.isNull()) return QualType();
@@ -8696,7 +8728,8 @@ static QualType DecodeTypeFromStr(const char *&Str, const ASTContext &Context,
       char *End;
       unsigned AddrSpace = strtoul(Str, &End, 10);
       if (End != Str && AddrSpace != 0) {
-        Type = Context.getAddrSpaceQualType(Type, AddrSpace);
+        Type = Context.getAddrSpaceQualType(Type, AddrSpace +
+            LangAS::Count);
         Str = End;
       }
       if (c == '*')
@@ -8788,7 +8821,7 @@ static GVALinkage basicGVALinkageForFunction(const ASTContext &Context,
   if (!FD->isExternallyVisible())
     return GVA_Internal;
 
-  GVALinkage External = GVA_StrongExternal;
+  GVALinkage External;
   switch (FD->getTemplateSpecializationKind()) {
   case TSK_Undeclared:
   case TSK_ExplicitSpecialization:
@@ -8860,8 +8893,22 @@ static GVALinkage adjustGVALinkageForAttributes(const ASTContext &Context,
 }
 
 GVALinkage ASTContext::GetGVALinkageForFunction(const FunctionDecl *FD) const {
-  return adjustGVALinkageForAttributes(
+  auto L = adjustGVALinkageForAttributes(
       *this, basicGVALinkageForFunction(*this, FD), FD);
+  auto EK = ExternalASTSource::EK_ReplyHazy;
+  if (auto *Ext = getExternalSource())
+    EK = Ext->hasExternalDefinitions(FD);
+  switch (EK) {
+  case ExternalASTSource::EK_Never:
+    if (L == GVA_DiscardableODR)
+      return GVA_StrongODR;
+    break;
+  case ExternalASTSource::EK_Always:
+    return GVA_AvailableExternally;
+  case ExternalASTSource::EK_ReplyHazy:
+    break;
+  }
+  return L;
 }
 
 static GVALinkage basicGVALinkageForVariable(const ASTContext &Context,
@@ -8870,22 +8917,30 @@ static GVALinkage basicGVALinkageForVariable(const ASTContext &Context,
     return GVA_Internal;
 
   if (VD->isStaticLocal()) {
-    GVALinkage StaticLocalLinkage = GVA_DiscardableODR;
     const DeclContext *LexicalContext = VD->getParentFunctionOrMethod();
     while (LexicalContext && !isa<FunctionDecl>(LexicalContext))
       LexicalContext = LexicalContext->getLexicalParent();
 
-    // Let the static local variable inherit its linkage from the nearest
-    // enclosing function.
-    if (LexicalContext)
-      StaticLocalLinkage =
-          Context.GetGVALinkageForFunction(cast<FunctionDecl>(LexicalContext));
+    // ObjC Blocks can create local variables that don't have a FunctionDecl
+    // LexicalContext.
+    if (!LexicalContext)
+      return GVA_DiscardableODR;
 
-    // GVA_StrongODR function linkage is stronger than what we need,
-    // downgrade to GVA_DiscardableODR.
-    // This allows us to discard the variable if we never end up needing it.
-    return StaticLocalLinkage == GVA_StrongODR ? GVA_DiscardableODR
-                                               : StaticLocalLinkage;
+    // Otherwise, let the static local variable inherit its linkage from the
+    // nearest enclosing function.
+    auto StaticLocalLinkage =
+        Context.GetGVALinkageForFunction(cast<FunctionDecl>(LexicalContext));
+
+    // Itanium ABI 5.2.2: "Each COMDAT group [for a static local variable] must
+    // be emitted in any object with references to the symbol for the object it
+    // contains, whether inline or out-of-line."
+    // Similar behavior is observed with MSVC. An alternative ABI could use
+    // StrongODR/AvailableExternally to match the function, but none are
+    // known/supported currently.
+    if (StaticLocalLinkage == GVA_StrongODR ||
+        StaticLocalLinkage == GVA_AvailableExternally)
+      return GVA_DiscardableODR;
+    return StaticLocalLinkage;
   }
 
   // MSVC treats in-class initialized static data members as definitions.
@@ -9002,10 +9057,12 @@ bool ASTContext::DeclMustBeEmitted(const Decl *D) {
       }
     }
 
+    GVALinkage Linkage = GetGVALinkageForFunction(FD);
+
     // static, static inline, always_inline, and extern inline functions can
     // always be deferred.  Normal inline functions can be deferred in C99/C++.
     // Implicit template instantiations can also be deferred in C++.
-    return !isDiscardableGVALinkage(GetGVALinkageForFunction(FD));
+    return !isDiscardableGVALinkage(Linkage);
   }
   
   const VarDecl *VD = cast<VarDecl>(D);
@@ -9488,6 +9545,18 @@ uint64_t ASTContext::getTargetNullPointerValue(QualType QT) const {
   return getTargetInfo().getNullPointerValue(AS);
 }
 
+unsigned ASTContext::getTargetAddressSpace(unsigned AS) const {
+  // For OpenCL, only function local variables are not explicitly marked with
+  // an address space in the AST, and these need to be the address space of
+  // alloca.
+  if (!AS && LangOpts.OpenCL)
+    return getTargetInfo().getDataLayout().getAllocaAddrSpace();
+  if (AS >= LangAS::Count)
+    return AS - LangAS::Count;
+  else
+    return (*AddrSpaceMap)[AS];
+}
+
 // Explicitly instantiate this in case a Redeclarable<T> is used from a TU that
 // doesn't include ASTContext.h
 template
diff --git a/lib/AST/ASTDumper.cpp b/lib/AST/ASTDumper.cpp
index 62261ccc905b..ef491ab06f8c 100644
--- a/lib/AST/ASTDumper.cpp
+++ b/lib/AST/ASTDumper.cpp
@@ -102,22 +102,26 @@ namespace  {
     /// Pending[i] is an action to dump an entity at level i.
     llvm::SmallVector<std::function<void(bool isLastChild)>, 32> Pending;
 
+    /// Indicates whether we should trigger deserialization of nodes that had
+    /// not already been loaded.
+    bool Deserialize = false;
+
     /// Indicates whether we're at the top level.
-    bool TopLevel;
+    bool TopLevel = true;
 
     /// Indicates if we're handling the first child after entering a new depth.
-    bool FirstChild;
+    bool FirstChild = true;
 
     /// Prefix for currently-being-dumped entity.
     std::string Prefix;
 
     /// Keep track of the last location we print out so that we can
     /// print out deltas from then on out.
-    const char *LastLocFilename;
-    unsigned LastLocLine;
+    const char *LastLocFilename = "";
+    unsigned LastLocLine = ~0U;
 
     /// The \c FullComment parent of the comment being dumped.
-    const FullComment *FC;
+    const FullComment *FC = nullptr;
 
     bool ShowColors;
 
@@ -203,15 +207,14 @@ namespace  {
   public:
     ASTDumper(raw_ostream &OS, const CommandTraits *Traits,
               const SourceManager *SM)
-      : OS(OS), Traits(Traits), SM(SM), TopLevel(true), FirstChild(true),
-        LastLocFilename(""), LastLocLine(~0U), FC(nullptr),
+      : OS(OS), Traits(Traits), SM(SM),
         ShowColors(SM && SM->getDiagnostics().getShowColors()) { }
 
     ASTDumper(raw_ostream &OS, const CommandTraits *Traits,
               const SourceManager *SM, bool ShowColors)
-      : OS(OS), Traits(Traits), SM(SM), TopLevel(true), FirstChild(true),
-        LastLocFilename(""), LastLocLine(~0U),
-        ShowColors(ShowColors) { }
+      : OS(OS), Traits(Traits), SM(SM), ShowColors(ShowColors) {}
+
+    void setDeserialize(bool D) { Deserialize = D; }
 
     void dumpDecl(const Decl *D);
     void dumpStmt(const Stmt *S);
@@ -764,14 +767,15 @@ bool ASTDumper::hasNodes(const DeclContext *DC) {
     return false;
 
   return DC->hasExternalLexicalStorage() ||
-         DC->noload_decls_begin() != DC->noload_decls_end();
+         (Deserialize ? DC->decls_begin() != DC->decls_end()
+                      : DC->noload_decls_begin() != DC->noload_decls_end());
 }
 
 void ASTDumper::dumpDeclContext(const DeclContext *DC) {
   if (!DC)
     return;
 
-  for (auto *D : DC->noload_decls())
+  for (auto *D : (Deserialize ? DC->decls() : DC->noload_decls()))
     dumpDecl(D);
 
   if (DC->hasExternalLexicalStorage()) {
@@ -795,11 +799,13 @@ void ASTDumper::dumpLookups(const DeclContext *DC, bool DumpDecls) {
 
     bool HasUndeserializedLookups = Primary->hasExternalVisibleStorage();
 
-    DeclContext::all_lookups_iterator I = Primary->noload_lookups_begin(),
-                                      E = Primary->noload_lookups_end();
-    while (I != E) {
+    for (auto I = Deserialize ? Primary->lookups_begin()
+                              : Primary->noload_lookups_begin(),
+              E = Deserialize ? Primary->lookups_end()
+                              : Primary->noload_lookups_end();
+         I != E; ++I) {
       DeclarationName Name = I.getLookupName();
-      DeclContextLookupResult R = *I++;
+      DeclContextLookupResult R = *I;
 
       dumpChild([=] {
         OS << "DeclarationName ";
@@ -1463,6 +1469,7 @@ void ASTDumper::VisitTemplateTypeParmDecl(const TemplateTypeParmDecl *D) {
     OS << " typename";
   else
     OS << " class";
+  OS << " depth " << D->getDepth() << " index " << D->getIndex();
   if (D->isParameterPack())
     OS << " ...";
   dumpName(D);
@@ -1472,6 +1479,7 @@ void ASTDumper::VisitTemplateTypeParmDecl(const TemplateTypeParmDecl *D) {
 
 void ASTDumper::VisitNonTypeTemplateParmDecl(const NonTypeTemplateParmDecl *D) {
   dumpType(D->getType());
+  OS << " depth " << D->getDepth() << " index " << D->getIndex();
   if (D->isParameterPack())
     OS << " ...";
   dumpName(D);
@@ -1481,6 +1489,7 @@ void ASTDumper::VisitNonTypeTemplateParmDecl(const NonTypeTemplateParmDecl *D) {
 
 void ASTDumper::VisitTemplateTemplateParmDecl(
     const TemplateTemplateParmDecl *D) {
+  OS << " depth " << D->getDepth() << " index " << D->getIndex();
   if (D->isParameterPack())
     OS << " ...";
   dumpName(D);
@@ -2504,9 +2513,10 @@ LLVM_DUMP_METHOD void Type::dump(llvm::raw_ostream &OS) const {
 
 LLVM_DUMP_METHOD void Decl::dump() const { dump(llvm::errs()); }
 
-LLVM_DUMP_METHOD void Decl::dump(raw_ostream &OS) const {
+LLVM_DUMP_METHOD void Decl::dump(raw_ostream &OS, bool Deserialize) const {
   ASTDumper P(OS, &getASTContext().getCommentCommandTraits(),
               &getASTContext().getSourceManager());
+  P.setDeserialize(Deserialize);
   P.dumpDecl(this);
 }
 
@@ -2521,12 +2531,14 @@ LLVM_DUMP_METHOD void DeclContext::dumpLookups() const {
 }
 
 LLVM_DUMP_METHOD void DeclContext::dumpLookups(raw_ostream &OS,
-                                               bool DumpDecls) const {
+                                               bool DumpDecls,
+                                               bool Deserialize) const {
   const DeclContext *DC = this;
   while (!DC->isTranslationUnit())
     DC = DC->getParent();
   ASTContext &Ctx = cast<TranslationUnitDecl>(DC)->getASTContext();
   ASTDumper P(OS, &Ctx.getCommentCommandTraits(), &Ctx.getSourceManager());
+  P.setDeserialize(Deserialize);
   P.dumpLookups(this, DumpDecls);
 }
 
diff --git a/lib/AST/ASTImporter.cpp b/lib/AST/ASTImporter.cpp
index 1ccb746633a7..95492825eb9d 100644
--- a/lib/AST/ASTImporter.cpp
+++ b/lib/AST/ASTImporter.cpp
@@ -72,7 +72,7 @@ namespace clang {
     QualType VisitEnumType(const EnumType *T);
     QualType VisitAttributedType(const AttributedType *T);
     QualType VisitTemplateTypeParmType(const TemplateTypeParmType *T);
-    // FIXME: SubstTemplateTypeParmType
+    QualType VisitSubstTemplateTypeParmType(const SubstTemplateTypeParmType *T);
     QualType VisitTemplateSpecializationType(const TemplateSpecializationType *T);
     QualType VisitElaboratedType(const ElaboratedType *T);
     // FIXME: DependentNameType
@@ -278,6 +278,8 @@ namespace clang {
     Expr *VisitArrayInitIndexExpr(ArrayInitIndexExpr *E);
     Expr *VisitCXXDefaultInitExpr(CXXDefaultInitExpr *E);
     Expr *VisitCXXNamedCastExpr(CXXNamedCastExpr *E);
+    Expr *VisitSubstNonTypeTemplateParmExpr(SubstNonTypeTemplateParmExpr *E);
+
 
     template<typename IIter, typename OIter>
     void ImportArray(IIter Ibegin, IIter Iend, OIter Obegin) {
@@ -397,6 +399,9 @@ static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
                                      QualType T1, QualType T2);
 static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
                                      Decl *D1, Decl *D2);
+static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
+                                     const TemplateArgument &Arg1,
+                                     const TemplateArgument &Arg2);
 
 /// \brief Determine structural equivalence of two expressions.
 static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
@@ -421,8 +426,103 @@ static bool IsStructurallyEquivalent(const IdentifierInfo *Name1,
 static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
                                      NestedNameSpecifier *NNS1,
                                      NestedNameSpecifier *NNS2) {
-  // FIXME: Implement!
-  return true;
+  if (NNS1->getKind() != NNS2->getKind())
+    return false;
+
+  NestedNameSpecifier *Prefix1 = NNS1->getPrefix(),
+      *Prefix2 = NNS2->getPrefix();
+  if ((bool)Prefix1 != (bool)Prefix2)
+    return false;
+
+  if (Prefix1)
+    if (!IsStructurallyEquivalent(Context, Prefix1, Prefix2))
+      return false;
+
+  switch (NNS1->getKind()) {
+  case NestedNameSpecifier::Identifier:
+    return IsStructurallyEquivalent(NNS1->getAsIdentifier(),
+                                    NNS2->getAsIdentifier());
+  case NestedNameSpecifier::Namespace:
+    return IsStructurallyEquivalent(Context, NNS1->getAsNamespace(),
+                                    NNS2->getAsNamespace());
+  case NestedNameSpecifier::NamespaceAlias:
+    return IsStructurallyEquivalent(Context, NNS1->getAsNamespaceAlias(),
+                                    NNS2->getAsNamespaceAlias());
+  case NestedNameSpecifier::TypeSpec:
+  case NestedNameSpecifier::TypeSpecWithTemplate:
+    return IsStructurallyEquivalent(Context, QualType(NNS1->getAsType(), 0),
+                                    QualType(NNS2->getAsType(), 0));
+  case NestedNameSpecifier::Global:
+    return true;
+  case NestedNameSpecifier::Super:
+    return IsStructurallyEquivalent(Context, NNS1->getAsRecordDecl(),
+                                    NNS2->getAsRecordDecl());
+  }
+  return false;
+}
+
+static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
+                                     const TemplateName &N1,
+                                     const TemplateName &N2) {
+  if (N1.getKind() != N2.getKind())
+    return false;
+  switch (N1.getKind()) {
+  case TemplateName::Template:
+    return IsStructurallyEquivalent(Context, N1.getAsTemplateDecl(),
+                                    N2.getAsTemplateDecl());
+
+  case TemplateName::OverloadedTemplate: {
+    OverloadedTemplateStorage *OS1 = N1.getAsOverloadedTemplate(),
+        *OS2 = N2.getAsOverloadedTemplate();
+    OverloadedTemplateStorage::iterator I1 = OS1->begin(), I2 = OS2->begin(),
+        E1 = OS1->end(), E2 = OS2->end();
+    for (; I1 != E1 && I2 != E2; ++I1, ++I2)
+      if (!IsStructurallyEquivalent(Context, *I1, *I2))
+        return false;
+    return I1 == E1 && I2 == E2;
+  }
+
+  case TemplateName::QualifiedTemplate: {
+    QualifiedTemplateName *QN1 = N1.getAsQualifiedTemplateName(),
+        *QN2 = N2.getAsQualifiedTemplateName();
+    return IsStructurallyEquivalent(Context, QN1->getDecl(), QN2->getDecl()) &&
+        IsStructurallyEquivalent(Context, QN1->getQualifier(),
+                                 QN2->getQualifier());
+  }
+
+  case TemplateName::DependentTemplate: {
+      DependentTemplateName *DN1 = N1.getAsDependentTemplateName(),
+          *DN2 = N2.getAsDependentTemplateName();
+      if (!IsStructurallyEquivalent(Context, DN1->getQualifier(),
+                                    DN2->getQualifier()))
+        return false;
+      if (DN1->isIdentifier() && DN2->isIdentifier())
+        return IsStructurallyEquivalent(DN1->getIdentifier(),
+                                        DN2->getIdentifier());
+      else if (DN1->isOverloadedOperator() && DN2->isOverloadedOperator())
+        return DN1->getOperator() == DN2->getOperator();
+      return false;
+  }
+
+  case TemplateName::SubstTemplateTemplateParm: {
+    SubstTemplateTemplateParmStorage *TS1 = N1.getAsSubstTemplateTemplateParm(),
+        *TS2 = N2.getAsSubstTemplateTemplateParm();
+    return IsStructurallyEquivalent(Context, TS1->getParameter(),
+                                    TS2->getParameter()) &&
+        IsStructurallyEquivalent(Context, TS1->getReplacement(),
+                                 TS2->getReplacement());
+    }
+  case TemplateName::SubstTemplateTemplateParmPack: {
+    SubstTemplateTemplateParmPackStorage
+        *P1 = N1.getAsSubstTemplateTemplateParmPack(),
+        *P2 = N2.getAsSubstTemplateTemplateParmPack();
+    return IsStructurallyEquivalent(Context, P1->getArgumentPack(),
+                                    P2->getArgumentPack()) &&
+        IsStructurallyEquivalent(Context, P1->getParameterPack(),
+                                 P2->getParameterPack());
+  }
+  }
+  return false;
 }
 
 /// \brief Determine whether two template arguments are equivalent.
@@ -783,6 +883,20 @@ static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
       return false;
     break;
 
+  case Type::DeducedTemplateSpecialization: {
+    auto *DT1 = cast<DeducedTemplateSpecializationType>(T1);
+    auto *DT2 = cast<DeducedTemplateSpecializationType>(T2);
+    if (!IsStructurallyEquivalent(Context,
+                                  DT1->getTemplateName(),
+                                  DT2->getTemplateName()))
+      return false;
+    if (!IsStructurallyEquivalent(Context,
+                                  DT1->getDeducedType(),
+                                  DT2->getDeducedType()))
+      return false;
+    break;
+  }
+
   case Type::Record:
   case Type::Enum:
     if (!IsStructurallyEquivalent(Context,
@@ -1976,6 +2090,23 @@ QualType ASTNodeImporter::VisitTemplateTypeParmType(
         T->getDepth(), T->getIndex(), T->isParameterPack(), ParmDecl);
 }
 
+QualType ASTNodeImporter::VisitSubstTemplateTypeParmType(
+    const SubstTemplateTypeParmType *T) {
+  const TemplateTypeParmType *Replaced =
+      cast_or_null<TemplateTypeParmType>(Importer.Import(
+        QualType(T->getReplacedParameter(), 0)).getTypePtr());
+  if (!Replaced)
+    return QualType();
+
+  QualType Replacement = Importer.Import(T->getReplacementType());
+  if (Replacement.isNull())
+    return QualType();
+  Replacement = Replacement.getCanonicalType();
+
+  return Importer.getToContext().getSubstTemplateTypeParmType(
+        Replaced, Replacement);
+}
+
 QualType ASTNodeImporter::VisitTemplateSpecializationType(
                                        const TemplateSpecializationType *T) {
   TemplateName ToTemplate = Importer.Import(T->getTemplateName());
@@ -2133,6 +2264,7 @@ ASTNodeImporter::ImportDeclarationNameLoc(const DeclarationNameInfo &From,
   case DeclarationName::ObjCOneArgSelector:
   case DeclarationName::ObjCMultiArgSelector:
   case DeclarationName::CXXUsingDirective:
+  case DeclarationName::CXXDeductionGuideName:
     return;
 
   case DeclarationName::CXXOperatorName: {
@@ -2231,8 +2363,10 @@ bool ASTNodeImporter::ImportDefinition(RecordDecl *From, RecordDecl *To,
     ToData.UserProvidedDefaultConstructor
       = FromData.UserProvidedDefaultConstructor;
     ToData.DeclaredSpecialMembers = FromData.DeclaredSpecialMembers;
-    ToData.ImplicitCopyConstructorHasConstParam
-      = FromData.ImplicitCopyConstructorHasConstParam;
+    ToData.ImplicitCopyConstructorCanHaveConstParamForVBase
+      = FromData.ImplicitCopyConstructorCanHaveConstParamForVBase;
+    ToData.ImplicitCopyConstructorCanHaveConstParamForNonVBase
+      = FromData.ImplicitCopyConstructorCanHaveConstParamForNonVBase;
     ToData.ImplicitCopyAssignmentHasConstParam
       = FromData.ImplicitCopyAssignmentHasConstParam;
     ToData.HasDeclaredCopyConstructorWithConstParam
@@ -2785,7 +2919,7 @@ Decl *ASTNodeImporter::VisitEnumDecl(EnumDecl *D) {
   if (!DC->isFunctionOrMethod() && SearchName) {
     SmallVector<NamedDecl *, 4> ConflictingDecls;
     SmallVector<NamedDecl *, 2> FoundDecls;
-    DC->getRedeclContext()->localUncachedLookup(Name, FoundDecls);
+    DC->getRedeclContext()->localUncachedLookup(SearchName, FoundDecls);
     for (unsigned I = 0, N = FoundDecls.size(); I != N; ++I) {
       if (!FoundDecls[I]->isInIdentifierNamespace(IDNS))
         continue;
@@ -2874,7 +3008,7 @@ Decl *ASTNodeImporter::VisitRecordDecl(RecordDecl *D) {
   if (!DC->isFunctionOrMethod()) {
     SmallVector<NamedDecl *, 4> ConflictingDecls;
     SmallVector<NamedDecl *, 2> FoundDecls;
-    DC->getRedeclContext()->localUncachedLookup(Name, FoundDecls);
+    DC->getRedeclContext()->localUncachedLookup(SearchName, FoundDecls);
     for (unsigned I = 0, N = FoundDecls.size(); I != N; ++I) {
       if (!FoundDecls[I]->isInIdentifierNamespace(IDNS))
         continue;
@@ -3671,6 +3805,9 @@ Decl *ASTNodeImporter::VisitVarDecl(VarDecl *D) {
   if (ImportDefinition(D, ToVar))
     return nullptr;
 
+  if (D->isConstexpr())
+    ToVar->setConstexpr(true);
+
   return ToVar;
 }
 
@@ -3724,8 +3861,27 @@ Decl *ASTNodeImporter::VisitParmVarDecl(ParmVarDecl *D) {
                                      Importer.Import(D->getInnerLocStart()),
                                             Loc, Name.getAsIdentifierInfo(),
                                             T, TInfo, D->getStorageClass(),
-                                            /*FIXME: Default argument*/nullptr);
+                                            /*DefaultArg*/ nullptr);
+
+  // Set the default argument.
   ToParm->setHasInheritedDefaultArg(D->hasInheritedDefaultArg());
+  ToParm->setKNRPromoted(D->isKNRPromoted());
+
+  Expr *ToDefArg = nullptr;
+  Expr *FromDefArg = nullptr;
+  if (D->hasUninstantiatedDefaultArg()) {
+    FromDefArg = D->getUninstantiatedDefaultArg();
+    ToDefArg = Importer.Import(FromDefArg);
+    ToParm->setUninstantiatedDefaultArg(ToDefArg);
+  } else if (D->hasUnparsedDefaultArg()) {
+    ToParm->setUnparsedDefaultArg();
+  } else if (D->hasDefaultArg()) {
+    FromDefArg = D->getDefaultArg();
+    ToDefArg = Importer.Import(FromDefArg);
+    ToParm->setDefaultArg(ToDefArg);
+  }
+  if (FromDefArg && !ToDefArg)
+    return nullptr;
 
   if (D->isUsed())
     ToParm->setIsUsed();
@@ -4424,8 +4580,10 @@ Decl *ASTNodeImporter::VisitObjCPropertyDecl(ObjCPropertyDecl *D) {
   ToProperty->setPropertyAttributes(D->getPropertyAttributes());
   ToProperty->setPropertyAttributesAsWritten(
                                       D->getPropertyAttributesAsWritten());
-  ToProperty->setGetterName(Importer.Import(D->getGetterName()));
-  ToProperty->setSetterName(Importer.Import(D->getSetterName()));
+  ToProperty->setGetterName(Importer.Import(D->getGetterName()),
+                            Importer.Import(D->getGetterNameLoc()));
+  ToProperty->setSetterName(Importer.Import(D->getSetterName()),
+                            Importer.Import(D->getSetterNameLoc()));
   ToProperty->setGetterMethodDecl(
      cast_or_null<ObjCMethodDecl>(Importer.Import(D->getGetterMethodDecl())));
   ToProperty->setSetterMethodDecl(
@@ -4753,12 +4911,46 @@ Decl *ASTNodeImporter::VisitClassTemplateSpecializationDecl(
     }
   } else {
     // Create a new specialization.
-    D2 = ClassTemplateSpecializationDecl::Create(Importer.getToContext(), 
-                                                 D->getTagKind(), DC, 
-                                                 StartLoc, IdLoc,
-                                                 ClassTemplate,
-                                                 TemplateArgs,
-                                                 /*PrevDecl=*/nullptr);
+    if (ClassTemplatePartialSpecializationDecl *PartialSpec =
+        dyn_cast<ClassTemplatePartialSpecializationDecl>(D)) {
+
+      // Import TemplateArgumentListInfo
+      TemplateArgumentListInfo ToTAInfo;
+      auto &ASTTemplateArgs = *PartialSpec->getTemplateArgsAsWritten();
+      for (unsigned I = 0, E = ASTTemplateArgs.NumTemplateArgs; I < E; ++I) {
+        bool Error = false;
+        auto ToLoc = ImportTemplateArgumentLoc(ASTTemplateArgs[I], Error);
+        if (Error)
+          return nullptr;
+        ToTAInfo.addArgument(ToLoc);
+      }
+
+      QualType CanonInjType = Importer.Import(
+            PartialSpec->getInjectedSpecializationType());
+      if (CanonInjType.isNull())
+        return nullptr;
+      CanonInjType = CanonInjType.getCanonicalType();
+
+      TemplateParameterList *ToTPList = ImportTemplateParameterList(
+            PartialSpec->getTemplateParameters());
+      if (!ToTPList && PartialSpec->getTemplateParameters())
+        return nullptr;
+
+      D2 = ClassTemplatePartialSpecializationDecl::Create(
+            Importer.getToContext(), D->getTagKind(), DC, StartLoc, IdLoc,
+            ToTPList, ClassTemplate,
+            llvm::makeArrayRef(TemplateArgs.data(), TemplateArgs.size()),
+            ToTAInfo, CanonInjType, nullptr);
+
+    } else {
+      D2 = ClassTemplateSpecializationDecl::Create(Importer.getToContext(),
+                                                   D->getTagKind(), DC,
+                                                   StartLoc, IdLoc,
+                                                   ClassTemplate,
+                                                   TemplateArgs,
+                                                   /*PrevDecl=*/nullptr);
+    }
+
     D2->setSpecializationKind(D->getSpecializationKind());
 
     // Add this specialization to the class template.
@@ -4766,13 +4958,31 @@ Decl *ASTNodeImporter::VisitClassTemplateSpecializationDecl(
     
     // Import the qualifier, if any.
     D2->setQualifierInfo(Importer.Import(D->getQualifierLoc()));
-    
+
+    Importer.Imported(D, D2);
+
+    if (auto *TSI = D->getTypeAsWritten()) {
+      TypeSourceInfo *TInfo = Importer.Import(TSI);
+      if (!TInfo)
+        return nullptr;
+      D2->setTypeAsWritten(TInfo);
+      D2->setTemplateKeywordLoc(Importer.Import(D->getTemplateKeywordLoc()));
+      D2->setExternLoc(Importer.Import(D->getExternLoc()));
+    }
+
+    SourceLocation POI = Importer.Import(D->getPointOfInstantiation());
+    if (POI.isValid())
+      D2->setPointOfInstantiation(POI);
+    else if (D->getPointOfInstantiation().isValid())
+      return nullptr;
+
+    D2->setTemplateSpecializationKind(D->getTemplateSpecializationKind());
+
     // Add the specialization to this context.
     D2->setLexicalDeclContext(LexicalDC);
     LexicalDC->addDeclInternal(D2);
   }
   Importer.Imported(D, D2);
-  
   if (D->isCompleteDefinition() && ImportDefinition(D, D2))
     return nullptr;
 
@@ -5010,13 +5220,17 @@ Stmt *ASTNodeImporter::VisitGCCAsmStmt(GCCAsmStmt *S) {
   SmallVector<IdentifierInfo *, 4> Names;
   for (unsigned I = 0, E = S->getNumOutputs(); I != E; I++) {
     IdentifierInfo *ToII = Importer.Import(S->getOutputIdentifier(I));
-    if (!ToII)
+    // ToII is nullptr when no symbolic name is given for output operand
+    // see ParseStmtAsm::ParseAsmOperandsOpt
+    if (!ToII && S->getOutputIdentifier(I))
       return nullptr;
     Names.push_back(ToII);
   }
   for (unsigned I = 0, E = S->getNumInputs(); I != E; I++) {
     IdentifierInfo *ToII = Importer.Import(S->getInputIdentifier(I));
-    if (!ToII)
+    // ToII is nullptr when no symbolic name is given for input operand
+    // see ParseStmtAsm::ParseAsmOperandsOpt
+    if (!ToII && S->getInputIdentifier(I))
       return nullptr;
     Names.push_back(ToII);
   }
@@ -5860,7 +6074,7 @@ Expr *ASTNodeImporter::VisitBinaryOperator(BinaryOperator *E) {
                                                       T, E->getValueKind(),
                                                       E->getObjectKind(),
                                            Importer.Import(E->getOperatorLoc()),
-                                                      E->isFPContractable());
+                                                      E->getFPFeatures());
 }
 
 Expr *ASTNodeImporter::VisitConditionalOperator(ConditionalOperator *E) {
@@ -6010,7 +6224,7 @@ Expr *ASTNodeImporter::VisitCompoundAssignOperator(CompoundAssignOperator *E) {
                                                E->getObjectKind(),
                                                CompLHSType, CompResultType,
                                            Importer.Import(E->getOperatorLoc()),
-                                               E->isFPContractable());
+                                               E->getFPFeatures());
 }
 
 bool ASTNodeImporter::ImportCastPath(CastExpr *CE, CXXCastPath &Path) {
@@ -6633,6 +6847,27 @@ Expr *ASTNodeImporter::VisitCXXNamedCastExpr(CXXNamedCastExpr *E) {
   }
 }
 
+
+Expr *ASTNodeImporter::VisitSubstNonTypeTemplateParmExpr(
+    SubstNonTypeTemplateParmExpr *E) {
+  QualType T = Importer.Import(E->getType());
+  if (T.isNull())
+    return nullptr;
+
+  NonTypeTemplateParmDecl *Param = cast_or_null<NonTypeTemplateParmDecl>(
+        Importer.Import(E->getParameter()));
+  if (!Param)
+    return nullptr;
+
+  Expr *Replacement = Importer.Import(E->getReplacement());
+  if (!Replacement)
+    return nullptr;
+
+  return new (Importer.getToContext()) SubstNonTypeTemplateParmExpr(
+        T, E->getValueKind(), Importer.Import(E->getExprLoc()), Param,
+        Replacement);
+}
+
 ASTImporter::ASTImporter(ASTContext &ToContext, FileManager &ToFileManager,
                          ASTContext &FromContext, FileManager &FromFileManager,
                          bool MinimalImport)
@@ -6839,14 +7074,14 @@ NestedNameSpecifier *ASTImporter::Import(NestedNameSpecifier *FromNNS) {
 
   case NestedNameSpecifier::Namespace:
     if (NamespaceDecl *NS = 
-          cast<NamespaceDecl>(Import(FromNNS->getAsNamespace()))) {
+          cast_or_null<NamespaceDecl>(Import(FromNNS->getAsNamespace()))) {
       return NestedNameSpecifier::Create(ToContext, prefix, NS);
     }
     return nullptr;
 
   case NestedNameSpecifier::NamespaceAlias:
     if (NamespaceAliasDecl *NSAD = 
-          cast<NamespaceAliasDecl>(Import(FromNNS->getAsNamespaceAlias()))) {
+          cast_or_null<NamespaceAliasDecl>(Import(FromNNS->getAsNamespaceAlias()))) {
       return NestedNameSpecifier::Create(ToContext, prefix, NSAD);
     }
     return nullptr;
@@ -6856,7 +7091,7 @@ NestedNameSpecifier *ASTImporter::Import(NestedNameSpecifier *FromNNS) {
 
   case NestedNameSpecifier::Super:
     if (CXXRecordDecl *RD =
-            cast<CXXRecordDecl>(Import(FromNNS->getAsRecordDecl()))) {
+            cast_or_null<CXXRecordDecl>(Import(FromNNS->getAsRecordDecl()))) {
       return NestedNameSpecifier::SuperSpecifier(ToContext, RD);
     }
     return nullptr;
@@ -6878,8 +7113,74 @@ NestedNameSpecifier *ASTImporter::Import(NestedNameSpecifier *FromNNS) {
 }
 
 NestedNameSpecifierLoc ASTImporter::Import(NestedNameSpecifierLoc FromNNS) {
-  // FIXME: Implement!
-  return NestedNameSpecifierLoc();
+  // Copied from NestedNameSpecifier mostly.
+  SmallVector<NestedNameSpecifierLoc , 8> NestedNames;
+  NestedNameSpecifierLoc NNS = FromNNS;
+
+  // Push each of the nested-name-specifiers's onto a stack for
+  // serialization in reverse order.
+  while (NNS) {
+    NestedNames.push_back(NNS);
+    NNS = NNS.getPrefix();
+  }
+
+  NestedNameSpecifierLocBuilder Builder;
+
+  while (!NestedNames.empty()) {
+    NNS = NestedNames.pop_back_val();
+    NestedNameSpecifier *Spec = Import(NNS.getNestedNameSpecifier());
+    if (!Spec)
+      return NestedNameSpecifierLoc();
+
+    NestedNameSpecifier::SpecifierKind Kind = Spec->getKind();
+    switch (Kind) {
+    case NestedNameSpecifier::Identifier:
+      Builder.Extend(getToContext(),
+                     Spec->getAsIdentifier(),
+                     Import(NNS.getLocalBeginLoc()),
+                     Import(NNS.getLocalEndLoc()));
+      break;
+
+    case NestedNameSpecifier::Namespace:
+      Builder.Extend(getToContext(),
+                     Spec->getAsNamespace(),
+                     Import(NNS.getLocalBeginLoc()),
+                     Import(NNS.getLocalEndLoc()));
+      break;
+
+    case NestedNameSpecifier::NamespaceAlias:
+      Builder.Extend(getToContext(),
+                     Spec->getAsNamespaceAlias(),
+                     Import(NNS.getLocalBeginLoc()),
+                     Import(NNS.getLocalEndLoc()));
+      break;
+
+    case NestedNameSpecifier::TypeSpec:
+    case NestedNameSpecifier::TypeSpecWithTemplate: {
+      TypeSourceInfo *TSI = getToContext().getTrivialTypeSourceInfo(
+            QualType(Spec->getAsType(), 0));
+      Builder.Extend(getToContext(),
+                     Import(NNS.getLocalBeginLoc()),
+                     TSI->getTypeLoc(),
+                     Import(NNS.getLocalEndLoc()));
+      break;
+    }
+
+    case NestedNameSpecifier::Global:
+      Builder.MakeGlobal(getToContext(), Import(NNS.getLocalBeginLoc()));
+      break;
+
+    case NestedNameSpecifier::Super: {
+      SourceRange ToRange = Import(NNS.getSourceRange());
+      Builder.MakeSuper(getToContext(),
+                        Spec->getAsRecordDecl(),
+                        ToRange.getBegin(),
+                        ToRange.getEnd());
+    }
+  }
+  }
+
+  return Builder.getWithLocInContext(getToContext());
 }
 
 TemplateName ASTImporter::Import(TemplateName From) {
@@ -7175,6 +7476,14 @@ DeclarationName ASTImporter::Import(DeclarationName FromName) {
                                                ToContext.getCanonicalType(T));
   }
 
+  case DeclarationName::CXXDeductionGuideName: {
+    TemplateDecl *Template = cast_or_null<TemplateDecl>(
+        Import(FromName.getCXXDeductionGuideTemplate()));
+    if (!Template)
+      return DeclarationName();
+    return ToContext.DeclarationNames.getCXXDeductionGuideName(Template);
+  }
+
   case DeclarationName::CXXConversionFunctionName: {
     QualType T = Import(FromName.getCXXNameType());
     if (T.isNull())
diff --git a/lib/AST/CMakeLists.txt b/lib/AST/CMakeLists.txt
index e28bd2e16d17..13bf352c2f21 100644
--- a/lib/AST/CMakeLists.txt
+++ b/lib/AST/CMakeLists.txt
@@ -31,6 +31,7 @@ add_clang_library(clangAST
   ExprConstant.cpp
   ExprCXX.cpp
   ExprObjC.cpp
+  ExternalASTMerger.cpp
   ExternalASTSource.cpp
   InheritViz.cpp
   ItaniumCXXABI.cpp
@@ -40,6 +41,7 @@ add_clang_library(clangAST
   MicrosoftMangle.cpp
   NestedNameSpecifier.cpp
   NSAPI.cpp
+  ODRHash.cpp
   OpenMPClause.cpp
   ParentMap.cpp
   RawCommentList.cpp
diff --git a/lib/AST/CXXInheritance.cpp b/lib/AST/CXXInheritance.cpp
index a97d6a22e7b3..56fb0464078f 100644
--- a/lib/AST/CXXInheritance.cpp
+++ b/lib/AST/CXXInheritance.cpp
@@ -88,7 +88,7 @@ bool CXXRecordDecl::isDerivedFrom(const CXXRecordDecl *Base,
   const CXXRecordDecl *BaseDecl = Base->getCanonicalDecl();
   // FIXME: Capturing 'this' is a workaround for name lookup bugs in GCC 4.7.
   return lookupInBases(
-      [this, BaseDecl](const CXXBaseSpecifier *Specifier, CXXBasePath &Path) {
+      [BaseDecl](const CXXBaseSpecifier *Specifier, CXXBasePath &Path) {
         return FindBaseClass(Specifier, Path, BaseDecl);
       },
       Paths);
@@ -109,7 +109,7 @@ bool CXXRecordDecl::isVirtuallyDerivedFrom(const CXXRecordDecl *Base) const {
   const CXXRecordDecl *BaseDecl = Base->getCanonicalDecl();
   // FIXME: Capturing 'this' is a workaround for name lookup bugs in GCC 4.7.
   return lookupInBases(
-      [this, BaseDecl](const CXXBaseSpecifier *Specifier, CXXBasePath &Path) {
+      [BaseDecl](const CXXBaseSpecifier *Specifier, CXXBasePath &Path) {
         return FindVirtualBaseClass(Specifier, Path, BaseDecl);
       },
       Paths);
diff --git a/lib/AST/Decl.cpp b/lib/AST/Decl.cpp
index 81f08787d515..2b22e5bb50a5 100644
--- a/lib/AST/Decl.cpp
+++ b/lib/AST/Decl.cpp
@@ -1414,6 +1414,11 @@ void NamedDecl::printQualifiedName(raw_ostream &OS,
                                    const PrintingPolicy &P) const {
   const DeclContext *Ctx = getDeclContext();
 
+  // For ObjC methods, look through categories and use the interface as context.
+  if (auto *MD = dyn_cast<ObjCMethodDecl>(this))
+    if (auto *ID = MD->getClassInterface())
+      Ctx = ID;
+
   if (Ctx->isFunctionOrMethod()) {
     printName(OS);
     return;
@@ -2143,13 +2148,6 @@ APValue *VarDecl::evaluateValue() const {
   return evaluateValue(Notes);
 }
 
-namespace {
-// Destroy an APValue that was allocated in an ASTContext.
-void DestroyAPValue(void* UntypedValue) {
-  static_cast<APValue*>(UntypedValue)->~APValue();
-}
-} // namespace
-
 APValue *VarDecl::evaluateValue(
     SmallVectorImpl<PartialDiagnosticAt> &Notes) const {
   EvaluatedStmt *Eval = ensureEvaluatedStmt();
@@ -2181,7 +2179,7 @@ APValue *VarDecl::evaluateValue(
   if (!Result)
     Eval->Evaluated = APValue();
   else if (Eval->Evaluated.needsCleanup())
-    getASTContext().AddDeallocation(DestroyAPValue, &Eval->Evaluated);
+    getASTContext().addDestruction(&Eval->Evaluated);
 
   Eval->IsEvaluating = false;
   Eval->WasEvaluated = true;
@@ -2510,7 +2508,7 @@ bool FunctionDecl::isVariadic() const {
 
 bool FunctionDecl::hasBody(const FunctionDecl *&Definition) const {
   for (auto I : redecls()) {
-    if (I->Body || I->IsLateTemplateParsed) {
+    if (I->doesThisDeclarationHaveABody()) {
       Definition = I;
       return true;
     }
@@ -3744,6 +3742,20 @@ void EnumDecl::completeDefinition(QualType NewType,
   TagDecl::completeDefinition();
 }
 
+bool EnumDecl::isClosed() const {
+  if (const auto *A = getAttr<EnumExtensibilityAttr>())
+    return A->getExtensibility() == EnumExtensibilityAttr::Closed;
+  return true;
+}
+
+bool EnumDecl::isClosedFlag() const {
+  return isClosed() && hasAttr<FlagEnumAttr>();
+}
+
+bool EnumDecl::isClosedNonFlag() const {
+  return isClosed() && !hasAttr<FlagEnumAttr>();
+}
+
 TemplateSpecializationKind EnumDecl::getTemplateSpecializationKind() const {
   if (MemberSpecializationInfo *MSI = getMemberSpecializationInfo())
     return MSI->getTemplateSpecializationKind();
@@ -4230,6 +4242,30 @@ TagDecl *TypedefNameDecl::getAnonDeclWithTypedefName(bool AnyRedecl) const {
   return nullptr;
 }
 
+bool TypedefNameDecl::isTransparentTagSlow() const {
+  auto determineIsTransparent = [&]() {
+    if (auto *TT = getUnderlyingType()->getAs<TagType>()) {
+      if (auto *TD = TT->getDecl()) {
+        if (TD->getName() != getName())
+          return false;
+        SourceLocation TTLoc = getLocation();
+        SourceLocation TDLoc = TD->getLocation();
+        if (!TTLoc.isMacroID() || !TDLoc.isMacroID())
+          return false;
+        SourceManager &SM = getASTContext().getSourceManager();
+        return SM.getSpellingLoc(TTLoc) == SM.getSpellingLoc(TDLoc);
+      }
+    }
+    return false;
+  };
+
+  bool isTransparent = determineIsTransparent();
+  CacheIsTransparentTag = 1;
+  if (isTransparent)
+    CacheIsTransparentTag |= 0x2;
+  return isTransparent;
+}
+
 TypedefDecl *TypedefDecl::CreateDeserialized(ASTContext &C, unsigned ID) {
   return new (C, ID) TypedefDecl(C, nullptr, SourceLocation(), SourceLocation(),
                                  nullptr, nullptr);
diff --git a/lib/AST/DeclBase.cpp b/lib/AST/DeclBase.cpp
index 6111abab646e..cda70c5edcd4 100644
--- a/lib/AST/DeclBase.cpp
+++ b/lib/AST/DeclBase.cpp
@@ -619,6 +619,7 @@ bool Decl::isWeakImported() const {
 unsigned Decl::getIdentifierNamespaceForKind(Kind DeclKind) {
   switch (DeclKind) {
     case Function:
+    case CXXDeductionGuide:
     case CXXMethod:
     case CXXConstructor:
     case ConstructorUsingShadow:
diff --git a/lib/AST/DeclCXX.cpp b/lib/AST/DeclCXX.cpp
index a9db65a51518..2e5cec9c108f 100644
--- a/lib/AST/DeclCXX.cpp
+++ b/lib/AST/DeclCXX.cpp
@@ -18,6 +18,7 @@
 #include "clang/AST/DeclTemplate.h"
 #include "clang/AST/Expr.h"
 #include "clang/AST/ExprCXX.h"
+#include "clang/AST/ODRHash.h"
 #include "clang/AST/TypeLoc.h"
 #include "clang/Basic/IdentifierTable.h"
 #include "llvm/ADT/STLExtras.h"
@@ -67,12 +68,14 @@ CXXRecordDecl::DefinitionData::DefinitionData(CXXRecordDecl *D)
       HasConstexprDefaultConstructor(false),
       HasNonLiteralTypeFieldsOrBases(false), ComputedVisibleConversions(false),
       UserProvidedDefaultConstructor(false), DeclaredSpecialMembers(0),
-      ImplicitCopyConstructorHasConstParam(true),
+      ImplicitCopyConstructorCanHaveConstParamForVBase(true),
+      ImplicitCopyConstructorCanHaveConstParamForNonVBase(true),
       ImplicitCopyAssignmentHasConstParam(true),
       HasDeclaredCopyConstructorWithConstParam(false),
       HasDeclaredCopyAssignmentWithConstParam(false), IsLambda(false),
-      IsParsingBaseSpecifiers(false), NumBases(0), NumVBases(0), Bases(),
-      VBases(), Definition(D), FirstFriend() {}
+      IsParsingBaseSpecifiers(false), HasODRHash(false), ODRHash(0),
+      NumBases(0), NumVBases(0), Bases(), VBases(), Definition(D),
+      FirstFriend() {}
 
 CXXBaseSpecifier *CXXRecordDecl::DefinitionData::getBasesSlowCase() const {
   return Bases.get(Definition->getASTContext().getExternalSource());
@@ -226,7 +229,7 @@ CXXRecordDecl::setBases(CXXBaseSpecifier const * const *Bases,
         //   'const B&' or 'const volatile B&' [...]
         if (CXXRecordDecl *VBaseDecl = VBase.getType()->getAsCXXRecordDecl())
           if (!VBaseDecl->hasCopyConstructorWithConstParam())
-            data().ImplicitCopyConstructorHasConstParam = false;
+            data().ImplicitCopyConstructorCanHaveConstParamForVBase = false;
 
         // C++1z [dcl.init.agg]p1:
         //   An aggregate is a class with [...] no virtual base classes
@@ -263,6 +266,14 @@ CXXRecordDecl::setBases(CXXBaseSpecifier const * const *Bases,
       //   In the definition of a constexpr constructor [...]
       //    -- the class shall not have any virtual base classes
       data().DefaultedDefaultConstructorIsConstexpr = false;
+
+      // C++1z [class.copy]p8:
+      //   The implicitly-declared copy constructor for a class X will have
+      //   the form 'X::X(const X&)' if each potentially constructed subobject
+      //   has a copy constructor whose first parameter is of type
+      //   'const B&' or 'const volatile B&' [...]
+      if (!BaseClassDecl->hasCopyConstructorWithConstParam())
+        data().ImplicitCopyConstructorCanHaveConstParamForVBase = false;
     } else {
       // C++ [class.ctor]p5:
       //   A default constructor is trivial [...] if:
@@ -305,6 +316,14 @@ CXXRecordDecl::setBases(CXXBaseSpecifier const * const *Bases,
       //   default constructor is constexpr.
       if (!BaseClassDecl->hasConstexprDefaultConstructor())
         data().DefaultedDefaultConstructorIsConstexpr = false;
+
+      // C++1z [class.copy]p8:
+      //   The implicitly-declared copy constructor for a class X will have
+      //   the form 'X::X(const X&)' if each potentially constructed subobject
+      //   has a copy constructor whose first parameter is of type
+      //   'const B&' or 'const volatile B&' [...]
+      if (!BaseClassDecl->hasCopyConstructorWithConstParam())
+        data().ImplicitCopyConstructorCanHaveConstParamForNonVBase = false;
     }
 
     // C++ [class.ctor]p3:
@@ -324,14 +343,6 @@ CXXRecordDecl::setBases(CXXBaseSpecifier const * const *Bases,
     if (!BaseClassDecl->hasCopyAssignmentWithConstParam())
       data().ImplicitCopyAssignmentHasConstParam = false;
 
-    // C++11 [class.copy]p8:
-    //   The implicitly-declared copy constructor for a class X will have
-    //   the form 'X::X(const X&)' if each direct [...] base class B of X
-    //   has a copy constructor whose first parameter is of type
-    //   'const B&' or 'const volatile B&' [...]
-    if (!BaseClassDecl->hasCopyConstructorWithConstParam())
-      data().ImplicitCopyConstructorHasConstParam = false;
-
     // A class has an Objective-C object member if... or any of its bases
     // has an Objective-C object member.
     if (BaseClassDecl->hasObjectMember())
@@ -371,6 +382,23 @@ CXXRecordDecl::setBases(CXXBaseSpecifier const * const *Bases,
   data().IsParsingBaseSpecifiers = false;
 }
 
+unsigned CXXRecordDecl::getODRHash() const {
+  assert(hasDefinition() && "ODRHash only for records with definitions");
+
+  // Previously calculated hash is stored in DefinitionData.
+  if (DefinitionData->HasODRHash)
+    return DefinitionData->ODRHash;
+
+  // Only calculate hash on first call of getODRHash per record.
+  ODRHash Hash;
+  Hash.AddCXXRecordDecl(getDefinition());
+  DefinitionData->HasODRHash = true;
+  DefinitionData->ODRHash = Hash.CalculateHash();
+
+  return DefinitionData->ODRHash;
+}
+
+
 void CXXRecordDecl::addedClassSubobject(CXXRecordDecl *Subobj) {
   // C++11 [class.copy]p11:
   //   A defaulted copy/move constructor for a class X is defined as
@@ -702,9 +730,7 @@ void CXXRecordDecl::addedMember(Decl *D) {
     ASTContext &Context = getASTContext();
     QualType T = Context.getBaseElementType(Field->getType());
     if (T->isObjCRetainableType() || T.isObjCGCStrong()) {
-      if (!Context.getLangOpts().ObjCAutoRefCount) {
-        setHasObjectMember(true);
-      } else if (T.getObjCLifetime() != Qualifiers::OCL_ExplicitNone) {
+      if (T.hasNonTrivialObjCLifetime()) {
         // Objective-C Automatic Reference Counting:
         //   If a class has a non-static data member of Objective-C pointer
         //   type (or array thereof), it is a non-POD type and its
@@ -716,6 +742,8 @@ void CXXRecordDecl::addedMember(Decl *D) {
         Data.PlainOldData = false;
         Data.HasTrivialSpecialMembers = 0;
         Data.HasIrrelevantDestructor = false;
+      } else if (!Context.getLangOpts().ObjCAutoRefCount) {
+        setHasObjectMember(true);
       }
     } else if (!T.isCXX98PODType(Context))
       data().PlainOldData = false;
@@ -905,12 +933,11 @@ void CXXRecordDecl::addedMember(Decl *D) {
 
         // C++11 [class.copy]p8:
         //   The implicitly-declared copy constructor for a class X will have
-        //   the form 'X::X(const X&)' if [...] for all the non-static data
-        //   members of X that are of a class type M (or array thereof), each
-        //   such class type has a copy constructor whose first parameter is
-        //   of type 'const M&' or 'const volatile M&'.
+        //   the form 'X::X(const X&)' if each potentially constructed subobject
+        //   of a class type M (or array thereof) has a copy constructor whose
+        //   first parameter is of type 'const M&' or 'const volatile M&'.
         if (!FieldRec->hasCopyConstructorWithConstParam())
-          data().ImplicitCopyConstructorHasConstParam = false;
+          data().ImplicitCopyConstructorCanHaveConstParamForNonVBase = false;
 
         // C++11 [class.copy]p18:
         //   The implicitly-declared copy assignment oeprator for a class X will
@@ -1472,6 +1499,23 @@ bool CXXRecordDecl::mayBeAbstract() const {
   return false;
 }
 
+void CXXDeductionGuideDecl::anchor() { }
+
+CXXDeductionGuideDecl *CXXDeductionGuideDecl::Create(
+    ASTContext &C, DeclContext *DC, SourceLocation StartLoc, bool IsExplicit,
+    const DeclarationNameInfo &NameInfo, QualType T, TypeSourceInfo *TInfo,
+    SourceLocation EndLocation) {
+  return new (C, DC) CXXDeductionGuideDecl(C, DC, StartLoc, IsExplicit,
+                                           NameInfo, T, TInfo, EndLocation);
+}
+
+CXXDeductionGuideDecl *CXXDeductionGuideDecl::CreateDeserialized(ASTContext &C,
+                                                                 unsigned ID) {
+  return new (C, ID) CXXDeductionGuideDecl(C, nullptr, SourceLocation(), false,
+                                           DeclarationNameInfo(), QualType(),
+                                           nullptr, SourceLocation());
+}
+
 void CXXMethodDecl::anchor() { }
 
 bool CXXMethodDecl::isStatic() const {
diff --git a/lib/AST/DeclObjC.cpp b/lib/AST/DeclObjC.cpp
index 60d05f682e6e..9218eb537a60 100644
--- a/lib/AST/DeclObjC.cpp
+++ b/lib/AST/DeclObjC.cpp
@@ -162,10 +162,10 @@ ObjCPropertyDecl::findPropertyDecl(const DeclContext *DC,
         return nullptr;
   }
 
-  // If context is class, then lookup property in its extensions.
+  // If context is class, then lookup property in its visible extensions.
   // This comes before property is looked up in primary class.
   if (auto *IDecl = dyn_cast<ObjCInterfaceDecl>(DC)) {
-    for (const auto *Ext : IDecl->known_extensions())
+    for (const auto *Ext : IDecl->visible_extensions())
       if (ObjCPropertyDecl *PD = ObjCPropertyDecl::findPropertyDecl(Ext,
                                                        propertyID,
                                                        queryKind))
diff --git a/lib/AST/DeclPrinter.cpp b/lib/AST/DeclPrinter.cpp
index b8ebe1c568c7..5d841a197f26 100644
--- a/lib/AST/DeclPrinter.cpp
+++ b/lib/AST/DeclPrinter.cpp
@@ -464,6 +464,7 @@ void DeclPrinter::VisitRecordDecl(RecordDecl *D) {
 
 void DeclPrinter::VisitEnumConstantDecl(EnumConstantDecl *D) {
   Out << *D;
+  prettyPrintAttributes(D);
   if (Expr *Init = D->getInitExpr()) {
     Out << " = ";
     Init->printPretty(Out, nullptr, Policy, Indentation);
@@ -480,6 +481,7 @@ void DeclPrinter::VisitFunctionDecl(FunctionDecl *D) {
 
   CXXConstructorDecl *CDecl = dyn_cast<CXXConstructorDecl>(D);
   CXXConversionDecl *ConversionDecl = dyn_cast<CXXConversionDecl>(D);
+  CXXDeductionGuideDecl *GuideDecl = dyn_cast<CXXDeductionGuideDecl>(D);
   if (!Policy.SuppressSpecifiers) {
     switch (D->getStorageClass()) {
     case SC_None: break;
@@ -495,13 +497,23 @@ void DeclPrinter::VisitFunctionDecl(FunctionDecl *D) {
     if (D->isModulePrivate())    Out << "__module_private__ ";
     if (D->isConstexpr() && !D->isExplicitlyDefaulted()) Out << "constexpr ";
     if ((CDecl && CDecl->isExplicitSpecified()) ||
-        (ConversionDecl && ConversionDecl->isExplicit()))
+        (ConversionDecl && ConversionDecl->isExplicitSpecified()) ||
+        (GuideDecl && GuideDecl->isExplicitSpecified()))
       Out << "explicit ";
   }
 
   PrintingPolicy SubPolicy(Policy);
   SubPolicy.SuppressSpecifiers = false;
-  std::string Proto = D->getNameInfo().getAsString();
+  std::string Proto;
+  if (!Policy.SuppressScope) {
+    if (const NestedNameSpecifier *NS = D->getQualifier()) {
+      llvm::raw_string_ostream OS(Proto);
+      NS->print(OS, Policy);
+    }
+  }
+  Proto += D->getNameInfo().getAsString();
+  if (GuideDecl)
+    Proto = GuideDecl->getDeducedTemplate()->getDeclName().getAsString();
   if (const TemplateArgumentList *TArgs = D->getTemplateSpecializationArgs()) {
     llvm::raw_string_ostream POut(Proto);
     DeclPrinter TArgPrinter(POut, SubPolicy, Indentation);
@@ -651,7 +663,9 @@ void DeclPrinter::VisitFunctionDecl(FunctionDecl *D) {
       }
     } else if (!ConversionDecl && !isa<CXXDestructorDecl>(D)) {
       if (FT && FT->hasTrailingReturn()) {
-        Out << "auto " << Proto << " -> ";
+        if (!GuideDecl)
+          Out << "auto ";
+        Out << Proto << " -> ";
         Proto.clear();
       }
       AFT->getReturnType().print(Out, Policy, Proto);
@@ -1043,7 +1057,10 @@ void DeclPrinter::VisitFunctionTemplateDecl(FunctionTemplateDecl *D) {
   prettyPrintPragmas(D->getTemplatedDecl());
   VisitRedeclarableTemplateDecl(D);
 
-  if (PrintInstantiation) {
+  // Never print "instantiations" for deduction guides (they don't really
+  // have them).
+  if (PrintInstantiation &&
+      !isa<CXXDeductionGuideDecl>(D->getTemplatedDecl())) {
     FunctionDecl *PrevDecl = D->getTemplatedDecl();
     const FunctionDecl *Def;
     if (PrevDecl->isDefined(Def) && Def != PrevDecl)
diff --git a/lib/AST/DeclTemplate.cpp b/lib/AST/DeclTemplate.cpp
index a5fbb0a3baec..00a6739478bd 100644
--- a/lib/AST/DeclTemplate.cpp
+++ b/lib/AST/DeclTemplate.cpp
@@ -208,10 +208,6 @@ void RedeclarableTemplateDecl::addSpecializationImpl(
 // FunctionTemplateDecl Implementation
 //===----------------------------------------------------------------------===//
 
-void FunctionTemplateDecl::DeallocateCommon(void *Ptr) {
-  static_cast<Common *>(Ptr)->~Common();
-}
-
 FunctionTemplateDecl *FunctionTemplateDecl::Create(ASTContext &C,
                                                    DeclContext *DC,
                                                    SourceLocation L,
@@ -231,7 +227,7 @@ FunctionTemplateDecl *FunctionTemplateDecl::CreateDeserialized(ASTContext &C,
 RedeclarableTemplateDecl::CommonBase *
 FunctionTemplateDecl::newCommon(ASTContext &C) const {
   Common *CommonPtr = new (C) Common;
-  C.AddDeallocation(DeallocateCommon, CommonPtr);
+  C.addDestruction(CommonPtr);
   return CommonPtr;
 }
 
@@ -288,19 +284,23 @@ ArrayRef<TemplateArgument> FunctionTemplateDecl::getInjectedTemplateArgs() {
 // ClassTemplateDecl Implementation
 //===----------------------------------------------------------------------===//
 
-void ClassTemplateDecl::DeallocateCommon(void *Ptr) {
-  static_cast<Common *>(Ptr)->~Common();
-}
-
 ClassTemplateDecl *ClassTemplateDecl::Create(ASTContext &C,
                                              DeclContext *DC,
                                              SourceLocation L,
                                              DeclarationName Name,
                                              TemplateParameterList *Params,
-                                             NamedDecl *Decl) {
+                                             NamedDecl *Decl,
+                                             Expr *AssociatedConstraints) {
   AdoptTemplateParameterList(Params, cast<DeclContext>(Decl));
-  ClassTemplateDecl *New = new (C, DC) ClassTemplateDecl(C, DC, L, Name,
-                                                         Params, Decl);
+
+  if (!AssociatedConstraints) {
+    return new (C, DC) ClassTemplateDecl(C, DC, L, Name, Params, Decl);
+  }
+
+  ConstrainedTemplateDeclInfo *const CTDI = new (C) ConstrainedTemplateDeclInfo;
+  ClassTemplateDecl *const New =
+      new (C, DC) ClassTemplateDecl(CTDI, C, DC, L, Name, Params, Decl);
+  New->setAssociatedConstraints(AssociatedConstraints);
   return New;
 }
 
@@ -340,7 +340,7 @@ ClassTemplateDecl::getPartialSpecializations() {
 RedeclarableTemplateDecl::CommonBase *
 ClassTemplateDecl::newCommon(ASTContext &C) const {
   Common *CommonPtr = new (C) Common;
-  C.AddDeallocation(DeallocateCommon, CommonPtr);
+  C.addDestruction(CommonPtr);
   return CommonPtr;
 }
 
@@ -880,13 +880,10 @@ TypeAliasTemplateDecl *TypeAliasTemplateDecl::CreateDeserialized(ASTContext &C,
                                            DeclarationName(), nullptr, nullptr);
 }
 
-void TypeAliasTemplateDecl::DeallocateCommon(void *Ptr) {
-  static_cast<Common *>(Ptr)->~Common();
-}
 RedeclarableTemplateDecl::CommonBase *
 TypeAliasTemplateDecl::newCommon(ASTContext &C) const {
   Common *CommonPtr = new (C) Common;
-  C.AddDeallocation(DeallocateCommon, CommonPtr);
+  C.addDestruction(CommonPtr);
   return CommonPtr;
 }
 
@@ -907,10 +904,6 @@ ClassScopeFunctionSpecializationDecl::CreateDeserialized(ASTContext &C,
 // VarTemplateDecl Implementation
 //===----------------------------------------------------------------------===//
 
-void VarTemplateDecl::DeallocateCommon(void *Ptr) {
-  static_cast<Common *>(Ptr)->~Common();
-}
-
 VarTemplateDecl *VarTemplateDecl::getDefinition() {
   VarTemplateDecl *CurD = this;
   while (CurD) {
@@ -966,7 +959,7 @@ VarTemplateDecl::getPartialSpecializations() {
 RedeclarableTemplateDecl::CommonBase *
 VarTemplateDecl::newCommon(ASTContext &C) const {
   Common *CommonPtr = new (C) Common;
-  C.AddDeallocation(DeallocateCommon, CommonPtr);
+  C.addDestruction(CommonPtr);
   return CommonPtr;
 }
 
diff --git a/lib/AST/DeclarationName.cpp b/lib/AST/DeclarationName.cpp
index 52791e51d2dc..1f8e26deda97 100644
--- a/lib/AST/DeclarationName.cpp
+++ b/lib/AST/DeclarationName.cpp
@@ -14,6 +14,7 @@
 #include "clang/AST/DeclarationName.h"
 #include "clang/AST/ASTContext.h"
 #include "clang/AST/DeclCXX.h"
+#include "clang/AST/DeclTemplate.h"
 #include "clang/AST/Type.h"
 #include "clang/AST/TypeLoc.h"
 #include "clang/AST/TypeOrdering.h"
@@ -43,6 +44,22 @@ public:
   }
 };
 
+/// Contains extra information for the name of a C++ deduction guide.
+class CXXDeductionGuideNameExtra : public DeclarationNameExtra,
+                                   public llvm::FoldingSetNode {
+public:
+  /// The template named by the deduction guide.
+  TemplateDecl *Template;
+
+  /// FETokenInfo - Extra information associated with this operator
+  /// name that can be used by the front end.
+  void *FETokenInfo;
+
+  void Profile(llvm::FoldingSetNodeID &ID) {
+    ID.AddPointer(Template);
+  }
+};
+
 /// CXXOperatorIdName - Contains extra information for the name of an
 /// overloaded operator in C++, such as "operator+.
 class CXXOperatorIdName : public DeclarationNameExtra {
@@ -122,7 +139,13 @@ int DeclarationName::compare(DeclarationName LHS, DeclarationName RHS) {
     if (QualTypeOrdering()(RHS.getCXXNameType(), LHS.getCXXNameType()))
       return 1;
     return 0;
-              
+
+  case DeclarationName::CXXDeductionGuideName:
+    // We never want to compare deduction guide names for templates from
+    // different scopes, so just compare the template-name.
+    return compare(LHS.getCXXDeductionGuideTemplate()->getDeclName(),
+                   RHS.getCXXDeductionGuideTemplate()->getDeclName());
+
   case DeclarationName::CXXOperatorName:
     return compareInt(LHS.getCXXOverloadedOperator(),
                       RHS.getCXXOverloadedOperator());
@@ -179,6 +202,12 @@ void DeclarationName::print(raw_ostream &OS, const PrintingPolicy &Policy) {
     return printCXXConstructorDestructorName(N.getCXXNameType(), OS, Policy);
   }
 
+  case DeclarationName::CXXDeductionGuideName:
+    OS << "<deduction guide for ";
+    getCXXDeductionGuideTemplate()->getDeclName().print(OS, Policy);
+    OS << '>';
+    return;
+
   case DeclarationName::CXXOperatorName: {
     static const char* const OperatorNames[NUM_OVERLOADED_OPERATORS] = {
       nullptr,
@@ -243,6 +272,9 @@ DeclarationName::NameKind DeclarationName::getNameKind() const {
     case DeclarationNameExtra::CXXDestructor:
       return CXXDestructorName;
 
+    case DeclarationNameExtra::CXXDeductionGuide:
+      return CXXDeductionGuideName;
+
     case DeclarationNameExtra::CXXConversionFunction:
       return CXXConversionFunctionName;
 
@@ -268,7 +300,15 @@ DeclarationName::NameKind DeclarationName::getNameKind() const {
 
 bool DeclarationName::isDependentName() const {
   QualType T = getCXXNameType();
-  return !T.isNull() && T->isDependentType();
+  if (!T.isNull() && T->isDependentType())
+    return true;
+
+  // A class-scope deduction guide in a dependent context has a dependent name.
+  auto *TD = getCXXDeductionGuideTemplate();
+  if (TD && TD->getDeclContext()->isDependentContext())
+    return true;
+
+  return false;
 }
 
 std::string DeclarationName::getAsString() const {
@@ -285,6 +325,12 @@ QualType DeclarationName::getCXXNameType() const {
     return QualType();
 }
 
+TemplateDecl *DeclarationName::getCXXDeductionGuideTemplate() const {
+  if (auto *Guide = getAsCXXDeductionGuideNameExtra())
+    return Guide->Template;
+  return nullptr;
+}
+
 OverloadedOperatorKind DeclarationName::getCXXOverloadedOperator() const {
   if (CXXOperatorIdName *CXXOp = getAsCXXOperatorIdName()) {
     unsigned value
@@ -312,6 +358,9 @@ void *DeclarationName::getFETokenInfoAsVoidSlow() const {
   case CXXConversionFunctionName:
     return getAsCXXSpecialName()->FETokenInfo;
 
+  case CXXDeductionGuideName:
+    return getAsCXXDeductionGuideNameExtra()->FETokenInfo;
+
   case CXXOperatorName:
     return getAsCXXOperatorIdName()->FETokenInfo;
 
@@ -335,6 +384,10 @@ void DeclarationName::setFETokenInfo(void *T) {
     getAsCXXSpecialName()->FETokenInfo = T;
     break;
 
+  case CXXDeductionGuideName:
+    getAsCXXDeductionGuideNameExtra()->FETokenInfo = T;
+    break;
+
   case CXXOperatorName:
     getAsCXXOperatorIdName()->FETokenInfo = T;
     break;
@@ -366,6 +419,7 @@ LLVM_DUMP_METHOD void DeclarationName::dump() const {
 DeclarationNameTable::DeclarationNameTable(const ASTContext &C) : Ctx(C) {
   CXXSpecialNamesImpl = new llvm::FoldingSet<CXXSpecialName>;
   CXXLiteralOperatorNames = new llvm::FoldingSet<CXXLiteralOperatorIdName>;
+  CXXDeductionGuideNames = new llvm::FoldingSet<CXXDeductionGuideNameExtra>;
 
   // Initialize the overloaded operator names.
   CXXOperatorNames = new (Ctx) CXXOperatorIdName[NUM_OVERLOADED_OPERATORS];
@@ -377,14 +431,18 @@ DeclarationNameTable::DeclarationNameTable(const ASTContext &C) : Ctx(C) {
 }
 
 DeclarationNameTable::~DeclarationNameTable() {
-  llvm::FoldingSet<CXXSpecialName> *SpecialNames =
-    static_cast<llvm::FoldingSet<CXXSpecialName>*>(CXXSpecialNamesImpl);
-  llvm::FoldingSet<CXXLiteralOperatorIdName> *LiteralNames
-    = static_cast<llvm::FoldingSet<CXXLiteralOperatorIdName>*>
-        (CXXLiteralOperatorNames);
+  auto *SpecialNames =
+      static_cast<llvm::FoldingSet<CXXSpecialName> *>(CXXSpecialNamesImpl);
+  auto *LiteralNames =
+      static_cast<llvm::FoldingSet<CXXLiteralOperatorIdName> *>(
+          CXXLiteralOperatorNames);
+  auto *DeductionGuideNames =
+      static_cast<llvm::FoldingSet<CXXDeductionGuideNameExtra> *>(
+          CXXDeductionGuideNames);
 
   delete SpecialNames;
   delete LiteralNames;
+  delete DeductionGuideNames;
 }
 
 DeclarationName DeclarationNameTable::getCXXConstructorName(CanQualType Ty) {
@@ -398,6 +456,30 @@ DeclarationName DeclarationNameTable::getCXXDestructorName(CanQualType Ty) {
 }
 
 DeclarationName
+DeclarationNameTable::getCXXDeductionGuideName(TemplateDecl *Template) {
+  Template = cast<TemplateDecl>(Template->getCanonicalDecl());
+
+  auto *DeductionGuideNames =
+      static_cast<llvm::FoldingSet<CXXDeductionGuideNameExtra> *>(
+          CXXDeductionGuideNames);
+
+  llvm::FoldingSetNodeID ID;
+  ID.AddPointer(Template);
+
+  void *InsertPos = nullptr;
+  if (auto *Name = DeductionGuideNames->FindNodeOrInsertPos(ID, InsertPos))
+    return DeclarationName(Name);
+
+  auto *Name = new (Ctx) CXXDeductionGuideNameExtra;
+  Name->ExtraKindOrNumArgs = DeclarationNameExtra::CXXDeductionGuide;
+  Name->Template = Template;
+  Name->FETokenInfo = nullptr;
+
+  DeductionGuideNames->InsertNode(Name, InsertPos);
+  return DeclarationName(Name);
+}
+
+DeclarationName
 DeclarationNameTable::getCXXConversionFunctionName(CanQualType Ty) {
   return getCXXSpecialName(DeclarationName::CXXConversionFunctionName, Ty);
 }
@@ -477,6 +559,7 @@ DeclarationNameTable::getCXXLiteralOperatorName(IdentifierInfo *II) {
 DeclarationNameLoc::DeclarationNameLoc(DeclarationName Name) {
   switch (Name.getNameKind()) {
   case DeclarationName::Identifier:
+  case DeclarationName::CXXDeductionGuideName:
     break;
   case DeclarationName::CXXConstructorName:
   case DeclarationName::CXXDestructorName:
@@ -509,6 +592,7 @@ bool DeclarationNameInfo::containsUnexpandedParameterPack() const {
   case DeclarationName::CXXOperatorName:
   case DeclarationName::CXXLiteralOperatorName:
   case DeclarationName::CXXUsingDirective:
+  case DeclarationName::CXXDeductionGuideName:
     return false;
 
   case DeclarationName::CXXConstructorName:
@@ -531,6 +615,7 @@ bool DeclarationNameInfo::isInstantiationDependent() const {
   case DeclarationName::CXXOperatorName:
   case DeclarationName::CXXLiteralOperatorName:
   case DeclarationName::CXXUsingDirective:
+  case DeclarationName::CXXDeductionGuideName:
     return false;
     
   case DeclarationName::CXXConstructorName:
@@ -560,6 +645,7 @@ void DeclarationNameInfo::printName(raw_ostream &OS) const {
   case DeclarationName::CXXOperatorName:
   case DeclarationName::CXXLiteralOperatorName:
   case DeclarationName::CXXUsingDirective:
+  case DeclarationName::CXXDeductionGuideName:
     OS << Name;
     return;
 
@@ -574,7 +660,9 @@ void DeclarationNameInfo::printName(raw_ostream &OS) const {
       LangOptions LO;
       LO.CPlusPlus = true;
       LO.Bool = true;
-      OS << TInfo->getType().getAsString(PrintingPolicy(LO));
+      PrintingPolicy PP(LO);
+      PP.SuppressScope = true;
+      OS << TInfo->getType().getAsString(PP);
     } else
       OS << Name;
     return;
@@ -585,6 +673,7 @@ void DeclarationNameInfo::printName(raw_ostream &OS) const {
 SourceLocation DeclarationNameInfo::getEndLoc() const {
   switch (Name.getNameKind()) {
   case DeclarationName::Identifier:
+  case DeclarationName::CXXDeductionGuideName:
     return NameLoc;
 
   case DeclarationName::CXXOperatorName: {
diff --git a/lib/AST/Expr.cpp b/lib/AST/Expr.cpp
index 14f31d0c6b8c..d523a0f93cf6 100644
--- a/lib/AST/Expr.cpp
+++ b/lib/AST/Expr.cpp
@@ -987,7 +987,7 @@ void StringLiteral::outputString(raw_ostream &OS) const {
 void StringLiteral::setString(const ASTContext &C, StringRef Str,
                               StringKind Kind, bool IsPascal) {
   //FIXME: we assume that the string data comes from a target that uses the same
-  // code unit size and endianess for the type of string.
+  // code unit size and endianness for the type of string.
   this->Kind = Kind;
   this->IsPascal = IsPascal;
   
@@ -1571,8 +1571,9 @@ bool CastExpr::CastConsistency() const {
     goto CheckNoBasePath;
 
   case CK_AddressSpaceConversion:
-    assert(getType()->isPointerType());
-    assert(getSubExpr()->getType()->isPointerType());
+    assert(getType()->isPointerType() || getType()->isBlockPointerType());
+    assert(getSubExpr()->getType()->isPointerType() ||
+           getSubExpr()->getType()->isBlockPointerType());
     assert(getType()->getPointeeType().getAddressSpace() !=
            getSubExpr()->getType()->getPointeeType().getAddressSpace());
   // These should not have an inheritance path.
@@ -1881,6 +1882,11 @@ bool InitListExpr::isTransparent() const {
   if (getNumInits() != 1 || !getInit(0))
     return false;
 
+  // Don't confuse aggregate initialization of a struct X { X &x; }; with a
+  // transparent struct copy.
+  if (!getInit(0)->isRValue() && getType()->isRecordType())
+    return false;
+
   return getType().getCanonicalType() ==
          getInit(0)->getType().getCanonicalType();
 }
@@ -2952,6 +2958,7 @@ bool Expr::HasSideEffects(const ASTContext &Ctx,
   case CXXNewExprClass:
   case CXXDeleteExprClass:
   case CoawaitExprClass:
+  case DependentCoawaitExprClass:
   case CoyieldExprClass:
     // These always have a side-effect.
     return true;
@@ -3880,16 +3887,22 @@ PseudoObjectExpr::PseudoObjectExpr(QualType type, ExprValueKind VK,
 
