diff --git a/clang/lib/Sema/SemaExprCXX.cpp b/clang/lib/Sema/SemaExprCXX.cpp
@@ -1415,42 +1415,26 @@ bool Sema::CheckCXXThisCapture(SourceLocation Loc, const bool Explicit,
 }
 
 ExprResult Sema::ActOnCXXThis(SourceLocation Loc) {
-  // C++20 [expr.prim.this]p1:
-  //   The keyword this names a pointer to the object for which an
-  //   implicit object member function is invoked or a non-static
-  //   data member's initializer is evaluated.
+  /// C++ 9.3.2: In the body of a non-static member function, the keyword this
+  /// is a non-lvalue expression whose value is the address of the object for
+  /// which the function is called.
   QualType ThisTy = getCurrentThisType();
 
-  if (CheckCXXThisType(Loc, ThisTy))
-    return ExprError();
+  if (ThisTy.isNull()) {
+    DeclContext *DC = getFunctionLevelDeclContext();
 
-  return BuildCXXThisExpr(Loc, ThisTy, /*IsImplicit=*/false);
-}
+    if (const auto *Method = dyn_cast<CXXMethodDecl>(DC);
+        Method && Method->isExplicitObjectMemberFunction()) {
+      return Diag(Loc, diag::err_invalid_this_use) << 1;
+    }
 
-bool Sema::CheckCXXThisType(SourceLocation Loc, QualType Type) {
-  if (!Type.isNull())
-    return false;
+    if (isLambdaCallWithExplicitObjectParameter(CurContext))
+      return Diag(Loc, diag::err_invalid_this_use) << 1;
 
-  // C++20 [expr.prim.this]p3:
-  //   If a declaration declares a member function or member function template
-  //   of a class X, the expression this is a prvalue of type
-  //   "pointer to cv-qualifier-seq X" wherever X is the current class between
-  //   the optional cv-qualifier-seq and the end of the function-definition,
-  //   member-declarator, or declarator. It shall not appear within the
-  //   declaration of either a static member function or an explicit object
-  //   member function of the current class (although its type and value
-  //   category are defined within such member functions as they are within
-  //   an implicit object member function).
-  DeclContext *DC = getFunctionLevelDeclContext();
-  if (const auto *Method = dyn_cast<CXXMethodDecl>(DC);
-      Method && Method->isExplicitObjectMemberFunction()) {
-    Diag(Loc, diag::err_invalid_this_use) << 1;
-  } else if (isLambdaCallWithExplicitObjectParameter(CurContext)) {
-    Diag(Loc, diag::err_invalid_this_use) << 1;
-  } else {
-    Diag(Loc, diag::err_invalid_this_use) << 0;
+    return Diag(Loc, diag::err_invalid_this_use) << 0;
   }
-  return true;
+
+  return BuildCXXThisExpr(Loc, ThisTy, /*IsImplicit=*/false);
 }
 
 Expr *Sema::BuildCXXThisExpr(SourceLocation Loc, QualType Type,
@@ -5895,7 +5879,8 @@ static bool EvaluateBinaryTypeTrait(Sema &Self, TypeTrait BTT, const TypeSourceI
         return false;
 
       if (Self.RequireCompleteType(
-              KeyLoc, RhsT, diag::err_incomplete_type_used_in_type_trait_expr))
+              Rhs->getTypeLoc().getBeginLoc(), RhsT,
+              diag::err_incomplete_type_used_in_type_trait_expr))
         return false;
 
       return BaseInterface->isSuperClassOf(DerivedInterface);
@@ -5918,8 +5903,9 @@ static bool EvaluateBinaryTypeTrait(Sema &Self, TypeTrait BTT, const TypeSourceI
     //   If Base and Derived are class types and are different types
     //   (ignoring possible cv-qualifiers) then Derived shall be a
     //   complete type.
-    if (Self.RequireCompleteType(KeyLoc, RhsT,
-                          diag::err_incomplete_type_used_in_type_trait_expr))
+    if (Self.RequireCompleteType(
+            Rhs->getTypeLoc().getBeginLoc(), RhsT,
+            diag::err_incomplete_type_used_in_type_trait_expr))
       return false;
 
     return cast<CXXRecordDecl>(rhsRecord->getDecl())
@@ -5971,7 +5957,8 @@ static bool EvaluateBinaryTypeTrait(Sema &Self, TypeTrait BTT, const TypeSourceI
       return LhsT->isVoidType();
 
     // A function definition requires a complete, non-abstract return type.
-    if (!Self.isCompleteType(KeyLoc, RhsT) || Self.isAbstractType(KeyLoc, RhsT))
+    if (!Self.isCompleteType(Rhs->getTypeLoc().getBeginLoc(), RhsT) ||
+        Self.isAbstractType(Rhs->getTypeLoc().getBeginLoc(), RhsT))
       return false;
 
