Vendor import of clang release_90 branch r369369:vendor/clang/clang-release_90-r369369

https://llvm.org/svn/llvm-project/cfe/branches/release_90@369369

41 files changed, 831 insertions, 468 deletions

diff --git a/include/clang/AST/ExprCXX.h b/include/clang/AST/ExprCXX.h
index 28ed6cdfde14..faddbc6d9675 100644
--- a/include/clang/AST/ExprCXX.h
+++ b/include/clang/AST/ExprCXX.h
@@ -185,15 +185,20 @@ public:
   static CXXMemberCallExpr *CreateEmpty(const ASTContext &Ctx, unsigned NumArgs,
                                         EmptyShell Empty);
 
-  /// Retrieves the implicit object argument for the member call.
+  /// Retrieve the implicit object argument for the member call.
   ///
   /// For example, in "x.f(5)", this returns the sub-expression "x".
   Expr *getImplicitObjectArgument() const;
 
-  /// Retrieves the declaration of the called method.
+  /// Retrieve the type of the object argument.
+  ///
+  /// Note that this always returns a non-pointer type.
+  QualType getObjectType() const;
+
+  /// Retrieve the declaration of the called method.
   CXXMethodDecl *getMethodDecl() const;
 
-  /// Retrieves the CXXRecordDecl for the underlying type of
+  /// Retrieve the CXXRecordDecl for the underlying type of
   /// the implicit object argument.
   ///
   /// Note that this is may not be the same declaration as that of the class
diff --git a/include/clang/Basic/DiagnosticSemaKinds.td b/include/clang/Basic/DiagnosticSemaKinds.td
index effcbad78b23..275c4e4365d1 100644
--- a/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/include/clang/Basic/DiagnosticSemaKinds.td
@@ -598,6 +598,10 @@ def ext_implicit_lib_function_decl : ExtWarn<
 def note_include_header_or_declare : Note<
   "include the header <%0> or explicitly provide a declaration for '%1'">;
 def note_previous_builtin_declaration : Note<"%0 is a builtin with type %1">;
+def warn_implicit_decl_no_jmp_buf
+    : Warning<"declaration of built-in function '%0' requires the declaration"
+    " of the 'jmp_buf' type, commonly provided in the header <setjmp.h>.">,
+      InGroup<DiagGroup<"incomplete-setjmp-declaration">>;
 def warn_implicit_decl_requires_sysheader : Warning<
   "declaration of built-in function '%1' requires inclusion of the header <%0>">,
   InGroup<BuiltinRequiresHeader>;
diff --git a/include/clang/Basic/TargetInfo.h b/include/clang/Basic/TargetInfo.h
index 7a8384f5fbc0..c6c966dfbe2c 100644
--- a/include/clang/Basic/TargetInfo.h
+++ b/include/clang/Basic/TargetInfo.h
@@ -1249,15 +1249,9 @@ public:
   bool isBigEndian() const { return BigEndian; }
   bool isLittleEndian() const { return !BigEndian; }
 
-  enum CallingConvMethodType {
-    CCMT_Unknown,
-    CCMT_Member,
-    CCMT_NonMember
-  };
-
   /// Gets the default calling convention for the given target and
   /// declaration context.
-  virtual CallingConv getDefaultCallingConv(CallingConvMethodType MT) const {
+  virtual CallingConv getDefaultCallingConv() const {
     // Not all targets will specify an explicit calling convention that we can
     // express.  This will always do the right thing, even though it's not
     // an explicit calling convention.
diff --git a/include/clang/Driver/Options.td b/include/clang/Driver/Options.td
index dfd27fab796e..4ea8bfff0973 100644
--- a/include/clang/Driver/Options.td
+++ b/include/clang/Driver/Options.td
@@ -518,7 +518,7 @@ def cl_mad_enable : Flag<["-"], "cl-mad-enable">, Group<opencl_Group>, Flags<[CC
 def cl_no_signed_zeros : Flag<["-"], "cl-no-signed-zeros">, Group<opencl_Group>, Flags<[CC1Option]>,
   HelpText<"OpenCL only. Allow use of less precise no signed zeros computations in the generated binary.">;
 def cl_std_EQ : Joined<["-"], "cl-std=">, Group<opencl_Group>, Flags<[CC1Option]>,
-  HelpText<"OpenCL language standard to compile for.">, Values<"cl,CL,cl1.1,CL1.1,cl1.2,CL1.2,cl2.0,CL2.0,c++">;
+  HelpText<"OpenCL language standard to compile for.">, Values<"cl,CL,cl1.1,CL1.1,cl1.2,CL1.2,cl2.0,CL2.0,clc++,CLC++">;
 def cl_denorms_are_zero : Flag<["-"], "cl-denorms-are-zero">, Group<opencl_Group>, Flags<[CC1Option]>,
   HelpText<"OpenCL only. Allow denormals to be flushed to zero.">;
 def cl_fp32_correctly_rounded_divide_sqrt : Flag<["-"], "cl-fp32-correctly-rounded-divide-sqrt">, Group<opencl_Group>, Flags<[CC1Option]>,
diff --git a/include/clang/Frontend/LangStandards.def b/include/clang/Frontend/LangStandards.def
index 0964e9b90a03..72ea23562ebd 100644
--- a/include/clang/Frontend/LangStandards.def
+++ b/include/clang/Frontend/LangStandards.def
@@ -174,6 +174,7 @@ LANGSTANDARD_ALIAS_DEPR(opencl10, "CL")
 LANGSTANDARD_ALIAS_DEPR(opencl11, "CL1.1")
 LANGSTANDARD_ALIAS_DEPR(opencl12, "CL1.2")
 LANGSTANDARD_ALIAS_DEPR(opencl20, "CL2.0")
+LANGSTANDARD_ALIAS_DEPR(openclcpp, "CLC++")
 
 // CUDA
 LANGSTANDARD(cuda, "cuda", CUDA, "NVIDIA CUDA(tm)",
diff --git a/include/clang/Sema/Sema.h b/include/clang/Sema/Sema.h
index af762f74d745..e6c63fd9c015 100644
--- a/include/clang/Sema/Sema.h
+++ b/include/clang/Sema/Sema.h
@@ -11165,6 +11165,7 @@ public:
   // Emitting members of dllexported classes is delayed until the class
   // (including field initializers) is fully parsed.
   SmallVector<CXXRecordDecl*, 4> DelayedDllExportClasses;
+  SmallVector<CXXMethodDecl*, 4> DelayedDllExportMemberFunctions;
 
 private:
   class SavePendingParsedClassStateRAII {
diff --git a/lib/AST/ASTContext.cpp b/lib/AST/ASTContext.cpp
index 0d69eb90abaf..468c7f47657d 100644
--- a/lib/AST/ASTContext.cpp
+++ b/lib/AST/ASTContext.cpp
@@ -10035,7 +10035,7 @@ CallingConv ASTContext::getDefaultCallingConvention(bool IsVariadic,
       break;
     }
   }
-  return Target->getDefaultCallingConv(TargetInfo::CCMT_Unknown);
+  return Target->getDefaultCallingConv();
 }
 
 bool ASTContext::isNearlyEmpty(const CXXRecordDecl *RD) const {
diff --git a/lib/AST/ExprCXX.cpp b/lib/AST/ExprCXX.cpp
index b30f785ba8f5..c5f86a4cc12b 100644
--- a/lib/AST/ExprCXX.cpp
+++ b/lib/AST/ExprCXX.cpp
@@ -651,6 +651,13 @@ Expr *CXXMemberCallExpr::getImplicitObjectArgument() const {
   return nullptr;
 }
 
+QualType CXXMemberCallExpr::getObjectType() const {
+  QualType Ty = getImplicitObjectArgument()->getType();
+  if (Ty->isPointerType())
+    Ty = Ty->getPointeeType();
+  return Ty;
+}
+
 CXXMethodDecl *CXXMemberCallExpr::getMethodDecl() const {
   if (const auto *MemExpr = dyn_cast<MemberExpr>(getCallee()->IgnoreParens()))
     return cast<CXXMethodDecl>(MemExpr->getMemberDecl());
diff --git a/lib/AST/ItaniumCXXABI.cpp b/lib/AST/ItaniumCXXABI.cpp
index 727a905d08a1..77fb5a1d33b3 100644
--- a/lib/AST/ItaniumCXXABI.cpp
+++ b/lib/AST/ItaniumCXXABI.cpp
@@ -177,7 +177,7 @@ public:
     if (!isVariadic && T.isWindowsGNUEnvironment() &&
         T.getArch() == llvm::Triple::x86)
       return CC_X86ThisCall;
-    return CC_C;
+    return Context.getTargetInfo().getDefaultCallingConv();
   }
 
   // We cheat and just check that the class has a vtable pointer, and that it's
diff --git a/lib/AST/MicrosoftCXXABI.cpp b/lib/AST/MicrosoftCXXABI.cpp
index 4dc4156df9ca..444e55f777fa 100644
--- a/lib/AST/MicrosoftCXXABI.cpp
+++ b/lib/AST/MicrosoftCXXABI.cpp
@@ -82,7 +82,7 @@ public:
     if (!isVariadic &&
         Context.getTargetInfo().getTriple().getArch() == llvm::Triple::x86)
       return CC_X86ThisCall;
-    return CC_C;
+    return Context.getTargetInfo().getDefaultCallingConv();
   }
 
   bool isNearlyEmpty(const CXXRecordDecl *RD) const override {
diff --git a/lib/Basic/Targets/AArch64.cpp b/lib/Basic/Targets/AArch64.cpp
index 74ac69ab8946..25f2b7b35f41 100644
--- a/lib/Basic/Targets/AArch64.cpp
+++ b/lib/Basic/Targets/AArch64.cpp
@@ -196,9 +196,6 @@ void AArch64TargetInfo::getTargetDefines(const LangOptions &Opts,
     Builder.defineMacro("__ARM_NEON_FP", "0xE");
   }
 
-  if (FPU & SveMode)
-    Builder.defineMacro("__ARM_FEATURE_SVE", "1");
-
   if (HasCRC)
     Builder.defineMacro("__ARM_FEATURE_CRC32", "1");
 
@@ -351,10 +348,19 @@ const char *const AArch64TargetInfo::GCCRegNames[] = {
     "d12", "d13", "d14", "d15", "d16", "d17", "d18", "d19", "d20", "d21", "d22",
     "d23", "d24", "d25", "d26", "d27", "d28", "d29", "d30", "d31",
 
-    // Vector registers
+    // Neon vector registers
     "v0", "v1", "v2", "v3", "v4", "v5", "v6", "v7", "v8", "v9", "v10", "v11",
     "v12", "v13", "v14", "v15", "v16", "v17", "v18", "v19", "v20", "v21", "v22",
-    "v23", "v24", "v25", "v26", "v27", "v28", "v29", "v30", "v31"
+    "v23", "v24", "v25", "v26", "v27", "v28", "v29", "v30", "v31",
+
+    // SVE vector registers
+    "z0",  "z1",  "z2",  "z3",  "z4",  "z5",  "z6",  "z7",  "z8",  "z9",  "z10",
+    "z11", "z12", "z13", "z14", "z15", "z16", "z17", "z18", "z19", "z20", "z21",
+    "z22", "z23", "z24", "z25", "z26", "z27", "z28", "z29", "z30", "z31",
+
+    // SVE predicate registers
+    "p0",  "p1",  "p2",  "p3",  "p4",  "p5",  "p6",  "p7",  "p8",  "p9",  "p10",
+    "p11", "p12", "p13", "p14", "p15"
 };
 
 ArrayRef<const char *> AArch64TargetInfo::getGCCRegNames() const {
diff --git a/lib/Basic/Targets/OSTargets.h b/lib/Basic/Targets/OSTargets.h
index 8542311ffa41..c0373ffaa444 100644
--- a/lib/Basic/Targets/OSTargets.h
+++ b/lib/Basic/Targets/OSTargets.h
@@ -618,8 +618,11 @@ protected:
       Builder.defineMacro("_XOPEN_SOURCE", "600");
     else
       Builder.defineMacro("_XOPEN_SOURCE", "500");
-    if (Opts.CPlusPlus)
+    if (Opts.CPlusPlus) {
       Builder.defineMacro("__C99FEATURES__");
+      Builder.defineMacro("_FILE_OFFSET_BITS", "64");
+    }
+    // GCC restricts the next two to C++.
     Builder.defineMacro("_LARGEFILE_SOURCE");
     Builder.defineMacro("_LARGEFILE64_SOURCE");
     Builder.defineMacro("__EXTENSIONS__");
diff --git a/lib/Basic/Targets/RISCV.cpp b/lib/Basic/Targets/RISCV.cpp
index f800bb0b25da..939ac46d671b 100644
--- a/lib/Basic/Targets/RISCV.cpp
+++ b/lib/Basic/Targets/RISCV.cpp
@@ -56,6 +56,10 @@ bool RISCVTargetInfo::validateAsmConstraint(
     // A 5-bit unsigned immediate for CSR access instructions.
     Info.setRequiresImmediate(0, 31);
     return true;
+  case 'f':
+    // A floating-point register.
+    Info.setAllowsRegister();
+    return true;
   }
 }
 
@@ -65,9 +69,18 @@ void RISCVTargetInfo::getTargetDefines(const LangOptions &Opts,
   Builder.defineMacro("__riscv");
   bool Is64Bit = getTriple().getArch() == llvm::Triple::riscv64;
   Builder.defineMacro("__riscv_xlen", Is64Bit ? "64" : "32");
-  // TODO: modify when more code models and ABIs are supported.
+  // TODO: modify when more code models are supported.
   Builder.defineMacro("__riscv_cmodel_medlow");
-  Builder.defineMacro("__riscv_float_abi_soft");
+
+  StringRef ABIName = getABI();
+  if (ABIName == "ilp32f" || ABIName == "lp64f")
+    Builder.defineMacro("__riscv_float_abi_single");
+  else if (ABIName == "ilp32d" || ABIName == "lp64d")
+    Builder.defineMacro("__riscv_float_abi_double");
+  else if (ABIName == "ilp32e")
+    Builder.defineMacro("__riscv_abi_rve");
+  else
+    Builder.defineMacro("__riscv_float_abi_soft");
 
   if (HasM) {
     Builder.defineMacro("__riscv_mul");
diff --git a/lib/Basic/Targets/RISCV.h b/lib/Basic/Targets/RISCV.h
index bc814b79ce51..ce193feaeb98 100644
--- a/lib/Basic/Targets/RISCV.h
+++ b/lib/Basic/Targets/RISCV.h
@@ -87,8 +87,7 @@ public:
   }
 
   bool setABI(const std::string &Name) override {
-    // TODO: support ilp32f and ilp32d ABIs.
-    if (Name == "ilp32") {
+    if (Name == "ilp32" || Name == "ilp32f" || Name == "ilp32d") {
       ABI = Name;
       return true;
     }
@@ -105,8 +104,7 @@ public:
   }
 
   bool setABI(const std::string &Name) override {
-    // TODO: support lp64f and lp64d ABIs.
-    if (Name == "lp64") {
+    if (Name == "lp64" || Name == "lp64f" || Name == "lp64d") {
       ABI = Name;
       return true;
     }
diff --git a/lib/Basic/Targets/SPIR.h b/lib/Basic/Targets/SPIR.h
index 6023c868dbdc..802ccf8b671e 100644
--- a/lib/Basic/Targets/SPIR.h
+++ b/lib/Basic/Targets/SPIR.h
@@ -88,7 +88,7 @@ public:
                                                             : CCCR_Warning;
   }
 
-  CallingConv getDefaultCallingConv(CallingConvMethodType MT) const override {
+  CallingConv getDefaultCallingConv() const override {
     return CC_SpirFunction;
   }
 
diff --git a/lib/Basic/Targets/X86.h b/lib/Basic/Targets/X86.h
index 588b6d3da1d6..dd1e7db6c81e 100644
--- a/lib/Basic/Targets/X86.h
+++ b/lib/Basic/Targets/X86.h
@@ -320,8 +320,8 @@ public:
     }
   }
 
-  CallingConv getDefaultCallingConv(CallingConvMethodType MT) const override {
-    return MT == CCMT_Member ? CC_X86ThisCall : CC_C;
+  CallingConv getDefaultCallingConv() const override {
+    return CC_C;
   }
 
   bool hasSjLjLowering() const override { return true; }
@@ -659,7 +659,7 @@ public:
     }
   }
 
-  CallingConv getDefaultCallingConv(CallingConvMethodType MT) const override {
+  CallingConv getDefaultCallingConv() const override {
     return CC_C;
   }
 
diff --git a/lib/Basic/Version.cpp b/lib/Basic/Version.cpp
index 3006ca33f213..9bf8687a8765 100644
--- a/lib/Basic/Version.cpp
+++ b/lib/Basic/Version.cpp
@@ -35,7 +35,7 @@ std::string getClangRepositoryPath() {
 