 // UnaryExprOrTypeTraitExpr
 Stmt::child_range UnaryExprOrTypeTraitExpr::children() {
+  const_child_range CCR =
+      const_cast<const UnaryExprOrTypeTraitExpr *>(this)->children();
+  return child_range(cast_away_const(CCR.begin()), cast_away_const(CCR.end()));
+}
+
+Stmt::const_child_range UnaryExprOrTypeTraitExpr::children() const {
   // If this is of a type and the type is a VLA type (and not a typedef), the
   // size expression of the VLA needs to be treated as an executable expression.
   // Why isn't this weirdness documented better in StmtIterator?
   if (isArgumentType()) {
-    if (const VariableArrayType* T = dyn_cast<VariableArrayType>(
-                                   getArgumentType().getTypePtr()))
-      return child_range(child_iterator(T), child_iterator());
-    return child_range(child_iterator(), child_iterator());
+    if (const VariableArrayType *T =
+            dyn_cast<VariableArrayType>(getArgumentType().getTypePtr()))
+      return const_child_range(const_child_iterator(T), const_child_iterator());
+    return const_child_range(const_child_iterator(), const_child_iterator());
   }
-  return child_range(&Argument.Ex, &Argument.Ex + 1);
+  return const_child_range(&Argument.Ex, &Argument.Ex + 1);
 }
 
 AtomicExpr::AtomicExpr(SourceLocation BLoc, ArrayRef<Expr*> args,
diff --git a/lib/AST/ExprCXX.cpp b/lib/AST/ExprCXX.cpp
index ad510e0070e6..6713fca04571 100644
--- a/lib/AST/ExprCXX.cpp
+++ b/lib/AST/ExprCXX.cpp
@@ -734,23 +734,23 @@ CXXBindTemporaryExpr *CXXBindTemporaryExpr::Create(const ASTContext &C,
 
 CXXTemporaryObjectExpr::CXXTemporaryObjectExpr(const ASTContext &C,
                                                CXXConstructorDecl *Cons,
-                                               TypeSourceInfo *Type,
+                                               QualType Type,
+                                               TypeSourceInfo *TSI,
                                                ArrayRef<Expr*> Args,
                                                SourceRange ParenOrBraceRange,
                                                bool HadMultipleCandidates,
                                                bool ListInitialization,
                                                bool StdInitListInitialization,
                                                bool ZeroInitialization)
-  : CXXConstructExpr(C, CXXTemporaryObjectExprClass, 
-                     Type->getType().getNonReferenceType(), 
-                     Type->getTypeLoc().getBeginLoc(),
+  : CXXConstructExpr(C, CXXTemporaryObjectExprClass, Type,
+                     TSI->getTypeLoc().getBeginLoc(),
                      Cons, false, Args,
                      HadMultipleCandidates,
                      ListInitialization,
                      StdInitListInitialization,
                      ZeroInitialization,
                      CXXConstructExpr::CK_Complete, ParenOrBraceRange),
-    Type(Type) {
+    Type(TSI) {
 }
 
 SourceLocation CXXTemporaryObjectExpr::getLocStart() const {
diff --git a/lib/AST/ExprClassification.cpp b/lib/AST/ExprClassification.cpp
index adb74b80b198..c035a42439a3 100644
--- a/lib/AST/ExprClassification.cpp
+++ b/lib/AST/ExprClassification.cpp
@@ -129,6 +129,7 @@ static Cl::Kinds ClassifyInternal(ASTContext &Ctx, const Expr *E) {
   case Expr::UnresolvedLookupExprClass:
   case Expr::UnresolvedMemberExprClass:
   case Expr::TypoExprClass:
+  case Expr::DependentCoawaitExprClass:
   case Expr::CXXDependentScopeMemberExprClass:
   case Expr::DependentScopeDeclRefExprClass:
     // ObjC instance variables are lvalues
@@ -626,7 +627,8 @@ static Cl::ModifiableType IsModifiable(ASTContext &Ctx, const Expr *E,
   // Const stuff is obviously not modifiable.
   if (CT.isConstQualified())
     return Cl::CM_ConstQualified;
-  if (CT.getQualifiers().getAddressSpace() == LangAS::opencl_constant)
+  if (Ctx.getLangOpts().OpenCL &&
+      CT.getQualifiers().getAddressSpace() == LangAS::opencl_constant)
     return Cl::CM_ConstAddrSpace;
 
   // Arrays are not modifiable, only their elements are.
diff --git a/lib/AST/ExprConstant.cpp b/lib/AST/ExprConstant.cpp
index 2c0fce91844c..2fafa4876758 100644
--- a/lib/AST/ExprConstant.cpp
+++ b/lib/AST/ExprConstant.cpp
@@ -350,36 +350,49 @@ namespace {
       MostDerivedArraySize = 2;
       MostDerivedPathLength = Entries.size();
     }
-    void diagnosePointerArithmetic(EvalInfo &Info, const Expr *E, uint64_t N);
+    void diagnosePointerArithmetic(EvalInfo &Info, const Expr *E,
+                                   const APSInt &N);
     /// Add N to the address of this subobject.
-    void adjustIndex(EvalInfo &Info, const Expr *E, uint64_t N) {
-      if (Invalid) return;
+    void adjustIndex(EvalInfo &Info, const Expr *E, APSInt N) {
+      if (Invalid || !N) return;
+      uint64_t TruncatedN = N.extOrTrunc(64).getZExtValue();
       if (isMostDerivedAnUnsizedArray()) {
         // Can't verify -- trust that the user is doing the right thing (or if
         // not, trust that the caller will catch the bad behavior).
-        Entries.back().ArrayIndex += N;
-        return;
-      }
-      if (MostDerivedPathLength == Entries.size() &&
-          MostDerivedIsArrayElement) {
-        Entries.back().ArrayIndex += N;
-        if (Entries.back().ArrayIndex > getMostDerivedArraySize()) {
-          diagnosePointerArithmetic(Info, E, Entries.back().ArrayIndex);
-          setInvalid();
-        }
+        // FIXME: Should we reject if this overflows, at least?
+        Entries.back().ArrayIndex += TruncatedN;
         return;
       }
+
       // [expr.add]p4: For the purposes of these operators, a pointer to a
       // nonarray object behaves the same as a pointer to the first element of
       // an array of length one with the type of the object as its element type.
-      if (IsOnePastTheEnd && N == (uint64_t)-1)
-        IsOnePastTheEnd = false;
-      else if (!IsOnePastTheEnd && N == 1)
-        IsOnePastTheEnd = true;
-      else if (N != 0) {
-        diagnosePointerArithmetic(Info, E, uint64_t(IsOnePastTheEnd) + N);
+      bool IsArray = MostDerivedPathLength == Entries.size() &&
+                     MostDerivedIsArrayElement;
+      uint64_t ArrayIndex =
+          IsArray ? Entries.back().ArrayIndex : (uint64_t)IsOnePastTheEnd;
+      uint64_t ArraySize =
+          IsArray ? getMostDerivedArraySize() : (uint64_t)1;
+
+      if (N < -(int64_t)ArrayIndex || N > ArraySize - ArrayIndex) {
+        // Calculate the actual index in a wide enough type, so we can include
+        // it in the note.
+        N = N.extend(std::max<unsigned>(N.getBitWidth() + 1, 65));
+        (llvm::APInt&)N += ArrayIndex;
+        assert(N.ugt(ArraySize) && "bounds check failed for in-bounds index");
+        diagnosePointerArithmetic(Info, E, N);
         setInvalid();
+        return;
       }
+
+      ArrayIndex += TruncatedN;
+      assert(ArrayIndex <= ArraySize &&
+             "bounds check succeeded for out-of-bounds index");
+
+      if (IsArray)
+        Entries.back().ArrayIndex = ArrayIndex;
+      else
+        IsOnePastTheEnd = (ArrayIndex != 0);
     }
   };
 
@@ -413,6 +426,17 @@ namespace {
     /// Index - The call index of this call.
     unsigned Index;
 
+    // FIXME: Adding this to every 'CallStackFrame' may have a nontrivial impact
+    // on the overall stack usage of deeply-recursing constexpr evaluataions.
+    // (We should cache this map rather than recomputing it repeatedly.)
+    // But let's try this and see how it goes; we can look into caching the map
+    // as a later change.
+
+    /// LambdaCaptureFields - Mapping from captured variables/this to
+    /// corresponding data members in the closure class.
+    llvm::DenseMap<const VarDecl *, FieldDecl *> LambdaCaptureFields;
+    FieldDecl *LambdaThisCaptureField;
+
     CallStackFrame(EvalInfo &Info, SourceLocation CallLoc,
                    const FunctionDecl *Callee, const LValue *This,
                    APValue *Arguments);
@@ -1048,16 +1072,17 @@ bool SubobjectDesignator::checkSubobject(EvalInfo &Info, const Expr *E,
 }
 
 void SubobjectDesignator::diagnosePointerArithmetic(EvalInfo &Info,
-                                                    const Expr *E, uint64_t N) {
+                                                    const Expr *E,
+                                                    const APSInt &N) {
   // If we're complaining, we must be able to statically determine the size of
   // the most derived array.
   if (MostDerivedPathLength == Entries.size() && MostDerivedIsArrayElement)
     Info.CCEDiag(E, diag::note_constexpr_array_index)
-      << static_cast<int>(N) << /*array*/ 0
+      << N << /*array*/ 0
       << static_cast<unsigned>(getMostDerivedArraySize());
   else
     Info.CCEDiag(E, diag::note_constexpr_array_index)
-      << static_cast<int>(N) << /*non-array*/ 1;
+      << N << /*non-array*/ 1;
   setInvalid();
 }
 
@@ -1273,14 +1298,24 @@ namespace {
     void clearIsNullPointer() {
       IsNullPtr = false;
     }
-    void adjustOffsetAndIndex(EvalInfo &Info, const Expr *E, uint64_t Index,
-                              CharUnits ElementSize) {
-      // Compute the new offset in the appropriate width.
-      Offset += Index * ElementSize;
-      if (Index && checkNullPointer(Info, E, CSK_ArrayIndex))
+    void adjustOffsetAndIndex(EvalInfo &Info, const Expr *E,
+                              const APSInt &Index, CharUnits ElementSize) {
+      // An index of 0 has no effect. (In C, adding 0 to a null pointer is UB,
+      // but we're not required to diagnose it and it's valid in C++.)
+      if (!Index)
+        return;
+
+      // Compute the new offset in the appropriate width, wrapping at 64 bits.
+      // FIXME: When compiling for a 32-bit target, we should use 32-bit
+      // offsets.
+      uint64_t Offset64 = Offset.getQuantity();
+      uint64_t ElemSize64 = ElementSize.getQuantity();
+      uint64_t Index64 = Index.extOrTrunc(64).getZExtValue();
+      Offset = CharUnits::fromQuantity(Offset64 + ElemSize64 * Index64);
+
+      if (checkNullPointer(Info, E, CSK_ArrayIndex))
         Designator.adjustIndex(Info, E, Index);
-      if (Index)
-        clearIsNullPointer();
+      clearIsNullPointer();
     }
     void adjustOffset(CharUnits N) {
       Offset += N;
@@ -1404,13 +1439,24 @@ static bool EvaluateIntegerOrLValue(const Expr *E, APValue &Result,
                                     EvalInfo &Info);
 static bool EvaluateFloat(const Expr *E, APFloat &Result, EvalInfo &Info);
 static bool EvaluateComplex(const Expr *E, ComplexValue &Res, EvalInfo &Info);
-static bool EvaluateAtomic(const Expr *E, APValue &Result, EvalInfo &Info);
+static bool EvaluateAtomic(const Expr *E, const LValue *This, APValue &Result,
+                           EvalInfo &Info);
 static bool EvaluateAsRValue(EvalInfo &Info, const Expr *E, APValue &Result);
 
 //===----------------------------------------------------------------------===//
 // Misc utilities
 //===----------------------------------------------------------------------===//
 
+/// Negate an APSInt in place, converting it to a signed form if necessary, and
+/// preserving its value (by extending by up to one bit as needed).
+static void negateAsSigned(APSInt &Int) {
+  if (Int.isUnsigned() || Int.isMinSignedValue()) {
+    Int = Int.extend(Int.getBitWidth() + 1);
+    Int.setIsSigned(true);
+  }
+  Int = -Int;
+}
+
 /// Produce a string describing the given constexpr call.
 static void describeCall(CallStackFrame *Frame, raw_ostream &Out) {
   unsigned ArgIndex = 0;
@@ -1458,13 +1504,6 @@ static bool EvaluateIgnoredValue(EvalInfo &Info, const Expr *E) {
   return true;
 }
 
-/// Sign- or zero-extend a value to 64 bits. If it's already 64 bits, just
-/// return its existing value.
-static int64_t getExtValue(const APSInt &Value) {
-  return Value.isSigned() ? Value.getSExtValue()
-                          : static_cast<int64_t>(Value.getZExtValue());
-}
-
 /// Should this call expression be treated as a string literal?
 static bool IsStringLiteralCall(const CallExpr *E) {
   unsigned Builtin = E->getBuiltinCallee();
@@ -2220,7 +2259,7 @@ static bool HandleSizeof(EvalInfo &Info, SourceLocation Loc,
 /// \param Adjustment - The adjustment, in objects of type EltTy, to add.
 static bool HandleLValueArrayAdjustment(EvalInfo &Info, const Expr *E,
                                         LValue &LVal, QualType EltTy,
-                                        int64_t Adjustment) {
+                                        APSInt Adjustment) {
   CharUnits SizeOfPointee;
   if (!HandleSizeof(Info, E->getExprLoc(), EltTy, SizeOfPointee))
     return false;
@@ -2229,6 +2268,13 @@ static bool HandleLValueArrayAdjustment(EvalInfo &Info, const Expr *E,
   return true;
 }
 
+static bool HandleLValueArrayAdjustment(EvalInfo &Info, const Expr *E,
+                                        LValue &LVal, QualType EltTy,
+                                        int64_t Adjustment) {
+  return HandleLValueArrayAdjustment(Info, E, LVal, EltTy,
+                                     APSInt::get(Adjustment));
+}
+
 /// Update an lvalue to refer to a component of a complex number.
 /// \param Info - Information about the ongoing evaluation.
 /// \param LVal - The lvalue to be updated.
@@ -2247,6 +2293,10 @@ static bool HandleLValueComplexElement(EvalInfo &Info, const Expr *E,
   return true;
 }
 
+static bool handleLValueToRValueConversion(EvalInfo &Info, const Expr *Conv,
+                                           QualType Type, const LValue &LVal,
+                                           APValue &RVal);
+
 /// Try to evaluate the initializer for a variable declaration.
 ///
 /// \param Info   Information about the ongoing evaluation.
@@ -2258,6 +2308,7 @@ static bool HandleLValueComplexElement(EvalInfo &Info, const Expr *E,
 static bool evaluateVarDeclInit(EvalInfo &Info, const Expr *E,
                                 const VarDecl *VD, CallStackFrame *Frame,
                                 APValue *&Result) {
+
   // If this is a parameter to an active constexpr function call, perform
   // argument substitution.
   if (const ParmVarDecl *PVD = dyn_cast<ParmVarDecl>(VD)) {
@@ -3191,9 +3242,9 @@ struct CompoundAssignSubobjectHandler {
       return false;
     }
 
-    int64_t Offset = getExtValue(RHS.getInt());
+    APSInt Offset = RHS.getInt();
     if (Opcode == BO_Sub)
-      Offset = -Offset;
+      negateAsSigned(Offset);
 
     LValue LVal;
     LVal.setFrom(Info.Ctx, Subobj);
@@ -4148,6 +4199,10 @@ static bool HandleFunctionCall(SourceLocation CallLoc,
       return false;
     This->moveInto(Result);
     return true;
+  } else if (MD && isLambdaCallOperator(MD)) {
+    // We're in a lambda; determine the lambda capture field maps.
+    MD->getParent()->getCaptureFields(Frame.LambdaCaptureFields,
+                                      Frame.LambdaThisCaptureField);
   }
 
   StmtResult Ret = {Result, ResultSlot};
@@ -4691,7 +4746,10 @@ public:
 
     case CK_AtomicToNonAtomic: {
       APValue AtomicVal;
-      if (!EvaluateAtomic(E->getSubExpr(), AtomicVal, Info))
+      // This does not need to be done in place even for class/array types:
+      // atomic-to-non-atomic conversion implies copying the object
+      // representation.
+      if (!Evaluate(AtomicVal, Info, E->getSubExpr()))
         return false;
       return DerivedSuccess(AtomicVal, E);
     }
@@ -5009,6 +5067,33 @@ bool LValueExprEvaluator::VisitDeclRefExpr(const DeclRefExpr *E) {
 
 
 bool LValueExprEvaluator::VisitVarDecl(const Expr *E, const VarDecl *VD) {
+
+  // If we are within a lambda's call operator, check whether the 'VD' referred
+  // to within 'E' actually represents a lambda-capture that maps to a
+  // data-member/field within the closure object, and if so, evaluate to the
+  // field or what the field refers to.
+  if (Info.CurrentCall && isLambdaCallOperator(Info.CurrentCall->Callee)) {
+    if (auto *FD = Info.CurrentCall->LambdaCaptureFields.lookup(VD)) {
+      if (Info.checkingPotentialConstantExpression())
+        return false;
+      // Start with 'Result' referring to the complete closure object...
+      Result = *Info.CurrentCall->This;
+      // ... then update it to refer to the field of the closure object
+      // that represents the capture.
+      if (!HandleLValueMember(Info, E, Result, FD))
+        return false;
+      // And if the field is of reference type, update 'Result' to refer to what
+      // the field refers to.
+      if (FD->getType()->isReferenceType()) {
+        APValue RVal;
+        if (!handleLValueToRValueConversion(Info, E, FD->getType(), Result,
+                                            RVal))
+          return false;
+        Result.setFrom(Info.Ctx, RVal);
+      }
+      return true;
+    }
+  }
   CallStackFrame *Frame = nullptr;
   if (VD->hasLocalStorage() && Info.CurrentCall->Index > 1) {
     // Only if a local variable was declared in the function currently being
@@ -5162,8 +5247,7 @@ bool LValueExprEvaluator::VisitArraySubscriptExpr(const ArraySubscriptExpr *E) {
   if (!EvaluateInteger(E->getIdx(), Index, Info))
     return false;
 
-  return HandleLValueArrayAdjustment(Info, E, Result, E->getType(),
-                                     getExtValue(Index));
+  return HandleLValueArrayAdjustment(Info, E, Result, E->getType(), Index);
 }
 
 bool LValueExprEvaluator::VisitUnaryDeref(const UnaryOperator *E) {
@@ -5409,6 +5493,27 @@ public:
       return false;
     }
     Result = *Info.CurrentCall->This;
+    // If we are inside a lambda's call operator, the 'this' expression refers
+    // to the enclosing '*this' object (either by value or reference) which is
+    // either copied into the closure object's field that represents the '*this'
+    // or refers to '*this'.
+    if (isLambdaCallOperator(Info.CurrentCall->Callee)) {
+      // Update 'Result' to refer to the data member/field of the closure object
+      // that represents the '*this' capture.
+      if (!HandleLValueMember(Info, E, Result,
+                             Info.CurrentCall->LambdaThisCaptureField)) 
+        return false;
+      // If we captured '*this' by reference, replace the field with its referent.
+      if (Info.CurrentCall->LambdaThisCaptureField->getType()
+              ->isPointerType()) {
+        APValue RVal;
+        if (!handleLValueToRValueConversion(Info, E, E->getType(), Result,
+                                            RVal))
+          return false;
+
+        Result.setFrom(Info.Ctx, RVal);
+      }
+    }
     return true;
   }
 
@@ -5440,13 +5545,11 @@ bool PointerExprEvaluator::VisitBinaryOperator(const BinaryOperator *E) {
   if (!EvaluateInteger(IExp, Offset, Info) || !EvalPtrOK)
     return false;
 
-  int64_t AdditionalOffset = getExtValue(Offset);
   if (E->getOpcode() == BO_Sub)
-    AdditionalOffset = -AdditionalOffset;
+    negateAsSigned(Offset);
 
   QualType Pointee = PExp->getType()->castAs<PointerType>()->getPointeeType();
-  return HandleLValueArrayAdjustment(Info, E, Result, Pointee,
-                                     AdditionalOffset);
+  return HandleLValueArrayAdjustment(Info, E, Result, Pointee, Offset);
 }
 
 bool PointerExprEvaluator::VisitUnaryAddrOf(const UnaryOperator *E) {
@@ -5576,6 +5679,8 @@ static CharUnits GetAlignOfType(EvalInfo &Info, QualType T) {
     T = Ref->getPointeeType();
 
   // __alignof is defined to return the preferred alignment.
+  if (T.getQualifiers().hasUnaligned())
+    return CharUnits::One();
   return Info.Ctx.toCharUnitsFromBits(
     Info.Ctx.getPreferredTypeAlign(T.getTypePtr()));
 }
@@ -5640,14 +5745,14 @@ bool PointerExprEvaluator::VisitBuiltinCallExpr(const CallExpr *E,
     APSInt Alignment;
     if (!EvaluateInteger(E->getArg(1), Alignment, Info))
       return false;
-    CharUnits Align = CharUnits::fromQuantity(getExtValue(Alignment));
+    CharUnits Align = CharUnits::fromQuantity(Alignment.getZExtValue());
 
     if (E->getNumArgs() > 2) {
       APSInt Offset;
       if (!EvaluateInteger(E->getArg(2), Offset, Info))
         return false;
 
-      int64_t AdditionalOffset = -getExtValue(Offset);
+      int64_t AdditionalOffset = -Offset.getZExtValue();
       OffsetResult.Offset += CharUnits::fromQuantity(AdditionalOffset);
     }
 
@@ -5664,12 +5769,11 @@ bool PointerExprEvaluator::VisitBuiltinCallExpr(const CallExpr *E,
 
       if (BaseAlignment < Align) {
         Result.Designator.setInvalid();
-        // FIXME: Quantities here cast to integers because the plural modifier
-        // does not work on APSInts yet.
+        // FIXME: Add support to Diagnostic for long / long long.
         CCEDiag(E->getArg(0),
                 diag::note_constexpr_baa_insufficient_alignment) << 0
-          << (int) BaseAlignment.getQuantity()
-          << (unsigned) getExtValue(Alignment);
+          << (unsigned)BaseAlignment.getQuantity()
+          << (unsigned)Align.getQuantity();
         return false;
       }
     }
@@ -5677,18 +5781,14 @@ bool PointerExprEvaluator::VisitBuiltinCallExpr(const CallExpr *E,
     // The offset must also have the correct alignment.
     if (OffsetResult.Offset.alignTo(Align) != OffsetResult.Offset) {
       Result.Designator.setInvalid();
-      APSInt Offset(64, false);
-      Offset = OffsetResult.Offset.getQuantity();
-
-      if (OffsetResult.Base)
-        CCEDiag(E->getArg(0),
-                diag::note_constexpr_baa_insufficient_alignment) << 1
-          << (int) getExtValue(Offset) << (unsigned) getExtValue(Alignment);
-      else
-        CCEDiag(E->getArg(0),
-                diag::note_constexpr_baa_value_insufficient_alignment)
-          << Offset << (unsigned) getExtValue(Alignment);
 
+      (OffsetResult.Base
+           ? CCEDiag(E->getArg(0),
+                     diag::note_constexpr_baa_insufficient_alignment) << 1
+           : CCEDiag(E->getArg(0),
+                     diag::note_constexpr_baa_value_insufficient_alignment))
+        << (int)OffsetResult.Offset.getQuantity()
+        << (unsigned)Align.getQuantity();
       return false;
     }
 
@@ -6245,14 +6345,40 @@ bool RecordExprEvaluator::VisitLambdaExpr(const LambdaExpr *E) {
   if (ClosureClass->isInvalidDecl()) return false;
 
   if (Info.checkingPotentialConstantExpression()) return true;
-  if (E->capture_size()) {
-    Info.FFDiag(E, diag::note_unimplemented_constexpr_lambda_feature_ast)
-        << "can not evaluate lambda expressions with captures";
-    return false;
+  
+  const size_t NumFields =
+      std::distance(ClosureClass->field_begin(), ClosureClass->field_end());
+
+  assert(NumFields == (size_t)std::distance(E->capture_init_begin(),
+                                            E->capture_init_end()) &&
+         "The number of lambda capture initializers should equal the number of "
+         "fields within the closure type");
+
+  Result = APValue(APValue::UninitStruct(), /*NumBases*/0, NumFields);
+  // Iterate through all the lambda's closure object's fields and initialize
+  // them.
+  auto *CaptureInitIt = E->capture_init_begin();
+  const LambdaCapture *CaptureIt = ClosureClass->captures_begin();
+  bool Success = true;
+  for (const auto *Field : ClosureClass->fields()) {
+    assert(CaptureInitIt != E->capture_init_end());
+    // Get the initializer for this field
+    Expr *const CurFieldInit = *CaptureInitIt++;
+    
+    // If there is no initializer, either this is a VLA or an error has
+    // occurred.
+    if (!CurFieldInit)
+      return Error(E);
+
+    APValue &FieldVal = Result.getStructField(Field->getFieldIndex());
+    if (!EvaluateInPlace(FieldVal, Info, This, CurFieldInit)) {
+      if (!Info.keepEvaluatingAfterFailure())
+        return false;
+      Success = false;
+    }
+    ++CaptureIt;
   }
-  // FIXME: Implement captures.
-  Result = APValue(APValue::UninitStruct(), /*NumBases*/0, /*NumFields*/0);
-  return true;
+  return Success;
 }
 
 static bool EvaluateRecord(const Expr *E, const LValue &This,
@@ -6971,7 +7097,6 @@ static int EvaluateBuiltinClassifyType(const CallExpr *E,
     case BuiltinType::OCLEvent:
     case BuiltinType::OCLClkEvent:
     case BuiltinType::OCLQueue:
-    case BuiltinType::OCLNDRange:
     case BuiltinType::OCLReserveID:
     case BuiltinType::Dependent:
       llvm_unreachable("CallExpr::isBuiltinClassifyType(): unimplemented type");
@@ -7030,6 +7155,7 @@ static int EvaluateBuiltinClassifyType(const CallExpr *E,
   case Type::Vector:
   case Type::ExtVector:
   case Type::Auto:
+  case Type::DeducedTemplateSpecialization:
   case Type::ObjCObject:
   case Type::ObjCInterface:
   case Type::ObjCObjectPointer:
@@ -7948,6 +8074,19 @@ bool DataRecursiveIntBinOpEvaluator::
   return true;
 }
 
+static void addOrSubLValueAsInteger(APValue &LVal, const APSInt &Index,
+                                    bool IsSub) {
+  // Compute the new offset in the appropriate width, wrapping at 64 bits.
+  // FIXME: When compiling for a 32-bit target, we should use 32-bit
+  // offsets.
+  assert(!LVal.hasLValuePath() && "have designator for integer lvalue");
+  CharUnits &Offset = LVal.getLValueOffset();
+  uint64_t Offset64 = Offset.getQuantity();
+  uint64_t Index64 = Index.extOrTrunc(64).getZExtValue();
+  Offset = CharUnits::fromQuantity(IsSub ? Offset64 - Index64
+                                         : Offset64 + Index64);
+}
+
 bool DataRecursiveIntBinOpEvaluator::
        VisitBinOp(const EvalResult &LHSResult, const EvalResult &RHSResult,
                   const BinaryOperator *E, APValue &Result) {
@@ -7994,12 +8133,7 @@ bool DataRecursiveIntBinOpEvaluator::
   // Handle cases like (unsigned long)&a + 4.
   if (E->isAdditiveOp() && LHSVal.isLValue() && RHSVal.isInt()) {
     Result = LHSVal;
-    CharUnits AdditionalOffset =
-        CharUnits::fromQuantity(RHSVal.getInt().getZExtValue());
-    if (E->getOpcode() == BO_Add)
-      Result.getLValueOffset() += AdditionalOffset;
-    else
-      Result.getLValueOffset() -= AdditionalOffset;
+    addOrSubLValueAsInteger(Result, RHSVal.getInt(), E->getOpcode() == BO_Sub);
     return true;
   }
   
@@ -8007,8 +8141,7 @@ bool DataRecursiveIntBinOpEvaluator::
   if (E->getOpcode() == BO_Add &&
       RHSVal.isLValue() && LHSVal.isInt()) {
     Result = RHSVal;
-    Result.getLValueOffset() +=
-        CharUnits::fromQuantity(LHSVal.getInt().getZExtValue());
+    addOrSubLValueAsInteger(Result, LHSVal.getInt(), /*IsSub*/false);
     return true;
   }
   
@@ -9565,10 +9698,11 @@ bool ComplexExprEvaluator::VisitInitListExpr(const InitListExpr *E) {
 namespace {
 class AtomicExprEvaluator :
     public ExprEvaluatorBase<AtomicExprEvaluator> {
+  const LValue *This;
   APValue &Result;
 public:
-  AtomicExprEvaluator(EvalInfo &Info, APValue &Result)
-      : ExprEvaluatorBaseTy(Info), Result(Result) {}
+  AtomicExprEvaluator(EvalInfo &Info, const LValue *This, APValue &Result)
+      : ExprEvaluatorBaseTy(Info), This(This), Result(Result) {}
 
   bool Success(const APValue &V, const Expr *E) {
     Result = V;
@@ -9578,7 +9712,10 @@ public:
   bool ZeroInitialization(const Expr *E) {
     ImplicitValueInitExpr VIE(
         E->getType()->castAs<AtomicType>()->getValueType());
-    return Evaluate(Result, Info, &VIE);
+    // For atomic-qualified class (and array) types in C++, initialize the
+    // _Atomic-wrapped subobject directly, in-place.
+    return This ? EvaluateInPlace(Result, Info, *This, &VIE)
+                : Evaluate(Result, Info, &VIE);
   }
 
   bool VisitCastExpr(const CastExpr *E) {
@@ -9586,15 +9723,17 @@ public:
     default:
       return ExprEvaluatorBaseTy::VisitCastExpr(E);
     case CK_NonAtomicToAtomic:
-      return Evaluate(Result, Info, E->getSubExpr());
+      return This ? EvaluateInPlace(Result, Info, *This, E->getSubExpr())
+                  : Evaluate(Result, Info, E->getSubExpr());
     }
   }
 };
 } // end anonymous namespace
 
-static bool EvaluateAtomic(const Expr *E, APValue &Result, EvalInfo &Info) {
+static bool EvaluateAtomic(const Expr *E, const LValue *This, APValue &Result,
+                           EvalInfo &Info) {
   assert(E->isRValue() && E->getType()->isAtomicType());
-  return AtomicExprEvaluator(Info, Result).Visit(E);
+  return AtomicExprEvaluator(Info, This, Result).Visit(E);
 }
 
 //===----------------------------------------------------------------------===//
@@ -9699,8 +9838,17 @@ static bool Evaluate(APValue &Result, EvalInfo &Info, const Expr *E) {
     if (!EvaluateVoid(E, Info))
       return false;
   } else if (T->isAtomicType()) {
-    if (!EvaluateAtomic(E, Result, Info))
-      return false;
+    QualType Unqual = T.getAtomicUnqualifiedType();
+    if (Unqual->isArrayType() || Unqual->isRecordType()) {
+      LValue LV;
+      LV.set(E, Info.CurrentCall->Index);
+      APValue &Value = Info.CurrentCall->createTemporary(E, false);
+      if (!EvaluateAtomic(E, &LV, Value, Info))
+        return false;
+    } else {
+      if (!EvaluateAtomic(E, nullptr, Result, Info))
+        return false;
+    }
   } else if (Info.getLangOpts().CPlusPlus11) {
     Info.FFDiag(E, diag::note_constexpr_nonliteral) << E->getType();
     return false;
@@ -9725,10 +9873,16 @@ static bool EvaluateInPlace(APValue &Result, EvalInfo &Info, const LValue &This,
   if (E->isRValue()) {
     // Evaluate arrays and record types in-place, so that later initializers can
     // refer to earlier-initialized members of the object.
-    if (E->getType()->isArrayType())
+    QualType T = E->getType();
+    if (T->isArrayType())
       return EvaluateArray(E, This, Result, Info);
-    else if (E->getType()->isRecordType())
+    else if (T->isRecordType())
       return EvaluateRecord(E, This, Result, Info);
+    else if (T->isAtomicType()) {
+      QualType Unqual = T.getAtomicUnqualifiedType();
+      if (Unqual->isArrayType() || Unqual->isRecordType())
+        return EvaluateAtomic(E, &This, Result, Info);
+    }
   }
 
   // For any other type, in-place evaluation is unimportant.
@@ -9945,7 +10099,7 @@ bool Expr::EvalResult::isGlobalLValue() const {
 // Note that to reduce code duplication, this helper does no evaluation
 // itself; the caller checks whether the expression is evaluatable, and
 // in the rare cases where CheckICE actually cares about the evaluated
-// value, it calls into Evalute.
+// value, it calls into Evaluate.
 
 namespace {
 
@@ -10067,6 +10221,7 @@ static ICEDiag CheckICE(const Expr* E, const ASTContext &Ctx) {
   case Expr::LambdaExprClass:
   case Expr::CXXFoldExprClass:
   case Expr::CoawaitExprClass:
+  case Expr::DependentCoawaitExprClass:
   case Expr::CoyieldExprClass:
     return ICEDiag(IK_NotICE, E->getLocStart());
 
diff --git a/lib/AST/ExternalASTMerger.cpp b/lib/AST/ExternalASTMerger.cpp
new file mode 100644
index 000000000000..2d4d0185ff2a
--- /dev/null
+++ b/lib/AST/ExternalASTMerger.cpp
@@ -0,0 +1,185 @@
+//===- ExternalASTMerger.cpp - Merging External AST Interface ---*- C++ -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+//  This file implements the ExternalASTMerger, which vends a combination of
+//  ASTs from several different ASTContext/FileManager pairs
+//
+//===----------------------------------------------------------------------===//
+
+#include "clang/AST/ASTContext.h"
+#include "clang/AST/Decl.h"
+#include "clang/AST/DeclObjC.h"
+#include "clang/AST/ExternalASTMerger.h"
+
+using namespace clang;
+
+namespace {
+
+template <typename T> struct Source {
+  T t;
+  Source(T t) : t(t) {}
+  operator T() { return t; }
+  template <typename U = T> U &get() { return t; }
+  template <typename U = T> const U &get() const { return t; }
+  template <typename U> operator Source<U>() { return Source<U>(t); }
+};
+
+typedef std::pair<Source<NamedDecl *>, ASTImporter *> Candidate;
+
+class LazyASTImporter : public ASTImporter {
+public:
+  LazyASTImporter(ASTContext &ToContext, FileManager &ToFileManager,
+                  ASTContext &FromContext, FileManager &FromFileManager)
+      : ASTImporter(ToContext, ToFileManager, FromContext, FromFileManager,
+                    /*MinimalImport=*/true) {}
+  Decl *Imported(Decl *From, Decl *To) override {
+    if (auto ToTag = dyn_cast<TagDecl>(To)) {
+      ToTag->setHasExternalLexicalStorage();
+    } else if (auto ToNamespace = dyn_cast<NamespaceDecl>(To)) {
+      ToNamespace->setHasExternalVisibleStorage();
+    }
+    return ASTImporter::Imported(From, To);
+  }
+};
+
+Source<const DeclContext *>
+LookupSameContext(Source<TranslationUnitDecl *> SourceTU, const DeclContext *DC,
+                  ASTImporter &ReverseImporter) {
+  if (DC->isTranslationUnit()) {
+    return SourceTU;
+  }
+  Source<const DeclContext *> SourceParentDC =
+      LookupSameContext(SourceTU, DC->getParent(), ReverseImporter);
+  if (!SourceParentDC) {
+    // If we couldn't find the parent DC in this TranslationUnit, give up.
+    return nullptr;
+  }
+  auto ND = cast<NamedDecl>(DC);
+  DeclarationName Name = ND->getDeclName();
+  Source<DeclarationName> SourceName = ReverseImporter.Import(Name);
+  DeclContext::lookup_result SearchResult =
+      SourceParentDC.get()->lookup(SourceName.get());
+  size_t SearchResultSize = SearchResult.size();
+  // Handle multiple candidates once we have a test for it.
+  // This may turn up when we import template specializations correctly.
+  assert(SearchResultSize < 2);
+  if (SearchResultSize == 0) {
+    // couldn't find the name, so we have to give up
+    return nullptr;
+  } else {
+    NamedDecl *SearchResultDecl = SearchResult[0];
+    return dyn_cast<DeclContext>(SearchResultDecl);
+  }
+}
+
+bool IsForwardDeclaration(Decl *D) {
+  assert(!isa<ObjCInterfaceDecl>(D)); // TODO handle this case
+  if (auto TD = dyn_cast<TagDecl>(D)) {
+    return !TD->isThisDeclarationADefinition();
+  } else if (auto FD = dyn_cast<FunctionDecl>(D)) {
+    return !FD->isThisDeclarationADefinition();
+  } else {
+    return false;
+  }
+}
+
+void ForEachMatchingDC(
+    const DeclContext *DC,
+    llvm::ArrayRef<ExternalASTMerger::ImporterPair> Importers,
+    std::function<void(const ExternalASTMerger::ImporterPair &IP,
+                       Source<const DeclContext *> SourceDC)>
+        Callback) {
+  for (const ExternalASTMerger::ImporterPair &IP : Importers) {
+    Source<TranslationUnitDecl *> SourceTU(
+        IP.Forward->getFromContext().getTranslationUnitDecl());
+    Source<const DeclContext *> SourceDC =
+        LookupSameContext(SourceTU, DC, *IP.Reverse);
+    if (SourceDC.get()) {
+      Callback(IP, SourceDC);
+    }
+  }
+}
+
+bool HasDeclOfSameType(llvm::ArrayRef<Candidate> Decls, const Candidate &C) {
+  return std::any_of(Decls.begin(), Decls.end(), [&C](const Candidate &D) {
+    return C.first.get()->getKind() == D.first.get()->getKind();
+  });
+}
+} // end namespace
+
+ExternalASTMerger::ExternalASTMerger(const ImporterEndpoint &Target,
+                                     llvm::ArrayRef<ImporterEndpoint> Sources) {
+  for (const ImporterEndpoint &S : Sources) {
+    Importers.push_back(
+        {llvm::make_unique<LazyASTImporter>(Target.AST, Target.FM, S.AST, S.FM),
+         llvm::make_unique<ASTImporter>(S.AST, S.FM, Target.AST, Target.FM,
+                                        /*MinimalImport=*/true)});
+  }
+}
+
+bool ExternalASTMerger::FindExternalVisibleDeclsByName(const DeclContext *DC,
+                                                       DeclarationName Name) {
+  llvm::SmallVector<NamedDecl *, 1> Decls;
+  llvm::SmallVector<Candidate, 4> CompleteDecls;
+  llvm::SmallVector<Candidate, 4> ForwardDecls;
+
+  auto FilterFoundDecl = [&CompleteDecls, &ForwardDecls](const Candidate &C) {
+    if (IsForwardDeclaration(C.first.get())) {
+      if (!HasDeclOfSameType(ForwardDecls, C)) {
+        ForwardDecls.push_back(C);
+      }
+    } else {
+      CompleteDecls.push_back(C);
+    }
+  };
+
+  ForEachMatchingDC(DC, Importers, [Name, &FilterFoundDecl](
+                                       const ImporterPair &IP,
+                                       Source<const DeclContext *> SourceDC) {
+    DeclarationName FromName = IP.Reverse->Import(Name);
+    DeclContextLookupResult Result = SourceDC.get()->lookup(FromName);
+    for (NamedDecl *FromD : Result) {
+      FilterFoundDecl(std::make_pair(FromD, IP.Forward.get()));
+    }
+  });
+
+  llvm::ArrayRef<Candidate> DeclsToReport =
+      CompleteDecls.empty() ? ForwardDecls : CompleteDecls;
+
+  if (DeclsToReport.empty()) {
+    return false;
+  }
+
+  Decls.reserve(DeclsToReport.size());
+  for (const Candidate &C : DeclsToReport) {
+    NamedDecl *d = cast<NamedDecl>(C.second->Import(C.first.get()));
+    assert(d);
+    Decls.push_back(d);
+  }
+  SetExternalVisibleDeclsForName(DC, Name, Decls);
+  return true;
+}
+
+void ExternalASTMerger::FindExternalLexicalDecls(
+    const DeclContext *DC, llvm::function_ref<bool(Decl::Kind)> IsKindWeWant,
+    SmallVectorImpl<Decl *> &Result) {
+  ForEachMatchingDC(
+      DC, Importers, [DC, IsKindWeWant](const ImporterPair &IP,
+                                        Source<const DeclContext *> SourceDC) {
+        for (const Decl *SourceDecl : SourceDC.get()->decls()) {
+          if (IsKindWeWant(SourceDecl->getKind())) {
+            Decl *ImportedDecl =
+                IP.Forward->Import(const_cast<Decl *>(SourceDecl));
+            assert(ImportedDecl->getDeclContext() == DC);
+            (void)ImportedDecl;
+            (void)DC;
+          }
+        }
+      });
+}
diff --git a/lib/AST/ExternalASTSource.cpp b/lib/AST/ExternalASTSource.cpp
index e3de8c5fefa2..182d38242f59 100644
--- a/lib/AST/ExternalASTSource.cpp
+++ b/lib/AST/ExternalASTSource.cpp
@@ -28,6 +28,11 @@ ExternalASTSource::getSourceDescriptor(unsigned ID) {
   return None;
 }
 
+ExternalASTSource::ExtKind
+ExternalASTSource::hasExternalDefinitions(const Decl *D) {
+  return EK_ReplyHazy;
+}
+
 ExternalASTSource::ASTSourceDescriptor::ASTSourceDescriptor(const Module &M)
   : Signature(M.Signature), ClangModule(&M) {
   if (M.Directory)
diff --git a/lib/AST/ItaniumMangle.cpp b/lib/AST/ItaniumMangle.cpp
index ab3e49d903cf..29fcdd7be924 100644
--- a/lib/AST/ItaniumMangle.cpp
+++ b/lib/AST/ItaniumMangle.cpp
@@ -982,9 +982,8 @@ void CXXNameMangler::mangleFloat(const llvm::APFloat &f) {
     unsigned digitBitIndex = 4 * (numCharacters - stringIndex - 1);
 
     // Project out 4 bits starting at 'digitIndex'.
-    llvm::integerPart hexDigit
-      = valueBits.getRawData()[digitBitIndex / llvm::integerPartWidth];
-    hexDigit >>= (digitBitIndex % llvm::integerPartWidth);
+    uint64_t hexDigit = valueBits.getRawData()[digitBitIndex / 64];
+    hexDigit >>= (digitBitIndex % 64);
     hexDigit &= 0xF;
 
     // Map that over to a lowercase hex digit.
@@ -1190,6 +1189,8 @@ void CXXNameMangler::mangleUnresolvedName(
       llvm_unreachable("Can't mangle a constructor name!");
     case DeclarationName::CXXUsingDirective:
       llvm_unreachable("Can't mangle a using directive name!");
+    case DeclarationName::CXXDeductionGuideName:
+      llvm_unreachable("Can't mangle a deduction guide name!");
     case DeclarationName::ObjCMultiArgSelector:
     case DeclarationName::ObjCOneArgSelector:
     case DeclarationName::ObjCZeroArgSelector:
@@ -1419,6 +1420,9 @@ void CXXNameMangler::mangleUnqualifiedName(const NamedDecl *ND,
     writeAbiTags(ND, AdditionalAbiTags);
     break;
 
+  case DeclarationName::CXXDeductionGuideName:
+    llvm_unreachable("Can't mangle a deduction guide name!");
+
   case DeclarationName::CXXUsingDirective:
     llvm_unreachable("Can't mangle a using directive name!");
   }
@@ -1870,6 +1874,7 @@ bool CXXNameMangler::mangleUnresolvedTypeOrSimpleId(QualType Ty,
   case Type::Paren:
   case Type::Attributed:
   case Type::Auto:
+  case Type::DeducedTemplateSpecialization:
   case Type::PackExpansion:
   case Type::ObjCObject:
   case Type::ObjCInterface:
@@ -1996,6 +2001,7 @@ void CXXNameMangler::mangleOperatorName(DeclarationName Name, unsigned Arity) {
   switch (Name.getNameKind()) {
   case DeclarationName::CXXConstructorName:
   case DeclarationName::CXXDestructorName:
+  case DeclarationName::CXXDeductionGuideName:
   case DeclarationName::CXXUsingDirective:
   case DeclarationName::Identifier:
   case DeclarationName::ObjCMultiArgSelector:
@@ -2152,10 +2158,12 @@ void CXXNameMangler::mangleQualifiers(Qualifiers Quals) {
     } else {
       switch (AS) {
       default: llvm_unreachable("Not a language specific address space");
-      //  <OpenCL-addrspace> ::= "CL" [ "global" | "local" | "constant" ]
+      //  <OpenCL-addrspace> ::= "CL" [ "global" | "local" | "constant |
+      //                                "generic" ]
       case LangAS::opencl_global:   ASString = "CLglobal";   break;
       case LangAS::opencl_local:    ASString = "CLlocal";    break;
       case LangAS::opencl_constant: ASString = "CLconstant"; break;
+      case LangAS::opencl_generic:  ASString = "CLgeneric";  break;
       //  <CUDA-addrspace> ::= "CU" [ "device" | "constant" | "shared" ]
       case LangAS::cuda_device:     ASString = "CUdevice";   break;
       case LangAS::cuda_constant:   ASString = "CUconstant"; break;
@@ -2493,9 +2501,6 @@ void CXXNameMangler::mangleType(const BuiltinType *T) {
   case BuiltinType::OCLQueue:
     Out << "9ocl_queue";
     break;
-  case BuiltinType::OCLNDRange:
-    Out << "11ocl_ndrange";
-    break;
   case BuiltinType::OCLReserveID:
     Out << "13ocl_reserveid";
     break;
@@ -3043,6 +3048,7 @@ void CXXNameMangler::mangleType(const DependentNameType *T) {
   //                   ::= Te <name> # dependent elaborated type specifier using
   //                                 # 'enum'
   switch (T->getKeyword()) {
+    case ETK_None:
     case ETK_Typename:
       break;
     case ETK_Struct:
@@ -3056,8 +3062,6 @@ void CXXNameMangler::mangleType(const DependentNameType *T) {
     case ETK_Enum:
       Out << "Te";
       break;
-    default:
-      llvm_unreachable("unexpected keyword for dependent type name");
   }
   // Typename types are always nested
   Out << 'N';
@@ -3146,6 +3150,16 @@ void CXXNameMangler::mangleType(const AutoType *T) {
     mangleType(D);
 }
 
+void CXXNameMangler::mangleType(const DeducedTemplateSpecializationType *T) {
+  // FIXME: This is not the right mangling. We also need to include a scope
+  // here in some cases.
+  QualType D = T->getDeducedType();
+  if (D.isNull())
+    mangleUnscopedTemplateName(T->getTemplateName(), nullptr);
+  else
+    mangleType(D);
+}
+
 void CXXNameMangler::mangleType(const AtomicType *T) {
   // <type> ::= U <source-name> <type>  # vendor extended type qualifier
   // (Until there's a standardized mangling...)
@@ -4021,6 +4035,12 @@ recurse:
     mangleExpression(cast<CoawaitExpr>(E)->getOperand());
     break;
 
+  case Expr::DependentCoawaitExprClass:
+    // FIXME: Propose a non-vendor mangling.
+    Out << "v18co_await";
+    mangleExpression(cast<DependentCoawaitExpr>(E)->getOperand());
+    break;
+
   case Expr::CoyieldExprClass:
     // FIXME: Propose a non-vendor mangling.
     Out << "v18co_yield";
diff --git a/lib/AST/Mangle.cpp b/lib/AST/Mangle.cpp
index 05dd886adcef..00d50c0e3bdf 100644
--- a/lib/AST/Mangle.cpp
+++ b/lib/AST/Mangle.cpp
@@ -262,9 +262,13 @@ void MangleContext::mangleObjCMethodNameWithoutSize(const ObjCMethodDecl *MD,
   const ObjCContainerDecl *CD =
   dyn_cast<ObjCContainerDecl>(MD->getDeclContext());
   assert (CD && "Missing container decl in GetNameForMethod");
-  OS << (MD->isInstanceMethod() ? '-' : '+') << '[' << CD->getName();
-  if (const ObjCCategoryImplDecl *CID = dyn_cast<ObjCCategoryImplDecl>(CD))
+  OS << (MD->isInstanceMethod() ? '-' : '+') << '[';
+  if (const ObjCCategoryImplDecl *CID = dyn_cast<ObjCCategoryImplDecl>(CD)) {
+    OS << CID->getClassInterface()->getName();
     OS << '(' << *CID << ')';
+  } else {
+    OS << CD->getName();
+  }
   OS << ' ';
   MD->getSelector().print(OS);
   OS << ']';
diff --git a/lib/AST/MicrosoftMangle.cpp b/lib/AST/MicrosoftMangle.cpp
index 76c368d7f04c..6e14dd055cf8 100644
--- a/lib/AST/MicrosoftMangle.cpp
+++ b/lib/AST/MicrosoftMangle.cpp
@@ -942,6 +942,9 @@ void MicrosoftCXXNameMangler::mangleUnqualifiedName(const NamedDecl *ND,
       break;
     }
 
+    case DeclarationName::CXXDeductionGuideName:
+      llvm_unreachable("Can't mangle a deduction guide name!");
+
     case DeclarationName::CXXUsingDirective:
       llvm_unreachable("Can't mangle a using directive name!");
   }
@@ -1797,10 +1800,6 @@ void MicrosoftCXXNameMangler::mangleType(const BuiltinType *T, Qualifiers,
     Out << "PA";
     mangleArtificalTagType(TTK_Struct, "ocl_queue");
     break;
-  case BuiltinType::OCLNDRange:
-    Out << "PA";
-    mangleArtificalTagType(TTK_Struct, "ocl_ndrange");
-    break;
   case BuiltinType::OCLReserveID:
     Out << "PA";
     mangleArtificalTagType(TTK_Struct, "ocl_reserveid");
@@ -1887,14 +1886,18 @@ void MicrosoftCXXNameMangler::mangleFunctionType(const FunctionType *T,
   // <return-type> ::= <type>
   //               ::= @ # structors (they have no declared return type)
   if (IsStructor) {
-    if (isa<CXXDestructorDecl>(D) && isStructorDecl(D) &&
-        StructorType == Dtor_Deleting) {
-      // The scalar deleting destructor takes an extra int argument.
-      // However, the FunctionType generated has 0 arguments.
-      // FIXME: This is a temporary hack.
-      // Maybe should fix the FunctionType creation instead?
-      Out << (PointersAre64Bit ? "PEAXI@Z" : "PAXI@Z");
-      return;
+    if (isa<CXXDestructorDecl>(D) && isStructorDecl(D)) {
+      // The scalar deleting destructor takes an extra int argument which is not
+      // reflected in the AST.
+      if (StructorType == Dtor_Deleting) {
+        Out << (PointersAre64Bit ? "PEAXI@Z" : "PAXI@Z");
+        return;
+      }
+      // The vbase destructor returns void which is not reflected in the AST.
+      if (StructorType == Dtor_Complete) {
+        Out << "XXZ";
+        return;
+      }
     }
     if (IsCtorClosure) {
       // Default constructor closure and copy constructor closure both return
@@ -1954,7 +1957,7 @@ void MicrosoftCXXNameMangler::mangleFunctionType(const FunctionType *T,
     // Happens for function pointer type arguments for example.
     for (unsigned I = 0, E = Proto->getNumParams(); I != E; ++I) {
       mangleArgumentType(Proto->getParamType(I), Range);
-      // Mangle each pass_object_size parameter as if it's a paramater of enum
+      // Mangle each pass_object_size parameter as if it's a parameter of enum
       // type passed directly after the parameter with the pass_object_size
       // attribute. The aforementioned enum's name is __pass_object_size, and we
       // pretend it resides in a top-level namespace called __clang.
@@ -2002,13 +2005,20 @@ void MicrosoftCXXNameMangler::mangleFunctionClass(const FunctionDecl *FD) {
   // <global-function> ::= Y # global near
   //                   ::= Z # global far
   if (const CXXMethodDecl *MD = dyn_cast<CXXMethodDecl>(FD)) {
+    bool IsVirtual = MD->isVirtual();
+    // When mangling vbase destructor variants, ignore whether or not the
+    // underlying destructor was defined to be virtual.
+    if (isa<CXXDestructorDecl>(MD) && isStructorDecl(MD) &&
+        StructorType == Dtor_Complete) {
+      IsVirtual = false;
+    }
     switch (MD->getAccess()) {
       case AS_none:
         llvm_unreachable("Unsupported access specifier");
       case AS_private:
         if (MD->isStatic())
           Out << 'C';
-        else if (MD->isVirtual())
+        else if (IsVirtual)
           Out << 'E';
         else
           Out << 'A';
@@ -2016,7 +2026,7 @@ void MicrosoftCXXNameMangler::mangleFunctionClass(const FunctionDecl *FD) {
       case AS_protected:
         if (MD->isStatic())
           Out << 'K';
-        else if (MD->isVirtual())
+        else if (IsVirtual)
           Out << 'M';
         else
           Out << 'I';
@@ -2024,7 +2034,7 @@ void MicrosoftCXXNameMangler::mangleFunctionClass(const FunctionDecl *FD) {
       case AS_public:
         if (MD->isStatic())
           Out << 'S';
-        else if (MD->isVirtual())
+        else if (IsVirtual)
           Out << 'U';
         else
           Out << 'Q';
@@ -2474,6 +2484,17 @@ void MicrosoftCXXNameMangler::mangleType(const AutoType *T, Qualifiers,
     << Range;
 }
 
+void MicrosoftCXXNameMangler::mangleType(
+    const DeducedTemplateSpecializationType *T, Qualifiers, SourceRange Range) {
+  assert(T->getDeducedType().isNull() && "expecting a dependent type!");
+
+  DiagnosticsEngine &Diags = Context.getDiags();
+  unsigned DiagID = Diags.getCustomDiagID(DiagnosticsEngine::Error,
+    "cannot mangle this deduced class template specialization type yet");
+  Diags.Report(Range.getBegin(), DiagID)
+    << Range;
+}
+
 void MicrosoftCXXNameMangler::mangleType(const AtomicType *T, Qualifiers,
                                          SourceRange Range) {
   QualType ValueType = T->getValueType();
@@ -2997,14 +3018,14 @@ void MicrosoftMangleContextImpl::mangleStringLiteral(const StringLiteral *SL,
   // N.B. The length is in terms of bytes, not characters.
   Mangler.mangleNumber(SL->getByteLength() + SL->getCharByteWidth());
 
-  auto GetLittleEndianByte = [&Mangler, &SL](unsigned Index) {
+  auto GetLittleEndianByte = [&SL](unsigned Index) {
     unsigned CharByteWidth = SL->getCharByteWidth();
     uint32_t CodeUnit = SL->getCodeUnit(Index / CharByteWidth);
     unsigned OffsetInCodeUnit = Index % CharByteWidth;
     return static_cast<char>((CodeUnit >> (8 * OffsetInCodeUnit)) & 0xff);
   };
 
-  auto GetBigEndianByte = [&Mangler, &SL](unsigned Index) {
+  auto GetBigEndianByte = [&SL](unsigned Index) {
     unsigned CharByteWidth = SL->getCharByteWidth();
     uint32_t CodeUnit = SL->getCodeUnit(Index / CharByteWidth);
     unsigned OffsetInCodeUnit = (CharByteWidth - 1) - (Index % CharByteWidth);
diff --git a/lib/AST/NSAPI.cpp b/lib/AST/NSAPI.cpp
index ac2a8d354247..e7c8c16b0145 100644
--- a/lib/AST/NSAPI.cpp
+++ b/lib/AST/NSAPI.cpp
@@ -453,7 +453,6 @@ NSAPI::getNSNumberFactoryMethodKind(QualType T) const {
   case BuiltinType::OCLEvent:
   case BuiltinType::OCLClkEvent:
   case BuiltinType::OCLQueue:
-  case BuiltinType::OCLNDRange:
   case BuiltinType::OCLReserveID:
   case BuiltinType::BoundMember:
   case BuiltinType::Dependent:
diff --git a/lib/AST/ODRHash.cpp b/lib/AST/ODRHash.cpp
new file mode 100644
index 000000000000..d72eebbe8e48
--- /dev/null
+++ b/lib/AST/ODRHash.cpp
@@ -0,0 +1,361 @@
+//===-- ODRHash.cpp - Hashing to diagnose ODR failures ----------*- C++ -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file implements the ODRHash class, which calculates a hash based
+/// on AST nodes, which is stable across different runs.
+///
+//===----------------------------------------------------------------------===//
+
+#include "clang/AST/ODRHash.h"
+
+#include "clang/AST/DeclVisitor.h"
+#include "clang/AST/NestedNameSpecifier.h"
+#include "clang/AST/StmtVisitor.h"
+#include "clang/AST/TypeVisitor.h"
+
+using namespace clang;
+
+void ODRHash::AddStmt(const Stmt *S) {
+  assert(S && "Expecting non-null pointer.");
+  S->ProcessODRHash(ID, *this);
+}
+
+void ODRHash::AddIdentifierInfo(const IdentifierInfo *II) {
+  assert(II && "Expecting non-null pointer.");
+  ID.AddString(II->getName());
+}
+
+void ODRHash::AddDeclarationName(DeclarationName Name) {
+  AddBoolean(Name.isEmpty());
+  if (Name.isEmpty())
+    return;
+
+  auto Kind = Name.getNameKind();
+  ID.AddInteger(Kind);
+  switch (Kind) {
+  case DeclarationName::Identifier:
+    AddIdentifierInfo(Name.getAsIdentifierInfo());
+    break;
+  case DeclarationName::ObjCZeroArgSelector:
+  case DeclarationName::ObjCOneArgSelector:
+  case DeclarationName::ObjCMultiArgSelector: {
+    Selector S = Name.getObjCSelector();
+    AddBoolean(S.isNull());
+    AddBoolean(S.isKeywordSelector());
+    AddBoolean(S.isUnarySelector());
+    unsigned NumArgs = S.getNumArgs();
+    for (unsigned i = 0; i < NumArgs; ++i) {
+      AddIdentifierInfo(S.getIdentifierInfoForSlot(i));
+    }
+    break;
+  }
+  case DeclarationName::CXXConstructorName:
+  case DeclarationName::CXXDestructorName:
+    AddQualType(Name.getCXXNameType());
+    break;
+  case DeclarationName::CXXOperatorName:
+    ID.AddInteger(Name.getCXXOverloadedOperator());
+    break;
+  case DeclarationName::CXXLiteralOperatorName:
+    AddIdentifierInfo(Name.getCXXLiteralIdentifier());
+    break;
+  case DeclarationName::CXXConversionFunctionName:
+    AddQualType(Name.getCXXNameType());
+    break;
+  case DeclarationName::CXXUsingDirective:
+    break;
+  case DeclarationName::CXXDeductionGuideName: {
+    auto *Template = Name.getCXXDeductionGuideTemplate();
+    AddBoolean(Template);
+    if (Template) {
+      AddDecl(Template);
+    }
+  }
+  }
+}
+
+void ODRHash::AddNestedNameSpecifier(const NestedNameSpecifier *NNS) {}
+void ODRHash::AddTemplateName(TemplateName Name) {}
+void ODRHash::AddTemplateArgument(TemplateArgument TA) {}
+void ODRHash::AddTemplateParameterList(const TemplateParameterList *TPL) {}
+
+void ODRHash::clear() {
+  DeclMap.clear();
+  TypeMap.clear();
+  Bools.clear();
+  ID.clear();
+}
+
+unsigned ODRHash::CalculateHash() {
+  // Append the bools to the end of the data segment backwards.  This allows
+  // for the bools data to be compressed 32 times smaller compared to using
+  // ID.AddBoolean
+  const unsigned unsigned_bits = sizeof(unsigned) * CHAR_BIT;
+  const unsigned size = Bools.size();
+  const unsigned remainder = size % unsigned_bits;
+  const unsigned loops = size / unsigned_bits;
+  auto I = Bools.rbegin();
+  unsigned value = 0;
+  for (unsigned i = 0; i < remainder; ++i) {
+    value <<= 1;
+    value |= *I;
+    ++I;
+  }
+  ID.AddInteger(value);
+
+  for (unsigned i = 0; i < loops; ++i) {
+    value = 0;
+    for (unsigned j = 0; j < unsigned_bits; ++j) {
+      value <<= 1;
+      value |= *I;
+      ++I;
+    }
+    ID.AddInteger(value);
+  }
+
+  assert(I == Bools.rend());
+  Bools.clear();
+  return ID.ComputeHash();
+}
+
+// Process a Decl pointer.  Add* methods call back into ODRHash while Visit*
+// methods process the relevant parts of the Decl.
+class ODRDeclVisitor : public ConstDeclVisitor<ODRDeclVisitor> {
+  typedef ConstDeclVisitor<ODRDeclVisitor> Inherited;
+  llvm::FoldingSetNodeID &ID;
+  ODRHash &Hash;
+
+public:
+  ODRDeclVisitor(llvm::FoldingSetNodeID &ID, ODRHash &Hash)
+      : ID(ID), Hash(Hash) {}
+
+  void AddStmt(const Stmt *S) {
+    Hash.AddBoolean(S);
+    if (S) {
+      Hash.AddStmt(S);
+    }
+  }
+
+  void AddIdentifierInfo(const IdentifierInfo *II) {
+    Hash.AddBoolean(II);
+    if (II) {
+      Hash.AddIdentifierInfo(II);
+    }
+  }
+
+  void AddQualType(QualType T) {
+    Hash.AddQualType(T);
+  }
+
+  void Visit(const Decl *D) {
+    ID.AddInteger(D->getKind());
+    Inherited::Visit(D);
+  }
+
+  void VisitNamedDecl(const NamedDecl *D) {
+    Hash.AddDeclarationName(D->getDeclName());
+    Inherited::VisitNamedDecl(D);
+  }
+
+  void VisitValueDecl(const ValueDecl *D) {
+    AddQualType(D->getType());
+    Inherited::VisitValueDecl(D);
+  }
+
+  void VisitAccessSpecDecl(const AccessSpecDecl *D) {
+    ID.AddInteger(D->getAccess());
+    Inherited::VisitAccessSpecDecl(D);
+  }
+
+  void VisitStaticAssertDecl(const StaticAssertDecl *D) {
+    AddStmt(D->getAssertExpr());
+    AddStmt(D->getMessage());
+
+    Inherited::VisitStaticAssertDecl(D);
+  }
+
+  void VisitFieldDecl(const FieldDecl *D) {
+    const bool IsBitfield = D->isBitField();
+    Hash.AddBoolean(IsBitfield);
+
+    if (IsBitfield) {
+      AddStmt(D->getBitWidth());
+    }
+
+    Hash.AddBoolean(D->isMutable());
+    AddStmt(D->getInClassInitializer());
+
+    Inherited::VisitFieldDecl(D);
+  }
+
+  void VisitFunctionDecl(const FunctionDecl *D) {
+    ID.AddInteger(D->getStorageClass());
+    Hash.AddBoolean(D->isInlineSpecified());
+    Hash.AddBoolean(D->isVirtualAsWritten());
+    Hash.AddBoolean(D->isPure());
+    Hash.AddBoolean(D->isDeletedAsWritten());
+
+    Inherited::VisitFunctionDecl(D);
+  }
+
+  void VisitCXXMethodDecl(const CXXMethodDecl *D) {
+    Hash.AddBoolean(D->isConst());
+    Hash.AddBoolean(D->isVolatile());
+
+    Inherited::VisitCXXMethodDecl(D);
+  }
+
+  void VisitTypedefNameDecl(const TypedefNameDecl *D) {
+    AddQualType(D->getUnderlyingType());
+
+    Inherited::VisitTypedefNameDecl(D);
+  }
+
+  void VisitTypedefDecl(const TypedefDecl *D) {
+    Inherited::VisitTypedefDecl(D);
+  }
+
+  void VisitTypeAliasDecl(const TypeAliasDecl *D) {
+    Inherited::VisitTypeAliasDecl(D);
+  }
+};
+
+// Only allow a small portion of Decl's to be processed.  Remove this once
+// all Decl's can be handled.
+bool ODRHash::isWhitelistedDecl(const Decl *D, const CXXRecordDecl *Parent) {
+  if (D->isImplicit()) return false;
+  if (D->getDeclContext() != Parent) return false;
+
+  switch (D->getKind()) {
+    default:
+      return false;
+    case Decl::AccessSpec:
+    case Decl::CXXMethod:
+    case Decl::Field:
+    case Decl::StaticAssert:
+    case Decl::TypeAlias:
+    case Decl::Typedef:
+      return true;
+  }
+}
+
+void ODRHash::AddSubDecl(const Decl *D) {
+  assert(D && "Expecting non-null pointer.");
+  AddDecl(D);
+
+  ODRDeclVisitor(ID, *this).Visit(D);
+}
+
+void ODRHash::AddCXXRecordDecl(const CXXRecordDecl *Record) {
+  assert(Record && Record->hasDefinition() &&
+         "Expected non-null record to be a definition.");
+  AddDecl(Record);
+
+  // Filter out sub-Decls which will not be processed in order to get an
+  // accurate count of Decl's.
+  llvm::SmallVector<const Decl *, 16> Decls;
+  for (const Decl *SubDecl : Record->decls()) {
+    if (isWhitelistedDecl(SubDecl, Record)) {
+      Decls.push_back(SubDecl);
+    }
+  }
+
+  ID.AddInteger(Decls.size());
+  for (auto SubDecl : Decls) {
+    AddSubDecl(SubDecl);
+  }
+}
+
+void ODRHash::AddDecl(const Decl *D) {
+  assert(D && "Expecting non-null pointer.");
+  auto Result = DeclMap.insert(std::make_pair(D, DeclMap.size()));
+  ID.AddInteger(Result.first->second);
+  // On first encounter of a Decl pointer, process it.  Every time afterwards,
+  // only the index value is needed.
+  if (!Result.second) {
+    return;
+  }
+
+  ID.AddInteger(D->getKind());
+
+  if (const NamedDecl *ND = dyn_cast<NamedDecl>(D)) {
+    AddDeclarationName(ND->getDeclName());
+  }
+}
+
+// Process a Type pointer.  Add* methods call back into ODRHash while Visit*
+// methods process the relevant parts of the Type.
+class ODRTypeVisitor : public TypeVisitor<ODRTypeVisitor> {
+  typedef TypeVisitor<ODRTypeVisitor> Inherited;
+  llvm::FoldingSetNodeID &ID;
+  ODRHash &Hash;
+
+public:
+  ODRTypeVisitor(llvm::FoldingSetNodeID &ID, ODRHash &Hash)
+      : ID(ID), Hash(Hash) {}
+
+  void AddStmt(Stmt *S) {
+    Hash.AddBoolean(S);
+    if (S) {
+      Hash.AddStmt(S);
+    }
+  }
+
+  void AddDecl(Decl *D) {
+    Hash.AddBoolean(D);
+    if (D) {
+      Hash.AddDecl(D);
+    }
+  }
+
+  void Visit(const Type *T) {
+    ID.AddInteger(T->getTypeClass());
+    Inherited::Visit(T);
+  }
+
+  void VisitType(const Type *T) {}
+
+  void VisitBuiltinType(const BuiltinType *T) {
+    ID.AddInteger(T->getKind());
+    VisitType(T);
+  }
+
+  void VisitTypedefType(const TypedefType *T) {
+    AddDecl(T->getDecl());
+    Hash.AddQualType(T->getDecl()->getUnderlyingType());
+    VisitType(T);
+  }
+};
+
+void ODRHash::AddType(const Type *T) {
+  assert(T && "Expecting non-null pointer.");
+  auto Result = TypeMap.insert(std::make_pair(T, TypeMap.size()));
+  ID.AddInteger(Result.first->second);
+  // On first encounter of a Type pointer, process it.  Every time afterwards,
+  // only the index value is needed.
+  if (!Result.second) {
+    return;
+  }
+
+  ODRTypeVisitor(ID, *this).Visit(T);
+}
+
+void ODRHash::AddQualType(QualType T) {
+  AddBoolean(T.isNull());
+  if (T.isNull())
+    return;
+  SplitQualType split = T.split();
+  ID.AddInteger(split.Quals.getAsOpaqueValue());
+  AddType(split.Ty);
+}
+
+void ODRHash::AddBoolean(bool Value) {
+  Bools.push_back(Value);
+}
diff --git a/lib/AST/OpenMPClause.cpp b/lib/AST/OpenMPClause.cpp
index a28b9f3b6d64..77470a9b76d0 100644
--- a/lib/AST/OpenMPClause.cpp
+++ b/lib/AST/OpenMPClause.cpp
@@ -48,11 +48,17 @@ const OMPClauseWithPreInit *OMPClauseWithPreInit::get(const OMPClause *C) {
     return static_cast<const OMPReductionClause *>(C);
   case OMPC_linear:
     return static_cast<const OMPLinearClause *>(C);
+  case OMPC_if:
+    return static_cast<const OMPIfClause *>(C);
+  case OMPC_num_threads:
+    return static_cast<const OMPNumThreadsClause *>(C);
+  case OMPC_num_teams:
+    return static_cast<const OMPNumTeamsClause *>(C);
+  case OMPC_thread_limit:
+    return static_cast<const OMPThreadLimitClause *>(C);
   case OMPC_default:
   case OMPC_proc_bind:
-  case OMPC_if:
   case OMPC_final:
-  case OMPC_num_threads:
   case OMPC_safelen:
   case OMPC_simdlen:
   case OMPC_collapse:
@@ -77,8 +83,6 @@ const OMPClauseWithPreInit *OMPClauseWithPreInit::get(const OMPClause *C) {
   case OMPC_threads:
   case OMPC_simd:
   case OMPC_map:
-  case OMPC_num_teams:
-  case OMPC_thread_limit:
   case OMPC_priority:
   case OMPC_grainsize:
   case OMPC_nogroup:
diff --git a/lib/AST/Stmt.cpp b/lib/AST/Stmt.cpp
index 697cdc3fb360..69e65f558f89 100644
--- a/lib/AST/Stmt.cpp
+++ b/lib/AST/Stmt.cpp
@@ -1083,7 +1083,7 @@ CapturedStmt *CapturedStmt::CreateDeserialized(const ASTContext &Context,
 }
 
 Stmt::child_range CapturedStmt::children() {
-  // Children are captured field initilizers.
+  // Children are captured field initializers.
   return child_range(getStoredStmts(), getStoredStmts() + NumCaptures);
 }
 
diff --git a/lib/AST/StmtCXX.cpp b/lib/AST/StmtCXX.cpp
index 4a04fc211262..aade13ed3bd4 100644
--- a/lib/AST/StmtCXX.cpp
+++ b/lib/AST/StmtCXX.cpp
@@ -86,3 +86,30 @@ VarDecl *CXXForRangeStmt::getLoopVariable() {
 const VarDecl *CXXForRangeStmt::getLoopVariable() const {
   return const_cast<CXXForRangeStmt *>(this)->getLoopVariable();
 }
+
+CoroutineBodyStmt *CoroutineBodyStmt::Create(
+    const ASTContext &C, CoroutineBodyStmt::CtorArgs const& Args) {
+  std::size_t Size = totalSizeToAlloc<Stmt *>(
+      CoroutineBodyStmt::FirstParamMove + Args.ParamMoves.size());
+
+  void *Mem = C.Allocate(Size, alignof(CoroutineBodyStmt));
+  return new (Mem) CoroutineBodyStmt(Args);
+}
+
+CoroutineBodyStmt::CoroutineBodyStmt(CoroutineBodyStmt::CtorArgs const &Args)
+    : Stmt(CoroutineBodyStmtClass), NumParams(Args.ParamMoves.size()) {
+  Stmt **SubStmts = getStoredStmts();
+  SubStmts[CoroutineBodyStmt::Body] = Args.Body;
+  SubStmts[CoroutineBodyStmt::Promise] = Args.Promise;
+  SubStmts[CoroutineBodyStmt::InitSuspend] = Args.InitialSuspend;
+  SubStmts[CoroutineBodyStmt::FinalSuspend] = Args.FinalSuspend;
+  SubStmts[CoroutineBodyStmt::OnException] = Args.OnException;
+  SubStmts[CoroutineBodyStmt::OnFallthrough] = Args.OnFallthrough;
+  SubStmts[CoroutineBodyStmt::Allocate] = Args.Allocate;
+  SubStmts[CoroutineBodyStmt::Deallocate] = Args.Deallocate;
+  SubStmts[CoroutineBodyStmt::ReturnValue] = Args.ReturnValue;
+  SubStmts[CoroutineBodyStmt::ReturnStmtOnAllocFailure] =
+      Args.ReturnStmtOnAllocFailure;
+  std::copy(Args.ParamMoves.begin(), Args.ParamMoves.end(),
+            const_cast<Stmt **>(getParamMoves().data()));
+}
diff --git a/lib/AST/StmtOpenMP.cpp b/lib/AST/StmtOpenMP.cpp
index 880817a1339b..a812884cd927 100644
--- a/lib/AST/StmtOpenMP.cpp
+++ b/lib/AST/StmtOpenMP.cpp
@@ -149,6 +149,8 @@ OMPForDirective::Create(const ASTContext &C, SourceLocation StartLoc,
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -201,6 +203,8 @@ OMPForSimdDirective::Create(const ASTContext &C, SourceLocation StartLoc,
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -366,6 +370,8 @@ OMPParallelForDirective *OMPParallelForDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -417,6 +423,8 @@ OMPParallelForSimdDirective *OMPParallelForSimdDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -753,6 +761,8 @@ OMPTargetParallelForDirective *OMPTargetParallelForDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -896,6 +906,8 @@ OMPTaskLoopDirective *OMPTaskLoopDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -947,6 +959,8 @@ OMPTaskLoopSimdDirective *OMPTaskLoopSimdDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -997,6 +1011,8 @@ OMPDistributeDirective *OMPDistributeDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1071,6 +1087,8 @@ OMPDistributeParallelForDirective *OMPDistributeParallelForDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1127,6 +1145,8 @@ OMPDistributeParallelForSimdDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1182,6 +1202,8 @@ OMPDistributeSimdDirective *OMPDistributeSimdDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1236,6 +1258,8 @@ OMPTargetParallelForSimdDirective *OMPTargetParallelForSimdDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1330,6 +1354,8 @@ OMPTeamsDistributeDirective *OMPTeamsDistributeDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1383,6 +1409,8 @@ OMPTeamsDistributeSimdDirective *OMPTeamsDistributeSimdDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1438,6 +1466,8 @@ OMPTeamsDistributeParallelForSimdDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1496,6 +1526,8 @@ OMPTeamsDistributeParallelForDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1576,6 +1608,8 @@ OMPTargetTeamsDistributeDirective *OMPTargetTeamsDistributeDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1634,6 +1668,8 @@ OMPTargetTeamsDistributeParallelForDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1695,6 +1731,8 @@ OMPTargetTeamsDistributeParallelForSimdDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
@@ -1753,6 +1791,8 @@ OMPTargetTeamsDistributeSimdDirective::Create(
   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);
diff --git a/lib/AST/StmtPrinter.cpp b/lib/AST/StmtPrinter.cpp
index 1ba1aa40ec5c..21f5259c3ca8 100644
--- a/lib/AST/StmtPrinter.cpp
+++ b/lib/AST/StmtPrinter.cpp
@@ -2475,6 +2475,13 @@ void StmtPrinter::VisitCoawaitExpr(CoawaitExpr *S) {
   PrintExpr(S->getOperand());
 }
 
+
+void StmtPrinter::VisitDependentCoawaitExpr(DependentCoawaitExpr *S) {
+  OS << "co_await ";
+  PrintExpr(S->getOperand());
+}
+
+
 void StmtPrinter::VisitCoyieldExpr(CoyieldExpr *S) {
   OS << "co_yield ";
   PrintExpr(S->getOperand());
diff --git a/lib/AST/StmtProfile.cpp b/lib/AST/StmtProfile.cpp
index bcd2e96875e7..f1fbe2806b5d 100644
--- a/lib/AST/StmtProfile.cpp
+++ b/lib/AST/StmtProfile.cpp
@@ -19,20 +19,22 @@
 #include "clang/AST/ExprCXX.h"
 #include "clang/AST/ExprObjC.h"
 #include "clang/AST/ExprOpenMP.h"
+#include "clang/AST/ODRHash.h"
 #include "clang/AST/StmtVisitor.h"
 #include "llvm/ADT/FoldingSet.h"
 using namespace clang;
 
 namespace {
   class StmtProfiler : public ConstStmtVisitor<StmtProfiler> {
+  protected:
     llvm::FoldingSetNodeID &ID;
-    const ASTContext &Context;
     bool Canonical;
 
   public:
-    StmtProfiler(llvm::FoldingSetNodeID &ID, const ASTContext &Context,
-                 bool Canonical)
-      : ID(ID), Context(Context), Canonical(Canonical) { }
+    StmtProfiler(llvm::FoldingSetNodeID &ID, bool Canonical)
+        : ID(ID), Canonical(Canonical) {}
+
+    virtual ~StmtProfiler() {}
 
     void VisitStmt(const Stmt *S);
 