     // Compute the result of add_rvalue_reference.
@@ -6021,12 +6008,14 @@ static bool EvaluateBinaryTypeTrait(Sema &Self, TypeTrait BTT, const TypeSourceI
     //   For both, T and U shall be complete types, (possibly cv-qualified)
     //   void, or arrays of unknown bound.
     if (!LhsT->isVoidType() && !LhsT->isIncompleteArrayType() &&
-        Self.RequireCompleteType(KeyLoc, LhsT,
-          diag::err_incomplete_type_used_in_type_trait_expr))
+        Self.RequireCompleteType(
+            Lhs->getTypeLoc().getBeginLoc(), LhsT,
+            diag::err_incomplete_type_used_in_type_trait_expr))
       return false;
     if (!RhsT->isVoidType() && !RhsT->isIncompleteArrayType() &&
-        Self.RequireCompleteType(KeyLoc, RhsT,
-          diag::err_incomplete_type_used_in_type_trait_expr))
+        Self.RequireCompleteType(
+            Rhs->getTypeLoc().getBeginLoc(), RhsT,
+            diag::err_incomplete_type_used_in_type_trait_expr))
       return false;
 
     // cv void is never assignable.
@@ -6081,12 +6070,17 @@ static bool EvaluateBinaryTypeTrait(Sema &Self, TypeTrait BTT, const TypeSourceI
   }
   case BTT_IsLayoutCompatible: {
     if (!LhsT->isVoidType() && !LhsT->isIncompleteArrayType())
-      Self.RequireCompleteType(KeyLoc, LhsT, diag::err_incomplete_type);
+      Self.RequireCompleteType(Lhs->getTypeLoc().getBeginLoc(), LhsT,
+                               diag::err_incomplete_type);
     if (!RhsT->isVoidType() && !RhsT->isIncompleteArrayType())
-      Self.RequireCompleteType(KeyLoc, RhsT, diag::err_incomplete_type);
+      Self.RequireCompleteType(Rhs->getTypeLoc().getBeginLoc(), RhsT,
+                               diag::err_incomplete_type);
 
-    if (LhsT->isVariableArrayType() || RhsT->isVariableArrayType())
-      Self.Diag(KeyLoc, diag::err_vla_unsupported)
+    if (LhsT->isVariableArrayType())
+      Self.Diag(Lhs->getTypeLoc().getBeginLoc(), diag::err_vla_unsupported)
+          << 1 << tok::kw___is_layout_compatible;
+    if (RhsT->isVariableArrayType())
+      Self.Diag(Rhs->getTypeLoc().getBeginLoc(), diag::err_vla_unsupported)
           << 1 << tok::kw___is_layout_compatible;
     return Self.IsLayoutCompatible(LhsT, RhsT);
   }

diff --git a/clang/lib/Sema/SemaExprMember.cpp b/clang/lib/Sema/SemaExprMember.cpp
@@ -61,10 +61,6 @@ enum IMAKind {
   /// The reference is a contextually-permitted abstract member reference.
   IMA_Abstract,
 
-  /// Whether the context is static is dependent on the enclosing template (i.e.
-  /// in a dependent class scope explicit specialization).
-  IMA_Dependent,
-
   /// The reference may be to an unresolved using declaration and the
   /// context is not an instance method.
   IMA_Unresolved_StaticOrExplicitContext,
@@ -95,18 +91,10 @@ static IMAKind ClassifyImplicitMemberAccess(Sema &SemaRef,
 
   DeclContext *DC = SemaRef.getFunctionLevelDeclContext();
 
-  bool couldInstantiateToStatic = false;
-  bool isStaticOrExplicitContext = SemaRef.CXXThisTypeOverride.isNull();
-
-  if (auto *MD = dyn_cast<CXXMethodDecl>(DC)) {
-    if (MD->isImplicitObjectMemberFunction()) {
-      isStaticOrExplicitContext = false;
-      // A dependent class scope function template explicit specialization
-      // that is neither declared 'static' nor with an explicit object
-      // parameter could instantiate to a static or non-static member function.
-      couldInstantiateToStatic = MD->getDependentSpecializationInfo();
-    }
-  }
+  bool isStaticOrExplicitContext =
+      SemaRef.CXXThisTypeOverride.isNull() &&
+      (!isa<CXXMethodDecl>(DC) || cast<CXXMethodDecl>(DC)->isStatic() ||
+       cast<CXXMethodDecl>(DC)->isExplicitObjectMemberFunction());
 
   if (R.isUnresolvableResult())
     return isStaticOrExplicitContext ? IMA_Unresolved_StaticOrExplicitContext
@@ -135,9 +123,6 @@ static IMAKind ClassifyImplicitMemberAccess(Sema &SemaRef,
   if (Classes.empty())
     return IMA_Static;
 