   // If the CLANG_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/trunk/lib/Basic/Version.cpp $");
+  StringRef SVNRepository("$URL: https://llvm.org/svn/llvm-project/cfe/branches/release_90/lib/Basic/Version.cpp $");
   if (URL.empty()) {
     URL = SVNRepository.slice(SVNRepository.find(':'),
                               SVNRepository.find("/lib/Basic"));
diff --git a/lib/CodeGen/CGBuiltin.cpp b/lib/CodeGen/CGBuiltin.cpp
index a300bab49f9c..cadce507412b 100644
--- a/lib/CodeGen/CGBuiltin.cpp
+++ b/lib/CodeGen/CGBuiltin.cpp
@@ -8011,6 +8011,151 @@ Value *CodeGenFunction::EmitAArch64BuiltinExpr(unsigned BuiltinID,
     return Builder.CreateExtractElement(Ops[0], EmitScalarExpr(E->getArg(1)),
                                         "vgetq_lane");
   }
+  case AArch64::BI_BitScanForward:
+  case AArch64::BI_BitScanForward64:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_BitScanForward, E);
+  case AArch64::BI_BitScanReverse:
+  case AArch64::BI_BitScanReverse64:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_BitScanReverse, E);
+  case AArch64::BI_InterlockedAnd64:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedAnd, E);
+  case AArch64::BI_InterlockedExchange64:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchange, E);
+  case AArch64::BI_InterlockedExchangeAdd64:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeAdd, E);
+  case AArch64::BI_InterlockedExchangeSub64:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeSub, E);
+  case AArch64::BI_InterlockedOr64:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr, E);
+  case AArch64::BI_InterlockedXor64:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedXor, E);
+  case AArch64::BI_InterlockedDecrement64:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedDecrement, E);
+  case AArch64::BI_InterlockedIncrement64:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedIncrement, E);
+  case AArch64::BI_InterlockedExchangeAdd8_acq:
+  case AArch64::BI_InterlockedExchangeAdd16_acq:
+  case AArch64::BI_InterlockedExchangeAdd_acq:
+  case AArch64::BI_InterlockedExchangeAdd64_acq:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeAdd_acq, E);
+  case AArch64::BI_InterlockedExchangeAdd8_rel:
+  case AArch64::BI_InterlockedExchangeAdd16_rel:
+  case AArch64::BI_InterlockedExchangeAdd_rel:
+  case AArch64::BI_InterlockedExchangeAdd64_rel:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeAdd_rel, E);
+  case AArch64::BI_InterlockedExchangeAdd8_nf:
+  case AArch64::BI_InterlockedExchangeAdd16_nf:
+  case AArch64::BI_InterlockedExchangeAdd_nf:
+  case AArch64::BI_InterlockedExchangeAdd64_nf:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeAdd_nf, E);
+  case AArch64::BI_InterlockedExchange8_acq:
+  case AArch64::BI_InterlockedExchange16_acq:
+  case AArch64::BI_InterlockedExchange_acq:
+  case AArch64::BI_InterlockedExchange64_acq:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchange_acq, E);
+  case AArch64::BI_InterlockedExchange8_rel:
+  case AArch64::BI_InterlockedExchange16_rel:
+  case AArch64::BI_InterlockedExchange_rel:
+  case AArch64::BI_InterlockedExchange64_rel:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchange_rel, E);
+  case AArch64::BI_InterlockedExchange8_nf:
+  case AArch64::BI_InterlockedExchange16_nf:
+  case AArch64::BI_InterlockedExchange_nf:
+  case AArch64::BI_InterlockedExchange64_nf:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchange_nf, E);
+  case AArch64::BI_InterlockedCompareExchange8_acq:
+  case AArch64::BI_InterlockedCompareExchange16_acq:
+  case AArch64::BI_InterlockedCompareExchange_acq:
+  case AArch64::BI_InterlockedCompareExchange64_acq:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedCompareExchange_acq, E);
+  case AArch64::BI_InterlockedCompareExchange8_rel:
+  case AArch64::BI_InterlockedCompareExchange16_rel:
+  case AArch64::BI_InterlockedCompareExchange_rel:
+  case AArch64::BI_InterlockedCompareExchange64_rel:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedCompareExchange_rel, E);
+  case AArch64::BI_InterlockedCompareExchange8_nf:
+  case AArch64::BI_InterlockedCompareExchange16_nf:
+  case AArch64::BI_InterlockedCompareExchange_nf:
+  case AArch64::BI_InterlockedCompareExchange64_nf:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedCompareExchange_nf, E);
+  case AArch64::BI_InterlockedOr8_acq:
+  case AArch64::BI_InterlockedOr16_acq:
+  case AArch64::BI_InterlockedOr_acq:
+  case AArch64::BI_InterlockedOr64_acq:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_acq, E);
+  case AArch64::BI_InterlockedOr8_rel:
+  case AArch64::BI_InterlockedOr16_rel:
+  case AArch64::BI_InterlockedOr_rel:
+  case AArch64::BI_InterlockedOr64_rel:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_rel, E);
+  case AArch64::BI_InterlockedOr8_nf:
+  case AArch64::BI_InterlockedOr16_nf:
+  case AArch64::BI_InterlockedOr_nf:
+  case AArch64::BI_InterlockedOr64_nf:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_nf, E);
+  case AArch64::BI_InterlockedXor8_acq:
+  case AArch64::BI_InterlockedXor16_acq:
+  case AArch64::BI_InterlockedXor_acq:
+  case AArch64::BI_InterlockedXor64_acq:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedXor_acq, E);
+  case AArch64::BI_InterlockedXor8_rel:
+  case AArch64::BI_InterlockedXor16_rel:
+  case AArch64::BI_InterlockedXor_rel:
+  case AArch64::BI_InterlockedXor64_rel:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedXor_rel, E);
+  case AArch64::BI_InterlockedXor8_nf:
+  case AArch64::BI_InterlockedXor16_nf:
+  case AArch64::BI_InterlockedXor_nf:
+  case AArch64::BI_InterlockedXor64_nf:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedXor_nf, E);
+  case AArch64::BI_InterlockedAnd8_acq:
+  case AArch64::BI_InterlockedAnd16_acq:
+  case AArch64::BI_InterlockedAnd_acq:
+  case AArch64::BI_InterlockedAnd64_acq:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedAnd_acq, E);
+  case AArch64::BI_InterlockedAnd8_rel:
+  case AArch64::BI_InterlockedAnd16_rel:
+  case AArch64::BI_InterlockedAnd_rel:
+  case AArch64::BI_InterlockedAnd64_rel:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedAnd_rel, E);
+  case AArch64::BI_InterlockedAnd8_nf:
+  case AArch64::BI_InterlockedAnd16_nf:
+  case AArch64::BI_InterlockedAnd_nf:
+  case AArch64::BI_InterlockedAnd64_nf:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedAnd_nf, E);
+  case AArch64::BI_InterlockedIncrement16_acq:
+  case AArch64::BI_InterlockedIncrement_acq:
+  case AArch64::BI_InterlockedIncrement64_acq:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedIncrement_acq, E);
+  case AArch64::BI_InterlockedIncrement16_rel:
+  case AArch64::BI_InterlockedIncrement_rel:
+  case AArch64::BI_InterlockedIncrement64_rel:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedIncrement_rel, E);
+  case AArch64::BI_InterlockedIncrement16_nf:
+  case AArch64::BI_InterlockedIncrement_nf:
+  case AArch64::BI_InterlockedIncrement64_nf:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedIncrement_nf, E);
+  case AArch64::BI_InterlockedDecrement16_acq:
+  case AArch64::BI_InterlockedDecrement_acq:
+  case AArch64::BI_InterlockedDecrement64_acq:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedDecrement_acq, E);
+  case AArch64::BI_InterlockedDecrement16_rel:
+  case AArch64::BI_InterlockedDecrement_rel:
+  case AArch64::BI_InterlockedDecrement64_rel:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedDecrement_rel, E);
+  case AArch64::BI_InterlockedDecrement16_nf:
+  case AArch64::BI_InterlockedDecrement_nf:
+  case AArch64::BI_InterlockedDecrement64_nf:
+    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedDecrement_nf, E);
+
+  case AArch64::BI_InterlockedAdd: {
+    Value *Arg0 = EmitScalarExpr(E->getArg(0));
+    Value *Arg1 = EmitScalarExpr(E->getArg(1));
+    AtomicRMWInst *RMWI = Builder.CreateAtomicRMW(
+      AtomicRMWInst::Add, Arg0, Arg1,
+      llvm::AtomicOrdering::SequentiallyConsistent);
+    return Builder.CreateAdd(RMWI, Arg1);
+  }
   }
 
   llvm::VectorType *VTy = GetNeonType(this, Type);
@@ -9128,151 +9273,6 @@ Value *CodeGenFunction::EmitAArch64BuiltinExpr(unsigned BuiltinID,
     Int = Intrinsic::aarch64_neon_suqadd;
     return EmitNeonCall(CGM.getIntrinsic(Int, Ty), Ops, "vuqadd");
   }
-  case AArch64::BI_BitScanForward:
-  case AArch64::BI_BitScanForward64:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_BitScanForward, E);
-  case AArch64::BI_BitScanReverse:
-  case AArch64::BI_BitScanReverse64:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_BitScanReverse, E);
-  case AArch64::BI_InterlockedAnd64:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedAnd, E);
-  case AArch64::BI_InterlockedExchange64:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchange, E);
-  case AArch64::BI_InterlockedExchangeAdd64:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeAdd, E);
-  case AArch64::BI_InterlockedExchangeSub64:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeSub, E);
-  case AArch64::BI_InterlockedOr64:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr, E);
-  case AArch64::BI_InterlockedXor64:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedXor, E);
-  case AArch64::BI_InterlockedDecrement64:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedDecrement, E);
-  case AArch64::BI_InterlockedIncrement64:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedIncrement, E);
-  case AArch64::BI_InterlockedExchangeAdd8_acq:
-  case AArch64::BI_InterlockedExchangeAdd16_acq:
-  case AArch64::BI_InterlockedExchangeAdd_acq:
-  case AArch64::BI_InterlockedExchangeAdd64_acq:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeAdd_acq, E);
-  case AArch64::BI_InterlockedExchangeAdd8_rel:
-  case AArch64::BI_InterlockedExchangeAdd16_rel:
-  case AArch64::BI_InterlockedExchangeAdd_rel:
-  case AArch64::BI_InterlockedExchangeAdd64_rel:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeAdd_rel, E);
-  case AArch64::BI_InterlockedExchangeAdd8_nf:
-  case AArch64::BI_InterlockedExchangeAdd16_nf:
-  case AArch64::BI_InterlockedExchangeAdd_nf:
-  case AArch64::BI_InterlockedExchangeAdd64_nf:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeAdd_nf, E);
-  case AArch64::BI_InterlockedExchange8_acq:
-  case AArch64::BI_InterlockedExchange16_acq:
-  case AArch64::BI_InterlockedExchange_acq:
-  case AArch64::BI_InterlockedExchange64_acq:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchange_acq, E);
-  case AArch64::BI_InterlockedExchange8_rel:
-  case AArch64::BI_InterlockedExchange16_rel:
-  case AArch64::BI_InterlockedExchange_rel:
-  case AArch64::BI_InterlockedExchange64_rel:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchange_rel, E);
-  case AArch64::BI_InterlockedExchange8_nf:
-  case AArch64::BI_InterlockedExchange16_nf:
-  case AArch64::BI_InterlockedExchange_nf:
-  case AArch64::BI_InterlockedExchange64_nf:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchange_nf, E);
-  case AArch64::BI_InterlockedCompareExchange8_acq:
-  case AArch64::BI_InterlockedCompareExchange16_acq:
-  case AArch64::BI_InterlockedCompareExchange_acq:
-  case AArch64::BI_InterlockedCompareExchange64_acq:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedCompareExchange_acq, E);
-  case AArch64::BI_InterlockedCompareExchange8_rel:
-  case AArch64::BI_InterlockedCompareExchange16_rel:
-  case AArch64::BI_InterlockedCompareExchange_rel:
-  case AArch64::BI_InterlockedCompareExchange64_rel:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedCompareExchange_rel, E);
-  case AArch64::BI_InterlockedCompareExchange8_nf:
-  case AArch64::BI_InterlockedCompareExchange16_nf:
-  case AArch64::BI_InterlockedCompareExchange_nf:
-  case AArch64::BI_InterlockedCompareExchange64_nf:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedCompareExchange_nf, E);
-  case AArch64::BI_InterlockedOr8_acq:
-  case AArch64::BI_InterlockedOr16_acq:
-  case AArch64::BI_InterlockedOr_acq:
-  case AArch64::BI_InterlockedOr64_acq:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_acq, E);
-  case AArch64::BI_InterlockedOr8_rel:
-  case AArch64::BI_InterlockedOr16_rel:
-  case AArch64::BI_InterlockedOr_rel:
-  case AArch64::BI_InterlockedOr64_rel:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_rel, E);
-  case AArch64::BI_InterlockedOr8_nf:
-  case AArch64::BI_InterlockedOr16_nf:
-  case AArch64::BI_InterlockedOr_nf:
-  case AArch64::BI_InterlockedOr64_nf:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr_nf, E);
-  case AArch64::BI_InterlockedXor8_acq:
-  case AArch64::BI_InterlockedXor16_acq:
-  case AArch64::BI_InterlockedXor_acq:
-  case AArch64::BI_InterlockedXor64_acq:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedXor_acq, E);
-  case AArch64::BI_InterlockedXor8_rel:
-  case AArch64::BI_InterlockedXor16_rel:
-  case AArch64::BI_InterlockedXor_rel:
-  case AArch64::BI_InterlockedXor64_rel:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedXor_rel, E);
-  case AArch64::BI_InterlockedXor8_nf:
-  case AArch64::BI_InterlockedXor16_nf:
-  case AArch64::BI_InterlockedXor_nf:
-  case AArch64::BI_InterlockedXor64_nf:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedXor_nf, E);
-  case AArch64::BI_InterlockedAnd8_acq:
-  case AArch64::BI_InterlockedAnd16_acq:
-  case AArch64::BI_InterlockedAnd_acq:
-  case AArch64::BI_InterlockedAnd64_acq:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedAnd_acq, E);
-  case AArch64::BI_InterlockedAnd8_rel:
-  case AArch64::BI_InterlockedAnd16_rel:
-  case AArch64::BI_InterlockedAnd_rel:
-  case AArch64::BI_InterlockedAnd64_rel:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedAnd_rel, E);
-  case AArch64::BI_InterlockedAnd8_nf:
-  case AArch64::BI_InterlockedAnd16_nf:
-  case AArch64::BI_InterlockedAnd_nf:
-  case AArch64::BI_InterlockedAnd64_nf:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedAnd_nf, E);
-  case AArch64::BI_InterlockedIncrement16_acq:
-  case AArch64::BI_InterlockedIncrement_acq:
-  case AArch64::BI_InterlockedIncrement64_acq:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedIncrement_acq, E);
-  case AArch64::BI_InterlockedIncrement16_rel:
-  case AArch64::BI_InterlockedIncrement_rel:
-  case AArch64::BI_InterlockedIncrement64_rel:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedIncrement_rel, E);
-  case AArch64::BI_InterlockedIncrement16_nf:
-  case AArch64::BI_InterlockedIncrement_nf:
-  case AArch64::BI_InterlockedIncrement64_nf:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedIncrement_nf, E);
-  case AArch64::BI_InterlockedDecrement16_acq:
-  case AArch64::BI_InterlockedDecrement_acq:
-  case AArch64::BI_InterlockedDecrement64_acq:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedDecrement_acq, E);
-  case AArch64::BI_InterlockedDecrement16_rel:
-  case AArch64::BI_InterlockedDecrement_rel:
-  case AArch64::BI_InterlockedDecrement64_rel:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedDecrement_rel, E);
-  case AArch64::BI_InterlockedDecrement16_nf:
-  case AArch64::BI_InterlockedDecrement_nf:
-  case AArch64::BI_InterlockedDecrement64_nf:
-    return EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedDecrement_nf, E);
-
-  case AArch64::BI_InterlockedAdd: {
-    Value *Arg0 = EmitScalarExpr(E->getArg(0));
-    Value *Arg1 = EmitScalarExpr(E->getArg(1));
-    AtomicRMWInst *RMWI = Builder.CreateAtomicRMW(
-      AtomicRMWInst::Add, Arg0, Arg1,
-      llvm::AtomicOrdering::SequentiallyConsistent);
-    return Builder.CreateAdd(RMWI, Arg1);
-  }
   }
 }
 
diff --git a/lib/CodeGen/CGExprAgg.cpp b/lib/CodeGen/CGExprAgg.cpp
index 695facd50b67..0a57870a7c58 100644
--- a/lib/CodeGen/CGExprAgg.cpp
+++ b/lib/CodeGen/CGExprAgg.cpp
@@ -1495,6 +1495,13 @@ void AggExprEmitter::VisitInitListExpr(InitListExpr *E) {
   // initializers throws an exception.
   SmallVector<EHScopeStack::stable_iterator, 16> cleanups;
   llvm::Instruction *cleanupDominator = nullptr;
+  auto addCleanup = [&](const EHScopeStack::stable_iterator &cleanup) {
+    cleanups.push_back(cleanup);
+    if (!cleanupDominator) // create placeholder once needed
+      cleanupDominator = CGF.Builder.CreateAlignedLoad(
+          CGF.Int8Ty, llvm::Constant::getNullValue(CGF.Int8PtrTy),
+          CharUnits::One());
+  };
 
   unsigned curInitIndex = 0;
 