@@ -41,22 +43,25 @@ namespace {
 
     /// \brief Visit a declaration that is referenced within an expression
     /// or statement.
-    void VisitDecl(const Decl *D);
+    virtual void VisitDecl(const Decl *D) = 0;
 
     /// \brief Visit a type that is referenced within an expression or
     /// statement.
-    void VisitType(QualType T);
+    virtual void VisitType(QualType T) = 0;
 
     /// \brief Visit a name that occurs within an expression or statement.
-    void VisitName(DeclarationName Name);
+    virtual void VisitName(DeclarationName Name) = 0;
+
+    /// \brief Visit identifiers that are not in Decl's or Type's.
+    virtual void VisitIdentifierInfo(IdentifierInfo *II) = 0;
 
     /// \brief Visit a nested-name-specifier that occurs within an expression
     /// or statement.
-    void VisitNestedNameSpecifier(NestedNameSpecifier *NNS);
+    virtual void VisitNestedNameSpecifier(NestedNameSpecifier *NNS) = 0;
 
     /// \brief Visit a template name that occurs within an expression or
     /// statement.
-    void VisitTemplateName(TemplateName Name);
+    virtual void VisitTemplateName(TemplateName Name) = 0;
 
     /// \brief Visit template arguments that occur within an expression or
     /// statement.
@@ -66,6 +71,127 @@ namespace {
     /// \brief Visit a single template argument.
     void VisitTemplateArgument(const TemplateArgument &Arg);
   };
+
+  class StmtProfilerWithPointers : public StmtProfiler {
+    const ASTContext &Context;
+
+  public:
+    StmtProfilerWithPointers(llvm::FoldingSetNodeID &ID,
+                             const ASTContext &Context, bool Canonical)
+        : StmtProfiler(ID, Canonical), Context(Context) {}
+  private:
+    void VisitDecl(const Decl *D) override {
+      ID.AddInteger(D ? D->getKind() : 0);
+
+      if (Canonical && D) {
+        if (const NonTypeTemplateParmDecl *NTTP =
+                dyn_cast<NonTypeTemplateParmDecl>(D)) {
+          ID.AddInteger(NTTP->getDepth());
+          ID.AddInteger(NTTP->getIndex());
+          ID.AddBoolean(NTTP->isParameterPack());
+          VisitType(NTTP->getType());
+          return;
+        }
+
+        if (const ParmVarDecl *Parm = dyn_cast<ParmVarDecl>(D)) {
+          // The Itanium C++ ABI uses the type, scope depth, and scope
+          // index of a parameter when mangling expressions that involve
+          // function parameters, so we will use the parameter's type for
+          // establishing function parameter identity. That way, our
+          // definition of "equivalent" (per C++ [temp.over.link]) is at
+          // least as strong as the definition of "equivalent" used for
+          // name mangling.
+          VisitType(Parm->getType());
+          ID.AddInteger(Parm->getFunctionScopeDepth());
+          ID.AddInteger(Parm->getFunctionScopeIndex());
+          return;
+        }
+
+        if (const TemplateTypeParmDecl *TTP =
+                dyn_cast<TemplateTypeParmDecl>(D)) {
+          ID.AddInteger(TTP->getDepth());
+          ID.AddInteger(TTP->getIndex());
+          ID.AddBoolean(TTP->isParameterPack());
+          return;
+        }
+
+        if (const TemplateTemplateParmDecl *TTP =
+                dyn_cast<TemplateTemplateParmDecl>(D)) {
+          ID.AddInteger(TTP->getDepth());
+          ID.AddInteger(TTP->getIndex());
+          ID.AddBoolean(TTP->isParameterPack());
+          return;
+        }
+      }
+
+      ID.AddPointer(D ? D->getCanonicalDecl() : nullptr);
+    }
+
+    void VisitType(QualType T) override {
+      if (Canonical && !T.isNull())
+        T = Context.getCanonicalType(T);
+
+      ID.AddPointer(T.getAsOpaquePtr());
+    }
+
+    void VisitName(DeclarationName Name) override {
+      ID.AddPointer(Name.getAsOpaquePtr());
+    }
+
+    void VisitIdentifierInfo(IdentifierInfo *II) override {
+      ID.AddPointer(II);
+    }
+
+    void VisitNestedNameSpecifier(NestedNameSpecifier *NNS) override {
+      if (Canonical)
+        NNS = Context.getCanonicalNestedNameSpecifier(NNS);
+      ID.AddPointer(NNS);
+    }
+
+    void VisitTemplateName(TemplateName Name) override {
+      if (Canonical)
+        Name = Context.getCanonicalTemplateName(Name);
+
+      Name.Profile(ID);
+    }
+  };
+
+  class StmtProfilerWithoutPointers : public StmtProfiler {
+    ODRHash &Hash;
+  public:
+    StmtProfilerWithoutPointers(llvm::FoldingSetNodeID &ID, ODRHash &Hash)
+        : StmtProfiler(ID, false), Hash(Hash) {}
+
+  private:
+    void VisitType(QualType T) override {
+      Hash.AddQualType(T);
+    }
+
+    void VisitName(DeclarationName Name) override {
+      Hash.AddDeclarationName(Name);
+    }
+    void VisitIdentifierInfo(IdentifierInfo *II) override {
+      ID.AddBoolean(II);
+      if (II) {
+        Hash.AddIdentifierInfo(II);
+      }
+    }
+    void VisitDecl(const Decl *D) override {
+      ID.AddBoolean(D);
+      if (D) {
+        Hash.AddDecl(D);
+      }
+    }
+    void VisitTemplateName(TemplateName Name) override {
+      Hash.AddTemplateName(Name);
+    }
+    void VisitNestedNameSpecifier(NestedNameSpecifier *NNS) override {
+      ID.AddBoolean(NNS);
+      if (NNS) {
+        Hash.AddNestedNameSpecifier(NNS);
+      }
+    }
+  };
 }
 
 void StmtProfiler::VisitStmt(const Stmt *S) {
@@ -283,6 +409,7 @@ void OMPClauseProfiler::VistOMPClauseWithPostUpdate(
 }
 
 void OMPClauseProfiler::VisitOMPIfClause(const OMPIfClause *C) {
+  VistOMPClauseWithPreInit(C);
   if (C->getCondition())
     Profiler->VisitStmt(C->getCondition());
 }
@@ -293,6 +420,7 @@ void OMPClauseProfiler::VisitOMPFinalClause(const OMPFinalClause *C) {
 }
 
 void OMPClauseProfiler::VisitOMPNumThreadsClause(const OMPNumThreadsClause *C) {
+  VistOMPClauseWithPreInit(C);
   if (C->getNumThreads())
     Profiler->VisitStmt(C->getNumThreads());
 }
@@ -495,11 +623,13 @@ void OMPClauseProfiler::VisitOMPMapClause(const OMPMapClause *C) {
   VisitOMPClauseList(C);
 }
 void OMPClauseProfiler::VisitOMPNumTeamsClause(const OMPNumTeamsClause *C) {
+  VistOMPClauseWithPreInit(C);
   if (C->getNumTeams())
     Profiler->VisitStmt(C->getNumTeams());
 }
 void OMPClauseProfiler::VisitOMPThreadLimitClause(
     const OMPThreadLimitClause *C) {
+  VistOMPClauseWithPreInit(C);
   if (C->getThreadLimit())
     Profiler->VisitStmt(C->getThreadLimit());
 }
@@ -849,7 +979,7 @@ void StmtProfiler::VisitOffsetOfExpr(const OffsetOfExpr *S) {
       break;
 
     case OffsetOfNode::Identifier:
-      ID.AddPointer(ON.getFieldName());
+      VisitIdentifierInfo(ON.getFieldName());
       break;
 
     case OffsetOfNode::Base:
@@ -857,7 +987,7 @@ void StmtProfiler::VisitOffsetOfExpr(const OffsetOfExpr *S) {
       break;
     }
   }
-  
+
   VisitExpr(S);
 }
 
@@ -1447,7 +1577,7 @@ StmtProfiler::VisitCXXPseudoDestructorExpr(const CXXPseudoDestructorExpr *S) {
   if (S->getDestroyedTypeInfo())
     VisitType(S->getDestroyedType());
   else
-    ID.AddPointer(S->getDestroyedTypeIdentifier());
+    VisitIdentifierInfo(S->getDestroyedTypeIdentifier());
 }
 
 void StmtProfiler::VisitOverloadExpr(const OverloadExpr *S) {
@@ -1595,6 +1725,10 @@ void StmtProfiler::VisitCoawaitExpr(const CoawaitExpr *S) {
   VisitExpr(S);
 }
 
+void StmtProfiler::VisitDependentCoawaitExpr(const DependentCoawaitExpr *S) {
+  VisitExpr(S);
+}
+
 void StmtProfiler::VisitCoyieldExpr(const CoyieldExpr *S) {
   VisitExpr(S);
 }
@@ -1697,77 +1831,6 @@ void StmtProfiler::VisitObjCAvailabilityCheckExpr(
   VisitExpr(S);
 }
 
-void StmtProfiler::VisitDecl(const Decl *D) {
-  ID.AddInteger(D? D->getKind() : 0);
-
-  if (Canonical && D) {
-    if (const NonTypeTemplateParmDecl *NTTP =
-          dyn_cast<NonTypeTemplateParmDecl>(D)) {
-      ID.AddInteger(NTTP->getDepth());
-      ID.AddInteger(NTTP->getIndex());
-      ID.AddBoolean(NTTP->isParameterPack());
-      VisitType(NTTP->getType());
-      return;
-    }
-
-    if (const ParmVarDecl *Parm = dyn_cast<ParmVarDecl>(D)) {
-      // The Itanium C++ ABI uses the type, scope depth, and scope
-      // index of a parameter when mangling expressions that involve
-      // function parameters, so we will use the parameter's type for
-      // establishing function parameter identity. That way, our
-      // definition of "equivalent" (per C++ [temp.over.link]) is at
-      // least as strong as the definition of "equivalent" used for
-      // name mangling.
-      VisitType(Parm->getType());
-      ID.AddInteger(Parm->getFunctionScopeDepth());
-      ID.AddInteger(Parm->getFunctionScopeIndex());
-      return;
-    }
-
-    if (const TemplateTypeParmDecl *TTP =
-          dyn_cast<TemplateTypeParmDecl>(D)) {
-      ID.AddInteger(TTP->getDepth());
-      ID.AddInteger(TTP->getIndex());
-      ID.AddBoolean(TTP->isParameterPack());
-      return;
-    }
-
-    if (const TemplateTemplateParmDecl *TTP =
-          dyn_cast<TemplateTemplateParmDecl>(D)) {
-      ID.AddInteger(TTP->getDepth());
-      ID.AddInteger(TTP->getIndex());
-      ID.AddBoolean(TTP->isParameterPack());
-      return;
-    }
-  }
-
-  ID.AddPointer(D? D->getCanonicalDecl() : nullptr);
-}
-
-void StmtProfiler::VisitType(QualType T) {
-  if (Canonical)
-    T = Context.getCanonicalType(T);
-
-  ID.AddPointer(T.getAsOpaquePtr());
-}
-
-void StmtProfiler::VisitName(DeclarationName Name) {
-  ID.AddPointer(Name.getAsOpaquePtr());
-}
-
-void StmtProfiler::VisitNestedNameSpecifier(NestedNameSpecifier *NNS) {
-  if (Canonical)
-    NNS = Context.getCanonicalNestedNameSpecifier(NNS);
-  ID.AddPointer(NNS);
-}
-
-void StmtProfiler::VisitTemplateName(TemplateName Name) {
-  if (Canonical)
-    Name = Context.getCanonicalTemplateName(Name);
-
-  Name.Profile(ID);
-}
-
 void StmtProfiler::VisitTemplateArguments(const TemplateArgumentLoc *Args,
                                           unsigned NumArgs) {
   ID.AddInteger(NumArgs);
@@ -1817,6 +1880,12 @@ void StmtProfiler::VisitTemplateArgument(const TemplateArgument &Arg) {
 
 void Stmt::Profile(llvm::FoldingSetNodeID &ID, const ASTContext &Context,
                    bool Canonical) const {
-  StmtProfiler Profiler(ID, Context, Canonical);
+  StmtProfilerWithPointers Profiler(ID, Context, Canonical);
+  Profiler.Visit(this);
+}
+
+void Stmt::ProcessODRHash(llvm::FoldingSetNodeID &ID,
+                          class ODRHash &Hash) const {
+  StmtProfilerWithoutPointers Profiler(ID, Hash);
   Profiler.Visit(this);
 }
diff --git a/lib/AST/TemplateBase.cpp b/lib/AST/TemplateBase.cpp
index 099f939c7a75..e4998c37a4ef 100644
--- a/lib/AST/TemplateBase.cpp
+++ b/lib/AST/TemplateBase.cpp
@@ -453,10 +453,6 @@ LLVM_DUMP_METHOD void TemplateArgument::dump() const { dump(llvm::errs()); }
 // TemplateArgumentLoc Implementation
 //===----------------------------------------------------------------------===//
 
-TemplateArgumentLocInfo::TemplateArgumentLocInfo() {
-  memset((void*)this, 0, sizeof(TemplateArgumentLocInfo));
-}
-
 SourceRange TemplateArgumentLoc::getSourceRange() const {
   switch (Argument.getKind()) {
   case TemplateArgument::Expression:
diff --git a/lib/AST/Type.cpp b/lib/AST/Type.cpp
index 0d0cd2e305be..df26233b4796 100644
--- a/lib/AST/Type.cpp
+++ b/lib/AST/Type.cpp
@@ -1559,61 +1559,79 @@ TagDecl *Type::getAsTagDecl() const {
 }
 
 namespace {
-  class GetContainedAutoVisitor :
-    public TypeVisitor<GetContainedAutoVisitor, AutoType*> {
+  class GetContainedDeducedTypeVisitor :
+    public TypeVisitor<GetContainedDeducedTypeVisitor, Type*> {
+    bool Syntactic;
   public:
-    using TypeVisitor<GetContainedAutoVisitor, AutoType*>::Visit;
-    AutoType *Visit(QualType T) {
+    GetContainedDeducedTypeVisitor(bool Syntactic = false)
+        : Syntactic(Syntactic) {}
+
+    using TypeVisitor<GetContainedDeducedTypeVisitor, Type*>::Visit;
+    Type *Visit(QualType T) {
       if (T.isNull())
         return nullptr;
       return Visit(T.getTypePtr());
     }
 
-    // The 'auto' type itself.
-    AutoType *VisitAutoType(const AutoType *AT) {
-      return const_cast<AutoType*>(AT);
+    // The deduced type itself.
+    Type *VisitDeducedType(const DeducedType *AT) {
+      return const_cast<DeducedType*>(AT);
     }
 
     // Only these types can contain the desired 'auto' type.
-    AutoType *VisitPointerType(const PointerType *T) {
+    Type *VisitElaboratedType(const ElaboratedType *T) {
+      return Visit(T->getNamedType());
+    }
+    Type *VisitPointerType(const PointerType *T) {
       return Visit(T->getPointeeType());
     }
-    AutoType *VisitBlockPointerType(const BlockPointerType *T) {
+    Type *VisitBlockPointerType(const BlockPointerType *T) {
       return Visit(T->getPointeeType());
     }
-    AutoType *VisitReferenceType(const ReferenceType *T) {
+    Type *VisitReferenceType(const ReferenceType *T) {
       return Visit(T->getPointeeTypeAsWritten());
     }
-    AutoType *VisitMemberPointerType(const MemberPointerType *T) {
+    Type *VisitMemberPointerType(const MemberPointerType *T) {
       return Visit(T->getPointeeType());
     }
-    AutoType *VisitArrayType(const ArrayType *T) {
+    Type *VisitArrayType(const ArrayType *T) {
       return Visit(T->getElementType());
     }
-    AutoType *VisitDependentSizedExtVectorType(
+    Type *VisitDependentSizedExtVectorType(
       const DependentSizedExtVectorType *T) {
       return Visit(T->getElementType());
     }
-    AutoType *VisitVectorType(const VectorType *T) {
+    Type *VisitVectorType(const VectorType *T) {
       return Visit(T->getElementType());
     }
-    AutoType *VisitFunctionType(const FunctionType *T) {
+    Type *VisitFunctionProtoType(const FunctionProtoType *T) {
+      if (Syntactic && T->hasTrailingReturn())
+        return const_cast<FunctionProtoType*>(T);
+      return VisitFunctionType(T);
+    }
+    Type *VisitFunctionType(const FunctionType *T) {
       return Visit(T->getReturnType());
     }
-    AutoType *VisitParenType(const ParenType *T) {
+    Type *VisitParenType(const ParenType *T) {
       return Visit(T->getInnerType());
     }
-    AutoType *VisitAttributedType(const AttributedType *T) {
+    Type *VisitAttributedType(const AttributedType *T) {
       return Visit(T->getModifiedType());
     }
-    AutoType *VisitAdjustedType(const AdjustedType *T) {
+    Type *VisitAdjustedType(const AdjustedType *T) {
       return Visit(T->getOriginalType());
     }
   };
 }
 
-AutoType *Type::getContainedAutoType() const {
-  return GetContainedAutoVisitor().Visit(this);
+DeducedType *Type::getContainedDeducedType() const {
+  return cast_or_null<DeducedType>(
+      GetContainedDeducedTypeVisitor().Visit(this));
+}
+
+bool Type::hasAutoForTrailingReturnType() const {
+  return dyn_cast_or_null<FunctionType>(
+      GetContainedDeducedTypeVisitor(true).Visit(this));
 }
 
 bool Type::hasIntegerRepresentation() const {
@@ -2005,20 +2023,8 @@ bool QualType::isCXX98PODType(const ASTContext &Context) const {
   if ((*this)->isIncompleteType())
     return false;
 
-  if (Context.getLangOpts().ObjCAutoRefCount) {
-    switch (getObjCLifetime()) {
-    case Qualifiers::OCL_ExplicitNone:
-      return true;
-      
-    case Qualifiers::OCL_Strong:
-    case Qualifiers::OCL_Weak:
-    case Qualifiers::OCL_Autoreleasing:
-      return false;
-
-    case Qualifiers::OCL_None:
-      break;
-    }        
-  }
+  if (hasNonTrivialObjCLifetime())
+    return false;
   
   QualType CanonicalType = getTypePtr()->CanonicalType;
   switch (CanonicalType->getTypeClass()) {
@@ -2067,22 +2073,8 @@ bool QualType::isTrivialType(const ASTContext &Context) const {
   if ((*this)->isIncompleteType())
     return false;
   
-  if (Context.getLangOpts().ObjCAutoRefCount) {
-    switch (getObjCLifetime()) {
-    case Qualifiers::OCL_ExplicitNone:
-      return true;
-      
-    case Qualifiers::OCL_Strong:
-    case Qualifiers::OCL_Weak:
-    case Qualifiers::OCL_Autoreleasing:
-      return false;
-      
-    case Qualifiers::OCL_None:
-      if ((*this)->isObjCLifetimeType())
-        return false;
-      break;
-    }        
-  }
+  if (hasNonTrivialObjCLifetime())
+    return false;
   
   QualType CanonicalType = getTypePtr()->CanonicalType;
   if (CanonicalType->isDependentType())
@@ -2119,22 +2111,8 @@ bool QualType::isTriviallyCopyableType(const ASTContext &Context) const {
   if ((*this)->isArrayType())
     return Context.getBaseElementType(*this).isTriviallyCopyableType(Context);
 
-  if (Context.getLangOpts().ObjCAutoRefCount) {
-    switch (getObjCLifetime()) {
-    case Qualifiers::OCL_ExplicitNone:
-      return true;
-      
-    case Qualifiers::OCL_Strong:
-    case Qualifiers::OCL_Weak:
-    case Qualifiers::OCL_Autoreleasing:
-      return false;
-      
-    case Qualifiers::OCL_None:
-      if ((*this)->isObjCLifetimeType())
-        return false;
-      break;
-    }        
-  }
+  if (hasNonTrivialObjCLifetime())
+    return false;
 
   // C++11 [basic.types]p9
   //   Scalar types, trivially copyable class types, arrays of such types, and
@@ -2170,7 +2148,11 @@ bool QualType::isTriviallyCopyableType(const ASTContext &Context) const {
   return false;
 }
 
-
+bool QualType::isNonWeakInMRRWithObjCWeak(const ASTContext &Context) const {
+  return !Context.getLangOpts().ObjCAutoRefCount &&
+         Context.getLangOpts().ObjCWeak &&
+         getObjCLifetime() != Qualifiers::OCL_Weak;
+}
 
 bool Type::isLiteralType(const ASTContext &Ctx) const {
   if (isDependentType())
@@ -2280,20 +2262,8 @@ bool QualType::isCXX11PODType(const ASTContext &Context) const {
   if (ty->isDependentType())
     return false;
 
-  if (Context.getLangOpts().ObjCAutoRefCount) {
-    switch (getObjCLifetime()) {
-    case Qualifiers::OCL_ExplicitNone:
-      return true;
-      
-    case Qualifiers::OCL_Strong:
-    case Qualifiers::OCL_Weak:
-    case Qualifiers::OCL_Autoreleasing:
-      return false;
-
-    case Qualifiers::OCL_None:
-      break;
-    }        
-  }
+  if (hasNonTrivialObjCLifetime())
+    return false;
 
   // C++11 [basic.types]p9:
   //   Scalar types, POD classes, arrays of such types, and cv-qualified
@@ -2630,8 +2600,6 @@ StringRef BuiltinType::getName(const PrintingPolicy &Policy) const {
     return "clk_event_t";
   case OCLQueue:
     return "queue_t";
-  case OCLNDRange:
-    return "ndrange_t";
   case OCLReserveID:
     return "reserve_id_t";
   case OMPArraySection:
@@ -3365,6 +3333,7 @@ static CachedProperties computeCachedProperties(const Type *T) {
     return CachedProperties(ExternalLinkage, false);
 
   case Type::Auto:
+  case Type::DeducedTemplateSpecialization:
     // Give non-deduced 'auto' types external linkage. We should only see them
     // here in error recovery.
     return CachedProperties(ExternalLinkage, false);
@@ -3472,6 +3441,7 @@ static LinkageInfo computeLinkageInfo(const Type *T) {
     return LinkageInfo::external();
 
   case Type::Auto:
+  case Type::DeducedTemplateSpecialization:
     return LinkageInfo::external();
 
   case Type::Record:
@@ -3608,7 +3578,8 @@ bool Type::canHaveNullability() const {
 
   // auto is considered dependent when it isn't deduced.
   case Type::Auto:
-    return !cast<AutoType>(type.getTypePtr())->isDeduced();
+  case Type::DeducedTemplateSpecialization:
+    return !cast<DeducedType>(type.getTypePtr())->isDeduced();
 
   case Type::Builtin:
     switch (cast<BuiltinType>(type.getTypePtr())->getKind()) {
@@ -3640,7 +3611,6 @@ bool Type::canHaveNullability() const {
     case BuiltinType::OCLEvent:
     case BuiltinType::OCLClkEvent:
     case BuiltinType::OCLQueue:
-    case BuiltinType::OCLNDRange:
     case BuiltinType::OCLReserveID:
     case BuiltinType::BuiltinFn:
     case BuiltinType::NullPtr:
diff --git a/lib/AST/TypeLoc.cpp b/lib/AST/TypeLoc.cpp
index 7242858f21e6..c9a268655723 100644
--- a/lib/AST/TypeLoc.cpp
+++ b/lib/AST/TypeLoc.cpp
@@ -340,7 +340,6 @@ TypeSpecifierType BuiltinTypeLoc::getWrittenTypeSpec() const {
   case BuiltinType::OCLEvent:
   case BuiltinType::OCLClkEvent:
   case BuiltinType::OCLQueue:
-  case BuiltinType::OCLNDRange:
   case BuiltinType::OCLReserveID:
   case BuiltinType::BuiltinFn:
   case BuiltinType::OMPArraySection:
diff --git a/lib/AST/TypePrinter.cpp b/lib/AST/TypePrinter.cpp
index cccc90876321..5268a2901ad9 100644
--- a/lib/AST/TypePrinter.cpp
+++ b/lib/AST/TypePrinter.cpp
@@ -96,7 +96,7 @@ namespace {
 
     static bool canPrefixQualifiers(const Type *T, bool &NeedARCStrongQualifier);
     void spaceBeforePlaceHolder(raw_ostream &OS);
-    void printTypeSpec(const NamedDecl *D, raw_ostream &OS);
+    void printTypeSpec(NamedDecl *D, raw_ostream &OS);
 
     void printBefore(const Type *ty, Qualifiers qs, raw_ostream &OS);
     void printBefore(QualType T, raw_ostream &OS);
@@ -189,6 +189,7 @@ bool TypePrinter::canPrefixQualifiers(const Type *T,
     case Type::Elaborated:
     case Type::TemplateTypeParm:
     case Type::SubstTemplateTypeParmPack:
+    case Type::DeducedTemplateSpecialization:
     case Type::TemplateSpecialization:
     case Type::InjectedClassName:
     case Type::DependentName:
@@ -797,7 +798,14 @@ void TypePrinter::printFunctionNoProtoAfter(const FunctionNoProtoType *T,
   printAfter(T->getReturnType(), OS);
 }
 
-void TypePrinter::printTypeSpec(const NamedDecl *D, raw_ostream &OS) {
+void TypePrinter::printTypeSpec(NamedDecl *D, raw_ostream &OS) {
+
+  // Compute the full nested-name-specifier for this type.
+  // In C, this will always be empty except when the type
+  // being printed is anonymous within other Record.
+  if (!Policy.SuppressScope)
+    AppendScope(D->getDeclContext(), OS);
+
   IdentifierInfo *II = D->getIdentifier();
   OS << II->getName();
   spaceBeforePlaceHolder(OS);
@@ -888,6 +896,24 @@ void TypePrinter::printAutoAfter(const AutoType *T, raw_ostream &OS) {
     printAfter(T->getDeducedType(), OS);
 }
 
+void TypePrinter::printDeducedTemplateSpecializationBefore(
+    const DeducedTemplateSpecializationType *T, raw_ostream &OS) {
+  // If the type has been deduced, print the deduced type.
+  if (!T->getDeducedType().isNull()) {
+    printBefore(T->getDeducedType(), OS);
+  } else {
+    IncludeStrongLifetimeRAII Strong(Policy);
+    T->getTemplateName().print(OS, Policy);
+    spaceBeforePlaceHolder(OS);
+  }
+}
+void TypePrinter::printDeducedTemplateSpecializationAfter(
+    const DeducedTemplateSpecializationType *T, raw_ostream &OS) {
+  // If the type has been deduced, print the deduced type.
+  if (!T->getDeducedType().isNull())
+    printAfter(T->getDeducedType(), OS);
+}
+
 void TypePrinter::printAtomicBefore(const AtomicType *T, raw_ostream &OS) {
   IncludeStrongLifetimeRAII Strong(Policy);
 
@@ -1627,14 +1653,22 @@ void Qualifiers::print(raw_ostream &OS, const PrintingPolicy& Policy,
         OS << "__local";
         break;
       case LangAS::opencl_constant:
+      case LangAS::cuda_constant:
         OS << "__constant";
         break;
       case LangAS::opencl_generic:
         OS << "__generic";
         break;
+      case LangAS::cuda_device:
+        OS << "__device";
+        break;
+      case LangAS::cuda_shared:
+        OS << "__shared";
+        break;
       default:
+        assert(addrspace >= LangAS::Count);
         OS << "__attribute__((address_space(";
-        OS << addrspace;
+        OS << addrspace - LangAS::Count;
         OS << ")))";
     }
   }
diff --git a/lib/ASTMatchers/Dynamic/Registry.cpp b/lib/ASTMatchers/Dynamic/Registry.cpp
index d1cab80c1a53..9fa693038194 100644
--- a/lib/ASTMatchers/Dynamic/Registry.cpp
+++ b/lib/ASTMatchers/Dynamic/Registry.cpp
@@ -330,6 +330,7 @@ RegistryMaps::RegistryMaps() {
   REGISTER_MATCHER(isPublic);
   REGISTER_MATCHER(isPure);
   REGISTER_MATCHER(isSignedInteger);
+  REGISTER_MATCHER(isStaticStorageClass);
   REGISTER_MATCHER(isStruct);
   REGISTER_MATCHER(isTemplateInstantiation);
   REGISTER_MATCHER(isUnion);
@@ -359,9 +360,14 @@ RegistryMaps::RegistryMaps() {
   REGISTER_MATCHER(nullStmt);
   REGISTER_MATCHER(numSelectorArgs);
   REGISTER_MATCHER(ofClass);
+  REGISTER_MATCHER(objcCategoryDecl);
   REGISTER_MATCHER(objcInterfaceDecl);
+  REGISTER_MATCHER(objcIvarDecl);
   REGISTER_MATCHER(objcMessageExpr);
+  REGISTER_MATCHER(objcMethodDecl);
   REGISTER_MATCHER(objcObjectPointerType);
+  REGISTER_MATCHER(objcPropertyDecl);
+  REGISTER_MATCHER(objcProtocolDecl);
   REGISTER_MATCHER(on);
   REGISTER_MATCHER(onImplicitObjectArgument);
   REGISTER_MATCHER(opaqueValueExpr);
@@ -412,6 +418,7 @@ RegistryMaps::RegistryMaps() {
   REGISTER_MATCHER(typedefNameDecl);
   REGISTER_MATCHER(typedefType);
   REGISTER_MATCHER(typeAliasDecl);
+  REGISTER_MATCHER(typeAliasTemplateDecl);
   REGISTER_MATCHER(typeLoc);
   REGISTER_MATCHER(unaryExprOrTypeTraitExpr);
   REGISTER_MATCHER(unaryOperator);
diff --git a/lib/Analysis/BodyFarm.cpp b/lib/Analysis/BodyFarm.cpp
index 56c812c34c50..59127246105d 100644
--- a/lib/Analysis/BodyFarm.cpp
+++ b/lib/Analysis/BodyFarm.cpp
@@ -87,7 +87,7 @@ BinaryOperator *ASTMaker::makeAssignment(const Expr *LHS, const Expr *RHS,
                                          QualType Ty) {
  return new (C) BinaryOperator(const_cast<Expr*>(LHS), const_cast<Expr*>(RHS),
                                BO_Assign, Ty, VK_RValue,
-                               OK_Ordinary, SourceLocation(), false);
+                               OK_Ordinary, SourceLocation(), FPOptions());
 }
 
 BinaryOperator *ASTMaker::makeComparison(const Expr *LHS, const Expr *RHS,
@@ -99,7 +99,7 @@ BinaryOperator *ASTMaker::makeComparison(const Expr *LHS, const Expr *RHS,
                                 Op,
                                 C.getLogicalOperationType(),
                                 VK_RValue,
-                                OK_Ordinary, SourceLocation(), false);
+                                OK_Ordinary, SourceLocation(), FPOptions());
 }
 
 CompoundStmt *ASTMaker::makeCompound(ArrayRef<Stmt *> Stmts) {
diff --git a/lib/Analysis/CFG.cpp b/lib/Analysis/CFG.cpp
index d56e0e8fa1d0..2a2b3d73b5ca 100644
--- a/lib/Analysis/CFG.cpp
+++ b/lib/Analysis/CFG.cpp
@@ -1390,7 +1390,7 @@ LocalScope* CFGBuilder::addLocalScopeForVarDecl(VarDecl *VD,
 
   // Check if type is a C++ class with non-trivial destructor.
   if (const CXXRecordDecl *CD = QT->getAsCXXRecordDecl())
-    if (!CD->hasTrivialDestructor()) {
+    if (CD->hasDefinition() && !CD->hasTrivialDestructor()) {
       // Add the variable to scope
       Scope = createOrReuseLocalScope(Scope);
       Scope->addVar(VD);
diff --git a/lib/Analysis/CallGraph.cpp b/lib/Analysis/CallGraph.cpp
index 8c126b09d057..6d9530bf0c68 100644
--- a/lib/Analysis/CallGraph.cpp
+++ b/lib/Analysis/CallGraph.cpp
@@ -62,6 +62,7 @@ public:
   void VisitCallExpr(CallExpr *CE) {
     if (Decl *D = getDeclFromCall(CE))
       addCalledDecl(D);
+    VisitChildren(CE);
   }
 
   // Adds may-call edges for the ObjC message sends.
diff --git a/lib/Analysis/CloneDetection.cpp b/lib/Analysis/CloneDetection.cpp
index e761738214c6..5bbcbe4e5722 100644
--- a/lib/Analysis/CloneDetection.cpp
+++ b/lib/Analysis/CloneDetection.cpp
@@ -24,27 +24,27 @@
 
 using namespace clang;
 
-StmtSequence::StmtSequence(const CompoundStmt *Stmt, ASTContext &Context,
+StmtSequence::StmtSequence(const CompoundStmt *Stmt, const Decl *D,
                            unsigned StartIndex, unsigned EndIndex)
-    : S(Stmt), Context(&Context), StartIndex(StartIndex), EndIndex(EndIndex) {
+    : S(Stmt), D(D), StartIndex(StartIndex), EndIndex(EndIndex) {
   assert(Stmt && "Stmt must not be a nullptr");
   assert(StartIndex < EndIndex && "Given array should not be empty");
   assert(EndIndex <= Stmt->size() && "Given array too big for this Stmt");
 }
 
-StmtSequence::StmtSequence(const Stmt *Stmt, ASTContext &Context)
-    : S(Stmt), Context(&Context), StartIndex(0), EndIndex(0) {}
+StmtSequence::StmtSequence(const Stmt *Stmt, const Decl *D)
+    : S(Stmt), D(D), StartIndex(0), EndIndex(0) {}
 
 StmtSequence::StmtSequence()
-    : S(nullptr), Context(nullptr), StartIndex(0), EndIndex(0) {}
+    : S(nullptr), D(nullptr), StartIndex(0), EndIndex(0) {}
 
 bool StmtSequence::contains(const StmtSequence &Other) const {
-  // If both sequences reside in different translation units, they can never
-  // contain each other.
-  if (Context != Other.Context)
+  // If both sequences reside in different declarations, they can never contain
+  // each other.
+  if (D != Other.D)
     return false;
 
-  const SourceManager &SM = Context->getSourceManager();
+  const SourceManager &SM = getASTContext().getSourceManager();
 
   // Otherwise check if the start and end locations of the current sequence
   // surround the other sequence.
@@ -76,6 +76,11 @@ StmtSequence::iterator StmtSequence::end() const {
   return CS->body_begin() + EndIndex;
 }
 
+ASTContext &StmtSequence::getASTContext() const {
+  assert(D);
+  return D->getASTContext();
+}
+
 SourceLocation StmtSequence::getStartLoc() const {
   return front()->getLocStart();
 }
@@ -86,168 +91,8 @@ SourceRange StmtSequence::getSourceRange() const {
   return SourceRange(getStartLoc(), getEndLoc());
 }
 
-namespace {
-
-/// \brief Analyzes the pattern of the referenced variables in a statement.
-class VariablePattern {
-
-  /// \brief Describes an occurence of a variable reference in a statement.
-  struct VariableOccurence {
-    /// The index of the associated VarDecl in the Variables vector.
-    size_t KindID;
-    /// The statement in the code where the variable was referenced.
-    const Stmt *Mention;
-
-    VariableOccurence(size_t KindID, const Stmt *Mention)
-        : KindID(KindID), Mention(Mention) {}
-  };
-
-  /// All occurences of referenced variables in the order of appearance.
-  std::vector<VariableOccurence> Occurences;
-  /// List of referenced variables in the order of appearance.
-  /// Every item in this list is unique.
-  std::vector<const VarDecl *> Variables;
-
-  /// \brief Adds a new variable referenced to this pattern.
-  /// \param VarDecl The declaration of the variable that is referenced.
-  /// \param Mention The SourceRange where this variable is referenced.
-  void addVariableOccurence(const VarDecl *VarDecl, const Stmt *Mention) {
-    // First check if we already reference this variable
-    for (size_t KindIndex = 0; KindIndex < Variables.size(); ++KindIndex) {
-      if (Variables[KindIndex] == VarDecl) {
-        // If yes, add a new occurence that points to the existing entry in
-        // the Variables vector.
-        Occurences.emplace_back(KindIndex, Mention);
-        return;
-      }
-    }
-    // If this variable wasn't already referenced, add it to the list of
-    // referenced variables and add a occurence that points to this new entry.
-    Occurences.emplace_back(Variables.size(), Mention);
-    Variables.push_back(VarDecl);
-  }
-
-  /// \brief Adds each referenced variable from the given statement.
-  void addVariables(const Stmt *S) {
-    // Sometimes we get a nullptr (such as from IfStmts which often have nullptr
-    // children). We skip such statements as they don't reference any
-    // variables.
-    if (!S)
-      return;
-
-    // Check if S is a reference to a variable. If yes, add it to the pattern.
-    if (auto D = dyn_cast<DeclRefExpr>(S)) {
-      if (auto VD = dyn_cast<VarDecl>(D->getDecl()->getCanonicalDecl()))
-        addVariableOccurence(VD, D);
-    }
-
-    // Recursively check all children of the given statement.
-    for (const Stmt *Child : S->children()) {
-      addVariables(Child);
-    }
-  }
-
-public:
-  /// \brief Creates an VariablePattern object with information about the given
-  ///        StmtSequence.
-  VariablePattern(const StmtSequence &Sequence) {
-    for (const Stmt *S : Sequence)
-      addVariables(S);
-  }
-
-  /// \brief Counts the differences between this pattern and the given one.
-  /// \param Other The given VariablePattern to compare with.
-  /// \param FirstMismatch Output parameter that will be filled with information
-  ///        about the first difference between the two patterns. This parameter
-  ///        can be a nullptr, in which case it will be ignored.
-  /// \return Returns the number of differences between the pattern this object
-  ///         is following and the given VariablePattern.
-  ///
-  /// For example, the following statements all have the same pattern and this
-  /// function would return zero:
-  ///
-  ///   if (a < b) return a; return b;
-  ///   if (x < y) return x; return y;
-  ///   if (u2 < u1) return u2; return u1;
-  ///
-  /// But the following statement has a different pattern (note the changed
-  /// variables in the return statements) and would have two differences when
-  /// compared with one of the statements above.
-  ///
-  ///   if (a < b) return b; return a;
-  ///
-  /// This function should only be called if the related statements of the given
-  /// pattern and the statements of this objects are clones of each other.
-  unsigned countPatternDifferences(
-      const VariablePattern &Other,
-      CloneDetector::SuspiciousClonePair *FirstMismatch = nullptr) {
-    unsigned NumberOfDifferences = 0;
-
-    assert(Other.Occurences.size() == Occurences.size());
-    for (unsigned i = 0; i < Occurences.size(); ++i) {
-      auto ThisOccurence = Occurences[i];
-      auto OtherOccurence = Other.Occurences[i];
-      if (ThisOccurence.KindID == OtherOccurence.KindID)
-        continue;
-
-      ++NumberOfDifferences;
-
-      // If FirstMismatch is not a nullptr, we need to store information about
-      // the first difference between the two patterns.
-      if (FirstMismatch == nullptr)
-        continue;
-
-      // Only proceed if we just found the first difference as we only store
-      // information about the first difference.
-      if (NumberOfDifferences != 1)
-        continue;
-
-      const VarDecl *FirstSuggestion = nullptr;
-      // If there is a variable available in the list of referenced variables
-      // which wouldn't break the pattern if it is used in place of the
-      // current variable, we provide this variable as the suggested fix.
-      if (OtherOccurence.KindID < Variables.size())
-        FirstSuggestion = Variables[OtherOccurence.KindID];
-
-      // Store information about the first clone.
-      FirstMismatch->FirstCloneInfo =
-          CloneDetector::SuspiciousClonePair::SuspiciousCloneInfo(
-              Variables[ThisOccurence.KindID], ThisOccurence.Mention,
-              FirstSuggestion);
-
-      // Same as above but with the other clone. We do this for both clones as
-      // we don't know which clone is the one containing the unintended
-      // pattern error.
-      const VarDecl *SecondSuggestion = nullptr;
-      if (ThisOccurence.KindID < Other.Variables.size())
-        SecondSuggestion = Other.Variables[ThisOccurence.KindID];
-
-      // Store information about the second clone.
-      FirstMismatch->SecondCloneInfo =
-          CloneDetector::SuspiciousClonePair::SuspiciousCloneInfo(
-              Other.Variables[OtherOccurence.KindID], OtherOccurence.Mention,
-              SecondSuggestion);
-
-      // SuspiciousClonePair guarantees that the first clone always has a
-      // suggested variable associated with it. As we know that one of the two
-      // clones in the pair always has suggestion, we swap the two clones
-      // in case the first clone has no suggested variable which means that
-      // the second clone has a suggested variable and should be first.
-      if (!FirstMismatch->FirstCloneInfo.Suggestion)
-        std::swap(FirstMismatch->FirstCloneInfo,
-                  FirstMismatch->SecondCloneInfo);
-
-      // This ensures that we always have at least one suggestion in a pair.
-      assert(FirstMismatch->FirstCloneInfo.Suggestion);
-    }
-
-    return NumberOfDifferences;
-  }
-};
-}
-
-/// \brief Prints the macro name that contains the given SourceLocation into
-///        the given raw_string_ostream.
+/// Prints the macro name that contains the given SourceLocation into the given
+/// raw_string_ostream.
 static void printMacroName(llvm::raw_string_ostream &MacroStack,
                            ASTContext &Context, SourceLocation Loc) {
   MacroStack << Lexer::getImmediateMacroName(Loc, Context.getSourceManager(),
@@ -258,8 +103,8 @@ static void printMacroName(llvm::raw_string_ostream &MacroStack,
   MacroStack << " ";
 }
 
-/// \brief Returns a string that represents all macro expansions that
-///        expanded into the given SourceLocation.
+/// Returns a string that represents all macro expansions that expanded into the
+/// given SourceLocation.
 ///
 /// If 'getMacroStack(A) == getMacroStack(B)' is true, then the SourceLocations
 /// A and B are expanded from the same macros in the same order.
@@ -279,7 +124,9 @@ static std::string getMacroStack(SourceLocation Loc, ASTContext &Context) {
 }
 
 namespace {
-/// \brief Collects the data of a single Stmt.
+typedef unsigned DataPiece;
+
+/// Collects the data of a single Stmt.
 ///
 /// This class defines what a code clone is: If it collects for two statements
 /// the same data, then those two statements are considered to be clones of each
@@ -292,11 +139,11 @@ template <typename T>
 class StmtDataCollector : public ConstStmtVisitor<StmtDataCollector<T>> {
 
   ASTContext &Context;
-  /// \brief The data sink to which all data is forwarded.
+  /// The data sink to which all data is forwarded.
   T &DataConsumer;
 
 public:
-  /// \brief Collects data of the given Stmt.
+  /// Collects data of the given Stmt.
   /// \param S The given statement.
   /// \param Context The ASTContext of S.
   /// \param DataConsumer The data sink to which all data is forwarded.
@@ -307,7 +154,7 @@ public:
 
   // Below are utility methods for appending different data to the vector.
 
-  void addData(CloneDetector::DataPiece Integer) {
+  void addData(DataPiece Integer) {
     DataConsumer.update(
         StringRef(reinterpret_cast<char *>(&Integer), sizeof(Integer)));
   }
@@ -425,7 +272,7 @@ public:
   })
   DEF_ADD_DATA(DeclStmt, {
     auto numDecls = std::distance(S->decl_begin(), S->decl_end());
-    addData(static_cast<CloneDetector::DataPiece>(numDecls));
+    addData(static_cast<DataPiece>(numDecls));
     for (const Decl *D : S->decls()) {
       if (const VarDecl *VD = dyn_cast<VarDecl>(D)) {
         addData(VD->getType());
@@ -454,199 +301,131 @@ public:
 };
 } // end anonymous namespace
 
-namespace {
-/// Generates CloneSignatures for a set of statements and stores the results in
-/// a CloneDetector object.
-class CloneSignatureGenerator {
+void CloneDetector::analyzeCodeBody(const Decl *D) {
+  assert(D);
+  assert(D->hasBody());
 
-  CloneDetector &CD;
-  ASTContext &Context;
+  Sequences.push_back(StmtSequence(D->getBody(), D));
+}
 
-  /// \brief Generates CloneSignatures for all statements in the given statement
-  /// tree and stores them in the CloneDetector.
-  ///
-  /// \param S The root of the given statement tree.
-  /// \param ParentMacroStack A string representing the macros that generated
-  ///                         the parent statement or an empty string if no
-  ///                         macros generated the parent statement.
-  ///                         See getMacroStack() for generating such a string.
-  /// \return The CloneSignature of the root statement.
-  CloneDetector::CloneSignature
-  generateSignatures(const Stmt *S, const std::string &ParentMacroStack) {
-    // Create an empty signature that will be filled in this method.
-    CloneDetector::CloneSignature Signature;
-
-    llvm::MD5 Hash;
-
-    // Collect all relevant data from S and hash it.
-    StmtDataCollector<llvm::MD5>(S, Context, Hash);
-
-    // Look up what macros expanded into the current statement.
-    std::string StartMacroStack = getMacroStack(S->getLocStart(), Context);
-    std::string EndMacroStack = getMacroStack(S->getLocEnd(), Context);
-
-    // First, check if ParentMacroStack is not empty which means we are currently
-    // dealing with a parent statement which was expanded from a macro.
-    // If this parent statement was expanded from the same macros as this
-    // statement, we reduce the initial complexity of this statement to zero.
-    // This causes that a group of statements that were generated by a single
-    // macro expansion will only increase the total complexity by one.
-    // Note: This is not the final complexity of this statement as we still
-    // add the complexity of the child statements to the complexity value.
-    if (!ParentMacroStack.empty() && (StartMacroStack == ParentMacroStack &&
-                                      EndMacroStack == ParentMacroStack)) {
-      Signature.Complexity = 0;
-    }
+/// Returns true if and only if \p Stmt contains at least one other
+/// sequence in the \p Group.
+static bool containsAnyInGroup(StmtSequence &Seq,
+                               CloneDetector::CloneGroup &Group) {
+  for (StmtSequence &GroupSeq : Group) {
+    if (Seq.contains(GroupSeq))
+      return true;
+  }
+  return false;
+}
 
-    // Storage for the signatures of the direct child statements. This is only
-    // needed if the current statement is a CompoundStmt.
-    std::vector<CloneDetector::CloneSignature> ChildSignatures;
-    const CompoundStmt *CS = dyn_cast<const CompoundStmt>(S);
+/// Returns true if and only if all sequences in \p OtherGroup are
+/// contained by a sequence in \p Group.
+static bool containsGroup(CloneDetector::CloneGroup &Group,
+                          CloneDetector::CloneGroup &OtherGroup) {
+  // We have less sequences in the current group than we have in the other,
+  // so we will never fulfill the requirement for returning true. This is only
+  // possible because we know that a sequence in Group can contain at most
+  // one sequence in OtherGroup.
+  if (Group.size() < OtherGroup.size())
+    return false;
 
-    // The signature of a statement includes the signatures of its children.
-    // Therefore we create the signatures for every child and add them to the
-    // current signature.
-    for (const Stmt *Child : S->children()) {
-      // Some statements like 'if' can have nullptr children that we will skip.
-      if (!Child)
-        continue;
+  for (StmtSequence &Stmt : Group) {
+    if (!containsAnyInGroup(Stmt, OtherGroup))
+      return false;
+  }
+  return true;
+}
 
-      // Recursive call to create the signature of the child statement. This
-      // will also create and store all clone groups in this child statement.
-      // We pass only the StartMacroStack along to keep things simple.
-      auto ChildSignature = generateSignatures(Child, StartMacroStack);
+void OnlyLargestCloneConstraint::constrain(
+    std::vector<CloneDetector::CloneGroup> &Result) {
+  std::vector<unsigned> IndexesToRemove;
 
-      // Add the collected data to the signature of the current statement.
-      Signature.Complexity += ChildSignature.Complexity;
-      Hash.update(StringRef(reinterpret_cast<char *>(&ChildSignature.Hash),
-                            sizeof(ChildSignature.Hash)));
+  // Compare every group in the result with the rest. If one groups contains
+  // another group, we only need to return the bigger group.
+  // Note: This doesn't scale well, so if possible avoid calling any heavy
+  // function from this loop to minimize the performance impact.
+  for (unsigned i = 0; i < Result.size(); ++i) {
+    for (unsigned j = 0; j < Result.size(); ++j) {
+      // Don't compare a group with itself.
+      if (i == j)
+        continue;
 
-      // If the current statement is a CompoundStatement, we need to store the
-      // signature for the generation of the sub-sequences.
-      if (CS)
-        ChildSignatures.push_back(ChildSignature);
+      if (containsGroup(Result[j], Result[i])) {
+        IndexesToRemove.push_back(i);
+        break;
+      }
     }
+  }
 
-    // If the current statement is a CompoundStmt, we also need to create the
-    // clone groups from the sub-sequences inside the children.
-    if (CS)
-      handleSubSequences(CS, ChildSignatures);
+  // Erasing a list of indexes from the vector should be done with decreasing
+  // indexes. As IndexesToRemove is constructed with increasing values, we just
+  // reverse iterate over it to get the desired order.
+  for (auto I = IndexesToRemove.rbegin(); I != IndexesToRemove.rend(); ++I) {
+    Result.erase(Result.begin() + *I);
+  }
+}
 
-    // Create the final hash code for the current signature.
-    llvm::MD5::MD5Result HashResult;
-    Hash.final(HashResult);
+static size_t createHash(llvm::MD5 &Hash) {
+  size_t HashCode;
 
-    // Copy as much of the generated hash code to the signature's hash code.
-    std::memcpy(&Signature.Hash, &HashResult,
-                std::min(sizeof(Signature.Hash), sizeof(HashResult)));
+  // Create the final hash code for the current Stmt.
+  llvm::MD5::MD5Result HashResult;
+  Hash.final(HashResult);
 
-    // Save the signature for the current statement in the CloneDetector object.
-    CD.add(StmtSequence(S, Context), Signature);
+  // Copy as much as possible of the generated hash code to the Stmt's hash
+  // code.
+  std::memcpy(&HashCode, &HashResult,
+              std::min(sizeof(HashCode), sizeof(HashResult)));
 
-    return Signature;
-  }
+  return HashCode;
+}
 
-  /// \brief Adds all possible sub-sequences in the child array of the given
-  ///        CompoundStmt to the CloneDetector.
-  /// \param CS The given CompoundStmt.
-  /// \param ChildSignatures A list of calculated signatures for each child in
-  ///                        the given CompoundStmt.
-  void handleSubSequences(
-      const CompoundStmt *CS,
-      const std::vector<CloneDetector::CloneSignature> &ChildSignatures) {
+size_t RecursiveCloneTypeIIConstraint::saveHash(
+    const Stmt *S, const Decl *D,
+    std::vector<std::pair<size_t, StmtSequence>> &StmtsByHash) {
+  llvm::MD5 Hash;
+  ASTContext &Context = D->getASTContext();
 
-    // FIXME: This function has quadratic runtime right now. Check if skipping
-    // this function for too long CompoundStmts is an option.
+  StmtDataCollector<llvm::MD5>(S, Context, Hash);
 
-    // The length of the sub-sequence. We don't need to handle sequences with
-    // the length 1 as they are already handled in CollectData().
+  auto CS = dyn_cast<CompoundStmt>(S);
+  SmallVector<size_t, 8> ChildHashes;
+
+  for (const Stmt *Child : S->children()) {
+    if (Child == nullptr) {
+      ChildHashes.push_back(0);
+      continue;
+    }
+    size_t ChildHash = saveHash(Child, D, StmtsByHash);
+    Hash.update(
+        StringRef(reinterpret_cast<char *>(&ChildHash), sizeof(ChildHash)));
+    ChildHashes.push_back(ChildHash);
+  }
+
+  if (CS) {
     for (unsigned Length = 2; Length <= CS->size(); ++Length) {
-      // The start index in the body of the CompoundStmt. We increase the
-      // position until the end of the sub-sequence reaches the end of the
-      // CompoundStmt body.
       for (unsigned Pos = 0; Pos <= CS->size() - Length; ++Pos) {
-        // Create an empty signature and add the signatures of all selected
-        // child statements to it.
-        CloneDetector::CloneSignature SubSignature;
-        llvm::MD5 SubHash;
-
+        llvm::MD5 Hash;
         for (unsigned i = Pos; i < Pos + Length; ++i) {
-          SubSignature.Complexity += ChildSignatures[i].Complexity;
-          size_t ChildHash = ChildSignatures[i].Hash;
-
-          SubHash.update(StringRef(reinterpret_cast<char *>(&ChildHash),
+          size_t ChildHash = ChildHashes[i];
+          Hash.update(StringRef(reinterpret_cast<char *>(&ChildHash),
                                 sizeof(ChildHash)));
         }
-
-        // Create the final hash code for the current signature.
-        llvm::MD5::MD5Result HashResult;
-        SubHash.final(HashResult);
-
-        // Copy as much of the generated hash code to the signature's hash code.
-        std::memcpy(&SubSignature.Hash, &HashResult,
-                    std::min(sizeof(SubSignature.Hash), sizeof(HashResult)));
-
-        // Save the signature together with the information about what children
-        // sequence we selected.
-        CD.add(StmtSequence(CS, Context, Pos, Pos + Length), SubSignature);
+        StmtsByHash.push_back(std::make_pair(
+            createHash(Hash), StmtSequence(CS, D, Pos, Pos + Length)));
       }
     }
   }
 
-public:
-  explicit CloneSignatureGenerator(CloneDetector &CD, ASTContext &Context)
-      : CD(CD), Context(Context) {}
-
-  /// \brief Generates signatures for all statements in the given function body.
-  void consumeCodeBody(const Stmt *S) { generateSignatures(S, ""); }
-};
-} // end anonymous namespace
-
-void CloneDetector::analyzeCodeBody(const Decl *D) {
-  assert(D);
-  assert(D->hasBody());
-  CloneSignatureGenerator Generator(*this, D->getASTContext());
-  Generator.consumeCodeBody(D->getBody());
-}
-
-void CloneDetector::add(const StmtSequence &S,
-                        const CloneSignature &Signature) {
-  Sequences.push_back(std::make_pair(Signature, S));
+  size_t HashCode = createHash(Hash);
+  StmtsByHash.push_back(std::make_pair(HashCode, StmtSequence(S, D)));
+  return HashCode;
 }
 
 namespace {
-/// \brief Returns true if and only if \p Stmt contains at least one other
-/// sequence in the \p Group.
-bool containsAnyInGroup(StmtSequence &Stmt, CloneDetector::CloneGroup &Group) {
-  for (StmtSequence &GroupStmt : Group.Sequences) {
-    if (Stmt.contains(GroupStmt))
-      return true;
-  }
-  return false;
-}
-
-/// \brief Returns true if and only if all sequences in \p OtherGroup are
-/// contained by a sequence in \p Group.
-bool containsGroup(CloneDetector::CloneGroup &Group,
-                   CloneDetector::CloneGroup &OtherGroup) {
-  // We have less sequences in the current group than we have in the other,
-  // so we will never fulfill the requirement for returning true. This is only
-  // possible because we know that a sequence in Group can contain at most
-  // one sequence in OtherGroup.
-  if (Group.Sequences.size() < OtherGroup.Sequences.size())
-    return false;
-
-  for (StmtSequence &Stmt : Group.Sequences) {
-    if (!containsAnyInGroup(Stmt, OtherGroup))
-      return false;
-  }
-  return true;
-}
-} // end anonymous namespace
-
-namespace {
-/// \brief Wrapper around FoldingSetNodeID that it can be used as the template
-///        argument of the StmtDataCollector.
+/// Wrapper around FoldingSetNodeID that it can be used as the template
+/// argument of the StmtDataCollector.
 class FoldingSetNodeIDWrapper {
 
   llvm::FoldingSetNodeID &FS;
@@ -658,8 +437,8 @@ public:
 };
 } // end anonymous namespace
 
-/// \brief Writes the relevant data from all statements and child statements
-///        in the given StmtSequence into the given FoldingSetNodeID.
+/// Writes the relevant data from all statements and child statements
+/// in the given StmtSequence into the given FoldingSetNodeID.
 static void CollectStmtSequenceData(const StmtSequence &Sequence,
                                     FoldingSetNodeIDWrapper &OutputData) {
   for (const Stmt *S : Sequence) {
@@ -670,13 +449,13 @@ static void CollectStmtSequenceData(const StmtSequence &Sequence,
       if (!Child)
         continue;
 
-      CollectStmtSequenceData(StmtSequence(Child, Sequence.getASTContext()),
+      CollectStmtSequenceData(StmtSequence(Child, Sequence.getContainingDecl()),
                               OutputData);
     }
   }
 }
 
-/// \brief Returns true if both sequences are clones of each other.
+/// Returns true if both sequences are clones of each other.
 static bool areSequencesClones(const StmtSequence &LHS,
                                const StmtSequence &RHS) {
   // We collect the data from all statements in the sequence as we did before
@@ -693,202 +472,272 @@ static bool areSequencesClones(const StmtSequence &LHS,
   return DataLHS == DataRHS;
 }
 
-/// \brief Finds all actual clone groups in a single group of presumed clones.
-/// \param Result Output parameter to which all found groups are added.
-/// \param Group A group of presumed clones. The clones are allowed to have a
-///              different variable pattern and may not be actual clones of each
-///              other.
-/// \param CheckVariablePattern If true, every clone in a group that was added
-///              to the output follows the same variable pattern as the other
-///              clones in its group.
-static void createCloneGroups(std::vector<CloneDetector::CloneGroup> &Result,
-                              const CloneDetector::CloneGroup &Group,
-                              bool CheckVariablePattern) {
-  // We remove the Sequences one by one, so a list is more appropriate.
-  std::list<StmtSequence> UnassignedSequences(Group.Sequences.begin(),
-                                              Group.Sequences.end());
-
-  // Search for clones as long as there could be clones in UnassignedSequences.
-  while (UnassignedSequences.size() > 1) {
-
-    // Pick the first Sequence as a protoype for a new clone group.
-    StmtSequence Prototype = UnassignedSequences.front();
-    UnassignedSequences.pop_front();
-
-    CloneDetector::CloneGroup FilteredGroup(Prototype, Group.Signature);
-
-    // Analyze the variable pattern of the prototype. Every other StmtSequence
-    // needs to have the same pattern to get into the new clone group.
-    VariablePattern PrototypeFeatures(Prototype);
-
-    // Search all remaining StmtSequences for an identical variable pattern
-    // and assign them to our new clone group.
-    auto I = UnassignedSequences.begin(), E = UnassignedSequences.end();
-    while (I != E) {
-      // If the sequence doesn't fit to the prototype, we have encountered
-      // an unintended hash code collision and we skip it.
-      if (!areSequencesClones(Prototype, *I)) {
-        ++I;
-        continue;
-      }
+void RecursiveCloneTypeIIConstraint::constrain(
+    std::vector<CloneDetector::CloneGroup> &Sequences) {
+  // FIXME: Maybe we can do this in-place and don't need this additional vector.
+  std::vector<CloneDetector::CloneGroup> Result;
 
-      // If we weren't asked to check for a matching variable pattern in clone
-      // groups we can add the sequence now to the new clone group.
-      // If we were asked to check for matching variable pattern, we first have
-      // to check that there are no differences between the two patterns and
-      // only proceed if they match.
-      if (!CheckVariablePattern ||
-          VariablePattern(*I).countPatternDifferences(PrototypeFeatures) == 0) {
-        FilteredGroup.Sequences.push_back(*I);
-        I = UnassignedSequences.erase(I);
-        continue;
-      }
+  for (CloneDetector::CloneGroup &Group : Sequences) {
+    // We assume in the following code that the Group is non-empty, so we
+    // skip all empty groups.
+    if (Group.empty())
+      continue;
+
+    std::vector<std::pair<size_t, StmtSequence>> StmtsByHash;
 
-      // We didn't found a matching variable pattern, so we continue with the
-      // next sequence.
-      ++I;
+    // Generate hash codes for all children of S and save them in StmtsByHash.
+    for (const StmtSequence &S : Group) {
+      saveHash(S.front(), S.getContainingDecl(), StmtsByHash);
     }
 
-    // Add a valid clone group to the list of found clone groups.
-    if (!FilteredGroup.isValid())
-      continue;
+    // Sort hash_codes in StmtsByHash.
+    std::stable_sort(StmtsByHash.begin(), StmtsByHash.end(),
+                     [](std::pair<size_t, StmtSequence> LHS,
+                            std::pair<size_t, StmtSequence> RHS) {
+                       return LHS.first < RHS.first;
+                     });
+
+    // Check for each StmtSequence if its successor has the same hash value.
+    // We don't check the last StmtSequence as it has no successor.
+    // Note: The 'size - 1 ' in the condition is safe because we check for an
+    // empty Group vector at the beginning of this function.
+    for (unsigned i = 0; i < StmtsByHash.size() - 1; ++i) {
+      const auto Current = StmtsByHash[i];
+
+      // It's likely that we just found an sequence of StmtSequences that
+      // represent a CloneGroup, so we create a new group and start checking and
+      // adding the StmtSequences in this sequence.
+      CloneDetector::CloneGroup NewGroup;
+
+      size_t PrototypeHash = Current.first;
+
+      for (; i < StmtsByHash.size(); ++i) {
+        // A different hash value means we have reached the end of the sequence.
+        if (PrototypeHash != StmtsByHash[i].first ||
+            !areSequencesClones(StmtsByHash[i].second, Current.second)) {
+          // The current sequence could be the start of a new CloneGroup. So we
+          // decrement i so that we visit it again in the outer loop.
+          // Note: i can never be 0 at this point because we are just comparing
+          // the hash of the Current StmtSequence with itself in the 'if' above.
+          assert(i != 0);
+          --i;
+          break;
+        }
+        // Same hash value means we should add the StmtSequence to the current
+        // group.
+        NewGroup.push_back(StmtsByHash[i].second);
+      }
 
-    Result.push_back(FilteredGroup);
+      // We created a new clone group with matching hash codes and move it to
+      // the result vector.
+      Result.push_back(NewGroup);
+    }
   }
+  // Sequences is the output parameter, so we copy our result into it.
+  Sequences = Result;
 }
 
-void CloneDetector::findClones(std::vector<CloneGroup> &Result,
-                               unsigned MinGroupComplexity,
-                               bool CheckPatterns) {
-  // A shortcut (and necessary for the for-loop later in this function).
-  if (Sequences.empty())
-    return;
+size_t MinComplexityConstraint::calculateStmtComplexity(
+    const StmtSequence &Seq, const std::string &ParentMacroStack) {
+  if (Seq.empty())
+    return 0;
+
+  size_t Complexity = 1;
+
+  ASTContext &Context = Seq.getASTContext();
+
+  // Look up what macros expanded into the current statement.
+  std::string StartMacroStack = getMacroStack(Seq.getStartLoc(), Context);
+  std::string EndMacroStack = getMacroStack(Seq.getEndLoc(), Context);
+
+  // First, check if ParentMacroStack is not empty which means we are currently
+  // dealing with a parent statement which was expanded from a macro.
+  // If this parent statement was expanded from the same macros as this
+  // statement, we reduce the initial complexity of this statement to zero.
+  // This causes that a group of statements that were generated by a single
+  // macro expansion will only increase the total complexity by one.
+  // Note: This is not the final complexity of this statement as we still
+  // add the complexity of the child statements to the complexity value.
+  if (!ParentMacroStack.empty() && (StartMacroStack == ParentMacroStack &&
+                                    EndMacroStack == ParentMacroStack)) {
+    Complexity = 0;
+  }
 
-  // We need to search for groups of StmtSequences with the same hash code to
-  // create our initial clone groups. By sorting all known StmtSequences by
-  // their hash value we make sure that StmtSequences with the same hash code
-  // are grouped together in the Sequences vector.
-  // Note: We stable sort here because the StmtSequences are added in the order
-  // in which they appear in the source file. We want to preserve that order
-  // because we also want to report them in that order in the CloneChecker.
-  std::stable_sort(Sequences.begin(), Sequences.end(),
-                   [](std::pair<CloneSignature, StmtSequence> LHS,
-                      std::pair<CloneSignature, StmtSequence> RHS) {
-                     return LHS.first.Hash < RHS.first.Hash;
-                   });
-
-  std::vector<CloneGroup> CloneGroups;
-
-  // Check for each CloneSignature if its successor has the same hash value.
-  // We don't check the last CloneSignature as it has no successor.
-  // Note: The 'size - 1' in the condition is safe because we check for an empty
-  // Sequences vector at the beginning of this function.
-  for (unsigned i = 0; i < Sequences.size() - 1; ++i) {
-    const auto Current = Sequences[i];
-    const auto Next = Sequences[i + 1];
-
-    if (Current.first.Hash != Next.first.Hash)
-      continue;
+  // Iterate over the Stmts in the StmtSequence and add their complexity values
+  // to the current complexity value.
+  if (Seq.holdsSequence()) {
+    for (const Stmt *S : Seq) {
+      Complexity += calculateStmtComplexity(
+          StmtSequence(S, Seq.getContainingDecl()), StartMacroStack);
+    }
+  } else {
+    for (const Stmt *S : Seq.front()->children()) {
+      Complexity += calculateStmtComplexity(
+          StmtSequence(S, Seq.getContainingDecl()), StartMacroStack);
+    }
+  }
+  return Complexity;
+}
 
-    // It's likely that we just found an sequence of CloneSignatures that
-    // represent a CloneGroup, so we create a new group and start checking and
-    // adding the CloneSignatures in this sequence.
-    CloneGroup Group;
-    Group.Signature = Current.first;
-
-    for (; i < Sequences.size(); ++i) {
-      const auto &Signature = Sequences[i];
-
-      // A different hash value means we have reached the end of the sequence.
-      if (Current.first.Hash != Signature.first.Hash) {
-        // The current Signature could be the start of a new CloneGroup. So we
-        // decrement i so that we visit it again in the outer loop.
-        // Note: i can never be 0 at this point because we are just comparing
-        // the hash of the Current CloneSignature with itself in the 'if' above.
-        assert(i != 0);
-        --i;
-        break;
-      }
+void MatchingVariablePatternConstraint::constrain(
+    std::vector<CloneDetector::CloneGroup> &CloneGroups) {
+  CloneConstraint::splitCloneGroups(
+      CloneGroups, [](const StmtSequence &A, const StmtSequence &B) {
+        VariablePattern PatternA(A);
+        VariablePattern PatternB(B);
+        return PatternA.countPatternDifferences(PatternB) == 0;
+      });
+}
 
-      // Skip CloneSignatures that won't pass the complexity requirement.
-      if (Signature.first.Complexity < MinGroupComplexity)
+void CloneConstraint::splitCloneGroups(
+    std::vector<CloneDetector::CloneGroup> &CloneGroups,
+    std::function<bool(const StmtSequence &, const StmtSequence &)> Compare) {
+  std::vector<CloneDetector::CloneGroup> Result;
+  for (auto &HashGroup : CloneGroups) {
+    // Contains all indexes in HashGroup that were already added to a
+    // CloneGroup.
+    std::vector<char> Indexes;
+    Indexes.resize(HashGroup.size());
+
+    for (unsigned i = 0; i < HashGroup.size(); ++i) {
+      // Skip indexes that are already part of a CloneGroup.
+      if (Indexes[i])
         continue;
 
-      Group.Sequences.push_back(Signature.second);
-    }
+      // Pick the first unhandled StmtSequence and consider it as the
+      // beginning
+      // of a new CloneGroup for now.
+      // We don't add i to Indexes because we never iterate back.
+      StmtSequence Prototype = HashGroup[i];
+      CloneDetector::CloneGroup PotentialGroup = {Prototype};
+      ++Indexes[i];
+
+      // Check all following StmtSequences for clones.
+      for (unsigned j = i + 1; j < HashGroup.size(); ++j) {
+        // Skip indexes that are already part of a CloneGroup.
+        if (Indexes[j])
+          continue;
+
+        // If a following StmtSequence belongs to our CloneGroup, we add it to
+        // it.
+        const StmtSequence &Candidate = HashGroup[j];
+
+        if (!Compare(Prototype, Candidate))
+          continue;
+
+        PotentialGroup.push_back(Candidate);
+        // Make sure we never visit this StmtSequence again.
+        ++Indexes[j];
+      }
 
-    // There is a chance that we haven't found more than two fitting
-    // CloneSignature because not enough CloneSignatures passed the complexity
-    // requirement. As a CloneGroup with less than two members makes no sense,
-    // we ignore this CloneGroup and won't add it to the result.
-    if (!Group.isValid())
-      continue;
+      // Otherwise, add it to the result and continue searching for more
+      // groups.
+      Result.push_back(PotentialGroup);
+    }
 
-    CloneGroups.push_back(Group);
+    assert(std::all_of(Indexes.begin(), Indexes.end(),
+                       [](char c) { return c == 1; }));
   }
+  CloneGroups = Result;
+}
 
-  // Add every valid clone group that fulfills the complexity requirement.
-  for (const CloneGroup &Group : CloneGroups) {
-    createCloneGroups(Result, Group, CheckPatterns);
+void VariablePattern::addVariableOccurence(const VarDecl *VarDecl,
+                                           const Stmt *Mention) {
+  // First check if we already reference this variable
+  for (size_t KindIndex = 0; KindIndex < Variables.size(); ++KindIndex) {
+    if (Variables[KindIndex] == VarDecl) {
+      // If yes, add a new occurence that points to the existing entry in
+      // the Variables vector.
+      Occurences.emplace_back(KindIndex, Mention);
+      return;
+    }
   }
+  // If this variable wasn't already referenced, add it to the list of
+  // referenced variables and add a occurence that points to this new entry.
+  Occurences.emplace_back(Variables.size(), Mention);
+  Variables.push_back(VarDecl);
+}
 
-  std::vector<unsigned> IndexesToRemove;
-
-  // Compare every group in the result with the rest. If one groups contains
-  // another group, we only need to return the bigger group.
-  // Note: This doesn't scale well, so if possible avoid calling any heavy
-  // function from this loop to minimize the performance impact.
-  for (unsigned i = 0; i < Result.size(); ++i) {
-    for (unsigned j = 0; j < Result.size(); ++j) {
-      // Don't compare a group with itself.
-      if (i == j)
-        continue;
+void VariablePattern::addVariables(const Stmt *S) {
+  // Sometimes we get a nullptr (such as from IfStmts which often have nullptr
+  // children). We skip such statements as they don't reference any
+  // variables.
+  if (!S)
+    return;
 
-      if (containsGroup(Result[j], Result[i])) {
-        IndexesToRemove.push_back(i);
-        break;
-      }
-    }
+  // Check if S is a reference to a variable. If yes, add it to the pattern.
+  if (auto D = dyn_cast<DeclRefExpr>(S)) {
+    if (auto VD = dyn_cast<VarDecl>(D->getDecl()->getCanonicalDecl()))
+      addVariableOccurence(VD, D);
   }
 
-  // Erasing a list of indexes from the vector should be done with decreasing
-  // indexes. As IndexesToRemove is constructed with increasing values, we just
-  // reverse iterate over it to get the desired order.
-  for (auto I = IndexesToRemove.rbegin(); I != IndexesToRemove.rend(); ++I) {
-    Result.erase(Result.begin() + *I);
+  // Recursively check all children of the given statement.
+  for (const Stmt *Child : S->children()) {
+    addVariables(Child);
   }
 }
 
-void CloneDetector::findSuspiciousClones(
-    std::vector<CloneDetector::SuspiciousClonePair> &Result,
-    unsigned MinGroupComplexity) {
-  std::vector<CloneGroup> Clones;
-  // Reuse the normal search for clones but specify that the clone groups don't
-  // need to have a common referenced variable pattern so that we can manually
-  // search for the kind of pattern errors this function is supposed to find.
-  findClones(Clones, MinGroupComplexity, false);
-
-  for (const CloneGroup &Group : Clones) {
-    for (unsigned i = 0; i < Group.Sequences.size(); ++i) {
-      VariablePattern PatternA(Group.Sequences[i]);
-
-      for (unsigned j = i + 1; j < Group.Sequences.size(); ++j) {
-        VariablePattern PatternB(Group.Sequences[j]);
-
-        CloneDetector::SuspiciousClonePair ClonePair;
-        // For now, we only report clones which break the variable pattern just
-        // once because multiple differences in a pattern are an indicator that
-        // those differences are maybe intended (e.g. because it's actually
-        // a different algorithm).
-        // TODO: In very big clones even multiple variables can be unintended,
-        // so replacing this number with a percentage could better handle such
-        // cases. On the other hand it could increase the false-positive rate
-        // for all clones if the percentage is too high.
-        if (PatternA.countPatternDifferences(PatternB, &ClonePair) == 1) {
-          Result.push_back(ClonePair);
-          break;
-        }
-      }
-    }
+unsigned VariablePattern::countPatternDifferences(
+    const VariablePattern &Other,
+    VariablePattern::SuspiciousClonePair *FirstMismatch) {
+  unsigned NumberOfDifferences = 0;
+
+  assert(Other.Occurences.size() == Occurences.size());
+  for (unsigned i = 0; i < Occurences.size(); ++i) {
+    auto ThisOccurence = Occurences[i];
+    auto OtherOccurence = Other.Occurences[i];
+    if (ThisOccurence.KindID == OtherOccurence.KindID)
+      continue;
+
+    ++NumberOfDifferences;
+
+    // If FirstMismatch is not a nullptr, we need to store information about
+    // the first difference between the two patterns.
+    if (FirstMismatch == nullptr)
+      continue;
+
+    // Only proceed if we just found the first difference as we only store
+    // information about the first difference.
+    if (NumberOfDifferences != 1)
+      continue;
+
+    const VarDecl *FirstSuggestion = nullptr;
+    // If there is a variable available in the list of referenced variables
+    // which wouldn't break the pattern if it is used in place of the
+    // current variable, we provide this variable as the suggested fix.
+    if (OtherOccurence.KindID < Variables.size())
+      FirstSuggestion = Variables[OtherOccurence.KindID];
+
+    // Store information about the first clone.
+    FirstMismatch->FirstCloneInfo =
+        VariablePattern::SuspiciousClonePair::SuspiciousCloneInfo(
+            Variables[ThisOccurence.KindID], ThisOccurence.Mention,
+            FirstSuggestion);
+
+    // Same as above but with the other clone. We do this for both clones as
+    // we don't know which clone is the one containing the unintended
+    // pattern error.
+    const VarDecl *SecondSuggestion = nullptr;
+    if (ThisOccurence.KindID < Other.Variables.size())
+      SecondSuggestion = Other.Variables[ThisOccurence.KindID];
+
+    // Store information about the second clone.
+    FirstMismatch->SecondCloneInfo =
+        VariablePattern::SuspiciousClonePair::SuspiciousCloneInfo(
+            Other.Variables[OtherOccurence.KindID], OtherOccurence.Mention,
+            SecondSuggestion);
+
+    // SuspiciousClonePair guarantees that the first clone always has a
+    // suggested variable associated with it. As we know that one of the two
+    // clones in the pair always has suggestion, we swap the two clones
+    // in case the first clone has no suggested variable which means that
+    // the second clone has a suggested variable and should be first.
+    if (!FirstMismatch->FirstCloneInfo.Suggestion)
+      std::swap(FirstMismatch->FirstCloneInfo, FirstMismatch->SecondCloneInfo);
+
+    // This ensures that we always have at least one suggestion in a pair.
+    assert(FirstMismatch->FirstCloneInfo.Suggestion);
   }
+
+  return NumberOfDifferences;
 }
diff --git a/lib/Analysis/OSLog.cpp b/lib/Analysis/OSLog.cpp
index 3e13a153c65f..b2983932ea22 100644
--- a/lib/Analysis/OSLog.cpp
+++ b/lib/Analysis/OSLog.cpp
@@ -10,11 +10,11 @@
 #include "llvm/ADT/SmallBitVector.h"
 
 using namespace clang;
-using llvm::APInt;
 
 using clang::analyze_os_log::OSLogBufferItem;
 using clang::analyze_os_log::OSLogBufferLayout;
 