-  if (couldInstantiateToStatic)
-    return IMA_Dependent;
-
   // C++11 [expr.prim.general]p12:
   //   An id-expression that denotes a non-static data member or non-static
   //   member function of a class can only be used:
@@ -283,30 +268,27 @@ ExprResult Sema::BuildPossibleImplicitMemberExpr(
     const CXXScopeSpec &SS, SourceLocation TemplateKWLoc, LookupResult &R,
     const TemplateArgumentListInfo *TemplateArgs, const Scope *S,
     UnresolvedLookupExpr *AsULE) {
-  switch (IMAKind Classification = ClassifyImplicitMemberAccess(*this, R)) {
+  switch (ClassifyImplicitMemberAccess(*this, R)) {
   case IMA_Instance:
+    return BuildImplicitMemberExpr(SS, TemplateKWLoc, R, TemplateArgs, true, S);
+
   case IMA_Mixed:
   case IMA_Mixed_Unrelated:
   case IMA_Unresolved:
-    return BuildImplicitMemberExpr(
-        SS, TemplateKWLoc, R, TemplateArgs,
-        /*IsKnownInstance=*/Classification == IMA_Instance, S);
+    return BuildImplicitMemberExpr(SS, TemplateKWLoc, R, TemplateArgs, false,
+                                   S);
+
   case IMA_Field_Uneval_Context:
     Diag(R.getNameLoc(), diag::warn_cxx98_compat_non_static_member_use)
       << R.getLookupNameInfo().getName();
     [[fallthrough]];
   case IMA_Static:
   case IMA_Abstract:
-  case IMA_Dependent:
   case IMA_Mixed_StaticOrExplicitContext:
   case IMA_Unresolved_StaticOrExplicitContext:
     if (TemplateArgs || TemplateKWLoc.isValid())
-      return BuildTemplateIdExpr(SS, TemplateKWLoc, R, /*RequiresADL=*/false,
-                                 TemplateArgs);
-    return AsULE ? AsULE
-                 : BuildDeclarationNameExpr(
-                       SS, R, /*NeedsADL=*/false, /*AcceptInvalidDecl=*/false,
-                       /*NeedUnresolved=*/Classification == IMA_Dependent);
+      return BuildTemplateIdExpr(SS, TemplateKWLoc, R, false, TemplateArgs);
+    return AsULE ? AsULE : BuildDeclarationNameExpr(SS, R, false);
 
   case IMA_Error_StaticOrExplicitContext:
   case IMA_Error_Unrelated:

diff --git a/clang/lib/Sema/SemaOpenACC.cpp b/clang/lib/Sema/SemaOpenACC.cpp
@@ -39,11 +39,27 @@ bool diagnoseConstructAppertainment(SemaOpenACC &S, OpenACCDirectiveKind K,
 
 bool doesClauseApplyToDirective(OpenACCDirectiveKind DirectiveKind,
                                 OpenACCClauseKind ClauseKind) {
-  // FIXME: For each clause as we implement them, we can add the
-  // 'legalization' list here.
-
-  // Do nothing so we can go to the 'unimplemented' diagnostic instead.
-  return true;
+  switch (ClauseKind) {
+    // FIXME: For each clause as we implement them, we can add the
+    // 'legalization' list here.
+  case OpenACCClauseKind::Default:
+    switch (DirectiveKind) {
+    case OpenACCDirectiveKind::Parallel:
+    case OpenACCDirectiveKind::Serial:
+    case OpenACCDirectiveKind::Kernels:
+    case OpenACCDirectiveKind::ParallelLoop:
+    case OpenACCDirectiveKind::SerialLoop:
+    case OpenACCDirectiveKind::KernelsLoop:
+    case OpenACCDirectiveKind::Data:
+      return true;
+    default:
+      return false;
+    }
+  default:
+    // Do nothing so we can go to the 'unimplemented' diagnostic instead.
+    return true;
+  }
+  llvm_unreachable("Invalid clause kind");
 }
 } // namespace
 
@@ -63,8 +79,43 @@ SemaOpenACC::ActOnClause(ArrayRef<const OpenACCClause *> ExistingClauses,
     return nullptr;
   }
 
-  // TODO OpenACC: Switch over the clauses we implement here and 'create'
-  // them.
+  switch (Clause.getClauseKind()) {
+  case OpenACCClauseKind::Default: {
+    // Restrictions only properly implemented on 'compute' constructs, and
+    // 'compute' constructs are the only construct that can do anything with
+    // this yet, so skip/treat as unimplemented in this case.
+    if (Clause.getDirectiveKind() != OpenACCDirectiveKind::Parallel &&
+        Clause.getDirectiveKind() != OpenACCDirectiveKind::Serial &&
+        Clause.getDirectiveKind() != OpenACCDirectiveKind::Kernels)
+      break;
+
+    // Don't add an invalid clause to the AST.
+    if (Clause.getDefaultClauseKind() == OpenACCDefaultClauseKind::Invalid)
+      return nullptr;
+
+    // OpenACC 3.3, Section 2.5.4:
+    // At most one 'default' clause may appear, and it must have a value of
+    // either 'none' or 'present'.
+    // Second half of the sentence is diagnosed during parsing.
+    auto Itr = llvm::find_if(ExistingClauses, [](const OpenACCClause *C) {
+      return C->getClauseKind() == OpenACCClauseKind::Default;
+    });
+
+    if (Itr != ExistingClauses.end()) {
+      Diag(Clause.getBeginLoc(),
+                   diag::err_acc_duplicate_clause_disallowed)
+          << Clause.getDirectiveKind() << Clause.getClauseKind();
+      Diag((*Itr)->getBeginLoc(), diag::note_acc_previous_clause_here);
+      return nullptr;
+    }
+
+    return OpenACCDefaultClause::Create(
+        getASTContext(), Clause.getDefaultClauseKind(), Clause.getBeginLoc(),
+        Clause.getLParenLoc(), Clause.getEndLoc());
+  }
+  default:
+    break;
+  }
 