@@ -1519,7 +1526,7 @@ void AggExprEmitter::VisitInitListExpr(InitListExpr *E) {
       if (QualType::DestructionKind dtorKind =
               Base.getType().isDestructedType()) {
         CGF.pushDestroy(dtorKind, V, Base.getType());
-        cleanups.push_back(CGF.EHStack.stable_begin());
+        addCleanup(CGF.EHStack.stable_begin());
       }
     }
   }
@@ -1596,15 +1603,9 @@ void AggExprEmitter::VisitInitListExpr(InitListExpr *E) {
           = field->getType().isDestructedType()) {
       assert(LV.isSimple());
       if (CGF.needsEHCleanup(dtorKind)) {
-        if (!cleanupDominator)
-          cleanupDominator = CGF.Builder.CreateAlignedLoad(
-              CGF.Int8Ty,
-              llvm::Constant::getNullValue(CGF.Int8PtrTy),
-              CharUnits::One()); // placeholder
-
         CGF.pushDestroy(EHCleanup, LV.getAddress(), field->getType(),
                         CGF.getDestroyer(dtorKind), false);
-        cleanups.push_back(CGF.EHStack.stable_begin());
+        addCleanup(CGF.EHStack.stable_begin());
         pushedCleanup = true;
       }
     }
@@ -1620,6 +1621,8 @@ void AggExprEmitter::VisitInitListExpr(InitListExpr *E) {
 
   // Deactivate all the partial cleanups in reverse order, which
   // generally means popping them.
+  assert((cleanupDominator || cleanups.empty()) &&
+         "Missing cleanupDominator before deactivating cleanup blocks");
   for (unsigned i = cleanups.size(); i != 0; --i)
     CGF.DeactivateCleanupBlock(cleanups[i-1], cleanupDominator);
 
diff --git a/lib/CodeGen/CGStmt.cpp b/lib/CodeGen/CGStmt.cpp
index dd0dea5b94a0..40ab79509f98 100644
--- a/lib/CodeGen/CGStmt.cpp
+++ b/lib/CodeGen/CGStmt.cpp
@@ -1846,11 +1846,9 @@ llvm::Value* CodeGenFunction::EmitAsmInput(
       InputExpr->EvaluateAsRValue(EVResult, getContext(), true);
 
       llvm::APSInt IntResult;
-      if (!EVResult.Val.toIntegralConstant(IntResult, InputExpr->getType(),
-                                           getContext()))
-        llvm_unreachable("Invalid immediate constant!");
-
-      return llvm::ConstantInt::get(getLLVMContext(), IntResult);
+      if (EVResult.Val.toIntegralConstant(IntResult, InputExpr->getType(),
+                                          getContext()))
+        return llvm::ConstantInt::get(getLLVMContext(), IntResult);
     }
 
     Expr::EvalResult Result;
diff --git a/lib/CodeGen/ItaniumCXXABI.cpp b/lib/CodeGen/ItaniumCXXABI.cpp
index 3b2413d960d6..51a2561a4552 100644
--- a/lib/CodeGen/ItaniumCXXABI.cpp
+++ b/lib/CodeGen/ItaniumCXXABI.cpp
@@ -1755,10 +1755,11 @@ llvm::Value *ItaniumCXXABI::EmitVirtualDestructorCall(
   CGCallee Callee = CGCallee::forVirtual(CE, GD, This, Ty);
 
   QualType ThisTy;
-  if (CE)
-    ThisTy = CE->getImplicitObjectArgument()->getType()->getPointeeType();
-  else
+  if (CE) {
+    ThisTy = CE->getObjectType();
+  } else {
     ThisTy = D->getDestroyedType();
+  }
 
   CGF.EmitCXXDestructorCall(GD, Callee, This.getPointer(), ThisTy, nullptr,
                             QualType(), nullptr);
diff --git a/lib/CodeGen/MicrosoftCXXABI.cpp b/lib/CodeGen/MicrosoftCXXABI.cpp
index fa34414de5da..ca06ad3f042b 100644
--- a/lib/CodeGen/MicrosoftCXXABI.cpp
+++ b/lib/CodeGen/MicrosoftCXXABI.cpp
@@ -1921,10 +1921,11 @@ llvm::Value *MicrosoftCXXABI::EmitVirtualDestructorCall(
       DtorType == Dtor_Deleting);
 
   QualType ThisTy;
-  if (CE)
-    ThisTy = CE->getImplicitObjectArgument()->getType()->getPointeeType();
-  else
+  if (CE) {
+    ThisTy = CE->getObjectType();
+  } else {
     ThisTy = D->getDestroyedType();
+  }
 
   This = adjustThisArgumentForVirtualFunctionCall(CGF, GD, This, true);
   RValue RV = CGF.EmitCXXDestructorCall(GD, Callee, This.getPointer(), ThisTy,
diff --git a/lib/CodeGen/TargetInfo.cpp b/lib/CodeGen/TargetInfo.cpp
index 5da988fb8a3c..1e1038dbfe95 100644
--- a/lib/CodeGen/TargetInfo.cpp
+++ b/lib/CodeGen/TargetInfo.cpp
@@ -9188,25 +9188,45 @@ static bool getTypeString(SmallStringEnc &Enc, const Decl *D,
 namespace {
 class RISCVABIInfo : public DefaultABIInfo {
 private:
-  unsigned XLen; // Size of the integer ('x') registers in bits.
+  // Size of the integer ('x') registers in bits.
+  unsigned XLen;
+  // Size of the floating point ('f') registers in bits. Note that the target
+  // ISA might have a wider FLen than the selected ABI (e.g. an RV32IF target
+  // with soft float ABI has FLen==0).
+  unsigned FLen;
   static const int NumArgGPRs = 8;
+  static const int NumArgFPRs = 8;
+  bool detectFPCCEligibleStructHelper(QualType Ty, CharUnits CurOff,
+                                      llvm::Type *&Field1Ty,
+                                      CharUnits &Field1Off,
+                                      llvm::Type *&Field2Ty,
+                                      CharUnits &Field2Off) const;
 
 public:
-  RISCVABIInfo(CodeGen::CodeGenTypes &CGT, unsigned XLen)
-      : DefaultABIInfo(CGT), XLen(XLen) {}
+  RISCVABIInfo(CodeGen::CodeGenTypes &CGT, unsigned XLen, unsigned FLen)
+      : DefaultABIInfo(CGT), XLen(XLen), FLen(FLen) {}
 
   // DefaultABIInfo's classifyReturnType and classifyArgumentType are
   // non-virtual, but computeInfo is virtual, so we overload it.
   void computeInfo(CGFunctionInfo &FI) const override;
 
-  ABIArgInfo classifyArgumentType(QualType Ty, bool IsFixed,
-                                  int &ArgGPRsLeft) const;
+  ABIArgInfo classifyArgumentType(QualType Ty, bool IsFixed, int &ArgGPRsLeft,
+                                  int &ArgFPRsLeft) const;
   ABIArgInfo classifyReturnType(QualType RetTy) const;
 
   Address EmitVAArg(CodeGenFunction &CGF, Address VAListAddr,
                     QualType Ty) const override;
 
   ABIArgInfo extendType(QualType Ty) const;
+
+  bool detectFPCCEligibleStruct(QualType Ty, llvm::Type *&Field1Ty,
+                                CharUnits &Field1Off, llvm::Type *&Field2Ty,
+                                CharUnits &Field2Off, int &NeededArgGPRs,
+                                int &NeededArgFPRs) const;
+  ABIArgInfo coerceAndExpandFPCCEligibleStruct(llvm::Type *Field1Ty,
+                                               CharUnits Field1Off,
+                                               llvm::Type *Field2Ty,
+                                               CharUnits Field2Off) const;
 };
 } // end anonymous namespace
 
@@ -9228,18 +9248,215 @@ void RISCVABIInfo::computeInfo(CGFunctionInfo &FI) const {
   // different for variadic arguments, we must also track whether we are
   // examining a vararg or not.
   int ArgGPRsLeft = IsRetIndirect ? NumArgGPRs - 1 : NumArgGPRs;
+  int ArgFPRsLeft = FLen ? NumArgFPRs : 0;
   int NumFixedArgs = FI.getNumRequiredArgs();
 
   int ArgNum = 0;
   for (auto &ArgInfo : FI.arguments()) {
     bool IsFixed = ArgNum < NumFixedArgs;
-    ArgInfo.info = classifyArgumentType(ArgInfo.type, IsFixed, ArgGPRsLeft);
+    ArgInfo.info =
+        classifyArgumentType(ArgInfo.type, IsFixed, ArgGPRsLeft, ArgFPRsLeft);
     ArgNum++;
   }
 }
 
+// Returns true if the struct is a potential candidate for the floating point
+// calling convention. If this function returns true, the caller is
+// responsible for checking that if there is only a single field then that
+// field is a float.
+bool RISCVABIInfo::detectFPCCEligibleStructHelper(QualType Ty, CharUnits CurOff,
+                                                  llvm::Type *&Field1Ty,
+                                                  CharUnits &Field1Off,
+                                                  llvm::Type *&Field2Ty,
+                                                  CharUnits &Field2Off) const {
+  bool IsInt = Ty->isIntegralOrEnumerationType();
+  bool IsFloat = Ty->isRealFloatingType();
+
+  if (IsInt || IsFloat) {
+    uint64_t Size = getContext().getTypeSize(Ty);
+    if (IsInt && Size > XLen)
+      return false;
+    // Can't be eligible if larger than the FP registers. Half precision isn't
+    // currently supported on RISC-V and the ABI hasn't been confirmed, so
+    // default to the integer ABI in that case.
+    if (IsFloat && (Size > FLen || Size < 32))
+      return false;
+    // Can't be eligible if an integer type was already found (int+int pairs
+    // are not eligible).
+    if (IsInt && Field1Ty && Field1Ty->isIntegerTy())
+      return false;
+    if (!Field1Ty) {
+      Field1Ty = CGT.ConvertType(Ty);
+      Field1Off = CurOff;
+      return true;
+    }
+    if (!Field2Ty) {
+      Field2Ty = CGT.ConvertType(Ty);
+      Field2Off = CurOff;
+      return true;
+    }
+    return false;
+  }
+
+  if (auto CTy = Ty->getAs<ComplexType>()) {
+    if (Field1Ty)
+      return false;
+    QualType EltTy = CTy->getElementType();
+    if (getContext().getTypeSize(EltTy) > FLen)
+      return false;
+    Field1Ty = CGT.ConvertType(EltTy);
+    Field1Off = CurOff;
+    assert(CurOff.isZero() && "Unexpected offset for first field");
+    Field2Ty = Field1Ty;
+    Field2Off = Field1Off + getContext().getTypeSizeInChars(EltTy);
+    return true;
+  }
+
+  if (const ConstantArrayType *ATy = getContext().getAsConstantArrayType(Ty)) {
+    uint64_t ArraySize = ATy->getSize().getZExtValue();
+    QualType EltTy = ATy->getElementType();
+    CharUnits EltSize = getContext().getTypeSizeInChars(EltTy);
+    for (uint64_t i = 0; i < ArraySize; ++i) {
+      bool Ret = detectFPCCEligibleStructHelper(EltTy, CurOff, Field1Ty,
+                                                Field1Off, Field2Ty, Field2Off);
+      if (!Ret)
+        return false;
+      CurOff += EltSize;
+    }
+    return true;
+  }
+
+  if (const auto *RTy = Ty->getAs<RecordType>()) {
+    // Structures with either a non-trivial destructor or a non-trivial
+    // copy constructor are not eligible for the FP calling convention.
+    if (CGCXXABI::RecordArgABI RAA = getRecordArgABI(Ty, CGT.getCXXABI()))
+      return false;
+    if (isEmptyRecord(getContext(), Ty, true))
+      return true;
+    const RecordDecl *RD = RTy->getDecl();
+    // Unions aren't eligible unless they're empty (which is caught above).
+    if (RD->isUnion())
+      return false;
+    int ZeroWidthBitFieldCount = 0;
+    for (const FieldDecl *FD : RD->fields()) {
+      const ASTRecordLayout &Layout = getContext().getASTRecordLayout(RD);
+      uint64_t FieldOffInBits = Layout.getFieldOffset(FD->getFieldIndex());
+      QualType QTy = FD->getType();
+      if (FD->isBitField()) {
+        unsigned BitWidth = FD->getBitWidthValue(getContext());
+        // Allow a bitfield with a type greater than XLen as long as the
+        // bitwidth is XLen or less.
+        if (getContext().getTypeSize(QTy) > XLen && BitWidth <= XLen)
+          QTy = getContext().getIntTypeForBitwidth(XLen, false);
+        if (BitWidth == 0) {
+          ZeroWidthBitFieldCount++;
+          continue;
+        }
+      }
+
+      bool Ret = detectFPCCEligibleStructHelper(
+          QTy, CurOff + getContext().toCharUnitsFromBits(FieldOffInBits),
+          Field1Ty, Field1Off, Field2Ty, Field2Off);
+      if (!Ret)
+        return false;
+
+      // As a quirk of the ABI, zero-width bitfields aren't ignored for fp+fp
+      // or int+fp structs, but are ignored for a struct with an fp field and
+      // any number of zero-width bitfields.
+      if (Field2Ty && ZeroWidthBitFieldCount > 0)
+        return false;
+    }
+    return Field1Ty != nullptr;
+  }
+
+  return false;
+}
+
+// Determine if a struct is eligible for passing according to the floating
+// point calling convention (i.e., when flattened it contains a single fp
+// value, fp+fp, or int+fp of appropriate size). If so, NeededArgFPRs and
+// NeededArgGPRs are incremented appropriately.
+bool RISCVABIInfo::detectFPCCEligibleStruct(QualType Ty, llvm::Type *&Field1Ty,
+                                            CharUnits &Field1Off,
+                                            llvm::Type *&Field2Ty,
+                                            CharUnits &Field2Off,
+                                            int &NeededArgGPRs,
+                                            int &NeededArgFPRs) const {
+  Field1Ty = nullptr;
+  Field2Ty = nullptr;
+  NeededArgGPRs = 0;
+  NeededArgFPRs = 0;
+  bool IsCandidate = detectFPCCEligibleStructHelper(
+      Ty, CharUnits::Zero(), Field1Ty, Field1Off, Field2Ty, Field2Off);
+  // Not really a candidate if we have a single int but no float.
+  if (Field1Ty && !Field2Ty && !Field1Ty->isFloatingPointTy())
+    return IsCandidate = false;
+  if (!IsCandidate)
+    return false;
+  if (Field1Ty && Field1Ty->isFloatingPointTy())
+    NeededArgFPRs++;
+  else if (Field1Ty)
+    NeededArgGPRs++;
+  if (Field2Ty && Field2Ty->isFloatingPointTy())
+    NeededArgFPRs++;
+  else if (Field2Ty)
+    NeededArgGPRs++;
+  return IsCandidate;
+}
+
+// Call getCoerceAndExpand for the two-element flattened struct described by
+// Field1Ty, Field1Off, Field2Ty, Field2Off. This method will create an
+// appropriate coerceToType and unpaddedCoerceToType.
+ABIArgInfo RISCVABIInfo::coerceAndExpandFPCCEligibleStruct(
+    llvm::Type *Field1Ty, CharUnits Field1Off, llvm::Type *Field2Ty,
+    CharUnits Field2Off) const {
+  SmallVector<llvm::Type *, 3> CoerceElts;
+  SmallVector<llvm::Type *, 2> UnpaddedCoerceElts;
+  if (!Field1Off.isZero())
+    CoerceElts.push_back(llvm::ArrayType::get(
+        llvm::Type::getInt8Ty(getVMContext()), Field1Off.getQuantity()));
+
+  CoerceElts.push_back(Field1Ty);
+  UnpaddedCoerceElts.push_back(Field1Ty);
+
+  if (!Field2Ty) {
+    return ABIArgInfo::getCoerceAndExpand(
+        llvm::StructType::get(getVMContext(), CoerceElts, !Field1Off.isZero()),
+        UnpaddedCoerceElts[0]);
+  }
+
+  CharUnits Field2Align =
+      CharUnits::fromQuantity(getDataLayout().getABITypeAlignment(Field2Ty));
+  CharUnits Field1Size =
+      CharUnits::fromQuantity(getDataLayout().getTypeStoreSize(Field1Ty));
+  CharUnits Field2OffNoPadNoPack = Field1Size.alignTo(Field2Align);
+
+  CharUnits Padding = CharUnits::Zero();
+  if (Field2Off > Field2OffNoPadNoPack)
+    Padding = Field2Off - Field2OffNoPadNoPack;
+  else if (Field2Off != Field2Align && Field2Off > Field1Size)
+    Padding = Field2Off - Field1Size;
+
+  bool IsPacked = !Field2Off.isMultipleOf(Field2Align);
+
+  if (!Padding.isZero())
+    CoerceElts.push_back(llvm::ArrayType::get(
+        llvm::Type::getInt8Ty(getVMContext()), Padding.getQuantity()));
+
+  CoerceElts.push_back(Field2Ty);
+  UnpaddedCoerceElts.push_back(Field2Ty);
+
+  auto CoerceToType =
+      llvm::StructType::get(getVMContext(), CoerceElts, IsPacked);
+  auto UnpaddedCoerceToType =
+      llvm::StructType::get(getVMContext(), UnpaddedCoerceElts, IsPacked);
+
+  return ABIArgInfo::getCoerceAndExpand(CoerceToType, UnpaddedCoerceToType);
+}
+
 ABIArgInfo RISCVABIInfo::classifyArgumentType(QualType Ty, bool IsFixed,
-                                              int &ArgGPRsLeft) const {
+                                              int &ArgGPRsLeft,
+                                              int &ArgFPRsLeft) const {
   assert(ArgGPRsLeft <= NumArgGPRs && "Arg GPR tracking underflow");
   Ty = useFirstFieldIfTransparentUnion(Ty);
 