+namespace {
 class OSLogFormatStringHandler
     : public analyze_format_string::FormatStringHandler {
 private:
@@ -166,6 +166,7 @@ public:
     }
   }
 };
+} // end anonymous namespace
 
 bool clang::analyze_os_log::computeOSLogBufferLayout(
     ASTContext &Ctx, const CallExpr *E, OSLogBufferLayout &Layout) {
diff --git a/lib/Analysis/ReachableCode.cpp b/lib/Analysis/ReachableCode.cpp
index a2f3203762f7..60724ea60b25 100644
--- a/lib/Analysis/ReachableCode.cpp
+++ b/lib/Analysis/ReachableCode.cpp
@@ -58,6 +58,14 @@ static bool isTrivialDoWhile(const CFGBlock *B, const Stmt *S) {
   return false;
 }
 
+static bool isBuiltinUnreachable(const Stmt *S) {
+  if (const auto *DRE = dyn_cast<DeclRefExpr>(S))
+    if (const auto *FDecl = dyn_cast<FunctionDecl>(DRE->getDecl()))
+      return FDecl->getIdentifier() &&
+             FDecl->getBuiltinID() == Builtin::BI__builtin_unreachable;
+  return false;
+}
+
 static bool isDeadReturn(const CFGBlock *B, const Stmt *S) {
   // Look to see if the current control flow ends with a 'return', and see if
   // 'S' is a substatement. The 'return' may not be the last element in the
@@ -132,15 +140,21 @@ static bool isExpandedFromConfigurationMacro(const Stmt *S,
   // so that we can refine it later.
   SourceLocation L = S->getLocStart();
   if (L.isMacroID()) {
+    SourceManager &SM = PP.getSourceManager();
     if (IgnoreYES_NO) {
       // The Objective-C constant 'YES' and 'NO'
       // are defined as macros.  Do not treat them
       // as configuration values.
-      SourceManager &SM = PP.getSourceManager();
       SourceLocation TopL = getTopMostMacro(L, SM);
       StringRef MacroName = PP.getImmediateMacroName(TopL);
       if (MacroName == "YES" || MacroName == "NO")
         return false;
+    } else if (!PP.getLangOpts().CPlusPlus) {
+      // Do not treat C 'false' and 'true' macros as configuration values.
+      SourceLocation TopL = getTopMostMacro(L, SM);
+      StringRef MacroName = PP.getImmediateMacroName(TopL);
+      if (MacroName == "false" || MacroName == "true")
+        return false;
     }
     return true;
   }
@@ -586,8 +600,7 @@ void DeadCodeScan::reportDeadCode(const CFGBlock *B,
 
   if (isa<BreakStmt>(S)) {
     UK = reachable_code::UK_Break;
-  }
-  else if (isTrivialDoWhile(B, S)) {
+  } else if (isTrivialDoWhile(B, S) || isBuiltinUnreachable(S)) {
     return;
   }
   else if (isDeadReturn(B, S)) {
diff --git a/lib/Analysis/ThreadSafetyTIL.cpp b/lib/Analysis/ThreadSafetyTIL.cpp
index 2923f7e66a61..83aa90435e2a 100644
--- a/lib/Analysis/ThreadSafetyTIL.cpp
+++ b/lib/Analysis/ThreadSafetyTIL.cpp
@@ -186,8 +186,8 @@ int BasicBlock::topologicalSort(SimpleArray<BasicBlock*>& Blocks, int ID) {
 //
 // This sort assumes that (1) dominators have been computed, (2) there are no
 // critical edges, and (3) the entry block is reachable from the exit block
-// and no blocks are accessable via traversal of back-edges from the exit that
-// weren't accessable via forward edges from the entry.
+// and no blocks are accessible via traversal of back-edges from the exit that
+// weren't accessible via forward edges from the entry.
 int BasicBlock::topologicalFinalSort(SimpleArray<BasicBlock*>& Blocks, int ID) {
   // Visited is assumed to have been set by the topologicalSort.  This pass
   // assumes !Visited means that we've visited this node before.
diff --git a/lib/Basic/CMakeLists.txt b/lib/Basic/CMakeLists.txt
index 8929ec30ff7b..2feb31851cfe 100644
--- a/lib/Basic/CMakeLists.txt
+++ b/lib/Basic/CMakeLists.txt
@@ -74,6 +74,7 @@ add_clang_library(clangBasic
   FileSystemStatCache.cpp
   IdentifierTable.cpp
   LangOptions.cpp
+  MemoryBufferCache.cpp
   Module.cpp
   ObjCRuntime.cpp
   OpenMPKinds.cpp
@@ -89,6 +90,7 @@ add_clang_library(clangBasic
   VersionTuple.cpp
   VirtualFileSystem.cpp
   Warnings.cpp
+  XRayLists.cpp
   ${version_inc}
   )
 
diff --git a/lib/Basic/Diagnostic.cpp b/lib/Basic/Diagnostic.cpp
index 7529c475d6b9..350d5477751c 100644
--- a/lib/Basic/Diagnostic.cpp
+++ b/lib/Basic/Diagnostic.cpp
@@ -16,6 +16,7 @@
 #include "clang/Basic/DiagnosticOptions.h"
 #include "clang/Basic/IdentifierTable.h"
 #include "clang/Basic/PartialDiagnostic.h"
+#include "clang/Basic/SourceManager.h"
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/StringExtras.h"
 #include "llvm/Support/CrashRecoveryContext.h"
@@ -131,13 +132,13 @@ void DiagnosticsEngine::Reset() {
 
   // Clear state related to #pragma diagnostic.
   DiagStates.clear();
-  DiagStatePoints.clear();
+  DiagStatesByLoc.clear();
   DiagStateOnPushStack.clear();
 
   // Create a DiagState and DiagStatePoint representing diagnostic changes
   // through command-line.
   DiagStates.emplace_back();
-  DiagStatePoints.push_back(DiagStatePoint(&DiagStates.back(), FullSourceLoc()));
+  DiagStatesByLoc.appendFirst(&DiagStates.back());
 }
 
 void DiagnosticsEngine::SetDelayedDiagnostic(unsigned DiagID, StringRef Arg1,
@@ -157,27 +158,94 @@ void DiagnosticsEngine::ReportDelayed() {
   DelayedDiagArg2.clear();
 }
 
-DiagnosticsEngine::DiagStatePointsTy::iterator
-DiagnosticsEngine::GetDiagStatePointForLoc(SourceLocation L) const {
-  assert(!DiagStatePoints.empty());
-  assert(DiagStatePoints.front().Loc.isInvalid() &&
-         "Should have created a DiagStatePoint for command-line");
+void DiagnosticsEngine::DiagStateMap::appendFirst(
+                                             DiagState *State) {
+  assert(Files.empty() && "not first");
+  FirstDiagState = CurDiagState = State;
+  CurDiagStateLoc = SourceLocation();
+}
+
+void DiagnosticsEngine::DiagStateMap::append(SourceManager &SrcMgr,
+                                             SourceLocation Loc,
+                                             DiagState *State) {
+  CurDiagState = State;
+  CurDiagStateLoc = Loc;
+
+  std::pair<FileID, unsigned> Decomp = SrcMgr.getDecomposedLoc(Loc);
+  unsigned Offset = Decomp.second;
+  for (File *F = getFile(SrcMgr, Decomp.first); F;
+       Offset = F->ParentOffset, F = F->Parent) {
+    F->HasLocalTransitions = true;
+    auto &Last = F->StateTransitions.back();
+    assert(Last.Offset <= Offset && "state transitions added out of order");
+
+    if (Last.Offset == Offset) {
+      if (Last.State == State)
+        break;
+      Last.State = State;
+      continue;
+    }
+
+    F->StateTransitions.push_back({State, Offset});
+  }
+}
 
-  if (!SourceMgr)
-    return DiagStatePoints.end() - 1;
+DiagnosticsEngine::DiagState *
+DiagnosticsEngine::DiagStateMap::lookup(SourceManager &SrcMgr,
+                                        SourceLocation Loc) const {
+  // Common case: we have not seen any diagnostic pragmas.
+  if (Files.empty())
+    return FirstDiagState;
 
-  FullSourceLoc Loc(L, *SourceMgr);
-  if (Loc.isInvalid())
-    return DiagStatePoints.end() - 1;
+  std::pair<FileID, unsigned> Decomp = SrcMgr.getDecomposedLoc(Loc);
+  const File *F = getFile(SrcMgr, Decomp.first);
+  return F->lookup(Decomp.second);
+}
+
+DiagnosticsEngine::DiagState *
+DiagnosticsEngine::DiagStateMap::File::lookup(unsigned Offset) const {
+  auto OnePastIt = std::upper_bound(
+      StateTransitions.begin(), StateTransitions.end(), Offset,
+      [](unsigned Offset, const DiagStatePoint &P) {
+        return Offset < P.Offset;
+      });
+  assert(OnePastIt != StateTransitions.begin() && "missing initial state");
+  return OnePastIt[-1].State;
+}
+
+DiagnosticsEngine::DiagStateMap::File *
+DiagnosticsEngine::DiagStateMap::getFile(SourceManager &SrcMgr,
+                                         FileID ID) const {
+  // Get or insert the File for this ID.
+  auto Range = Files.equal_range(ID);
+  if (Range.first != Range.second)
+    return &Range.first->second;
+  auto &F = Files.insert(Range.first, std::make_pair(ID, File()))->second;
+
+  // We created a new File; look up the diagnostic state at the start of it and
+  // initialize it.
+  if (ID.isValid()) {
+    std::pair<FileID, unsigned> Decomp = SrcMgr.getDecomposedIncludedLoc(ID);
+    F.Parent = getFile(SrcMgr, Decomp.first);
+    F.ParentOffset = Decomp.second;
+    F.StateTransitions.push_back({F.Parent->lookup(Decomp.second), 0});
+  } else {
+    // This is the (imaginary) root file into which we pretend all top-level
+    // files are included; it descends from the initial state.
+    //
+    // FIXME: This doesn't guarantee that we use the same ordering as
+    // isBeforeInTranslationUnit in the cases where someone invented another
+    // top-level file and added diagnostic pragmas to it. See the code at the
+    // end of isBeforeInTranslationUnit for the quirks it deals with.
+    F.StateTransitions.push_back({FirstDiagState, 0});
+  }
+  return &F;
+}
 
-  DiagStatePointsTy::iterator Pos = DiagStatePoints.end();
-  FullSourceLoc LastStateChangePos = DiagStatePoints.back().Loc;
-  if (LastStateChangePos.isValid() &&
-      Loc.isBeforeInTranslationUnitThan(LastStateChangePos))
-    Pos = std::upper_bound(DiagStatePoints.begin(), DiagStatePoints.end(),
-                           DiagStatePoint(nullptr, Loc));
-  --Pos;
-  return Pos;
+void DiagnosticsEngine::PushDiagStatePoint(DiagState *State,
+                                           SourceLocation Loc) {
+  assert(Loc.isValid() && "Adding invalid loc point");
+  DiagStatesByLoc.append(*SourceMgr, Loc, State);
 }
 
 void DiagnosticsEngine::setSeverity(diag::kind Diag, diag::Severity Map,
@@ -187,65 +255,38 @@ void DiagnosticsEngine::setSeverity(diag::kind Diag, diag::Severity Map,
   assert((Diags->isBuiltinWarningOrExtension(Diag) ||
           (Map == diag::Severity::Fatal || Map == diag::Severity::Error)) &&
          "Cannot map errors into warnings!");
-  assert(!DiagStatePoints.empty());
   assert((L.isInvalid() || SourceMgr) && "No SourceMgr for valid location");
 
-  FullSourceLoc Loc = SourceMgr? FullSourceLoc(L, *SourceMgr) : FullSourceLoc();
-  FullSourceLoc LastStateChangePos = DiagStatePoints.back().Loc;
   // Don't allow a mapping to a warning override an error/fatal mapping.
+  bool WasUpgradedFromWarning = false;
   if (Map == diag::Severity::Warning) {
     DiagnosticMapping &Info = GetCurDiagState()->getOrAddMapping(Diag);
     if (Info.getSeverity() == diag::Severity::Error ||
-        Info.getSeverity() == diag::Severity::Fatal)
+        Info.getSeverity() == diag::Severity::Fatal) {
       Map = Info.getSeverity();
+      WasUpgradedFromWarning = true;
+    }
   }
   DiagnosticMapping Mapping = makeUserMapping(Map, L);
+  Mapping.setUpgradedFromWarning(WasUpgradedFromWarning);
 
   // Common case; setting all the diagnostics of a group in one place.
-  if (Loc.isInvalid() || Loc == LastStateChangePos) {
-    GetCurDiagState()->setMapping(Diag, Mapping);
-    return;
-  }
-
-  // Another common case; modifying diagnostic state in a source location
-  // after the previous one.
-  if ((Loc.isValid() && LastStateChangePos.isInvalid()) ||
-      LastStateChangePos.isBeforeInTranslationUnitThan(Loc)) {
-    // A diagnostic pragma occurred, create a new DiagState initialized with
-    // the current one and a new DiagStatePoint to record at which location
-    // the new state became active.
-    DiagStates.push_back(*GetCurDiagState());
-    PushDiagStatePoint(&DiagStates.back(), Loc);
-    GetCurDiagState()->setMapping(Diag, Mapping);
-    return;
-  }
-
-  // We allow setting the diagnostic state in random source order for
-  // completeness but it should not be actually happening in normal practice.
-
-  DiagStatePointsTy::iterator Pos = GetDiagStatePointForLoc(Loc);
-  assert(Pos != DiagStatePoints.end());
-
-  // Update all diagnostic states that are active after the given location.
-  for (DiagStatePointsTy::iterator
-         I = Pos+1, E = DiagStatePoints.end(); I != E; ++I) {
-    I->State->setMapping(Diag, Mapping);
-  }
-
-  // If the location corresponds to an existing point, just update its state.
-  if (Pos->Loc == Loc) {
-    Pos->State->setMapping(Diag, Mapping);
+  if ((L.isInvalid() || L == DiagStatesByLoc.getCurDiagStateLoc()) &&
+      DiagStatesByLoc.getCurDiagState()) {
+    // FIXME: This is theoretically wrong: if the current state is shared with
+    // some other location (via push/pop) we will change the state for that
+    // other location as well. This cannot currently happen, as we can't update
+    // the diagnostic state at the same location at which we pop.
+    DiagStatesByLoc.getCurDiagState()->setMapping(Diag, Mapping);
     return;
   }
 
-  // Create a new state/point and fit it into the vector of DiagStatePoints
-  // so that the vector is always ordered according to location.
-  assert(Pos->Loc.isBeforeInTranslationUnitThan(Loc));
-  DiagStates.push_back(*Pos->State);
-  DiagState *NewState = &DiagStates.back();
-  NewState->setMapping(Diag, Mapping);
-  DiagStatePoints.insert(Pos+1, DiagStatePoint(NewState,
-                                               FullSourceLoc(Loc, *SourceMgr)));
+  // A diagnostic pragma occurred, create a new DiagState initialized with
+  // the current one and a new DiagStatePoint to record at which location
+  // the new state became active.
+  DiagStates.push_back(*GetCurDiagState());
+  DiagStates.back().setMapping(Diag, Mapping);
+  PushDiagStatePoint(&DiagStates.back(), L);
 }
 
 bool DiagnosticsEngine::setSeverityForGroup(diag::Flavor Flavor,
diff --git a/lib/Basic/DiagnosticIDs.cpp b/lib/Basic/DiagnosticIDs.cpp
index 3c370f67fa32..e0580af45b50 100644
--- a/lib/Basic/DiagnosticIDs.cpp
+++ b/lib/Basic/DiagnosticIDs.cpp
@@ -411,11 +411,8 @@ DiagnosticIDs::getDiagnosticSeverity(unsigned DiagID, SourceLocation Loc,
   // to error.  Errors can only be mapped to fatal.
   diag::Severity Result = diag::Severity::Fatal;
 
-  DiagnosticsEngine::DiagStatePointsTy::iterator
-    Pos = Diag.GetDiagStatePointForLoc(Loc);
-  DiagnosticsEngine::DiagState *State = Pos->State;
-
   // Get the mapping information, or compute it lazily.
+  DiagnosticsEngine::DiagState *State = Diag.GetDiagStateForLoc(Loc);
   DiagnosticMapping &Mapping = State->getOrAddMapping((diag::kind)DiagID);
 
   // TODO: Can a null severity really get here?
diff --git a/lib/Basic/FileManager.cpp b/lib/Basic/FileManager.cpp
index 50050d0a519b..0c10b5f4d14c 100644
--- a/lib/Basic/FileManager.cpp
+++ b/lib/Basic/FileManager.cpp
@@ -50,14 +50,14 @@ using namespace clang;
 
 FileManager::FileManager(const FileSystemOptions &FSO,
                          IntrusiveRefCntPtr<vfs::FileSystem> FS)
-  : FS(FS), FileSystemOpts(FSO),
-    SeenDirEntries(64), SeenFileEntries(64), NextFileUID(0) {
+    : FS(std::move(FS)), FileSystemOpts(FSO), SeenDirEntries(64),
+      SeenFileEntries(64), NextFileUID(0) {
   NumDirLookups = NumFileLookups = 0;
   NumDirCacheMisses = NumFileCacheMisses = 0;
 
   // If the caller doesn't provide a virtual file system, just grab the real
   // file system.
-  if (!FS)
+  if (!this->FS)
     this->FS = vfs::getRealFileSystem();
 }
 
@@ -386,6 +386,7 @@ FileManager::getVirtualFile(StringRef Filename, off_t Size,
   UFE->ModTime = ModificationTime;
   UFE->Dir     = DirInfo;
   UFE->UID     = NextFileUID++;
+  UFE->IsValid = true;
   UFE->File.reset();
   return UFE;
 }
diff --git a/lib/Basic/IdentifierTable.cpp b/lib/Basic/IdentifierTable.cpp
index af424cd92390..74c85376c7db 100644
--- a/lib/Basic/IdentifierTable.cpp
+++ b/lib/Basic/IdentifierTable.cpp
@@ -244,7 +244,7 @@ static KeywordStatus getTokenKwStatus(const LangOptions &LangOpts,
 
 /// \brief Returns true if the identifier represents a keyword in the
 /// specified language.
-bool IdentifierInfo::isKeyword(const LangOptions &LangOpts) {
+bool IdentifierInfo::isKeyword(const LangOptions &LangOpts) const {
   switch (getTokenKwStatus(LangOpts, getTokenID())) {
   case KS_Enabled:
   case KS_Extension:
@@ -254,6 +254,19 @@ bool IdentifierInfo::isKeyword(const LangOptions &LangOpts) {
   }
 }
 
+/// \brief Returns true if the identifier represents a C++ keyword in the
+/// specified language.
+bool IdentifierInfo::isCPlusPlusKeyword(const LangOptions &LangOpts) const {
+  if (!LangOpts.CPlusPlus || !isKeyword(LangOpts))
+    return false;
+  // This is a C++ keyword if this identifier is not a keyword when checked
+  // using LangOptions without C++ support.
+  LangOptions LangOptsNoCPP = LangOpts;
+  LangOptsNoCPP.CPlusPlus = false;
+  LangOptsNoCPP.CPlusPlus11 = false;
+  return !isKeyword(LangOptsNoCPP);
+}
+
 tok::PPKeywordKind IdentifierInfo::getPPKeywordID() const {
   // We use a perfect hash function here involving the length of the keyword,
   // the first and third character.  For preprocessor ID's there are no
@@ -487,8 +500,10 @@ ObjCMethodFamily Selector::getMethodFamilyImpl(Selector sel) {
     if (name == "self") return OMF_self;
     if (name == "initialize") return OMF_initialize;
   }
- 
-  if (name == "performSelector") return OMF_performSelector;
+
+  if (name == "performSelector" || name == "performSelectorInBackground" ||
+      name == "performSelectorOnMainThread")
+    return OMF_performSelector;
 
   // The other method families may begin with a prefix of underscores.
   while (!name.empty() && name.front() == '_')
diff --git a/lib/Basic/LangOptions.cpp b/lib/Basic/LangOptions.cpp
index ff10a773a97c..c8a774311efe 100644
--- a/lib/Basic/LangOptions.cpp
+++ b/lib/Basic/LangOptions.cpp
@@ -33,6 +33,8 @@ void LangOptions::resetNonModularOptions() {
   // sanitizer options (this affects __has_feature(address_sanitizer) etc).
   Sanitize.clear();
   SanitizerBlacklistFiles.clear();
+  XRayAlwaysInstrumentFiles.clear();
+  XRayNeverInstrumentFiles.clear();
 
   CurrentModule.clear();
   IsHeaderFile = false;
diff --git a/lib/Basic/MemoryBufferCache.cpp b/lib/Basic/MemoryBufferCache.cpp
new file mode 100644
index 000000000000..c1fc571ec9b3
--- /dev/null
+++ b/lib/Basic/MemoryBufferCache.cpp
@@ -0,0 +1,48 @@
+//===- MemoryBufferCache.cpp - Cache for loaded memory buffers ------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+
+#include "clang/Basic/MemoryBufferCache.h"
+#include "llvm/Support/MemoryBuffer.h"
+
+using namespace clang;
+
+llvm::MemoryBuffer &
+MemoryBufferCache::addBuffer(llvm::StringRef Filename,
+                             std::unique_ptr<llvm::MemoryBuffer> Buffer) {
+  auto Insertion =
+      Buffers.insert({Filename, BufferEntry{std::move(Buffer), NextIndex++}});
+  assert(Insertion.second && "Already has a buffer");
+  return *Insertion.first->second.Buffer;
+}
+
+llvm::MemoryBuffer *MemoryBufferCache::lookupBuffer(llvm::StringRef Filename) {
+  auto I = Buffers.find(Filename);
+  if (I == Buffers.end())
+    return nullptr;
+  return I->second.Buffer.get();
+}
+
+bool MemoryBufferCache::isBufferFinal(llvm::StringRef Filename) {
+  auto I = Buffers.find(Filename);
+  if (I == Buffers.end())
+    return false;
+  return I->second.Index < FirstRemovableIndex;
+}
+
+bool MemoryBufferCache::tryToRemoveBuffer(llvm::StringRef Filename) {
+  auto I = Buffers.find(Filename);
+  assert(I != Buffers.end() && "No buffer to remove...");
+  if (I->second.Index < FirstRemovableIndex)
+    return true;
+
+  Buffers.erase(I);
+  return false;
+}
+
+void MemoryBufferCache::finalizeCurrentBuffers() { FirstRemovableIndex = NextIndex; }
diff --git a/lib/Basic/Module.cpp b/lib/Basic/Module.cpp
index 80bbc24f3db3..a6fd931cb174 100644
--- a/lib/Basic/Module.cpp
+++ b/lib/Basic/Module.cpp
@@ -27,7 +27,7 @@ using namespace clang;
 Module::Module(StringRef Name, SourceLocation DefinitionLoc, Module *Parent,
                bool IsFramework, bool IsExplicit, unsigned VisibilityID)
     : Name(Name), DefinitionLoc(DefinitionLoc), Parent(Parent), Directory(),
-      Umbrella(), Signature(0), ASTFile(nullptr), VisibilityID(VisibilityID),
+      Umbrella(), ASTFile(nullptr), VisibilityID(VisibilityID),
       IsMissingRequirement(false), HasIncompatibleModuleFile(false),
       IsAvailable(true), IsFromModuleFile(false), IsFramework(IsFramework),
       IsExplicit(IsExplicit), IsSystem(false), IsExternC(false),
diff --git a/lib/Basic/OpenMPKinds.cpp b/lib/Basic/OpenMPKinds.cpp
index 905c3693d378..5a8bb61eaadf 100644
--- a/lib/Basic/OpenMPKinds.cpp
+++ b/lib/Basic/OpenMPKinds.cpp
@@ -863,3 +863,76 @@ bool clang::isOpenMPLoopBoundSharingDirective(OpenMPDirectiveKind Kind) {
          Kind == OMPD_target_teams_distribute_parallel_for_simd ||
          Kind == OMPD_target_teams_distribute_simd;
 }
+
+void clang::getOpenMPCaptureRegions(
+    SmallVectorImpl<OpenMPDirectiveKind> &CaptureRegions,
+    OpenMPDirectiveKind DKind) {
+  assert(DKind <= OMPD_unknown);
+  switch (DKind) {
+  case OMPD_parallel:
+  case OMPD_parallel_for:
+  case OMPD_parallel_for_simd:
+  case OMPD_parallel_sections:
+  case OMPD_distribute_parallel_for:
+    CaptureRegions.push_back(OMPD_parallel);
+    break;
+  case OMPD_target_teams:
+    CaptureRegions.push_back(OMPD_target);
+    CaptureRegions.push_back(OMPD_teams);
+    break;
+  case OMPD_teams:
+  case OMPD_simd:
+  case OMPD_for:
+  case OMPD_for_simd:
+  case OMPD_sections:
+  case OMPD_section:
+  case OMPD_single:
+  case OMPD_master:
+  case OMPD_critical:
+  case OMPD_taskgroup:
+  case OMPD_distribute:
+  case OMPD_ordered:
+  case OMPD_atomic:
+  case OMPD_target_data:
+  case OMPD_target:
+  case OMPD_target_parallel_for:
+  case OMPD_target_parallel_for_simd:
+  case OMPD_target_simd:
+  case OMPD_task:
+  case OMPD_taskloop:
+  case OMPD_taskloop_simd:
+  case OMPD_distribute_parallel_for_simd:
+  case OMPD_distribute_simd:
+  case OMPD_teams_distribute:
+  case OMPD_teams_distribute_simd:
+  case OMPD_teams_distribute_parallel_for_simd:
+  case OMPD_teams_distribute_parallel_for:
+  case OMPD_target_teams_distribute:
+  case OMPD_target_teams_distribute_parallel_for:
+  case OMPD_target_teams_distribute_parallel_for_simd:
+  case OMPD_target_teams_distribute_simd:
+    CaptureRegions.push_back(DKind);
+    break;
+  case OMPD_target_parallel:
+    CaptureRegions.push_back(OMPD_target);
+    CaptureRegions.push_back(OMPD_parallel);
+    break;
+  case OMPD_threadprivate:
+  case OMPD_taskyield:
+  case OMPD_barrier:
+  case OMPD_taskwait:
+  case OMPD_cancellation_point:
+  case OMPD_cancel:
+  case OMPD_flush:
+  case OMPD_target_enter_data:
+  case OMPD_target_exit_data:
+  case OMPD_declare_reduction:
+  case OMPD_declare_simd:
+  case OMPD_declare_target:
+  case OMPD_end_declare_target:
+  case OMPD_target_update:
+    llvm_unreachable("OpenMP Directive is not allowed");
+  case OMPD_unknown:
+    llvm_unreachable("Unknown OpenMP directive");
+  }
+}
diff --git a/lib/Basic/SourceManager.cpp b/lib/Basic/SourceManager.cpp
index 380ca373e69b..156aa0b11f7d 100644
--- a/lib/Basic/SourceManager.cpp
+++ b/lib/Basic/SourceManager.cpp
@@ -1136,6 +1136,7 @@ unsigned SourceManager::getColumnNumber(FileID FID, unsigned FilePos,
     return 1;
   }
 
+  const char *Buf = MemBuf->getBufferStart();
   // See if we just calculated the line number for this FilePos and can use
   // that to lookup the start of the line instead of searching for it.
   if (LastLineNoFileIDQuery == FID &&
@@ -1144,11 +1145,19 @@ unsigned SourceManager::getColumnNumber(FileID FID, unsigned FilePos,
     unsigned *SourceLineCache = LastLineNoContentCache->SourceLineCache;
     unsigned LineStart = SourceLineCache[LastLineNoResult - 1];
     unsigned LineEnd = SourceLineCache[LastLineNoResult];
-    if (FilePos >= LineStart && FilePos < LineEnd)
+    if (FilePos >= LineStart && FilePos < LineEnd) {
+      // LineEnd is the LineStart of the next line.
+      // A line ends with separator LF or CR+LF on Windows.
+      // FilePos might point to the last separator,
+      // but we need a column number at most 1 + the last column.
+      if (FilePos + 1 == LineEnd && FilePos > LineStart) {
+        if (Buf[FilePos - 1] == '\r' || Buf[FilePos - 1] == '\n')
+          --FilePos;
+      }
       return FilePos - LineStart + 1;
+    }
   }
 
-  const char *Buf = MemBuf->getBufferStart();
   unsigned LineStart = FilePos;
   while (LineStart && Buf[LineStart-1] != '\n' && Buf[LineStart-1] != '\r')
     --LineStart;
diff --git a/lib/Basic/TargetInfo.cpp b/lib/Basic/TargetInfo.cpp
index b1b01e5f584f..e19404dc54cb 100644
--- a/lib/Basic/TargetInfo.cpp
+++ b/lib/Basic/TargetInfo.cpp
@@ -282,8 +282,9 @@ bool TargetInfo::isTypeSigned(IntType T) {
 
 /// adjust - Set forced language options.
 /// Apply changes to the target information with respect to certain
-/// language options which change the target configuration.
-void TargetInfo::adjust(const LangOptions &Opts) {
+/// language options which change the target configuration and adjust
+/// the language based on the target options where applicable.
+void TargetInfo::adjust(LangOptions &Opts) {
   if (Opts.NoBitFieldTypeAlign)
     UseBitFieldTypeAlignment = false;
   if (Opts.ShortWChar)
diff --git a/lib/Basic/Targets.cpp b/lib/Basic/Targets.cpp
index 1a95ff26816e..a457f6deee75 100644
--- a/lib/Basic/Targets.cpp
+++ b/lib/Basic/Targets.cpp
@@ -545,6 +545,8 @@ protected:
     Builder.defineMacro("__ELF__");
     if (Opts.POSIXThreads)
       Builder.defineMacro("_REENTRANT");
+    if (this->HasFloat128)
+      Builder.defineMacro("__FLOAT128__");
   }
 public:
   OpenBSDTargetInfo(const llvm::Triple &Triple, const TargetOptions &Opts)
@@ -552,11 +554,11 @@ public:
     this->TLSSupported = false;
 
       switch (Triple.getArch()) {
-        default:
         case llvm::Triple::x86:
         case llvm::Triple::x86_64:
-        case llvm::Triple::arm:
-        case llvm::Triple::sparc:
+          this->HasFloat128 = true;
+          // FALLTHROUGH
+        default:
           this->MCountName = "__mcount";
           break;
         case llvm::Triple::mips64:
@@ -886,6 +888,7 @@ class PPCTargetInfo : public TargetInfo {
   std::string CPU;
 
   // Target cpu features.
+  bool HasAltivec;
   bool HasVSX;
   bool HasP8Vector;
   bool HasP8Crypto;
@@ -901,9 +904,10 @@ protected:
 
 public:
   PPCTargetInfo(const llvm::Triple &Triple, const TargetOptions &)
-    : TargetInfo(Triple), HasVSX(false), HasP8Vector(false),
+    : TargetInfo(Triple), HasAltivec(false), HasVSX(false), HasP8Vector(false),
       HasP8Crypto(false), HasDirectMove(false), HasQPX(false), HasHTM(false),
       HasBPERMD(false), HasExtDiv(false), HasP9Vector(false) {
+    SuitableAlign = 128;
     SimdDefaultAlign = 128;
     LongDoubleWidth = LongDoubleAlign = 128;
     LongDoubleFormat = &llvm::APFloat::PPCDoubleDouble();
@@ -930,6 +934,13 @@ public:
     ArchDefineA2q   = 1 << 15
   } ArchDefineTypes;
 
+  // Set the language option for altivec based on our value.
+  void adjust(LangOptions &Opts) override {
+    if (HasAltivec)
+      Opts.AltiVec = 1;
+    TargetInfo::adjust(Opts);
+  }
+
   // Note: GCC recognizes the following additional cpus:
   //  401, 403, 405, 405fp, 440fp, 464, 464fp, 476, 476fp, 505, 740, 801,
   //  821, 823, 8540, 8548, e300c2, e300c3, e500mc64, e6500, 860, cell,
@@ -1164,7 +1175,9 @@ const Builtin::Info PPCTargetInfo::BuiltinInfo[] = {
 bool PPCTargetInfo::handleTargetFeatures(std::vector<std::string> &Features,
                                          DiagnosticsEngine &Diags) {
   for (const auto &Feature : Features) {
-    if (Feature == "+vsx") {
+    if (Feature == "+altivec") {
+      HasAltivec = true;
+    } else if (Feature == "+vsx") {
       HasVSX = true;
     } else if (Feature == "+bpermd") {
       HasBPERMD = true;
@@ -1224,82 +1237,100 @@ void PPCTargetInfo::getTargetDefines(const LangOptions &Opts,
   if (ABI == "elfv2")
     Builder.defineMacro("_CALL_ELF", "2");
 
+  // This typically is only for a new enough linker (bfd >= 2.16.2 or gold), but
+  // our suppport post-dates this and it should work on all 64-bit ppc linux
+  // platforms. It is guaranteed to work on all elfv2 platforms.
+  if (getTriple().getOS() == llvm::Triple::Linux && PointerWidth == 64)
+    Builder.defineMacro("_CALL_LINUX", "1");
+
   // Subtarget options.
   Builder.defineMacro("__NATURAL_ALIGNMENT__");
   Builder.defineMacro("__REGISTER_PREFIX__", "");
 
   // FIXME: Should be controlled by command line option.
-  if (LongDoubleWidth == 128)
+  if (LongDoubleWidth == 128) {
     Builder.defineMacro("__LONG_DOUBLE_128__");
-
-  if (Opts.AltiVec) {
-    Builder.defineMacro("__VEC__", "10206");
-    Builder.defineMacro("__ALTIVEC__");
+    Builder.defineMacro("__LONGDOUBLE128");
   }
 
+  // Define this for elfv2 (64-bit only) or 64-bit darwin.
+  if (ABI == "elfv2" ||
+      (getTriple().getOS() == llvm::Triple::Darwin && PointerWidth == 64))
+    Builder.defineMacro("__STRUCT_PARM_ALIGN__", "16");
+
   // CPU identification.
-  ArchDefineTypes defs = (ArchDefineTypes)llvm::StringSwitch<int>(CPU)
-    .Case("440",   ArchDefineName)
-    .Case("450",   ArchDefineName | ArchDefine440)
-    .Case("601",   ArchDefineName)
-    .Case("602",   ArchDefineName | ArchDefinePpcgr)
-    .Case("603",   ArchDefineName | ArchDefinePpcgr)
-    .Case("603e",  ArchDefineName | ArchDefine603 | ArchDefinePpcgr)
-    .Case("603ev", ArchDefineName | ArchDefine603 | ArchDefinePpcgr)
-    .Case("604",   ArchDefineName | ArchDefinePpcgr)
-    .Case("604e",  ArchDefineName | ArchDefine604 | ArchDefinePpcgr)
-    .Case("620",   ArchDefineName | ArchDefinePpcgr)
-    .Case("630",   ArchDefineName | ArchDefinePpcgr)
-    .Case("7400",  ArchDefineName | ArchDefinePpcgr)
-    .Case("7450",  ArchDefineName | ArchDefinePpcgr)
-    .Case("750",   ArchDefineName | ArchDefinePpcgr)
-    .Case("970",   ArchDefineName | ArchDefinePwr4 | ArchDefinePpcgr
-                     | ArchDefinePpcsq)
-    .Case("a2",    ArchDefineA2)
-    .Case("a2q",   ArchDefineName | ArchDefineA2 | ArchDefineA2q)
-    .Case("pwr3",  ArchDefinePpcgr)
-    .Case("pwr4",  ArchDefineName | ArchDefinePpcgr | ArchDefinePpcsq)
-    .Case("pwr5",  ArchDefineName | ArchDefinePwr4 | ArchDefinePpcgr
-                     | ArchDefinePpcsq)
-    .Case("pwr5x", ArchDefineName | ArchDefinePwr5 | ArchDefinePwr4
-                     | ArchDefinePpcgr | ArchDefinePpcsq)
-    .Case("pwr6",  ArchDefineName | ArchDefinePwr5x | ArchDefinePwr5
-                     | ArchDefinePwr4 | ArchDefinePpcgr | ArchDefinePpcsq)
-    .Case("pwr6x", ArchDefineName | ArchDefinePwr6 | ArchDefinePwr5x
-                     | ArchDefinePwr5 | ArchDefinePwr4 | ArchDefinePpcgr
-                     | ArchDefinePpcsq)
-    .Case("pwr7",  ArchDefineName | ArchDefinePwr6x | ArchDefinePwr6
-                     | ArchDefinePwr5x | ArchDefinePwr5 | ArchDefinePwr4
-                     | ArchDefinePpcgr | ArchDefinePpcsq)
-    .Case("pwr8",  ArchDefineName | ArchDefinePwr7 | ArchDefinePwr6x
-                     | ArchDefinePwr6 | ArchDefinePwr5x | ArchDefinePwr5
-                     | ArchDefinePwr4 | ArchDefinePpcgr | ArchDefinePpcsq)
-    .Case("pwr9",  ArchDefineName | ArchDefinePwr8 | ArchDefinePwr7
-                     | ArchDefinePwr6x | ArchDefinePwr6 | ArchDefinePwr5x
-                     | ArchDefinePwr5 | ArchDefinePwr4 | ArchDefinePpcgr
-                     | ArchDefinePpcsq)
-    .Case("power3",  ArchDefinePpcgr)
-    .Case("power4",  ArchDefinePwr4 | ArchDefinePpcgr | ArchDefinePpcsq)
-    .Case("power5",  ArchDefinePwr5 | ArchDefinePwr4 | ArchDefinePpcgr
-                       | ArchDefinePpcsq)
-    .Case("power5x", ArchDefinePwr5x | ArchDefinePwr5 | ArchDefinePwr4
-                       | ArchDefinePpcgr | ArchDefinePpcsq)
-    .Case("power6",  ArchDefinePwr6 | ArchDefinePwr5x | ArchDefinePwr5
-                       | ArchDefinePwr4 | ArchDefinePpcgr | ArchDefinePpcsq)
-    .Case("power6x", ArchDefinePwr6x | ArchDefinePwr6 | ArchDefinePwr5x
-                       | ArchDefinePwr5 | ArchDefinePwr4 | ArchDefinePpcgr
-                       | ArchDefinePpcsq)
-    .Case("power7",  ArchDefinePwr7 | ArchDefinePwr6x | ArchDefinePwr6
-                       | ArchDefinePwr5x | ArchDefinePwr5 | ArchDefinePwr4
-                       | ArchDefinePpcgr | ArchDefinePpcsq)
-    .Case("power8",  ArchDefinePwr8 | ArchDefinePwr7 | ArchDefinePwr6x
-                       | ArchDefinePwr6 | ArchDefinePwr5x | ArchDefinePwr5
-                       | ArchDefinePwr4 | ArchDefinePpcgr | ArchDefinePpcsq)
-    .Case("power9",  ArchDefinePwr9 | ArchDefinePwr8 | ArchDefinePwr7
-                       | ArchDefinePwr6x | ArchDefinePwr6 | ArchDefinePwr5x
-                       | ArchDefinePwr5 | ArchDefinePwr4 | ArchDefinePpcgr
-                       | ArchDefinePpcsq)
-    .Default(ArchDefineNone);
+  ArchDefineTypes defs =
+      (ArchDefineTypes)llvm::StringSwitch<int>(CPU)
+          .Case("440", ArchDefineName)
+          .Case("450", ArchDefineName | ArchDefine440)
+          .Case("601", ArchDefineName)
+          .Case("602", ArchDefineName | ArchDefinePpcgr)
+          .Case("603", ArchDefineName | ArchDefinePpcgr)
+          .Case("603e", ArchDefineName | ArchDefine603 | ArchDefinePpcgr)
+          .Case("603ev", ArchDefineName | ArchDefine603 | ArchDefinePpcgr)
+          .Case("604", ArchDefineName | ArchDefinePpcgr)
+          .Case("604e", ArchDefineName | ArchDefine604 | ArchDefinePpcgr)
+          .Case("620", ArchDefineName | ArchDefinePpcgr)
+          .Case("630", ArchDefineName | ArchDefinePpcgr)
+          .Case("7400", ArchDefineName | ArchDefinePpcgr)
+          .Case("7450", ArchDefineName | ArchDefinePpcgr)
+          .Case("750", ArchDefineName | ArchDefinePpcgr)
+          .Case("970", ArchDefineName | ArchDefinePwr4 | ArchDefinePpcgr |
+                           ArchDefinePpcsq)
+          .Case("a2", ArchDefineA2)
+          .Case("a2q", ArchDefineName | ArchDefineA2 | ArchDefineA2q)
+          .Case("pwr3", ArchDefinePpcgr)
+          .Case("pwr4", ArchDefineName | ArchDefinePpcgr | ArchDefinePpcsq)
+          .Case("pwr5", ArchDefineName | ArchDefinePwr4 | ArchDefinePpcgr |
+                            ArchDefinePpcsq)
+          .Case("pwr5x", ArchDefineName | ArchDefinePwr5 | ArchDefinePwr4 |
+                             ArchDefinePpcgr | ArchDefinePpcsq)
+          .Case("pwr6", ArchDefineName | ArchDefinePwr5x | ArchDefinePwr5 |
+                            ArchDefinePwr4 | ArchDefinePpcgr | ArchDefinePpcsq)
+          .Case("pwr6x", ArchDefineName | ArchDefinePwr6 | ArchDefinePwr5x |
+                             ArchDefinePwr5 | ArchDefinePwr4 | ArchDefinePpcgr |
+                             ArchDefinePpcsq)
+          .Case("pwr7", ArchDefineName | ArchDefinePwr6x | ArchDefinePwr6 |
+                            ArchDefinePwr5x | ArchDefinePwr5 | ArchDefinePwr4 |
+                            ArchDefinePpcgr | ArchDefinePpcsq)
+          .Case("pwr8", ArchDefineName | ArchDefinePwr7 | ArchDefinePwr6x |
+                            ArchDefinePwr6 | ArchDefinePwr5x | ArchDefinePwr5 |
+                            ArchDefinePwr4 | ArchDefinePpcgr | ArchDefinePpcsq)
+          .Case("pwr9", ArchDefineName | ArchDefinePwr8 | ArchDefinePwr7 |
+                            ArchDefinePwr6x | ArchDefinePwr6 | ArchDefinePwr5x |
+                            ArchDefinePwr5 | ArchDefinePwr4 | ArchDefinePpcgr |
+                            ArchDefinePpcsq)
+          .Case("power3", ArchDefinePpcgr)
+          .Case("power4", ArchDefinePwr4 | ArchDefinePpcgr | ArchDefinePpcsq)
+          .Case("power5", ArchDefinePwr5 | ArchDefinePwr4 | ArchDefinePpcgr |
+                              ArchDefinePpcsq)
+          .Case("power5x", ArchDefinePwr5x | ArchDefinePwr5 | ArchDefinePwr4 |
+                               ArchDefinePpcgr | ArchDefinePpcsq)
+          .Case("power6", ArchDefinePwr6 | ArchDefinePwr5x | ArchDefinePwr5 |
+                              ArchDefinePwr4 | ArchDefinePpcgr |
+                              ArchDefinePpcsq)
+          .Case("power6x", ArchDefinePwr6x | ArchDefinePwr6 | ArchDefinePwr5x |
+                               ArchDefinePwr5 | ArchDefinePwr4 |
+                               ArchDefinePpcgr | ArchDefinePpcsq)
+          .Case("power7", ArchDefinePwr7 | ArchDefinePwr6x | ArchDefinePwr6 |
+                              ArchDefinePwr5x | ArchDefinePwr5 |
+                              ArchDefinePwr4 | ArchDefinePpcgr |
+                              ArchDefinePpcsq)
+          .Case("power8", ArchDefinePwr8 | ArchDefinePwr7 | ArchDefinePwr6x |
+                              ArchDefinePwr6 | ArchDefinePwr5x |
+                              ArchDefinePwr5 | ArchDefinePwr4 |
+                              ArchDefinePpcgr | ArchDefinePpcsq)
+          .Case("power9", ArchDefinePwr9 | ArchDefinePwr8 | ArchDefinePwr7 |
+                              ArchDefinePwr6x | ArchDefinePwr6 |
+                              ArchDefinePwr5x | ArchDefinePwr5 |
+                              ArchDefinePwr4 | ArchDefinePpcgr |
+                              ArchDefinePpcsq)
+          // powerpc64le automatically defaults to at least power8.
+          .Case("ppc64le", ArchDefinePwr8 | ArchDefinePwr7 | ArchDefinePwr6x |
+                               ArchDefinePwr6 | ArchDefinePwr5x |
+                               ArchDefinePwr5 | ArchDefinePwr4 |
+                               ArchDefinePpcgr | ArchDefinePpcsq)
+          .Default(ArchDefineNone);
 
   if (defs & ArchDefineName)
     Builder.defineMacro(Twine("_ARCH_", StringRef(CPU).upper()));
@@ -1343,6 +1374,10 @@ void PPCTargetInfo::getTargetDefines(const LangOptions &Opts,
     Builder.defineMacro("__TOS_BGQ__");
   }
 
+  if (HasAltivec) {
+    Builder.defineMacro("__VEC__", "10206");
+    Builder.defineMacro("__ALTIVEC__");
+  }
   if (HasVSX)
     Builder.defineMacro("__VSX__");
   if (HasP8Vector)
@@ -1362,6 +1397,9 @@ void PPCTargetInfo::getTargetDefines(const LangOptions &Opts,
   if (PointerWidth == 64)
     Builder.defineMacro("__GCC_HAVE_SYNC_COMPARE_AND_SWAP_8");
 
+  // We have support for the bswap intrinsics so we can define this.
+  Builder.defineMacro("__HAVE_BSWAP__", "1");
+
   // FIXME: The following are not yet generated here by Clang, but are
   //        generated by GCC:
   //
@@ -1374,8 +1412,6 @@ void PPCTargetInfo::getTargetDefines(const LangOptions &Opts,
   //   __RSQRTEF__
   //   _SOFT_DOUBLE_
   //   __NO_LWSYNC__
-  //   __HAVE_BSWAP__
-  //   __LONGDOUBLE128
   //   __CMODEL_MEDIUM__
   //   __CMODEL_LARGE__
   //   _CALL_SYSV
@@ -1481,6 +1517,11 @@ bool PPCTargetInfo::initFeatureMap(
     .Case("pwr8", true)
     .Case("pwr7", true)
     .Default(false);
+  Features["htm"] = llvm::StringSwitch<bool>(CPU)
+    .Case("ppc64le", true)
+    .Case("pwr9", true)
+    .Case("pwr8", true)
+    .Default(false);
 
   if (!ppcUserFeaturesCheck(Diags, FeaturesVec))
     return false;
@@ -1490,44 +1531,47 @@ bool PPCTargetInfo::initFeatureMap(
 
 bool PPCTargetInfo::hasFeature(StringRef Feature) const {
   return llvm::StringSwitch<bool>(Feature)
-    .Case("powerpc", true)
-    .Case("vsx", HasVSX)
-    .Case("power8-vector", HasP8Vector)
-    .Case("crypto", HasP8Crypto)
-    .Case("direct-move", HasDirectMove)
-    .Case("qpx", HasQPX)
-    .Case("htm", HasHTM)
-    .Case("bpermd", HasBPERMD)
-    .Case("extdiv", HasExtDiv)
-    .Case("float128", HasFloat128)
-    .Case("power9-vector", HasP9Vector)
-    .Default(false);
+      .Case("powerpc", true)
+      .Case("altivec", HasAltivec)
+      .Case("vsx", HasVSX)
+      .Case("power8-vector", HasP8Vector)
+      .Case("crypto", HasP8Crypto)
+      .Case("direct-move", HasDirectMove)
+      .Case("qpx", HasQPX)
+      .Case("htm", HasHTM)
+      .Case("bpermd", HasBPERMD)
+      .Case("extdiv", HasExtDiv)
+      .Case("float128", HasFloat128)
+      .Case("power9-vector", HasP9Vector)
+      .Default(false);
 }
 
 void PPCTargetInfo::setFeatureEnabled(llvm::StringMap<bool> &Features,
                                       StringRef Name, bool Enabled) const {
-  // If we're enabling direct-move or power8-vector go ahead and enable vsx
-  // as well. Do the inverse if we're disabling vsx. We'll diagnose any user
-  // incompatible options.
   if (Enabled) {
-    if (Name == "direct-move" ||
-        Name == "power8-vector" ||
-        Name == "float128" ||
-        Name == "power9-vector") {
-      // power9-vector is really a superset of power8-vector so encode that.
-      Features[Name] = Features["vsx"] = true;
-      if (Name == "power9-vector")
-        Features["power8-vector"] = true;
-    } else {
-      Features[Name] = true;
-    }
+    // If we're enabling any of the vsx based features then enable vsx and
+    // altivec. We'll diagnose any problems later.
+    bool FeatureHasVSX = llvm::StringSwitch<bool>(Name)
+                             .Case("vsx", true)
+                             .Case("direct-move", true)
+                             .Case("power8-vector", true)
+                             .Case("power9-vector", true)
+                             .Case("float128", true)
+                             .Default(false);
+    if (FeatureHasVSX)
+      Features["vsx"] = Features["altivec"] = true;
+    if (Name == "power9-vector")
+      Features["power8-vector"] = true;
+    Features[Name] = true;
   } else {
-    if (Name == "vsx") {
-      Features[Name] = Features["direct-move"] = Features["power8-vector"] =
+    // If we're disabling altivec or vsx go ahead and disable all of the vsx
+    // features.
+    if ((Name == "altivec") || (Name == "vsx"))
+      Features["vsx"] = Features["direct-move"] = Features["power8-vector"] =
           Features["float128"] = Features["power9-vector"] = false;
-    } else {
-      Features[Name] = false;
-    }
+    if (Name == "power8-vector")
+      Features["power9-vector"] = false;
+    Features[Name] = false;
   }
 }
 
@@ -1718,7 +1762,6 @@ public:
     BoolWidth = BoolAlign = 32; //XXX support -mone-byte-bool?
     PtrDiffType = SignedInt; // for http://llvm.org/bugs/show_bug.cgi?id=15726
     LongLongAlign = 32;
-    SuitableAlign = 128;
     resetDataLayout("E-m:o-p:32:32-f64:32:64-n32");
   }
   BuiltinVaListKind getBuiltinVaListKind() const override {
@@ -1731,12 +1774,12 @@ public:
   DarwinPPC64TargetInfo(const llvm::Triple &Triple, const TargetOptions &Opts)
       : DarwinTargetInfo<PPC64TargetInfo>(Triple, Opts) {
     HasAlignMac68kSupport = true;
-    SuitableAlign = 128;
     resetDataLayout("E-m:o-i64:64-n32:64");
   }
 };
 
 static const unsigned NVPTXAddrSpaceMap[] = {
+    0, // Default
     1, // opencl_global
     3, // opencl_local
     4, // opencl_constant
@@ -1990,14 +2033,25 @@ ArrayRef<const char *> NVPTXTargetInfo::getGCCRegNames() const {
   return llvm::makeArrayRef(GCCRegNames);
 }
 
-static const unsigned AMDGPUAddrSpaceMap[] = {
-  1,    // opencl_global
-  3,    // opencl_local
-  2,    // opencl_constant
-  4,    // opencl_generic
-  1,    // cuda_device
-  2,    // cuda_constant
-  3     // cuda_shared
+static const LangAS::Map AMDGPUPrivateIsZeroMap = {
+    4,  // Default
+    1,  // opencl_global
+    3,  // opencl_local
+    2,  // opencl_constant
+    4,  // opencl_generic
+    1,  // cuda_device
+    2,  // cuda_constant
+    3   // cuda_shared
+};
+static const LangAS::Map AMDGPUGenericIsZeroMap = {
+    0,  // Default
+    1,  // opencl_global
+    3,  // opencl_local
+    2,  // opencl_constant
+    0,  // opencl_generic
+    1,  // cuda_device
+    2,  // cuda_constant
+    3   // cuda_shared
 };
 
 // If you edit the description strings, make sure you update
@@ -2007,15 +2061,39 @@ static const char *const DataLayoutStringR600 =
   "e-p:32:32-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128"
   "-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64";
 
-static const char *const DataLayoutStringSI =
+static const char *const DataLayoutStringSIPrivateIsZero =
   "e-p:32:32-p1:64:64-p2:64:64-p3:32:32-p4:64:64-p5:32:32"
   "-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128"
   "-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64";
 
+static const char *const DataLayoutStringSIGenericIsZero =
+  "e-p:64:64-p1:64:64-p2:64:64-p3:32:32-p4:32:32-p5:32:32"
+  "-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128"
+  "-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64-A5";
+
 class AMDGPUTargetInfo final : public TargetInfo {
   static const Builtin::Info BuiltinInfo[];
   static const char * const GCCRegNames[];
 
+  struct AddrSpace {
+    unsigned Generic, Global, Local, Constant, Private;
+    AddrSpace(bool IsGenericZero_ = false){
+      if (IsGenericZero_) {
+        Generic   = 0;
+        Global    = 1;
+        Local     = 3;
+        Constant  = 2;
+        Private   = 5;
+      } else {
+        Generic   = 4;
+        Global    = 1;
+        Local     = 3;
+        Constant  = 2;
+        Private   = 0;
+      }
+    }
+  };
+
   /// \brief The GPU profiles supported by the AMDGPU target.
   enum GPUKind {
     GK_NONE,
@@ -2029,18 +2107,27 @@ class AMDGPUTargetInfo final : public TargetInfo {
     GK_CAYMAN,
     GK_GFX6,
     GK_GFX7,
-    GK_GFX8
+    GK_GFX8,
+    GK_GFX9
   } GPU;
 
   bool hasFP64:1;
   bool hasFMAF:1;
   bool hasLDEXPF:1;
-  bool hasFullSpeedFP32Denorms:1;
+  const AddrSpace AS;
+
+  static bool hasFullSpeedFMAF32(StringRef GPUName) {
+    return parseAMDGCNName(GPUName) >= GK_GFX9;
+  }
 
   static bool isAMDGCN(const llvm::Triple &TT) {
     return TT.getArch() == llvm::Triple::amdgcn;
   }
 
+  static bool isGenericZero(const llvm::Triple &TT) {
+    return TT.getEnvironmentName() == "amdgiz" ||
+        TT.getEnvironmentName() == "amdgizcl";
+  }
 public:
   AMDGPUTargetInfo(const llvm::Triple &Triple, const TargetOptions &Opts)
     : TargetInfo(Triple) ,
@@ -2048,17 +2135,21 @@ public:
       hasFP64(false),
       hasFMAF(false),
       hasLDEXPF(false),
-      hasFullSpeedFP32Denorms(false){
+      AS(isGenericZero(Triple)){
     if (getTriple().getArch() == llvm::Triple::amdgcn) {
       hasFP64 = true;
       hasFMAF = true;
       hasLDEXPF = true;
     }
-
+    auto IsGenericZero = isGenericZero(Triple);
     resetDataLayout(getTriple().getArch() == llvm::Triple::amdgcn ?
-                    DataLayoutStringSI : DataLayoutStringR600);
+                    (IsGenericZero ? DataLayoutStringSIGenericIsZero :
+                        DataLayoutStringSIPrivateIsZero)
+                    : DataLayoutStringR600);
+    assert(DataLayout->getAllocaAddrSpace() == AS.Private);
 
-    AddrSpaceMap = &AMDGPUAddrSpaceMap;
+    AddrSpaceMap = IsGenericZero ? &AMDGPUGenericIsZeroMap :
+        &AMDGPUPrivateIsZeroMap;
     UseAddrSpaceMapMangling = true;
   }
 
@@ -2066,14 +2157,10 @@ public:
     if (GPU <= GK_CAYMAN)
       return 32;
 
-    switch(AddrSpace) {
-      default:
-        return 64;
-      case 0:
-      case 3:
-      case 5:
-        return 32;
+    if (AddrSpace == AS.Private || AddrSpace == AS.Local) {
+      return 32;
     }
+    return 64;
   }
 
   uint64_t getMaxPointerWidth() const override {
@@ -2113,15 +2200,16 @@ public:
     for (auto &I : TargetOpts.FeaturesAsWritten) {
       if (I == "+fp32-denormals" || I == "-fp32-denormals")
         hasFP32Denormals = true;
-      if (I == "+fp64-denormals" || I == "-fp64-denormals")
+      if (I == "+fp64-fp16-denormals" || I == "-fp64-fp16-denormals")
         hasFP64Denormals = true;
     }
     if (!hasFP32Denormals)
-      TargetOpts.Features.push_back((Twine(hasFullSpeedFP32Denorms &&
+      TargetOpts.Features.push_back(
+          (Twine(hasFullSpeedFMAF32(TargetOpts.CPU) &&
           !CGOpts.FlushDenorm ? '+' : '-') + Twine("fp32-denormals")).str());
-    // Always do not flush fp64 denorms.
+    // Always do not flush fp64 or fp16 denorms.
     if (!hasFP64Denormals && hasFP64)
-      TargetOpts.Features.push_back("+fp64-denormals");
+      TargetOpts.Features.push_back("+fp64-fp16-denormals");
   }
 
   ArrayRef<Builtin::Info> getTargetBuiltins() const override {
@@ -2206,6 +2294,8 @@ public:
       .Case("gfx803",    GK_GFX8)
       .Case("gfx804",    GK_GFX8)
       .Case("gfx810",    GK_GFX8)
+      .Case("gfx900",    GK_GFX9)
+      .Case("gfx901",    GK_GFX9)
       .Default(GK_NONE);
   }
 
@@ -2248,6 +2338,33 @@ public:
     return LangAS::opencl_constant;
   }
 
+  /// \returns Target specific vtbl ptr address space.
+  unsigned getVtblPtrAddressSpace() const override {
+    // \todo: We currently have address spaces defined in AMDGPU Backend. It
+    // would be nice if we could use it here instead of using bare numbers (same
+    // applies to getDWARFAddressSpace).
+    return 2; // constant.
+  }
+
+  /// \returns If a target requires an address within a target specific address
+  /// space \p AddressSpace to be converted in order to be used, then return the
+  /// corresponding target specific DWARF address space.
+  ///
+  /// \returns Otherwise return None and no conversion will be emitted in the
+  /// DWARF.
+  Optional<unsigned> getDWARFAddressSpace(
+      unsigned AddressSpace) const override {
+    const unsigned DWARF_Private = 1;
+    const unsigned DWARF_Local   = 2;
+    if (AddressSpace == AS.Private) {
+      return DWARF_Private;
+    } else if (AddressSpace == AS.Local) {
+      return DWARF_Local;
+    } else {
+      return None;
+    }
+  }
+
   CallingConvCheckResult checkCallingConvention(CallingConv CC) const override {
     switch (CC) {
       default:
@@ -2262,7 +2379,7 @@ public:
   // address space has value 0 but in private and local address space has
   // value ~0.
   uint64_t getNullPointerValue(unsigned AS) const override {
-    return AS != LangAS::opencl_local && AS != 0 ? 0 : ~0;
+    return AS == LangAS::opencl_local ? ~0 : 0;
   }
 };
 
@@ -2345,9 +2462,13 @@ bool AMDGPUTargetInfo::initFeatureMap(
     case GK_GFX7:
       break;
 
+    case GK_GFX9:
+      Features["gfx9-insts"] = true;
+      LLVM_FALLTHROUGH;
     case GK_GFX8:
       Features["s-memrealtime"] = true;
       Features["16-bit-insts"] = true;
+      Features["dpp"] = true;
       break;
 
     case GK_NONE:
@@ -2489,11 +2610,10 @@ class X86TargetInfo : public TargetInfo {
   bool HasXSAVEC = false;
   bool HasXSAVES = false;
   bool HasMWAITX = false;
+  bool HasCLZERO = false;
   bool HasPKU = false;
   bool HasCLFLUSHOPT = false;
-  bool HasPCOMMIT = false;
   bool HasCLWB = false;
-  bool HasUMIP = false;
   bool HasMOVBE = false;
   bool HasPREFETCHWT1 = false;
 
@@ -3067,8 +3187,7 @@ bool X86TargetInfo::initFeatureMap(
     setFeatureEnabledImpl(Features, "avx512ifma", true);
     setFeatureEnabledImpl(Features, "avx512vbmi", true);
     setFeatureEnabledImpl(Features, "sha", true);
-    setFeatureEnabledImpl(Features, "umip", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_SkylakeServer:
     setFeatureEnabledImpl(Features, "avx512f", true);
     setFeatureEnabledImpl(Features, "avx512cd", true);
@@ -3076,44 +3195,43 @@ bool X86TargetInfo::initFeatureMap(
     setFeatureEnabledImpl(Features, "avx512bw", true);
     setFeatureEnabledImpl(Features, "avx512vl", true);
     setFeatureEnabledImpl(Features, "pku", true);
-    setFeatureEnabledImpl(Features, "pcommit", true);
     setFeatureEnabledImpl(Features, "clwb", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_SkylakeClient:
     setFeatureEnabledImpl(Features, "xsavec", true);
     setFeatureEnabledImpl(Features, "xsaves", true);
     setFeatureEnabledImpl(Features, "mpx", true);
     setFeatureEnabledImpl(Features, "sgx", true);
     setFeatureEnabledImpl(Features, "clflushopt", true);
-    // FALLTHROUGH
+    setFeatureEnabledImpl(Features, "rtm", true);
+    LLVM_FALLTHROUGH;
   case CK_Broadwell:
     setFeatureEnabledImpl(Features, "rdseed", true);
     setFeatureEnabledImpl(Features, "adx", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_Haswell:
     setFeatureEnabledImpl(Features, "avx2", true);
     setFeatureEnabledImpl(Features, "lzcnt", true);
     setFeatureEnabledImpl(Features, "bmi", true);
     setFeatureEnabledImpl(Features, "bmi2", true);
-    setFeatureEnabledImpl(Features, "rtm", true);
     setFeatureEnabledImpl(Features, "fma", true);
     setFeatureEnabledImpl(Features, "movbe", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_IvyBridge:
     setFeatureEnabledImpl(Features, "rdrnd", true);
     setFeatureEnabledImpl(Features, "f16c", true);
     setFeatureEnabledImpl(Features, "fsgsbase", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_SandyBridge:
     setFeatureEnabledImpl(Features, "avx", true);
     setFeatureEnabledImpl(Features, "xsave", true);
     setFeatureEnabledImpl(Features, "xsaveopt", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_Westmere:
   case CK_Silvermont:
     setFeatureEnabledImpl(Features, "aes", true);
     setFeatureEnabledImpl(Features, "pclmul", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_Nehalem:
     setFeatureEnabledImpl(Features, "sse4.2", true);
     setFeatureEnabledImpl(Features, "fxsr", true);
@@ -3173,7 +3291,7 @@ bool X86TargetInfo::initFeatureMap(
     setFeatureEnabledImpl(Features, "sse4a", true);
     setFeatureEnabledImpl(Features, "lzcnt", true);
     setFeatureEnabledImpl(Features, "popcnt", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_K8SSE3:
   case CK_OpteronSSE3:
   case CK_Athlon64SSE3:
@@ -3188,7 +3306,7 @@ bool X86TargetInfo::initFeatureMap(
     setFeatureEnabledImpl(Features, "bmi", true);
     setFeatureEnabledImpl(Features, "f16c", true);
     setFeatureEnabledImpl(Features, "xsaveopt", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_BTVER1:
     setFeatureEnabledImpl(Features, "ssse3", true);
     setFeatureEnabledImpl(Features, "sse4a", true);
@@ -3205,6 +3323,7 @@ bool X86TargetInfo::initFeatureMap(
     setFeatureEnabledImpl(Features, "bmi", true);
     setFeatureEnabledImpl(Features, "bmi2", true);
     setFeatureEnabledImpl(Features, "clflushopt", true);
+    setFeatureEnabledImpl(Features, "clzero", true);
     setFeatureEnabledImpl(Features, "cx16", true);
     setFeatureEnabledImpl(Features, "f16c", true);
     setFeatureEnabledImpl(Features, "fma", true);
@@ -3229,17 +3348,17 @@ bool X86TargetInfo::initFeatureMap(
     setFeatureEnabledImpl(Features, "avx2", true);
     setFeatureEnabledImpl(Features, "bmi2", true);
     setFeatureEnabledImpl(Features, "mwaitx", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_BDVER3:
     setFeatureEnabledImpl(Features, "fsgsbase", true);
     setFeatureEnabledImpl(Features, "xsaveopt", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_BDVER2:
     setFeatureEnabledImpl(Features, "bmi", true);
     setFeatureEnabledImpl(Features, "fma", true);
     setFeatureEnabledImpl(Features, "f16c", true);
     setFeatureEnabledImpl(Features, "tbm", true);
-    // FALLTHROUGH
+    LLVM_FALLTHROUGH;
   case CK_BDVER1:
     // xop implies avx, sse4a and fma4.
     setFeatureEnabledImpl(Features, "xop", true);
@@ -3560,14 +3679,12 @@ bool X86TargetInfo::handleTargetFeatures(std::vector<std::string> &Features,
       HasPKU = true;
     } else if (Feature == "+clflushopt") {
       HasCLFLUSHOPT = true;
-    } else if (Feature == "+pcommit") {
-      HasPCOMMIT = true;
     } else if (Feature == "+clwb") {
       HasCLWB = true;
-    } else if (Feature == "+umip") {
-      HasUMIP = true;
     } else if (Feature == "+prefetchwt1") {
       HasPREFETCHWT1 = true;
+    } else if (Feature == "+clzero") {
+      HasCLZERO = true;
     }
 
     X86SSEEnum Level = llvm::StringSwitch<X86SSEEnum>(Feature)
@@ -3885,6 +4002,18 @@ void X86TargetInfo::getTargetDefines(const LangOptions &Opts,
     Builder.defineMacro("__PKU__");
   if (HasCX16)
     Builder.defineMacro("__GCC_HAVE_SYNC_COMPARE_AND_SWAP_16");
+  if (HasCLFLUSHOPT)
+    Builder.defineMacro("__CLFLUSHOPT__");
+  if (HasCLWB)
+    Builder.defineMacro("__CLWB__");
+  if (HasMPX)
+    Builder.defineMacro("__MPX__");
+  if (HasSGX)
+    Builder.defineMacro("__SGX__");
+  if (HasPREFETCHWT1)
+    Builder.defineMacro("__PREFETCHWT1__");
+  if (HasCLZERO)
+    Builder.defineMacro("__CLZERO__");
 
   // Each case falls through to the previous one here.
   switch (SSELevel) {
@@ -3971,6 +4100,7 @@ bool X86TargetInfo::hasFeature(StringRef Feature) const {
       .Case("bmi2", HasBMI2)
       .Case("clflushopt", HasCLFLUSHOPT)
       .Case("clwb", HasCLWB)
+      .Case("clzero", HasCLZERO)
       .Case("cx16", HasCX16)
       .Case("f16c", HasF16C)
       .Case("fma", HasFMA)
@@ -3984,7 +4114,6 @@ bool X86TargetInfo::hasFeature(StringRef Feature) const {
       .Case("movbe", HasMOVBE)
       .Case("mpx", HasMPX)
       .Case("pclmul", HasPCLMUL)
-      .Case("pcommit", HasPCOMMIT)
       .Case("pku", HasPKU)
       .Case("popcnt", HasPOPCNT)
       .Case("prefetchwt1", HasPREFETCHWT1)
@@ -4002,7 +4131,6 @@ bool X86TargetInfo::hasFeature(StringRef Feature) const {
       .Case("sse4.2", SSELevel >= SSE42)
       .Case("sse4a", XOPLevel >= SSE4A)
       .Case("tbm", HasTBM)
-      .Case("umip", HasUMIP)
       .Case("x86", true)
       .Case("x86_32", getTriple().getArch() == llvm::Triple::x86)
       .Case("x86_64", getTriple().getArch() == llvm::Triple::x86_64)
@@ -5056,6 +5184,8 @@ class ARMTargetInfo : public TargetInfo {
       return "7M";
     case llvm::ARM::AK_ARMV7EM:
       return "7EM";
+    case llvm::ARM::AK_ARMV7VE:
+      return "7VE";
     case llvm::ARM::AK_ARMV8A:
       return "8A";
     case llvm::ARM::AK_ARMV8_1A:
@@ -5144,6 +5274,8 @@ public:
       default:
         if (Triple.getOS() == llvm::Triple::NetBSD)
           setABI("apcs-gnu");
+        else if (Triple.getOS() == llvm::Triple::OpenBSD)
+          setABI("aapcs-linux");
         else
           setABI("aapcs");
         break;
@@ -5925,7 +6057,8 @@ class AArch64TargetInfo : public TargetInfo {
 public:
   AArch64TargetInfo(const llvm::Triple &Triple, const TargetOptions &Opts)
       : TargetInfo(Triple), ABI("aapcs") {
-    if (getTriple().getOS() == llvm::Triple::NetBSD) {
+    if (getTriple().getOS() == llvm::Triple::NetBSD ||
+        getTriple().getOS() == llvm::Triple::OpenBSD) {
       WCharType = SignedInt;
 
       // NetBSD apparently prefers consistency across ARM targets to consistency
@@ -5960,8 +6093,9 @@ public:
     // AArch64 targets default to using the ARM C++ ABI.
     TheCXXABI.set(TargetCXXABI::GenericAArch64);
 
-    if (Triple.getOS() == llvm::Triple::Linux ||
-        Triple.getOS() == llvm::Triple::UnknownOS)
+    if (Triple.getOS() == llvm::Triple::Linux)
+      this->MCountName = "\01_mcount";
+    else if (Triple.getOS() == llvm::Triple::UnknownOS)
       this->MCountName = Opts.EABIVersion == "gnu" ? "\01_mcount" : "mcount";
   }
 
@@ -6408,6 +6542,7 @@ public:
       .Case("hexagonv5", "5")
       .Case("hexagonv55", "55")
       .Case("hexagonv60", "60")
+      .Case("hexagonv62", "62")
       .Default(nullptr);
   }
 
@@ -6452,6 +6587,9 @@ void HexagonTargetInfo::getTargetDefines(const LangOptions &Opts,
     Builder.defineMacro("__HEXAGON_ARCH__", "60");
     Builder.defineMacro("__QDSP6_V60__");
     Builder.defineMacro("__QDSP6_ARCH__", "60");
+  } else if (CPU == "hexagonv62") {
+    Builder.defineMacro("__HEXAGON_V62__");
+    Builder.defineMacro("__HEXAGON_ARCH__", "62");
   }
 
   if (hasFeature("hvx")) {
@@ -7037,6 +7175,15 @@ public:
     Builder.defineMacro("__zarch__");
     Builder.defineMacro("__LONG_DOUBLE_128__");
 
+    const std::string ISARev = llvm::StringSwitch<std::string>(CPU)
+                                   .Cases("arch8", "z10", "8")
+                                   .Cases("arch9", "z196", "9")
+                                   .Cases("arch10", "zEC12", "10")
+                                   .Cases("arch11", "z13", "11")
+                                   .Default("");
+    if (!ISARev.empty())
+      Builder.defineMacro("__ARCH__", ISARev);
+
     Builder.defineMacro("__GCC_HAVE_SYNC_COMPARE_AND_SWAP_1");
     Builder.defineMacro("__GCC_HAVE_SYNC_COMPARE_AND_SWAP_2");
     Builder.defineMacro("__GCC_HAVE_SYNC_COMPARE_AND_SWAP_4");
@@ -7044,6 +7191,8 @@ public:
 
     if (HasTransactionalExecution)
       Builder.defineMacro("__HTM__");
+    if (HasVector)
+      Builder.defineMacro("__VX__");
     if (Opts.ZVector)
       Builder.defineMacro("__VEC__", "10301");
   }
@@ -7268,6 +7417,7 @@ ArrayRef<const char *> MSP430TargetInfo::getGCCRegNames() const {
 // publicly available in http://tce.cs.tut.fi
 
 static const unsigned TCEOpenCLAddrSpaceMap[] = {
+    0, // Default
     3, // opencl_global
     4, // opencl_local
     5, // opencl_constant
@@ -7434,6 +7584,8 @@ class MipsTargetInfo : public TargetInfo {
   bool IsMicromips;
   bool IsNan2008;
   bool IsSingleFloat;
+  bool IsNoABICalls;
+  bool CanUseBSDABICalls;
   enum MipsFloatABI {
     HardFloat, SoftFloat
   } FloatABI;
@@ -7449,8 +7601,9 @@ protected:
 public:
   MipsTargetInfo(const llvm::Triple &Triple, const TargetOptions &)
       : TargetInfo(Triple), IsMips16(false), IsMicromips(false),
-        IsNan2008(false), IsSingleFloat(false), FloatABI(HardFloat),
-        DspRev(NoDSP), HasMSA(false), HasFP64(false) {
+        IsNan2008(false), IsSingleFloat(false), IsNoABICalls(false),
+        CanUseBSDABICalls(false), FloatABI(HardFloat), DspRev(NoDSP),
+        HasMSA(false), HasFP64(false) {
     TheCXXABI.set(TargetCXXABI::GenericMIPS);
 
     setABI((getTriple().getArch() == llvm::Triple::mips ||
@@ -7459,6 +7612,9 @@ public:
                : "n64");
 
     CPU = ABI == "o32" ? "mips32r2" : "mips64r2";
+
+    CanUseBSDABICalls = Triple.getOS() == llvm::Triple::FreeBSD ||
+                        Triple.getOS() == llvm::Triple::OpenBSD;
   }
 
   bool isNaN2008Default() const {
@@ -7535,7 +7691,11 @@ public:
 
   void setN64ABITypes() {
     setN32N64ABITypes();
-    Int64Type = SignedLong;
+    if (getTriple().getOS() == llvm::Triple::OpenBSD) {
+      Int64Type = SignedLongLong;
+    } else {
+      Int64Type = SignedLong;
+    }
     IntMaxType = Int64Type;
     LongWidth = LongAlign = 64;
     PointerWidth = PointerAlign = 64;
@@ -7639,6 +7799,12 @@ public:
     } else
       llvm_unreachable("Invalid ABI.");
 
+    if (!IsNoABICalls) {
+      Builder.defineMacro("__mips_abicalls");
+      if (CanUseBSDABICalls)
+        Builder.defineMacro("__ABICALLS__");
+    }
+
     Builder.defineMacro("__REGISTER_PREFIX__", "");
 
     switch (FloatABI) {
@@ -7853,6 +8019,8 @@ public:
         IsNan2008 = true;
       else if (Feature == "-nan2008")
         IsNan2008 = false;
+      else if (Feature == "+noabicalls")
+        IsNoABICalls = true;
     }
 
     setDataLayout();
@@ -8216,6 +8384,7 @@ const Builtin::Info Le64TargetInfo::BuiltinInfo[] = {
 };
 
 static const unsigned SPIRAddrSpaceMap[] = {
+    0, // Default
     1, // opencl_global
     3, // opencl_local
     2, // opencl_constant
@@ -8435,6 +8604,254 @@ public:
   }
 };
 