   Diag(Clause.getBeginLoc(), diag::warn_acc_clause_unimplemented)
       << Clause.getClauseKind();

diff --git a/clang/lib/Sema/SemaOpenMP.cpp b/clang/lib/Sema/SemaOpenMP.cpp
@@ -4478,6 +4478,8 @@ void Sema::ActOnOpenMPRegionStart(OpenMPDirectiveKind DKind, Scope *CurScope) {
                              Params);
     break;
   }
+  // For 'target teams loop', collect all captured regions so codegen can
+  // later decide the best IR to emit given the associated loop-nest.
   case OMPD_target_teams_loop:
   case OMPD_target_teams_distribute_parallel_for:
   case OMPD_target_teams_distribute_parallel_for_simd: {
@@ -6135,6 +6137,79 @@ processImplicitMapsWithDefaultMappers(Sema &S, DSAStackTy *Stack,
   }
 }
 
+namespace {
+/// A 'teams loop' with a nested 'loop bind(parallel)' or generic function
+/// call in the associated loop-nest cannot be a 'parallel for'.
+class TeamsLoopChecker final : public ConstStmtVisitor<TeamsLoopChecker> {
+  Sema &SemaRef;
+
+public:
+  bool teamsLoopCanBeParallelFor() const { return TeamsLoopCanBeParallelFor; }
+
+  // Is there a nested OpenMP loop bind(parallel)
+  void VisitOMPExecutableDirective(const OMPExecutableDirective *D) {
+    if (D->getDirectiveKind() == llvm::omp::Directive::OMPD_loop) {
+      if (const auto *C = D->getSingleClause<OMPBindClause>())
+        if (C->getBindKind() == OMPC_BIND_parallel) {
+          TeamsLoopCanBeParallelFor = false;
+          // No need to continue visiting any more
+          return;
+        }
+    }
+    for (const Stmt *Child : D->children())
+      if (Child)
+        Visit(Child);
+  }
+
+  void VisitCallExpr(const CallExpr *C) {
+    // Function calls inhibit parallel loop translation of 'target teams loop'
+    // unless the assume-no-nested-parallelism flag has been specified.
+    // OpenMP API runtime library calls do not inhibit parallel loop
+    // translation, regardless of the assume-no-nested-parallelism.
+    if (C) {
+      bool IsOpenMPAPI = false;
+      auto *FD = dyn_cast_or_null<FunctionDecl>(C->getCalleeDecl());
+      if (FD) {
+        std::string Name = FD->getNameInfo().getAsString();
+        IsOpenMPAPI = Name.find("omp_") == 0;
+      }
+      TeamsLoopCanBeParallelFor =
+          IsOpenMPAPI || SemaRef.getLangOpts().OpenMPNoNestedParallelism;
+      if (!TeamsLoopCanBeParallelFor)
+        return;
+    }
+    for (const Stmt *Child : C->children())
+      if (Child)
+        Visit(Child);
+  }
+
+  void VisitCapturedStmt(const CapturedStmt *S) {
+    if (!S)
+      return;
+    Visit(S->getCapturedDecl()->getBody());
+  }
+
+  void VisitStmt(const Stmt *S) {
+    if (!S)
+      return;
+    for (const Stmt *Child : S->children())
+      if (Child)
+        Visit(Child);
+  }
+  explicit TeamsLoopChecker(Sema &SemaRef)
+      : SemaRef(SemaRef), TeamsLoopCanBeParallelFor(true) {}
+
+private:
+  bool TeamsLoopCanBeParallelFor;
+};
+} // namespace
+
+static bool teamsLoopCanBeParallelFor(Stmt *AStmt, Sema &SemaRef) {
+  TeamsLoopChecker Checker(SemaRef);
+  Checker.Visit(AStmt);
+  return Checker.teamsLoopCanBeParallelFor();
+}
+
 bool Sema::mapLoopConstruct(llvm::SmallVector<OMPClause *> &ClausesWithoutBind,
                             ArrayRef<OMPClause *> Clauses,
                             OpenMPBindClauseKind &BindKind,
@@ -10895,7 +10970,8 @@ StmtResult Sema::ActOnOpenMPTargetTeamsGenericLoopDirective(
   setFunctionHasBranchProtectedScope();
 
   return OMPTargetTeamsGenericLoopDirective::Create(
-      Context, StartLoc, EndLoc, NestedLoopCount, Clauses, AStmt, B);
+      Context, StartLoc, EndLoc, NestedLoopCount, Clauses, AStmt, B,
+      teamsLoopCanBeParallelFor(AStmt, *this));
 }
 