@@ -9257,6 +9474,42 @@ ABIArgInfo RISCVABIInfo::classifyArgumentType(QualType Ty, bool IsFixed,
     return ABIArgInfo::getIgnore();
 
   uint64_t Size = getContext().getTypeSize(Ty);
+
+  // Pass floating point values via FPRs if possible.
+  if (IsFixed && Ty->isFloatingType() && FLen >= Size && ArgFPRsLeft) {
+    ArgFPRsLeft--;
+    return ABIArgInfo::getDirect();
+  }
+
+  // Complex types for the hard float ABI must be passed direct rather than
+  // using CoerceAndExpand.
+  if (IsFixed && Ty->isComplexType() && FLen && ArgFPRsLeft >= 2) {
+    QualType EltTy = Ty->getAs<ComplexType>()->getElementType();
+    if (getContext().getTypeSize(EltTy) <= FLen) {
+      ArgFPRsLeft -= 2;
+      return ABIArgInfo::getDirect();
+    }
+  }
+
+  if (IsFixed && FLen && Ty->isStructureOrClassType()) {
+    llvm::Type *Field1Ty = nullptr;
+    llvm::Type *Field2Ty = nullptr;
+    CharUnits Field1Off = CharUnits::Zero();
+    CharUnits Field2Off = CharUnits::Zero();
+    int NeededArgGPRs;
+    int NeededArgFPRs;
+    bool IsCandidate =
+        detectFPCCEligibleStruct(Ty, Field1Ty, Field1Off, Field2Ty, Field2Off,
+                                 NeededArgGPRs, NeededArgFPRs);
+    if (IsCandidate && NeededArgGPRs <= ArgGPRsLeft &&
+        NeededArgFPRs <= ArgFPRsLeft) {
+      ArgGPRsLeft -= NeededArgGPRs;
+      ArgFPRsLeft -= NeededArgFPRs;
+      return coerceAndExpandFPCCEligibleStruct(Field1Ty, Field1Off, Field2Ty,
+                                               Field2Off);
+    }
+  }
+
   uint64_t NeededAlign = getContext().getTypeAlign(Ty);
   bool MustUseStack = false;
   // Determine the number of GPRs needed to pass the current argument
@@ -9315,10 +9568,12 @@ ABIArgInfo RISCVABIInfo::classifyReturnType(QualType RetTy) const {
     return ABIArgInfo::getIgnore();
 
   int ArgGPRsLeft = 2;
+  int ArgFPRsLeft = FLen ? 2 : 0;
 
   // The rules for return and argument types are the same, so defer to
   // classifyArgumentType.
-  return classifyArgumentType(RetTy, /*IsFixed=*/true, ArgGPRsLeft);
+  return classifyArgumentType(RetTy, /*IsFixed=*/true, ArgGPRsLeft,
+                              ArgFPRsLeft);
 }
 
 Address RISCVABIInfo::EmitVAArg(CodeGenFunction &CGF, Address VAListAddr,
@@ -9353,8 +9608,9 @@ ABIArgInfo RISCVABIInfo::extendType(QualType Ty) const {
 namespace {
 class RISCVTargetCodeGenInfo : public TargetCodeGenInfo {
 public:
-  RISCVTargetCodeGenInfo(CodeGen::CodeGenTypes &CGT, unsigned XLen)
-      : TargetCodeGenInfo(new RISCVABIInfo(CGT, XLen)) {}
+  RISCVTargetCodeGenInfo(CodeGen::CodeGenTypes &CGT, unsigned XLen,
+                         unsigned FLen)
+      : TargetCodeGenInfo(new RISCVABIInfo(CGT, XLen, FLen)) {}
 
   void setTargetAttributes(const Decl *D, llvm::GlobalValue *GV,
                            CodeGen::CodeGenModule &CGM) const override {
@@ -9493,9 +9749,16 @@ const TargetCodeGenInfo &CodeGenModule::getTargetCodeGenInfo() {
     return SetCGInfo(new MSP430TargetCodeGenInfo(Types));
 
   case llvm::Triple::riscv32:
-    return SetCGInfo(new RISCVTargetCodeGenInfo(Types, 32));
-  case llvm::Triple::riscv64:
-    return SetCGInfo(new RISCVTargetCodeGenInfo(Types, 64));
+  case llvm::Triple::riscv64: {
+    StringRef ABIStr = getTarget().getABI();
+    unsigned XLen = getTarget().getPointerWidth(0);
+    unsigned ABIFLen = 0;
+    if (ABIStr.endswith("f"))
+      ABIFLen = 32;
+    else if (ABIStr.endswith("d"))
+      ABIFLen = 64;
+    return SetCGInfo(new RISCVTargetCodeGenInfo(Types, XLen, ABIFLen));
+  }
 
   case llvm::Triple::systemz: {
     bool HasVector = getTarget().getABI() == "vector";
diff --git a/lib/DirectoryWatcher/linux/DirectoryWatcher-linux.cpp b/lib/DirectoryWatcher/linux/DirectoryWatcher-linux.cpp
index 6d7d69da4db5..1a66faeb3239 100644
--- a/lib/DirectoryWatcher/linux/DirectoryWatcher-linux.cpp
+++ b/lib/DirectoryWatcher/linux/DirectoryWatcher-linux.cpp
@@ -24,7 +24,6 @@
 #include <vector>
 
 #include <fcntl.h>
-#include <linux/version.h>
 #include <sys/epoll.h>
 #include <sys/inotify.h>
 #include <unistd.h>
@@ -335,7 +334,7 @@ std::unique_ptr<DirectoryWatcher> clang::DirectoryWatcher::create(
       InotifyFD, Path.str().c_str(),
       IN_CREATE | IN_DELETE | IN_DELETE_SELF | IN_MODIFY |
       IN_MOVED_FROM | IN_MOVE_SELF | IN_MOVED_TO | IN_ONLYDIR | IN_IGNORED
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,36)
+#ifdef IN_EXCL_UNLINK
       | IN_EXCL_UNLINK
 #endif
       );
diff --git a/lib/Driver/ToolChains/Clang.cpp b/lib/Driver/ToolChains/Clang.cpp
index cb861f27aeda..2508178423bf 100644
--- a/lib/Driver/ToolChains/Clang.cpp
+++ b/lib/Driver/ToolChains/Clang.cpp
@@ -501,8 +501,6 @@ static codegenoptions::DebugInfoKind DebugLevelToInfoKind(const Arg &A) {
   return codegenoptions::LimitedDebugInfo;
 }
 
-enum class FramePointerKind { None, NonLeaf, All };
-
 static bool mustUseNonLeafFramePointerForTarget(const llvm::Triple &Triple) {
   switch (Triple.getArch()){
   default:
@@ -517,9 +515,6 @@ static bool mustUseNonLeafFramePointerForTarget(const llvm::Triple &Triple) {
 
 static bool useFramePointerForTargetByDefault(const ArgList &Args,
                                               const llvm::Triple &Triple) {
-  if (Args.hasArg(options::OPT_pg))
-    return true;
-
   switch (Triple.getArch()) {
   case llvm::Triple::xcore:
   case llvm::Triple::wasm32:
@@ -579,22 +574,32 @@ static bool useFramePointerForTargetByDefault(const ArgList &Args,
   return true;
 }
 
-static FramePointerKind getFramePointerKind(const ArgList &Args,
-                                            const llvm::Triple &Triple) {
-  Arg *A = Args.getLastArg(options::OPT_fomit_frame_pointer,
-                           options::OPT_fno_omit_frame_pointer);
-  bool OmitFP = A && A->getOption().matches(options::OPT_fomit_frame_pointer);
-  bool NoOmitFP =
-      A && A->getOption().matches(options::OPT_fno_omit_frame_pointer);
-  if (NoOmitFP || mustUseNonLeafFramePointerForTarget(Triple) ||
-      (!OmitFP && useFramePointerForTargetByDefault(Args, Triple))) {
-    if (Args.hasFlag(options::OPT_momit_leaf_frame_pointer,
-                     options::OPT_mno_omit_leaf_frame_pointer,
-                     Triple.isPS4CPU()))
-      return FramePointerKind::NonLeaf;
-    return FramePointerKind::All;
-  }
-  return FramePointerKind::None;
+static bool shouldUseFramePointer(const ArgList &Args,
+                                  const llvm::Triple &Triple) {
+  if (Arg *A = Args.getLastArg(options::OPT_fno_omit_frame_pointer,
+                               options::OPT_fomit_frame_pointer))
+    return A->getOption().matches(options::OPT_fno_omit_frame_pointer) ||
+           mustUseNonLeafFramePointerForTarget(Triple);
+
+  if (Args.hasArg(options::OPT_pg))
+    return true;
+
+  return useFramePointerForTargetByDefault(Args, Triple);
+}
+
+static bool shouldUseLeafFramePointer(const ArgList &Args,
+                                      const llvm::Triple &Triple) {
+  if (Arg *A = Args.getLastArg(options::OPT_mno_omit_leaf_frame_pointer,
+                               options::OPT_momit_leaf_frame_pointer))
+    return A->getOption().matches(options::OPT_mno_omit_leaf_frame_pointer);
+
+  if (Args.hasArg(options::OPT_pg))
+    return true;
+
+  if (Triple.isPS4CPU())
+    return false;
+
+  return useFramePointerForTargetByDefault(Args, Triple);
 }
 
 /// Add a CC1 option to specify the debug compilation directory.
@@ -3946,12 +3951,8 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
   if (Args.hasFlag(options::OPT_mrtd, options::OPT_mno_rtd, false))
     CmdArgs.push_back("-fdefault-calling-conv=stdcall");
 
-  FramePointerKind FPKeepKind = getFramePointerKind(Args, RawTriple);
-  if (FPKeepKind != FramePointerKind::None) {
+  if (shouldUseFramePointer(Args, RawTriple))
     CmdArgs.push_back("-mdisable-fp-elim");
-    if (FPKeepKind == FramePointerKind::NonLeaf)
-      CmdArgs.push_back("-momit-leaf-frame-pointer");
-  }
   if (!Args.hasFlag(options::OPT_fzero_initialized_in_bss,
                     options::OPT_fno_zero_initialized_in_bss))
     CmdArgs.push_back("-mno-zero-initialized-in-bss");
@@ -4136,6 +4137,9 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
     CmdArgs.push_back(A->getValue());
   }
 
+  if (!shouldUseLeafFramePointer(Args, RawTriple))
+    CmdArgs.push_back("-momit-leaf-frame-pointer");
+
   // Explicitly error on some things we know we don't support and can't just
   // ignore.
   if (!Args.hasArg(options::OPT_fallow_unsupported)) {
@@ -5489,7 +5493,7 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
   }
 
   if (Arg *A = Args.getLastArg(options::OPT_pg))
-    if (FPKeepKind == FramePointerKind::None)
+    if (!shouldUseFramePointer(Args, Triple))
       D.Diag(diag::err_drv_argument_not_allowed_with) << "-fomit-frame-pointer"
                                                       << A->getAsString(Args);
 
diff --git a/lib/Frontend/ASTUnit.cpp b/lib/Frontend/ASTUnit.cpp
index 7445a94cfe59..783d1f9d0919 100644
--- a/lib/Frontend/ASTUnit.cpp
+++ b/lib/Frontend/ASTUnit.cpp
@@ -435,7 +435,6 @@ void ASTUnit::CacheCodeCompletionResults() {
           | (1LL << CodeCompletionContext::CCC_UnionTag)
           | (1LL << CodeCompletionContext::CCC_ClassOrStructTag)
           | (1LL << CodeCompletionContext::CCC_Type)
-          | (1LL << CodeCompletionContext::CCC_Symbol)
           | (1LL << CodeCompletionContext::CCC_SymbolOrNewName)
           | (1LL << CodeCompletionContext::CCC_ParenthesizedExpression);
 
diff --git a/lib/Frontend/CompilerInvocation.cpp b/lib/Frontend/CompilerInvocation.cpp
index 8a9844096f08..bc54e38a1a63 100644
--- a/lib/Frontend/CompilerInvocation.cpp
+++ b/lib/Frontend/CompilerInvocation.cpp
@@ -2408,7 +2408,7 @@ static void ParseLangArgs(LangOptions &Opts, ArgList &Args, InputKind IK,
         .Cases("cl1.1", "CL1.1", LangStandard::lang_opencl11)
         .Cases("cl1.2", "CL1.2", LangStandard::lang_opencl12)
         .Cases("cl2.0", "CL2.0", LangStandard::lang_opencl20)
-        .Case("c++", LangStandard::lang_openclcpp)
+        .Cases("clc++", "CLC++", LangStandard::lang_openclcpp)
         .Default(LangStandard::lang_unspecified);
 
     if (OpenCLLangStd == LangStandard::lang_unspecified) {
diff --git a/lib/Frontend/InitPreprocessor.cpp b/lib/Frontend/InitPreprocessor.cpp
index 3906e2ae1b98..6feb7bcbd4b7 100644
--- a/lib/Frontend/InitPreprocessor.cpp
+++ b/lib/Frontend/InitPreprocessor.cpp
@@ -437,17 +437,17 @@ static void InitializeStandardPredefinedMacros(const TargetInfo &TI,
       default:
         llvm_unreachable("Unsupported OpenCL version");
       }
-      Builder.defineMacro("CL_VERSION_1_0", "100");
-      Builder.defineMacro("CL_VERSION_1_1", "110");
-      Builder.defineMacro("CL_VERSION_1_2", "120");
-      Builder.defineMacro("CL_VERSION_2_0", "200");
+    }
+    Builder.defineMacro("CL_VERSION_1_0", "100");
+    Builder.defineMacro("CL_VERSION_1_1", "110");
+    Builder.defineMacro("CL_VERSION_1_2", "120");
+    Builder.defineMacro("CL_VERSION_2_0", "200");
 
-      if (TI.isLittleEndian())
-        Builder.defineMacro("__ENDIAN_LITTLE__");
+    if (TI.isLittleEndian())
+      Builder.defineMacro("__ENDIAN_LITTLE__");
 
-      if (LangOpts.FastRelaxedMath)
-        Builder.defineMacro("__FAST_RELAXED_MATH__");
-    }
+    if (LangOpts.FastRelaxedMath)
+      Builder.defineMacro("__FAST_RELAXED_MATH__");
   }
   // Not "standard" per se, but available even with the -undef flag.
   if (LangOpts.AsmPreprocessor)
diff --git a/lib/Headers/emmintrin.h b/lib/Headers/emmintrin.h
index 3d55f5f2710f..c8fefdfc792a 100644
--- a/lib/Headers/emmintrin.h
+++ b/lib/Headers/emmintrin.h
@@ -4029,7 +4029,7 @@ _mm_storeu_si128(__m128i_u *__p, __m128i __b)
 /// \param __b
 ///    A 128-bit integer vector containing the value to be stored.
 static __inline__ void __DEFAULT_FN_ATTRS
-_mm_storeu_si64(void const *__p, __m128i __b)
+_mm_storeu_si64(void *__p, __m128i __b)
 {
   struct __storeu_si64 {
     long long __v;
@@ -4050,7 +4050,7 @@ _mm_storeu_si64(void const *__p, __m128i __b)
 /// \param __b
 ///    A 128-bit integer vector containing the value to be stored.
 static __inline__ void __DEFAULT_FN_ATTRS
-_mm_storeu_si32(void const *__p, __m128i __b)
+_mm_storeu_si32(void *__p, __m128i __b)
 {
   struct __storeu_si32 {
     int __v;
@@ -4071,7 +4071,7 @@ _mm_storeu_si32(void const *__p, __m128i __b)
 /// \param __b
 ///    A 128-bit integer vector containing the value to be stored.
 static __inline__ void __DEFAULT_FN_ATTRS
-_mm_storeu_si16(void const *__p, __m128i __b)
+_mm_storeu_si16(void *__p, __m128i __b)
 {
   struct __storeu_si16 {
     short __v;
diff --git a/lib/Headers/opencl-c-base.h b/lib/Headers/opencl-c-base.h
index a82954ddd326..9a23333a33e6 100644
--- a/lib/Headers/opencl-c-base.h
+++ b/lib/Headers/opencl-c-base.h
@@ -126,7 +126,7 @@ typedef double double8 __attribute__((ext_vector_type(8)));
 typedef double double16 __attribute__((ext_vector_type(16)));
 #endif
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 #define NULL ((void*)0)
 #endif
 