+/// Information about a specific microcontroller.
+struct MCUInfo {
+  const char *Name;
+  const char *DefineName;
+};
+
+// This list should be kept up-to-date with AVRDevices.td in LLVM.
+static ArrayRef<MCUInfo> AVRMcus = {
+  { "at90s1200", "__AVR_AT90S1200__" },
+  { "attiny11", "__AVR_ATtiny11__" },
+  { "attiny12", "__AVR_ATtiny12__" },
+  { "attiny15", "__AVR_ATtiny15__" },
+  { "attiny28", "__AVR_ATtiny28__" },
+  { "at90s2313", "__AVR_AT90S2313__" },
+  { "at90s2323", "__AVR_AT90S2323__" },
+  { "at90s2333", "__AVR_AT90S2333__" },
+  { "at90s2343", "__AVR_AT90S2343__" },
+  { "attiny22", "__AVR_ATtiny22__" },
+  { "attiny26", "__AVR_ATtiny26__" },
+  { "at86rf401", "__AVR_AT86RF401__" },
+  { "at90s4414", "__AVR_AT90S4414__" },
+  { "at90s4433", "__AVR_AT90S4433__" },
+  { "at90s4434", "__AVR_AT90S4434__" },
+  { "at90s8515", "__AVR_AT90S8515__" },
+  { "at90c8534", "__AVR_AT90c8534__" },
+  { "at90s8535", "__AVR_AT90S8535__" },
+  { "ata5272", "__AVR_ATA5272__" },
+  { "attiny13", "__AVR_ATtiny13__" },
+  { "attiny13a", "__AVR_ATtiny13A__" },
+  { "attiny2313", "__AVR_ATtiny2313__" },
+  { "attiny2313a", "__AVR_ATtiny2313A__" },
+  { "attiny24", "__AVR_ATtiny24__" },
+  { "attiny24a", "__AVR_ATtiny24A__" },
+  { "attiny4313", "__AVR_ATtiny4313__" },
+  { "attiny44", "__AVR_ATtiny44__" },
+  { "attiny44a", "__AVR_ATtiny44A__" },
+  { "attiny84", "__AVR_ATtiny84__" },
+  { "attiny84a", "__AVR_ATtiny84A__" },
+  { "attiny25", "__AVR_ATtiny25__" },
+  { "attiny45", "__AVR_ATtiny45__" },
+  { "attiny85", "__AVR_ATtiny85__" },
+  { "attiny261", "__AVR_ATtiny261__" },
+  { "attiny261a", "__AVR_ATtiny261A__" },
+  { "attiny461", "__AVR_ATtiny461__" },
+  { "attiny461a", "__AVR_ATtiny461A__" },
+  { "attiny861", "__AVR_ATtiny861__" },
+  { "attiny861a", "__AVR_ATtiny861A__" },
+  { "attiny87", "__AVR_ATtiny87__" },
+  { "attiny43u", "__AVR_ATtiny43U__" },
+  { "attiny48", "__AVR_ATtiny48__" },
+  { "attiny88", "__AVR_ATtiny88__" },
+  { "attiny828", "__AVR_ATtiny828__" },
+  { "at43usb355", "__AVR_AT43USB355__" },
+  { "at76c711", "__AVR_AT76C711__" },
+  { "atmega103", "__AVR_ATmega103__" },
+  { "at43usb320", "__AVR_AT43USB320__" },
+  { "attiny167", "__AVR_ATtiny167__" },
+  { "at90usb82", "__AVR_AT90USB82__" },
+  { "at90usb162", "__AVR_AT90USB162__" },
+  { "ata5505", "__AVR_ATA5505__" },
+  { "atmega8u2", "__AVR_ATmega8U2__" },
+  { "atmega16u2", "__AVR_ATmega16U2__" },
+  { "atmega32u2", "__AVR_ATmega32U2__" },
+  { "attiny1634", "__AVR_ATtiny1634__" },
+  { "atmega8", "__AVR_ATmega8__" },
+  { "ata6289", "__AVR_ATA6289__" },
+  { "atmega8a", "__AVR_ATmega8A__" },
+  { "ata6285", "__AVR_ATA6285__" },
+  { "ata6286", "__AVR_ATA6286__" },
+  { "atmega48", "__AVR_ATmega48__" },
+  { "atmega48a", "__AVR_ATmega48A__" },
+  { "atmega48pa", "__AVR_ATmega48PA__" },
+  { "atmega48p", "__AVR_ATmega48P__" },
+  { "atmega88", "__AVR_ATmega88__" },
+  { "atmega88a", "__AVR_ATmega88A__" },
+  { "atmega88p", "__AVR_ATmega88P__" },
+  { "atmega88pa", "__AVR_ATmega88PA__" },
+  { "atmega8515", "__AVR_ATmega8515__" },
+  { "atmega8535", "__AVR_ATmega8535__" },
+  { "atmega8hva", "__AVR_ATmega8HVA__" },
+  { "at90pwm1", "__AVR_AT90PWM1__" },
+  { "at90pwm2", "__AVR_AT90PWM2__" },
+  { "at90pwm2b", "__AVR_AT90PWM2B__" },
+  { "at90pwm3", "__AVR_AT90PWM3__" },
+  { "at90pwm3b", "__AVR_AT90PWM3B__" },
+  { "at90pwm81", "__AVR_AT90PWM81__" },
+  { "ata5790", "__AVR_ATA5790__" },
+  { "ata5795", "__AVR_ATA5795__" },
+  { "atmega16", "__AVR_ATmega16__" },
+  { "atmega16a", "__AVR_ATmega16A__" },
+  { "atmega161", "__AVR_ATmega161__" },
+  { "atmega162", "__AVR_ATmega162__" },
+  { "atmega163", "__AVR_ATmega163__" },
+  { "atmega164a", "__AVR_ATmega164A__" },
+  { "atmega164p", "__AVR_ATmega164P__" },
+  { "atmega164pa", "__AVR_ATmega164PA__" },
+  { "atmega165", "__AVR_ATmega165__" },
+  { "atmega165a", "__AVR_ATmega165A__" },
+  { "atmega165p", "__AVR_ATmega165P__" },
+  { "atmega165pa", "__AVR_ATmega165PA__" },
+  { "atmega168", "__AVR_ATmega168__" },
+  { "atmega168a", "__AVR_ATmega168A__" },
+  { "atmega168p", "__AVR_ATmega168P__" },
+  { "atmega168pa", "__AVR_ATmega168PA__" },
+  { "atmega169", "__AVR_ATmega169__" },
+  { "atmega169a", "__AVR_ATmega169A__" },
+  { "atmega169p", "__AVR_ATmega169P__" },
+  { "atmega169pa", "__AVR_ATmega169PA__" },
+  { "atmega32", "__AVR_ATmega32__" },
+  { "atmega32a", "__AVR_ATmega32A__" },
+  { "atmega323", "__AVR_ATmega323__" },
+  { "atmega324a", "__AVR_ATmega324A__" },
+  { "atmega324p", "__AVR_ATmega324P__" },
+  { "atmega324pa", "__AVR_ATmega324PA__" },
+  { "atmega325", "__AVR_ATmega325__" },
+  { "atmega325a", "__AVR_ATmega325A__" },
+  { "atmega325p", "__AVR_ATmega325P__" },
+  { "atmega325pa", "__AVR_ATmega325PA__" },
+  { "atmega3250", "__AVR_ATmega3250__" },
+  { "atmega3250a", "__AVR_ATmega3250A__" },
+  { "atmega3250p", "__AVR_ATmega3250P__" },
+  { "atmega3250pa", "__AVR_ATmega3250PA__" },
+  { "atmega328", "__AVR_ATmega328__" },
+  { "atmega328p", "__AVR_ATmega328P__" },
+  { "atmega329", "__AVR_ATmega329__" },
+  { "atmega329a", "__AVR_ATmega329A__" },
+  { "atmega329p", "__AVR_ATmega329P__" },
+  { "atmega329pa", "__AVR_ATmega329PA__" },
+  { "atmega3290", "__AVR_ATmega3290__" },
+  { "atmega3290a", "__AVR_ATmega3290A__" },
+  { "atmega3290p", "__AVR_ATmega3290P__" },
+  { "atmega3290pa", "__AVR_ATmega3290PA__" },
+  { "atmega406", "__AVR_ATmega406__" },
+  { "atmega64", "__AVR_ATmega64__" },
+  { "atmega64a", "__AVR_ATmega64A__" },
+  { "atmega640", "__AVR_ATmega640__" },
+  { "atmega644", "__AVR_ATmega644__" },
+  { "atmega644a", "__AVR_ATmega644A__" },
+  { "atmega644p", "__AVR_ATmega644P__" },
+  { "atmega644pa", "__AVR_ATmega644PA__" },
+  { "atmega645", "__AVR_ATmega645__" },
+  { "atmega645a", "__AVR_ATmega645A__" },
+  { "atmega645p", "__AVR_ATmega645P__" },
+  { "atmega649", "__AVR_ATmega649__" },
+  { "atmega649a", "__AVR_ATmega649A__" },
+  { "atmega649p", "__AVR_ATmega649P__" },
+  { "atmega6450", "__AVR_ATmega6450__" },
+  { "atmega6450a", "__AVR_ATmega6450A__" },
+  { "atmega6450p", "__AVR_ATmega6450P__" },
+  { "atmega6490", "__AVR_ATmega6490__" },
+  { "atmega6490a", "__AVR_ATmega6490A__" },
+  { "atmega6490p", "__AVR_ATmega6490P__" },
+  { "atmega64rfr2", "__AVR_ATmega64RFR2__" },
+  { "atmega644rfr2", "__AVR_ATmega644RFR2__" },
+  { "atmega16hva", "__AVR_ATmega16HVA__" },
+  { "atmega16hva2", "__AVR_ATmega16HVA2__" },
+  { "atmega16hvb", "__AVR_ATmega16HVB__" },
+  { "atmega16hvbrevb", "__AVR_ATmega16HVBREVB__" },
+  { "atmega32hvb", "__AVR_ATmega32HVB__" },
+  { "atmega32hvbrevb", "__AVR_ATmega32HVBREVB__" },
+  { "atmega64hve", "__AVR_ATmega64HVE__" },
+  { "at90can32", "__AVR_AT90CAN32__" },
+  { "at90can64", "__AVR_AT90CAN64__" },
+  { "at90pwm161", "__AVR_AT90PWM161__" },
+  { "at90pwm216", "__AVR_AT90PWM216__" },
+  { "at90pwm316", "__AVR_AT90PWM316__" },
+  { "atmega32c1", "__AVR_ATmega32C1__" },
+  { "atmega64c1", "__AVR_ATmega64C1__" },
+  { "atmega16m1", "__AVR_ATmega16M1__" },
+  { "atmega32m1", "__AVR_ATmega32M1__" },
+  { "atmega64m1", "__AVR_ATmega64M1__" },
+  { "atmega16u4", "__AVR_ATmega16U4__" },
+  { "atmega32u4", "__AVR_ATmega32U4__" },
+  { "atmega32u6", "__AVR_ATmega32U6__" },
+  { "at90usb646", "__AVR_AT90USB646__" },
+  { "at90usb647", "__AVR_AT90USB647__" },
+  { "at90scr100", "__AVR_AT90SCR100__" },
+  { "at94k", "__AVR_AT94K__" },
+  { "m3000", "__AVR_AT000__" },
+  { "atmega128", "__AVR_ATmega128__" },
+  { "atmega128a", "__AVR_ATmega128A__" },
+  { "atmega1280", "__AVR_ATmega1280__" },
+  { "atmega1281", "__AVR_ATmega1281__" },
+  { "atmega1284", "__AVR_ATmega1284__" },
+  { "atmega1284p", "__AVR_ATmega1284P__" },
+  { "atmega128rfa1", "__AVR_ATmega128RFA1__" },
+  { "atmega128rfr2", "__AVR_ATmega128RFR2__" },
+  { "atmega1284rfr2", "__AVR_ATmega1284RFR2__" },
+  { "at90can128", "__AVR_AT90CAN128__" },
+  { "at90usb1286", "__AVR_AT90USB1286__" },
+  { "at90usb1287", "__AVR_AT90USB1287__" },
+  { "atmega2560", "__AVR_ATmega2560__" },
+  { "atmega2561", "__AVR_ATmega2561__" },
+  { "atmega256rfr2", "__AVR_ATmega256RFR2__" },
+  { "atmega2564rfr2", "__AVR_ATmega2564RFR2__" },
+  { "atxmega16a4", "__AVR_ATxmega16A4__" },
+  { "atxmega16a4u", "__AVR_ATxmega16a4U__" },
+  { "atxmega16c4", "__AVR_ATxmega16C4__" },
+  { "atxmega16d4", "__AVR_ATxmega16D4__" },
+  { "atxmega32a4", "__AVR_ATxmega32A4__" },
+  { "atxmega32a4u", "__AVR_ATxmega32A4U__" },
+  { "atxmega32c4", "__AVR_ATxmega32C4__" },
+  { "atxmega32d4", "__AVR_ATxmega32D4__" },
+  { "atxmega32e5", "__AVR_ATxmega32E5__" },
+  { "atxmega16e5", "__AVR_ATxmega16E5__" },
+  { "atxmega8e5", "__AVR_ATxmega8E5__" },
+  { "atxmega32x1", "__AVR_ATxmega32X1__" },
+  { "atxmega64a3", "__AVR_ATxmega64A3__" },
+  { "atxmega64a3u", "__AVR_ATxmega64A3U__" },
+  { "atxmega64a4u", "__AVR_ATxmega64A4U__" },
+  { "atxmega64b1", "__AVR_ATxmega64B1__" },
+  { "atxmega64b3", "__AVR_ATxmega64B3__" },
+  { "atxmega64c3", "__AVR_ATxmega64C3__" },
+  { "atxmega64d3", "__AVR_ATxmega64D3__" },
+  { "atxmega64d4", "__AVR_ATxmega64D4__" },
+  { "atxmega64a1", "__AVR_ATxmega64A1__" },
+  { "atxmega64a1u", "__AVR_ATxmega64A1U__" },
+  { "atxmega128a3", "__AVR_ATxmega128A3__" },
+  { "atxmega128a3u", "__AVR_ATxmega128A3U__" },
+  { "atxmega128b1", "__AVR_ATxmega128B1__" },
+  { "atxmega128b3", "__AVR_ATxmega128B3__" },
+  { "atxmega128c3", "__AVR_ATxmega128C3__" },
+  { "atxmega128d3", "__AVR_ATxmega128D3__" },
+  { "atxmega128d4", "__AVR_ATxmega128D4__" },
+  { "atxmega192a3", "__AVR_ATxmega192A3__" },
+  { "atxmega192a3u", "__AVR_ATxmega192A3U__" },
+  { "atxmega192c3", "__AVR_ATxmega192C3__" },
+  { "atxmega192d3", "__AVR_ATxmega192D3__" },
+  { "atxmega256a3", "__AVR_ATxmega256A3__" },
+  { "atxmega256a3u", "__AVR_ATxmega256A3U__" },
+  { "atxmega256a3b", "__AVR_ATxmega256A3B__" },
+  { "atxmega256a3bu", "__AVR_ATxmega256A3BU__" },
+  { "atxmega256c3", "__AVR_ATxmega256C3__" },
+  { "atxmega256d3", "__AVR_ATxmega256D3__" },
+  { "atxmega384c3", "__AVR_ATxmega384C3__" },
+  { "atxmega384d3", "__AVR_ATxmega384D3__" },
+  { "atxmega128a1", "__AVR_ATxmega128A1__" },
+  { "atxmega128a1u", "__AVR_ATxmega128A1U__" },
+  { "atxmega128a4u", "__AVR_ATxmega128a4U__" },
+  { "attiny4", "__AVR_ATtiny4__" },
+  { "attiny5", "__AVR_ATtiny5__" },
+  { "attiny9", "__AVR_ATtiny9__" },
+  { "attiny10", "__AVR_ATtiny10__" },
+  { "attiny20", "__AVR_ATtiny20__" },
+  { "attiny40", "__AVR_ATtiny40__" },
+  { "attiny102", "__AVR_ATtiny102__" },
+  { "attiny104", "__AVR_ATtiny104__" },
+};
 
 // AVR Target
 class AVRTargetInfo : public TargetInfo {
@@ -8476,7 +8893,17 @@ public:
 
   void getTargetDefines(const LangOptions &Opts,
                         MacroBuilder &Builder) const override {
+    Builder.defineMacro("AVR");
+    Builder.defineMacro("__AVR");
     Builder.defineMacro("__AVR__");
+
+    if (!this->CPU.empty()) {
+      auto It = std::find_if(AVRMcus.begin(), AVRMcus.end(),
+        [&](const MCUInfo &Info) { return Info.Name == this->CPU; });
+
+      if (It != AVRMcus.end())
+        Builder.defineMacro(It->DefineName);
+    }
   }
 
   ArrayRef<Builtin::Info> getTargetBuiltins() const override {
@@ -8517,6 +8944,57 @@ public:
 
   bool validateAsmConstraint(const char *&Name,
                              TargetInfo::ConstraintInfo &Info) const override {
+    // There aren't any multi-character AVR specific constraints.
+    if (StringRef(Name).size() > 1) return false;
+
+    switch (*Name) {
+      default: return false;
+      case 'a': // Simple upper registers
+      case 'b': // Base pointer registers pairs
+      case 'd': // Upper register
+      case 'l': // Lower registers
+      case 'e': // Pointer register pairs
+      case 'q': // Stack pointer register
+      case 'r': // Any register
+      case 'w': // Special upper register pairs
+      case 't': // Temporary register
+      case 'x': case 'X': // Pointer register pair X
+      case 'y': case 'Y': // Pointer register pair Y
+      case 'z': case 'Z': // Pointer register pair Z
+        Info.setAllowsRegister();
+        return true;
+      case 'I': // 6-bit positive integer constant
+        Info.setRequiresImmediate(0, 63);
+        return true;
+      case 'J': // 6-bit negative integer constant
+        Info.setRequiresImmediate(-63, 0);
+        return true;
+      case 'K': // Integer constant (Range: 2)
+        Info.setRequiresImmediate(2);
+        return true;
+      case 'L': // Integer constant (Range: 0)
+        Info.setRequiresImmediate(0);
+        return true;
+      case 'M': // 8-bit integer constant
+        Info.setRequiresImmediate(0, 0xff);
+        return true;
+      case 'N': // Integer constant (Range: -1)
+        Info.setRequiresImmediate(-1);
+        return true;
+      case 'O': // Integer constant (Range: 8, 16, 24)
+        Info.setRequiresImmediate({8, 16, 24});
+        return true;
+      case 'P': // Integer constant (Range: 1)
+        Info.setRequiresImmediate(1);
+        return true;
+      case 'R': // Integer constant (Range: -6 to 5)
+        Info.setRequiresImmediate(-6, 5);
+        return true;
+      case 'G': // Floating point constant
+      case 'Q': // A memory address based on Y or Z pointer with displacement.
+        return true;
+    }
+
     return false;
   }
 
@@ -8534,6 +9012,41 @@ public:
                ? (IsSigned ? SignedInt : UnsignedInt)
                : TargetInfo::getLeastIntTypeByWidth(BitWidth, IsSigned);
   }
+
+  bool setCPU(const std::string &Name) override {
+    bool IsFamily = llvm::StringSwitch<bool>(Name)
+      .Case("avr1", true)
+      .Case("avr2", true)
+      .Case("avr25", true)
+      .Case("avr3", true)
+      .Case("avr31", true)
+      .Case("avr35", true)
+      .Case("avr4", true)
+      .Case("avr5", true)
+      .Case("avr51", true)
+      .Case("avr6", true)
+      .Case("avrxmega1", true)
+      .Case("avrxmega2", true)
+      .Case("avrxmega3", true)
+      .Case("avrxmega4", true)
+      .Case("avrxmega5", true)
+      .Case("avrxmega6", true)
+      .Case("avrxmega7", true)
+      .Case("avrtiny", true)
+      .Default(false);
+
+    if (IsFamily) this->CPU = Name;
+
+    bool IsMCU = std::find_if(AVRMcus.begin(), AVRMcus.end(),
+      [&](const MCUInfo &Info) { return Info.Name == Name; }) != AVRMcus.end();
+
+    if (IsMCU) this->CPU = Name;
+
+    return IsFamily || IsMCU;
+  }
+
+protected:
+  std::string CPU;
 };
 
 } // end anonymous namespace
@@ -8574,6 +9087,8 @@ static TargetInfo *AllocateTarget(const llvm::Triple &Triple,
       return new LinuxTargetInfo<AArch64leTargetInfo>(Triple, Opts);
     case llvm::Triple::NetBSD:
       return new NetBSDTargetInfo<AArch64leTargetInfo>(Triple, Opts);
+    case llvm::Triple::OpenBSD:
+      return new OpenBSDTargetInfo<AArch64leTargetInfo>(Triple, Opts);
     default:
       return new AArch64leTargetInfo(Triple, Opts);
     }
@@ -8604,8 +9119,6 @@ static TargetInfo *AllocateTarget(const llvm::Triple &Triple,
       return new LinuxTargetInfo<ARMleTargetInfo>(Triple, Opts);
     case llvm::Triple::FreeBSD:
       return new FreeBSDTargetInfo<ARMleTargetInfo>(Triple, Opts);
-    case llvm::Triple::Fuchsia:
-      return new FuchsiaTargetInfo<ARMleTargetInfo>(Triple, Opts);
     case llvm::Triple::NetBSD:
       return new NetBSDTargetInfo<ARMleTargetInfo>(Triple, Opts);
     case llvm::Triple::OpenBSD:
@@ -8642,8 +9155,6 @@ static TargetInfo *AllocateTarget(const llvm::Triple &Triple,
       return new LinuxTargetInfo<ARMbeTargetInfo>(Triple, Opts);
     case llvm::Triple::FreeBSD:
       return new FreeBSDTargetInfo<ARMbeTargetInfo>(Triple, Opts);
-    case llvm::Triple::Fuchsia:
-      return new FuchsiaTargetInfo<ARMbeTargetInfo>(Triple, Opts);
     case llvm::Triple::NetBSD:
       return new NetBSDTargetInfo<ARMbeTargetInfo>(Triple, Opts);
     case llvm::Triple::OpenBSD:
@@ -8879,8 +9390,6 @@ static TargetInfo *AllocateTarget(const llvm::Triple &Triple,
       return new BitrigI386TargetInfo(Triple, Opts);
     case llvm::Triple::FreeBSD:
       return new FreeBSDTargetInfo<X86_32TargetInfo>(Triple, Opts);
-    case llvm::Triple::Fuchsia:
-      return new FuchsiaTargetInfo<X86_32TargetInfo>(Triple, Opts);
     case llvm::Triple::KFreeBSD:
       return new KFreeBSDTargetInfo<X86_32TargetInfo>(Triple, Opts);
     case llvm::Triple::Minix:
@@ -8976,11 +9485,19 @@ static TargetInfo *AllocateTarget(const llvm::Triple &Triple,
     return new SPIR64TargetInfo(Triple, Opts);
   }
   case llvm::Triple::wasm32:
-    if (!(Triple == llvm::Triple("wasm32-unknown-unknown")))
+    if (Triple.getSubArch() != llvm::Triple::NoSubArch ||
+        Triple.getVendor() != llvm::Triple::UnknownVendor ||
+        Triple.getOS() != llvm::Triple::UnknownOS ||
+        Triple.getEnvironment() != llvm::Triple::UnknownEnvironment ||
+        !(Triple.isOSBinFormatELF() || Triple.isOSBinFormatWasm()))
       return nullptr;
     return new WebAssemblyOSTargetInfo<WebAssembly32TargetInfo>(Triple, Opts);
   case llvm::Triple::wasm64:
-    if (!(Triple == llvm::Triple("wasm64-unknown-unknown")))
+    if (Triple.getSubArch() != llvm::Triple::NoSubArch ||
+        Triple.getVendor() != llvm::Triple::UnknownVendor ||
+        Triple.getOS() != llvm::Triple::UnknownOS ||
+        Triple.getEnvironment() != llvm::Triple::UnknownEnvironment ||
+        !(Triple.isOSBinFormatELF() || Triple.isOSBinFormatWasm()))
       return nullptr;
     return new WebAssemblyOSTargetInfo<WebAssembly64TargetInfo>(Triple, Opts);
 
diff --git a/lib/Basic/Version.cpp b/lib/Basic/Version.cpp
index 97e75a9cd45e..a1a67c2bc144 100644
--- a/lib/Basic/Version.cpp
+++ b/lib/Basic/Version.cpp
@@ -36,7 +36,7 @@ std::string getClangRepositoryPath() {
 
   // If the SVN_REPOSITORY is empty, try to use the SVN keyword. This helps us
   // pick up a tag in an SVN export, for example.
-  StringRef SVNRepository("$URL: https://llvm.org/svn/llvm-project/cfe/tags/RELEASE_400/final/lib/Basic/Version.cpp $");
+  StringRef SVNRepository("$URL: https://llvm.org/svn/llvm-project/cfe/trunk/lib/Basic/Version.cpp $");
   if (URL.empty()) {
     URL = SVNRepository.slice(SVNRepository.find(':'),
                               SVNRepository.find("/lib/Basic"));
diff --git a/lib/Basic/VirtualFileSystem.cpp b/lib/Basic/VirtualFileSystem.cpp
index 50fcb22faf53..f5db717866a9 100644
--- a/lib/Basic/VirtualFileSystem.cpp
+++ b/lib/Basic/VirtualFileSystem.cpp
@@ -27,13 +27,6 @@
 #include <memory>
 #include <utility>
 
-// For chdir.
-#ifdef LLVM_ON_WIN32
-#  include <direct.h>
-#else
-#  include <unistd.h>
-#endif
-
 using namespace clang;
 using namespace clang::vfs;
 using namespace llvm;
@@ -235,11 +228,7 @@ std::error_code RealFileSystem::setCurrentWorkingDirectory(const Twine &Path) {
   // difference for example on network filesystems, where symlinks might be
   // switched during runtime of the tool. Fixing this depends on having a
   // file system abstraction that allows openat() style interactions.
-  SmallString<256> Storage;
-  StringRef Dir = Path.toNullTerminatedStringRef(Storage);
-  if (int Err = ::chdir(Dir.data()))
-    return std::error_code(Err, std::generic_category());
-  return std::error_code();
+  return llvm::sys::fs::set_current_path(Path);
 }
 
 IntrusiveRefCntPtr<FileSystem> vfs::getRealFileSystem() {
@@ -249,16 +238,13 @@ IntrusiveRefCntPtr<FileSystem> vfs::getRealFileSystem() {
 
 namespace {
 class RealFSDirIter : public clang::vfs::detail::DirIterImpl {
-  std::string Path;
   llvm::sys::fs::directory_iterator Iter;
 public:
-  RealFSDirIter(const Twine &_Path, std::error_code &EC)
-      : Path(_Path.str()), Iter(Path, EC) {
+  RealFSDirIter(const Twine &Path, std::error_code &EC) : Iter(Path, EC) {
     if (!EC && Iter != llvm::sys::fs::directory_iterator()) {
       llvm::sys::fs::file_status S;
       EC = Iter->status(S);
-      if (!EC)
-        CurrentEntry = Status::copyWithNewName(S, Iter->path());
+      CurrentEntry = Status::copyWithNewName(S, Iter->path());
     }
   }
 
@@ -1869,7 +1855,7 @@ vfs::recursive_directory_iterator::recursive_directory_iterator(FileSystem &FS_,
                                                            std::error_code &EC)
     : FS(&FS_) {
   directory_iterator I = FS->dir_begin(Path, EC);
-  if (!EC && I != directory_iterator()) {
+  if (I != directory_iterator()) {
     State = std::make_shared<IterState>();
     State->push(I);
   }
@@ -1882,8 +1868,6 @@ recursive_directory_iterator::increment(std::error_code &EC) {
   vfs::directory_iterator End;
   if (State->top()->isDirectory()) {
     vfs::directory_iterator I = FS->dir_begin(State->top()->getName(), EC);
-    if (EC)
-      return *this;
     if (I != End) {
       State->push(I);
       return *this;
diff --git a/lib/Basic/XRayLists.cpp b/lib/Basic/XRayLists.cpp
new file mode 100644
index 000000000000..dccf3baa75e2
--- /dev/null
+++ b/lib/Basic/XRayLists.cpp
@@ -0,0 +1,53 @@
+//===--- XRayFunctionFilter.cpp - XRay automatic-attribution --------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// User-provided filters for always/never XRay instrumenting certain functions.
+//
+//===----------------------------------------------------------------------===//
+#include "clang/Basic/XRayLists.h"
+
+using namespace clang;
+
+XRayFunctionFilter::XRayFunctionFilter(
+    ArrayRef<std::string> AlwaysInstrumentPaths,
+    ArrayRef<std::string> NeverInstrumentPaths, SourceManager &SM)
+    : AlwaysInstrument(
+          llvm::SpecialCaseList::createOrDie(AlwaysInstrumentPaths)),
+      NeverInstrument(llvm::SpecialCaseList::createOrDie(NeverInstrumentPaths)),
+      SM(SM) {}
+
+XRayFunctionFilter::ImbueAttribute
+XRayFunctionFilter::shouldImbueFunction(StringRef FunctionName) const {
+  // First apply the always instrument list, than if it isn't an "always" see
+  // whether it's treated as a "never" instrument function.
+  if (AlwaysInstrument->inSection("fun", FunctionName))
+    return ImbueAttribute::ALWAYS;
+  if (NeverInstrument->inSection("fun", FunctionName))
+    return ImbueAttribute::NEVER;
+  return ImbueAttribute::NONE;
+}
+
+XRayFunctionFilter::ImbueAttribute
+XRayFunctionFilter::shouldImbueFunctionsInFile(StringRef Filename,
+                                               StringRef Category) const {
+  if (AlwaysInstrument->inSection("src", Filename, Category))
+    return ImbueAttribute::ALWAYS;
+  if (NeverInstrument->inSection("src", Filename, Category))
+    return ImbueAttribute::NEVER;
+  return ImbueAttribute::NONE;
+}
+
+XRayFunctionFilter::ImbueAttribute
+XRayFunctionFilter::shouldImbueLocation(SourceLocation Loc,
+                                        StringRef Category) const {
+  if (!Loc.isValid())
+    return ImbueAttribute::NONE;
+  return this->shouldImbueFunctionsInFile(SM.getFilename(SM.getFileLoc(Loc)),
+                                          Category);
+}
diff --git a/lib/CodeGen/ABIInfo.h b/lib/CodeGen/ABIInfo.h
index ac31dfdaf3e4..c0be60ef53bc 100644
--- a/lib/CodeGen/ABIInfo.h
+++ b/lib/CodeGen/ABIInfo.h
@@ -10,6 +10,7 @@
 #ifndef LLVM_CLANG_LIB_CODEGEN_ABIINFO_H
 #define LLVM_CLANG_LIB_CODEGEN_ABIINFO_H
 
+#include "clang/AST/CharUnits.h"
 #include "clang/AST/Type.h"
 #include "llvm/IR/CallingConv.h"
 #include "llvm/IR/Type.h"
diff --git a/lib/CodeGen/BackendUtil.cpp b/lib/CodeGen/BackendUtil.cpp
index d2ce6ea48e41..855d6795b9d6 100644
--- a/lib/CodeGen/BackendUtil.cpp
+++ b/lib/CodeGen/BackendUtil.cpp
@@ -61,6 +61,9 @@ using namespace llvm;
 
 namespace {
 
+// Default filename used for profile generation.
+static constexpr StringLiteral DefaultProfileGenName = "default_%m.profraw";
+
 class EmitAssemblyHelper {
   DiagnosticsEngine &Diags;
   const HeaderSearchOptions &HSOpts;
@@ -73,7 +76,6 @@ class EmitAssemblyHelper {
 
   std::unique_ptr<raw_pwrite_stream> OS;
 
-private:
   TargetIRAnalysis getTargetIRAnalysis() const {
     if (TM)
       return TM->getTargetIRAnalysis();
@@ -262,7 +264,7 @@ static TargetLibraryInfoImpl *createTLII(llvm::Triple &TargetTriple,
     TLII->disableAllFunctions();
   else {
     // Disable individual libc/libm calls in TargetLibraryInfo.
-    LibFunc::Func F;
+    LibFunc F;
     for (auto &FuncName : CodeGenOpts.getNoBuiltinFuncs())
       if (TLII->getLibFunc(FuncName, F))
         TLII->setUnavailable(F);
@@ -292,6 +294,140 @@ static void addSymbolRewriterPass(const CodeGenOptions &Opts,
   MPM->add(createRewriteSymbolsPass(DL));
 }
 
+static CodeGenOpt::Level getCGOptLevel(const CodeGenOptions &CodeGenOpts) {
+  switch (CodeGenOpts.OptimizationLevel) {
+  default:
+    llvm_unreachable("Invalid optimization level!");
+  case 0:
+    return CodeGenOpt::None;
+  case 1:
+    return CodeGenOpt::Less;
+  case 2:
+    return CodeGenOpt::Default; // O2/Os/Oz
+  case 3:
+    return CodeGenOpt::Aggressive;
+  }
+}
+
+static llvm::CodeModel::Model getCodeModel(const CodeGenOptions &CodeGenOpts) {
+  unsigned CodeModel =
+      llvm::StringSwitch<unsigned>(CodeGenOpts.CodeModel)
+      .Case("small", llvm::CodeModel::Small)
+      .Case("kernel", llvm::CodeModel::Kernel)
+      .Case("medium", llvm::CodeModel::Medium)
+      .Case("large", llvm::CodeModel::Large)
+      .Case("default", llvm::CodeModel::Default)
+      .Default(~0u);
+  assert(CodeModel != ~0u && "invalid code model!");
+  return static_cast<llvm::CodeModel::Model>(CodeModel);
+}
+
+static llvm::Reloc::Model getRelocModel(const CodeGenOptions &CodeGenOpts) {
+  // Keep this synced with the equivalent code in
+  // lib/Frontend/CompilerInvocation.cpp
+  llvm::Optional<llvm::Reloc::Model> RM;
+  RM = llvm::StringSwitch<llvm::Reloc::Model>(CodeGenOpts.RelocationModel)
+      .Case("static", llvm::Reloc::Static)
+      .Case("pic", llvm::Reloc::PIC_)
+      .Case("ropi", llvm::Reloc::ROPI)
+      .Case("rwpi", llvm::Reloc::RWPI)
+      .Case("ropi-rwpi", llvm::Reloc::ROPI_RWPI)
+      .Case("dynamic-no-pic", llvm::Reloc::DynamicNoPIC);
+  assert(RM.hasValue() && "invalid PIC model!");
+  return *RM;
+}
+
+static TargetMachine::CodeGenFileType getCodeGenFileType(BackendAction Action) {
+  if (Action == Backend_EmitObj)
+    return TargetMachine::CGFT_ObjectFile;
+  else if (Action == Backend_EmitMCNull)
+    return TargetMachine::CGFT_Null;
+  else {
+    assert(Action == Backend_EmitAssembly && "Invalid action!");
+    return TargetMachine::CGFT_AssemblyFile;
+  }
+}
+
+static void initTargetOptions(llvm::TargetOptions &Options,
+                              const CodeGenOptions &CodeGenOpts,
+                              const clang::TargetOptions &TargetOpts,
+                              const LangOptions &LangOpts,
+                              const HeaderSearchOptions &HSOpts) {
+  Options.ThreadModel =
+      llvm::StringSwitch<llvm::ThreadModel::Model>(CodeGenOpts.ThreadModel)
+          .Case("posix", llvm::ThreadModel::POSIX)
+          .Case("single", llvm::ThreadModel::Single);
+
+  // Set float ABI type.
+  assert((CodeGenOpts.FloatABI == "soft" || CodeGenOpts.FloatABI == "softfp" ||
+          CodeGenOpts.FloatABI == "hard" || CodeGenOpts.FloatABI.empty()) &&
+         "Invalid Floating Point ABI!");
+  Options.FloatABIType =
+      llvm::StringSwitch<llvm::FloatABI::ABIType>(CodeGenOpts.FloatABI)
+          .Case("soft", llvm::FloatABI::Soft)
+          .Case("softfp", llvm::FloatABI::Soft)
+          .Case("hard", llvm::FloatABI::Hard)
+          .Default(llvm::FloatABI::Default);
+
+  // Set FP fusion mode.
+  switch (LangOpts.getDefaultFPContractMode()) {
+  case LangOptions::FPC_Off:
+    Options.AllowFPOpFusion = llvm::FPOpFusion::Strict;
+    break;
+  case LangOptions::FPC_On:
+    Options.AllowFPOpFusion = llvm::FPOpFusion::Standard;
+    break;
+  case LangOptions::FPC_Fast:
+    Options.AllowFPOpFusion = llvm::FPOpFusion::Fast;
+    break;
+  }
+
+  Options.UseInitArray = CodeGenOpts.UseInitArray;
+  Options.DisableIntegratedAS = CodeGenOpts.DisableIntegratedAS;
+  Options.CompressDebugSections = CodeGenOpts.CompressDebugSections;
+  Options.RelaxELFRelocations = CodeGenOpts.RelaxELFRelocations;
+
+  // Set EABI version.
+  Options.EABIVersion = llvm::StringSwitch<llvm::EABI>(TargetOpts.EABIVersion)
+                            .Case("4", llvm::EABI::EABI4)
+                            .Case("5", llvm::EABI::EABI5)
+                            .Case("gnu", llvm::EABI::GNU)
+                            .Default(llvm::EABI::Default);
+
+  if (LangOpts.SjLjExceptions)
+    Options.ExceptionModel = llvm::ExceptionHandling::SjLj;
+
+  Options.NoInfsFPMath = CodeGenOpts.NoInfsFPMath;
+  Options.NoNaNsFPMath = CodeGenOpts.NoNaNsFPMath;
+  Options.NoZerosInBSS = CodeGenOpts.NoZeroInitializedInBSS;
+  Options.UnsafeFPMath = CodeGenOpts.UnsafeFPMath;
+  Options.StackAlignmentOverride = CodeGenOpts.StackAlignment;
+  Options.FunctionSections = CodeGenOpts.FunctionSections;
+  Options.DataSections = CodeGenOpts.DataSections;
+  Options.UniqueSectionNames = CodeGenOpts.UniqueSectionNames;
+  Options.EmulatedTLS = CodeGenOpts.EmulatedTLS;
+  Options.DebuggerTuning = CodeGenOpts.getDebuggerTuning();
+
+  Options.MCOptions.MCRelaxAll = CodeGenOpts.RelaxAll;
+  Options.MCOptions.MCSaveTempLabels = CodeGenOpts.SaveTempLabels;
+  Options.MCOptions.MCUseDwarfDirectory = !CodeGenOpts.NoDwarfDirectoryAsm;
+  Options.MCOptions.MCNoExecStack = CodeGenOpts.NoExecStack;
+  Options.MCOptions.MCIncrementalLinkerCompatible =
+      CodeGenOpts.IncrementalLinkerCompatible;
+  Options.MCOptions.MCPIECopyRelocations = CodeGenOpts.PIECopyRelocations;
+  Options.MCOptions.MCFatalWarnings = CodeGenOpts.FatalWarnings;
+  Options.MCOptions.AsmVerbose = CodeGenOpts.AsmVerbose;
+  Options.MCOptions.PreserveAsmComments = CodeGenOpts.PreserveAsmComments;
+  Options.MCOptions.ABIName = TargetOpts.ABI;
+  for (const auto &Entry : HSOpts.UserEntries)
+    if (!Entry.IsFramework &&
+        (Entry.Group == frontend::IncludeDirGroup::Quoted ||
+         Entry.Group == frontend::IncludeDirGroup::Angled ||
+         Entry.Group == frontend::IncludeDirGroup::System))
+      Options.MCOptions.IASSearchPaths.push_back(
+          Entry.IgnoreSysRoot ? Entry.Path : HSOpts.Sysroot + Entry.Path);
+}
+
 void EmitAssemblyHelper::CreatePasses(legacy::PassManager &MPM,
                                       legacy::FunctionPassManager &FPM) {
   // Handle disabling of all LLVM passes, where we want to preserve the
@@ -316,8 +452,13 @@ void EmitAssemblyHelper::CreatePasses(legacy::PassManager &MPM,
                                      !CodeGenOpts.DisableLifetimeMarkers);
     PMBuilder.Inliner = createAlwaysInlinerLegacyPass(InsertLifetimeIntrinsics);
   } else {
+    // We do not want to inline hot callsites for SamplePGO module-summary build
+    // because profile annotation will happen again in ThinLTO backend, and we
+    // want the IR of the hot path to match the profile.
     PMBuilder.Inliner = createFunctionInliningPass(
-        CodeGenOpts.OptimizationLevel, CodeGenOpts.OptimizeSize);
+        CodeGenOpts.OptimizationLevel, CodeGenOpts.OptimizeSize,
+        (!CodeGenOpts.SampleProfileFile.empty() &&
+         CodeGenOpts.EmitSummaryIndex));
   }
 
   PMBuilder.OptLevel = CodeGenOpts.OptimizationLevel;
@@ -334,16 +475,13 @@ void EmitAssemblyHelper::CreatePasses(legacy::PassManager &MPM,
 
   MPM.add(new TargetLibraryInfoWrapperPass(*TLII));
 
-  // Add target-specific passes that need to run as early as possible.
   if (TM)
-    PMBuilder.addExtension(
-        PassManagerBuilder::EP_EarlyAsPossible,
-        [&](const PassManagerBuilder &, legacy::PassManagerBase &PM) {
-          TM->addEarlyAsPossiblePasses(PM);
-        });
+    TM->adjustPassManager(PMBuilder);
 
-  PMBuilder.addExtension(PassManagerBuilder::EP_EarlyAsPossible,
-                         addAddDiscriminatorsPass);
+  if (CodeGenOpts.DebugInfoForProfiling ||
+      !CodeGenOpts.SampleProfileFile.empty())
+    PMBuilder.addExtension(PassManagerBuilder::EP_EarlyAsPossible,
+                           addAddDiscriminatorsPass);
 
   // In ObjC ARC mode, add the main ARC optimization passes.
   if (LangOpts.ObjCAutoRefCount) {
@@ -454,7 +592,7 @@ void EmitAssemblyHelper::CreatePasses(legacy::PassManager &MPM,
     if (!CodeGenOpts.InstrProfileOutput.empty())
       PMBuilder.PGOInstrGen = CodeGenOpts.InstrProfileOutput;
     else
-      PMBuilder.PGOInstrGen = "default_%m.profraw";
+      PMBuilder.PGOInstrGen = DefaultProfileGenName;
   }
   if (CodeGenOpts.hasProfileIRUse())
     PMBuilder.PGOInstrUse = CodeGenOpts.ProfileInstrumentUsePath;
@@ -495,126 +633,14 @@ void EmitAssemblyHelper::CreateTargetMachine(bool MustCreateTM) {
     return;
   }
 
-  unsigned CodeModel =
-    llvm::StringSwitch<unsigned>(CodeGenOpts.CodeModel)
-      .Case("small", llvm::CodeModel::Small)
-      .Case("kernel", llvm::CodeModel::Kernel)
-      .Case("medium", llvm::CodeModel::Medium)
-      .Case("large", llvm::CodeModel::Large)
-      .Case("default", llvm::CodeModel::Default)
-      .Default(~0u);
-  assert(CodeModel != ~0u && "invalid code model!");
-  llvm::CodeModel::Model CM = static_cast<llvm::CodeModel::Model>(CodeModel);
-
+  llvm::CodeModel::Model CM  = getCodeModel(CodeGenOpts);
   std::string FeaturesStr =
       llvm::join(TargetOpts.Features.begin(), TargetOpts.Features.end(), ",");
-
-  // Keep this synced with the equivalent code in tools/driver/cc1as_main.cpp.
-  llvm::Optional<llvm::Reloc::Model> RM;
-  RM = llvm::StringSwitch<llvm::Reloc::Model>(CodeGenOpts.RelocationModel)
-           .Case("static", llvm::Reloc::Static)
-           .Case("pic", llvm::Reloc::PIC_)
-           .Case("ropi", llvm::Reloc::ROPI)
-           .Case("rwpi", llvm::Reloc::RWPI)
-           .Case("ropi-rwpi", llvm::Reloc::ROPI_RWPI)
-           .Case("dynamic-no-pic", llvm::Reloc::DynamicNoPIC);
-  assert(RM.hasValue() && "invalid PIC model!");
-
-  CodeGenOpt::Level OptLevel;
-  switch (CodeGenOpts.OptimizationLevel) {
-  default:
-    llvm_unreachable("Invalid optimization level!");
-  case 0:
-    OptLevel = CodeGenOpt::None;
-    break;
-  case 1:
-    OptLevel = CodeGenOpt::Less;
-    break;
-  case 2:
-    OptLevel = CodeGenOpt::Default;
-    break; // O2/Os/Oz
-  case 3:
-    OptLevel = CodeGenOpt::Aggressive;
-    break;
-  }
+  llvm::Reloc::Model RM = getRelocModel(CodeGenOpts);
+  CodeGenOpt::Level OptLevel = getCGOptLevel(CodeGenOpts);
 
   llvm::TargetOptions Options;
-
-  Options.ThreadModel =
-    llvm::StringSwitch<llvm::ThreadModel::Model>(CodeGenOpts.ThreadModel)
-      .Case("posix", llvm::ThreadModel::POSIX)
-      .Case("single", llvm::ThreadModel::Single);
-
-  // Set float ABI type.
-  assert((CodeGenOpts.FloatABI == "soft" || CodeGenOpts.FloatABI == "softfp" ||
-          CodeGenOpts.FloatABI == "hard" || CodeGenOpts.FloatABI.empty()) &&
-         "Invalid Floating Point ABI!");
-  Options.FloatABIType =
-      llvm::StringSwitch<llvm::FloatABI::ABIType>(CodeGenOpts.FloatABI)
-          .Case("soft", llvm::FloatABI::Soft)
-          .Case("softfp", llvm::FloatABI::Soft)
-          .Case("hard", llvm::FloatABI::Hard)
-          .Default(llvm::FloatABI::Default);
-
-  // Set FP fusion mode.
-  switch (CodeGenOpts.getFPContractMode()) {
-  case CodeGenOptions::FPC_Off:
-    Options.AllowFPOpFusion = llvm::FPOpFusion::Strict;
-    break;
-  case CodeGenOptions::FPC_On:
-    Options.AllowFPOpFusion = llvm::FPOpFusion::Standard;
-    break;
-  case CodeGenOptions::FPC_Fast:
-    Options.AllowFPOpFusion = llvm::FPOpFusion::Fast;
-    break;
-  }
-
-  Options.UseInitArray = CodeGenOpts.UseInitArray;
-  Options.DisableIntegratedAS = CodeGenOpts.DisableIntegratedAS;
-  Options.CompressDebugSections = CodeGenOpts.CompressDebugSections;
-  Options.RelaxELFRelocations = CodeGenOpts.RelaxELFRelocations;
-
-  // Set EABI version.
-  Options.EABIVersion = llvm::StringSwitch<llvm::EABI>(TargetOpts.EABIVersion)
-                            .Case("4", llvm::EABI::EABI4)
-                            .Case("5", llvm::EABI::EABI5)
-                            .Case("gnu", llvm::EABI::GNU)
-                            .Default(llvm::EABI::Default);
-
-  if (LangOpts.SjLjExceptions)
-    Options.ExceptionModel = llvm::ExceptionHandling::SjLj;
-
-  Options.LessPreciseFPMADOption = CodeGenOpts.LessPreciseFPMAD;
-  Options.NoInfsFPMath = CodeGenOpts.NoInfsFPMath;
-  Options.NoNaNsFPMath = CodeGenOpts.NoNaNsFPMath;
-  Options.NoZerosInBSS = CodeGenOpts.NoZeroInitializedInBSS;
-  Options.UnsafeFPMath = CodeGenOpts.UnsafeFPMath;
-  Options.StackAlignmentOverride = CodeGenOpts.StackAlignment;
-  Options.FunctionSections = CodeGenOpts.FunctionSections;
-  Options.DataSections = CodeGenOpts.DataSections;
-  Options.UniqueSectionNames = CodeGenOpts.UniqueSectionNames;
-  Options.EmulatedTLS = CodeGenOpts.EmulatedTLS;
-  Options.DebuggerTuning = CodeGenOpts.getDebuggerTuning();
-
-  Options.MCOptions.MCRelaxAll = CodeGenOpts.RelaxAll;
-  Options.MCOptions.MCSaveTempLabels = CodeGenOpts.SaveTempLabels;
-  Options.MCOptions.MCUseDwarfDirectory = !CodeGenOpts.NoDwarfDirectoryAsm;
-  Options.MCOptions.MCNoExecStack = CodeGenOpts.NoExecStack;
-  Options.MCOptions.MCIncrementalLinkerCompatible =
-      CodeGenOpts.IncrementalLinkerCompatible;
-  Options.MCOptions.MCPIECopyRelocations = CodeGenOpts.PIECopyRelocations;
-  Options.MCOptions.MCFatalWarnings = CodeGenOpts.FatalWarnings;
-  Options.MCOptions.AsmVerbose = CodeGenOpts.AsmVerbose;
-  Options.MCOptions.PreserveAsmComments = CodeGenOpts.PreserveAsmComments;
-  Options.MCOptions.ABIName = TargetOpts.ABI;
-  for (const auto &Entry : HSOpts.UserEntries)
-    if (!Entry.IsFramework &&
-        (Entry.Group == frontend::IncludeDirGroup::Quoted ||
-         Entry.Group == frontend::IncludeDirGroup::Angled ||
-         Entry.Group == frontend::IncludeDirGroup::System))
-      Options.MCOptions.IASSearchPaths.push_back(
-          Entry.IgnoreSysRoot ? Entry.Path : HSOpts.Sysroot + Entry.Path);
-
+  initTargetOptions(Options, CodeGenOpts, TargetOpts, LangOpts, HSOpts);
   TM.reset(TheTarget->createTargetMachine(Triple, TargetOpts.CPU, FeaturesStr,
                                           Options, RM, CM, OptLevel));
 }
@@ -630,13 +656,7 @@ bool EmitAssemblyHelper::AddEmitPasses(legacy::PassManager &CodeGenPasses,
 
   // Normal mode, emit a .s or .o file by running the code generator. Note,
   // this also adds codegenerator level optimization passes.
-  TargetMachine::CodeGenFileType CGFT = TargetMachine::CGFT_AssemblyFile;
-  if (Action == Backend_EmitObj)
-    CGFT = TargetMachine::CGFT_ObjectFile;
-  else if (Action == Backend_EmitMCNull)
-    CGFT = TargetMachine::CGFT_Null;
-  else
-    assert(Action == Backend_EmitAssembly && "Invalid action!");
+  TargetMachine::CodeGenFileType CGFT = getCodeGenFileType(Action);
 
   // Add ObjC ARC final-cleanup optimizations. This is done as part of the
   // "codegen" passes so that it isn't run multiple times when there is
@@ -683,14 +703,31 @@ void EmitAssemblyHelper::EmitAssembly(BackendAction Action,
   CodeGenPasses.add(
       createTargetTransformInfoWrapperPass(getTargetIRAnalysis()));
 
+  std::unique_ptr<raw_fd_ostream> ThinLinkOS;
+
   switch (Action) {
   case Backend_EmitNothing:
     break;
 
   case Backend_EmitBC:
-    PerModulePasses.add(createBitcodeWriterPass(
-        *OS, CodeGenOpts.EmitLLVMUseLists, CodeGenOpts.EmitSummaryIndex,
-        CodeGenOpts.EmitSummaryIndex));
+    if (CodeGenOpts.EmitSummaryIndex) {
+      if (!CodeGenOpts.ThinLinkBitcodeFile.empty()) {
+        std::error_code EC;
+        ThinLinkOS.reset(new llvm::raw_fd_ostream(
+            CodeGenOpts.ThinLinkBitcodeFile, EC,
+            llvm::sys::fs::F_None));
+        if (EC) {
+          Diags.Report(diag::err_fe_unable_to_open_output) << CodeGenOpts.ThinLinkBitcodeFile
+                                                           << EC.message();
+          return;
+        }
+      }
+      PerModulePasses.add(
+          createWriteThinLTOBitcodePass(*OS, ThinLinkOS.get()));
+    }
+    else
+      PerModulePasses.add(
+          createBitcodeWriterPass(*OS, CodeGenOpts.EmitLLVMUseLists));
     break;
 
   case Backend_EmitLL:
@@ -779,7 +816,24 @@ void EmitAssemblyHelper::EmitAssemblyWithNewPassManager(
     return;
   TheModule->setDataLayout(TM->createDataLayout());
 
-  PassBuilder PB(TM.get());
+  PGOOptions PGOOpt;
+
+  // -fprofile-generate.
+  PGOOpt.RunProfileGen = CodeGenOpts.hasProfileIRInstr();
+  if (PGOOpt.RunProfileGen)
+    PGOOpt.ProfileGenFile = CodeGenOpts.InstrProfileOutput.empty() ?
+      DefaultProfileGenName : CodeGenOpts.InstrProfileOutput;
+
+  // -fprofile-use.
+  if (CodeGenOpts.hasProfileIRUse())
+    PGOOpt.ProfileUseFile = CodeGenOpts.ProfileInstrumentUsePath;
+
+  // Only pass a PGO options struct if -fprofile-generate or
+  // -fprofile-use were passed on the cmdline.
+  PassBuilder PB(TM.get(),
+    (PGOOpt.RunProfileGen ||
+      !PGOOpt.ProfileUseFile.empty()) ?
+        Optional<PGOOptions>(PGOOpt) : None);
 
   LoopAnalysisManager LAM;
   FunctionAnalysisManager FAM;
@@ -861,8 +915,31 @@ void EmitAssemblyHelper::EmitAssemblyWithNewPassManager(
   }
 }
 
+Expected<BitcodeModule> clang::FindThinLTOModule(MemoryBufferRef MBRef) {
+  Expected<std::vector<BitcodeModule>> BMsOrErr = getBitcodeModuleList(MBRef);
+  if (!BMsOrErr)
+    return BMsOrErr.takeError();
+
+  // The bitcode file may contain multiple modules, we want the one with a
+  // summary.
+  for (BitcodeModule &BM : *BMsOrErr) {
+    Expected<bool> HasSummary = BM.hasSummary();
+    if (HasSummary && *HasSummary)
+      return BM;
+  }
+
+  return make_error<StringError>("Could not find module summary",
+                                 inconvertibleErrorCode());
+}
+
 static void runThinLTOBackend(ModuleSummaryIndex *CombinedIndex, Module *M,
-                              std::unique_ptr<raw_pwrite_stream> OS) {
+                              const HeaderSearchOptions &HeaderOpts,
+                              const CodeGenOptions &CGOpts,
+                              const clang::TargetOptions &TOpts,
+                              const LangOptions &LOpts,
+                              std::unique_ptr<raw_pwrite_stream> OS,
+                              std::string SampleProfile,
+                              BackendAction Action) {
   StringMap<std::map<GlobalValue::GUID, GlobalValueSummary *>>
       ModuleToDefinedGVSummaries;
   CombinedIndex->collectDefinedGVSummariesPerModule(ModuleToDefinedGVSummaries);
@@ -897,32 +974,15 @@ static void runThinLTOBackend(ModuleSummaryIndex *CombinedIndex, Module *M,
       return;
     }
 
-    Expected<std::vector<BitcodeModule>> BMsOrErr =
-        getBitcodeModuleList(**MBOrErr);
-    if (!BMsOrErr) {
-      handleAllErrors(BMsOrErr.takeError(), [&](ErrorInfoBase &EIB) {
+    Expected<BitcodeModule> BMOrErr = FindThinLTOModule(**MBOrErr);
+    if (!BMOrErr) {
+      handleAllErrors(BMOrErr.takeError(), [&](ErrorInfoBase &EIB) {
         errs() << "Error loading imported file '" << I.first()
                << "': " << EIB.message() << '\n';
       });
       return;
     }
-
-    // The bitcode file may contain multiple modules, we want the one with a
-    // summary.
-    bool FoundModule = false;
-    for (BitcodeModule &BM : *BMsOrErr) {
-      Expected<bool> HasSummary = BM.hasSummary();
-      if (HasSummary && *HasSummary) {
-        ModuleMap.insert({I.first(), BM});
-        FoundModule = true;
-        break;
-      }
-    }
-    if (!FoundModule) {
-      errs() << "Error loading imported file '" << I.first()
-             << "': Could not find module summary\n";
-      return;
-    }
+    ModuleMap.insert({I.first(), *BMOrErr});
 
     OwnedImports.push_back(std::move(*MBOrErr));
   }
@@ -930,6 +990,35 @@ static void runThinLTOBackend(ModuleSummaryIndex *CombinedIndex, Module *M,
     return llvm::make_unique<lto::NativeObjectStream>(std::move(OS));
   };
   lto::Config Conf;
+  Conf.CPU = TOpts.CPU;
+  Conf.CodeModel = getCodeModel(CGOpts);
+  Conf.MAttrs = TOpts.Features;
+  Conf.RelocModel = getRelocModel(CGOpts);
+  Conf.CGOptLevel = getCGOptLevel(CGOpts);
+  initTargetOptions(Conf.Options, CGOpts, TOpts, LOpts, HeaderOpts);
+  Conf.SampleProfile = std::move(SampleProfile);
+  switch (Action) {
+  case Backend_EmitNothing:
+    Conf.PreCodeGenModuleHook = [](size_t Task, const Module &Mod) {
+      return false;
+    };
+    break;
+  case Backend_EmitLL:
+    Conf.PreCodeGenModuleHook = [&](size_t Task, const Module &Mod) {
+      M->print(*OS, nullptr, CGOpts.EmitLLVMUseLists);
+      return false;
+    };
+    break;
+  case Backend_EmitBC:
+    Conf.PreCodeGenModuleHook = [&](size_t Task, const Module &Mod) {
+      WriteBitcodeToFile(M, *OS, CGOpts.EmitLLVMUseLists);
+      return false;
+    };
+    break;
+  default:
+    Conf.CGFileType = getCodeGenFileType(Action);
+    break;
+  }
   if (Error E = thinBackend(
           Conf, 0, AddStream, *M, *CombinedIndex, ImportList,
           ModuleToDefinedGVSummaries[M->getModuleIdentifier()], ModuleMap)) {
@@ -965,7 +1054,8 @@ void clang::EmitBackendOutput(DiagnosticsEngine &Diags,
     // of an error).
     bool DoThinLTOBackend = CombinedIndex != nullptr;
     if (DoThinLTOBackend) {
-      runThinLTOBackend(CombinedIndex.get(), M, std::move(OS));
+      runThinLTOBackend(CombinedIndex.get(), M, HeaderOpts, CGOpts, TOpts,
+                        LOpts, std::move(OS), CGOpts.SampleProfileFile, Action);
       return;
     }
   }
@@ -996,6 +1086,7 @@ static const char* getSectionNameForBitcode(const Triple &T) {
     return "__LLVM,__bitcode";
   case Triple::COFF:
   case Triple::ELF:
+  case Triple::Wasm:
   case Triple::UnknownObjectFormat:
     return ".llvmbc";
   }
@@ -1008,6 +1099,7 @@ static const char* getSectionNameForCommandline(const Triple &T) {
     return "__LLVM,__cmdline";
   case Triple::COFF:
   case Triple::ELF:
+  case Triple::Wasm:
   case Triple::UnknownObjectFormat:
     return ".llvmcmd";
   }
diff --git a/lib/CodeGen/CGAtomic.cpp b/lib/CodeGen/CGAtomic.cpp
index 9287e46127bd..28e20b53d656 100644
--- a/lib/CodeGen/CGAtomic.cpp
+++ b/lib/CodeGen/CGAtomic.cpp
@@ -1181,7 +1181,7 @@ RValue AtomicInfo::convertAtomicTempToRValue(Address addr,
   if (LVal.isBitField())
     return CGF.EmitLoadOfBitfieldLValue(
         LValue::MakeBitfield(addr, LVal.getBitFieldInfo(), LVal.getType(),
-                             LVal.getAlignmentSource()));
+                             LVal.getAlignmentSource()), loc);
   if (LVal.isVectorElt())
     return CGF.EmitLoadOfLValue(
         LValue::MakeVectorElt(addr, LVal.getVectorIdx(), LVal.getType(),
diff --git a/lib/CodeGen/CGBlocks.cpp b/lib/CodeGen/CGBlocks.cpp
index b250b9a32b18..1a57b3e6608d 100644
--- a/lib/CodeGen/CGBlocks.cpp
+++ b/lib/CodeGen/CGBlocks.cpp
@@ -16,7 +16,7 @@
 #include "CGObjCRuntime.h"
 #include "CodeGenFunction.h"
 #include "CodeGenModule.h"
-#include "ConstantBuilder.h"
+#include "clang/CodeGen/ConstantInitBuilder.h"
 #include "clang/AST/DeclObjC.h"
 #include "llvm/ADT/SmallSet.h"
 #include "llvm/IR/CallSite.h"
@@ -266,7 +266,7 @@ static bool isSafeForCXXConstantCapture(QualType type) {
 static llvm::Constant *tryCaptureAsConstant(CodeGenModule &CGM,
                                             CodeGenFunction *CGF,
                                             const VarDecl *var) {
-  // Return if this is a function paramter. We shouldn't try to
+  // Return if this is a function parameter. We shouldn't try to
   // rematerialize default arguments of function parameters.
   if (isa<ParmVarDecl>(var))
     return nullptr;
@@ -318,6 +318,19 @@ static void initializeForBlockHeader(CodeGenModule &CGM, CGBlockInfo &info,
   elementTypes.push_back(CGM.getBlockDescriptorType());
 }
 
+static QualType getCaptureFieldType(const CodeGenFunction &CGF,
+                                    const BlockDecl::Capture &CI) {
+  const VarDecl *VD = CI.getVariable();
+
+  // If the variable is captured by an enclosing block or lambda expression,
+  // use the type of the capture field.
+  if (CGF.BlockInfo && CI.isNested())
+    return CGF.BlockInfo->getCapture(VD).fieldType();
+  if (auto *FD = CGF.LambdaCaptureFields.lookup(VD))
+    return FD->getType();
+  return VD->getType();
+}
+
 /// Compute the layout of the given block.  Attempts to lay the block
 /// out with minimal space requirements.
 static void computeBlockInfo(CodeGenModule &CGM, CodeGenFunction *CGF,
@@ -432,15 +445,7 @@ static void computeBlockInfo(CodeGenModule &CGM, CodeGenFunction *CGF,
       }
     }
 
-    QualType VT = variable->getType();
-
-    // If the variable is captured by an enclosing block or lambda expression,
-    // use the type of the capture field.
-    if (CGF->BlockInfo && CI.isNested())
-      VT = CGF->BlockInfo->getCapture(variable).fieldType();
-    else if (auto *FD = CGF->LambdaCaptureFields.lookup(variable))
-      VT = FD->getType();
-
+    QualType VT = getCaptureFieldType(*CGF, CI);
     CharUnits size = C.getTypeSizeInChars(VT);
     CharUnits align = C.getDeclAlign(variable);
     
@@ -606,8 +611,8 @@ static void enterBlockScope(CodeGenFunction &CGF, BlockDecl *block) {
     if (capture.isConstant()) continue;
 
     // Ignore objects that aren't destructed.
-    QualType::DestructionKind dtorKind =
-      variable->getType().isDestructedType();
+    QualType VT = getCaptureFieldType(CGF, CI);
+    QualType::DestructionKind dtorKind = VT.isDestructedType();
     if (dtorKind == QualType::DK_none) continue;
 
     CodeGenFunction::Destroyer *destroyer;
@@ -634,7 +639,7 @@ static void enterBlockScope(CodeGenFunction &CGF, BlockDecl *block) {
     if (useArrayEHCleanup) 
       cleanupKind = InactiveNormalAndEHCleanup;
 
-    CGF.pushDestroy(cleanupKind, addr, variable->getType(),
+    CGF.pushDestroy(cleanupKind, addr, VT,
                     destroyer, useArrayEHCleanup);
 
     // Remember where that cleanup was.
@@ -718,7 +723,12 @@ llvm::Value *CodeGenFunction::EmitBlockLiteral(const CGBlockInfo &blockInfo) {
 
   // Otherwise, we have to emit this as a local block.
 
-  llvm::Constant *isa = CGM.getNSConcreteStackBlock();
+  llvm::Constant *isa =
+      (!CGM.getContext().getLangOpts().OpenCL)
+          ? CGM.getNSConcreteStackBlock()
+          : CGM.getNullPointer(cast<llvm::PointerType>(
+                                   CGM.getNSConcreteStackBlock()->getType()),
+                               QualType(getContext().VoidPtrTy));
   isa = llvm::ConstantExpr::getBitCast(isa, VoidPtrTy);
 
   // Build the block descriptor.
@@ -906,9 +916,8 @@ llvm::Value *CodeGenFunction::EmitBlockLiteral(const CGBlockInfo &blockInfo) {
 
   // Cast to the converted block-pointer type, which happens (somewhat
   // unfortunately) to be a pointer to function type.
-  llvm::Value *result =
-    Builder.CreateBitCast(blockAddr.getPointer(),
-                          ConvertType(blockInfo.getBlockExpr()->getType()));
+  llvm::Value *result = Builder.CreatePointerCast(
+      blockAddr.getPointer(), ConvertType(blockInfo.getBlockExpr()->getType()));
 
   return result;
 }
@@ -976,21 +985,41 @@ RValue CodeGenFunction::EmitBlockCallExpr(const CallExpr *E,
   llvm::Value *BlockPtr = EmitScalarExpr(E->getCallee());
 
   // Get a pointer to the generic block literal.
+  // For OpenCL we generate generic AS void ptr to be able to reuse the same
+  // block definition for blocks with captures generated as private AS local
+  // variables and without captures generated as global AS program scope
+  // variables.
+  unsigned AddrSpace = 0;
+  if (getLangOpts().OpenCL)
+    AddrSpace = getContext().getTargetAddressSpace(LangAS::opencl_generic);
+
   llvm::Type *BlockLiteralTy =
-    llvm::PointerType::getUnqual(CGM.getGenericBlockLiteralType());
+      llvm::PointerType::get(CGM.getGenericBlockLiteralType(), AddrSpace);
 
   // Bitcast the callee to a block literal.
-  BlockPtr = Builder.CreateBitCast(BlockPtr, BlockLiteralTy, "block.literal");
+  BlockPtr =
+      Builder.CreatePointerCast(BlockPtr, BlockLiteralTy, "block.literal");
 
   // Get the function pointer from the literal.
   llvm::Value *FuncPtr =
     Builder.CreateStructGEP(CGM.getGenericBlockLiteralType(), BlockPtr, 3);
 
-  BlockPtr = Builder.CreateBitCast(BlockPtr, VoidPtrTy);
 
   // Add the block literal.
   CallArgList Args;
-  Args.add(RValue::get(BlockPtr), getContext().VoidPtrTy);
+
+  QualType VoidPtrQualTy = getContext().VoidPtrTy;
+  llvm::Type *GenericVoidPtrTy = VoidPtrTy;
+  if (getLangOpts().OpenCL) {
+    GenericVoidPtrTy = Builder.getInt8PtrTy(
+        getContext().getTargetAddressSpace(LangAS::opencl_generic));
+    VoidPtrQualTy =
+        getContext().getPointerType(getContext().getAddrSpaceQualType(
+            getContext().VoidTy, LangAS::opencl_generic));
+  }
+
+  BlockPtr = Builder.CreatePointerCast(BlockPtr, GenericVoidPtrTy);
+  Args.add(RValue::get(BlockPtr), VoidPtrQualTy);
 
   QualType FnType = BPT->getPointeeType();
 
@@ -1097,7 +1126,12 @@ static llvm::Constant *buildGlobalBlock(CodeGenModule &CGM,
   auto fields = builder.beginStruct();
 
   // isa
-  fields.add(CGM.getNSConcreteGlobalBlock());
+  fields.add(
+      (!CGM.getContext().getLangOpts().OpenCL)
+          ? CGM.getNSConcreteGlobalBlock()
+          : CGM.getNullPointer(cast<llvm::PointerType>(
+                                   CGM.getNSConcreteGlobalBlock()->getType()),
+                               QualType(CGM.getContext().VoidPtrTy)));
 
   // __flags
   BlockFlags flags = BLOCK_IS_GLOBAL | BLOCK_HAS_SIGNATURE;
@@ -1114,16 +1148,19 @@ static llvm::Constant *buildGlobalBlock(CodeGenModule &CGM,
   // Descriptor
   fields.add(buildBlockDescriptor(CGM, blockInfo));
 
-  llvm::Constant *literal =
-    fields.finishAndCreateGlobal("__block_literal_global",
-                                 blockInfo.BlockAlign,
-                                 /*constant*/ true);
+  unsigned AddrSpace = 0;
+  if (CGM.getContext().getLangOpts().OpenCL)
+    AddrSpace = CGM.getContext().getTargetAddressSpace(LangAS::opencl_global);
+
+  llvm::Constant *literal = fields.finishAndCreateGlobal(
+      "__block_literal_global", blockInfo.BlockAlign,
+      /*constant*/ true, llvm::GlobalVariable::InternalLinkage, AddrSpace);
 
   // Return a constant of the appropriately-casted type.
   llvm::Type *RequiredType =
     CGM.getTypes().ConvertType(blockInfo.getBlockExpr()->getType());
   llvm::Constant *Result =
-      llvm::ConstantExpr::getBitCast(literal, RequiredType);
+      llvm::ConstantExpr::getPointerCast(literal, RequiredType);
   CGM.setAddrOfGlobalBlock(blockInfo.BlockExpression, Result);
   return Result;
 }
@@ -1155,9 +1192,13 @@ void CodeGenFunction::setBlockContextParameter(const ImplicitParamDecl *D,
 
   // Instead of messing around with LocalDeclMap, just set the value
   // directly as BlockPointer.
-  BlockPointer = Builder.CreateBitCast(arg,
-                                       BlockInfo->StructureType->getPointerTo(),
-                                       "block");
+  BlockPointer = Builder.CreatePointerCast(
+      arg,
+      BlockInfo->StructureType->getPointerTo(
+          getContext().getLangOpts().OpenCL
+              ? getContext().getTargetAddressSpace(LangAS::opencl_generic)
+              : 0),
+      "block");
 }
 
 Address CodeGenFunction::LoadBlockStruct() {
@@ -1196,6 +1237,15 @@ CodeGenFunction::GenerateBlockFunction(GlobalDecl GD,
   // The first argument is the block pointer.  Just take it as a void*
   // and cast it later.
   QualType selfTy = getContext().VoidPtrTy;
+
+  // For OpenCL passed block pointer can be private AS local variable or
+  // global AS program scope variable (for the case with and without captures).
+  // Generic AS is used therefore to be able to accomodate both private and
+  // generic AS in one implementation.
+  if (getLangOpts().OpenCL)
+    selfTy = getContext().getPointerType(getContext().getAddrSpaceQualType(
+        getContext().VoidTy, LangAS::opencl_generic));
+
   IdentifierInfo *II = &CGM.getContext().Idents.get(".block_descriptor");
 
   ImplicitParamDecl selfDecl(getContext(), const_cast<BlockDecl*>(blockDecl),
@@ -1323,23 +1373,102 @@ CodeGenFunction::GenerateBlockFunction(GlobalDecl GD,
   return fn;
 }
 
-/*
-    notes.push_back(HelperInfo());
-    HelperInfo &note = notes.back();
-    note.index = capture.getIndex();
-    note.RequiresCopying = (ci->hasCopyExpr() || BlockRequiresCopying(type));
-    note.cxxbar_import = ci->getCopyExpr();
-
-    if (ci->isByRef()) {
-      note.flag = BLOCK_FIELD_IS_BYREF;
-      if (type.isObjCGCWeak())
-        note.flag |= BLOCK_FIELD_IS_WEAK;
-    } else if (type->isBlockPointerType()) {
-      note.flag = BLOCK_FIELD_IS_BLOCK;
-    } else {
-      note.flag = BLOCK_FIELD_IS_OBJECT;
-    }
- */
+namespace {
+
+/// Represents a type of copy/destroy operation that should be performed for an
+/// entity that's captured by a block.
+enum class BlockCaptureEntityKind {
+  CXXRecord, // Copy or destroy
+  ARCWeak,
+  ARCStrong,
+  BlockObject, // Assign or release
+  None
+};
+
+/// Represents a captured entity that requires extra operations in order for
+/// this entity to be copied or destroyed correctly.
+struct BlockCaptureManagedEntity {
+  BlockCaptureEntityKind Kind;
+  BlockFieldFlags Flags;
+  const BlockDecl::Capture &CI;
+  const CGBlockInfo::Capture &Capture;
+
+  BlockCaptureManagedEntity(BlockCaptureEntityKind Type, BlockFieldFlags Flags,
+                            const BlockDecl::Capture &CI,
+                            const CGBlockInfo::Capture &Capture)
+      : Kind(Type), Flags(Flags), CI(CI), Capture(Capture) {}
+};
+
+} // end anonymous namespace
+
+static std::pair<BlockCaptureEntityKind, BlockFieldFlags>
+computeCopyInfoForBlockCapture(const BlockDecl::Capture &CI, QualType T,
+                               const LangOptions &LangOpts) {
+  if (CI.getCopyExpr()) {
+    assert(!CI.isByRef());
+    // don't bother computing flags
+    return std::make_pair(BlockCaptureEntityKind::CXXRecord, BlockFieldFlags());
+  }
+  BlockFieldFlags Flags;
+  if (CI.isByRef()) {
+    Flags = BLOCK_FIELD_IS_BYREF;
+    if (T.isObjCGCWeak())
+      Flags |= BLOCK_FIELD_IS_WEAK;
+    return std::make_pair(BlockCaptureEntityKind::BlockObject, Flags);
+  }
+  if (!T->isObjCRetainableType())
+    // For all other types, the memcpy is fine.
+    return std::make_pair(BlockCaptureEntityKind::None, Flags);
+
+  Flags = BLOCK_FIELD_IS_OBJECT;
+  bool isBlockPointer = T->isBlockPointerType();
+  if (isBlockPointer)
+    Flags = BLOCK_FIELD_IS_BLOCK;
+
+  // Special rules for ARC captures:
+  Qualifiers QS = T.getQualifiers();
+
+  // We need to register __weak direct captures with the runtime.
+  if (QS.getObjCLifetime() == Qualifiers::OCL_Weak)
+    return std::make_pair(BlockCaptureEntityKind::ARCWeak, Flags);
+
+  // We need to retain the copied value for __strong direct captures.
+  if (QS.getObjCLifetime() == Qualifiers::OCL_Strong) {
+    // If it's a block pointer, we have to copy the block and
+    // assign that to the destination pointer, so we might as
+    // well use _Block_object_assign.  Otherwise we can avoid that.
+    return std::make_pair(!isBlockPointer ? BlockCaptureEntityKind::ARCStrong
+                                          : BlockCaptureEntityKind::BlockObject,
+                          Flags);
+  }
+
+  // Non-ARC captures of retainable pointers are strong and
+  // therefore require a call to _Block_object_assign.
+  if (!QS.getObjCLifetime() && !LangOpts.ObjCAutoRefCount)
+    return std::make_pair(BlockCaptureEntityKind::BlockObject, Flags);
+
+  // Otherwise the memcpy is fine.
+  return std::make_pair(BlockCaptureEntityKind::None, Flags);
+}
+
+/// Find the set of block captures that need to be explicitly copied or destroy.
+static void findBlockCapturedManagedEntities(
+    const CGBlockInfo &BlockInfo, const LangOptions &LangOpts,
+    SmallVectorImpl<BlockCaptureManagedEntity> &ManagedCaptures,
+    llvm::function_ref<std::pair<BlockCaptureEntityKind, BlockFieldFlags>(
+        const BlockDecl::Capture &, QualType, const LangOptions &)>
+        Predicate) {
+  for (const auto &CI : BlockInfo.getBlockDecl()->captures()) {
+    const VarDecl *Variable = CI.getVariable();
+    const CGBlockInfo::Capture &Capture = BlockInfo.getCapture(Variable);
+    if (Capture.isConstant())
+      continue;
+
+    auto Info = Predicate(CI, Variable->getType(), LangOpts);
+    if (Info.first != BlockCaptureEntityKind::None)
+      ManagedCaptures.emplace_back(Info.first, Info.second, CI, Capture);
+  }
+}
 
 /// Generate the copy-helper function for a block closure object:
 ///   static void block_copy_helper(block_t *dst, block_t *src);
@@ -1399,78 +1528,28 @@ CodeGenFunction::GenerateCopyHelperFunction(const CGBlockInfo &blockInfo) {
   dst = Address(Builder.CreateLoad(dst), blockInfo.BlockAlign);
   dst = Builder.CreateBitCast(dst, structPtrTy, "block.dest");
 
-  const BlockDecl *blockDecl = blockInfo.getBlockDecl();
+  SmallVector<BlockCaptureManagedEntity, 4> CopiedCaptures;
+  findBlockCapturedManagedEntities(blockInfo, getLangOpts(), CopiedCaptures,
+                                   computeCopyInfoForBlockCapture);
 