 StmtResult Sema::ActOnOpenMPParallelGenericLoopDirective(
@@ -15645,14 +15721,19 @@ static OpenMPDirectiveKind getOpenMPCaptureRegionForClause(
       if (NameModifier == OMPD_unknown || NameModifier == OMPD_parallel)
         CaptureRegion = OMPD_target;
       break;
+    case OMPD_teams_loop:
+    case OMPD_target_teams_loop:
+      // For [target] teams loop, assume capture region is 'teams' so it's
+      // available for codegen later to use if/when necessary.
+      CaptureRegion = OMPD_teams;
+      break;
     case OMPD_target_teams_distribute_parallel_for_simd:
       if (OpenMPVersion >= 50 &&
           (NameModifier == OMPD_unknown || NameModifier == OMPD_simd)) {
         CaptureRegion = OMPD_parallel;
         break;
       }
       [[fallthrough]];
-    case OMPD_target_teams_loop:
     case OMPD_target_teams_distribute_parallel_for:
       // If this clause applies to the nested 'parallel' region, capture within
       // the 'teams' region, otherwise do not capture.
@@ -15775,7 +15856,6 @@ static OpenMPDirectiveKind getOpenMPCaptureRegionForClause(
     case OMPD_declare_target:
     case OMPD_end_declare_target:
     case OMPD_loop:
-    case OMPD_teams_loop:
     case OMPD_teams:
     case OMPD_tile:
     case OMPD_unroll:

diff --git a/clang/lib/Sema/SemaSYCL.cpp b/clang/lib/Sema/SemaSYCL.cpp
@@ -8,6 +8,7 @@
 // This implements Semantic Analysis for SYCL constructs.
 //===----------------------------------------------------------------------===//
 
+#include "clang/Sema/SemaSYCL.h"
 #include "clang/AST/Mangle.h"
 #include "clang/Sema/Sema.h"
 #include "clang/Sema/SemaDiagnostic.h"
@@ -18,28 +19,30 @@ using namespace clang;
 // SYCL device specific diagnostics implementation
 // -----------------------------------------------------------------------------
 
-Sema::SemaDiagnosticBuilder Sema::SYCLDiagIfDeviceCode(SourceLocation Loc,
+SemaSYCL::SemaSYCL(Sema &S) : SemaBase(S) {}
+
+Sema::SemaDiagnosticBuilder SemaSYCL::DiagIfDeviceCode(SourceLocation Loc,
                                                        unsigned DiagID) {
   assert(getLangOpts().SYCLIsDevice &&
          "Should only be called during SYCL compilation");
-  FunctionDecl *FD = dyn_cast<FunctionDecl>(getCurLexicalContext());
+  FunctionDecl *FD = dyn_cast<FunctionDecl>(SemaRef.getCurLexicalContext());
   SemaDiagnosticBuilder::Kind DiagKind = [this, FD] {
     if (!FD)
       return SemaDiagnosticBuilder::K_Nop;
-    if (getEmissionStatus(FD) == Sema::FunctionEmissionStatus::Emitted)
+    if (SemaRef.getEmissionStatus(FD) == Sema::FunctionEmissionStatus::Emitted)
       return SemaDiagnosticBuilder::K_ImmediateWithCallStack;
     return SemaDiagnosticBuilder::K_Deferred;
   }();
-  return SemaDiagnosticBuilder(DiagKind, Loc, DiagID, FD, *this);
+  return SemaDiagnosticBuilder(DiagKind, Loc, DiagID, FD, SemaRef);
 }
 
-static bool isZeroSizedArray(Sema &SemaRef, QualType Ty) {
-  if (const auto *CAT = SemaRef.getASTContext().getAsConstantArrayType(Ty))
+static bool isZeroSizedArray(SemaSYCL &S, QualType Ty) {
+  if (const auto *CAT = S.getASTContext().getAsConstantArrayType(Ty))
     return CAT->isZeroSize();
   return false;
 }
 
-void Sema::deepTypeCheckForSYCLDevice(SourceLocation UsedAt,
+void SemaSYCL::deepTypeCheckForDevice(SourceLocation UsedAt,
                                       llvm::DenseSet<QualType> Visited,
                                       ValueDecl *DeclToCheck) {
   assert(getLangOpts().SYCLIsDevice &&
@@ -51,18 +54,18 @@ void Sema::deepTypeCheckForSYCLDevice(SourceLocation UsedAt,
   auto Check = [&](QualType TypeToCheck, const ValueDecl *D) {
     bool ErrorFound = false;
     if (isZeroSizedArray(*this, TypeToCheck)) {
-      SYCLDiagIfDeviceCode(UsedAt, diag::err_typecheck_zero_array_size) << 1;
+      DiagIfDeviceCode(UsedAt, diag::err_typecheck_zero_array_size) << 1;
       ErrorFound = true;
     }
     // Checks for other types can also be done here.
     if (ErrorFound) {
       if (NeedToEmitNotes) {
         if (auto *FD = dyn_cast<FieldDecl>(D))
-          SYCLDiagIfDeviceCode(FD->getLocation(),
-                               diag::note_illegal_field_declared_here)
+          DiagIfDeviceCode(FD->getLocation(),
+                           diag::note_illegal_field_declared_here)
               << FD->getType()->isPointerType() << FD->getType();
         else
-          SYCLDiagIfDeviceCode(D->getLocation(), diag::note_declared_at);
+          DiagIfDeviceCode(D->getLocation(), diag::note_declared_at);
       }
     }
 