@@ -276,7 +276,7 @@ typedef uint cl_mem_fence_flags;
  */
 #define CLK_GLOBAL_MEM_FENCE   0x02
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 typedef enum memory_scope {
   memory_scope_work_item = __OPENCL_MEMORY_SCOPE_WORK_ITEM,
@@ -288,9 +288,6 @@ typedef enum memory_scope {
 #endif
 } memory_scope;
 
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
-
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
 /**
  * Queue a memory fence to ensure correct ordering of memory
  * operations between work-items of a work-group to
@@ -313,7 +310,7 @@ typedef enum memory_order
   memory_order_seq_cst = __ATOMIC_SEQ_CST
 } memory_order;
 
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // OpenCL v1.1 s6.11.3, v1.2 s6.12.14, v2.0 s6.13.14 - Image Read and Write Functions
 
@@ -389,14 +386,10 @@ typedef enum memory_order
 #endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
 
 // OpenCL v2.0 s6.13.16 - Pipe Functions
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 #define CLK_NULL_RESERVE_ID (__builtin_astype(((void*)(__SIZE_MAX__)), reserve_id_t))
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
-
 
 // OpenCL v2.0 s6.13.17 - Enqueue Kernels
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
-
 #define CL_COMPLETE                                 0x0
 #define CL_RUNNING                                  0x1
 #define CL_SUBMITTED                                0x2
@@ -435,7 +428,7 @@ typedef struct {
   size_t localWorkSize[MAX_WORK_DIM];
 } ndrange_t;
 
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 #ifdef cl_intel_device_side_avc_motion_estimation
 #pragma OPENCL EXTENSION cl_intel_device_side_avc_motion_estimation : begin
diff --git a/lib/Headers/opencl-c.h b/lib/Headers/opencl-c.h
index 4207c53ccedb..8741bccec9ad 100644
--- a/lib/Headers/opencl-c.h
+++ b/lib/Headers/opencl-c.h
@@ -11,11 +11,11 @@
 
 #include "opencl-c-base.h"
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 #ifndef cl_khr_depth_images
 #define cl_khr_depth_images
 #endif //cl_khr_depth_images
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 #if __OPENCL_C_VERSION__ < CL_VERSION_2_0
 #ifdef cl_khr_3d_image_writes
@@ -23,10 +23,10 @@
 #endif //cl_khr_3d_image_writes
 #endif //__OPENCL_C_VERSION__ < CL_VERSION_2_0
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 #pragma OPENCL EXTENSION cl_intel_planar_yuv : begin
 #pragma OPENCL EXTENSION cl_intel_planar_yuv : end
-#endif // __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 
 #define __ovld __attribute__((overloadable))
 #define __conv __attribute__((convergent))
@@ -6517,11 +6517,11 @@ size_t __ovld __cnfn get_group_id(uint dimindx);
  */
 size_t __ovld __cnfn get_global_offset(uint dimindx);
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 size_t __ovld get_enqueued_local_size(uint dimindx);
 size_t __ovld get_global_linear_id(void);
 size_t __ovld get_local_linear_id(void);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // OpenCL v1.1 s6.11.2, v1.2 s6.12.2, v2.0 s6.13.2 - Math functions
 
@@ -7352,7 +7352,7 @@ half16 __ovld __cnfn fmod(half16 x, half16 y);
  * Returns fmin(x - floor (x), 0x1.fffffep-1f ).
  * floor(x) is returned in iptr.
  */
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 float __ovld fract(float x, float *iptr);
 float2 __ovld fract(float2 x, float2 *iptr);
 float3 __ovld fract(float3 x, float3 *iptr);
@@ -7434,7 +7434,7 @@ half4 __ovld fract(half4 x, __private half4 *iptr);
 half8 __ovld fract(half8 x, __private half8 *iptr);
 half16 __ovld fract(half16 x, __private half16 *iptr);
 #endif //cl_khr_fp16
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Extract mantissa and exponent from x. For each
@@ -7442,7 +7442,7 @@ half16 __ovld fract(half16 x, __private half16 *iptr);
  * magnitude in the interval [1/2, 1) or 0. Each
  * component of x equals mantissa returned * 2^exp.
  */
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 float __ovld frexp(float x, int *exp);
 float2 __ovld frexp(float2 x, int2 *exp);
 float3 __ovld frexp(float3 x, int3 *exp);
@@ -7524,7 +7524,7 @@ half4 __ovld frexp(half4 x, __private int4 *exp);
 half8 __ovld frexp(half8 x, __private int8 *exp);
 half16 __ovld frexp(half16 x, __private int16 *exp);
 #endif //cl_khr_fp16
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Compute the value of the square root of x^2 + y^2
@@ -7649,7 +7649,7 @@ half8 __ovld __cnfn lgamma(half8 x);
 half16 __ovld __cnfn lgamma(half16 x);
 #endif //cl_khr_fp16
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 float __ovld lgamma_r(float x, int *signp);
 float2 __ovld lgamma_r(float2 x, int2 *signp);
 float3 __ovld lgamma_r(float3 x, int3 *signp);
@@ -7731,7 +7731,7 @@ half4 __ovld lgamma_r(half4 x, __private int4 *signp);
 half8 __ovld lgamma_r(half8 x, __private int8 *signp);
 half16 __ovld lgamma_r(half16 x, __private int16 *signp);
 #endif //cl_khr_fp16
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Compute natural logarithm.
@@ -7955,7 +7955,7 @@ half16 __ovld __cnfn minmag(half16 x, half16 y);
  * the argument. It stores the integral part in the object
  * pointed to by iptr.
  */
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 float __ovld modf(float x, float *iptr);
 float2 __ovld modf(float2 x, float2 *iptr);
 float3 __ovld modf(float3 x, float3 *iptr);
@@ -8037,7 +8037,7 @@ half4 __ovld modf(half4 x, __private half4 *iptr);
 half8 __ovld modf(half8 x, __private half8 *iptr);
 half16 __ovld modf(half16 x, __private half16 *iptr);
 #endif //cl_khr_fp16
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Returns a quiet NaN. The nancode may be placed
@@ -8215,7 +8215,7 @@ half16 __ovld __cnfn remainder(half16 x, half16 y);
  * sign as x/y. It stores this signed value in the object
  * pointed to by quo.
  */
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 float __ovld remquo(float x, float y, int *quo);
 float2 __ovld remquo(float2 x, float2 y, int2 *quo);
 float3 __ovld remquo(float3 x, float3 y, int3 *quo);
@@ -8298,7 +8298,7 @@ half4 __ovld remquo(half4 x, half4 y, __private int4 *quo);
 half8 __ovld remquo(half8 x, half8 y, __private int8 *quo);
 half16 __ovld remquo(half16 x, half16 y, __private int16 *quo);
 #endif //cl_khr_fp16
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 /**
  * Round to integral value (using round to nearest
  * even rounding mode) in floating-point format.
@@ -8439,7 +8439,7 @@ half16 __ovld __cnfn sin(half16);
  * is the return value and computed cosine is returned
  * in cosval.
  */
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 float __ovld sincos(float x, float *cosval);
 float2 __ovld sincos(float2 x, float2 *cosval);
 float3 __ovld sincos(float3 x, float3 *cosval);
@@ -8521,7 +8521,7 @@ half4 __ovld sincos(half4 x, __private half4 *cosval);
 half8 __ovld sincos(half8 x, __private half8 *cosval);
 half16 __ovld sincos(half16 x, __private half16 *cosval);
 #endif //cl_khr_fp16
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Compute hyperbolic sine.
@@ -9446,7 +9446,7 @@ ulong16 __ovld __cnfn clz(ulong16 x);
  * returns the size in bits of the type of x or
  * component type of x, if x is a vector.
  */
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 char __ovld ctz(char x);
 uchar __ovld ctz(uchar x);
 char2 __ovld ctz(char2 x);
@@ -9495,7 +9495,7 @@ long8 __ovld ctz(long8 x);
 ulong8 __ovld ctz(ulong8 x);
 long16 __ovld ctz(long16 x);
 ulong16 __ovld ctz(ulong16 x);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Returns mul_hi(a, b) + c.
@@ -11340,7 +11340,7 @@ half8 __ovld vload8(size_t offset, const __constant half *p);
 half16 __ovld vload16(size_t offset, const __constant half *p);
 #endif //cl_khr_fp16
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 char2 __ovld vload2(size_t offset, const char *p);
 uchar2 __ovld vload2(size_t offset, const uchar *p);
 short2 __ovld vload2(size_t offset, const short *p);
@@ -11578,9 +11578,9 @@ half4 __ovld vload4(size_t offset, const __private half *p);
 half8 __ovld vload8(size_t offset, const __private half *p);
 half16 __ovld vload16(size_t offset, const __private half *p);
 #endif //cl_khr_fp16
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 void __ovld vstore2(char2 data, size_t offset, char *p);
 void __ovld vstore2(uchar2 data, size_t offset, uchar *p);
 void __ovld vstore2(short2 data, size_t offset, short *p);
@@ -11814,7 +11814,7 @@ void __ovld vstore4(half4 data, size_t offset, __private half *p);
 void __ovld vstore8(half8 data, size_t offset, __private half *p);
 void __ovld vstore16(half16 data, size_t offset, __private half *p);
 #endif //cl_khr_fp16
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Read sizeof (half) bytes of data from address
@@ -11825,13 +11825,13 @@ void __ovld vstore16(half16 data, size_t offset, __private half *p);
  * must be 16-bit aligned.
  */
 float __ovld vload_half(size_t offset, const __constant half *p);
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 float __ovld vload_half(size_t offset, const half *p);
 #else
 float __ovld vload_half(size_t offset, const __global half *p);
 float __ovld vload_half(size_t offset, const __local half *p);
 float __ovld vload_half(size_t offset, const __private half *p);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Read sizeof (halfn) bytes of data from address
@@ -11846,7 +11846,7 @@ float3 __ovld vload_half3(size_t offset, const __constant half *p);
 float4 __ovld vload_half4(size_t offset, const __constant half *p);
 float8 __ovld vload_half8(size_t offset, const __constant half *p);
 float16 __ovld vload_half16(size_t offset, const __constant half *p);
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 float2 __ovld vload_half2(size_t offset, const half *p);
 float3 __ovld vload_half3(size_t offset, const half *p);
 float4 __ovld vload_half4(size_t offset, const half *p);
@@ -11868,7 +11868,7 @@ float3 __ovld vload_half3(size_t offset, const __private half *p);
 float4 __ovld vload_half4(size_t offset, const __private half *p);
 float8 __ovld vload_half8(size_t offset, const __private half *p);
 float16 __ovld vload_half16(size_t offset, const __private half *p);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * The float value given by data is first
@@ -11881,7 +11881,7 @@ float16 __ovld vload_half16(size_t offset, const __private half *p);
  * The default current rounding mode is round to
  * nearest even.
  */
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 void __ovld vstore_half(float data, size_t offset, half *p);
 void __ovld vstore_half_rte(float data, size_t offset, half *p);
 void __ovld vstore_half_rtz(float data, size_t offset, half *p);
@@ -11927,7 +11927,7 @@ void __ovld vstore_half_rtz(double data, size_t offset, __private half *p);
 void __ovld vstore_half_rtp(double data, size_t offset, __private half *p);
 void __ovld vstore_half_rtn(double data, size_t offset, __private half *p);
 #endif //cl_khr_fp64
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * The floatn value given by data is converted to
@@ -11940,7 +11940,7 @@ void __ovld vstore_half_rtn(double data, size_t offset, __private half *p);
  * The default current rounding mode is round to
  * nearest even.
  */
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 void __ovld vstore_half2(float2 data, size_t offset, half *p);
 void __ovld vstore_half3(float3 data, size_t offset, half *p);
 void __ovld vstore_half4(float4 data, size_t offset, half *p);
@@ -12146,7 +12146,7 @@ void __ovld vstore_half4_rtn(double4 data, size_t offset, __private half *p);
 void __ovld vstore_half8_rtn(double8 data, size_t offset, __private half *p);
 void __ovld vstore_half16_rtn(double16 data, size_t offset, __private half *p);
 #endif //cl_khr_fp64
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * For n = 1, 2, 4, 8 and 16 read sizeof (halfn)
@@ -12167,7 +12167,7 @@ float3 __ovld vloada_half3(size_t offset, const __constant half *p);
 float4 __ovld vloada_half4(size_t offset, const __constant half *p);
 float8 __ovld vloada_half8(size_t offset, const __constant half *p);
 float16 __ovld vloada_half16(size_t offset, const __constant half *p);
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 float __ovld vloada_half(size_t offset, const half *p);
 float2 __ovld vloada_half2(size_t offset, const half *p);
 float3 __ovld vloada_half3(size_t offset, const half *p);
@@ -12193,7 +12193,7 @@ float3 __ovld vloada_half3(size_t offset, const __private half *p);
 float4 __ovld vloada_half4(size_t offset, const __private half *p);
 float8 __ovld vloada_half8(size_t offset, const __private half *p);
 float16 __ovld vloada_half16(size_t offset, const __private half *p);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * The floatn value given by data is converted to
@@ -12211,7 +12211,7 @@ float16 __ovld vloada_half16(size_t offset, const __private half *p);
  * mode. The default current rounding mode is
  * round to nearest even.
  */
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 void __ovld vstorea_half(float data, size_t offset, half *p);
 void __ovld vstorea_half2(float2 data, size_t offset, half *p);
 void __ovld vstorea_half3(float3 data, size_t offset, half *p);
@@ -12496,7 +12496,7 @@ void __ovld vstorea_half4_rtn(double4 data,size_t offset, __private half *p);
 void __ovld vstorea_half8_rtn(double8 data,size_t offset, __private half *p);
 void __ovld vstorea_half16_rtn(double16 data,size_t offset, __private half *p);
 #endif //cl_khr_fp64
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // OpenCL v1.1 s6.11.8, v1.2 s6.12.8, v2.0 s6.13.8 - Synchronization Functions
 
@@ -12532,10 +12532,10 @@ void __ovld vstorea_half16_rtn(double16 data,size_t offset, __private half *p);
 
 void __ovld __conv barrier(cl_mem_fence_flags flags);
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 void __ovld __conv work_group_barrier(cl_mem_fence_flags flags, memory_scope scope);
 void __ovld __conv work_group_barrier(cl_mem_fence_flags flags);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // OpenCL v1.1 s6.11.9, v1.2 s6.12.9 - Explicit Memory Fence Functions
 
@@ -12580,7 +12580,7 @@ void __ovld write_mem_fence(cl_mem_fence_flags flags);
 
 // OpenCL v2.0 s6.13.9 - Address Space Qualifier Functions
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 cl_mem_fence_flags __ovld get_fence(const void *ptr);
 cl_mem_fence_flags __ovld get_fence(void *ptr);
 
@@ -12591,7 +12591,7 @@ cl_mem_fence_flags __ovld get_fence(void *ptr);
  * where gentype is builtin type or user defined type.
  */
 
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // OpenCL v1.1 s6.11.10, v1.2 s6.12.10, v2.0 s6.13.10 - Async Copies from Global to Local Memory, Local to Global Memory, and Prefetch
 
@@ -13371,7 +13371,7 @@ unsigned long __ovld atom_xor(volatile __local unsigned long *p, unsigned long v
 
 // OpenCL v2.0 s6.13.11 - Atomics Functions
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // double atomics support requires extensions cl_khr_int64_base_atomics and cl_khr_int64_extended_atomics
 #if defined(cl_khr_int64_base_atomics) && defined(cl_khr_int64_extended_atomics)
@@ -13692,7 +13692,7 @@ void __ovld atomic_flag_clear(volatile atomic_flag *object);
 void __ovld atomic_flag_clear_explicit(volatile atomic_flag *object, memory_order order);
 void __ovld atomic_flag_clear_explicit(volatile atomic_flag *object, memory_order order, memory_scope scope);
 
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // OpenCL v1.1 s6.11.12, v1.2 s6.12.12, v2.0 s6.13.12 - Miscellaneous Vector Functions
 
@@ -14186,7 +14186,7 @@ half16 __ovld __cnfn shuffle2(half8 x, half8 y, ushort16 mask);
 half16 __ovld __cnfn shuffle2(half16 x, half16 y, ushort16 mask);
 #endif //cl_khr_fp16
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 // OpenCL v1.2 s6.12.13, v2.0 s6.13.13 - printf
 
 int printf(__constant const char* st, ...) __attribute__((format(printf, 1, 2)));
@@ -14307,7 +14307,7 @@ int4 __purefn __ovld read_imagei(read_only image3d_t image, sampler_t sampler, f
 uint4 __purefn __ovld read_imageui(read_only image3d_t image, sampler_t sampler, int4 coord);
 uint4 __purefn __ovld read_imageui(read_only image3d_t image, sampler_t sampler, float4 coord);
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 float4 __purefn __ovld read_imagef(read_only image2d_array_t image_array, sampler_t sampler, int4 coord);
 float4 __purefn __ovld read_imagef(read_only image2d_array_t image_array, sampler_t sampler, float4 coord);
 