-  for (const auto &CI : blockDecl->captures()) {
-    const VarDecl *variable = CI.getVariable();
-    QualType type = variable->getType();
-
-    const CGBlockInfo::Capture &capture = blockInfo.getCapture(variable);
-    if (capture.isConstant()) continue;
-
-    const Expr *copyExpr = CI.getCopyExpr();
-    BlockFieldFlags flags;
-
-    bool useARCWeakCopy = false;
-    bool useARCStrongCopy = false;
-
-    if (copyExpr) {
-      assert(!CI.isByRef());
-      // don't bother computing flags
-
-    } else if (CI.isByRef()) {
-      flags = BLOCK_FIELD_IS_BYREF;
-      if (type.isObjCGCWeak())
-        flags |= BLOCK_FIELD_IS_WEAK;
-
-    } else if (type->isObjCRetainableType()) {
-      flags = BLOCK_FIELD_IS_OBJECT;
-      bool isBlockPointer = type->isBlockPointerType();
-      if (isBlockPointer)
-        flags = BLOCK_FIELD_IS_BLOCK;
-
-      // Special rules for ARC captures:
-      Qualifiers qs = type.getQualifiers();
-
-      // We need to register __weak direct captures with the runtime.
-      if (qs.getObjCLifetime() == Qualifiers::OCL_Weak) {
-        useARCWeakCopy = true;
-
-      // We need to retain the copied value for __strong direct captures.
-      } else if (qs.getObjCLifetime() == Qualifiers::OCL_Strong) {
-        // If it's a block pointer, we have to copy the block and
-        // assign that to the destination pointer, so we might as
-        // well use _Block_object_assign.  Otherwise we can avoid that.
-        if (!isBlockPointer)
-          useARCStrongCopy = true;
-
-      // Non-ARC captures of retainable pointers are strong and
-      // therefore require a call to _Block_object_assign.
-      } else if (!qs.getObjCLifetime() && !getLangOpts().ObjCAutoRefCount) {
-        // fall through
-
-      // Otherwise the memcpy is fine.
-      } else {
-        continue;
-      }
-
-    // For all other types, the memcpy is fine.
-    } else {
-      continue;
-    }
+  for (const auto &CopiedCapture : CopiedCaptures) {
+    const BlockDecl::Capture &CI = CopiedCapture.CI;
+    const CGBlockInfo::Capture &capture = CopiedCapture.Capture;
+    BlockFieldFlags flags = CopiedCapture.Flags;
 
     unsigned index = capture.getIndex();
     Address srcField = Builder.CreateStructGEP(src, index, capture.getOffset());
     Address dstField = Builder.CreateStructGEP(dst, index, capture.getOffset());
 
     // If there's an explicit copy expression, we do that.
-    if (copyExpr) {
-      EmitSynthesizedCXXCopyCtor(dstField, srcField, copyExpr);
-    } else if (useARCWeakCopy) {
+    if (CI.getCopyExpr()) {
+      assert(CopiedCapture.Kind == BlockCaptureEntityKind::CXXRecord);
+      EmitSynthesizedCXXCopyCtor(dstField, srcField, CI.getCopyExpr());
+    } else if (CopiedCapture.Kind == BlockCaptureEntityKind::ARCWeak) {
       EmitARCCopyWeak(dstField, srcField);
     } else {
       llvm::Value *srcValue = Builder.CreateLoad(srcField, "blockcopy.src");
-      if (useARCStrongCopy) {
+      if (CopiedCapture.Kind == BlockCaptureEntityKind::ARCStrong) {
         // At -O0, store null into the destination field (so that the
         // storeStrong doesn't over-release) and then call storeStrong.
         // This is a workaround to not having an initStrong call.
@@ -1491,6 +1570,7 @@ CodeGenFunction::GenerateCopyHelperFunction(const CGBlockInfo &blockInfo) {
           cast<llvm::Instruction>(dstField.getPointer())->eraseFromParent();
         }
       } else {
+        assert(CopiedCapture.Kind == BlockCaptureEntityKind::BlockObject);
         srcValue = Builder.CreateBitCast(srcValue, VoidPtrTy);
         llvm::Value *dstAddr =
           Builder.CreateBitCast(dstField.getPointer(), VoidPtrTy);
@@ -1498,6 +1578,7 @@ CodeGenFunction::GenerateCopyHelperFunction(const CGBlockInfo &blockInfo) {
           dstAddr, srcValue, llvm::ConstantInt::get(Int32Ty, flags.getBitMask())
         };
 
+        const VarDecl *variable = CI.getVariable();
         bool copyCanThrow = false;
         if (CI.isByRef() && variable->getType()->getAsCXXRecordDecl()) {
           const Expr *copyExpr =
@@ -1521,6 +1602,52 @@ CodeGenFunction::GenerateCopyHelperFunction(const CGBlockInfo &blockInfo) {
   return llvm::ConstantExpr::getBitCast(Fn, VoidPtrTy);
 }
 
+static std::pair<BlockCaptureEntityKind, BlockFieldFlags>
+computeDestroyInfoForBlockCapture(const BlockDecl::Capture &CI, QualType T,
+                                  const LangOptions &LangOpts) {
+  BlockFieldFlags Flags;
+  if (CI.isByRef()) {
+    Flags = BLOCK_FIELD_IS_BYREF;
+    if (T.isObjCGCWeak())
+      Flags |= BLOCK_FIELD_IS_WEAK;
+    return std::make_pair(BlockCaptureEntityKind::BlockObject, Flags);
+  }
+
+  if (const CXXRecordDecl *Record = T->getAsCXXRecordDecl()) {
+    if (Record->hasTrivialDestructor())
+      return std::make_pair(BlockCaptureEntityKind::None, BlockFieldFlags());
+    return std::make_pair(BlockCaptureEntityKind::CXXRecord, BlockFieldFlags());
+  }
+
+  // Other types don't need to be destroy explicitly.
+  if (!T->isObjCRetainableType())
+    return std::make_pair(BlockCaptureEntityKind::None, Flags);
+
+  Flags = BLOCK_FIELD_IS_OBJECT;
+  if (T->isBlockPointerType())
+    Flags = BLOCK_FIELD_IS_BLOCK;
+
+  // Special rules for ARC captures.
+  Qualifiers QS = T.getQualifiers();
+
+  // Use objc_storeStrong for __strong direct captures; the
+  // dynamic tools really like it when we do this.
+  if (QS.getObjCLifetime() == Qualifiers::OCL_Strong)
+    return std::make_pair(BlockCaptureEntityKind::ARCStrong, Flags);
+
+  // Support __weak direct captures.
+  if (QS.getObjCLifetime() == Qualifiers::OCL_Weak)
+    return std::make_pair(BlockCaptureEntityKind::ARCWeak, Flags);
+
+  // Non-ARC captures are strong, and we need to use
+  // _Block_object_dispose.
+  if (!QS.hasObjCLifetime() && !LangOpts.ObjCAutoRefCount)
+    return std::make_pair(BlockCaptureEntityKind::BlockObject, Flags);
+
+  // Otherwise, we have nothing to do.
+  return std::make_pair(BlockCaptureEntityKind::None, Flags);
+}
+
 /// Generate the destroy-helper function for a block closure object:
 ///   static void block_destroy_helper(block_t *theBlock);
 ///
@@ -1570,79 +1697,39 @@ CodeGenFunction::GenerateDestroyHelperFunction(const CGBlockInfo &blockInfo) {
   src = Address(Builder.CreateLoad(src), blockInfo.BlockAlign);
   src = Builder.CreateBitCast(src, structPtrTy, "block");
 
-  const BlockDecl *blockDecl = blockInfo.getBlockDecl();
-
   CodeGenFunction::RunCleanupsScope cleanups(*this);
 
-  for (const auto &CI : blockDecl->captures()) {
-    const VarDecl *variable = CI.getVariable();
-    QualType type = variable->getType();
-
-    const CGBlockInfo::Capture &capture = blockInfo.getCapture(variable);
-    if (capture.isConstant()) continue;
-
-    BlockFieldFlags flags;
-    const CXXDestructorDecl *dtor = nullptr;
+  SmallVector<BlockCaptureManagedEntity, 4> DestroyedCaptures;
+  findBlockCapturedManagedEntities(blockInfo, getLangOpts(), DestroyedCaptures,
+                                   computeDestroyInfoForBlockCapture);
 
-    bool useARCWeakDestroy = false;
-    bool useARCStrongDestroy = false;
-
-    if (CI.isByRef()) {
-      flags = BLOCK_FIELD_IS_BYREF;
-      if (type.isObjCGCWeak())
-        flags |= BLOCK_FIELD_IS_WEAK;
-    } else if (const CXXRecordDecl *record = type->getAsCXXRecordDecl()) {
-      if (record->hasTrivialDestructor())
-        continue;
-      dtor = record->getDestructor();
-    } else if (type->isObjCRetainableType()) {
-      flags = BLOCK_FIELD_IS_OBJECT;
-      if (type->isBlockPointerType())
-        flags = BLOCK_FIELD_IS_BLOCK;
-
-      // Special rules for ARC captures.
-      Qualifiers qs = type.getQualifiers();
-
-      // Use objc_storeStrong for __strong direct captures; the
-      // dynamic tools really like it when we do this.
-      if (qs.getObjCLifetime() == Qualifiers::OCL_Strong) {
-        useARCStrongDestroy = true;
-
-      // Support __weak direct captures.
-      } else if (qs.getObjCLifetime() == Qualifiers::OCL_Weak) {
-        useARCWeakDestroy = true;
-
-      // Non-ARC captures are strong, and we need to use _Block_object_dispose.
-      } else if (!qs.hasObjCLifetime() && !getLangOpts().ObjCAutoRefCount) {
-        // fall through
-
-      // Otherwise, we have nothing to do.
-      } else {
-        continue;
-      }
-    } else {
-      continue;
-    }
+  for (const auto &DestroyedCapture : DestroyedCaptures) {
+    const BlockDecl::Capture &CI = DestroyedCapture.CI;
+    const CGBlockInfo::Capture &capture = DestroyedCapture.Capture;
+    BlockFieldFlags flags = DestroyedCapture.Flags;
 
     Address srcField =
       Builder.CreateStructGEP(src, capture.getIndex(), capture.getOffset());
 
-    // If there's an explicit copy expression, we do that.
-    if (dtor) {
-      PushDestructorCleanup(dtor, srcField);
+    // If the captured record has a destructor then call it.
+    if (DestroyedCapture.Kind == BlockCaptureEntityKind::CXXRecord) {
+      const auto *Dtor =
+          CI.getVariable()->getType()->getAsCXXRecordDecl()->getDestructor();
+      PushDestructorCleanup(Dtor, srcField);
 
-    // If this is a __weak capture, emit the release directly.
-    } else if (useARCWeakDestroy) {
+      // If this is a __weak capture, emit the release directly.
+    } else if (DestroyedCapture.Kind == BlockCaptureEntityKind::ARCWeak) {
       EmitARCDestroyWeak(srcField);
 
     // Destroy strong objects with a call if requested.
-    } else if (useARCStrongDestroy) {
+    } else if (DestroyedCapture.Kind == BlockCaptureEntityKind::ARCStrong) {
       EmitARCDestroyStrong(srcField, ARCImpreciseLifetime);
 
     // Otherwise we call _Block_object_dispose.  It wouldn't be too
     // hard to just emit this as a cleanup if we wanted to make sure
     // that things were done in reverse.
     } else {
+      assert(DestroyedCapture.Kind == BlockCaptureEntityKind::BlockObject);
       llvm::Value *value = Builder.CreateLoad(srcField);
       value = Builder.CreateBitCast(value, VoidPtrTy);
       BuildBlockRelease(value, flags);
diff --git a/lib/CodeGen/CGBuiltin.cpp b/lib/CodeGen/CGBuiltin.cpp
index b3d02f1f51c6..6ea0a325a429 100644
--- a/lib/CodeGen/CGBuiltin.cpp
+++ b/lib/CodeGen/CGBuiltin.cpp
@@ -420,10 +420,11 @@ getDefaultBuiltinObjectSizeResult(unsigned Type, llvm::IntegerType *ResType) {
 
 llvm::Value *
 CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const Expr *E, unsigned Type,
-                                                 llvm::IntegerType *ResType) {
+                                                 llvm::IntegerType *ResType,
+                                                 llvm::Value *EmittedE) {
   uint64_t ObjectSize;
   if (!E->tryEvaluateObjectSize(ObjectSize, getContext(), Type))
-    return emitBuiltinObjectSize(E, Type, ResType);
+    return emitBuiltinObjectSize(E, Type, ResType, EmittedE);
   return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
 }
 
@@ -432,9 +433,14 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const Expr *E, unsigned Type,
 ///   - A llvm::Argument (if E is a param with the pass_object_size attribute on
 ///     it)
 ///   - A call to the @llvm.objectsize intrinsic
+///
+/// EmittedE is the result of emitting `E` as a scalar expr. If it's non-null
+/// and we wouldn't otherwise try to reference a pass_object_size parameter,
+/// we'll call @llvm.objectsize on EmittedE, rather than emitting E.
 llvm::Value *
 CodeGenFunction::emitBuiltinObjectSize(const Expr *E, unsigned Type,
-                                       llvm::IntegerType *ResType) {
+                                       llvm::IntegerType *ResType,
+                                       llvm::Value *EmittedE) {
   // We need to reference an argument if the pointer is a parameter with the
   // pass_object_size attribute.
   if (auto *D = dyn_cast<DeclRefExpr>(E->IgnoreParenImpCasts())) {
@@ -457,16 +463,20 @@ CodeGenFunction::emitBuiltinObjectSize(const Expr *E, unsigned Type,
   // LLVM can't handle Type=3 appropriately, and __builtin_object_size shouldn't
   // evaluate E for side-effects. In either case, we shouldn't lower to
   // @llvm.objectsize.
-  if (Type == 3 || E->HasSideEffects(getContext()))
+  if (Type == 3 || (!EmittedE && E->HasSideEffects(getContext())))
     return getDefaultBuiltinObjectSizeResult(Type, ResType);
 
-  // LLVM only supports 0 and 2, make sure that we pass along that
-  // as a boolean.
-  auto *CI = ConstantInt::get(Builder.getInt1Ty(), (Type & 2) >> 1);
-  // FIXME: Get right address space.
-  llvm::Type *Tys[] = {ResType, Builder.getInt8PtrTy(0)};
-  Value *F = CGM.getIntrinsic(Intrinsic::objectsize, Tys);
-  return Builder.CreateCall(F, {EmitScalarExpr(E), CI});
+  Value *Ptr = EmittedE ? EmittedE : EmitScalarExpr(E);
+  assert(Ptr->getType()->isPointerTy() &&
+         "Non-pointer passed to __builtin_object_size?");
+
+  Value *F = CGM.getIntrinsic(Intrinsic::objectsize, {ResType, Ptr->getType()});
+
+  // LLVM only supports 0 and 2, make sure that we pass along that as a boolean.
+  Value *Min = Builder.getInt1((Type & 2) != 0);
+  // For GCC compatability, __builtin_object_size treat NULL as unknown size.
+  Value *NullIsUnknown = Builder.getTrue();
+  return Builder.CreateCall(F, {Ptr, Min, NullIsUnknown});
 }
 
 // Many of MSVC builtins are on both x64 and ARM; to avoid repeating code, we
@@ -482,10 +492,12 @@ enum class CodeGenFunction::MSVCIntrin {
   _InterlockedIncrement,
   _InterlockedOr,
   _InterlockedXor,
+  _interlockedbittestandset,
+  __fastfail,
 };
 
 Value *CodeGenFunction::EmitMSVCBuiltinExpr(MSVCIntrin BuiltinID,
-  const CallExpr *E) {
+                                            const CallExpr *E) {
   switch (BuiltinID) {
   case MSVCIntrin::_BitScanForward:
   case MSVCIntrin::_BitScanReverse: {
@@ -548,6 +560,22 @@ Value *CodeGenFunction::EmitMSVCBuiltinExpr(MSVCIntrin BuiltinID,
   case MSVCIntrin::_InterlockedXor:
     return MakeBinaryAtomicValue(*this, AtomicRMWInst::Xor, E);
 
+  case MSVCIntrin::_interlockedbittestandset: {
+    llvm::Value *Addr = EmitScalarExpr(E->getArg(0));
+    llvm::Value *Bit = EmitScalarExpr(E->getArg(1));
+    AtomicRMWInst *RMWI = Builder.CreateAtomicRMW(
+        AtomicRMWInst::Or, Addr,
+        Builder.CreateShl(ConstantInt::get(Bit->getType(), 1), Bit),
+        llvm::AtomicOrdering::SequentiallyConsistent);
+    // Shift the relevant bit to the least significant position, truncate to
+    // the result type, and test the low bit.
+    llvm::Value *Shifted = Builder.CreateLShr(RMWI, Bit);
+    llvm::Value *Truncated =
+        Builder.CreateTrunc(Shifted, ConvertType(E->getType()));
+    return Builder.CreateAnd(Truncated,
+                             ConstantInt::get(Truncated->getType(), 1));
+  }
+
   case MSVCIntrin::_InterlockedDecrement: {
     llvm::Type *IntTy = ConvertType(E->getType());
     AtomicRMWInst *RMWI = Builder.CreateAtomicRMW(
@@ -566,6 +594,37 @@ Value *CodeGenFunction::EmitMSVCBuiltinExpr(MSVCIntrin BuiltinID,
       llvm::AtomicOrdering::SequentiallyConsistent);
     return Builder.CreateAdd(RMWI, ConstantInt::get(IntTy, 1));
   }
+
+  case MSVCIntrin::__fastfail: {
+    // Request immediate process termination from the kernel. The instruction
+    // sequences to do this are documented on MSDN:
+    // https://msdn.microsoft.com/en-us/library/dn774154.aspx
+    llvm::Triple::ArchType ISA = getTarget().getTriple().getArch();
+    StringRef Asm, Constraints;
+    switch (ISA) {
+    default:
+      ErrorUnsupported(E, "__fastfail call for this architecture");
+      break;
+    case llvm::Triple::x86:
+    case llvm::Triple::x86_64:
+      Asm = "int $$0x29";
+      Constraints = "{cx}";
+      break;
+    case llvm::Triple::thumb:
+      Asm = "udf #251";
+      Constraints = "{r0}";
+      break;
+    }
+    llvm::FunctionType *FTy = llvm::FunctionType::get(VoidTy, {Int32Ty}, false);
+    llvm::InlineAsm *IA =
+        llvm::InlineAsm::get(FTy, Asm, Constraints, /*SideEffects=*/true);
+    llvm::AttributeList NoReturnAttr = llvm::AttributeList::get(
+        getLLVMContext(), llvm::AttributeList::FunctionIndex,
+        llvm::Attribute::NoReturn);
+    CallSite CS = Builder.CreateCall(IA, EmitScalarExpr(E->getArg(0)));
+    CS.setAttributes(NoReturnAttr);
+    return CS.getInstruction();
+  }
   }
   llvm_unreachable("Incorrect MSVC intrinsic!");
 }
@@ -932,7 +991,8 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
 
     // We pass this builtin onto the optimizer so that it can figure out the
     // object size in more complex cases.
-    return RValue::get(emitBuiltinObjectSize(E->getArg(0), Type, ResType));
+    return RValue::get(emitBuiltinObjectSize(E->getArg(0), Type, ResType,
+                                             /*EmittedE=*/nullptr));
   }
   case Builtin::BI__builtin_prefetch: {
     Value *Locality, *RW, *Address = EmitScalarExpr(E->getArg(0));
@@ -2195,16 +2255,9 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
   case Builtin::BI_InterlockedXor16:
   case Builtin::BI_InterlockedXor:
     return RValue::get(EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedXor, E));
-  case Builtin::BI__readfsdword: {
-    llvm::Type *IntTy = ConvertType(E->getType());
-    Value *IntToPtr =
-      Builder.CreateIntToPtr(EmitScalarExpr(E->getArg(0)),
-                             llvm::PointerType::get(IntTy, 257));
-    LoadInst *Load = Builder.CreateAlignedLoad(
-        IntTy, IntToPtr, getContext().getTypeAlignInChars(E->getType()));
-    Load->setVolatile(true);
-    return RValue::get(Load);
-  }
+  case Builtin::BI_interlockedbittestandset:
+    return RValue::get(
+        EmitMSVCBuiltinExpr(MSVCIntrin::_interlockedbittestandset, E));
 
   case Builtin::BI__exception_code:
   case Builtin::BI_exception_code:
@@ -2218,9 +2271,9 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
   case Builtin::BI_setjmpex: {
     if (getTarget().getTriple().isOSMSVCRT()) {
       llvm::Type *ArgTypes[] = {Int8PtrTy, Int8PtrTy};
-      llvm::AttributeSet ReturnsTwiceAttr =
-          AttributeSet::get(getLLVMContext(), llvm::AttributeSet::FunctionIndex,
-                            llvm::Attribute::ReturnsTwice);
+      llvm::AttributeList ReturnsTwiceAttr = llvm::AttributeList::get(
+          getLLVMContext(), llvm::AttributeList::FunctionIndex,
+          llvm::Attribute::ReturnsTwice);
       llvm::Constant *SetJmpEx = CGM.CreateRuntimeFunction(
           llvm::FunctionType::get(IntTy, ArgTypes, /*isVarArg=*/false),
           "_setjmpex", ReturnsTwiceAttr, /*Local=*/true);
@@ -2238,9 +2291,9 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
   }
   case Builtin::BI_setjmp: {
     if (getTarget().getTriple().isOSMSVCRT()) {
-      llvm::AttributeSet ReturnsTwiceAttr =
-          AttributeSet::get(getLLVMContext(), llvm::AttributeSet::FunctionIndex,
-                            llvm::Attribute::ReturnsTwice);
+      llvm::AttributeList ReturnsTwiceAttr = llvm::AttributeList::get(
+          getLLVMContext(), llvm::AttributeList::FunctionIndex,
+          llvm::Attribute::ReturnsTwice);
       llvm::Value *Buf = Builder.CreateBitOrPointerCast(
           EmitScalarExpr(E->getArg(0)), Int8PtrTy);
       llvm::CallSite CS;
@@ -2276,6 +2329,9 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
     break;
   }
 
+  case Builtin::BI__fastfail:
+    return RValue::get(EmitMSVCBuiltinExpr(MSVCIntrin::__fastfail, E));
+
   case Builtin::BI__builtin_coro_size: {
     auto & Context = getContext();
     auto SizeTy = Context.getSizeType();
@@ -2492,25 +2548,36 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
     unsigned NumArgs = E->getNumArgs();
 
     llvm::Type *QueueTy = ConvertType(getContext().OCLQueueTy);
-    llvm::Type *RangeTy = ConvertType(getContext().OCLNDRangeTy);
+    llvm::Type *GenericVoidPtrTy = Builder.getInt8PtrTy(
+        getContext().getTargetAddressSpace(LangAS::opencl_generic));
 
     llvm::Value *Queue = EmitScalarExpr(E->getArg(0));
     llvm::Value *Flags = EmitScalarExpr(E->getArg(1));
-    llvm::Value *Range = EmitScalarExpr(E->getArg(2));
+    LValue NDRangeL = EmitAggExprToLValue(E->getArg(2));
+    llvm::Value *Range = NDRangeL.getAddress().getPointer();
+    llvm::Type *RangeTy = NDRangeL.getAddress().getType();
 
     if (NumArgs == 4) {
       // The most basic form of the call with parameters:
       // queue_t, kernel_enqueue_flags_t, ndrange_t, block(void)
       Name = "__enqueue_kernel_basic";
-      llvm::Type *ArgTys[] = {QueueTy, Int32Ty, RangeTy, Int8PtrTy};
+      llvm::Type *ArgTys[] = {QueueTy, Int32Ty, RangeTy, GenericVoidPtrTy};
       llvm::FunctionType *FTy = llvm::FunctionType::get(
           Int32Ty, llvm::ArrayRef<llvm::Type *>(ArgTys, 4), false);
 
-      llvm::Value *Block =
-          Builder.CreateBitCast(EmitScalarExpr(E->getArg(3)), Int8PtrTy);
+      llvm::Value *Block = Builder.CreatePointerCast(
+          EmitScalarExpr(E->getArg(3)), GenericVoidPtrTy);
 
-      return RValue::get(Builder.CreateCall(
-          CGM.CreateRuntimeFunction(FTy, Name), {Queue, Flags, Range, Block}));
+      AttrBuilder B;
+      B.addAttribute(Attribute::ByVal);
+      llvm::AttributeList ByValAttrSet =
+          llvm::AttributeList::get(CGM.getModule().getContext(), 3U, B);
+
+      auto RTCall =
+          Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, Name, ByValAttrSet),
+                             {Queue, Flags, Range, Block});
+      RTCall->setAttributes(ByValAttrSet);
+      return RValue::get(RTCall);
     }
     assert(NumArgs >= 5 && "Invalid enqueue_kernel signature");
 
@@ -2518,14 +2585,14 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
     if (E->getArg(3)->getType()->isBlockPointerType()) {
       // No events passed, but has variadic arguments.
       Name = "__enqueue_kernel_vaargs";
-      llvm::Value *Block =
-          Builder.CreateBitCast(EmitScalarExpr(E->getArg(3)), Int8PtrTy);
+      llvm::Value *Block = Builder.CreatePointerCast(
+          EmitScalarExpr(E->getArg(3)), GenericVoidPtrTy);
       // Create a vector of the arguments, as well as a constant value to
       // express to the runtime the number of variadic arguments.
       std::vector<llvm::Value *> Args = {Queue, Flags, Range, Block,
                                          ConstantInt::get(IntTy, NumArgs - 4)};
-      std::vector<llvm::Type *> ArgTys = {QueueTy, IntTy, RangeTy, Int8PtrTy,
-                                          IntTy};
+      std::vector<llvm::Type *> ArgTys = {QueueTy, IntTy, RangeTy,
+                                          GenericVoidPtrTy, IntTy};
 
       // Each of the following arguments specifies the size of the corresponding
       // argument passed to the enqueued block.
@@ -2555,12 +2622,12 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
       // Convert to generic address space.
       EventList = Builder.CreatePointerCast(EventList, EventPtrTy);
       ClkEvent = Builder.CreatePointerCast(ClkEvent, EventPtrTy);
-      llvm::Value *Block =
-          Builder.CreateBitCast(EmitScalarExpr(E->getArg(6)), Int8PtrTy);
+      llvm::Value *Block = Builder.CreatePointerCast(
+          EmitScalarExpr(E->getArg(6)), GenericVoidPtrTy);
 
-      std::vector<llvm::Type *> ArgTys = {QueueTy,  Int32Ty,    RangeTy,
-                                          Int32Ty,  EventPtrTy, EventPtrTy,
-                                          Int8PtrTy};
+      std::vector<llvm::Type *> ArgTys = {
+          QueueTy,    Int32Ty,    RangeTy,         Int32Ty,
+          EventPtrTy, EventPtrTy, GenericVoidPtrTy};
 
       std::vector<llvm::Value *> Args = {Queue,     Flags,    Range, NumEvents,
                                          EventList, ClkEvent, Block};
@@ -2596,26 +2663,30 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
   // OpenCL v2.0 s6.13.17.6 - Kernel query functions need bitcast of block
   // parameter.
   case Builtin::BIget_kernel_work_group_size: {
+    llvm::Type *GenericVoidPtrTy = Builder.getInt8PtrTy(
+        getContext().getTargetAddressSpace(LangAS::opencl_generic));
     Value *Arg = EmitScalarExpr(E->getArg(0));
-    Arg = Builder.CreateBitCast(Arg, Int8PtrTy);
-    return RValue::get(
-        Builder.CreateCall(CGM.CreateRuntimeFunction(
-                               llvm::FunctionType::get(IntTy, Int8PtrTy, false),
-                               "__get_kernel_work_group_size_impl"),
-                           Arg));
+    Arg = Builder.CreatePointerCast(Arg, GenericVoidPtrTy);
+    return RValue::get(Builder.CreateCall(
+        CGM.CreateRuntimeFunction(
+            llvm::FunctionType::get(IntTy, GenericVoidPtrTy, false),
+            "__get_kernel_work_group_size_impl"),
+        Arg));
   }
   case Builtin::BIget_kernel_preferred_work_group_size_multiple: {
+    llvm::Type *GenericVoidPtrTy = Builder.getInt8PtrTy(
+        getContext().getTargetAddressSpace(LangAS::opencl_generic));
     Value *Arg = EmitScalarExpr(E->getArg(0));
-    Arg = Builder.CreateBitCast(Arg, Int8PtrTy);
+    Arg = Builder.CreatePointerCast(Arg, GenericVoidPtrTy);
     return RValue::get(Builder.CreateCall(
         CGM.CreateRuntimeFunction(
-            llvm::FunctionType::get(IntTy, Int8PtrTy, false),
+            llvm::FunctionType::get(IntTy, GenericVoidPtrTy, false),
             "__get_kernel_preferred_work_group_multiple_impl"),
         Arg));
   }
   case Builtin::BIprintf:
-    if (getLangOpts().CUDA && getLangOpts().CUDAIsDevice)
-      return EmitCUDADevicePrintfCallExpr(E, ReturnValue);
+    if (getTarget().getTriple().isNVPTX())
+      return EmitNVPTXDevicePrintfCallExpr(E, ReturnValue);
     break;
   case Builtin::BI__builtin_canonicalize:
   case Builtin::BI__builtin_canonicalizef:
@@ -7115,6 +7186,13 @@ static Value *EmitX86MinMax(CodeGenFunction &CGF, ICmpInst::Predicate Pred,
   return EmitX86Select(CGF, Ops[3], Res, Ops[2]);
 }
 
+static Value *EmitX86SExtMask(CodeGenFunction &CGF, Value *Op, 
+                              llvm::Type *DstTy) {
+  unsigned NumberOfElements = DstTy->getVectorNumElements();
+  Value *Mask = getMaskVecValue(CGF, Op, NumberOfElements);
+  return CGF.Builder.CreateSExt(Mask, DstTy, "vpmovm2");
+}
+
 Value *CodeGenFunction::EmitX86BuiltinExpr(unsigned BuiltinID,
                                            const CallExpr *E) {
   if (BuiltinID == X86::BI__builtin_ms_va_start ||
@@ -7321,7 +7399,12 @@ Value *CodeGenFunction::EmitX86BuiltinExpr(unsigned BuiltinID,
   case X86::BI__builtin_ia32_undef128:
   case X86::BI__builtin_ia32_undef256:
   case X86::BI__builtin_ia32_undef512:
-    return UndefValue::get(ConvertType(E->getType()));
+    // The x86 definition of "undef" is not the same as the LLVM definition
+    // (PR32176). We leave optimizing away an unnecessary zero constant to the
+    // IR optimizer and backend.
+    // TODO: If we had a "freeze" IR instruction to generate a fixed undef
+    // value, we should use that here instead of a zero.
+    return llvm::Constant::getNullValue(ConvertType(E->getType()));
   case X86::BI__builtin_ia32_vec_init_v8qi:
   case X86::BI__builtin_ia32_vec_init_v4hi:
   case X86::BI__builtin_ia32_vec_init_v2si:
@@ -7408,6 +7491,21 @@ Value *CodeGenFunction::EmitX86BuiltinExpr(unsigned BuiltinID,
   case X86::BI__builtin_ia32_storesd128_mask: {
     return EmitX86MaskedStore(*this, Ops, 16);
   }
+
+  case X86::BI__builtin_ia32_cvtmask2b128:
+  case X86::BI__builtin_ia32_cvtmask2b256:
+  case X86::BI__builtin_ia32_cvtmask2b512:
+  case X86::BI__builtin_ia32_cvtmask2w128:
+  case X86::BI__builtin_ia32_cvtmask2w256:
+  case X86::BI__builtin_ia32_cvtmask2w512:
+  case X86::BI__builtin_ia32_cvtmask2d128:
+  case X86::BI__builtin_ia32_cvtmask2d256:
+  case X86::BI__builtin_ia32_cvtmask2d512:
+  case X86::BI__builtin_ia32_cvtmask2q128:
+  case X86::BI__builtin_ia32_cvtmask2q256:
+  case X86::BI__builtin_ia32_cvtmask2q512:
+    return EmitX86SExtMask(*this, Ops[0], ConvertType(E->getType()));
+
   case X86::BI__builtin_ia32_movdqa32store128_mask:
   case X86::BI__builtin_ia32_movdqa64store128_mask:
   case X86::BI__builtin_ia32_storeaps128_mask:
@@ -7922,6 +8020,45 @@ Value *CodeGenFunction::EmitX86BuiltinExpr(unsigned BuiltinID,
     // instruction, but it will create a memset that won't be optimized away.
     return Builder.CreateMemSet(Ops[0], Ops[1], Ops[2], 1, true);
   }
+  case X86::BI__ud2:
+    // llvm.trap makes a ud2a instruction on x86.
+    return EmitTrapCall(Intrinsic::trap);
+  case X86::BI__int2c: {
+    // This syscall signals a driver assertion failure in x86 NT kernels.
+    llvm::FunctionType *FTy = llvm::FunctionType::get(VoidTy, false);
+    llvm::InlineAsm *IA =
+        llvm::InlineAsm::get(FTy, "int $$0x2c", "", /*SideEffects=*/true);
+    llvm::AttributeList NoReturnAttr = llvm::AttributeList::get(
+        getLLVMContext(), llvm::AttributeList::FunctionIndex,
+        llvm::Attribute::NoReturn);
+    CallSite CS = Builder.CreateCall(IA);
+    CS.setAttributes(NoReturnAttr);
+    return CS.getInstruction();
+  }
+  case X86::BI__readfsbyte:
+  case X86::BI__readfsword:
+  case X86::BI__readfsdword:
+  case X86::BI__readfsqword: {
+    llvm::Type *IntTy = ConvertType(E->getType());
+    Value *Ptr = Builder.CreateIntToPtr(EmitScalarExpr(E->getArg(0)),
+                                        llvm::PointerType::get(IntTy, 257));
+    LoadInst *Load = Builder.CreateAlignedLoad(
+        IntTy, Ptr, getContext().getTypeAlignInChars(E->getType()));
+    Load->setVolatile(true);
+    return Load;
+  }
+  case X86::BI__readgsbyte:
+  case X86::BI__readgsword:
+  case X86::BI__readgsdword:
+  case X86::BI__readgsqword: {
+    llvm::Type *IntTy = ConvertType(E->getType());
+    Value *Ptr = Builder.CreateIntToPtr(EmitScalarExpr(E->getArg(0)),
+                                        llvm::PointerType::get(IntTy, 256));
+    LoadInst *Load = Builder.CreateAlignedLoad(
+        IntTy, Ptr, getContext().getTypeAlignInChars(E->getType()));
+    Load->setVolatile(true);
+    return Load;
+  }
   }
 }
 
@@ -8326,6 +8463,14 @@ Value *CodeGenFunction::EmitAMDGPUBuiltinExpr(unsigned BuiltinID,
 
   case AMDGPU::BI__builtin_amdgcn_ds_swizzle:
     return emitBinaryBuiltin(*this, E, Intrinsic::amdgcn_ds_swizzle);
+  case AMDGPU::BI__builtin_amdgcn_mov_dpp: {
+    llvm::SmallVector<llvm::Value *, 5> Args;
+    for (unsigned I = 0; I != 5; ++I)
+      Args.push_back(EmitScalarExpr(E->getArg(I)));
+    Value *F = CGM.getIntrinsic(Intrinsic::amdgcn_mov_dpp,
+                                    Args[0]->getType());
+    return Builder.CreateCall(F, Args);
+  }
   case AMDGPU::BI__builtin_amdgcn_div_fixup:
   case AMDGPU::BI__builtin_amdgcn_div_fixupf:
   case AMDGPU::BI__builtin_amdgcn_div_fixuph:
@@ -8391,7 +8536,9 @@ Value *CodeGenFunction::EmitAMDGPUBuiltinExpr(unsigned BuiltinID,
   case AMDGPU::BI__builtin_amdgcn_classf:
   case AMDGPU::BI__builtin_amdgcn_classh:
     return emitFPIntBuiltin(*this, E, Intrinsic::amdgcn_class);
-
+  case AMDGPU::BI__builtin_amdgcn_fmed3f:
+  case AMDGPU::BI__builtin_amdgcn_fmed3h:
+    return emitTernaryBuiltin(*this, E, Intrinsic::amdgcn_fmed3);
   case AMDGPU::BI__builtin_amdgcn_read_exec: {
     CallInst *CI = cast<CallInst>(
       EmitSpecialRegisterBuiltin(*this, E, Int64Ty, Int64Ty, true, "exec"));
diff --git a/lib/CodeGen/CGCUDANV.cpp b/lib/CodeGen/CGCUDANV.cpp
index 83febcb4af8c..813cd7400186 100644
--- a/lib/CodeGen/CGCUDANV.cpp
+++ b/lib/CodeGen/CGCUDANV.cpp
@@ -15,7 +15,7 @@
 #include "CGCUDARuntime.h"
 #include "CodeGenFunction.h"
 #include "CodeGenModule.h"
-#include "ConstantBuilder.h"
+#include "clang/CodeGen/ConstantInitBuilder.h"
 #include "clang/AST/Decl.h"
 #include "llvm/IR/BasicBlock.h"
 #include "llvm/IR/CallSite.h"
diff --git a/lib/CodeGen/CGCXX.cpp b/lib/CodeGen/CGCXX.cpp
index 59010f4407c2..0f3141ab76d0 100644
--- a/lib/CodeGen/CGCXX.cpp
+++ b/lib/CodeGen/CGCXX.cpp
@@ -256,7 +256,7 @@ llvm::Constant *CodeGenModule::getAddrOfCXXStructor(
 
   return GetOrCreateLLVMFunction(
       getMangledName(GD), FnType, GD, /*ForVTable=*/false, DontDefer,
-      /*isThunk=*/false, /*ExtraAttrs=*/llvm::AttributeSet(), IsForDefinition);
+      /*isThunk=*/false, /*ExtraAttrs=*/llvm::AttributeList(), IsForDefinition);
 }
 
 static CGCallee BuildAppleKextVirtualCall(CodeGenFunction &CGF,
diff --git a/lib/CodeGen/CGCXXABI.h b/lib/CodeGen/CGCXXABI.h
index d53fd4cb63b2..7b912e3aca57 100644
--- a/lib/CodeGen/CGCXXABI.h
+++ b/lib/CodeGen/CGCXXABI.h
@@ -291,11 +291,26 @@ public:
   /// Emit constructor variants required by this ABI.
   virtual void EmitCXXConstructors(const CXXConstructorDecl *D) = 0;
 
+  /// Notes how many arguments were added to the beginning (Prefix) and ending
+  /// (Suffix) of an arg list.
+  ///
+  /// Note that Prefix actually refers to the number of args *after* the first
+  /// one: `this` arguments always come first.
+  struct AddedStructorArgs {
+    unsigned Prefix = 0;
+    unsigned Suffix = 0;
+    AddedStructorArgs() = default;
+    AddedStructorArgs(unsigned P, unsigned S) : Prefix(P), Suffix(S) {}
+    static AddedStructorArgs prefix(unsigned N) { return {N, 0}; }
+    static AddedStructorArgs suffix(unsigned N) { return {0, N}; }
+  };
+
   /// Build the signature of the given constructor or destructor variant by
   /// adding any required parameters.  For convenience, ArgTys has been
   /// initialized with the type of 'this'.
-  virtual void buildStructorSignature(const CXXMethodDecl *MD, StructorType T,
-                                      SmallVectorImpl<CanQualType> &ArgTys) = 0;
+  virtual AddedStructorArgs
+  buildStructorSignature(const CXXMethodDecl *MD, StructorType T,
+                         SmallVectorImpl<CanQualType> &ArgTys) = 0;
 
   /// Returns true if the given destructor type should be emitted as a linkonce
   /// delegating thunk, regardless of whether the dtor is defined in this TU or
@@ -355,9 +370,9 @@ public:
 
   /// Add any ABI-specific implicit arguments needed to call a constructor.
   ///
-  /// \return The number of args added to the call, which is typically zero or
-  /// one.
-  virtual unsigned
+  /// \return The number of arguments added at the beginning and end of the
+  /// call, which is typically zero or one.
+  virtual AddedStructorArgs
   addImplicitConstructorArgs(CodeGenFunction &CGF, const CXXConstructorDecl *D,
                              CXXCtorType Type, bool ForVirtualBase,
                              bool Delegating, CallArgList &Args) = 0;
@@ -377,7 +392,7 @@ public:
   isVirtualOffsetNeededForVTableField(CodeGenFunction &CGF,
                                       CodeGenFunction::VPtr Vptr) = 0;
 
-  /// Checks if ABI requires to initilize vptrs for given dynamic class.
+  /// Checks if ABI requires to initialize vptrs for given dynamic class.
   virtual bool doStructorsInitializeVPtrs(const CXXRecordDecl *VTableClass) = 0;
 
   /// Get the address point of the vtable for the given base subobject.
diff --git a/lib/CodeGen/CGCall.cpp b/lib/CodeGen/CGCall.cpp
index c7c61e0c8ecb..8af32055fc4c 100644
--- a/lib/CodeGen/CGCall.cpp
+++ b/lib/CodeGen/CGCall.cpp
@@ -101,39 +101,64 @@ CodeGenTypes::arrangeFreeFunctionType(CanQual<FunctionNoProtoType> FTNP) {
                                  FTNP->getExtInfo(), {}, RequiredArgs(0));
 }
 
+static void addExtParameterInfosForCall(
+         llvm::SmallVectorImpl<FunctionProtoType::ExtParameterInfo> &paramInfos,
+                                        const FunctionProtoType *proto,
+                                        unsigned prefixArgs,
+                                        unsigned totalArgs) {
+  assert(proto->hasExtParameterInfos());
+  assert(paramInfos.size() <= prefixArgs);
+  assert(proto->getNumParams() + prefixArgs <= totalArgs);
+
+  paramInfos.reserve(totalArgs);
+
+  // Add default infos for any prefix args that don't already have infos.
+  paramInfos.resize(prefixArgs);
+
+  // Add infos for the prototype.
+  for (const auto &ParamInfo : proto->getExtParameterInfos()) {
+    paramInfos.push_back(ParamInfo);
+    // pass_object_size params have no parameter info.
+    if (ParamInfo.hasPassObjectSize())
+      paramInfos.emplace_back();
+  }
+
+  assert(paramInfos.size() <= totalArgs &&
+         "Did we forget to insert pass_object_size args?");
+  // Add default infos for the variadic and/or suffix arguments.
+  paramInfos.resize(totalArgs);
+}
+
 /// Adds the formal paramaters in FPT to the given prefix. If any parameter in
 /// FPT has pass_object_size attrs, then we'll add parameters for those, too.
 static void appendParameterTypes(const CodeGenTypes &CGT,
                                  SmallVectorImpl<CanQualType> &prefix,
               SmallVectorImpl<FunctionProtoType::ExtParameterInfo> &paramInfos,
-                                 CanQual<FunctionProtoType> FPT,
-                                 const FunctionDecl *FD) {
-  // Fill out paramInfos.
-  if (FPT->hasExtParameterInfos() || !paramInfos.empty()) {
-    assert(paramInfos.size() <= prefix.size());
-    auto protoParamInfos = FPT->getExtParameterInfos();
-    paramInfos.reserve(prefix.size() + protoParamInfos.size());
-    paramInfos.resize(prefix.size());
-    paramInfos.append(protoParamInfos.begin(), protoParamInfos.end());
-  }
-
-  // Fast path: unknown target.
-  if (FD == nullptr) {
+                                 CanQual<FunctionProtoType> FPT) {
+  // Fast path: don't touch param info if we don't need to.
+  if (!FPT->hasExtParameterInfos()) {
+    assert(paramInfos.empty() &&
+           "We have paramInfos, but the prototype doesn't?");
     prefix.append(FPT->param_type_begin(), FPT->param_type_end());
     return;
   }
 
-  // In the vast majority cases, we'll have precisely FPT->getNumParams()
+  unsigned PrefixSize = prefix.size();
+  // In the vast majority of cases, we'll have precisely FPT->getNumParams()
   // parameters; the only thing that can change this is the presence of
   // pass_object_size. So, we preallocate for the common case.
   prefix.reserve(prefix.size() + FPT->getNumParams());
 
-  assert(FD->getNumParams() == FPT->getNumParams());
+  auto ExtInfos = FPT->getExtParameterInfos();
+  assert(ExtInfos.size() == FPT->getNumParams());
   for (unsigned I = 0, E = FPT->getNumParams(); I != E; ++I) {
     prefix.push_back(FPT->getParamType(I));
-    if (FD->getParamDecl(I)->hasAttr<PassObjectSizeAttr>())
+    if (ExtInfos[I].hasPassObjectSize())
       prefix.push_back(CGT.getContext().getSizeType());
   }
+
+  addExtParameterInfosForCall(paramInfos, FPT.getTypePtr(), PrefixSize,
+                              prefix.size());
 }
 
 /// Arrange the LLVM function layout for a value of the given function
@@ -147,7 +172,7 @@ arrangeLLVMFunctionInfo(CodeGenTypes &CGT, bool instanceMethod,
   RequiredArgs Required =
       RequiredArgs::forPrototypePlus(FTP, prefix.size(), FD);
   // FIXME: Kill copy.
-  appendParameterTypes(CGT, prefix, paramInfos, FTP, FD);
+  appendParameterTypes(CGT, prefix, paramInfos, FTP);
   CanQualType resultType = FTP->getReturnType().getUnqualifiedType();
 
   return CGT.arrangeLLVMFunctionInfo(resultType, instanceMethod,
@@ -286,9 +311,19 @@ CodeGenTypes::arrangeCXXStructorDeclaration(const CXXMethodDecl *MD,
 
   // Add the formal parameters.
   if (PassParams)
-    appendParameterTypes(*this, argTypes, paramInfos, FTP, MD);
-
-  TheCXXABI.buildStructorSignature(MD, Type, argTypes);
+    appendParameterTypes(*this, argTypes, paramInfos, FTP);
+
+  CGCXXABI::AddedStructorArgs AddedArgs =
+      TheCXXABI.buildStructorSignature(MD, Type, argTypes);
+  if (!paramInfos.empty()) {
+    // Note: prefix implies after the first param.
+    if (AddedArgs.Prefix)
+      paramInfos.insert(paramInfos.begin() + 1, AddedArgs.Prefix,
+                        FunctionProtoType::ExtParameterInfo{});
+    if (AddedArgs.Suffix)
+      paramInfos.append(AddedArgs.Suffix,
+                        FunctionProtoType::ExtParameterInfo{});
+  }
 
   RequiredArgs required =
       (PassParams && MD->isVariadic() ? RequiredArgs(argTypes.size())
@@ -321,26 +356,6 @@ getArgTypesForDeclaration(ASTContext &ctx, const FunctionArgList &args) {
   return argTypes;
 }
 
-static void addExtParameterInfosForCall(
-         llvm::SmallVectorImpl<FunctionProtoType::ExtParameterInfo> &paramInfos,
-                                        const FunctionProtoType *proto,
-                                        unsigned prefixArgs,
-                                        unsigned totalArgs) {
-  assert(proto->hasExtParameterInfos());
-  assert(paramInfos.size() <= prefixArgs);
-  assert(proto->getNumParams() + prefixArgs <= totalArgs);
-
-  // Add default infos for any prefix args that don't already have infos.
-  paramInfos.resize(prefixArgs);
-
-  // Add infos for the prototype.
-  auto protoInfos = proto->getExtParameterInfos();
-  paramInfos.append(protoInfos.begin(), protoInfos.end());
-
-  // Add default infos for the variadic arguments.
-  paramInfos.resize(totalArgs);
-}
-
 static llvm::SmallVector<FunctionProtoType::ExtParameterInfo, 16>
 getExtParameterInfosForCall(const FunctionProtoType *proto,
                             unsigned prefixArgs, unsigned totalArgs) {
@@ -352,18 +367,31 @@ getExtParameterInfosForCall(const FunctionProtoType *proto,
 }
 
 /// Arrange a call to a C++ method, passing the given arguments.
+///
+/// ExtraPrefixArgs is the number of ABI-specific args passed after the `this`
+/// parameter.
+/// ExtraSuffixArgs is the number of ABI-specific args passed at the end of
+/// args.
+/// PassProtoArgs indicates whether `args` has args for the parameters in the
+/// given CXXConstructorDecl.
 const CGFunctionInfo &
 CodeGenTypes::arrangeCXXConstructorCall(const CallArgList &args,
                                         const CXXConstructorDecl *D,
                                         CXXCtorType CtorKind,
-                                        unsigned ExtraArgs) {
+                                        unsigned ExtraPrefixArgs,
+                                        unsigned ExtraSuffixArgs,
+                                        bool PassProtoArgs) {
   // FIXME: Kill copy.
   SmallVector<CanQualType, 16> ArgTypes;
   for (const auto &Arg : args)
     ArgTypes.push_back(Context.getCanonicalParamType(Arg.Ty));
 
+  // +1 for implicit this, which should always be args[0].
+  unsigned TotalPrefixArgs = 1 + ExtraPrefixArgs;
+
   CanQual<FunctionProtoType> FPT = GetFormalType(D);
-  RequiredArgs Required = RequiredArgs::forPrototypePlus(FPT, 1 + ExtraArgs, D);
+  RequiredArgs Required =
+      RequiredArgs::forPrototypePlus(FPT, TotalPrefixArgs + ExtraSuffixArgs, D);
   GlobalDecl GD(D, CtorKind);
   CanQualType ResultType = TheCXXABI.HasThisReturn(GD)
                                ? ArgTypes.front()
@@ -372,8 +400,14 @@ CodeGenTypes::arrangeCXXConstructorCall(const CallArgList &args,
                                      : Context.VoidTy;
 
   FunctionType::ExtInfo Info = FPT->getExtInfo();
-  auto ParamInfos = getExtParameterInfosForCall(FPT.getTypePtr(), 1 + ExtraArgs,
-                                                ArgTypes.size());
+  llvm::SmallVector<FunctionProtoType::ExtParameterInfo, 16> ParamInfos;
+  // If the prototype args are elided, we should only have ABI-specific args,
+  // which never have param info.
+  if (PassProtoArgs && FPT->hasExtParameterInfos()) {
+    // ABI-specific suffix arguments are treated the same as variadic arguments.
+    addExtParameterInfosForCall(ParamInfos, FPT.getTypePtr(), TotalPrefixArgs,
+                                ArgTypes.size());
+  }
   return arrangeLLVMFunctionInfo(ResultType, /*instanceMethod=*/true,
                                  /*chainCall=*/false, ArgTypes, Info,
                                  ParamInfos, Required);
@@ -617,15 +651,20 @@ CodeGenTypes::arrangeBuiltinFunctionDeclaration(CanQualType resultType,
 }
 
 /// Arrange a call to a C++ method, passing the given arguments.
+///
+/// numPrefixArgs is the number of ABI-specific prefix arguments we have. It
+/// does not count `this`.
 const CGFunctionInfo &
 CodeGenTypes::arrangeCXXMethodCall(const CallArgList &args,
                                    const FunctionProtoType *proto,
-                                   RequiredArgs required) {
-  unsigned numRequiredArgs =
-    (proto->isVariadic() ? required.getNumRequiredArgs() : args.size());
-  unsigned numPrefixArgs = numRequiredArgs - proto->getNumParams();
+                                   RequiredArgs required,
+                                   unsigned numPrefixArgs) {
+  assert(numPrefixArgs + 1 <= args.size() &&
+         "Emitting a call with less args than the required prefix?");
+  // Add one to account for `this`. It's a bit awkward here, but we don't count
+  // `this` in similar places elsewhere.
   auto paramInfos =
-    getExtParameterInfosForCall(proto, numPrefixArgs, args.size());
+    getExtParameterInfosForCall(proto, numPrefixArgs + 1, args.size());
 
   // FIXME: Kill copy.
   auto argTypes = getArgTypesForCall(Context, args);
@@ -680,7 +719,7 @@ CodeGenTypes::arrangeLLVMFunctionInfo(CanQualType resultType,
                      ArrayRef<FunctionProtoType::ExtParameterInfo> paramInfos,
                                       RequiredArgs required) {
   assert(std::all_of(argTypes.begin(), argTypes.end(),
-                     std::mem_fun_ref(&CanQualType::isCanonicalAsParam)));
+                     [](CanQualType T) { return T.isCanonicalAsParam(); }));
 
   // Lookup or create unique function info.
   llvm::FoldingSetNodeID ID;
@@ -1620,15 +1659,113 @@ static void AddAttributesFromFunctionProtoType(ASTContext &Ctx,
     FuncAttrs.addAttribute(llvm::Attribute::NoUnwind);
 }
 
+void CodeGenModule::ConstructDefaultFnAttrList(StringRef Name, bool HasOptnone,
+                                               bool AttrOnCallSite,
+                                               llvm::AttrBuilder &FuncAttrs) {
+  // OptimizeNoneAttr takes precedence over -Os or -Oz. No warning needed.
+  if (!HasOptnone) {
+    if (CodeGenOpts.OptimizeSize)
+      FuncAttrs.addAttribute(llvm::Attribute::OptimizeForSize);
+    if (CodeGenOpts.OptimizeSize == 2)
+      FuncAttrs.addAttribute(llvm::Attribute::MinSize);
+  }
+
+  if (CodeGenOpts.DisableRedZone)
+    FuncAttrs.addAttribute(llvm::Attribute::NoRedZone);
+  if (CodeGenOpts.NoImplicitFloat)
+    FuncAttrs.addAttribute(llvm::Attribute::NoImplicitFloat);
+
+  if (AttrOnCallSite) {
+    // Attributes that should go on the call site only.
+    if (!CodeGenOpts.SimplifyLibCalls ||
+        CodeGenOpts.isNoBuiltinFunc(Name.data()))
+      FuncAttrs.addAttribute(llvm::Attribute::NoBuiltin);
+    if (!CodeGenOpts.TrapFuncName.empty())
+      FuncAttrs.addAttribute("trap-func-name", CodeGenOpts.TrapFuncName);
+  } else {
+    // Attributes that should go on the function, but not the call site.
+    if (!CodeGenOpts.DisableFPElim) {
+      FuncAttrs.addAttribute("no-frame-pointer-elim", "false");
+    } else if (CodeGenOpts.OmitLeafFramePointer) {
+      FuncAttrs.addAttribute("no-frame-pointer-elim", "false");
+      FuncAttrs.addAttribute("no-frame-pointer-elim-non-leaf");
+    } else {
+      FuncAttrs.addAttribute("no-frame-pointer-elim", "true");
+      FuncAttrs.addAttribute("no-frame-pointer-elim-non-leaf");
+    }
+
+    FuncAttrs.addAttribute("less-precise-fpmad",
+                           llvm::toStringRef(CodeGenOpts.LessPreciseFPMAD));
+
+    if (!CodeGenOpts.FPDenormalMode.empty())
+      FuncAttrs.addAttribute("denormal-fp-math", CodeGenOpts.FPDenormalMode);
+
+    FuncAttrs.addAttribute("no-trapping-math",
+                           llvm::toStringRef(CodeGenOpts.NoTrappingMath));
+
+    // TODO: Are these all needed?
+    // unsafe/inf/nan/nsz are handled by instruction-level FastMathFlags.
+    FuncAttrs.addAttribute("no-infs-fp-math",
+                           llvm::toStringRef(CodeGenOpts.NoInfsFPMath));
+    FuncAttrs.addAttribute("no-nans-fp-math",
+                           llvm::toStringRef(CodeGenOpts.NoNaNsFPMath));
+    FuncAttrs.addAttribute("unsafe-fp-math",
+                           llvm::toStringRef(CodeGenOpts.UnsafeFPMath));
+    FuncAttrs.addAttribute("use-soft-float",
+                           llvm::toStringRef(CodeGenOpts.SoftFloat));
+    FuncAttrs.addAttribute("stack-protector-buffer-size",
+                           llvm::utostr(CodeGenOpts.SSPBufferSize));
+    FuncAttrs.addAttribute("no-signed-zeros-fp-math",
+                           llvm::toStringRef(CodeGenOpts.NoSignedZeros));
+    FuncAttrs.addAttribute(
+        "correctly-rounded-divide-sqrt-fp-math",
+        llvm::toStringRef(CodeGenOpts.CorrectlyRoundedDivSqrt));
+
+    // TODO: Reciprocal estimate codegen options should apply to instructions?
+    std::vector<std::string> &Recips = getTarget().getTargetOpts().Reciprocals;
+    if (!Recips.empty())
+      FuncAttrs.addAttribute("reciprocal-estimates",
+                             llvm::join(Recips.begin(), Recips.end(), ","));
+
+    if (CodeGenOpts.StackRealignment)
+      FuncAttrs.addAttribute("stackrealign");
+    if (CodeGenOpts.Backchain)
+      FuncAttrs.addAttribute("backchain");
+  }
+
+  if (getLangOpts().CUDA && getLangOpts().CUDAIsDevice) {
+    // Conservatively, mark all functions and calls in CUDA as convergent
+    // (meaning, they may call an intrinsically convergent op, such as
+    // __syncthreads(), and so can't have certain optimizations applied around
+    // them).  LLVM will remove this attribute where it safely can.
+    FuncAttrs.addAttribute(llvm::Attribute::Convergent);
+
+    // Exceptions aren't supported in CUDA device code.
+    FuncAttrs.addAttribute(llvm::Attribute::NoUnwind);
+
+    // Respect -fcuda-flush-denormals-to-zero.
+    if (getLangOpts().CUDADeviceFlushDenormalsToZero)
+      FuncAttrs.addAttribute("nvptx-f32ftz", "true");
+  }
+}
+
+void CodeGenModule::AddDefaultFnAttrs(llvm::Function &F) {
+  llvm::AttrBuilder FuncAttrs;
+  ConstructDefaultFnAttrList(F.getName(),
+                             F.hasFnAttribute(llvm::Attribute::OptimizeNone),
+                             /* AttrOnCallsite = */ false, FuncAttrs);
+  llvm::AttributeList AS = llvm::AttributeList::get(
+      getLLVMContext(), llvm::AttributeList::FunctionIndex, FuncAttrs);
+  F.addAttributes(llvm::AttributeList::FunctionIndex, AS);
+}
+
 void CodeGenModule::ConstructAttributeList(
     StringRef Name, const CGFunctionInfo &FI, CGCalleeInfo CalleeInfo,
     AttributeListType &PAL, unsigned &CallingConv, bool AttrOnCallSite) {
   llvm::AttrBuilder FuncAttrs;
   llvm::AttrBuilder RetAttrs;
-  bool HasOptnone = false;
 
   CallingConv = FI.getEffectiveCallingConvention();
-
   if (FI.isNoReturn())
     FuncAttrs.addAttribute(llvm::Attribute::NoReturn);
 
@@ -1639,7 +1776,7 @@ void CodeGenModule::ConstructAttributeList(
 
   const Decl *TargetDecl = CalleeInfo.getCalleeDecl();
 
-  bool HasAnyX86InterruptAttr = false;
+  bool HasOptnone = false;
   // FIXME: handle sseregparm someday...
   if (TargetDecl) {
     if (TargetDecl->hasAttr<ReturnsTwiceAttr>())
@@ -1679,7 +1816,6 @@ void CodeGenModule::ConstructAttributeList(
     if (TargetDecl->hasAttr<ReturnsNonNullAttr>())
       RetAttrs.addAttribute(llvm::Attribute::NonNull);
 
-    HasAnyX86InterruptAttr = TargetDecl->hasAttr<AnyX86InterruptAttr>();
     HasOptnone = TargetDecl->hasAttr<OptimizeNoneAttr>();
     if (auto *AllocSize = TargetDecl->getAttr<AllocSizeAttr>()) {
       Optional<unsigned> NumElemsParam;
@@ -1691,86 +1827,19 @@ void CodeGenModule::ConstructAttributeList(
     }
   }
 
-  // OptimizeNoneAttr takes precedence over -Os or -Oz. No warning needed.
-  if (!HasOptnone) {
-    if (CodeGenOpts.OptimizeSize)
-      FuncAttrs.addAttribute(llvm::Attribute::OptimizeForSize);
-    if (CodeGenOpts.OptimizeSize == 2)
-      FuncAttrs.addAttribute(llvm::Attribute::MinSize);
-  }
+  ConstructDefaultFnAttrList(Name, HasOptnone, AttrOnCallSite, FuncAttrs);
 
-  if (CodeGenOpts.DisableRedZone)
-    FuncAttrs.addAttribute(llvm::Attribute::NoRedZone);
-  if (CodeGenOpts.NoImplicitFloat)
-    FuncAttrs.addAttribute(llvm::Attribute::NoImplicitFloat);
   if (CodeGenOpts.EnableSegmentedStacks &&
       !(TargetDecl && TargetDecl->hasAttr<NoSplitStackAttr>()))
     FuncAttrs.addAttribute("split-stack");
 
-  if (AttrOnCallSite) {
-    // Attributes that should go on the call site only.
-    if (!CodeGenOpts.SimplifyLibCalls ||
-        CodeGenOpts.isNoBuiltinFunc(Name.data()))
-      FuncAttrs.addAttribute(llvm::Attribute::NoBuiltin);
-    if (!CodeGenOpts.TrapFuncName.empty())
-      FuncAttrs.addAttribute("trap-func-name", CodeGenOpts.TrapFuncName);
-  } else {
-    // Attributes that should go on the function, but not the call site.
-    if (!CodeGenOpts.DisableFPElim) {
-      FuncAttrs.addAttribute("no-frame-pointer-elim", "false");
-    } else if (CodeGenOpts.OmitLeafFramePointer) {
-      FuncAttrs.addAttribute("no-frame-pointer-elim", "false");
-      FuncAttrs.addAttribute("no-frame-pointer-elim-non-leaf");
-    } else {
-      FuncAttrs.addAttribute("no-frame-pointer-elim", "true");
-      FuncAttrs.addAttribute("no-frame-pointer-elim-non-leaf");
-    }
-
+  if (!AttrOnCallSite) {
     bool DisableTailCalls =
-        CodeGenOpts.DisableTailCalls || HasAnyX86InterruptAttr ||
-        (TargetDecl && TargetDecl->hasAttr<DisableTailCallsAttr>());
-    FuncAttrs.addAttribute(
-        "disable-tail-calls",
-        llvm::toStringRef(DisableTailCalls));
-
-    FuncAttrs.addAttribute("less-precise-fpmad",
-                           llvm::toStringRef(CodeGenOpts.LessPreciseFPMAD));
-
-    if (!CodeGenOpts.FPDenormalMode.empty())
-      FuncAttrs.addAttribute("denormal-fp-math",
-                             CodeGenOpts.FPDenormalMode);
-
-    FuncAttrs.addAttribute("no-trapping-math",
-                           llvm::toStringRef(CodeGenOpts.NoTrappingMath));
-
-    // TODO: Are these all needed?
-    // unsafe/inf/nan/nsz are handled by instruction-level FastMathFlags.
-    FuncAttrs.addAttribute("no-infs-fp-math",
-                           llvm::toStringRef(CodeGenOpts.NoInfsFPMath));
-    FuncAttrs.addAttribute("no-nans-fp-math",
-                           llvm::toStringRef(CodeGenOpts.NoNaNsFPMath));
-    FuncAttrs.addAttribute("unsafe-fp-math",
-                           llvm::toStringRef(CodeGenOpts.UnsafeFPMath));
-    FuncAttrs.addAttribute("use-soft-float",
-                           llvm::toStringRef(CodeGenOpts.SoftFloat));
-    FuncAttrs.addAttribute("stack-protector-buffer-size",
-                           llvm::utostr(CodeGenOpts.SSPBufferSize));
-    FuncAttrs.addAttribute("no-signed-zeros-fp-math",
-                           llvm::toStringRef(CodeGenOpts.NoSignedZeros));
-    FuncAttrs.addAttribute(
-        "correctly-rounded-divide-sqrt-fp-math",
-        llvm::toStringRef(CodeGenOpts.CorrectlyRoundedDivSqrt));
-
-    // TODO: Reciprocal estimate codegen options should apply to instructions?
-    std::vector<std::string> &Recips = getTarget().getTargetOpts().Reciprocals;
-    if (!Recips.empty())
-      FuncAttrs.addAttribute("reciprocal-estimates",
-                             llvm::join(Recips.begin(), Recips.end(), ","));
-
-    if (CodeGenOpts.StackRealignment)
-      FuncAttrs.addAttribute("stackrealign");
-    if (CodeGenOpts.Backchain)
-      FuncAttrs.addAttribute("backchain");
+        CodeGenOpts.DisableTailCalls ||
+        (TargetDecl && (TargetDecl->hasAttr<DisableTailCallsAttr>() ||
+                        TargetDecl->hasAttr<AnyX86InterruptAttr>()));
+    FuncAttrs.addAttribute("disable-tail-calls",
+                           llvm::toStringRef(DisableTailCalls));
 
     // Add target-cpu and target-features attributes to functions. If
     // we have a decl for the function and it has a target attribute then
@@ -1819,21 +1888,6 @@ void CodeGenModule::ConstructAttributeList(
     }
   }
 
-  if (getLangOpts().CUDA && getLangOpts().CUDAIsDevice) {
-    // Conservatively, mark all functions and calls in CUDA as convergent
-    // (meaning, they may call an intrinsically convergent op, such as
-    // __syncthreads(), and so can't have certain optimizations applied around
-    // them).  LLVM will remove this attribute where it safely can.
-    FuncAttrs.addAttribute(llvm::Attribute::Convergent);
-
-    // Exceptions aren't supported in CUDA device code.
-    FuncAttrs.addAttribute(llvm::Attribute::NoUnwind);
-
-    // Respect -fcuda-flush-denormals-to-zero.
-    if (getLangOpts().CUDADeviceFlushDenormalsToZero)
-      FuncAttrs.addAttribute("nvptx-f32ftz", "true");
-  }
-
   ClangToLLVMArgMapping IRFunctionArgs(getContext(), FI);
 
   QualType RetTy = FI.getReturnType();
@@ -1878,8 +1932,8 @@ void CodeGenModule::ConstructAttributeList(
 
   // Attach return attributes.
   if (RetAttrs.hasAttributes()) {
-    PAL.push_back(llvm::AttributeSet::get(
-        getLLVMContext(), llvm::AttributeSet::ReturnIndex, RetAttrs));
+    PAL.push_back(llvm::AttributeList::get(
+        getLLVMContext(), llvm::AttributeList::ReturnIndex, RetAttrs));
   }
 
   bool hasUsedSRet = false;
@@ -1891,7 +1945,7 @@ void CodeGenModule::ConstructAttributeList(
     hasUsedSRet = true;
     if (RetAI.getInReg())
       SRETAttrs.addAttribute(llvm::Attribute::InReg);
-    PAL.push_back(llvm::AttributeSet::get(
+    PAL.push_back(llvm::AttributeList::get(
         getLLVMContext(), IRFunctionArgs.getSRetArgNo() + 1, SRETAttrs));
   }
 
@@ -1899,7 +1953,7 @@ void CodeGenModule::ConstructAttributeList(
   if (IRFunctionArgs.hasInallocaArg()) {
     llvm::AttrBuilder Attrs;
     Attrs.addAttribute(llvm::Attribute::InAlloca);
-    PAL.push_back(llvm::AttributeSet::get(
+    PAL.push_back(llvm::AttributeList::get(
         getLLVMContext(), IRFunctionArgs.getInallocaArgNo() + 1, Attrs));
   }
 
@@ -1914,7 +1968,7 @@ void CodeGenModule::ConstructAttributeList(
     // Add attribute for padding argument, if necessary.
     if (IRFunctionArgs.hasPaddingArg(ArgNo)) {
       if (AI.getPaddingInReg())
-        PAL.push_back(llvm::AttributeSet::get(
+        PAL.push_back(llvm::AttributeList::get(
             getLLVMContext(), IRFunctionArgs.getPaddingArgNo(ArgNo) + 1,
             llvm::Attribute::InReg));
     }
@@ -2031,17 +2085,15 @@ void CodeGenModule::ConstructAttributeList(
       unsigned FirstIRArg, NumIRArgs;
       std::tie(FirstIRArg, NumIRArgs) = IRFunctionArgs.getIRArgs(ArgNo);
       for (unsigned i = 0; i < NumIRArgs; i++)
-        PAL.push_back(llvm::AttributeSet::get(getLLVMContext(),
-                                              FirstIRArg + i + 1, Attrs));
+        PAL.push_back(llvm::AttributeList::get(getLLVMContext(),
+                                               FirstIRArg + i + 1, Attrs));
     }
   }
   assert(ArgNo == FI.arg_size());
 
   if (FuncAttrs.hasAttributes())
-    PAL.push_back(llvm::
-                  AttributeSet::get(getLLVMContext(),
-                                    llvm::AttributeSet::FunctionIndex,
-                                    FuncAttrs));
+    PAL.push_back(llvm::AttributeList::get(
+        getLLVMContext(), llvm::AttributeList::FunctionIndex, FuncAttrs));
 }
 
 /// An argument came in as a promoted argument; demote it back to its
@@ -2152,8 +2204,8 @@ void CodeGenFunction::EmitFunctionProlog(const CGFunctionInfo &FI,
   if (IRFunctionArgs.hasSRetArg()) {
     auto AI = cast<llvm::Argument>(FnArgs[IRFunctionArgs.getSRetArgNo()]);
     AI->setName("agg.result");
-    AI->addAttr(llvm::AttributeSet::get(getLLVMContext(), AI->getArgNo() + 1,
-                                        llvm::Attribute::NoAlias));
+    AI->addAttr(llvm::AttributeList::get(getLLVMContext(), AI->getArgNo() + 1,
+                                         llvm::Attribute::NoAlias));
   }
 
   // Track if we received the parameter as a pointer (indirect, byval, or
@@ -2244,9 +2296,9 @@ void CodeGenFunction::EmitFunctionProlog(const CGFunctionInfo &FI,
         if (const ParmVarDecl *PVD = dyn_cast<ParmVarDecl>(Arg)) {
           if (getNonNullAttr(CurCodeDecl, PVD, PVD->getType(),
                              PVD->getFunctionScopeIndex()))
-            AI->addAttr(llvm::AttributeSet::get(getLLVMContext(),
-                                                AI->getArgNo() + 1,
-                                                llvm::Attribute::NonNull));
+            AI->addAttr(llvm::AttributeList::get(getLLVMContext(),
+                                                 AI->getArgNo() + 1,
+                                                 llvm::Attribute::NonNull));
 
           QualType OTy = PVD->getOriginalType();
           if (const auto *ArrTy =
@@ -2263,12 +2315,12 @@ void CodeGenFunction::EmitFunctionProlog(const CGFunctionInfo &FI,
                 llvm::AttrBuilder Attrs;
                 Attrs.addDereferenceableAttr(
                   getContext().getTypeSizeInChars(ETy).getQuantity()*ArrSize);
-                AI->addAttr(llvm::AttributeSet::get(getLLVMContext(),
-                                                    AI->getArgNo() + 1, Attrs));
+                AI->addAttr(llvm::AttributeList::get(
+                    getLLVMContext(), AI->getArgNo() + 1, Attrs));
               } else if (getContext().getTargetAddressSpace(ETy) == 0) {
-                AI->addAttr(llvm::AttributeSet::get(getLLVMContext(),
-                                                    AI->getArgNo() + 1,
-                                                    llvm::Attribute::NonNull));
+                AI->addAttr(llvm::AttributeList::get(getLLVMContext(),
+                                                     AI->getArgNo() + 1,
+                                                     llvm::Attribute::NonNull));
               }
             }
           } else if (const auto *ArrTy =
@@ -2278,9 +2330,9 @@ void CodeGenFunction::EmitFunctionProlog(const CGFunctionInfo &FI,
             // we know that it must be nonnull.
             if (ArrTy->getSizeModifier() == VariableArrayType::Static &&
                 !getContext().getTargetAddressSpace(ArrTy->getElementType()))
-              AI->addAttr(llvm::AttributeSet::get(getLLVMContext(),
-                                                  AI->getArgNo() + 1,
-                                                  llvm::Attribute::NonNull));
+              AI->addAttr(llvm::AttributeList::get(getLLVMContext(),
+                                                   AI->getArgNo() + 1,
+                                                   llvm::Attribute::NonNull));
           }
 
           const auto *AVAttr = PVD->getAttr<AlignValueAttr>();
@@ -2298,15 +2350,14 @@ void CodeGenFunction::EmitFunctionProlog(const CGFunctionInfo &FI,
 
             llvm::AttrBuilder Attrs;
             Attrs.addAlignmentAttr(Alignment);
-            AI->addAttr(llvm::AttributeSet::get(getLLVMContext(),
-                                                AI->getArgNo() + 1, Attrs));
+            AI->addAttr(llvm::AttributeList::get(getLLVMContext(),
+                                                 AI->getArgNo() + 1, Attrs));
           }
         }
 
         if (Arg->getType().isRestrictQualified())
-          AI->addAttr(llvm::AttributeSet::get(getLLVMContext(),
-                                              AI->getArgNo() + 1,
-                                              llvm::Attribute::NoAlias));
+          AI->addAttr(llvm::AttributeList::get(
+              getLLVMContext(), AI->getArgNo() + 1, llvm::Attribute::NoAlias));
 
         // LLVM expects swifterror parameters to be used in very restricted
         // ways.  Copy the value into a less-restricted temporary.
@@ -2858,19 +2909,7 @@ void CodeGenFunction::EmitFunctionEpilog(const CGFunctionInfo &FI,
 
   llvm::Instruction *Ret;
   if (RV) {
-    if (CurCodeDecl && SanOpts.has(SanitizerKind::ReturnsNonnullAttribute)) {
-      if (auto RetNNAttr = CurCodeDecl->getAttr<ReturnsNonNullAttr>()) {
-        SanitizerScope SanScope(this);
-        llvm::Value *Cond = Builder.CreateICmpNE(
-            RV, llvm::Constant::getNullValue(RV->getType()));
-        llvm::Constant *StaticData[] = {
-            EmitCheckSourceLocation(EndLoc),
-            EmitCheckSourceLocation(RetNNAttr->getLocation()),
-        };
-        EmitCheck(std::make_pair(Cond, SanitizerKind::ReturnsNonnullAttribute),
-                  SanitizerHandler::NonnullReturn, StaticData, None);
-      }
-    }
+    EmitReturnValueCheck(RV, EndLoc);
     Ret = Builder.CreateRet(RV);
   } else {
     Ret = Builder.CreateRetVoid();
@@ -2880,6 +2919,63 @@ void CodeGenFunction::EmitFunctionEpilog(const CGFunctionInfo &FI,
     Ret->setDebugLoc(std::move(RetDbgLoc));
 }
 