@@ -93,8 +96,8 @@ void Sema::deepTypeCheckForSYCLDevice(SourceLocation UsedAt,
     auto EmitHistory = [&]() {
       // The first element is always nullptr.
       for (uint64_t Index = 1; Index < History.size(); ++Index) {
-        SYCLDiagIfDeviceCode(History[Index]->getLocation(),
-                             diag::note_within_field_of_type)
+        DiagIfDeviceCode(History[Index]->getLocation(),
+                         diag::note_within_field_of_type)
             << History[Index]->getType();
       }
     };
@@ -130,3 +133,26 @@ void Sema::deepTypeCheckForSYCLDevice(SourceLocation UsedAt,
     }
   } while (!StackForRecursion.empty());
 }
+
+ExprResult SemaSYCL::BuildUniqueStableNameExpr(SourceLocation OpLoc,
+                                               SourceLocation LParen,
+                                               SourceLocation RParen,
+                                               TypeSourceInfo *TSI) {
+  return SYCLUniqueStableNameExpr::Create(getASTContext(), OpLoc, LParen,
+                                          RParen, TSI);
+}
+
+ExprResult SemaSYCL::ActOnUniqueStableNameExpr(SourceLocation OpLoc,
+                                               SourceLocation LParen,
+                                               SourceLocation RParen,
+                                               ParsedType ParsedTy) {
+  TypeSourceInfo *TSI = nullptr;
+  QualType Ty = SemaRef.GetTypeFromParser(ParsedTy, &TSI);
+
+  if (Ty.isNull())
+    return ExprError();
+  if (!TSI)
+    TSI = getASTContext().getTrivialTypeSourceInfo(Ty, LParen);
+
+  return BuildUniqueStableNameExpr(OpLoc, LParen, RParen, TSI);
+}
diff --git a/clang/lib/Sema/SemaTemplateInstantiateDecl.cpp b/clang/lib/Sema/SemaTemplateInstantiateDecl.cpp
@@ -5093,14 +5093,6 @@ void Sema::InstantiateFunctionDefinition(SourceLocation PointOfInstantiation,
   EnterExpressionEvaluationContext EvalContext(
       *this, Sema::ExpressionEvaluationContext::PotentiallyEvaluated);
 
-  Qualifiers ThisTypeQuals;
-  CXXRecordDecl *ThisContext = nullptr;
-  if (CXXMethodDecl *Method = dyn_cast<CXXMethodDecl>(Function)) {
-    ThisContext = Method->getParent();
-    ThisTypeQuals = Method->getMethodQualifiers();
-  }
-  CXXThisScopeRAII ThisScope(*this, ThisContext, ThisTypeQuals);
-
   // Introduce a new scope where local variable instantiations will be
   // recorded, unless we're actually a member function within a local
   // class, in which case we need to merge our results with the parent

diff --git a/clang/lib/Sema/SemaTemplateVariadic.cpp b/clang/lib/Sema/SemaTemplateVariadic.cpp
@@ -1243,6 +1243,17 @@ std::optional<unsigned> Sema::getFullyPackExpandedSize(TemplateArgument Arg) {
     // expanded this pack expansion into the enclosing pack if we could.
     if (Elem.isPackExpansion())
       return std::nullopt;
+    // Don't guess the size of unexpanded packs. The pack within a template
+    // argument may have yet to be of a PackExpansion type before we see the
+    // ellipsis in the annotation stage.
+    //
+    // This doesn't mean we would invalidate the optimization: Arg can be an
+    // unexpanded pack regardless of Elem's dependence. For instance,
+    // A TemplateArgument that contains either a SubstTemplateTypeParmPackType
+    // or SubstNonTypeTemplateParmPackExpr is always considered Unexpanded, but
+    // the underlying TemplateArgument thereof may not.
+    if (Elem.containsUnexpandedParameterPack())
+      return std::nullopt;
   }
   return Pack.pack_size();
 }

diff --git a/clang/lib/Sema/TreeTransform.h b/clang/lib/Sema/TreeTransform.h
@@ -40,6 +40,7 @@
 #include "clang/Sema/SemaDiagnostic.h"
 #include "clang/Sema/SemaInternal.h"
 #include "clang/Sema/SemaOpenACC.h"
+#include "clang/Sema/SemaSYCL.h"
 #include "llvm/ADT/ArrayRef.h"
 #include "llvm/Support/ErrorHandling.h"
 #include <algorithm>
@@ -2632,7 +2633,8 @@ class TreeTransform {
                                              SourceLocation LParen,
                                              SourceLocation RParen,
                                              TypeSourceInfo *TSI) {
-    return getSema().BuildSYCLUniqueStableNameExpr(OpLoc, LParen, RParen, TSI);
+    return getSema().SYCL().BuildUniqueStableNameExpr(OpLoc, LParen, RParen,
+                                                      TSI);
   }
 
   /// Build a new predefined expression.
@@ -3307,13 +3309,12 @@ class TreeTransform {
 