@@ -14315,7 +14315,7 @@ int4 __purefn __ovld read_imagei(read_only image2d_array_t image_array, sampler_
 int4 __purefn __ovld read_imagei(read_only image2d_array_t image_array, sampler_t sampler, float4 coord);
 uint4 __purefn __ovld read_imageui(read_only image2d_array_t image_array, sampler_t sampler, int4 coord);
 uint4 __purefn __ovld read_imageui(read_only image2d_array_t image_array, sampler_t sampler, float4 coord);
-#endif // __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 
 float4 __purefn __ovld read_imagef(read_only image1d_t image, sampler_t sampler, int coord);
 float4 __purefn __ovld read_imagef(read_only image1d_t image, sampler_t sampler, float coord);
@@ -14325,7 +14325,7 @@ int4 __purefn __ovld read_imagei(read_only image1d_t image, sampler_t sampler, f
 uint4 __purefn __ovld read_imageui(read_only image1d_t image, sampler_t sampler, int coord);
 uint4 __purefn __ovld read_imageui(read_only image1d_t image, sampler_t sampler, float coord);
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 float4 __purefn __ovld read_imagef(read_only image1d_array_t image_array, sampler_t sampler, int2 coord);
 float4 __purefn __ovld read_imagef(read_only image1d_array_t image_array, sampler_t sampler, float2 coord);
 
@@ -14333,7 +14333,7 @@ int4 __purefn __ovld read_imagei(read_only image1d_array_t image_array, sampler_
 int4 __purefn __ovld read_imagei(read_only image1d_array_t image_array, sampler_t sampler, float2 coord);
 uint4 __purefn __ovld read_imageui(read_only image1d_array_t image_array, sampler_t sampler, int2 coord);
 uint4 __purefn __ovld read_imageui(read_only image1d_array_t image_array, sampler_t sampler, float2 coord);
-#endif // __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 
 #ifdef cl_khr_depth_images
 float __purefn __ovld read_imagef(read_only image2d_depth_t image, sampler_t sampler, float2 coord);
@@ -14358,7 +14358,7 @@ float __purefn __ovld read_imagef(read_only image2d_array_msaa_depth_t image, in
 #endif //cl_khr_gl_msaa_sharing
 
 // OpenCL Extension v2.0 s9.18 - Mipmaps
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 #ifdef cl_khr_mipmap_image
 
 float4 __purefn __ovld read_imagef(read_only image1d_t image, sampler_t sampler, float coord, float lod);
@@ -14410,9 +14410,9 @@ int4 __purefn __ovld read_imagei(read_only image3d_t image, sampler_t sampler, f
 uint4 __purefn __ovld read_imageui(read_only image3d_t image, sampler_t sampler, float4 coord, float4 gradientX, float4 gradientY);
 
 #endif //cl_khr_mipmap_image
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 
 /**
 * Sampler-less Image Access
@@ -14447,7 +14447,7 @@ float4 __purefn __ovld read_imagef(read_only image3d_t image, int4 coord);
 int4 __purefn __ovld read_imagei(read_only image3d_t image, int4 coord);
 uint4 __purefn __ovld read_imageui(read_only image3d_t image, int4 coord);
 
-#endif // __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 
 // Image read functions returning half4 type
 #ifdef cl_khr_fp16
@@ -14457,7 +14457,7 @@ half4 __purefn __ovld read_imageh(read_only image2d_t image, sampler_t sampler,
 half4 __purefn __ovld read_imageh(read_only image2d_t image, sampler_t sampler, float2 coord);
 half4 __purefn __ovld read_imageh(read_only image3d_t image, sampler_t sampler, int4 coord);
 half4 __purefn __ovld read_imageh(read_only image3d_t image, sampler_t sampler, float4 coord);
-#if __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 half4 __purefn __ovld read_imageh(read_only image1d_array_t image, sampler_t sampler, int2 coord);
 half4 __purefn __ovld read_imageh(read_only image1d_array_t image, sampler_t sampler, float2 coord);
 half4 __purefn __ovld read_imageh(read_only image2d_array_t image, sampler_t sampler, int4 coord);
@@ -14471,11 +14471,11 @@ half4 __purefn __ovld read_imageh(read_only image3d_t image, int4 coord);
 half4 __purefn __ovld read_imageh(read_only image1d_array_t image, int2 coord);
 half4 __purefn __ovld read_imageh(read_only image2d_array_t image, int4 coord);
 half4 __purefn __ovld read_imageh(read_only image1d_buffer_t image, int coord);
-#endif // __OPENCL_C_VERSION__ >= CL_VERSION_1_2
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
 #endif //cl_khr_fp16
 
 // Image read functions for read_write images
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 float4 __purefn __ovld read_imagef(read_write image1d_t image, int coord);
 int4 __purefn __ovld read_imagei(read_write image1d_t image, int coord);
 uint4 __purefn __ovld read_imageui(read_write image1d_t image, int coord);
@@ -14518,7 +14518,7 @@ float __purefn __ovld read_imagef(read_write image2d_msaa_depth_t image, int2 co
 float __purefn __ovld read_imagef(read_write image2d_array_msaa_depth_t image, int4 coord, int sample);
 #endif //cl_khr_gl_msaa_sharing
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 #ifdef cl_khr_mipmap_image
 float4 __purefn __ovld read_imagef(read_write image1d_t image, sampler_t sampler, float coord, float lod);
 int4 __purefn __ovld read_imagei(read_write image1d_t image, sampler_t sampler, float coord, float lod);
@@ -14569,7 +14569,7 @@ int4 __purefn __ovld read_imagei(read_write image3d_t image, sampler_t sampler,
 uint4 __purefn __ovld read_imageui(read_write image3d_t image, sampler_t sampler, float4 coord, float4 gradientX, float4 gradientY);
 
 #endif //cl_khr_mipmap_image
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // Image read functions returning half4 type
 #ifdef cl_khr_fp16
@@ -14580,7 +14580,7 @@ half4 __purefn __ovld read_imageh(read_write image1d_array_t image, int2 coord);
 half4 __purefn __ovld read_imageh(read_write image2d_array_t image, int4 coord);
 half4 __purefn __ovld read_imageh(read_write image1d_buffer_t image, int coord);
 #endif //cl_khr_fp16
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Write color value to location specified by coordinate
@@ -14681,7 +14681,7 @@ void __ovld write_imagef(write_only image2d_array_depth_t image, int4 coord, flo
 #endif //cl_khr_depth_images
 
 // OpenCL Extension v2.0 s9.18 - Mipmaps
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 #ifdef cl_khr_mipmap_image
 void __ovld write_imagef(write_only image1d_t image, int coord, int lod, float4 color);
 void __ovld write_imagei(write_only image1d_t image, int coord, int lod, int4 color);
@@ -14708,7 +14708,7 @@ void __ovld write_imagei(write_only image3d_t image, int4 coord, int lod, int4 c
 void __ovld write_imageui(write_only image3d_t image, int4 coord, int lod, uint4 color);
 #endif
 #endif //cl_khr_mipmap_image
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // Image write functions for half4 type
 #ifdef cl_khr_fp16
@@ -14723,7 +14723,7 @@ void __ovld write_imageh(write_only image1d_buffer_t image, int coord, half4 col
 #endif //cl_khr_fp16
 
 // Image write functions for read_write images
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 void __ovld write_imagef(read_write image2d_t image, int2 coord, float4 color);
 void __ovld write_imagei(read_write image2d_t image, int2 coord, int4 color);
 void __ovld write_imageui(read_write image2d_t image, int2 coord, uint4 color);
@@ -14755,7 +14755,7 @@ void __ovld write_imagef(read_write image2d_depth_t image, int2 coord, float col
 void __ovld write_imagef(read_write image2d_array_depth_t image, int4 coord, float color);
 #endif //cl_khr_depth_images
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 #ifdef cl_khr_mipmap_image
 void __ovld write_imagef(read_write image1d_t image, int coord, int lod, float4 color);
 void __ovld write_imagei(read_write image1d_t image, int coord, int lod, int4 color);
@@ -14782,7 +14782,7 @@ void __ovld write_imagei(read_write image3d_t image, int4 coord, int lod, int4 c
 void __ovld write_imageui(read_write image3d_t image, int4 coord, int lod, uint4 color);
 #endif
 #endif //cl_khr_mipmap_image
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // Image write functions for half4 type
 #ifdef cl_khr_fp16
@@ -14795,7 +14795,7 @@ void __ovld write_imageh(read_write image1d_array_t image, int2 coord, half4 col
 void __ovld write_imageh(read_write image2d_array_t image, int4 coord, half4 color);
 void __ovld write_imageh(read_write image1d_buffer_t image, int coord, half4 color);
 #endif //cl_khr_fp16
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // Note: In OpenCL v1.0/1.1/1.2, image argument of image query builtin functions does not have
 // access qualifier, which by default assume read_only access qualifier. Image query builtin
@@ -14843,7 +14843,7 @@ int __ovld __cnfn get_image_width(write_only image2d_array_msaa_t image);
 int __ovld __cnfn get_image_width(write_only image2d_array_msaa_depth_t image);
 #endif //cl_khr_gl_msaa_sharing
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 int __ovld __cnfn get_image_width(read_write image1d_t image);
 int __ovld __cnfn get_image_width(read_write image1d_buffer_t image);
 int __ovld __cnfn get_image_width(read_write image2d_t image);
@@ -14860,7 +14860,7 @@ int __ovld __cnfn get_image_width(read_write image2d_msaa_depth_t image);
 int __ovld __cnfn get_image_width(read_write image2d_array_msaa_t image);
 int __ovld __cnfn get_image_width(read_write image2d_array_msaa_depth_t image);
 #endif //cl_khr_gl_msaa_sharing
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Return the image height in pixels.
@@ -14895,7 +14895,7 @@ int __ovld __cnfn get_image_height(write_only image2d_array_msaa_t image);
 int __ovld __cnfn get_image_height(write_only image2d_array_msaa_depth_t image);
 #endif //cl_khr_gl_msaa_sharing
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 int __ovld __cnfn get_image_height(read_write image2d_t image);
 int __ovld __cnfn get_image_height(read_write image3d_t image);
 int __ovld __cnfn get_image_height(read_write image2d_array_t image);
@@ -14909,7 +14909,7 @@ int __ovld __cnfn get_image_height(read_write image2d_msaa_depth_t image);
 int __ovld __cnfn get_image_height(read_write image2d_array_msaa_t image);
 int __ovld __cnfn get_image_height(read_write image2d_array_msaa_depth_t image);
 #endif //cl_khr_gl_msaa_sharing
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Return the image depth in pixels.
@@ -14920,12 +14920,12 @@ int __ovld __cnfn get_image_depth(read_only image3d_t image);
 int __ovld __cnfn get_image_depth(write_only image3d_t image);
 #endif
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 int __ovld __cnfn get_image_depth(read_write image3d_t image);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // OpenCL Extension v2.0 s9.18 - Mipmaps
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 #ifdef cl_khr_mipmap_image
 /**
  * Return the image miplevels.
@@ -14961,7 +14961,7 @@ int __ovld get_image_num_mip_levels(read_write image2d_array_depth_t image);
 int __ovld get_image_num_mip_levels(read_write image2d_depth_t image);
 
 #endif //cl_khr_mipmap_image
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Return the channel data type. Valid values are:
@@ -15018,7 +15018,7 @@ int __ovld __cnfn get_image_channel_data_type(write_only image2d_array_msaa_t im
 int __ovld __cnfn get_image_channel_data_type(write_only image2d_array_msaa_depth_t image);
 #endif //cl_khr_gl_msaa_sharing
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 int __ovld __cnfn get_image_channel_data_type(read_write image1d_t image);
 int __ovld __cnfn get_image_channel_data_type(read_write image1d_buffer_t image);
 int __ovld __cnfn get_image_channel_data_type(read_write image2d_t image);
@@ -15035,7 +15035,7 @@ int __ovld __cnfn get_image_channel_data_type(read_write image2d_msaa_depth_t im
 int __ovld __cnfn get_image_channel_data_type(read_write image2d_array_msaa_t image);
 int __ovld __cnfn get_image_channel_data_type(read_write image2d_array_msaa_depth_t image);
 #endif //cl_khr_gl_msaa_sharing
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Return the image channel order. Valid values are:
@@ -15090,7 +15090,7 @@ int __ovld __cnfn get_image_channel_order(write_only image2d_array_msaa_t image)
 int __ovld __cnfn get_image_channel_order(write_only image2d_array_msaa_depth_t image);
 #endif //cl_khr_gl_msaa_sharing
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 int __ovld __cnfn get_image_channel_order(read_write image1d_t image);
 int __ovld __cnfn get_image_channel_order(read_write image1d_buffer_t image);
 int __ovld __cnfn get_image_channel_order(read_write image2d_t image);
@@ -15107,7 +15107,7 @@ int __ovld __cnfn get_image_channel_order(read_write image2d_msaa_depth_t image)
 int __ovld __cnfn get_image_channel_order(read_write image2d_array_msaa_t image);
 int __ovld __cnfn get_image_channel_order(read_write image2d_array_msaa_depth_t image);
 #endif //cl_khr_gl_msaa_sharing
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Return the 2D image width and height as an int2
@@ -15140,7 +15140,7 @@ int2 __ovld __cnfn get_image_dim(write_only image2d_array_msaa_t image);
 int2 __ovld __cnfn get_image_dim(write_only image2d_array_msaa_depth_t image);
 #endif //cl_khr_gl_msaa_sharing
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 int2 __ovld __cnfn get_image_dim(read_write image2d_t image);
 int2 __ovld __cnfn get_image_dim(read_write image2d_array_t image);
 #ifdef cl_khr_depth_images
@@ -15153,7 +15153,7 @@ int2 __ovld __cnfn get_image_dim(read_write image2d_msaa_depth_t image);
 int2 __ovld __cnfn get_image_dim(read_write image2d_array_msaa_t image);
 int2 __ovld __cnfn get_image_dim(read_write image2d_array_msaa_depth_t image);
 #endif //cl_khr_gl_msaa_sharing
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Return the 3D image width, height, and depth as an
@@ -15165,9 +15165,9 @@ int4 __ovld __cnfn get_image_dim(read_only image3d_t image);
 #ifdef cl_khr_3d_image_writes
 int4 __ovld __cnfn get_image_dim(write_only image3d_t image);
 #endif
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 int4 __ovld __cnfn get_image_dim(read_write image3d_t image);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
  * Return the image array size.
@@ -15193,7 +15193,7 @@ size_t __ovld __cnfn get_image_array_size(write_only image2d_array_msaa_t image_
 size_t __ovld __cnfn get_image_array_size(write_only image2d_array_msaa_depth_t image_array);
 #endif //cl_khr_gl_msaa_sharing
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 size_t __ovld __cnfn get_image_array_size(read_write image1d_array_t image_array);
 size_t __ovld __cnfn get_image_array_size(read_write image2d_array_t image_array);
 #ifdef cl_khr_depth_images
@@ -15203,7 +15203,7 @@ size_t __ovld __cnfn get_image_array_size(read_write image2d_array_depth_t image
 size_t __ovld __cnfn get_image_array_size(read_write image2d_array_msaa_t image_array);
 size_t __ovld __cnfn get_image_array_size(read_write image2d_array_msaa_depth_t image_array);
 #endif //cl_khr_gl_msaa_sharing
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 /**
 * Return the number of samples associated with image
@@ -15219,17 +15219,17 @@ int __ovld get_image_num_samples(write_only image2d_msaa_depth_t image);
 int __ovld get_image_num_samples(write_only image2d_array_msaa_t image);
 int __ovld get_image_num_samples(write_only image2d_array_msaa_depth_t image);
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 int __ovld get_image_num_samples(read_write image2d_msaa_t image);
 int __ovld get_image_num_samples(read_write image2d_msaa_depth_t image);
 int __ovld get_image_num_samples(read_write image2d_array_msaa_t image);
 int __ovld get_image_num_samples(read_write image2d_array_msaa_depth_t image);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 #endif
 
 // OpenCL v2.0 s6.13.15 - Work-group Functions
 
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 int __ovld __conv work_group_all(int predicate);
 int __ovld __conv work_group_any(int predicate);
 
@@ -15327,16 +15327,16 @@ double __ovld __conv work_group_scan_inclusive_min(double x);
 double __ovld __conv work_group_scan_inclusive_max(double x);
 #endif //cl_khr_fp64
 
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // OpenCL v2.0 s6.13.16 - Pipe Functions
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 bool __ovld is_valid_reserve_id(reserve_id_t reserve_id);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 
 // OpenCL v2.0 s6.13.17 - Enqueue Kernels
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 ndrange_t __ovld ndrange_1D(size_t);
 ndrange_t __ovld ndrange_1D(size_t, size_t);
@@ -15365,7 +15365,7 @@ bool __ovld is_valid_event (clk_event_t event);
 void __ovld capture_event_profiling_info(clk_event_t, clk_profiling_info, __global void* value);
 
 queue_t __ovld get_default_queue(void);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 // OpenCL Extension v2.0 s9.17 - Sub-groups
 