+void CodeGenFunction::EmitReturnValueCheck(llvm::Value *RV,
+                                           SourceLocation EndLoc) {
+  // A current decl may not be available when emitting vtable thunks.
+  if (!CurCodeDecl)
+    return;
+
+  ReturnsNonNullAttr *RetNNAttr = nullptr;
+  if (SanOpts.has(SanitizerKind::ReturnsNonnullAttribute))
+    RetNNAttr = CurCodeDecl->getAttr<ReturnsNonNullAttr>();
+
+  if (!RetNNAttr && !requiresReturnValueNullabilityCheck())
+    return;
+
+  // Prefer the returns_nonnull attribute if it's present.
+  SourceLocation AttrLoc;
+  SanitizerMask CheckKind;
+  SanitizerHandler Handler;
+  if (RetNNAttr) {
+    assert(!requiresReturnValueNullabilityCheck() &&
+           "Cannot check nullability and the nonnull attribute");
+    AttrLoc = RetNNAttr->getLocation();
+    CheckKind = SanitizerKind::ReturnsNonnullAttribute;
+    Handler = SanitizerHandler::NonnullReturn;
+  } else {
+    if (auto *DD = dyn_cast<DeclaratorDecl>(CurCodeDecl))
+      if (auto *TSI = DD->getTypeSourceInfo())
+        if (auto FTL = TSI->getTypeLoc().castAs<FunctionTypeLoc>())
+          AttrLoc = FTL.getReturnLoc().findNullabilityLoc();
+    CheckKind = SanitizerKind::NullabilityReturn;
+    Handler = SanitizerHandler::NullabilityReturn;
+  }
+
+  SanitizerScope SanScope(this);
+
+  llvm::BasicBlock *Check = nullptr;
+  llvm::BasicBlock *NoCheck = nullptr;
+  if (requiresReturnValueNullabilityCheck()) {
+    // Before doing the nullability check, make sure that the preconditions for
+    // the check are met.
+    Check = createBasicBlock("nullcheck");
+    NoCheck = createBasicBlock("no.nullcheck");
+    Builder.CreateCondBr(RetValNullabilityPrecondition, Check, NoCheck);
+    EmitBlock(Check);
+  }
+
+  // Now do the null check. If the returns_nonnull attribute is present, this
+  // is done unconditionally.
+  llvm::Value *Cond = Builder.CreateIsNotNull(RV);
+  llvm::Constant *StaticData[] = {
+      EmitCheckSourceLocation(EndLoc), EmitCheckSourceLocation(AttrLoc),
+  };
+  EmitCheck(std::make_pair(Cond, CheckKind), Handler, StaticData, None);
+
+  if (requiresReturnValueNullabilityCheck())
+    EmitBlock(NoCheck);
+}
+
 static bool isInAllocaArgument(CGCXXABI &ABI, QualType type) {
   const CXXRecordDecl *RD = type->getAsCXXRecordDecl();
   return RD && ABI.getRecordArgABI(RD) == CGCXXABI::RAA_DirectInMemory;
@@ -3188,50 +3284,63 @@ void CallArgList::freeArgumentMemory(CodeGenFunction &CGF) const {
 
 void CodeGenFunction::EmitNonNullArgCheck(RValue RV, QualType ArgType,
                                           SourceLocation ArgLoc,
-                                          const FunctionDecl *FD,
+                                          AbstractCallee AC,
                                           unsigned ParmNum) {
-  if (!SanOpts.has(SanitizerKind::NonnullAttribute) || !FD)
+  if (!AC.getDecl() || !(SanOpts.has(SanitizerKind::NonnullAttribute) ||
+                         SanOpts.has(SanitizerKind::NullabilityArg)))
     return;
-  auto PVD = ParmNum < FD->getNumParams() ? FD->getParamDecl(ParmNum) : nullptr;
+
+  // The param decl may be missing in a variadic function.
+  auto PVD = ParmNum < AC.getNumParams() ? AC.getParamDecl(ParmNum) : nullptr;
   unsigned ArgNo = PVD ? PVD->getFunctionScopeIndex() : ParmNum;
-  auto NNAttr = getNonNullAttr(FD, PVD, ArgType, ArgNo);
-  if (!NNAttr)
+
+  // Prefer the nonnull attribute if it's present. 
+  const NonNullAttr *NNAttr = nullptr;
+  if (SanOpts.has(SanitizerKind::NonnullAttribute))
+    NNAttr = getNonNullAttr(AC.getDecl(), PVD, ArgType, ArgNo);
+
+  bool CanCheckNullability = false;
+  if (SanOpts.has(SanitizerKind::NullabilityArg) && !NNAttr && PVD) {
+    auto Nullability = PVD->getType()->getNullability(getContext());
+    CanCheckNullability = Nullability &&
+                          *Nullability == NullabilityKind::NonNull &&
+                          PVD->getTypeSourceInfo();
+  }
+
+  if (!NNAttr && !CanCheckNullability)
     return;
+
+  SourceLocation AttrLoc;
+  SanitizerMask CheckKind;
+  SanitizerHandler Handler;
+  if (NNAttr) {
+    AttrLoc = NNAttr->getLocation();
+    CheckKind = SanitizerKind::NonnullAttribute;
+    Handler = SanitizerHandler::NonnullArg;
+  } else {
+    AttrLoc = PVD->getTypeSourceInfo()->getTypeLoc().findNullabilityLoc();
+    CheckKind = SanitizerKind::NullabilityArg;
+    Handler = SanitizerHandler::NullabilityArg;
+  }
+
   SanitizerScope SanScope(this);
   assert(RV.isScalar());
   llvm::Value *V = RV.getScalarVal();
   llvm::Value *Cond =
       Builder.CreateICmpNE(V, llvm::Constant::getNullValue(V->getType()));
   llvm::Constant *StaticData[] = {
-      EmitCheckSourceLocation(ArgLoc),
-      EmitCheckSourceLocation(NNAttr->getLocation()),
+      EmitCheckSourceLocation(ArgLoc), EmitCheckSourceLocation(AttrLoc),
       llvm::ConstantInt::get(Int32Ty, ArgNo + 1),
   };
-  EmitCheck(std::make_pair(Cond, SanitizerKind::NonnullAttribute),
-            SanitizerHandler::NonnullArg, StaticData, None);
+  EmitCheck(std::make_pair(Cond, CheckKind), Handler, StaticData, None);
 }
 
 void CodeGenFunction::EmitCallArgs(
     CallArgList &Args, ArrayRef<QualType> ArgTypes,
     llvm::iterator_range<CallExpr::const_arg_iterator> ArgRange,
-    const FunctionDecl *CalleeDecl, unsigned ParamsToSkip,
-    EvaluationOrder Order) {
+    AbstractCallee AC, unsigned ParamsToSkip, EvaluationOrder Order) {
   assert((int)ArgTypes.size() == (ArgRange.end() - ArgRange.begin()));
 
-  auto MaybeEmitImplicitObjectSize = [&](unsigned I, const Expr *Arg) {
-    if (CalleeDecl == nullptr || I >= CalleeDecl->getNumParams())
-      return;
-    auto *PS = CalleeDecl->getParamDecl(I)->getAttr<PassObjectSizeAttr>();
-    if (PS == nullptr)
-      return;
-
-    const auto &Context = getContext();
-    auto SizeTy = Context.getSizeType();
-    auto T = Builder.getIntNTy(Context.getTypeSize(SizeTy));
-    llvm::Value *V = evaluateOrEmitBuiltinObjectSize(Arg, PS->getType(), T);
-    Args.add(RValue::get(V), SizeTy);
-  };
-
   // We *have* to evaluate arguments from right to left in the MS C++ ABI,
   // because arguments are destroyed left to right in the callee. As a special
   // case, there are certain language constructs that require left-to-right
@@ -3242,6 +3351,27 @@ void CodeGenFunction::EmitCallArgs(
           ? Order == EvaluationOrder::ForceLeftToRight
           : Order != EvaluationOrder::ForceRightToLeft;
 
+  auto MaybeEmitImplicitObjectSize = [&](unsigned I, const Expr *Arg,
+                                         RValue EmittedArg) {
+    if (!AC.hasFunctionDecl() || I >= AC.getNumParams())
+      return;
+    auto *PS = AC.getParamDecl(I)->getAttr<PassObjectSizeAttr>();
+    if (PS == nullptr)
+      return;
+
+    const auto &Context = getContext();
+    auto SizeTy = Context.getSizeType();
+    auto T = Builder.getIntNTy(Context.getTypeSize(SizeTy));
+    assert(EmittedArg.getScalarVal() && "We emitted nothing for the arg?");
+    llvm::Value *V = evaluateOrEmitBuiltinObjectSize(Arg, PS->getType(), T,
+                                                     EmittedArg.getScalarVal());
+    Args.add(RValue::get(V), SizeTy);
+    // If we're emitting args in reverse, be sure to do so with
+    // pass_object_size, as well.
+    if (!LeftToRight)
+      std::swap(Args.back(), *(&Args.back() - 1));
+  };
+
   // Insert a stack save if we're going to need any inalloca args.
   bool HasInAllocaArgs = false;
   if (CGM.getTarget().getCXXABI().isMicrosoft()) {
@@ -3259,11 +3389,20 @@ void CodeGenFunction::EmitCallArgs(
   for (unsigned I = 0, E = ArgTypes.size(); I != E; ++I) {
     unsigned Idx = LeftToRight ? I : E - I - 1;
     CallExpr::const_arg_iterator Arg = ArgRange.begin() + Idx;
-    if (!LeftToRight) MaybeEmitImplicitObjectSize(Idx, *Arg);
+    unsigned InitialArgSize = Args.size();
     EmitCallArg(Args, *Arg, ArgTypes[Idx]);
-    EmitNonNullArgCheck(Args.back().RV, ArgTypes[Idx], (*Arg)->getExprLoc(),
-                        CalleeDecl, ParamsToSkip + Idx);
-    if (LeftToRight) MaybeEmitImplicitObjectSize(Idx, *Arg);
+    // In particular, we depend on it being the last arg in Args, and the
+    // objectsize bits depend on there only being one arg if !LeftToRight.
+    assert(InitialArgSize + 1 == Args.size() &&
+           "The code below depends on only adding one arg per EmitCallArg");
+    (void)InitialArgSize;
+    RValue RVArg = Args.back().RV;
+    EmitNonNullArgCheck(RVArg, ArgTypes[Idx], (*Arg)->getExprLoc(), AC,
+                        ParamsToSkip + Idx);
+    // @llvm.objectsize should never have side-effects and shouldn't need
+    // destruction/cleanups, so we can safely "emit" it after its arg,
+    // regardless of right-to-leftness
+    MaybeEmitImplicitObjectSize(Idx, *Arg, RVArg);
   }
 
   if (!LeftToRight) {
@@ -3571,12 +3710,14 @@ RValue CodeGenFunction::EmitCall(const CGFunctionInfo &CallInfo,
   Address ArgMemory = Address::invalid();
   const llvm::StructLayout *ArgMemoryLayout = nullptr;
   if (llvm::StructType *ArgStruct = CallInfo.getArgStruct()) {
-    ArgMemoryLayout = CGM.getDataLayout().getStructLayout(ArgStruct);
+    const llvm::DataLayout &DL = CGM.getDataLayout();
+    ArgMemoryLayout = DL.getStructLayout(ArgStruct);
     llvm::Instruction *IP = CallArgs.getStackBase();
     llvm::AllocaInst *AI;
     if (IP) {
       IP = IP->getNextNode();
-      AI = new llvm::AllocaInst(ArgStruct, "argmem", IP);
+      AI = new llvm::AllocaInst(ArgStruct, DL.getAllocaAddrSpace(),
+                                "argmem", IP);
     } else {
       AI = CreateTempAlloca(ArgStruct, "argmem");
     }
@@ -3977,8 +4118,8 @@ RValue CodeGenFunction::EmitCall(const CGFunctionInfo &CallInfo,
                              Callee.getAbstractInfo(),
                              AttributeList, CallingConv,
                              /*AttrOnCallSite=*/true);
-  llvm::AttributeSet Attrs = llvm::AttributeSet::get(getLLVMContext(),
-                                                     AttributeList);
+  llvm::AttributeList Attrs =
+      llvm::AttributeList::get(getLLVMContext(), AttributeList);
 
   // Apply some call-site-specific attributes.
   // TODO: work this into building the attribute set.
@@ -3989,15 +4130,14 @@ RValue CodeGenFunction::EmitCall(const CGFunctionInfo &CallInfo,
       !(Callee.getAbstractInfo().getCalleeDecl() &&
         Callee.getAbstractInfo().getCalleeDecl()->hasAttr<NoInlineAttr>())) {
     Attrs =
-        Attrs.addAttribute(getLLVMContext(),
-                           llvm::AttributeSet::FunctionIndex,
+        Attrs.addAttribute(getLLVMContext(), llvm::AttributeList::FunctionIndex,
                            llvm::Attribute::AlwaysInline);
   }
 
   // Disable inlining inside SEH __try blocks.
   if (isSEHTryScope()) {
     Attrs =
-        Attrs.addAttribute(getLLVMContext(), llvm::AttributeSet::FunctionIndex,
+        Attrs.addAttribute(getLLVMContext(), llvm::AttributeList::FunctionIndex,
                            llvm::Attribute::NoInline);
   }
 
@@ -4014,7 +4154,7 @@ RValue CodeGenFunction::EmitCall(const CGFunctionInfo &CallInfo,
     CannotThrow = true;
   } else {
     // Otherwise, nounwind call sites will never throw.
-    CannotThrow = Attrs.hasAttribute(llvm::AttributeSet::FunctionIndex,
+    CannotThrow = Attrs.hasAttribute(llvm::AttributeList::FunctionIndex,
                                      llvm::Attribute::NoUnwind);
   }
   llvm::BasicBlock *InvokeDest = CannotThrow ? nullptr : getInvokeDest();
@@ -4210,6 +4350,10 @@ RValue CodeGenFunction::EmitCall(const CGFunctionInfo &CallInfo,
       llvm::ConstantInt *AlignmentCI = cast<llvm::ConstantInt>(Alignment);
       EmitAlignmentAssumption(Ret.getScalarVal(), AlignmentCI->getZExtValue(),
                               OffsetValue);
+    } else if (const auto *AA = TargetDecl->getAttr<AllocAlignAttr>()) {
+      llvm::Value *ParamVal =
+          CallArgs[AA->getParamIndex() - 1].RV.getScalarVal();
+      EmitAlignmentAssumption(Ret.getScalarVal(), ParamVal);
     }
   }
 
diff --git a/lib/CodeGen/CGCall.h b/lib/CodeGen/CGCall.h
index 031ce831cb37..97221e20c195 100644
--- a/lib/CodeGen/CGCall.h
+++ b/lib/CodeGen/CGCall.h
@@ -25,10 +25,10 @@
 #include "ABIInfo.h"
 
 namespace llvm {
-  class AttributeSet;
-  class Function;
-  class Type;
-  class Value;
+class AttributeList;
+class Function;
+class Type;
+class Value;
 }
 
 namespace clang {
@@ -39,28 +39,28 @@ namespace clang {
   class VarDecl;
 
 namespace CodeGen {
-  typedef SmallVector<llvm::AttributeSet, 8> AttributeListType;
-
-  /// Abstract information about a function or function prototype.
-  class CGCalleeInfo {
-    /// \brief The function prototype of the callee.
-    const FunctionProtoType *CalleeProtoTy;
-    /// \brief The function declaration of the callee.
-    const Decl *CalleeDecl;
-
-  public:
-    explicit CGCalleeInfo() : CalleeProtoTy(nullptr), CalleeDecl(nullptr) {}
-    CGCalleeInfo(const FunctionProtoType *calleeProtoTy, const Decl *calleeDecl)
-        : CalleeProtoTy(calleeProtoTy), CalleeDecl(calleeDecl) {}
-    CGCalleeInfo(const FunctionProtoType *calleeProtoTy)
-        : CalleeProtoTy(calleeProtoTy), CalleeDecl(nullptr) {}
-    CGCalleeInfo(const Decl *calleeDecl)
-        : CalleeProtoTy(nullptr), CalleeDecl(calleeDecl) {}
-
-    const FunctionProtoType *getCalleeFunctionProtoType() const {
-      return CalleeProtoTy;
-    }
-    const Decl *getCalleeDecl() const { return CalleeDecl; }
+typedef SmallVector<llvm::AttributeList, 8> AttributeListType;
+
+/// Abstract information about a function or function prototype.
+class CGCalleeInfo {
+  /// \brief The function prototype of the callee.
+  const FunctionProtoType *CalleeProtoTy;
+  /// \brief The function declaration of the callee.
+  const Decl *CalleeDecl;
+
+public:
+  explicit CGCalleeInfo() : CalleeProtoTy(nullptr), CalleeDecl(nullptr) {}
+  CGCalleeInfo(const FunctionProtoType *calleeProtoTy, const Decl *calleeDecl)
+      : CalleeProtoTy(calleeProtoTy), CalleeDecl(calleeDecl) {}
+  CGCalleeInfo(const FunctionProtoType *calleeProtoTy)
+      : CalleeProtoTy(calleeProtoTy), CalleeDecl(nullptr) {}
+  CGCalleeInfo(const Decl *calleeDecl)
+      : CalleeProtoTy(nullptr), CalleeDecl(calleeDecl) {}
+
+  const FunctionProtoType *getCalleeFunctionProtoType() const {
+    return CalleeProtoTy;
+  }
+  const Decl *getCalleeDecl() const { return CalleeDecl; }
   };
 
   /// All available information about a concrete callee.
diff --git a/lib/CodeGen/CGClass.cpp b/lib/CodeGen/CGClass.cpp
index 05d056739524..7ba03a0d42dd 100644
--- a/lib/CodeGen/CGClass.cpp
+++ b/lib/CodeGen/CGClass.cpp
@@ -309,8 +309,10 @@ Address CodeGenFunction::GetAddressOfBaseClass(
   // just do a bitcast; null checks are unnecessary.
   if (NonVirtualOffset.isZero() && !VBase) {
     if (sanitizePerformTypeCheck()) {
+      SanitizerSet SkippedChecks;
+      SkippedChecks.set(SanitizerKind::Null, !NullCheckValue);
       EmitTypeCheck(TCK_Upcast, Loc, Value.getPointer(),
-                    DerivedTy, DerivedAlign, !NullCheckValue);
+                    DerivedTy, DerivedAlign, SkippedChecks);
     }
     return Builder.CreateBitCast(Value, BasePtrTy);
   }
@@ -331,8 +333,10 @@ Address CodeGenFunction::GetAddressOfBaseClass(
   }
 
   if (sanitizePerformTypeCheck()) {
+    SanitizerSet SkippedChecks;
+    SkippedChecks.set(SanitizerKind::Null, true);
     EmitTypeCheck(VBase ? TCK_UpcastToVirtualBase : TCK_Upcast, Loc,
-                  Value.getPointer(), DerivedTy, DerivedAlign, true);
+                  Value.getPointer(), DerivedTy, DerivedAlign, SkippedChecks);
   }
 
   // Compute the virtual offset.
@@ -685,7 +689,8 @@ void CodeGenFunction::EmitInitializerForField(FieldDecl *Field, LValue LHS,
 /// complete-to-base constructor delegation optimization, i.e.
 /// emitting the complete constructor as a simple call to the base
 /// constructor.
-static bool IsConstructorDelegationValid(const CXXConstructorDecl *Ctor) {
+bool CodeGenFunction::IsConstructorDelegationValid(
+    const CXXConstructorDecl *Ctor) {
 
   // Currently we disable the optimization for classes with virtual
   // bases because (1) the addresses of parameter variables need to be
@@ -1131,10 +1136,11 @@ namespace {
           RHS = EC->getSubExpr();
         if (!RHS)
           return nullptr;
-        MemberExpr *ME2 = dyn_cast<MemberExpr>(RHS);
-        if (dyn_cast<FieldDecl>(ME2->getMemberDecl()) != Field)
-          return nullptr;
-        return Field;
+        if (MemberExpr *ME2 = dyn_cast<MemberExpr>(RHS)) {
+          if (ME2->getMemberDecl() == Field)
+            return Field;
+        }
+        return nullptr;
       } else if (CXXMemberCallExpr *MCE = dyn_cast<CXXMemberCallExpr>(S)) {
         CXXMethodDecl *MD = dyn_cast<CXXMethodDecl>(MCE->getCalleeDecl());
         if (!(MD && isMemcpyEquivalentSpecialMember(MD)))
@@ -1384,6 +1390,20 @@ void CodeGenFunction::EmitDestructorBody(FunctionArgList &Args) {
   const CXXDestructorDecl *Dtor = cast<CXXDestructorDecl>(CurGD.getDecl());
   CXXDtorType DtorType = CurGD.getDtorType();
 
+  // For an abstract class, non-base destructors are never used (and can't
+  // be emitted in general, because vbase dtors may not have been validated
+  // by Sema), but the Itanium ABI doesn't make them optional and Clang may
+  // in fact emit references to them from other compilations, so emit them
+  // as functions containing a trap instruction.
+  if (DtorType != Dtor_Base && Dtor->getParent()->isAbstract()) {
+    llvm::CallInst *TrapCall = EmitTrapCall(llvm::Intrinsic::trap);
+    TrapCall->setDoesNotReturn();
+    TrapCall->setDoesNotThrow();
+    Builder.CreateUnreachable();
+    Builder.ClearInsertionPoint();
+    return;
+  }
+
   Stmt *Body = Dtor->getBody();
   if (Body)
     incrementProfileCounter(Body);
@@ -1416,9 +1436,7 @@ void CodeGenFunction::EmitDestructorBody(FunctionArgList &Args) {
   // we'd introduce *two* handler blocks.  In the Microsoft ABI, we
   // always delegate because we might not have a definition in this TU.
   switch (DtorType) {
-  case Dtor_Comdat:
-    llvm_unreachable("not expecting a COMDAT");
-
+  case Dtor_Comdat: llvm_unreachable("not expecting a COMDAT");
   case Dtor_Deleting: llvm_unreachable("already handled deleting case");
 
   case Dtor_Complete:
@@ -1433,7 +1451,9 @@ void CodeGenFunction::EmitDestructorBody(FunctionArgList &Args) {
                             /*Delegating=*/false, LoadCXXThisAddress());
       break;
     }
+
     // Fallthrough: act like we're in the base variant.
+    LLVM_FALLTHROUGH;
 
   case Dtor_Base:
     assert(Body);
@@ -1950,7 +1970,11 @@ void CodeGenFunction::EmitCXXConstructorCall(const CXXConstructorDecl *D,
 
   // Add the rest of the user-supplied arguments.
   const FunctionProtoType *FPT = D->getType()->castAs<FunctionProtoType>();
-  EmitCallArgs(Args, FPT, E->arguments(), E->getConstructor());
+  EvaluationOrder Order = E->isListInitialization()
+                              ? EvaluationOrder::ForceLeftToRight
+                              : EvaluationOrder::Default;
+  EmitCallArgs(Args, FPT, E->arguments(), E->getConstructor(),
+               /*ParamsToSkip*/ 0, Order);
 
   EmitCXXConstructorCall(D, Type, ForVirtualBase, Delegating, This, Args);
 }
@@ -1970,7 +1994,7 @@ static bool canEmitDelegateCallArgs(CodeGenFunction &CGF,
 
     // Likewise if they're inalloca.
     const CGFunctionInfo &Info =
-        CGF.CGM.getTypes().arrangeCXXConstructorCall(Args, Ctor, Type, 0);
+        CGF.CGM.getTypes().arrangeCXXConstructorCall(Args, Ctor, Type, 0, 0);
     if (Info.usesInAlloca())
       return false;
   }
@@ -2012,10 +2036,11 @@ void CodeGenFunction::EmitCXXConstructorCall(const CXXConstructorDecl *D,
     return;
   }
 
+  bool PassPrototypeArgs = true;
   // Check whether we can actually emit the constructor before trying to do so.
   if (auto Inherited = D->getInheritedConstructor()) {
-    if (getTypes().inheritingCtorHasParams(Inherited, Type) &&
-        !canEmitDelegateCallArgs(*this, D, Type, Args)) {
+    PassPrototypeArgs = getTypes().inheritingCtorHasParams(Inherited, Type);
+    if (PassPrototypeArgs && !canEmitDelegateCallArgs(*this, D, Type, Args)) {
       EmitInlinedInheritingCXXConstructorCall(D, Type, ForVirtualBase,
                                               Delegating, Args);
       return;
@@ -2023,14 +2048,15 @@ void CodeGenFunction::EmitCXXConstructorCall(const CXXConstructorDecl *D,
   }
 
   // Insert any ABI-specific implicit constructor arguments.
-  unsigned ExtraArgs = CGM.getCXXABI().addImplicitConstructorArgs(
-      *this, D, Type, ForVirtualBase, Delegating, Args);
+  CGCXXABI::AddedStructorArgs ExtraArgs =
+      CGM.getCXXABI().addImplicitConstructorArgs(*this, D, Type, ForVirtualBase,
+                                                 Delegating, Args);
 
   // Emit the call.
   llvm::Constant *CalleePtr =
     CGM.getAddrOfCXXStructor(D, getFromCtorType(Type));
-  const CGFunctionInfo &Info =
-    CGM.getTypes().arrangeCXXConstructorCall(Args, D, Type, ExtraArgs);
+  const CGFunctionInfo &Info = CGM.getTypes().arrangeCXXConstructorCall(
+      Args, D, Type, ExtraArgs.Prefix, ExtraArgs.Suffix, PassPrototypeArgs);
   CGCallee Callee = CGCallee::forDirect(CalleePtr, D);
   EmitCall(Info, Callee, ReturnValueSlot(), Args);
 
@@ -2102,7 +2128,9 @@ void CodeGenFunction::EmitInheritedCXXConstructorCall(
 void CodeGenFunction::EmitInlinedInheritingCXXConstructorCall(
     const CXXConstructorDecl *Ctor, CXXCtorType CtorType, bool ForVirtualBase,
     bool Delegating, CallArgList &Args) {
-  InlinedInheritingConstructorScope Scope(*this, GlobalDecl(Ctor, CtorType));
+  GlobalDecl GD(Ctor, CtorType);
+  InlinedInheritingConstructorScope Scope(*this, GD);
+  ApplyInlineDebugLocation DebugScope(*this, GD);
 
   // Save the arguments to be passed to the inherited constructor.
   CXXInheritedCtorInitExprArgs = Args;
diff --git a/lib/CodeGen/CGCleanup.cpp b/lib/CodeGen/CGCleanup.cpp
index 3666858e63d2..437ab7dd4649 100644
--- a/lib/CodeGen/CGCleanup.cpp
+++ b/lib/CodeGen/CGCleanup.cpp
@@ -418,11 +418,15 @@ void CodeGenFunction::ResolveBranchFixups(llvm::BasicBlock *Block) {
 }
 
 /// Pops cleanup blocks until the given savepoint is reached.
-void CodeGenFunction::PopCleanupBlocks(EHScopeStack::stable_iterator Old) {
+void CodeGenFunction::PopCleanupBlocks(
+    EHScopeStack::stable_iterator Old,
+    std::initializer_list<llvm::Value **> ValuesToReload) {
   assert(Old.isValid());
 
+  bool HadBranches = false;
   while (EHStack.stable_begin() != Old) {
     EHCleanupScope &Scope = cast<EHCleanupScope>(*EHStack.begin());
+    HadBranches |= Scope.hasBranches();
 
     // As long as Old strictly encloses the scope's enclosing normal
     // cleanup, we're going to emit another normal cleanup which
@@ -432,14 +436,41 @@ void CodeGenFunction::PopCleanupBlocks(EHScopeStack::stable_iterator Old) {
 
     PopCleanupBlock(FallThroughIsBranchThrough);
   }
+
+  // If we didn't have any branches, the insertion point before cleanups must
+  // dominate the current insertion point and we don't need to reload any
+  // values.
+  if (!HadBranches)
+    return;
+
+  // Spill and reload all values that the caller wants to be live at the current
+  // insertion point.
+  for (llvm::Value **ReloadedValue : ValuesToReload) {
+    auto *Inst = dyn_cast_or_null<llvm::Instruction>(*ReloadedValue);
+    if (!Inst)
+      continue;
+    Address Tmp =
+        CreateDefaultAlignTempAlloca(Inst->getType(), "tmp.exprcleanup");
+
+    // Find an insertion point after Inst and spill it to the temporary.
+    llvm::BasicBlock::iterator InsertBefore;
+    if (auto *Invoke = dyn_cast<llvm::InvokeInst>(Inst))
+      InsertBefore = Invoke->getNormalDest()->getFirstInsertionPt();
+    else
+      InsertBefore = std::next(Inst->getIterator());
+    CGBuilderTy(CGM, &*InsertBefore).CreateStore(Inst, Tmp);
+
+    // Reload the value at the current insertion point.
+    *ReloadedValue = Builder.CreateLoad(Tmp);
+  }
 }
 
 /// Pops cleanup blocks until the given savepoint is reached, then add the
 /// cleanups from the given savepoint in the lifetime-extended cleanups stack.
-void
-CodeGenFunction::PopCleanupBlocks(EHScopeStack::stable_iterator Old,
-                                  size_t OldLifetimeExtendedSize) {
-  PopCleanupBlocks(Old);
+void CodeGenFunction::PopCleanupBlocks(
+    EHScopeStack::stable_iterator Old, size_t OldLifetimeExtendedSize,
+    std::initializer_list<llvm::Value **> ValuesToReload) {
+  PopCleanupBlocks(Old, ValuesToReload);
 
   // Move our deferred cleanups onto the EH stack.
   for (size_t I = OldLifetimeExtendedSize,
@@ -578,7 +609,7 @@ static void destroyOptimisticNormalEntry(CodeGenFunction &CGF,
     llvm::SwitchInst *si = cast<llvm::SwitchInst>(use.getUser());
     if (si->getNumCases() == 1 && si->getDefaultDest() == unreachableBB) {
       // Replace the switch with a branch.
-      llvm::BranchInst::Create(si->case_begin().getCaseSuccessor(), si);
+      llvm::BranchInst::Create(si->case_begin()->getCaseSuccessor(), si);
 
       // The switch operand is a load from the cleanup-dest alloca.
       llvm::LoadInst *condition = cast<llvm::LoadInst>(si->getCondition());
diff --git a/lib/CodeGen/CGCoroutine.cpp b/lib/CodeGen/CGCoroutine.cpp
index 2fdb1279ece9..0ef680ef6609 100644
--- a/lib/CodeGen/CGCoroutine.cpp
+++ b/lib/CodeGen/CGCoroutine.cpp
@@ -12,28 +12,58 @@
 //===----------------------------------------------------------------------===//
 
 #include "CodeGenFunction.h"
+#include "llvm/ADT/ScopeExit.h"
 #include "clang/AST/StmtCXX.h"
 
 using namespace clang;
 using namespace CodeGen;
 
-namespace clang {
-namespace CodeGen {
+using llvm::Value;
+using llvm::BasicBlock;
+
+namespace {
+enum class AwaitKind { Init, Normal, Yield, Final };
+static constexpr llvm::StringLiteral AwaitKindStr[] = {"init", "await", "yield",
+                                                       "final"};
+}
+
+struct clang::CodeGen::CGCoroData {
+  // What is the current await expression kind and how many
+  // await/yield expressions were encountered so far.
+  // These are used to generate pretty labels for await expressions in LLVM IR.
+  AwaitKind CurrentAwaitKind = AwaitKind::Init;
+  unsigned AwaitNum = 0;
+  unsigned YieldNum = 0;
+
+  // How many co_return statements are in the coroutine. Used to decide whether
+  // we need to add co_return; equivalent at the end of the user authored body.
+  unsigned CoreturnCount = 0;
+
+  // A branch to this block is emitted when coroutine needs to suspend.
+  llvm::BasicBlock *SuspendBB = nullptr;
+
+  // Stores the jump destination just before the coroutine memory is freed.
+  // This is the destination that every suspend point jumps to for the cleanup
+  // branch.
+  CodeGenFunction::JumpDest CleanupJD;
+
+  // Stores the jump destination just before the final suspend. The co_return
+  // statements jumps to this point after calling return_xxx promise member.
+  CodeGenFunction::JumpDest FinalJD;
 
-struct CGCoroData {
   // Stores the llvm.coro.id emitted in the function so that we can supply it
   // as the first argument to coro.begin, coro.alloc and coro.free intrinsics.
   // Note: llvm.coro.id returns a token that cannot be directly expressed in a
   // builtin.
   llvm::CallInst *CoroId = nullptr;
+
   // If coro.id came from the builtin, remember the expression to give better
   // diagnostic. If CoroIdExpr is nullptr, the coro.id was created by
   // EmitCoroutineBody.
   CallExpr const *CoroIdExpr = nullptr;
 };
-}
-}
 
+// Defining these here allows to keep CGCoroData private to this file.
 clang::CodeGen::CodeGenFunction::CGCoroInfo::CGCoroInfo() {}
 CodeGenFunction::CGCoroInfo::~CGCoroInfo() {}
 
@@ -59,19 +89,250 @@ static void createCoroData(CodeGenFunction &CGF,
   CurCoro.Data->CoroIdExpr = CoroIdExpr;
 }
 
+// Synthesize a pretty name for a suspend point.
+static SmallString<32> buildSuspendPrefixStr(CGCoroData &Coro, AwaitKind Kind) {
+  unsigned No = 0;
+  switch (Kind) {
+  case AwaitKind::Init:
+  case AwaitKind::Final:
+    break;
+  case AwaitKind::Normal:
+    No = ++Coro.AwaitNum;
+    break;
+  case AwaitKind::Yield:
+    No = ++Coro.YieldNum;
+    break;
+  }
+  SmallString<32> Prefix(AwaitKindStr[static_cast<unsigned>(Kind)]);
+  if (No > 1) {
+    Twine(No).toVector(Prefix);
+  }
+  return Prefix;
+}
+
+// Emit suspend expression which roughly looks like:
+//
+//   auto && x = CommonExpr();
+//   if (!x.await_ready()) {
+//      llvm_coro_save();
+//      x.await_suspend(...);     (*)
+//      llvm_coro_suspend(); (**)
+//   }
+//   x.await_resume();
+//
+// where the result of the entire expression is the result of x.await_resume()
+//
+//   (*) If x.await_suspend return type is bool, it allows to veto a suspend:
+//      if (x.await_suspend(...))
+//        llvm_coro_suspend();
+//
+//  (**) llvm_coro_suspend() encodes three possible continuations as
+//       a switch instruction:
+//
+//  %where-to = call i8 @llvm.coro.suspend(...)
+//  switch i8 %where-to, label %coro.ret [ ; jump to epilogue to suspend
+//    i8 0, label %yield.ready   ; go here when resumed
+//    i8 1, label %yield.cleanup ; go here when destroyed
+//  ]
+//
+//  See llvm's docs/Coroutines.rst for more details.
+//
+static RValue emitSuspendExpression(CodeGenFunction &CGF, CGCoroData &Coro,
+                                    CoroutineSuspendExpr const &S,
+                                    AwaitKind Kind, AggValueSlot aggSlot,
+                                    bool ignoreResult) {
+  auto *E = S.getCommonExpr();
+  auto Binder =
+      CodeGenFunction::OpaqueValueMappingData::bind(CGF, S.getOpaqueValue(), E);
+  auto UnbindOnExit = llvm::make_scope_exit([&] { Binder.unbind(CGF); });
+
+  auto Prefix = buildSuspendPrefixStr(Coro, Kind);
+  BasicBlock *ReadyBlock = CGF.createBasicBlock(Prefix + Twine(".ready"));
+  BasicBlock *SuspendBlock = CGF.createBasicBlock(Prefix + Twine(".suspend"));
+  BasicBlock *CleanupBlock = CGF.createBasicBlock(Prefix + Twine(".cleanup"));
+
+  // If expression is ready, no need to suspend.
+  CGF.EmitBranchOnBoolExpr(S.getReadyExpr(), ReadyBlock, SuspendBlock, 0);
+
+  // Otherwise, emit suspend logic.
+  CGF.EmitBlock(SuspendBlock);
+
+  auto &Builder = CGF.Builder;
+  llvm::Function *CoroSave = CGF.CGM.getIntrinsic(llvm::Intrinsic::coro_save);
+  auto *NullPtr = llvm::ConstantPointerNull::get(CGF.CGM.Int8PtrTy);
+  auto *SaveCall = Builder.CreateCall(CoroSave, {NullPtr});
+
+  auto *SuspendRet = CGF.EmitScalarExpr(S.getSuspendExpr());
+  if (SuspendRet != nullptr) {
+    // Veto suspension if requested by bool returning await_suspend.
+    assert(SuspendRet->getType()->isIntegerTy(1) &&
+           "Sema should have already checked that it is void or bool");
+    BasicBlock *RealSuspendBlock =
+        CGF.createBasicBlock(Prefix + Twine(".suspend.bool"));
+    CGF.Builder.CreateCondBr(SuspendRet, RealSuspendBlock, ReadyBlock);
+    SuspendBlock = RealSuspendBlock;
+    CGF.EmitBlock(RealSuspendBlock);
+  }
+
+  // Emit the suspend point.
+  const bool IsFinalSuspend = (Kind == AwaitKind::Final);
+  llvm::Function *CoroSuspend =
+      CGF.CGM.getIntrinsic(llvm::Intrinsic::coro_suspend);
+  auto *SuspendResult = Builder.CreateCall(
+      CoroSuspend, {SaveCall, Builder.getInt1(IsFinalSuspend)});
+
+  // Create a switch capturing three possible continuations.
+  auto *Switch = Builder.CreateSwitch(SuspendResult, Coro.SuspendBB, 2);
+  Switch->addCase(Builder.getInt8(0), ReadyBlock);
+  Switch->addCase(Builder.getInt8(1), CleanupBlock);
+
+  // Emit cleanup for this suspend point.
+  CGF.EmitBlock(CleanupBlock);
+  CGF.EmitBranchThroughCleanup(Coro.CleanupJD);
+
+  // Emit await_resume expression.
+  CGF.EmitBlock(ReadyBlock);
+  return CGF.EmitAnyExpr(S.getResumeExpr(), aggSlot, ignoreResult);
+}
+
+RValue CodeGenFunction::EmitCoawaitExpr(const CoawaitExpr &E,
+                                        AggValueSlot aggSlot,
+                                        bool ignoreResult) {
+  return emitSuspendExpression(*this, *CurCoro.Data, E,
+                               CurCoro.Data->CurrentAwaitKind, aggSlot,
+                               ignoreResult);
+}
+RValue CodeGenFunction::EmitCoyieldExpr(const CoyieldExpr &E,
+                                        AggValueSlot aggSlot,
+                                        bool ignoreResult) {
+  return emitSuspendExpression(*this, *CurCoro.Data, E, AwaitKind::Yield,
+                               aggSlot, ignoreResult);
+}
+
+void CodeGenFunction::EmitCoreturnStmt(CoreturnStmt const &S) {
+  ++CurCoro.Data->CoreturnCount;
+  EmitStmt(S.getPromiseCall());
+  EmitBranchThroughCleanup(CurCoro.Data->FinalJD);
+}
+
+// For WinEH exception representation backend need to know what funclet coro.end
+// belongs to. That information is passed in a funclet bundle.
+static SmallVector<llvm::OperandBundleDef, 1>
+getBundlesForCoroEnd(CodeGenFunction &CGF) {
+  SmallVector<llvm::OperandBundleDef, 1> BundleList;
+
+  if (llvm::Instruction *EHPad = CGF.CurrentFuncletPad)
+    BundleList.emplace_back("funclet", EHPad);
+
+  return BundleList;
+}
+
+namespace {
+// We will insert coro.end to cut any of the destructors for objects that
+// do not need to be destroyed once the coroutine is resumed.
+// See llvm/docs/Coroutines.rst for more details about coro.end.
+struct CallCoroEnd final : public EHScopeStack::Cleanup {
+  void Emit(CodeGenFunction &CGF, Flags flags) override {
+    auto &CGM = CGF.CGM;
+    auto *NullPtr = llvm::ConstantPointerNull::get(CGF.Int8PtrTy);
+    llvm::Function *CoroEndFn = CGM.getIntrinsic(llvm::Intrinsic::coro_end);
+    // See if we have a funclet bundle to associate coro.end with. (WinEH)
+    auto Bundles = getBundlesForCoroEnd(CGF);
+    auto *CoroEnd = CGF.Builder.CreateCall(
+        CoroEndFn, {NullPtr, CGF.Builder.getTrue()}, Bundles);
+    if (Bundles.empty()) {
+      // Otherwise, (landingpad model), create a conditional branch that leads
+      // either to a cleanup block or a block with EH resume instruction.
+      auto *ResumeBB = CGF.getEHResumeBlock(/*cleanup=*/true);
+      auto *CleanupContBB = CGF.createBasicBlock("cleanup.cont");
+      CGF.Builder.CreateCondBr(CoroEnd, ResumeBB, CleanupContBB);
+      CGF.EmitBlock(CleanupContBB);
+    }
+  }
+};
+}
+
+namespace {
+// Make sure to call coro.delete on scope exit.
+struct CallCoroDelete final : public EHScopeStack::Cleanup {
+  Stmt *Deallocate;
+
+  // TODO: Wrap deallocate in if(coro.free(...)) Deallocate.
+  void Emit(CodeGenFunction &CGF, Flags) override {
+    // Note: That deallocation will be emitted twice: once for a normal exit and
+    // once for exceptional exit. This usage is safe because Deallocate does not
+    // contain any declarations. The SubStmtBuilder::makeNewAndDeleteExpr()
+    // builds a single call to a deallocation function which is safe to emit
+    // multiple times.
+    CGF.EmitStmt(Deallocate);
+  }
+  explicit CallCoroDelete(Stmt *DeallocStmt) : Deallocate(DeallocStmt) {}
+};
+}
+
 void CodeGenFunction::EmitCoroutineBody(const CoroutineBodyStmt &S) {
   auto *NullPtr = llvm::ConstantPointerNull::get(Builder.getInt8PtrTy());
   auto &TI = CGM.getContext().getTargetInfo();
   unsigned NewAlign = TI.getNewAlign() / TI.getCharWidth();
 
+  auto *FinalBB = createBasicBlock("coro.final");
+  auto *RetBB = createBasicBlock("coro.ret");
+
   auto *CoroId = Builder.CreateCall(
       CGM.getIntrinsic(llvm::Intrinsic::coro_id),
       {Builder.getInt32(NewAlign), NullPtr, NullPtr, NullPtr});
   createCoroData(*this, CurCoro, CoroId);
+  CurCoro.Data->SuspendBB = RetBB;
+
+  auto *AllocateCall = EmitScalarExpr(S.getAllocate());
+
+  // Handle allocation failure if 'ReturnStmtOnAllocFailure' was provided.
+  if (auto *RetOnAllocFailure = S.getReturnStmtOnAllocFailure()) {
+    auto *RetOnFailureBB = createBasicBlock("coro.ret.on.failure");
+    auto *InitBB = createBasicBlock("coro.init");
+
+    // See if allocation was successful.
+    auto *NullPtr = llvm::ConstantPointerNull::get(Int8PtrTy);
+    auto *Cond = Builder.CreateICmpNE(AllocateCall, NullPtr);
+    Builder.CreateCondBr(Cond, InitBB, RetOnFailureBB);
+
+    // If not, return OnAllocFailure object.
+    EmitBlock(RetOnFailureBB);
+    EmitStmt(RetOnAllocFailure);
+
+    EmitBlock(InitBB);
+  }
+
+  CurCoro.Data->CleanupJD = getJumpDestInCurrentScope(RetBB);
+  {
+    CodeGenFunction::RunCleanupsScope ResumeScope(*this);
+    EHStack.pushCleanup<CallCoroDelete>(NormalAndEHCleanup, S.getDeallocate());
+
+    EmitStmt(S.getPromiseDeclStmt());
+
+    EHStack.pushCleanup<CallCoroEnd>(EHCleanup);
+
+    CurCoro.Data->FinalJD = getJumpDestInCurrentScope(FinalBB);
+
+    // FIXME: Emit initial suspend and more before the body.
+
+    CurCoro.Data->CurrentAwaitKind = AwaitKind::Normal;
+    EmitStmt(S.getBody());
+
+    // See if we need to generate final suspend.
+    const bool CanFallthrough = Builder.GetInsertBlock();
+    const bool HasCoreturns = CurCoro.Data->CoreturnCount > 0;
+    if (CanFallthrough || HasCoreturns) {
+      EmitBlock(FinalBB);
+      // FIXME: Emit final suspend.
+    }
+  }
+
+  EmitBlock(RetBB);
+  llvm::Function *CoroEnd = CGM.getIntrinsic(llvm::Intrinsic::coro_end);
+  Builder.CreateCall(CoroEnd, {NullPtr, Builder.getFalse()});
 
-  EmitScalarExpr(S.getAllocate());
-  // FIXME: Emit the rest of the coroutine.
-  EmitStmt(S.getDeallocate());
+  // FIXME: Emit return for the coroutine return object.
 }
 
 // Emit coroutine intrinsic and patch up arguments of the token type.
diff --git a/lib/CodeGen/CGDebugInfo.cpp b/lib/CodeGen/CGDebugInfo.cpp
index 12a68036b09c..818b51543d30 100644
--- a/lib/CodeGen/CGDebugInfo.cpp
+++ b/lib/CodeGen/CGDebugInfo.cpp
@@ -107,8 +107,8 @@ void ApplyDebugLocation::init(SourceLocation TemporaryLocation,
 
   // Construct a location that has a valid scope, but no line info.
   assert(!DI->LexicalBlockStack.empty());
-  CGF->Builder.SetCurrentDebugLocation(
-      llvm::DebugLoc::get(0, 0, DI->LexicalBlockStack.back()));
+  CGF->Builder.SetCurrentDebugLocation(llvm::DebugLoc::get(
+      0, 0, DI->LexicalBlockStack.back(), DI->getInlinedAt()));
 }
 
 ApplyDebugLocation::ApplyDebugLocation(CodeGenFunction &CGF, const Expr *E)
@@ -134,6 +134,30 @@ ApplyDebugLocation::~ApplyDebugLocation() {
     CGF->Builder.SetCurrentDebugLocation(std::move(OriginalLocation));
 }
 
+ApplyInlineDebugLocation::ApplyInlineDebugLocation(CodeGenFunction &CGF,
+                                                   GlobalDecl InlinedFn)
+    : CGF(&CGF) {
+  if (!CGF.getDebugInfo()) {
+    this->CGF = nullptr;
+    return;
+  }
+  auto &DI = *CGF.getDebugInfo();
+  SavedLocation = DI.getLocation();
+  assert((DI.getInlinedAt() ==
+          CGF.Builder.getCurrentDebugLocation()->getInlinedAt()) &&
+         "CGDebugInfo and IRBuilder are out of sync");
+
+  DI.EmitInlineFunctionStart(CGF.Builder, InlinedFn);
+}
+
+ApplyInlineDebugLocation::~ApplyInlineDebugLocation() {
+  if (!CGF)
+    return;
+  auto &DI = *CGF->getDebugInfo();
+  DI.EmitInlineFunctionEnd(CGF->Builder);
+  DI.EmitLocation(CGF->Builder, SavedLocation);
+}
+
 void CGDebugInfo::setLocation(SourceLocation Loc) {
   // If the new location isn't valid return.
   if (Loc.isInvalid())
@@ -249,8 +273,8 @@ StringRef CGDebugInfo::getObjCMethodName(const ObjCMethodDecl *OMD) {
          << OC->getIdentifier()->getNameStart() << ')';
     }
   } else if (const auto *OCD = dyn_cast<ObjCCategoryImplDecl>(DC)) {
-    OS << ((const NamedDecl *)OCD)->getIdentifier()->getNameStart() << '('
-       << OCD->getIdentifier()->getNameStart() << ')';
+    OS << OCD->getClassInterface()->getName() << '('
+       << OCD->getName() << ')';
   } else if (isa<ObjCProtocolDecl>(DC)) {
     // We can extract the type of the class from the self pointer.
     if (ImplicitParamDecl *SelfDecl = OMD->getSelfDecl()) {
@@ -509,7 +533,8 @@ void CGDebugInfo::CreateCompileUnit() {
                                    Checksum),
       Producer, LO.Optimize, CGM.getCodeGenOpts().DwarfDebugFlags, RuntimeVers,
       CGM.getCodeGenOpts().SplitDwarfFile, EmissionKind, 0 /* DWOid */,
-      CGM.getCodeGenOpts().SplitDwarfInlining);
+      CGM.getCodeGenOpts().SplitDwarfInlining,
+      CGM.getCodeGenOpts().DebugInfoForProfiling);
 }
 
 llvm::DIType *CGDebugInfo::CreateType(const BuiltinType *BT) {
@@ -581,8 +606,6 @@ llvm::DIType *CGDebugInfo::CreateType(const BuiltinType *BT) {
     return getOrCreateStructPtrType("opencl_clk_event_t", OCLClkEventDITy);
   case BuiltinType::OCLQueue:
     return getOrCreateStructPtrType("opencl_queue_t", OCLQueueDITy);
-  case BuiltinType::OCLNDRange:
-    return getOrCreateStructPtrType("opencl_ndrange_t", OCLNDRangeDITy);
   case BuiltinType::OCLReserveID:
     return getOrCreateStructPtrType("opencl_reserve_id_t", OCLReserveIDDITy);
 
@@ -793,17 +816,19 @@ llvm::DIType *CGDebugInfo::CreatePointerLikeType(llvm::dwarf::Tag Tag,
   // Bit size, align and offset of the type.
   // Size is always the size of a pointer. We can't use getTypeSize here
   // because that does not return the correct value for references.
-  unsigned AS = CGM.getContext().getTargetAddressSpace(PointeeTy);
-  uint64_t Size = CGM.getTarget().getPointerWidth(AS);
+  unsigned AddressSpace = CGM.getContext().getTargetAddressSpace(PointeeTy);
+  uint64_t Size = CGM.getTarget().getPointerWidth(AddressSpace);
   auto Align = getTypeAlignIfRequired(Ty, CGM.getContext());
+  Optional<unsigned> DWARFAddressSpace =
+      CGM.getTarget().getDWARFAddressSpace(AddressSpace);
 
   if (Tag == llvm::dwarf::DW_TAG_reference_type ||
       Tag == llvm::dwarf::DW_TAG_rvalue_reference_type)
     return DBuilder.createReferenceType(Tag, getOrCreateType(PointeeTy, Unit),
-                                        Size, Align);
+                                        Size, Align, DWARFAddressSpace);
   else
     return DBuilder.createPointerType(getOrCreateType(PointeeTy, Unit), Size,
-                                      Align);
+                                      Align, DWARFAddressSpace);
 }
 
 llvm::DIType *CGDebugInfo::getOrCreateStructPtrType(StringRef Name,
@@ -1608,8 +1633,13 @@ llvm::DIType *CGDebugInfo::getOrCreateVTablePtrType(llvm::DIFile *Unit) {
   llvm::DITypeRefArray SElements = DBuilder.getOrCreateTypeArray(STy);
   llvm::DIType *SubTy = DBuilder.createSubroutineType(SElements);
   unsigned Size = Context.getTypeSize(Context.VoidPtrTy);
+  unsigned VtblPtrAddressSpace = CGM.getTarget().getVtblPtrAddressSpace();
+  Optional<unsigned> DWARFAddressSpace =
+      CGM.getTarget().getDWARFAddressSpace(VtblPtrAddressSpace);
+
   llvm::DIType *vtbl_ptr_type =
-      DBuilder.createPointerType(SubTy, Size, 0, "__vtbl_ptr_type");
+      DBuilder.createPointerType(SubTy, Size, 0, DWARFAddressSpace,
+                                 "__vtbl_ptr_type");
   VTablePtrType = DBuilder.createPointerType(vtbl_ptr_type, Size);
   return VTablePtrType;
 }
@@ -1648,10 +1678,14 @@ void CGDebugInfo::CollectVTableInfo(const CXXRecordDecl *RD, llvm::DIFile *Unit,
     unsigned VSlotCount =
         VFTLayout.vtable_components().size() - CGM.getLangOpts().RTTIData;
     unsigned VTableWidth = PtrWidth * VSlotCount;
+    unsigned VtblPtrAddressSpace = CGM.getTarget().getVtblPtrAddressSpace();
+    Optional<unsigned> DWARFAddressSpace =
+        CGM.getTarget().getDWARFAddressSpace(VtblPtrAddressSpace);
 
     // Create a very wide void* type and insert it directly in the element list.
     llvm::DIType *VTableType =
-        DBuilder.createPointerType(nullptr, VTableWidth, 0, "__vtbl_ptr_type");
+        DBuilder.createPointerType(nullptr, VTableWidth, 0, DWARFAddressSpace,
+                                   "__vtbl_ptr_type");
     EltTys.push_back(VTableType);
 
     // The vptr is a pointer to this special vtable type.
@@ -1714,7 +1748,27 @@ void CGDebugInfo::completeType(const RecordDecl *RD) {
     completeRequiredType(RD);
 }
 
+/// Return true if the class or any of its methods are marked dllimport.
+static bool isClassOrMethodDLLImport(const CXXRecordDecl *RD) {
+  if (RD->hasAttr<DLLImportAttr>())
+    return true;
+  for (const CXXMethodDecl *MD : RD->methods())
+    if (MD->hasAttr<DLLImportAttr>())
+      return true;
+  return false;
+}
+
 void CGDebugInfo::completeClassData(const RecordDecl *RD) {
+  if (auto *CXXRD = dyn_cast<CXXRecordDecl>(RD))
+    if (CXXRD->isDynamicClass() &&
+        CGM.getVTableLinkage(CXXRD) ==
+            llvm::GlobalValue::AvailableExternallyLinkage &&
+        !isClassOrMethodDLLImport(CXXRD))
+      return;
+  completeClass(RD);
+}
+
+void CGDebugInfo::completeClass(const RecordDecl *RD) {
   if (DebugKind <= codegenoptions::DebugLineTablesOnly)
     return;
   QualType Ty = CGM.getContext().getRecordType(RD);
@@ -1760,22 +1814,16 @@ static bool isDefinedInClangModule(const RecordDecl *RD) {
   return true;
 }
 
-/// Return true if the class or any of its methods are marked dllimport.
-static bool isClassOrMethodDLLImport(const CXXRecordDecl *RD) {
-  if (RD->hasAttr<DLLImportAttr>())
-    return true;
-  for (const CXXMethodDecl *MD : RD->methods())
-    if (MD->hasAttr<DLLImportAttr>())
-      return true;
-  return false;
-}
-
 static bool shouldOmitDefinition(codegenoptions::DebugInfoKind DebugKind,
                                  bool DebugTypeExtRefs, const RecordDecl *RD,
                                  const LangOptions &LangOpts) {
   if (DebugTypeExtRefs && isDefinedInClangModule(RD->getDefinition()))
     return true;
 
+  if (auto *ES = RD->getASTContext().getExternalSource())
+    if (ES->hasExternalDefinitions(RD) == ExternalASTSource::EK_Always)
+      return true;
+
   if (DebugKind > codegenoptions::LimitedDebugInfo)
     return false;
 
@@ -2009,7 +2057,11 @@ CGDebugInfo::getOrCreateModuleRef(ExternalASTSource::ASTSourceDescriptor Mod,
   if (CreateSkeletonCU && IsRootModule) {
     // PCH files don't have a signature field in the control block,
     // but LLVM detects skeleton CUs by looking for a non-zero DWO id.
-    uint64_t Signature = Mod.getSignature() ? Mod.getSignature() : ~1ULL;
+    // We use the lower 64 bits for debug info.
+    uint64_t Signature =
+        Mod.getSignature()
+            ? (uint64_t)Mod.getSignature()[1] << 32 | Mod.getSignature()[0]
+            : ~1ULL;
     llvm::DIBuilder DIB(CGM.getModule());
     DIB.createCompileUnit(TheCU->getSourceLanguage(),
                           DIB.createFile(Mod.getModuleName(), Mod.getPath()),
@@ -2408,6 +2460,21 @@ llvm::DIType *CGDebugInfo::CreateTypeDefinition(const EnumType *Ty) {
                                         FullName);
 }
 
+llvm::DIMacro *CGDebugInfo::CreateMacro(llvm::DIMacroFile *Parent,
+                                        unsigned MType, SourceLocation LineLoc,
+                                        StringRef Name, StringRef Value) {
+  unsigned Line = LineLoc.isInvalid() ? 0 : getLineNumber(LineLoc);
+  return DBuilder.createMacro(Parent, Line, MType, Name, Value);
+}
+
+llvm::DIMacroFile *CGDebugInfo::CreateTempMacroFile(llvm::DIMacroFile *Parent,
+                                                    SourceLocation LineLoc,
+                                                    SourceLocation FileLoc) {
+  llvm::DIFile *FName = getOrCreateFile(FileLoc);
+  unsigned Line = LineLoc.isInvalid() ? 0 : getLineNumber(LineLoc);
+  return DBuilder.createTempMacroFile(Parent, Line, FName);
+}
+
 static QualType UnwrapTypeForDebugInfo(QualType T, const ASTContext &C) {
   Qualifiers Quals;
   do {
@@ -2451,8 +2518,9 @@ static QualType UnwrapTypeForDebugInfo(QualType T, const ASTContext &C) {
     case Type::SubstTemplateTypeParm:
       T = cast<SubstTemplateTypeParmType>(T)->getReplacementType();
       break;
-    case Type::Auto: {
-      QualType DT = cast<AutoType>(T)->getDeducedType();
+    case Type::Auto:
+    case Type::DeducedTemplateSpecialization: {
+      QualType DT = cast<DeducedType>(T)->getDeducedType();
       assert(!DT.isNull() && "Undeduced types shouldn't reach here.");
       T = DT;
       break;
@@ -2488,11 +2556,17 @@ void CGDebugInfo::completeTemplateDefinition(
     const ClassTemplateSpecializationDecl &SD) {
   if (DebugKind <= codegenoptions::DebugLineTablesOnly)
     return;
+  completeUnusedClass(SD);
+}
+
+void CGDebugInfo::completeUnusedClass(const CXXRecordDecl &D) {
+  if (DebugKind <= codegenoptions::DebugLineTablesOnly)
+    return;
 
-  completeClassData(&SD);
+  completeClassData(&D);
   // In case this type has no member function definitions being emitted, ensure
   // it is retained
-  RetainedTypes.push_back(CGM.getContext().getRecordType(&SD).getAsOpaquePtr());
+  RetainedTypes.push_back(CGM.getContext().getRecordType(&D).getAsOpaquePtr());
 }
 
 llvm::DIType *CGDebugInfo::getOrCreateType(QualType Ty, llvm::DIFile *Unit) {
@@ -2618,6 +2692,7 @@ llvm::DIType *CGDebugInfo::CreateTypeNode(QualType Ty, llvm::DIFile *Unit) {
   case Type::Attributed:
   case Type::Adjusted:
   case Type::Decayed:
+  case Type::DeducedTemplateSpecialization:
   case Type::Elaborated:
   case Type::Paren:
   case Type::SubstTemplateTypeParm:
@@ -2774,9 +2849,10 @@ void CGDebugInfo::collectFunctionDeclProps(GlobalDecl GD, llvm::DIFile *Unit,
   }
   // No need to replicate the linkage name if it isn't different from the
   // subprogram name, no need to have it at all unless coverage is enabled or
-  // debug is set to more than just line tables.
+  // debug is set to more than just line tables or extra debug info is needed.
   if (LinkageName == Name || (!CGM.getCodeGenOpts().EmitGcovArcs &&
                               !CGM.getCodeGenOpts().EmitGcovNotes &&
+                              !CGM.getCodeGenOpts().DebugInfoForProfiling &&
                               DebugKind <= codegenoptions::DebugLineTablesOnly))
     LinkageName = StringRef();
 
@@ -2844,28 +2920,40 @@ void CGDebugInfo::collectVarDeclProps(const VarDecl *VD, llvm::DIFile *&Unit,
  VDContext = getContextDescriptor(cast<Decl>(DC), Mod ? Mod : TheCU);
 }
 
-llvm::DISubprogram *
-CGDebugInfo::getFunctionForwardDeclaration(const FunctionDecl *FD) {
+llvm::DISubprogram *CGDebugInfo::getFunctionFwdDeclOrStub(GlobalDecl GD,
+                                                          bool Stub) {
   llvm::DINodeArray TParamsArray;
   StringRef Name, LinkageName;
   llvm::DINode::DIFlags Flags = llvm::DINode::FlagZero;
-  SourceLocation Loc = FD->getLocation();
+  SourceLocation Loc = GD.getDecl()->getLocation();
   llvm::DIFile *Unit = getOrCreateFile(Loc);
   llvm::DIScope *DContext = Unit;
   unsigned Line = getLineNumber(Loc);
-
-  collectFunctionDeclProps(FD, Unit, Name, LinkageName, DContext,
+  collectFunctionDeclProps(GD, Unit, Name, LinkageName, DContext,
                            TParamsArray, Flags);
+  auto *FD = dyn_cast<FunctionDecl>(GD.getDecl());
+
   // Build function type.
   SmallVector<QualType, 16> ArgTypes;
-  for (const ParmVarDecl *Parm: FD->parameters())
-    ArgTypes.push_back(Parm->getType());
+  if (FD)
+    for (const ParmVarDecl *Parm : FD->parameters())
+      ArgTypes.push_back(Parm->getType());
   CallingConv CC = FD->getType()->castAs<FunctionType>()->getCallConv();
   QualType FnType = CGM.getContext().getFunctionType(
       FD->getReturnType(), ArgTypes, FunctionProtoType::ExtProtoInfo(CC));
+  if (Stub) {
+    return DBuilder.createFunction(
+        DContext, Name, LinkageName, Unit, Line,
+        getOrCreateFunctionType(GD.getDecl(), FnType, Unit),
+        !FD->isExternallyVisible(),
+        /* isDefinition = */ true, 0, Flags, CGM.getLangOpts().Optimize,
+        TParamsArray.get(), getFunctionDeclaration(FD));
+  }
+
   llvm::DISubprogram *SP = DBuilder.createTempFunctionFwdDecl(
       DContext, Name, LinkageName, Unit, Line,
-      getOrCreateFunctionType(FD, FnType, Unit), !FD->isExternallyVisible(),
+      getOrCreateFunctionType(GD.getDecl(), FnType, Unit),
+      !FD->isExternallyVisible(),
       /* isDefinition = */ false, 0, Flags, CGM.getLangOpts().Optimize,
       TParamsArray.get(), getFunctionDeclaration(FD));
   const auto *CanonDecl = cast<FunctionDecl>(FD->getCanonicalDecl());
@@ -2875,6 +2963,16 @@ CGDebugInfo::getFunctionForwardDeclaration(const FunctionDecl *FD) {
   return SP;
 }
 
+llvm::DISubprogram *
+CGDebugInfo::getFunctionForwardDeclaration(GlobalDecl GD) {
+  return getFunctionFwdDeclOrStub(GD, /* Stub = */ false);
+}
+
+llvm::DISubprogram *
+CGDebugInfo::getFunctionStub(GlobalDecl GD) {
+  return getFunctionFwdDeclOrStub(GD, /* Stub = */ true);
+}
+
 llvm::DIGlobalVariable *
 CGDebugInfo::getGlobalVariableForwardDeclaration(const VarDecl *VD) {
   QualType T;
@@ -3146,6 +3244,27 @@ void CGDebugInfo::EmitFunctionDecl(GlobalDecl GD, SourceLocation Loc,
       TParamsArray.get(), getFunctionDeclaration(D)));
 }
 
+void CGDebugInfo::EmitInlineFunctionStart(CGBuilderTy &Builder, GlobalDecl GD) {
+  const auto *FD = cast<FunctionDecl>(GD.getDecl());
+  // If there is a subprogram for this function available then use it.
+  auto FI = SPCache.find(FD->getCanonicalDecl());
+  llvm::DISubprogram *SP = nullptr;
+  if (FI != SPCache.end())
+    SP = dyn_cast_or_null<llvm::DISubprogram>(FI->second);
+  if (!SP)
+    SP = getFunctionStub(GD);
+  FnBeginRegionCount.push_back(LexicalBlockStack.size());
+  LexicalBlockStack.emplace_back(SP);
+  setInlinedAt(Builder.getCurrentDebugLocation());
+  EmitLocation(Builder, FD->getLocation());
+}
+
+void CGDebugInfo::EmitInlineFunctionEnd(CGBuilderTy &Builder) {
+  assert(CurInlinedAt && "unbalanced inline scope stack");
+  EmitFunctionEnd(Builder);
+  setInlinedAt(llvm::DebugLoc(CurInlinedAt).getInlinedAt());
+}
+
 void CGDebugInfo::EmitLocation(CGBuilderTy &Builder, SourceLocation Loc) {
   // Update our current location
   setLocation(Loc);
@@ -3155,7 +3274,7 @@ void CGDebugInfo::EmitLocation(CGBuilderTy &Builder, SourceLocation Loc) {
 
   llvm::MDNode *Scope = LexicalBlockStack.back();
   Builder.SetCurrentDebugLocation(llvm::DebugLoc::get(
-      getLineNumber(CurLoc), getColumnNumber(CurLoc), Scope));
+      getLineNumber(CurLoc), getColumnNumber(CurLoc), Scope, CurInlinedAt));
 }
 
 void CGDebugInfo::CreateLexicalBlock(SourceLocation Loc) {
@@ -3167,14 +3286,29 @@ void CGDebugInfo::CreateLexicalBlock(SourceLocation Loc) {
       getColumnNumber(CurLoc)));
 }
 
+void CGDebugInfo::AppendAddressSpaceXDeref(
+    unsigned AddressSpace,
+    SmallVectorImpl<int64_t> &Expr) const {
+  Optional<unsigned> DWARFAddressSpace =
+      CGM.getTarget().getDWARFAddressSpace(AddressSpace);
+  if (!DWARFAddressSpace)
+    return;
+
+  Expr.push_back(llvm::dwarf::DW_OP_constu);
+  Expr.push_back(DWARFAddressSpace.getValue());
+  Expr.push_back(llvm::dwarf::DW_OP_swap);
+  Expr.push_back(llvm::dwarf::DW_OP_xderef);
+}
+
 void CGDebugInfo::EmitLexicalBlockStart(CGBuilderTy &Builder,
                                         SourceLocation Loc) {
   // Set our current location.
   setLocation(Loc);
 
   // Emit a line table change for the current location inside the new scope.
-  Builder.SetCurrentDebugLocation(llvm::DebugLoc::get(
-      getLineNumber(Loc), getColumnNumber(Loc), LexicalBlockStack.back()));
+  Builder.SetCurrentDebugLocation(
+      llvm::DebugLoc::get(getLineNumber(Loc), getColumnNumber(Loc),
+                          LexicalBlockStack.back(), CurInlinedAt));
 
   if (DebugKind <= codegenoptions::DebugLineTablesOnly)
     return;
@@ -3316,13 +3450,16 @@ void CGDebugInfo::EmitDeclare(const VarDecl *VD, llvm::Value *Storage,
     Line = getLineNumber(VD->getLocation());
     Column = getColumnNumber(VD->getLocation());
   }
-  SmallVector<int64_t, 9> Expr;
+  SmallVector<int64_t, 13> Expr;
   llvm::DINode::DIFlags Flags = llvm::DINode::FlagZero;
   if (VD->isImplicit())
     Flags |= llvm::DINode::FlagArtificial;
 
   auto Align = getDeclAlignIfRequired(VD, CGM.getContext());
 
+  unsigned AddressSpace = CGM.getContext().getTargetAddressSpace(VD->getType());
+  AppendAddressSpaceXDeref(AddressSpace, Expr);
+
   // If this is the first argument and it is implicit then
   // give it an object pointer flag.
   // FIXME: There has to be a better way to do this, but for static
@@ -3360,9 +3497,10 @@ void CGDebugInfo::EmitDeclare(const VarDecl *VD, llvm::Value *Storage,
                                                   Line, Ty, Align);
 
       // Insert an llvm.dbg.declare into the current block.
-      DBuilder.insertDeclare(Storage, D, DBuilder.createExpression(Expr),
-                             llvm::DebugLoc::get(Line, Column, Scope),
-                             Builder.GetInsertBlock());
+      DBuilder.insertDeclare(
+          Storage, D, DBuilder.createExpression(Expr),
+          llvm::DebugLoc::get(Line, Column, Scope, CurInlinedAt),
+          Builder.GetInsertBlock());
       return;
     } else if (isa<VariableArrayType>(VD->getType()))
       Expr.push_back(llvm::dwarf::DW_OP_deref);
@@ -3393,9 +3531,10 @@ void CGDebugInfo::EmitDeclare(const VarDecl *VD, llvm::Value *Storage,
             Flags | llvm::DINode::FlagArtificial, FieldAlign);
 
         // Insert an llvm.dbg.declare into the current block.
-        DBuilder.insertDeclare(Storage, D, DBuilder.createExpression(Expr),
-                               llvm::DebugLoc::get(Line, Column, Scope),
-                               Builder.GetInsertBlock());
+        DBuilder.insertDeclare(
+            Storage, D, DBuilder.createExpression(Expr),
+            llvm::DebugLoc::get(Line, Column, Scope, CurInlinedAt),
+            Builder.GetInsertBlock());
       }
     }
   }
@@ -3411,7 +3550,7 @@ void CGDebugInfo::EmitDeclare(const VarDecl *VD, llvm::Value *Storage,
 
   // Insert an llvm.dbg.declare into the current block.
   DBuilder.insertDeclare(Storage, D, DBuilder.createExpression(Expr),
-                         llvm::DebugLoc::get(Line, Column, Scope),
+                         llvm::DebugLoc::get(Line, Column, Scope, CurInlinedAt),
                          Builder.GetInsertBlock());
 }
 
@@ -3492,7 +3631,8 @@ void CGDebugInfo::EmitDeclareOfBlockDeclRefVariable(
       Line, Ty, false, llvm::DINode::FlagZero, Align);
 
   // Insert an llvm.dbg.declare into the current block.
-  auto DL = llvm::DebugLoc::get(Line, Column, LexicalBlockStack.back());
+  auto DL =
+      llvm::DebugLoc::get(Line, Column, LexicalBlockStack.back(), CurInlinedAt);
   if (InsertPoint)
     DBuilder.insertDeclare(Storage, D, DBuilder.createExpression(addr), DL,
                            InsertPoint);
@@ -3660,12 +3800,13 @@ void CGDebugInfo::EmitDeclareOfBlockLiteralArgVariable(const CGBlockInfo &block,
     // Insert an llvm.dbg.value into the current block.
     DBuilder.insertDbgValueIntrinsic(
         LocalAddr, 0, debugVar, DBuilder.createExpression(),
-        llvm::DebugLoc::get(line, column, scope), Builder.GetInsertBlock());
+        llvm::DebugLoc::get(line, column, scope, CurInlinedAt),
+        Builder.GetInsertBlock());
   }
 
   // Insert an llvm.dbg.declare into the current block.
   DBuilder.insertDeclare(Arg, debugVar, DBuilder.createExpression(),
-                         llvm::DebugLoc::get(line, column, scope),
+                         llvm::DebugLoc::get(line, column, scope, CurInlinedAt),
                          Builder.GetInsertBlock());
 }
 
@@ -3747,9 +3888,16 @@ void CGDebugInfo::EmitGlobalVariable(llvm::GlobalVariable *Var,
     GVE = CollectAnonRecordDecls(RD, Unit, LineNo, LinkageName, Var, DContext);
   } else {
     auto Align = getDeclAlignIfRequired(D, CGM.getContext());
+
+    SmallVector<int64_t, 4> Expr;
+    unsigned AddressSpace =
+        CGM.getContext().getTargetAddressSpace(D->getType());
+    AppendAddressSpaceXDeref(AddressSpace, Expr);
+
     GVE = DBuilder.createGlobalVariableExpression(
         DContext, DeclName, LinkageName, Unit, LineNo, getOrCreateType(T, Unit),
-        Var->hasLocalLinkage(), /*Expr=*/nullptr,
+        Var->hasLocalLinkage(),
+        Expr.empty() ? nullptr : DBuilder.createExpression(Expr),
         getOrCreateStaticDataMemberDeclarationOrNull(D), Align);
     Var->addDebugInfo(GVE);
   }
diff --git a/lib/CodeGen/CGDebugInfo.h b/lib/CodeGen/CGDebugInfo.h
index ac2e8dd2e0a4..5050ca0ad3fa 100644
--- a/lib/CodeGen/CGDebugInfo.h
+++ b/lib/CodeGen/CGDebugInfo.h
@@ -61,6 +61,7 @@ class CGDebugInfo {
   ModuleMap *ClangModuleMap = nullptr;
   ExternalASTSource::ASTSourceDescriptor PCHDescriptor;
   SourceLocation CurLoc;
+  llvm::MDNode *CurInlinedAt = nullptr;
   llvm::DIType *VTablePtrType = nullptr;
   llvm::DIType *ClassTy = nullptr;
   llvm::DICompositeType *ObjTy = nullptr;
@@ -292,6 +293,15 @@ class CGDebugInfo {
   /// Create a new lexical block node and push it on the stack.
   void CreateLexicalBlock(SourceLocation Loc);
 
+  /// If target-specific LLVM \p AddressSpace directly maps to target-specific
+  /// DWARF address space, appends extended dereferencing mechanism to complex
+  /// expression \p Expr. Otherwise, does nothing.
+  ///
+  /// Extended dereferencing mechanism is has the following format:
+  ///     DW_OP_constu <DWARF Address Space> DW_OP_swap DW_OP_xderef
+  void AppendAddressSpaceXDeref(unsigned AddressSpace,
+                                SmallVectorImpl<int64_t> &Expr) const;
+
 public:
   CGDebugInfo(CodeGenModule &CGM);
   ~CGDebugInfo();
@@ -320,6 +330,17 @@ public:
   /// ignored.
   void setLocation(SourceLocation Loc);
 
+  /// Return the current source location. This does not necessarily correspond
+  /// to the IRBuilder's current DebugLoc.
+  SourceLocation getLocation() const { return CurLoc; }
+
+  /// Update the current inline scope. All subsequent calls to \p EmitLocation
+  /// will create a location with this inlinedAt field.
+  void setInlinedAt(llvm::MDNode *InlinedAt) { CurInlinedAt = InlinedAt; }
+
+  /// \return the current inline scope.
+  llvm::MDNode *getInlinedAt() const { return CurInlinedAt; }
+
   // Converts a SourceLocation to a DebugLoc
   llvm::DebugLoc SourceLocToDebugLoc(SourceLocation Loc);
 
@@ -336,6 +357,11 @@ public:
                          SourceLocation ScopeLoc, QualType FnType,
                          llvm::Function *Fn, CGBuilderTy &Builder);
 
+  /// Start a new scope for an inlined function.
+  void EmitInlineFunctionStart(CGBuilderTy &Builder, GlobalDecl GD);
+  /// End an inlined function scope.
+  void EmitInlineFunctionEnd(CGBuilderTy &Builder);
+
   /// Emit debug info for a function declaration.
   void EmitFunctionDecl(GlobalDecl GD, SourceLocation Loc, QualType FnType);
 
@@ -409,9 +435,21 @@ public:
   void completeType(const RecordDecl *RD);
   void completeRequiredType(const RecordDecl *RD);
   void completeClassData(const RecordDecl *RD);
+  void completeClass(const RecordDecl *RD);
 
   void completeTemplateDefinition(const ClassTemplateSpecializationDecl &SD);
-
+  void completeUnusedClass(const CXXRecordDecl &D);
+
+  /// Create debug info for a macro defined by a #define directive or a macro
+  /// undefined by a #undef directive.
+  llvm::DIMacro *CreateMacro(llvm::DIMacroFile *Parent, unsigned MType,
+                             SourceLocation LineLoc, StringRef Name,
+                             StringRef Value);
+
+  /// Create debug info for a file referenced by an #include directive.
+  llvm::DIMacroFile *CreateTempMacroFile(llvm::DIMacroFile *Parent,
+                                         SourceLocation LineLoc,
+                                         SourceLocation FileLoc);
 private:
   /// Emit call to llvm.dbg.declare for a variable declaration.
   void EmitDeclare(const VarDecl *decl, llvm::Value *AI,
@@ -491,11 +529,18 @@ private:
   llvm::DIDerivedType *
   getOrCreateStaticDataMemberDeclarationOrNull(const VarDecl *D);
 
+  /// Helper that either creates a forward declaration or a stub.
+  llvm::DISubprogram *getFunctionFwdDeclOrStub(GlobalDecl GD, bool Stub);
+
   /// Create a subprogram describing the forward declaration
-  /// represented in the given FunctionDecl.
-  llvm::DISubprogram *getFunctionForwardDeclaration(const FunctionDecl *FD);
+  /// represented in the given FunctionDecl wrapped in a GlobalDecl.
+  llvm::DISubprogram *getFunctionForwardDeclaration(GlobalDecl GD);
+
+  /// Create a DISubprogram describing the function
+  /// represented in the given FunctionDecl wrapped in a GlobalDecl.
+  llvm::DISubprogram *getFunctionStub(GlobalDecl GD);
 
-  /// Create a global variable describing the forward decalration
+  /// Create a global variable describing the forward declaration
   /// represented in the given VarDecl.
   llvm::DIGlobalVariable *
   getGlobalVariableForwardDeclaration(const VarDecl *VD);
@@ -622,6 +667,20 @@ public:
 
 };
 
+/// A scoped helper to set the current debug location to an inlined location.
+class ApplyInlineDebugLocation {
+  SourceLocation SavedLocation;
+  CodeGenFunction *CGF;
+
+public:
+  /// Set up the CodeGenFunction's DebugInfo to produce inline locations for the
+  /// function \p InlinedFn. The current debug location becomes the inlined call
+  /// site of the inlined function.
+  ApplyInlineDebugLocation(CodeGenFunction &CGF, GlobalDecl InlinedFn);
+  /// Restore everything back to the orginial state.
+  ~ApplyInlineDebugLocation();
+};
+
 } // namespace CodeGen
 } // namespace clang
 
diff --git a/lib/CodeGen/CGDecl.cpp b/lib/CodeGen/CGDecl.cpp
index 0a88b2310beb..0f959043a22e 100644
--- a/lib/CodeGen/CGDecl.cpp
+++ b/lib/CodeGen/CGDecl.cpp
@@ -50,6 +50,7 @@ void CodeGenFunction::EmitDecl(const Decl &D) {
   case Decl::TemplateTypeParm:
   case Decl::UnresolvedUsingValue:
   case Decl::NonTypeTemplateParm:
+  case Decl::CXXDeductionGuide:
   case Decl::CXXMethod:
   case Decl::CXXConstructor:
   case Decl::CXXDestructor:
@@ -671,6 +672,27 @@ static void drillIntoBlockVariable(CodeGenFunction &CGF,
   lvalue.setAddress(CGF.emitBlockByrefAddress(lvalue.getAddress(), var));
 }
 