@@ -15374,16 +15374,16 @@ queue_t __ovld get_default_queue(void);
 uint    __ovld get_sub_group_size(void);
 uint    __ovld get_max_sub_group_size(void);
 uint    __ovld get_num_sub_groups(void);
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 uint    __ovld get_enqueued_num_sub_groups(void);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 uint    __ovld get_sub_group_id(void);
 uint    __ovld get_sub_group_local_id(void);
 
 void    __ovld __conv sub_group_barrier(cl_mem_fence_flags flags);
-#if __OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 void    __ovld __conv sub_group_barrier(cl_mem_fence_flags flags, memory_scope scope);
-#endif //__OPENCL_C_VERSION__ >= CL_VERSION_2_0
+#endif //defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 int     __ovld __conv sub_group_all(int predicate);
 int     __ovld __conv sub_group_any(int predicate);
@@ -15573,12 +15573,12 @@ uint2   __ovld __conv intel_sub_group_block_read2( read_only image2d_t image, in
 uint4   __ovld __conv intel_sub_group_block_read4( read_only image2d_t image, int2 coord );
 uint8   __ovld __conv intel_sub_group_block_read8( read_only image2d_t image, int2 coord );
 
-#if (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 uint    __ovld __conv intel_sub_group_block_read(read_write image2d_t image, int2 coord);
 uint2   __ovld __conv intel_sub_group_block_read2(read_write image2d_t image, int2 coord);
 uint4   __ovld __conv intel_sub_group_block_read4(read_write image2d_t image, int2 coord);
 uint8   __ovld __conv intel_sub_group_block_read8(read_write image2d_t image, int2 coord);
-#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 uint    __ovld __conv intel_sub_group_block_read( const __global uint* p );
 uint2   __ovld __conv intel_sub_group_block_read2( const __global uint* p );
@@ -15590,12 +15590,12 @@ void    __ovld __conv intel_sub_group_block_write2(write_only image2d_t image, i
 void    __ovld __conv intel_sub_group_block_write4(write_only image2d_t image, int2 coord, uint4 data);
 void    __ovld __conv intel_sub_group_block_write8(write_only image2d_t image, int2 coord, uint8 data);
 
-#if (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 void    __ovld __conv intel_sub_group_block_write(read_write image2d_t image, int2 coord, uint data);
 void    __ovld __conv intel_sub_group_block_write2(read_write image2d_t image, int2 coord, uint2 data);
 void    __ovld __conv intel_sub_group_block_write4(read_write image2d_t image, int2 coord, uint4 data);
 void    __ovld __conv intel_sub_group_block_write8(read_write image2d_t image, int2 coord, uint8 data);
-#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 void    __ovld __conv intel_sub_group_block_write( __global uint* p, uint data );
 void    __ovld __conv intel_sub_group_block_write2( __global uint* p, uint2 data );
@@ -15713,12 +15713,12 @@ uint2      __ovld __conv intel_sub_group_block_read_ui2( read_only image2d_t ima
 uint4      __ovld __conv intel_sub_group_block_read_ui4( read_only image2d_t image, int2 byte_coord );
 uint8      __ovld __conv intel_sub_group_block_read_ui8( read_only image2d_t image, int2 byte_coord );
 
-#if (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 uint       __ovld __conv intel_sub_group_block_read_ui( read_write image2d_t image, int2 byte_coord );
 uint2      __ovld __conv intel_sub_group_block_read_ui2( read_write image2d_t image, int2 byte_coord );
 uint4      __ovld __conv intel_sub_group_block_read_ui4( read_write image2d_t image, int2 byte_coord );
 uint8      __ovld __conv intel_sub_group_block_read_ui8( read_write image2d_t image, int2 byte_coord );
-#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 uint       __ovld __conv intel_sub_group_block_read_ui( const __global uint* p );
 uint2      __ovld __conv intel_sub_group_block_read_ui2( const __global uint* p );
@@ -15730,12 +15730,12 @@ void       __ovld __conv intel_sub_group_block_write_ui2( read_only image2d_t im
 void       __ovld __conv intel_sub_group_block_write_ui4( read_only image2d_t image, int2 byte_coord, uint4 data );
 void       __ovld __conv intel_sub_group_block_write_ui8( read_only image2d_t image, int2 byte_coord, uint8 data );
 
-#if (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 void       __ovld __conv intel_sub_group_block_write_ui( read_write image2d_t image, int2 byte_coord, uint data );
 void       __ovld __conv intel_sub_group_block_write_ui2( read_write image2d_t image, int2 byte_coord, uint2 data );
 void       __ovld __conv intel_sub_group_block_write_ui4( read_write image2d_t image, int2 byte_coord, uint4 data );
 void       __ovld __conv intel_sub_group_block_write_ui8( read_write image2d_t image, int2 byte_coord, uint8 data );
-#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 void       __ovld __conv intel_sub_group_block_write_ui( __global uint* p, uint data );
 void       __ovld __conv intel_sub_group_block_write_ui2( __global uint* p, uint2 data );
@@ -15747,12 +15747,12 @@ ushort2     __ovld __conv intel_sub_group_block_read_us2( read_only image2d_t im
 ushort4     __ovld __conv intel_sub_group_block_read_us4( read_only image2d_t image, int2 coord );
 ushort8     __ovld __conv intel_sub_group_block_read_us8( read_only image2d_t image, int2 coord );
 
-#if (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 ushort      __ovld __conv intel_sub_group_block_read_us(read_write image2d_t image, int2 coord);
 ushort2     __ovld __conv intel_sub_group_block_read_us2(read_write image2d_t image, int2 coord);
 ushort4     __ovld __conv intel_sub_group_block_read_us4(read_write image2d_t image, int2 coord);
 ushort8     __ovld __conv intel_sub_group_block_read_us8(read_write image2d_t image, int2 coord);
-#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 ushort      __ovld __conv intel_sub_group_block_read_us(  const __global ushort* p );
 ushort2     __ovld __conv intel_sub_group_block_read_us2( const __global ushort* p );
@@ -15764,12 +15764,12 @@ void        __ovld __conv intel_sub_group_block_write_us2(write_only image2d_t i
 void        __ovld __conv intel_sub_group_block_write_us4(write_only image2d_t image, int2 coord, ushort4 data);
 void        __ovld __conv intel_sub_group_block_write_us8(write_only image2d_t image, int2 coord, ushort8 data);
 
-#if (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 void        __ovld __conv intel_sub_group_block_write_us(read_write image2d_t image, int2 coord, ushort  data);
 void        __ovld __conv intel_sub_group_block_write_us2(read_write image2d_t image, int2 coord, ushort2 data);
 void        __ovld __conv intel_sub_group_block_write_us4(read_write image2d_t image, int2 coord, ushort4 data);
 void        __ovld __conv intel_sub_group_block_write_us8(read_write image2d_t image, int2 coord, ushort8 data);
-#endif // (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
 void        __ovld __conv intel_sub_group_block_write_us(  __global ushort* p, ushort  data );
 void        __ovld __conv intel_sub_group_block_write_us2( __global ushort* p, ushort2 data );
diff --git a/lib/Lex/PPDirectives.cpp b/lib/Lex/PPDirectives.cpp
index 2756042f23eb..5658f46c99de 100644
--- a/lib/Lex/PPDirectives.cpp
+++ b/lib/Lex/PPDirectives.cpp
@@ -33,6 +33,7 @@
 #include "clang/Lex/Token.h"
 #include "clang/Lex/VariadicMacroSupport.h"
 #include "llvm/ADT/ArrayRef.h"
+#include "llvm/ADT/ScopeExit.h"
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/STLExtras.h"
@@ -2399,6 +2400,13 @@ MacroInfo *Preprocessor::ReadOptionalMacroParameterListAndBody(
   Token Tok;
   LexUnexpandedToken(Tok);
 
+  // Ensure we consume the rest of the macro body if errors occur.
+  auto _ = llvm::make_scope_exit([&]() {
+    // The flag indicates if we are still waiting for 'eod'.
+    if (CurLexer->ParsingPreprocessorDirective)
+      DiscardUntilEndOfDirective();
+  });
+
   // Used to un-poison and then re-poison identifiers of the __VA_ARGS__ ilk
   // within their appropriate context.
   VariadicMacroScopeGuard VariadicMacroScopeGuard(*this);
@@ -2420,12 +2428,8 @@ MacroInfo *Preprocessor::ReadOptionalMacroParameterListAndBody(
   } else if (Tok.is(tok::l_paren)) {
     // This is a function-like macro definition.  Read the argument list.
     MI->setIsFunctionLike();
-    if (ReadMacroParameterList(MI, LastTok)) {
-      // Throw away the rest of the line.
-      if (CurPPLexer->ParsingPreprocessorDirective)
-        DiscardUntilEndOfDirective();
+    if (ReadMacroParameterList(MI, LastTok))
       return nullptr;
-    }
 
     // If this is a definition of an ISO C/C++ variadic function-like macro (not
     // using the GNU named varargs extension) inform our variadic scope guard
diff --git a/lib/Sema/Sema.cpp b/lib/Sema/Sema.cpp
index 11fed28b52db..485d39e2c9e8 100644
--- a/lib/Sema/Sema.cpp
+++ b/lib/Sema/Sema.cpp
@@ -961,6 +961,7 @@ void Sema::ActOnEndOfTranslationUnit() {
 
   // All dllexport classes should have been processed already.
   assert(DelayedDllExportClasses.empty());
+  assert(DelayedDllExportMemberFunctions.empty());
 
   // Remove file scoped decls that turned out to be used.
   UnusedFileScopedDecls.erase(
diff --git a/lib/Sema/SemaDecl.cpp b/lib/Sema/SemaDecl.cpp
index a6c52b7d4b2b..8f19edbc4f36 100644
--- a/lib/Sema/SemaDecl.cpp
+++ b/lib/Sema/SemaDecl.cpp
@@ -1984,10 +1984,27 @@ NamedDecl *Sema::LazilyCreateBuiltin(IdentifierInfo *II, unsigned ID,
   ASTContext::GetBuiltinTypeError Error;
   QualType R = Context.GetBuiltinType(ID, Error);
   if (Error) {
-    if (ForRedeclaration)
-      Diag(Loc, diag::warn_implicit_decl_requires_sysheader)
-          << getHeaderName(Context.BuiltinInfo, ID, Error)
+    if (!ForRedeclaration)
+      return nullptr;
+
+    // If we have a builtin without an associated type we should not emit a
+    // warning when we were not able to find a type for it.
+    if (Error == ASTContext::GE_Missing_type)
+      return nullptr;
+
+    // If we could not find a type for setjmp it is because the jmp_buf type was
+    // not defined prior to the setjmp declaration.
+    if (Error == ASTContext::GE_Missing_setjmp) {
+      Diag(Loc, diag::warn_implicit_decl_no_jmp_buf)
           << Context.BuiltinInfo.getName(ID);
+      return nullptr;
+    }
+
+    // Generally, we emit a warning that the declaration requires the
+    // appropriate header.
+    Diag(Loc, diag::warn_implicit_decl_requires_sysheader)
+        << getHeaderName(Context.BuiltinInfo, ID, Error)
+        << Context.BuiltinInfo.getName(ID);
     return nullptr;
   }
 
@@ -11527,9 +11544,12 @@ void Sema::AddInitializerToDecl(Decl *RealDecl, Expr *Init, bool DirectInit) {
   // Check for self-references within variable initializers.
   // Variables declared within a function/method body (except for references)
   // are handled by a dataflow analysis.
-  if (!VDecl->hasLocalStorage() || VDecl->getType()->isRecordType() ||
-      VDecl->getType()->isReferenceType()) {
-    CheckSelfReference(*this, RealDecl, Init, DirectInit);
+  // This is undefined behavior in C++, but valid in C.
+  if (getLangOpts().CPlusPlus) {
+    if (!VDecl->hasLocalStorage() || VDecl->getType()->isRecordType() ||
+        VDecl->getType()->isReferenceType()) {
+      CheckSelfReference(*this, RealDecl, Init, DirectInit);
+    }
   }
 
   // If the type changed, it means we had an incomplete type that was
diff --git a/lib/Sema/SemaDeclCXX.cpp b/lib/Sema/SemaDeclCXX.cpp
index 9a6385f28319..15984f89e22d 100644
--- a/lib/Sema/SemaDeclCXX.cpp
+++ b/lib/Sema/SemaDeclCXX.cpp
@@ -1030,8 +1030,10 @@ static IsTupleLike isTupleLike(Sema &S, SourceLocation Loc, QualType T,
   TemplateArgumentListInfo Args(Loc, Loc);
   Args.addArgument(getTrivialTypeTemplateArgument(S, Loc, T));
 
-  // If there's no tuple_size specialization, it's not tuple-like.
-  if (lookupStdTypeTraitMember(S, R, Loc, "tuple_size", Args, /*DiagID*/0))
+  // If there's no tuple_size specialization or the lookup of 'value' is empty,
+  // it's not tuple-like.
+  if (lookupStdTypeTraitMember(S, R, Loc, "tuple_size", Args, /*DiagID*/ 0) ||
+      R.empty())
     return IsTupleLike::NotTupleLike;
 
   // If we get this far, we've committed to the tuple interpretation, but
@@ -1048,11 +1050,6 @@ static IsTupleLike isTupleLike(Sema &S, SourceLocation Loc, QualType T,
     }
   } Diagnoser(R, Args);
 
-  if (R.empty()) {
-    Diagnoser.diagnoseNotICE(S, Loc, SourceRange());
-    return IsTupleLike::Error;
-  }
-
   ExprResult E =
       S.BuildDeclarationNameExpr(CXXScopeSpec(), R, /*NeedsADL*/false);
   if (E.isInvalid())
@@ -6165,8 +6162,8 @@ void Sema::CheckCompletedCXXClass(CXXRecordDecl *Record) {
         M->dropAttr<DLLExportAttr>();
 
       if (M->hasAttr<DLLExportAttr>()) {
-        DefineImplicitSpecialMember(*this, M, M->getLocation());
-        ActOnFinishInlineFunctionDef(M);
+        // Define after any fields with in-class initializers have been parsed.
+        DelayedDllExportMemberFunctions.push_back(M);
       }
     }
   };
@@ -11419,6 +11416,21 @@ void Sema::ActOnFinishCXXMemberDecls() {
 
 void Sema::ActOnFinishCXXNonNestedClass(Decl *D) {
   referenceDLLExportedClassMethods();
+
+  if (!DelayedDllExportMemberFunctions.empty()) {
+    SmallVector<CXXMethodDecl*, 4> WorkList;
+    std::swap(DelayedDllExportMemberFunctions, WorkList);
+    for (CXXMethodDecl *M : WorkList) {
+      DefineImplicitSpecialMember(*this, M, M->getLocation());
+
+      // Pass the method to the consumer to get emitted. This is not necessary
+      // for explicit instantiation definitions, as they will get emitted
+      // anyway.
+      if (M->getParent()->getTemplateSpecializationKind() !=
+          TSK_ExplicitInstantiationDefinition)
+        ActOnFinishInlineFunctionDef(M);
+    }
+  }
 }
 
 void Sema::referenceDLLExportedClassMethods() {
diff --git a/lib/Sema/SemaExprCXX.cpp b/lib/Sema/SemaExprCXX.cpp
index 705e3b9bd7fb..c1c08969c7bd 100644
--- a/lib/Sema/SemaExprCXX.cpp
+++ b/lib/Sema/SemaExprCXX.cpp
@@ -6794,14 +6794,10 @@ ExprResult Sema::ActOnStartCXXMemberReference(Scope *S, Expr *Base,
   // it's legal for the type to be incomplete if this is a pseudo-destructor
   // call.  We'll do more incomplete-type checks later in the lookup process,
   // so just skip this check for ObjC types.
-  if (BaseType->isObjCObjectOrInterfaceType()) {
+  if (!BaseType->isRecordType()) {
     ObjectType = ParsedType::make(BaseType);
     MayBePseudoDestructor = true;
     return Base;
-  } else if (!BaseType->isRecordType()) {
-    ObjectType = nullptr;
-    MayBePseudoDestructor = true;
-    return Base;
   }
 
   // The object type must be complete (or dependent), or
diff --git a/lib/Sema/SemaInit.cpp b/lib/Sema/SemaInit.cpp
index bc1069609336..60f34775c6b2 100644
--- a/lib/Sema/SemaInit.cpp
+++ b/lib/Sema/SemaInit.cpp
@@ -1289,7 +1289,16 @@ void InitListChecker::CheckSubElementType(const InitializedEntity &Entity,
     // FIXME: Better EqualLoc?
     InitializationKind Kind =
         InitializationKind::CreateCopy(expr->getBeginLoc(), SourceLocation());
-    InitializationSequence Seq(SemaRef, Entity, Kind, expr,
+
+    // Vector elements can be initialized from other vectors in which case
+    // we need initialization entity with a type of a vector (and not a vector
+    // element!) initializing multiple vector elements.
+    auto TmpEntity =
+        (ElemType->isExtVectorType() && !Entity.getType()->isExtVectorType())
+            ? InitializedEntity::InitializeTemporary(ElemType)
+            : Entity;
+
+    InitializationSequence Seq(SemaRef, TmpEntity, Kind, expr,
                                /*TopLevelOfInitList*/ true);
 
     // C++14 [dcl.init.aggr]p13:
@@ -1300,8 +1309,7 @@ void InitListChecker::CheckSubElementType(const InitializedEntity &Entity,
     // assignment-expression.
     if (Seq || isa<InitListExpr>(expr)) {
       if (!VerifyOnly) {
-        ExprResult Result =
-          Seq.Perform(SemaRef, Entity, Kind, expr);
+        ExprResult Result = Seq.Perform(SemaRef, TmpEntity, Kind, expr);
         if (Result.isInvalid())
           hadError = true;
 
@@ -8108,7 +8116,7 @@ ExprResult InitializationSequence::Perform(Sema &S,
       // argument passing.
       assert(Step->Type->isSamplerT() &&
              "Sampler initialization on non-sampler type.");
-      Expr *Init = CurInit.get();
+      Expr *Init = CurInit.get()->IgnoreParens();
       QualType SourceType = Init->getType();
       // Case 1
       if (Entity.isParameterKind()) {
diff --git a/lib/Sema/SemaOpenMP.cpp b/lib/Sema/SemaOpenMP.cpp
index bd68011c18b2..4ac87469bf44 100644
--- a/lib/Sema/SemaOpenMP.cpp
+++ b/lib/Sema/SemaOpenMP.cpp
@@ -139,6 +139,7 @@ private:
     /// clause, false otherwise.
     llvm::Optional<std::pair<const Expr *, OMPOrderedClause *>> OrderedRegion;
     unsigned AssociatedLoops = 1;
+    bool HasMutipleLoops = false;
     const Decl *PossiblyLoopCounter = nullptr;
     bool NowaitRegion = false;
     bool CancelRegion = false;
@@ -678,12 +679,19 @@ public:
   /// Set collapse value for the region.
   void setAssociatedLoops(unsigned Val) {
     getTopOfStack().AssociatedLoops = Val;
+    if (Val > 1)
+      getTopOfStack().HasMutipleLoops = true;
   }
   /// Return collapse value for region.
   unsigned getAssociatedLoops() const {
     const SharingMapTy *Top = getTopOfStackOrNull();
     return Top ? Top->AssociatedLoops : 0;
   }
+  /// Returns true if the construct is associated with multiple loops.
+  bool hasMutipleLoops() const {
+    const SharingMapTy *Top = getTopOfStackOrNull();
+    return Top ? Top->HasMutipleLoops : false;
+  }
 
   /// Marks current target region as one with closely nested teams
   /// region.
@@ -1874,6 +1882,13 @@ bool Sema::isOpenMPPrivateDecl(const ValueDecl *D, unsigned Level) const {
         !isOpenMPSimdDirective(DSAStack->getCurrentDirective()))
       return true;
   }
+  if (const auto *VD = dyn_cast<VarDecl>(D)) {
+    if (DSAStack->isThreadPrivate(const_cast<VarDecl *>(VD)) &&
+        DSAStack->isForceVarCapturing() &&
+        !DSAStack->hasExplicitDSA(
+            D, [](OpenMPClauseKind K) { return K == OMPC_copyin; }, Level))
+      return true;
+  }
   return DSAStack->hasExplicitDSA(
              D, [](OpenMPClauseKind K) { return K == OMPC_private; }, Level) ||
          (DSAStack->isClauseParsingMode() &&
@@ -5604,13 +5619,14 @@ void Sema::ActOnOpenMPLoopInitialization(SourceLocation ForLoc, Stmt *Init) {
     if (!ISC.checkAndSetInit(Init, /*EmitDiags=*/false)) {
       if (ValueDecl *D = ISC.getLoopDecl()) {
         auto *VD = dyn_cast<VarDecl>(D);
+        DeclRefExpr *PrivateRef = nullptr;
         if (!VD) {
           if (VarDecl *Private = isOpenMPCapturedDecl(D)) {
             VD = Private;
           } else {
-            DeclRefExpr *Ref = buildCapture(*this, D, ISC.getLoopDeclRefExpr(),
-                                            /*WithInit=*/false);
-            VD = cast<VarDecl>(Ref->getDecl());
+            PrivateRef = buildCapture(*this, D, ISC.getLoopDeclRefExpr(),
+                                      /*WithInit=*/false);
+            VD = cast<VarDecl>(PrivateRef->getDecl());
           }
         }
         DSAStack->addLoopControlVariable(D, VD);
@@ -5623,6 +5639,49 @@ void Sema::ActOnOpenMPLoopInitialization(SourceLocation ForLoc, Stmt *Init) {
                                  Var->getType().getNonLValueExprType(Context),
                                  ForLoc, /*RefersToCapture=*/true));
         }
+        OpenMPDirectiveKind DKind = DSAStack->getCurrentDirective();
+        // OpenMP [2.14.1.1, Data-sharing Attribute Rules for Variables
+        // Referenced in a Construct, C/C++]. The loop iteration variable in the
+        // associated for-loop of a simd construct with just one associated
+        // for-loop may be listed in a linear clause with a constant-linear-step
+        // that is the increment of the associated for-loop. The loop iteration
+        // variable(s) in the associated for-loop(s) of a for or parallel for
+        // construct may be listed in a private or lastprivate clause.
+        DSAStackTy::DSAVarData DVar =
+            DSAStack->getTopDSA(D, /*FromParent=*/false);
+        // If LoopVarRefExpr is nullptr it means the corresponding loop variable
+        // is declared in the loop and it is predetermined as a private.
+        Expr *LoopDeclRefExpr = ISC.getLoopDeclRefExpr();
+        OpenMPClauseKind PredeterminedCKind =
+            isOpenMPSimdDirective(DKind)
+                ? (DSAStack->hasMutipleLoops() ? OMPC_lastprivate : OMPC_linear)
+                : OMPC_private;
+        if (((isOpenMPSimdDirective(DKind) && DVar.CKind != OMPC_unknown &&
+              DVar.CKind != PredeterminedCKind && DVar.RefExpr &&
+              (LangOpts.OpenMP <= 45 || (DVar.CKind != OMPC_lastprivate &&
+                                         DVar.CKind != OMPC_private))) ||
+             ((isOpenMPWorksharingDirective(DKind) || DKind == OMPD_taskloop ||
+               isOpenMPDistributeDirective(DKind)) &&
+              !isOpenMPSimdDirective(DKind) && DVar.CKind != OMPC_unknown &&
+              DVar.CKind != OMPC_private && DVar.CKind != OMPC_lastprivate)) &&
+            (DVar.CKind != OMPC_private || DVar.RefExpr)) {
+          Diag(Init->getBeginLoc(), diag::err_omp_loop_var_dsa)
+              << getOpenMPClauseName(DVar.CKind)
+              << getOpenMPDirectiveName(DKind)
+              << getOpenMPClauseName(PredeterminedCKind);
+          if (DVar.RefExpr == nullptr)
+            DVar.CKind = PredeterminedCKind;
+          reportOriginalDsa(*this, DSAStack, D, DVar,
+                            /*IsLoopIterVar=*/true);
+        } else if (LoopDeclRefExpr) {
+          // Make the loop iteration variable private (for worksharing
+          // constructs), linear (for simd directives with the only one
+          // associated loop) or lastprivate (for simd directives with several
+          // collapsed or ordered loops).
+          if (DVar.CKind == OMPC_unknown)
+            DSAStack->addDSA(D, LoopDeclRefExpr, PredeterminedCKind,
+                             PrivateRef);
+        }
       }
     }
     DSAStack->setAssociatedLoops(AssociatedLoops - 1);
@@ -5677,8 +5736,6 @@ static bool checkOpenMPIterationSpace(
 
   // Check loop variable's type.
   if (ValueDecl *LCDecl = ISC.getLoopDecl()) {
-    Expr *LoopDeclRefExpr = ISC.getLoopDeclRefExpr();
-
     // OpenMP [2.6, Canonical Loop Form]
     // Var is one of the following:
     //   A variable of signed or unsigned integer type.
@@ -5704,46 +5761,6 @@ static bool checkOpenMPIterationSpace(
     // sharing attributes.
     VarsWithImplicitDSA.erase(LCDecl);
 
-    // OpenMP [2.14.1.1, Data-sharing Attribute Rules for Variables Referenced
-    // in a Construct, C/C++].
-    // The loop iteration variable in the associated for-loop of a simd
-    // construct with just one associated for-loop may be listed in a linear
-    // clause with a constant-linear-step that is the increment of the
-    // associated for-loop.
-    // The loop iteration variable(s) in the associated for-loop(s) of a for or
-    // parallel for construct may be listed in a private or lastprivate clause.
-    DSAStackTy::DSAVarData DVar = DSA.getTopDSA(LCDecl, false);
-    // If LoopVarRefExpr is nullptr it means the corresponding loop variable is
-    // declared in the loop and it is predetermined as a private.
-    OpenMPClauseKind PredeterminedCKind =
-        isOpenMPSimdDirective(DKind)
-            ? ((NestedLoopCount == 1) ? OMPC_linear : OMPC_lastprivate)
-            : OMPC_private;
-    if (((isOpenMPSimdDirective(DKind) && DVar.CKind != OMPC_unknown &&
-          DVar.CKind != PredeterminedCKind && DVar.RefExpr &&
-          (SemaRef.getLangOpts().OpenMP <= 45 ||
-           (DVar.CKind != OMPC_lastprivate && DVar.CKind != OMPC_private))) ||
-         ((isOpenMPWorksharingDirective(DKind) || DKind == OMPD_taskloop ||
-           isOpenMPDistributeDirective(DKind)) &&
-          !isOpenMPSimdDirective(DKind) && DVar.CKind != OMPC_unknown &&
-          DVar.CKind != OMPC_private && DVar.CKind != OMPC_lastprivate)) &&
-        (DVar.CKind != OMPC_private || DVar.RefExpr)) {
-      SemaRef.Diag(Init->getBeginLoc(), diag::err_omp_loop_var_dsa)
-          << getOpenMPClauseName(DVar.CKind) << getOpenMPDirectiveName(DKind)
-          << getOpenMPClauseName(PredeterminedCKind);
-      if (DVar.RefExpr == nullptr)
-        DVar.CKind = PredeterminedCKind;
-      reportOriginalDsa(SemaRef, &DSA, LCDecl, DVar, /*IsLoopIterVar=*/true);
-      HasErrors = true;
-    } else if (LoopDeclRefExpr != nullptr) {
-      // Make the loop iteration variable private (for worksharing constructs),
-      // linear (for simd directives with the only one associated loop) or
-      // lastprivate (for simd directives with several collapsed or ordered
-      // loops).
-      if (DVar.CKind == OMPC_unknown)
-        DSA.addDSA(LCDecl, LoopDeclRefExpr, PredeterminedCKind);
-    }
-
     assert(isOpenMPLoopDirective(DKind) && "DSA for non-loop vars");
 
     // Check test-expr.
diff --git a/lib/Sema/SemaStmtAsm.cpp b/lib/Sema/SemaStmtAsm.cpp
index b123a739a7ab..9b051e02d127 100644
--- a/lib/Sema/SemaStmtAsm.cpp
+++ b/lib/Sema/SemaStmtAsm.cpp
@@ -383,25 +383,19 @@ StmtResult Sema::ActOnGCCAsmStmt(SourceLocation AsmLoc, bool IsSimple,
     } else if (Info.requiresImmediateConstant() && !Info.allowsRegister()) {
       if (!InputExpr->isValueDependent()) {
         Expr::EvalResult EVResult;
-        if (!InputExpr->EvaluateAsRValue(EVResult, Context, true))
-          return StmtError(
-              Diag(InputExpr->getBeginLoc(), diag::err_asm_immediate_expected)
-              << Info.getConstraintStr() << InputExpr->getSourceRange());
-
-        // For compatibility with GCC, we also allow pointers that would be
-        // integral constant expressions if they were cast to int.
-        llvm::APSInt IntResult;
-        if (!EVResult.Val.toIntegralConstant(IntResult, InputExpr->getType(),
-                                             Context))
-          return StmtError(
-              Diag(InputExpr->getBeginLoc(), diag::err_asm_immediate_expected)
-              << Info.getConstraintStr() << InputExpr->getSourceRange());
-
-        if (!Info.isValidAsmImmediate(IntResult))
-          return StmtError(Diag(InputExpr->getBeginLoc(),
-                                diag::err_invalid_asm_value_for_constraint)
-                           << IntResult.toString(10) << Info.getConstraintStr()
-                           << InputExpr->getSourceRange());
+        if (InputExpr->EvaluateAsRValue(EVResult, Context, true)) {
+          // For compatibility with GCC, we also allow pointers that would be
+          // integral constant expressions if they were cast to int.
+          llvm::APSInt IntResult;
+          if (EVResult.Val.toIntegralConstant(IntResult, InputExpr->getType(),
+                                               Context))
+            if (!Info.isValidAsmImmediate(IntResult))
+              return StmtError(Diag(InputExpr->getBeginLoc(),
+                                    diag::err_invalid_asm_value_for_constraint)
+                               << IntResult.toString(10)
+                               << Info.getConstraintStr()
+                               << InputExpr->getSourceRange());
+        }
       }
 
     } else {
diff --git a/lib/Sema/SemaTemplate.cpp b/lib/Sema/SemaTemplate.cpp
index 3212281cc34d..ec4b63a2e508 100644
--- a/lib/Sema/SemaTemplate.cpp
+++ b/lib/Sema/SemaTemplate.cpp
@@ -362,13 +362,27 @@ bool Sema::LookupTemplateName(LookupResult &Found,
     // x->B::f, and we are looking into the type of the object.
     assert(!SS.isSet() && "ObjectType and scope specifier cannot coexist");
     LookupCtx = computeDeclContext(ObjectType);
-    IsDependent = !LookupCtx;
+    IsDependent = !LookupCtx && ObjectType->isDependentType();
     assert((IsDependent || !ObjectType->isIncompleteType() ||
             ObjectType->castAs<TagType>()->isBeingDefined()) &&
            "Caller should have completed object type");
 
-    // Template names cannot appear inside an Objective-C class or object type.
-    if (ObjectType->isObjCObjectOrInterfaceType()) {
+    // Template names cannot appear inside an Objective-C class or object type
+    // or a vector type.
+    //
+    // FIXME: This is wrong. For example:
+    //
+    //   template<typename T> using Vec = T __attribute__((ext_vector_type(4)));
+    //   Vec<int> vi;
+    //   vi.Vec<int>::~Vec<int>();
+    //
+    // ... should be accepted but we will not treat 'Vec' as a template name
+    // here. The right thing to do would be to check if the name is a valid
+    // vector component name, and look up a template name if not. And similarly
+    // for lookups into Objective-C class and object types, where the same
+    // problem can arise.
+    if (ObjectType->isObjCObjectOrInterfaceType() ||
+        ObjectType->isVectorType()) {
       Found.clear();
       return false;
     }
diff --git a/lib/StaticAnalyzer/Checkers/IteratorChecker.cpp b/lib/StaticAnalyzer/Checkers/IteratorChecker.cpp
index 6f1060b5f26d..600458a743ea 100644
--- a/lib/StaticAnalyzer/Checkers/IteratorChecker.cpp
+++ b/lib/StaticAnalyzer/Checkers/IteratorChecker.cpp
@@ -406,13 +406,15 @@ void IteratorChecker::checkPreCall(const CallEvent &Call,
       } else if (isRandomIncrOrDecrOperator(Func->getOverloadedOperator())) {
         if (const auto *InstCall = dyn_cast<CXXInstanceCall>(&Call)) {
           // Check for out-of-range incrementions and decrementions
-          if (Call.getNumArgs() >= 1) {
+          if (Call.getNumArgs() >= 1 &&
+              Call.getArgExpr(0)->getType()->isIntegralOrEnumerationType()) {
             verifyRandomIncrOrDecr(C, Func->getOverloadedOperator(),
                                    InstCall->getCXXThisVal(),
                                    Call.getArgSVal(0));
           }
         } else {
-          if (Call.getNumArgs() >= 2) {
+          if (Call.getNumArgs() >= 2 &&
+              Call.getArgExpr(1)->getType()->isIntegralOrEnumerationType()) {
             verifyRandomIncrOrDecr(C, Func->getOverloadedOperator(),
                                    Call.getArgSVal(0), Call.getArgSVal(1));
           }
@@ -590,14 +592,16 @@ void IteratorChecker::checkPostCall(const CallEvent &Call,
       return;
     } else if (isRandomIncrOrDecrOperator(Func->getOverloadedOperator())) {
       if (const auto *InstCall = dyn_cast<CXXInstanceCall>(&Call)) {
-        if (Call.getNumArgs() >= 1) {
+        if (Call.getNumArgs() >= 1 &&
+              Call.getArgExpr(0)->getType()->isIntegralOrEnumerationType()) {
           handleRandomIncrOrDecr(C, Func->getOverloadedOperator(),
                                  Call.getReturnValue(),
                                  InstCall->getCXXThisVal(), Call.getArgSVal(0));
           return;
         }
       } else {
-        if (Call.getNumArgs() >= 2) {
+        if (Call.getNumArgs() >= 2 &&
+              Call.getArgExpr(1)->getType()->isIntegralOrEnumerationType()) {
           handleRandomIncrOrDecr(C, Func->getOverloadedOperator(),
                                  Call.getReturnValue(), Call.getArgSVal(0),
                                  Call.getArgSVal(1));