+void CodeGenFunction::EmitNullabilityCheck(LValue LHS, llvm::Value *RHS,
+                                           SourceLocation Loc) {
+  if (!SanOpts.has(SanitizerKind::NullabilityAssign))
+    return;
+
+  auto Nullability = LHS.getType()->getNullability(getContext());
+  if (!Nullability || *Nullability != NullabilityKind::NonNull)
+    return;
+
+  // Check if the right hand side of the assignment is nonnull, if the left
+  // hand side must be nonnull.
+  SanitizerScope SanScope(this);
+  llvm::Value *IsNotNull = Builder.CreateIsNotNull(RHS);
+  llvm::Constant *StaticData[] = {
+      EmitCheckSourceLocation(Loc), EmitCheckTypeDescriptor(LHS.getType()),
+      llvm::ConstantInt::get(Int8Ty, 0), // The LogAlignment info is unused.
+      llvm::ConstantInt::get(Int8Ty, TCK_NonnullAssign)};
+  EmitCheck({{IsNotNull, SanitizerKind::NullabilityAssign}},
+            SanitizerHandler::TypeMismatch, StaticData, RHS);
+}
+
 void CodeGenFunction::EmitScalarInit(const Expr *init, const ValueDecl *D,
                                      LValue lvalue, bool capturedByInit) {
   Qualifiers::ObjCLifetime lifetime = lvalue.getObjCLifetime();
@@ -678,6 +700,7 @@ void CodeGenFunction::EmitScalarInit(const Expr *init, const ValueDecl *D,
     llvm::Value *value = EmitScalarExpr(init);
     if (capturedByInit)
       drillIntoBlockVariable(*this, lvalue, cast<VarDecl>(D));
+    EmitNullabilityCheck(lvalue, value, init->getExprLoc());
     EmitStoreThroughLValue(RValue::get(value), lvalue, true);
     return;
   }
@@ -766,6 +789,8 @@ void CodeGenFunction::EmitScalarInit(const Expr *init, const ValueDecl *D,
 
   if (capturedByInit) drillIntoBlockVariable(*this, lvalue, cast<VarDecl>(D));
 
+  EmitNullabilityCheck(lvalue, value, init->getExprLoc());
+
   // If the variable might have been accessed by its initializer, we
   // might have to initialize with a barrier.  We have to do this for
   // both __weak and __strong, but __weak got filtered out above.
@@ -1022,11 +1047,21 @@ CodeGenFunction::EmitAutoVarAlloca(const VarDecl &D) {
       // Emit a lifetime intrinsic if meaningful. There's no point in doing this
       // if we don't have a valid insertion point (?).
       if (HaveInsertPoint() && !IsMSCatchParam) {
-        // goto or switch-case statements can break lifetime into several
-        // regions which need more efforts to handle them correctly. PR28267
-        // This is rare case, but it's better just omit intrinsics than have
-        // them incorrectly placed.
-        if (!Bypasses.IsBypassed(&D)) {
+        // If there's a jump into the lifetime of this variable, its lifetime
+        // gets broken up into several regions in IR, which requires more work
+        // to handle correctly. For now, just omit the intrinsics; this is a
+        // rare case, and it's better to just be conservatively correct.
+        // PR28267.
+        //
+        // We have to do this in all language modes if there's a jump past the
+        // declaration. We also have to do it in C if there's a jump to an
+        // earlier point in the current block because non-VLA lifetimes begin as
+        // soon as the containing block is entered, not when its variables
+        // actually come into scope; suppressing the lifetime annotations
+        // completely in this case is unnecessarily pessimistic, but again, this
+        // is rare.
+        if (!Bypasses.IsBypassed(&D) &&
+            !(!getLangOpts().CPlusPlus && hasLabelBeenSeenInCurrentScope())) {
           uint64_t size = CGM.getDataLayout().getTypeAllocSize(allocaTy);
           emission.SizeForLifetimeMarkers =
               EmitLifetimeStart(size, address.getPointer());
@@ -1083,6 +1118,12 @@ CodeGenFunction::EmitAutoVarAlloca(const VarDecl &D) {
   if (D.hasAttr<AnnotateAttr>())
     EmitVarAnnotations(&D, address.getPointer());
 
+  // Make sure we call @llvm.lifetime.end.
+  if (emission.useLifetimeMarkers())
+    EHStack.pushCleanup<CallLifetimeEnd>(NormalEHLifetimeMarker,
+                                         emission.getAllocatedAddress(),
+                                         emission.getSizeForLifetimeMarkers());
+
   return emission;
 }
 
@@ -1373,13 +1414,6 @@ void CodeGenFunction::EmitAutoVarCleanups(const AutoVarEmission &emission) {
 
   const VarDecl &D = *emission.Variable;
 
-  // Make sure we call @llvm.lifetime.end.  This needs to happen
-  // *last*, so the cleanup needs to be pushed *first*.
-  if (emission.useLifetimeMarkers())
-    EHStack.pushCleanup<CallLifetimeEnd>(NormalEHLifetimeMarker,
-                                         emission.getAllocatedAddress(),
-                                         emission.getSizeForLifetimeMarkers());
-
   // Check the type for a cleanup.
   if (QualType::DestructionKind dtorKind = D.getType().isDestructedType())
     emitAutoVarTypeCleanup(emission, dtorKind);
@@ -1691,17 +1725,19 @@ void CodeGenFunction::pushRegularPartialArrayCleanup(llvm::Value *arrayBegin,
 
 /// Lazily declare the @llvm.lifetime.start intrinsic.
 llvm::Constant *CodeGenModule::getLLVMLifetimeStartFn() {
-  if (LifetimeStartFn) return LifetimeStartFn;
+  if (LifetimeStartFn)
+    return LifetimeStartFn;
   LifetimeStartFn = llvm::Intrinsic::getDeclaration(&getModule(),
-                                            llvm::Intrinsic::lifetime_start);
+    llvm::Intrinsic::lifetime_start, Int8PtrTy);
   return LifetimeStartFn;
 }
 
 /// Lazily declare the @llvm.lifetime.end intrinsic.
 llvm::Constant *CodeGenModule::getLLVMLifetimeEndFn() {
-  if (LifetimeEndFn) return LifetimeEndFn;
+  if (LifetimeEndFn)
+    return LifetimeEndFn;
   LifetimeEndFn = llvm::Intrinsic::getDeclaration(&getModule(),
-                                              llvm::Intrinsic::lifetime_end);
+    llvm::Intrinsic::lifetime_end, Int8PtrTy);
   return LifetimeEndFn;
 }
 
@@ -1869,6 +1905,19 @@ void CodeGenFunction::EmitParmDecl(const VarDecl &D, ParamValue Arg,
 
   if (D.hasAttr<AnnotateAttr>())
     EmitVarAnnotations(&D, DeclPtr.getPointer());
+
+  // We can only check return value nullability if all arguments to the
+  // function satisfy their nullability preconditions. This makes it necessary
+  // to emit null checks for args in the function body itself.
+  if (requiresReturnValueNullabilityCheck()) {
+    auto Nullability = Ty->getNullability(getContext());
+    if (Nullability && *Nullability == NullabilityKind::NonNull) {
+      SanitizerScope SanScope(this);
+      RetValNullabilityPrecondition =
+          Builder.CreateAnd(RetValNullabilityPrecondition,
+                            Builder.CreateIsNotNull(Arg.getAnyValue()));
+    }
+  }
 }
 
 void CodeGenModule::EmitOMPDeclareReduction(const OMPDeclareReductionDecl *D,
diff --git a/lib/CodeGen/CGDeclCXX.cpp b/lib/CodeGen/CGDeclCXX.cpp
index f56e18216931..f61d60a63a6a 100644
--- a/lib/CodeGen/CGDeclCXX.cpp
+++ b/lib/CodeGen/CGDeclCXX.cpp
@@ -237,7 +237,7 @@ void CodeGenFunction::registerGlobalDtorWithAtExit(const VarDecl &VD,
     llvm::FunctionType::get(IntTy, dtorStub->getType(), false);
 
   llvm::Constant *atexit =
-      CGM.CreateRuntimeFunction(atexitTy, "atexit", llvm::AttributeSet(),
+      CGM.CreateRuntimeFunction(atexitTy, "atexit", llvm::AttributeList(),
                                 /*Local=*/true);
   if (llvm::Function *atexitFn = dyn_cast<llvm::Function>(atexit))
     atexitFn->setDoesNotThrow();
diff --git a/lib/CodeGen/CGException.cpp b/lib/CodeGen/CGException.cpp
index f908bf2b3b0a..ca1535182ec1 100644
--- a/lib/CodeGen/CGException.cpp
+++ b/lib/CodeGen/CGException.cpp
@@ -180,8 +180,8 @@ static const EHPersonality &getObjCXXPersonality(const llvm::Triple &T,
   // The GCC runtime's personality function inherently doesn't support
   // mixed EH.  Use the C++ personality just to avoid returning null.
   case ObjCRuntime::GCC:
-  case ObjCRuntime::ObjFW: // XXX: this will change soon
-    return EHPersonality::GNU_ObjC;
+  case ObjCRuntime::ObjFW:
+    return getObjCPersonality(T, L);
   case ObjCRuntime::GNUstep:
     return EHPersonality::GNU_ObjCXX;
   }
@@ -231,7 +231,7 @@ static llvm::Constant *getPersonalityFn(CodeGenModule &CGM,
                                         const EHPersonality &Personality) {
   return CGM.CreateRuntimeFunction(llvm::FunctionType::get(CGM.Int32Ty, true),
                                    Personality.PersonalityFn,
-                                   llvm::AttributeSet(), /*Local=*/true);
+                                   llvm::AttributeList(), /*Local=*/true);
 }
 
 static llvm::Constant *getOpaquePersonalityFn(CodeGenModule &CGM,
@@ -1666,23 +1666,12 @@ void CodeGenFunction::startOutlinedSEHHelper(CodeGenFunction &ParentCGF,
 
   QualType RetTy = IsFilter ? getContext().LongTy : getContext().VoidTy;
 
-  llvm::Function *ParentFn = ParentCGF.CurFn;
   const CGFunctionInfo &FnInfo =
     CGM.getTypes().arrangeBuiltinFunctionDeclaration(RetTy, Args);
 
   llvm::FunctionType *FnTy = CGM.getTypes().GetFunctionType(FnInfo);
   llvm::Function *Fn = llvm::Function::Create(
       FnTy, llvm::GlobalValue::InternalLinkage, Name.str(), &CGM.getModule());
-  // The filter is either in the same comdat as the function, or it's internal.
-  if (llvm::Comdat *C = ParentFn->getComdat()) {
-    Fn->setComdat(C);
-  } else if (ParentFn->hasWeakLinkage() || ParentFn->hasLinkOnceLinkage()) {
-    llvm::Comdat *C = CGM.getModule().getOrInsertComdat(ParentFn->getName());
-    ParentFn->setComdat(C);
-    Fn->setComdat(C);
-  } else {
-    Fn->setLinkage(llvm::GlobalValue::InternalLinkage);
-  }
 
   IsOutlinedSEHHelper = true;
 
diff --git a/lib/CodeGen/CGExpr.cpp b/lib/CodeGen/CGExpr.cpp
index e5e34a5f3ed6..265ef27a46b0 100644
--- a/lib/CodeGen/CGExpr.cpp
+++ b/lib/CodeGen/CGExpr.cpp
@@ -71,7 +71,8 @@ Address CodeGenFunction::CreateTempAlloca(llvm::Type *Ty, CharUnits Align,
 /// block.
 llvm::AllocaInst *CodeGenFunction::CreateTempAlloca(llvm::Type *Ty,
                                                     const Twine &Name) {
-  return new llvm::AllocaInst(Ty, nullptr, Name, AllocaInsertPt);
+  return new llvm::AllocaInst(Ty, CGM.getDataLayout().getAllocaAddrSpace(),
+                              nullptr, Name, AllocaInsertPt);
 }
 
 /// CreateDefaultAlignTempAlloca - This creates an alloca with the
@@ -534,7 +535,8 @@ bool CodeGenFunction::sanitizePerformTypeCheck() const {
 
 void CodeGenFunction::EmitTypeCheck(TypeCheckKind TCK, SourceLocation Loc,
                                     llvm::Value *Ptr, QualType Ty,
-                                    CharUnits Alignment, bool SkipNullCheck) {
+                                    CharUnits Alignment,
+                                    SanitizerSet SkippedChecks) {
   if (!sanitizePerformTypeCheck())
     return;
 
@@ -552,7 +554,7 @@ void CodeGenFunction::EmitTypeCheck(TypeCheckKind TCK, SourceLocation Loc,
   bool AllowNullPointers = TCK == TCK_DowncastPointer || TCK == TCK_Upcast ||
                            TCK == TCK_UpcastToVirtualBase;
   if ((SanOpts.has(SanitizerKind::Null) || AllowNullPointers) &&
-      !SkipNullCheck) {
+      !SkippedChecks.has(SanitizerKind::Null)) {
     // The glvalue must not be an empty glvalue.
     llvm::Value *IsNonNull = Builder.CreateIsNotNull(Ptr);
 
@@ -568,7 +570,9 @@ void CodeGenFunction::EmitTypeCheck(TypeCheckKind TCK, SourceLocation Loc,
     }
   }
 
-  if (SanOpts.has(SanitizerKind::ObjectSize) && !Ty->isIncompleteType()) {
+  if (SanOpts.has(SanitizerKind::ObjectSize) &&
+      !SkippedChecks.has(SanitizerKind::ObjectSize) &&
+      !Ty->isIncompleteType()) {
     uint64_t Size = getContext().getTypeSizeInChars(Ty).getQuantity();
 
     // The glvalue must refer to a large enough storage region.
@@ -578,22 +582,24 @@ void CodeGenFunction::EmitTypeCheck(TypeCheckKind TCK, SourceLocation Loc,
     llvm::Type *Tys[2] = { IntPtrTy, Int8PtrTy };
     llvm::Value *F = CGM.getIntrinsic(llvm::Intrinsic::objectsize, Tys);
     llvm::Value *Min = Builder.getFalse();
+    llvm::Value *NullIsUnknown = Builder.getFalse();
     llvm::Value *CastAddr = Builder.CreateBitCast(Ptr, Int8PtrTy);
-    llvm::Value *LargeEnough =
-        Builder.CreateICmpUGE(Builder.CreateCall(F, {CastAddr, Min}),
-                              llvm::ConstantInt::get(IntPtrTy, Size));
+    llvm::Value *LargeEnough = Builder.CreateICmpUGE(
+        Builder.CreateCall(F, {CastAddr, Min, NullIsUnknown}),
+        llvm::ConstantInt::get(IntPtrTy, Size));
     Checks.push_back(std::make_pair(LargeEnough, SanitizerKind::ObjectSize));
   }
 
   uint64_t AlignVal = 0;
 
-  if (SanOpts.has(SanitizerKind::Alignment)) {
+  if (SanOpts.has(SanitizerKind::Alignment) &&
+      !SkippedChecks.has(SanitizerKind::Alignment)) {
     AlignVal = Alignment.getQuantity();
     if (!Ty->isIncompleteType() && !AlignVal)
       AlignVal = getContext().getTypeAlignInChars(Ty).getQuantity();
 
     // The glvalue must be suitably aligned.
-    if (AlignVal) {
+    if (AlignVal > 1) {
       llvm::Value *Align =
           Builder.CreateAnd(Builder.CreatePtrToInt(Ptr, IntPtrTy),
                             llvm::ConstantInt::get(IntPtrTy, AlignVal - 1));
@@ -624,6 +630,7 @@ void CodeGenFunction::EmitTypeCheck(TypeCheckKind TCK, SourceLocation Loc,
   //       or call a non-static member function
   CXXRecordDecl *RD = Ty->getAsCXXRecordDecl();
   if (SanOpts.has(SanitizerKind::Vptr) &&
+      !SkippedChecks.has(SanitizerKind::Vptr) &&
       (TCK == TCK_MemberAccess || TCK == TCK_MemberCall ||
        TCK == TCK_DowncastPointer || TCK == TCK_DowncastReference ||
        TCK == TCK_UpcastToVirtualBase) &&
@@ -947,15 +954,47 @@ LValue CodeGenFunction::EmitUnsupportedLValue(const Expr *E,
                         E->getType());
 }
 
+bool CodeGenFunction::IsWrappedCXXThis(const Expr *Obj) {
+  const Expr *Base = Obj;
+  while (!isa<CXXThisExpr>(Base)) {
+    // The result of a dynamic_cast can be null.
+    if (isa<CXXDynamicCastExpr>(Base))
+      return false;
+
+    if (const auto *CE = dyn_cast<CastExpr>(Base)) {
+      Base = CE->getSubExpr();
+    } else if (const auto *PE = dyn_cast<ParenExpr>(Base)) {
+      Base = PE->getSubExpr();
+    } else if (const auto *UO = dyn_cast<UnaryOperator>(Base)) {
+      if (UO->getOpcode() == UO_Extension)
+        Base = UO->getSubExpr();
+      else
+        return false;
+    } else {
+      return false;
+    }
+  }
+  return true;
+}
+
 LValue CodeGenFunction::EmitCheckedLValue(const Expr *E, TypeCheckKind TCK) {
   LValue LV;
   if (SanOpts.has(SanitizerKind::ArrayBounds) && isa<ArraySubscriptExpr>(E))
     LV = EmitArraySubscriptExpr(cast<ArraySubscriptExpr>(E), /*Accessed*/true);
   else
     LV = EmitLValue(E);
-  if (!isa<DeclRefExpr>(E) && !LV.isBitField() && LV.isSimple())
+  if (!isa<DeclRefExpr>(E) && !LV.isBitField() && LV.isSimple()) {
+    SanitizerSet SkippedChecks;
+    if (const auto *ME = dyn_cast<MemberExpr>(E)) {
+      bool IsBaseCXXThis = IsWrappedCXXThis(ME->getBase());
+      if (IsBaseCXXThis)
+        SkippedChecks.set(SanitizerKind::Alignment, true);
+      if (IsBaseCXXThis || isa<DeclRefExpr>(ME->getBase()))
+        SkippedChecks.set(SanitizerKind::Null, true);
+    }
     EmitTypeCheck(TCK, E->getExprLoc(), LV.getPointer(),
-                  E->getType(), LV.getAlignment());
+                  E->getType(), LV.getAlignment(), SkippedChecks);
+  }
   return LV;
 }
 
@@ -1033,7 +1072,19 @@ LValue CodeGenFunction::EmitLValue(const Expr *E) {
     const auto *cleanups = cast<ExprWithCleanups>(E);
     enterFullExpression(cleanups);
     RunCleanupsScope Scope(*this);
-    return EmitLValue(cleanups->getSubExpr());
+    LValue LV = EmitLValue(cleanups->getSubExpr());
+    if (LV.isSimple()) {
+      // Defend against branches out of gnu statement expressions surrounded by
+      // cleanups.
+      llvm::Value *V = LV.getPointer();
+      Scope.ForceCleanup({&V});
+      return LValue::MakeAddr(Address(V, LV.getAlignment()), LV.getType(),
+                              getContext(), LV.getAlignmentSource(),
+                              LV.getTBAAInfo());
+    }
+    // FIXME: Is it possible to create an ExprWithCleanups that produces a
+    // bitfield lvalue or some other non-simple lvalue?
+    return LV;
   }
 
   case Expr::CXXDefaultArgExprClass:
@@ -1265,6 +1316,53 @@ llvm::MDNode *CodeGenFunction::getRangeForLoadFromType(QualType Ty) {
   return MDHelper.createRange(Min, End);
 }
 
+bool CodeGenFunction::EmitScalarRangeCheck(llvm::Value *Value, QualType Ty,
+                                           SourceLocation Loc) {
+  bool HasBoolCheck = SanOpts.has(SanitizerKind::Bool);
+  bool HasEnumCheck = SanOpts.has(SanitizerKind::Enum);
+  if (!HasBoolCheck && !HasEnumCheck)
+    return false;
+
+  bool IsBool = hasBooleanRepresentation(Ty) ||
+                NSAPI(CGM.getContext()).isObjCBOOLType(Ty);
+  bool NeedsBoolCheck = HasBoolCheck && IsBool;
+  bool NeedsEnumCheck = HasEnumCheck && Ty->getAs<EnumType>();
+  if (!NeedsBoolCheck && !NeedsEnumCheck)
+    return false;
+
+  // Single-bit booleans don't need to be checked. Special-case this to avoid
+  // a bit width mismatch when handling bitfield values. This is handled by
+  // EmitFromMemory for the non-bitfield case.
+  if (IsBool &&
+      cast<llvm::IntegerType>(Value->getType())->getBitWidth() == 1)
+    return false;
+
+  llvm::APInt Min, End;
+  if (!getRangeForType(*this, Ty, Min, End, /*StrictEnums=*/true, IsBool))
+    return true;
+
+  SanitizerScope SanScope(this);
+  llvm::Value *Check;
+  --End;
+  if (!Min) {
+    Check = Builder.CreateICmpULE(
+        Value, llvm::ConstantInt::get(getLLVMContext(), End));
+  } else {
+    llvm::Value *Upper = Builder.CreateICmpSLE(
+        Value, llvm::ConstantInt::get(getLLVMContext(), End));
+    llvm::Value *Lower = Builder.CreateICmpSGE(
+        Value, llvm::ConstantInt::get(getLLVMContext(), Min));
+    Check = Builder.CreateAnd(Upper, Lower);
+  }
+  llvm::Constant *StaticArgs[] = {EmitCheckSourceLocation(Loc),
+                                  EmitCheckTypeDescriptor(Ty)};
+  SanitizerMask Kind =
+      NeedsEnumCheck ? SanitizerKind::Enum : SanitizerKind::Bool;
+  EmitCheck(std::make_pair(Check, Kind), SanitizerHandler::LoadInvalidValue,
+            StaticArgs, EmitCheckValue(Value));
+  return true;
+}
+
 llvm::Value *CodeGenFunction::EmitLoadOfScalar(Address Addr, bool Volatile,
                                                QualType Ty,
                                                SourceLocation Loc,
@@ -1273,26 +1371,28 @@ llvm::Value *CodeGenFunction::EmitLoadOfScalar(Address Addr, bool Volatile,
                                                QualType TBAABaseType,
                                                uint64_t TBAAOffset,
                                                bool isNontemporal) {
-  // For better performance, handle vector loads differently.
-  if (Ty->isVectorType()) {
-    const llvm::Type *EltTy = Addr.getElementType();
-
-    const auto *VTy = cast<llvm::VectorType>(EltTy);
-
-    // Handle vectors of size 3 like size 4 for better performance.
-    if (VTy->getNumElements() == 3) {
-
-      // Bitcast to vec4 type.
-      llvm::VectorType *vec4Ty = llvm::VectorType::get(VTy->getElementType(),
-                                                         4);
-      Address Cast = Builder.CreateElementBitCast(Addr, vec4Ty, "castToVec4");
-      // Now load value.
-      llvm::Value *V = Builder.CreateLoad(Cast, Volatile, "loadVec4");
-
-      // Shuffle vector to get vec3.
-      V = Builder.CreateShuffleVector(V, llvm::UndefValue::get(vec4Ty),
-                                      {0, 1, 2}, "extractVec");
-      return EmitFromMemory(V, Ty);
+  if (!CGM.getCodeGenOpts().PreserveVec3Type) {
+    // For better performance, handle vector loads differently.
+    if (Ty->isVectorType()) {
+      const llvm::Type *EltTy = Addr.getElementType();
+
+      const auto *VTy = cast<llvm::VectorType>(EltTy);
+
+      // Handle vectors of size 3 like size 4 for better performance.
+      if (VTy->getNumElements() == 3) {
+
+        // Bitcast to vec4 type.
+        llvm::VectorType *vec4Ty =
+            llvm::VectorType::get(VTy->getElementType(), 4);
+        Address Cast = Builder.CreateElementBitCast(Addr, vec4Ty, "castToVec4");
+        // Now load value.
+        llvm::Value *V = Builder.CreateLoad(Cast, Volatile, "loadVec4");
+
+        // Shuffle vector to get vec3.
+        V = Builder.CreateShuffleVector(V, llvm::UndefValue::get(vec4Ty),
+                                        {0, 1, 2}, "extractVec");
+        return EmitFromMemory(V, Ty);
+      }
     }
   }
 
@@ -1317,35 +1417,9 @@ llvm::Value *CodeGenFunction::EmitLoadOfScalar(Address Addr, bool Volatile,
                                       false /*ConvertTypeToTag*/);
   }
 
-  bool IsBool = hasBooleanRepresentation(Ty) ||
-                NSAPI(CGM.getContext()).isObjCBOOLType(Ty);
-  bool NeedsBoolCheck = SanOpts.has(SanitizerKind::Bool) && IsBool;
-  bool NeedsEnumCheck =
-      SanOpts.has(SanitizerKind::Enum) && Ty->getAs<EnumType>();
-  if (NeedsBoolCheck || NeedsEnumCheck) {
-    SanitizerScope SanScope(this);
-    llvm::APInt Min, End;
-    if (getRangeForType(*this, Ty, Min, End, /*StrictEnums=*/true, IsBool)) {
-      --End;
-      llvm::Value *Check;
-      if (!Min)
-        Check = Builder.CreateICmpULE(
-          Load, llvm::ConstantInt::get(getLLVMContext(), End));
-      else {
-        llvm::Value *Upper = Builder.CreateICmpSLE(
-          Load, llvm::ConstantInt::get(getLLVMContext(), End));
-        llvm::Value *Lower = Builder.CreateICmpSGE(
-          Load, llvm::ConstantInt::get(getLLVMContext(), Min));
-        Check = Builder.CreateAnd(Upper, Lower);
-      }
-      llvm::Constant *StaticArgs[] = {
-        EmitCheckSourceLocation(Loc),
-        EmitCheckTypeDescriptor(Ty)
-      };
-      SanitizerMask Kind = NeedsEnumCheck ? SanitizerKind::Enum : SanitizerKind::Bool;
-      EmitCheck(std::make_pair(Check, Kind), SanitizerHandler::LoadInvalidValue,
-                StaticArgs, EmitCheckValue(Load));
-    }
+  if (EmitScalarRangeCheck(Load, Ty, Loc)) {
+    // In order to prevent the optimizer from throwing away the check, don't
+    // attach range metadata to the load.
   } else if (CGM.getCodeGenOpts().OptimizationLevel > 0)
     if (llvm::MDNode *RangeInfo = getRangeForLoadFromType(Ty))
       Load->setMetadata(llvm::LLVMContext::MD_range, RangeInfo);
@@ -1386,24 +1460,25 @@ void CodeGenFunction::EmitStoreOfScalar(llvm::Value *Value, Address Addr,
                                         uint64_t TBAAOffset,
                                         bool isNontemporal) {
 
-  // Handle vectors differently to get better performance.
-  if (Ty->isVectorType()) {
-    llvm::Type *SrcTy = Value->getType();
-    auto *VecTy = cast<llvm::VectorType>(SrcTy);
-    // Handle vec3 special.
-    if (VecTy->getNumElements() == 3) {
-      // Our source is a vec3, do a shuffle vector to make it a vec4.
-      llvm::Constant *Mask[] = {Builder.getInt32(0), Builder.getInt32(1),
-                                Builder.getInt32(2),
-                                llvm::UndefValue::get(Builder.getInt32Ty())};
-      llvm::Value *MaskV = llvm::ConstantVector::get(Mask);
-      Value = Builder.CreateShuffleVector(Value,
-                                          llvm::UndefValue::get(VecTy),
-                                          MaskV, "extractVec");
-      SrcTy = llvm::VectorType::get(VecTy->getElementType(), 4);
-    }
-    if (Addr.getElementType() != SrcTy) {
-      Addr = Builder.CreateElementBitCast(Addr, SrcTy, "storetmp");
+  if (!CGM.getCodeGenOpts().PreserveVec3Type) {
+    // Handle vectors differently to get better performance.
+    if (Ty->isVectorType()) {
+      llvm::Type *SrcTy = Value->getType();
+      auto *VecTy = cast<llvm::VectorType>(SrcTy);
+      // Handle vec3 special.
+      if (VecTy->getNumElements() == 3) {
+        // Our source is a vec3, do a shuffle vector to make it a vec4.
+        llvm::Constant *Mask[] = {Builder.getInt32(0), Builder.getInt32(1),
+                                  Builder.getInt32(2),
+                                  llvm::UndefValue::get(Builder.getInt32Ty())};
+        llvm::Value *MaskV = llvm::ConstantVector::get(Mask);
+        Value = Builder.CreateShuffleVector(Value, llvm::UndefValue::get(VecTy),
+                                            MaskV, "extractVec");
+        SrcTy = llvm::VectorType::get(VecTy->getElementType(), 4);
+      }
+      if (Addr.getElementType() != SrcTy) {
+        Addr = Builder.CreateElementBitCast(Addr, SrcTy, "storetmp");
+      }
     }
   }
 
@@ -1487,10 +1562,11 @@ RValue CodeGenFunction::EmitLoadOfLValue(LValue LV, SourceLocation Loc) {
     return EmitLoadOfGlobalRegLValue(LV);
 
   assert(LV.isBitField() && "Unknown LValue type!");
-  return EmitLoadOfBitfieldLValue(LV);
+  return EmitLoadOfBitfieldLValue(LV, Loc);
 }
 
-RValue CodeGenFunction::EmitLoadOfBitfieldLValue(LValue LV) {
+RValue CodeGenFunction::EmitLoadOfBitfieldLValue(LValue LV,
+                                                 SourceLocation Loc) {
   const CGBitFieldInfo &Info = LV.getBitFieldInfo();
 
   // Get the output type.
@@ -1515,7 +1591,7 @@ RValue CodeGenFunction::EmitLoadOfBitfieldLValue(LValue LV) {
                               "bf.clear");
   }
   Val = Builder.CreateIntCast(Val, ResLTy, Info.IsSigned, "bf.cast");
-
+  EmitScalarRangeCheck(Val, LV.getType(), Loc);
   return RValue::get(Val);
 }
 
@@ -2545,8 +2621,8 @@ static void emitCheckHandlerCall(CodeGenFunction &CGF,
 
   llvm::Value *Fn = CGF.CGM.CreateRuntimeFunction(
       FnType, FnName,
-      llvm::AttributeSet::get(CGF.getLLVMContext(),
-                              llvm::AttributeSet::FunctionIndex, B),
+      llvm::AttributeList::get(CGF.getLLVMContext(),
+                               llvm::AttributeList::FunctionIndex, B),
       /*Local=*/true);
   llvm::CallInst *HandlerCall = CGF.EmitNounwindRuntimeCall(Fn, FnArgs);
   if (!MayReturn) {
@@ -2709,6 +2785,24 @@ void CodeGenFunction::EmitCfiSlowPathCheck(
   EmitBlock(Cont);
 }
 
+// Emit a stub for __cfi_check function so that the linker knows about this
+// symbol in LTO mode.
+void CodeGenFunction::EmitCfiCheckStub() {
+  llvm::Module *M = &CGM.getModule();
+  auto &Ctx = M->getContext();
+  llvm::Function *F = llvm::Function::Create(
+      llvm::FunctionType::get(VoidTy, {Int64Ty, Int8PtrTy, Int8PtrTy}, false),
+      llvm::GlobalValue::WeakAnyLinkage, "__cfi_check", M);
+  llvm::BasicBlock *BB = llvm::BasicBlock::Create(Ctx, "entry", F);
+  // FIXME: consider emitting an intrinsic call like
+  // call void @llvm.cfi_check(i64 %0, i8* %1, i8* %2)
+  // which can be lowered in CrossDSOCFI pass to the actual contents of
+  // __cfi_check. This would allow inlining of __cfi_check calls.
+  llvm::CallInst::Create(
+      llvm::Intrinsic::getDeclaration(M, llvm::Intrinsic::trap), "", BB);
+  llvm::ReturnInst::Create(Ctx, nullptr, BB);
+}
+
 // This function is basically a switch over the CFI failure kind, which is
 // extracted from CFICheckFailData (1st function argument). Each case is either
 // llvm.trap or a call to one of the two runtime handlers, based on
@@ -2821,7 +2915,7 @@ llvm::CallInst *CodeGenFunction::EmitTrapCall(llvm::Intrinsic::ID IntrID) {
   if (!CGM.getCodeGenOpts().TrapFuncName.empty()) {
     auto A = llvm::Attribute::get(getLLVMContext(), "trap-func-name",
                                   CGM.getCodeGenOpts().TrapFuncName);
-    TrapCall->addAttribute(llvm::AttributeSet::FunctionIndex, A);
+    TrapCall->addAttribute(llvm::AttributeList::FunctionIndex, A);
   }
 
   return TrapCall;
@@ -3335,7 +3429,14 @@ LValue CodeGenFunction::EmitMemberExpr(const MemberExpr *E) {
     AlignmentSource AlignSource;
     Address Addr = EmitPointerWithAlignment(BaseExpr, &AlignSource);
     QualType PtrTy = BaseExpr->getType()->getPointeeType();
-    EmitTypeCheck(TCK_MemberAccess, E->getExprLoc(), Addr.getPointer(), PtrTy);
+    SanitizerSet SkippedChecks;
+    bool IsBaseCXXThis = IsWrappedCXXThis(BaseExpr);
+    if (IsBaseCXXThis)
+      SkippedChecks.set(SanitizerKind::Alignment, true);
+    if (IsBaseCXXThis || isa<DeclRefExpr>(BaseExpr))
+      SkippedChecks.set(SanitizerKind::Null, true);
+    EmitTypeCheck(TCK_MemberAccess, E->getExprLoc(), Addr.getPointer(), PtrTy,
+                  /*Alignment=*/CharUnits::Zero(), SkippedChecks);
     BaseLV = MakeAddrLValue(Addr, PtrTy, AlignSource);
   } else
     BaseLV = EmitCheckedLValue(BaseExpr, TCK_MemberAccess);
diff --git a/lib/CodeGen/CGExprAgg.cpp b/lib/CodeGen/CGExprAgg.cpp
index 009244784e50..49bbb4808eaa 100644
--- a/lib/CodeGen/CGExprAgg.cpp
+++ b/lib/CodeGen/CGExprAgg.cpp
@@ -111,6 +111,13 @@ public:
   void VisitGenericSelectionExpr(GenericSelectionExpr *GE) {
     Visit(GE->getResultExpr());
   }
+  void VisitCoawaitExpr(CoawaitExpr *E) {
+    CGF.EmitCoawaitExpr(*E, Dest, IsResultUnused);
+  }
+  void VisitCoyieldExpr(CoyieldExpr *E) {
+    CGF.EmitCoyieldExpr(*E, Dest, IsResultUnused);
+  }
+  void VisitUnaryCoawait(UnaryOperator *E) { Visit(E->getSubExpr()); }
   void VisitUnaryExtension(UnaryOperator *E) { Visit(E->getSubExpr()); }
   void VisitSubstNonTypeTemplateParmExpr(SubstNonTypeTemplateParmExpr *E) {
     return Visit(E->getReplacement());
@@ -1267,7 +1274,7 @@ void AggExprEmitter::VisitInitListExpr(InitListExpr *E) {
       // Store the initializer into the field.
       EmitInitializationToLValue(E->getInit(curInitIndex++), LV);
     } else {
-      // We're out of initalizers; default-initialize to null
+      // We're out of initializers; default-initialize to null
       EmitNullInitializationToLValue(LV);
     }
 
diff --git a/lib/CodeGen/CGExprCXX.cpp b/lib/CodeGen/CGExprCXX.cpp
index 71c8fb8b7ae3..d02f074dd605 100644
--- a/lib/CodeGen/CGExprCXX.cpp
+++ b/lib/CodeGen/CGExprCXX.cpp
@@ -24,7 +24,15 @@
 using namespace clang;
 using namespace CodeGen;
 
-static RequiredArgs
+namespace {
+struct MemberCallInfo {
+  RequiredArgs ReqArgs;
+  // Number of prefix arguments for the call. Ignores the `this` pointer.
+  unsigned PrefixSize;
+};
+}
+
+static MemberCallInfo
 commonEmitCXXMemberOrOperatorCall(CodeGenFunction &CGF, const CXXMethodDecl *MD,
                                   llvm::Value *This, llvm::Value *ImplicitParam,
                                   QualType ImplicitParamTy, const CallExpr *CE,
@@ -48,6 +56,7 @@ commonEmitCXXMemberOrOperatorCall(CodeGenFunction &CGF, const CXXMethodDecl *MD,
 
   const FunctionProtoType *FPT = MD->getType()->castAs<FunctionProtoType>();
   RequiredArgs required = RequiredArgs::forPrototypePlus(FPT, Args.size(), MD);
+  unsigned PrefixSize = Args.size() - 1;
 
   // And the rest of the call args.
   if (RtlArgs) {
@@ -65,7 +74,7 @@ commonEmitCXXMemberOrOperatorCall(CodeGenFunction &CGF, const CXXMethodDecl *MD,
         FPT->getNumParams() == 0 &&
         "No CallExpr specified for function with non-zero number of arguments");
   }
-  return required;
+  return {required, PrefixSize};
 }
 
 RValue CodeGenFunction::EmitCXXMemberOrOperatorCall(
@@ -75,9 +84,10 @@ RValue CodeGenFunction::EmitCXXMemberOrOperatorCall(
     const CallExpr *CE, CallArgList *RtlArgs) {
   const FunctionProtoType *FPT = MD->getType()->castAs<FunctionProtoType>();
   CallArgList Args;
-  RequiredArgs required = commonEmitCXXMemberOrOperatorCall(
+  MemberCallInfo CallInfo = commonEmitCXXMemberOrOperatorCall(
       *this, MD, This, ImplicitParam, ImplicitParamTy, CE, Args, RtlArgs);
-  auto &FnInfo = CGM.getTypes().arrangeCXXMethodCall(Args, FPT, required);
+  auto &FnInfo = CGM.getTypes().arrangeCXXMethodCall(
+      Args, FPT, CallInfo.ReqArgs, CallInfo.PrefixSize);
   return EmitCall(FnInfo, Callee, ReturnValue, Args);
 }
 
@@ -290,10 +300,20 @@ RValue CodeGenFunction::EmitCXXMemberOrOperatorMemberCallExpr(
   if (CE)
     CallLoc = CE->getExprLoc();
 
-  EmitTypeCheck(isa<CXXConstructorDecl>(CalleeDecl)
-                ? CodeGenFunction::TCK_ConstructorCall
-                : CodeGenFunction::TCK_MemberCall,
-                CallLoc, This.getPointer(), C.getRecordType(CalleeDecl->getParent()));
+  SanitizerSet SkippedChecks;
+  if (const auto *CMCE = dyn_cast<CXXMemberCallExpr>(CE)) {
+    auto *IOA = CMCE->getImplicitObjectArgument();
+    bool IsImplicitObjectCXXThis = IsWrappedCXXThis(IOA);
+    if (IsImplicitObjectCXXThis)
+      SkippedChecks.set(SanitizerKind::Alignment, true);
+    if (IsImplicitObjectCXXThis || isa<DeclRefExpr>(IOA))
+      SkippedChecks.set(SanitizerKind::Null, true);
+  }
+  EmitTypeCheck(
+      isa<CXXConstructorDecl>(CalleeDecl) ? CodeGenFunction::TCK_ConstructorCall
+                                          : CodeGenFunction::TCK_MemberCall,
+      CallLoc, This.getPointer(), C.getRecordType(CalleeDecl->getParent()),
+      /*Alignment=*/CharUnits::Zero(), SkippedChecks);
 
   // FIXME: Uses of 'MD' past this point need to be audited. We may need to use
   // 'CalleeDecl' instead.
@@ -420,7 +440,8 @@ CodeGenFunction::EmitCXXMemberPointerCallExpr(const CXXMemberCallExpr *E,
 
   // And the rest of the call args
   EmitCallArgs(Args, FPT, E->arguments());
-  return EmitCall(CGM.getTypes().arrangeCXXMethodCall(Args, FPT, required),
+  return EmitCall(CGM.getTypes().arrangeCXXMethodCall(Args, FPT, required,
+                                                      /*PrefixSize=*/0),
                   Callee, ReturnValue, Args);
 }
 
@@ -659,7 +680,10 @@ static llvm::Value *EmitCXXNewAllocSize(CodeGenFunction &CGF,
   // Emit the array size expression.
   // We multiply the size of all dimensions for NumElements.
   // e.g for 'int[2][3]', ElemType is 'int' and NumElements is 6.
-  numElements = CGF.EmitScalarExpr(e->getArraySize());
+  numElements = CGF.CGM.EmitConstantExpr(e->getArraySize(),
+                                         CGF.getContext().getSizeType(), &CGF);
+  if (!numElements)
+    numElements = CGF.EmitScalarExpr(e->getArraySize());
   assert(isa<llvm::IntegerType>(numElements->getType()));
 
   // The number of elements can be have an arbitrary integer type;
@@ -1256,10 +1280,10 @@ static RValue EmitNewDeleteCall(CodeGenFunction &CGF,
       Fn && Fn->hasFnAttribute(llvm::Attribute::NoBuiltin)) {
     // FIXME: Add addAttribute to CallSite.
     if (llvm::CallInst *CI = dyn_cast<llvm::CallInst>(CallOrInvoke))
-      CI->addAttribute(llvm::AttributeSet::FunctionIndex,
+      CI->addAttribute(llvm::AttributeList::FunctionIndex,
                        llvm::Attribute::Builtin);
     else if (llvm::InvokeInst *II = dyn_cast<llvm::InvokeInst>(CallOrInvoke))
-      II->addAttribute(llvm::AttributeSet::FunctionIndex,
+      II->addAttribute(llvm::AttributeList::FunctionIndex,
                        llvm::Attribute::Builtin);
     else
       llvm_unreachable("unexpected kind of call instruction");
@@ -1560,7 +1584,7 @@ llvm::Value *CodeGenFunction::EmitCXXNewExpr(const CXXNewExpr *E) {
 
     // FIXME: Why do we not pass a CalleeDecl here?
     EmitCallArgs(allocatorArgs, allocatorType, E->placement_arguments(),
-                 /*CalleeDecl*/nullptr, /*ParamsToSkip*/ParamsToSkip);
+                 /*AC*/AbstractCallee(), /*ParamsToSkip*/ParamsToSkip);
 
     RValue RV =
       EmitNewDeleteCall(*this, allocator, allocatorType, allocatorArgs);
diff --git a/lib/CodeGen/CGExprComplex.cpp b/lib/CodeGen/CGExprComplex.cpp
index 59bc9cdbc056..980972370dc2 100644
--- a/lib/CodeGen/CGExprComplex.cpp
+++ b/lib/CodeGen/CGExprComplex.cpp
@@ -110,6 +110,16 @@ public:
   VisitSubstNonTypeTemplateParmExpr(SubstNonTypeTemplateParmExpr *PE) {
     return Visit(PE->getReplacement());
   }
+  ComplexPairTy VisitCoawaitExpr(CoawaitExpr *S) {
+    return CGF.EmitCoawaitExpr(*S).getComplexVal();
+  }
+  ComplexPairTy VisitCoyieldExpr(CoyieldExpr *S) {
+    return CGF.EmitCoyieldExpr(*S).getComplexVal();
+  }
+  ComplexPairTy VisitUnaryCoawait(const UnaryOperator *E) {
+    return Visit(E->getSubExpr());
+  }
+
 
   // l-values.
   ComplexPairTy VisitDeclRefExpr(DeclRefExpr *E) {
@@ -198,7 +208,11 @@ public:
   ComplexPairTy VisitExprWithCleanups(ExprWithCleanups *E) {
     CGF.enterFullExpression(E);
     CodeGenFunction::RunCleanupsScope Scope(CGF);
-    return Visit(E->getSubExpr());
+    ComplexPairTy Vals = Visit(E->getSubExpr());
+    // Defend against dominance problems caused by jumps out of expression
+    // evaluation through the shared cleanup block.
+    Scope.ForceCleanup({&Vals.first, &Vals.second});
+    return Vals;
   }
   ComplexPairTy VisitCXXScalarValueInitExpr(CXXScalarValueInitExpr *E) {
     assert(E->getType()->isAnyComplexType() && "Expected complex type!");
diff --git a/lib/CodeGen/CGExprScalar.cpp b/lib/CodeGen/CGExprScalar.cpp
index 1b85c45cd4be..a64303831171 100644
--- a/lib/CodeGen/CGExprScalar.cpp
+++ b/lib/CodeGen/CGExprScalar.cpp
@@ -12,6 +12,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "CodeGenFunction.h"
+#include "CGCleanup.h"
 #include "CGCXXABI.h"
 #include "CGDebugInfo.h"
 #include "CGObjCRuntime.h"
@@ -24,6 +25,7 @@
 #include "clang/AST/StmtVisitor.h"
 #include "clang/Basic/TargetInfo.h"
 #include "clang/Frontend/CodeGenOptions.h"
+#include "llvm/ADT/Optional.h"
 #include "llvm/IR/CFG.h"
 #include "llvm/IR/Constants.h"
 #include "llvm/IR/DataLayout.h"
@@ -47,7 +49,7 @@ struct BinOpInfo {
   Value *RHS;
   QualType Ty;  // Computation Type.
   BinaryOperator::Opcode Opcode; // Opcode of BinOp to perform
-  bool FPContractable;
+  FPOptions FPFeatures;
   const Expr *E;      // Entire expr, for error unsupported.  May not be binop.
 };
 
@@ -58,6 +60,75 @@ static bool MustVisitNullValue(const Expr *E) {
   return E->getType()->isNullPtrType();
 }
 
+/// If \p E is a widened promoted integer, get its base (unpromoted) type.
+static llvm::Optional<QualType> getUnwidenedIntegerType(const ASTContext &Ctx,
+                                                        const Expr *E) {
+  const Expr *Base = E->IgnoreImpCasts();
+  if (E == Base)
+    return llvm::None;
+
+  QualType BaseTy = Base->getType();
+  if (!BaseTy->isPromotableIntegerType() ||
+      Ctx.getTypeSize(BaseTy) >= Ctx.getTypeSize(E->getType()))
+    return llvm::None;
+
+  return BaseTy;
+}
+
+/// Check if \p E is a widened promoted integer.
+static bool IsWidenedIntegerOp(const ASTContext &Ctx, const Expr *E) {
+  return getUnwidenedIntegerType(Ctx, E).hasValue();
+}
+
+/// Check if we can skip the overflow check for \p Op.
+static bool CanElideOverflowCheck(const ASTContext &Ctx, const BinOpInfo &Op) {
+  assert((isa<UnaryOperator>(Op.E) || isa<BinaryOperator>(Op.E)) &&
+         "Expected a unary or binary operator");
+
+  if (const auto *UO = dyn_cast<UnaryOperator>(Op.E))
+    return IsWidenedIntegerOp(Ctx, UO->getSubExpr());
+
+  const auto *BO = cast<BinaryOperator>(Op.E);
+  auto OptionalLHSTy = getUnwidenedIntegerType(Ctx, BO->getLHS());
+  if (!OptionalLHSTy)
+    return false;
+
+  auto OptionalRHSTy = getUnwidenedIntegerType(Ctx, BO->getRHS());
+  if (!OptionalRHSTy)
+    return false;
+
+  QualType LHSTy = *OptionalLHSTy;
+  QualType RHSTy = *OptionalRHSTy;
+
+  // We usually don't need overflow checks for binary operations with widened
+  // operands. Multiplication with promoted unsigned operands is a special case.
+  if ((Op.Opcode != BO_Mul && Op.Opcode != BO_MulAssign) ||
+      !LHSTy->isUnsignedIntegerType() || !RHSTy->isUnsignedIntegerType())
+    return true;
+
+  // The overflow check can be skipped if either one of the unpromoted types
+  // are less than half the size of the promoted type.
+  unsigned PromotedSize = Ctx.getTypeSize(Op.E->getType());
+  return (2 * Ctx.getTypeSize(LHSTy)) < PromotedSize ||
+         (2 * Ctx.getTypeSize(RHSTy)) < PromotedSize;
+}
+
+/// Update the FastMathFlags of LLVM IR from the FPOptions in LangOptions.
+static void updateFastMathFlags(llvm::FastMathFlags &FMF,
+                                FPOptions FPFeatures) {
+  FMF.setAllowContract(FPFeatures.allowFPContractAcrossStatement());
+}
+
+/// Propagate fast-math flags from \p Op to the instruction in \p V.
+static Value *propagateFMFlags(Value *V, const BinOpInfo &Op) {
+  if (auto *I = dyn_cast<llvm::Instruction>(V)) {
+    llvm::FastMathFlags FMF = I->getFastMathFlags();
+    updateFastMathFlags(FMF, Op.FPFeatures);
+    I->setFastMathFlags(FMF);
+  }
+  return V;
+}
+
 class ScalarExprEmitter
   : public StmtVisitor<ScalarExprEmitter, Value*> {
   CodeGenFunction &CGF;
@@ -221,6 +292,15 @@ public:
   Value *VisitGenericSelectionExpr(GenericSelectionExpr *GE) {
     return Visit(GE->getResultExpr());
   }
+  Value *VisitCoawaitExpr(CoawaitExpr *S) {
+    return CGF.EmitCoawaitExpr(*S).getScalarVal();
+  }
+  Value *VisitCoyieldExpr(CoyieldExpr *S) {
+    return CGF.EmitCoyieldExpr(*S).getScalarVal();
+  }
+  Value *VisitUnaryCoawait(const UnaryOperator *E) {
+    return Visit(E->getSubExpr());
+  }
 
   // Leaves.
   Value *VisitIntegerLiteral(const IntegerLiteral *E) {
@@ -300,6 +380,24 @@ public:
     return V;
   }
 
+  Value *VisitObjCAvailabilityCheckExpr(ObjCAvailabilityCheckExpr *E) {
+    VersionTuple Version = E->getVersion();
+
+    // If we're checking for a platform older than our minimum deployment
+    // target, we can fold the check away.
+    if (Version <= CGF.CGM.getTarget().getPlatformMinVersion())
+      return llvm::ConstantInt::get(Builder.getInt1Ty(), 1);
+
+    Optional<unsigned> Min = Version.getMinor(), SMin = Version.getSubminor();
+    llvm::Value *Args[] = {
+        llvm::ConstantInt::get(CGF.CGM.Int32Ty, Version.getMajor()),
+        llvm::ConstantInt::get(CGF.CGM.Int32Ty, Min ? *Min : 0),
+        llvm::ConstantInt::get(CGF.CGM.Int32Ty, SMin ? *SMin : 0),
+    };
+
+    return CGF.EmitBuiltinAvailable(Args);
+  }
+
   Value *VisitArraySubscriptExpr(ArraySubscriptExpr *E);
   Value *VisitShuffleVectorExpr(ShuffleVectorExpr *E);
   Value *VisitConvertVectorExpr(ConvertVectorExpr *E);
@@ -405,11 +503,7 @@ public:
     return CGF.LoadCXXThis();
   }
 
-  Value *VisitExprWithCleanups(ExprWithCleanups *E) {
-    CGF.enterFullExpression(E);
-    CodeGenFunction::RunCleanupsScope Scope(CGF);
-    return Visit(E->getSubExpr());
-  }
+  Value *VisitExprWithCleanups(ExprWithCleanups *E);
   Value *VisitCXXNewExpr(const CXXNewExpr *E) {
     return CGF.EmitCXXNewExpr(E);
   }
@@ -464,16 +558,21 @@ public:
           return Builder.CreateNSWMul(Ops.LHS, Ops.RHS, "mul");
         // Fall through.
       case LangOptions::SOB_Trapping:
+        if (CanElideOverflowCheck(CGF.getContext(), Ops))
+          return Builder.CreateNSWMul(Ops.LHS, Ops.RHS, "mul");
         return EmitOverflowCheckedBinOp(Ops);
       }
     }
 
     if (Ops.Ty->isUnsignedIntegerType() &&
-        CGF.SanOpts.has(SanitizerKind::UnsignedIntegerOverflow))
+        CGF.SanOpts.has(SanitizerKind::UnsignedIntegerOverflow) &&
+        !CanElideOverflowCheck(CGF.getContext(), Ops))
       return EmitOverflowCheckedBinOp(Ops);
 
-    if (Ops.LHS->getType()->isFPOrFPVectorTy())
-      return Builder.CreateFMul(Ops.LHS, Ops.RHS, "mul");
+    if (Ops.LHS->getType()->isFPOrFPVectorTy()) {
+      Value *V = Builder.CreateFMul(Ops.LHS, Ops.RHS, "mul");
+      return propagateFMFlags(V, Ops);
+    }
     return Builder.CreateMul(Ops.LHS, Ops.RHS, "mul");
   }
   /// Create a binary op that checks for overflow.
@@ -1616,6 +1715,16 @@ Value *ScalarExprEmitter::VisitStmtExpr(const StmtExpr *E) {
                               E->getExprLoc());
 }
 
+Value *ScalarExprEmitter::VisitExprWithCleanups(ExprWithCleanups *E) {
+  CGF.enterFullExpression(E);
+  CodeGenFunction::RunCleanupsScope Scope(CGF);
+  Value *V = Visit(E->getSubExpr());
+  // Defend against dominance problems caused by jumps out of expression
+  // evaluation through the shared cleanup block.
+  Scope.ForceCleanup({&V});
+  return V;
+}
+
 //===----------------------------------------------------------------------===//
 //                             Unary Operators
 //===----------------------------------------------------------------------===//
@@ -1627,7 +1736,7 @@ static BinOpInfo createBinOpInfoFromIncDec(const UnaryOperator *E,
   BinOp.RHS = llvm::ConstantInt::get(InVal->getType(), 1, false);
   BinOp.Ty = E->getType();
   BinOp.Opcode = IsInc ? BO_Add : BO_Sub;
-  BinOp.FPContractable = false;
+  // FIXME: once UnaryOperator carries FPFeatures, copy it here.
   BinOp.E = E;
   return BinOp;
 }
@@ -1645,6 +1754,8 @@ llvm::Value *ScalarExprEmitter::EmitIncDecConsiderOverflowBehavior(
       return Builder.CreateNSWAdd(InVal, Amount, Name);
     // Fall through.
   case LangOptions::SOB_Trapping:
+    if (IsWidenedIntegerOp(CGF.getContext(), E->getSubExpr()))
+      return Builder.CreateNSWAdd(InVal, Amount, Name);
     return EmitOverflowCheckedBinOp(createBinOpInfoFromIncDec(E, InVal, IsInc));
   }
   llvm_unreachable("Unknown SignedOverflowBehaviorTy");
@@ -1891,7 +2002,7 @@ Value *ScalarExprEmitter::VisitUnaryMinus(const UnaryOperator *E) {
     BinOp.LHS = llvm::Constant::getNullValue(BinOp.RHS->getType());
   BinOp.Ty = E->getType();
   BinOp.Opcode = BO_Sub;
-  BinOp.FPContractable = false;
+  // FIXME: once UnaryOperator carries FPFeatures, copy it here.
   BinOp.E = E;
   return EmitSub(BinOp);
 }
@@ -2112,7 +2223,7 @@ BinOpInfo ScalarExprEmitter::EmitBinOps(const BinaryOperator *E) {
   Result.RHS = Visit(E->getRHS());
   Result.Ty  = E->getType();
   Result.Opcode = E->getOpcode();
-  Result.FPContractable = E->isFPContractable();
+  Result.FPFeatures = E->getFPFeatures();
   Result.E = E;
   return Result;
 }
@@ -2132,7 +2243,7 @@ LValue ScalarExprEmitter::EmitCompoundAssignLValue(
   OpInfo.RHS = Visit(E->getRHS());
   OpInfo.Ty = E->getComputationResultType();
   OpInfo.Opcode = E->getOpcode();
-  OpInfo.FPContractable = E->isFPContractable();
+  OpInfo.FPFeatures = E->getFPFeatures();
   OpInfo.E = E;
   // Load/convert the LHS.
   LValue LHSLV = EmitCheckedLValue(E->getLHS(), CodeGenFunction::TCK_Store);
@@ -2263,8 +2374,10 @@ void ScalarExprEmitter::EmitUndefinedBehaviorIntegerDivAndRemCheck(
                                     SanitizerKind::IntegerDivideByZero));
   }
 
+  const auto *BO = cast<BinaryOperator>(Ops.E);
   if (CGF.SanOpts.has(SanitizerKind::SignedIntegerOverflow) &&
-      Ops.Ty->hasSignedIntegerRepresentation()) {
+      Ops.Ty->hasSignedIntegerRepresentation() &&
+      !IsWidenedIntegerOp(CGF.getContext(), BO->getLHS())) {
     llvm::IntegerType *Ty = cast<llvm::IntegerType>(Zero->getType());
 
     llvm::Value *IntMin =
@@ -2324,12 +2437,12 @@ Value *ScalarExprEmitter::EmitDiv(const BinOpInfo &Ops) {
 
 Value *ScalarExprEmitter::EmitRem(const BinOpInfo &Ops) {
   // Rem in C can't be a floating point type: C99 6.5.5p2.
-  if (CGF.SanOpts.has(SanitizerKind::IntegerDivideByZero)) {
+  if ((CGF.SanOpts.has(SanitizerKind::IntegerDivideByZero) ||
+       CGF.SanOpts.has(SanitizerKind::SignedIntegerOverflow)) &&
+      Ops.Ty->isIntegerType()) {
     CodeGenFunction::SanitizerScope SanScope(&CGF);
     llvm::Value *Zero = llvm::Constant::getNullValue(ConvertType(Ops.Ty));
-
-    if (Ops.Ty->isIntegerType())
-      EmitUndefinedBehaviorIntegerDivAndRemCheck(Ops, Zero, false);
+    EmitUndefinedBehaviorIntegerDivAndRemCheck(Ops, Zero, false);
   }
 
   if (Ops.Ty->hasUnsignedIntegerRepresentation())
@@ -2577,12 +2690,7 @@ static Value* tryEmitFMulAdd(const BinOpInfo &op,
          "Only fadd/fsub can be the root of an fmuladd.");
 
   // Check whether this op is marked as fusable.
-  if (!op.FPContractable)
-    return nullptr;
-
-  // Check whether -ffp-contract=on. (If -ffp-contract=off/fast, fusing is
-  // either disabled, or handled entirely by the LLVM backend).
-  if (CGF.CGM.getCodeGenOpts().getFPContractMode() != CodeGenOptions::FPC_On)
+  if (!op.FPFeatures.allowFPContractWithinStatement())
     return nullptr;
 
   // We have a potentially fusable op. Look for a mul on one of the operands.
@@ -2616,12 +2724,15 @@ Value *ScalarExprEmitter::EmitAdd(const BinOpInfo &op) {
         return Builder.CreateNSWAdd(op.LHS, op.RHS, "add");
       // Fall through.
     case LangOptions::SOB_Trapping:
+      if (CanElideOverflowCheck(CGF.getContext(), op))
+        return Builder.CreateNSWAdd(op.LHS, op.RHS, "add");
       return EmitOverflowCheckedBinOp(op);
     }
   }
 
   if (op.Ty->isUnsignedIntegerType() &&
-      CGF.SanOpts.has(SanitizerKind::UnsignedIntegerOverflow))
+      CGF.SanOpts.has(SanitizerKind::UnsignedIntegerOverflow) &&
+      !CanElideOverflowCheck(CGF.getContext(), op))
     return EmitOverflowCheckedBinOp(op);
 
   if (op.LHS->getType()->isFPOrFPVectorTy()) {
@@ -2629,7 +2740,8 @@ Value *ScalarExprEmitter::EmitAdd(const BinOpInfo &op) {
     if (Value *FMulAdd = tryEmitFMulAdd(op, CGF, Builder))
       return FMulAdd;
 
-    return Builder.CreateFAdd(op.LHS, op.RHS, "add");
+    Value *V = Builder.CreateFAdd(op.LHS, op.RHS, "add");
+    return propagateFMFlags(V, op);
   }
 
   return Builder.CreateAdd(op.LHS, op.RHS, "add");
@@ -2647,19 +2759,23 @@ Value *ScalarExprEmitter::EmitSub(const BinOpInfo &op) {
           return Builder.CreateNSWSub(op.LHS, op.RHS, "sub");
         // Fall through.
       case LangOptions::SOB_Trapping:
+        if (CanElideOverflowCheck(CGF.getContext(), op))
+          return Builder.CreateNSWSub(op.LHS, op.RHS, "sub");
         return EmitOverflowCheckedBinOp(op);
       }
     }
 
     if (op.Ty->isUnsignedIntegerType() &&
-        CGF.SanOpts.has(SanitizerKind::UnsignedIntegerOverflow))
+        CGF.SanOpts.has(SanitizerKind::UnsignedIntegerOverflow) &&
+        !CanElideOverflowCheck(CGF.getContext(), op))
       return EmitOverflowCheckedBinOp(op);
 
     if (op.LHS->getType()->isFPOrFPVectorTy()) {
       // Try to form an fmuladd.
       if (Value *FMulAdd = tryEmitFMulAdd(op, CGF, Builder, true))
         return FMulAdd;
-      return Builder.CreateFSub(op.LHS, op.RHS, "sub");
+      Value *V = Builder.CreateFSub(op.LHS, op.RHS, "sub");
+      return propagateFMFlags(V, op);
     }
 
     return Builder.CreateSub(op.LHS, op.RHS, "sub");
@@ -2751,8 +2867,8 @@ Value *ScalarExprEmitter::EmitShl(const BinOpInfo &Ops) {
            isa<llvm::IntegerType>(Ops.LHS->getType())) {
     CodeGenFunction::SanitizerScope SanScope(&CGF);
     SmallVector<std::pair<Value *, SanitizerMask>, 2> Checks;
-    llvm::Value *WidthMinusOne = GetWidthMinusOneValue(Ops.LHS, RHS);
-    llvm::Value *ValidExponent = Builder.CreateICmpULE(RHS, WidthMinusOne);
+    llvm::Value *WidthMinusOne = GetWidthMinusOneValue(Ops.LHS, Ops.RHS);
+    llvm::Value *ValidExponent = Builder.CreateICmpULE(Ops.RHS, WidthMinusOne);
 
     if (SanitizeExponent) {
       Checks.push_back(
@@ -2767,12 +2883,14 @@ Value *ScalarExprEmitter::EmitShl(const BinOpInfo &Ops) {
       llvm::BasicBlock *Cont = CGF.createBasicBlock("cont");
       llvm::BasicBlock *CheckShiftBase = CGF.createBasicBlock("check");
       Builder.CreateCondBr(ValidExponent, CheckShiftBase, Cont);
+      llvm::Value *PromotedWidthMinusOne =
+          (RHS == Ops.RHS) ? WidthMinusOne
+                           : GetWidthMinusOneValue(Ops.LHS, RHS);
       CGF.EmitBlock(CheckShiftBase);
-      llvm::Value *BitsShiftedOff =
-        Builder.CreateLShr(Ops.LHS,
-                           Builder.CreateSub(WidthMinusOne, RHS, "shl.zeros",
-                                             /*NUW*/true, /*NSW*/true),
-                           "shl.check");
+      llvm::Value *BitsShiftedOff = Builder.CreateLShr(
+          Ops.LHS, Builder.CreateSub(PromotedWidthMinusOne, RHS, "shl.zeros",
+                                     /*NUW*/ true, /*NSW*/ true),
+          "shl.check");
       if (CGF.getLangOpts().CPlusPlus) {
         // In C99, we are not permitted to shift a 1 bit into the sign bit.
         // Under C++11's rules, shifting a 1 bit into the sign bit is
@@ -3038,10 +3156,12 @@ Value *ScalarExprEmitter::VisitBinAssign(const BinaryOperator *E) {
     // because the result is altered by the store, i.e., [C99 6.5.16p1]
     // 'An assignment expression has the value of the left operand after
     // the assignment...'.
-    if (LHS.isBitField())
+    if (LHS.isBitField()) {
       CGF.EmitStoreThroughBitfieldLValue(RValue::get(RHS), LHS, &RHS);
-    else
+    } else {
+      CGF.EmitNullabilityCheck(LHS, RHS, E->getExprLoc());
       CGF.EmitStoreThroughLValue(RValue::get(RHS), LHS);
+    }
   }
 
   // If the result is clearly ignored, return now.
@@ -3327,9 +3447,11 @@ VisitAbstractConditionalOperator(const AbstractConditionalOperator *E) {
   // safe to evaluate the LHS and RHS unconditionally.
   if (isCheapEnoughToEvaluateUnconditionally(lhsExpr, CGF) &&
       isCheapEnoughToEvaluateUnconditionally(rhsExpr, CGF)) {
-    CGF.incrementProfileCounter(E);
-
     llvm::Value *CondV = CGF.EvaluateExprAsBool(condExpr);
+    llvm::Value *StepV = Builder.CreateZExtOrBitCast(CondV, CGF.Int64Ty);
+
+    CGF.incrementProfileCounter(E, StepV);
+
     llvm::Value *LHS = Visit(lhsExpr);
     llvm::Value *RHS = Visit(rhsExpr);
     if (!LHS) {
@@ -3491,8 +3613,12 @@ Value *ScalarExprEmitter::VisitAsTypeExpr(AsTypeExpr *E) {
   // vector to get a vec4, then a bitcast if the target type is different.
   if (NumElementsSrc == 3 && NumElementsDst != 3) {
     Src = ConvertVec3AndVec4(Builder, CGF, Src, 4);
-    Src = createCastsForTypeOfSameSize(Builder, CGF.CGM.getDataLayout(), Src,
-                                       DstTy);
+
+    if (!CGF.CGM.getCodeGenOpts().PreserveVec3Type) {
+      Src = createCastsForTypeOfSameSize(Builder, CGF.CGM.getDataLayout(), Src,
+                                         DstTy);
+    }
+
     Src->setName("astype");
     return Src;
   }
@@ -3501,9 +3627,12 @@ Value *ScalarExprEmitter::VisitAsTypeExpr(AsTypeExpr *E) {
   // to vec4 if the original type is not vec4, then a shuffle vector to
   // get a vec3.
   if (NumElementsSrc != 3 && NumElementsDst == 3) {
-    auto Vec4Ty = llvm::VectorType::get(DstTy->getVectorElementType(), 4);
-    Src = createCastsForTypeOfSameSize(Builder, CGF.CGM.getDataLayout(), Src,
-                                       Vec4Ty);
+    if (!CGF.CGM.getCodeGenOpts().PreserveVec3Type) {
+      auto Vec4Ty = llvm::VectorType::get(DstTy->getVectorElementType(), 4);
+      Src = createCastsForTypeOfSameSize(Builder, CGF.CGM.getDataLayout(), Src,
+                                         Vec4Ty);
+    }
+
     Src = ConvertVec3AndVec4(Builder, CGF, Src, 3);
     Src->setName("astype");
     return Src;
diff --git a/lib/CodeGen/CGCUDABuiltin.cpp b/lib/CodeGen/CGGPUBuiltin.cpp
index 44dd003757ad..48156b1b26b7 100644
--- a/lib/CodeGen/CGCUDABuiltin.cpp
+++ b/lib/CodeGen/CGGPUBuiltin.cpp
@@ -1,4 +1,4 @@
-//===----- CGCUDABuiltin.cpp - Codegen for CUDA builtins ------------------===//
+//===------ CGGPUBuiltin.cpp - Codegen for GPU builtins -------------------===//
 //
 //                     The LLVM Compiler Infrastructure
 //
@@ -7,8 +7,8 @@
 //
 //===----------------------------------------------------------------------===//
 //
-// Generates code for built-in CUDA calls which are not runtime-specific.
-// (Runtime-specific codegen lives in CGCUDARuntime.)
+// Generates code for built-in GPU calls which are not runtime-specific.
+// (Runtime-specific codegen lives in programming model specific files.)
 //
 //===----------------------------------------------------------------------===//
 
@@ -67,10 +67,9 @@ static llvm::Function *GetVprintfDeclaration(llvm::Module &M) {
 // Note that by the time this function runs, E's args have already undergone the
 // standard C vararg promotion (short -> int, float -> double, etc.).
 RValue
-CodeGenFunction::EmitCUDADevicePrintfCallExpr(const CallExpr *E,
-                                              ReturnValueSlot ReturnValue) {
-  assert(getLangOpts().CUDA);
-  assert(getLangOpts().CUDAIsDevice);
+CodeGenFunction::EmitNVPTXDevicePrintfCallExpr(const CallExpr *E,
+                                               ReturnValueSlot ReturnValue) {
+  assert(getTarget().getTriple().isNVPTX());
   assert(E->getBuiltinCallee() == Builtin::BIprintf);
   assert(E->getNumArgs() >= 1); // printf always has at least one arg.
 
diff --git a/lib/CodeGen/CGObjC.cpp b/lib/CodeGen/CGObjC.cpp
index 932b8a129e6e..3a09a15dbc15 100644
--- a/lib/CodeGen/CGObjC.cpp
+++ b/lib/CodeGen/CGObjC.cpp
@@ -1,4 +1,4 @@
-//===---- CGBuiltin.cpp - Emit LLVM Code for builtins ---------------------===//
+//===---- CGObjC.cpp - Emit LLVM Code for Objective-C ---------------------===//
 //
 //                     The LLVM Compiler Infrastructure
 //
@@ -117,10 +117,22 @@ llvm::Value *CodeGenFunction::EmitObjCCollectionLiteral(const Expr *E,
   const ObjCArrayLiteral *ALE = dyn_cast<ObjCArrayLiteral>(E);
   if (!ALE)
     DLE = cast<ObjCDictionaryLiteral>(E);
-  
-  // Compute the type of the array we're initializing.
+
+  // Optimize empty collections by referencing constants, when available.
   uint64_t NumElements = 
     ALE ? ALE->getNumElements() : DLE->getNumElements();
+  if (NumElements == 0 && CGM.getLangOpts().ObjCRuntime.hasEmptyCollections()) {
+    StringRef ConstantName = ALE ? "__NSArray0__" : "__NSDictionary0__";
+    QualType IdTy(CGM.getContext().getObjCIdType());
+    llvm::Constant *Constant =
+        CGM.CreateRuntimeVariable(ConvertType(IdTy), ConstantName);
+    Address Addr(Constant, Context.getTypeAlignInChars(IdTy));
+    LValue LV = MakeAddrLValue(Addr, IdTy);
+    return Builder.CreateBitCast(EmitLoadOfScalar(LV, E->getLocStart()),
+                                 ConvertType(E->getType()));
+  }
+
+  // Compute the type of the array we're initializing.
   llvm::APInt APNumElements(Context.getTypeSize(Context.getSizeType()),
                             NumElements);
   QualType ElementType = Context.getObjCIdType().withConst();
@@ -427,7 +439,7 @@ RValue CodeGenFunction::EmitObjCMessageExpr(const ObjCMessageExpr *E,
   QualType ResultType = method ? method->getReturnType() : E->getType();
 
   CallArgList Args;
-  EmitCallArgs(Args, method, E->arguments());
+  EmitCallArgs(Args, method, E->arguments(), /*AC*/AbstractCallee(method));
 
   // For delegate init calls in ARC, do an unsafe store of null into
   // self.  This represents the call taking direct ownership of that
@@ -1316,7 +1328,7 @@ CodeGenFunction::generateObjCSetterBody(const ObjCImplementationDecl *classImpl,
 
   BinaryOperator assign(&ivarRef, finalArg, BO_Assign,
                         ivarRef.getType(), VK_RValue, OK_Ordinary,
-                        SourceLocation(), false);
+                        SourceLocation(), FPOptions());
   EmitStmt(&assign);
 }
 
@@ -1469,6 +1481,8 @@ void CodeGenFunction::EmitObjCForCollectionStmt(const ObjCForCollectionStmt &S){
   if (DI)
     DI->EmitLexicalBlockStart(Builder, S.getSourceRange().getBegin());
 
+  RunCleanupsScope ForScope(*this);
+
   // The local variable comes into scope immediately.
   AutoVarEmission variable = AutoVarEmission::invalid();
   if (const DeclStmt *SD = dyn_cast<DeclStmt>(S.getElement()))
@@ -1499,8 +1513,6 @@ void CodeGenFunction::EmitObjCForCollectionStmt(const ObjCForCollectionStmt &S){
                                       ArrayType::Normal, 0);
   Address ItemsPtr = CreateMemTemp(ItemsTy, "items.ptr");
 
-  RunCleanupsScope ForScope(*this);
-
   // Emit the collection pointer.  In ARC, we do a retain.
   llvm::Value *Collection;
   if (getLangOpts().ObjCAutoRefCount) {
@@ -1802,26 +1814,49 @@ void CodeGenFunction::EmitARCIntrinsicUse(ArrayRef<llvm::Value*> values) {
 }
 
 
+static bool IsForwarding(StringRef Name) {
+  return llvm::StringSwitch<bool>(Name)
+      .Cases("objc_autoreleaseReturnValue",             // ARCInstKind::AutoreleaseRV
+             "objc_autorelease",                        // ARCInstKind::Autorelease
+             "objc_retainAutoreleaseReturnValue",       // ARCInstKind::FusedRetainAutoreleaseRV
+             "objc_retainAutoreleasedReturnValue",      // ARCInstKind::RetainRV
+             "objc_retainAutorelease",                  // ARCInstKind::FusedRetainAutorelease
+             "objc_retainedObject",                     // ARCInstKind::NoopCast
+             "objc_retain",                             // ARCInstKind::Retain
+             "objc_unretainedObject",                   // ARCInstKind::NoopCast
+             "objc_unretainedPointer",                  // ARCInstKind::NoopCast
+             "objc_unsafeClaimAutoreleasedReturnValue", // ARCInstKind::ClaimRV
+             true)
+      .Default(false);
+}
+
 static llvm::Constant *createARCRuntimeFunction(CodeGenModule &CGM,
-                                                llvm::FunctionType *type,
-                                                StringRef fnName) {
-  llvm::Constant *fn = CGM.CreateRuntimeFunction(type, fnName);
+                                                llvm::FunctionType *FTy,
+                                                StringRef Name) {
+  llvm::Constant *RTF = CGM.CreateRuntimeFunction(FTy, Name);
 
-  if (llvm::Function *f = dyn_cast<llvm::Function>(fn)) {
+  if (auto *F = dyn_cast<llvm::Function>(RTF)) {
     // If the target runtime doesn't naturally support ARC, emit weak
     // references to the runtime support library.  We don't really
     // permit this to fail, but we need a particular relocation style.
     if (!CGM.getLangOpts().ObjCRuntime.hasNativeARC() &&
         !CGM.getTriple().isOSBinFormatCOFF()) {
-      f->setLinkage(llvm::Function::ExternalWeakLinkage);
-    } else if (fnName == "objc_retain" || fnName  == "objc_release") {
+      F->setLinkage(llvm::Function::ExternalWeakLinkage);
+    } else if (Name == "objc_retain" || Name  == "objc_release") {
       // If we have Native ARC, set nonlazybind attribute for these APIs for
       // performance.
-      f->addFnAttr(llvm::Attribute::NonLazyBind);
+      F->addFnAttr(llvm::Attribute::NonLazyBind);
+    }
+
+    if (IsForwarding(Name)) {
+      llvm::AttrBuilder B;
+      B.addAttribute(llvm::Attribute::Returned);
+
+      F->arg_begin()->addAttr(llvm::AttributeList::get(F->getContext(), 1, B));
     }
   }
 
-  return fn;
+  return RTF;
 }
 
 /// Perform an operation having the signature
@@ -1832,7 +1867,8 @@ static llvm::Value *emitARCValueOperation(CodeGenFunction &CGF,
                                           llvm::Constant *&fn,
                                           StringRef fnName,
                                           bool isTailCall = false) {
-  if (isa<llvm::ConstantPointerNull>(value)) return value;
+  if (isa<llvm::ConstantPointerNull>(value))
+    return value;
 
   if (!fn) {
     llvm::FunctionType *fnType =
@@ -3239,7 +3275,7 @@ CodeGenFunction::GenerateObjCAtomicSetterCopyHelperFunction(
   CallExpr *CalleeExp = cast<CallExpr>(PID->getSetterCXXAssignment());
   CXXOperatorCallExpr TheCall(C, OO_Equal, CalleeExp->getCallee(),
                               Args, DestTy->g