   /// Build a new C++ "this" expression.
   ///
-  /// By default, performs semantic analysis to build a new "this" expression.
-  /// Subclasses may override this routine to provide different behavior.
+  /// By default, builds a new "this" expression without performing any
+  /// semantic analysis. Subclasses may override this routine to provide
+  /// different behavior.
   ExprResult RebuildCXXThisExpr(SourceLocation ThisLoc,
                                 QualType ThisType,
                                 bool isImplicit) {
-    if (getSema().CheckCXXThisType(ThisLoc, ThisType))
-      return ExprError();
     return getSema().BuildCXXThisExpr(ThisLoc, ThisType, isImplicit);
   }
 
@@ -4034,6 +4035,11 @@ class TreeTransform {
   llvm::SmallVector<OpenACCClause *>
   TransformOpenACCClauseList(OpenACCDirectiveKind DirKind,
                              ArrayRef<const OpenACCClause *> OldClauses);
+
+  OpenACCClause *
+  TransformOpenACCClause(ArrayRef<const OpenACCClause *> ExistingClauses,
+                         OpenACCDirectiveKind DirKind,
+                         const OpenACCClause *OldClause);
 };
 
 template <typename Derived>
@@ -11074,13 +11080,44 @@ OMPClause *TreeTransform<Derived>::TransformOMPXBareClause(OMPXBareClause *C) {
 //===----------------------------------------------------------------------===//
 // OpenACC transformation
 //===----------------------------------------------------------------------===//
+template <typename Derived>
+OpenACCClause *TreeTransform<Derived>::TransformOpenACCClause(
+    ArrayRef<const OpenACCClause *> ExistingClauses,
+    OpenACCDirectiveKind DirKind, const OpenACCClause *OldClause) {
+
+  SemaOpenACC::OpenACCParsedClause ParsedClause(
+      DirKind, OldClause->getClauseKind(), OldClause->getBeginLoc());
+  ParsedClause.setEndLoc(OldClause->getEndLoc());
+
+  if (const auto *WithParms = dyn_cast<OpenACCClauseWithParams>(OldClause))
+    ParsedClause.setLParenLoc(WithParms->getLParenLoc());
+
+  switch (OldClause->getClauseKind()) {
+  case OpenACCClauseKind::Default:
+    // There is nothing to do here as nothing dependent can appear in this
+    // clause. So just set the values so Sema can set the right value.
+    ParsedClause.setDefaultDetails(
+        cast<OpenACCDefaultClause>(OldClause)->getDefaultClauseKind());
+    break;
+  default:
+    assert(false && "Unhandled OpenACC clause in TreeTransform");
+    return nullptr;
+  }
+
+  return getSema().OpenACC().ActOnClause(ExistingClauses, ParsedClause);
+}
+
 template <typename Derived>
 llvm::SmallVector<OpenACCClause *>
 TreeTransform<Derived>::TransformOpenACCClauseList(
     OpenACCDirectiveKind DirKind, ArrayRef<const OpenACCClause *> OldClauses) {
-  // TODO OpenACC: Ensure we loop through the list and transform the individual
-  // clauses.
-  return {};
+  llvm::SmallVector<OpenACCClause *> TransformedClauses;
+  for (const auto *Clause : OldClauses) {
+    if (OpenACCClause *TransformedClause = getDerived().TransformOpenACCClause(
+            TransformedClauses, DirKind, Clause))
+      TransformedClauses.push_back(TransformedClause);
+  }
+  return TransformedClauses;
 }
 
 template <typename Derived>

diff --git a/clang/lib/Serialization/ASTReader.cpp b/clang/lib/Serialization/ASTReader.cpp
@@ -6626,8 +6626,6 @@ void ASTReader::ReadPragmaDiagnosticMappings(DiagnosticsEngine &Diag) {
              "Invalid data, missing pragma diagnostic states");
       FileID FID = ReadFileID(F, Record, Idx);
       assert(FID.isValid() && "invalid FileID for transition");
-      // FIXME: Remove this once we don't need the side-effects.
-      (void)SourceMgr.getSLocEntryOrNull(FID);
       unsigned Transitions = Record[Idx++];
 
       // Note that we don't need to set up Parent/ParentOffset here, because
@@ -11756,14 +11754,16 @@ void ASTRecordReader::readOMPChildren(OMPChildren *Data) {
 
 OpenACCClause *ASTRecordReader::readOpenACCClause() {
   OpenACCClauseKind ClauseKind = readEnum<OpenACCClauseKind>();
-  // TODO OpenACC: We don't have these used anywhere, but eventually we should
-  // be constructing the Clauses with them, so these attributes can go away at
-  // that point.
-  [[maybe_unused]] SourceLocation BeginLoc = readSourceLocation();
-  [[maybe_unused]] SourceLocation EndLoc = readSourceLocation();
+  SourceLocation BeginLoc = readSourceLocation();
+  SourceLocation EndLoc = readSourceLocation();
 
   switch (ClauseKind) {
-  case OpenACCClauseKind::Default:
+  case OpenACCClauseKind::Default: {
+    SourceLocation LParenLoc = readSourceLocation();
+    OpenACCDefaultClauseKind DCK = readEnum<OpenACCDefaultClauseKind>();
+    return OpenACCDefaultClause::Create(getContext(), DCK, BeginLoc, LParenLoc,
+                                        EndLoc);
+  }
   case OpenACCClauseKind::Finalize:
   case OpenACCClauseKind::IfPresent:
   case OpenACCClauseKind::Seq:

diff --git a/clang/lib/Serialization/ASTReaderStmt.cpp b/clang/lib/Serialization/ASTReaderStmt.cpp
@@ -2769,6 +2769,7 @@ void ASTStmtReader::VisitOMPTeamsGenericLoopDirective(
 void ASTStmtReader::VisitOMPTargetTeamsGenericLoopDirective(
     OMPTargetTeamsGenericLoopDirective *D) {
   VisitOMPLoopDirective(D);
+  D->setCanBeParallelFor(Record.readBool());
 }
 
 void ASTStmtReader::VisitOMPParallelGenericLoopDirective(

diff --git a/clang/lib/Serialization/ASTWriter.cpp b/clang/lib/Serialization/ASTWriter.cpp
@@ -163,8 +163,13 @@ static TypeCode getTypeCodeForTypeClass(Type::TypeClass id) {
 
 namespace {
 
-std::set<const FileEntry *> GetAffectingModuleMaps(const Preprocessor &PP,
-                                                   Module *RootModule) {
+std::optional<std::set<const FileEntry *>>
+GetAffectingModuleMaps(const Preprocessor &PP, Module *RootModule) {
+  // Without implicit module map search, there's no good reason to know about
+  // any module maps that are not affecting.
+  if (!PP.getHeaderSearchInfo().getHeaderSearchOpts().ImplicitModuleMaps)
+    return std::nullopt;
+
   SmallVector<const Module *> ModulesToProcess{RootModule};
 
   const HeaderSearch &HS = PP.getHeaderSearchInfo();
@@ -4735,8 +4740,16 @@ void ASTWriter::computeNonAffectingInputFiles() {
     if (!Cache->OrigEntry)
       continue;
 
-    if (!isModuleMap(File.getFileCharacteristic()) ||
-        llvm::is_contained(AffectingModuleMaps, *Cache->OrigEntry))
+    // Don't prune anything other than module maps.
+    if (!isModuleMap(File.getFileCharacteristic()))
+      continue;
+
+    // Don't prune module maps if all are guaranteed to be affecting.
+    if (!AffectingModuleMaps)
+      continue;
+
+    // Don't prune module maps that are affecting.
+    if (llvm::is_contained(*AffectingModuleMaps, *Cache->OrigEntry))
       continue;
 
     IsSLocAffecting[I] = false;
@@ -7406,7 +7419,12 @@ void ASTRecordWriter::writeOpenACCClause(const OpenACCClause *C) {
   writeSourceLocation(C->getEndLoc());
 
   switch (C->getClauseKind()) {
-  case OpenACCClauseKind::Default:
+  case OpenACCClauseKind::Default: {
+    const auto *DC = cast<OpenACCDefaultClause>(C);
+    writeSourceLocation(DC->getLParenLoc());
+    writeEnum(DC->getDefaultClauseKind());
+    return;
+  }
   case OpenACCClauseKind::Finalize:
   case OpenACCClauseKind::IfPresent:
   case OpenACCClauseKind::Seq:

diff --git a/clang/lib/Serialization/ASTWriterStmt.cpp b/clang/lib/Serialization/ASTWriterStmt.cpp
@@ -2821,6 +2821,7 @@ void ASTStmtWriter::VisitOMPTeamsGenericLoopDirective(
 void ASTStmtWriter::VisitOMPTargetTeamsGenericLoopDirective(
     OMPTargetTeamsGenericLoopDirective *D) {
   VisitOMPLoopDirective(D);
+  Record.writeBool(D->canBeParallelFor());
   Code = serialization::STMT_OMP_TARGET_TEAMS_GENERIC_LOOP_DIRECTIVE;
 }
 

diff --git a/clang/test/APINotes/instancetype.m b/clang/test/APINotes/instancetype.m
@@ -1,3 +1,4 @@
+// RUN: rm -rf %t
 // RUN: %clang_cc1 -fmodules -fimplicit-module-maps -fmodules-cache-path=%t/ModulesCache -fapinotes-modules -fsyntax-only -I %S/Inputs/Headers -verify %s
 
 @import InstancetypeModule;

diff --git a/clang/test/AST/Interp/builtin-functions.cpp b/clang/test/AST/Interp/builtin-functions.cpp
@@ -482,6 +482,9 @@ void test_noexcept(int *i) {
 #undef TEST_TYPE
 } // end namespace test_launder
 
+
+/// FIXME: The commented out tests here use a IntAP value and fail.
+/// This currently means we will leak the IntAP value since nothing cleans it up.
 namespace clz {
   char clz1[__builtin_clz(1) == BITSIZE(int) - 1 ? 1 : -1];
   char clz2[__builtin_clz(7) == BITSIZE(int) - 3 ? 1 : -1];
@@ -492,6 +495,63 @@ namespace clz {
   char clz7[__builtin_clzs(0x1) == BITSIZE(short) - 1 ? 1 : -1];
   char clz8[__builtin_clzs(0xf) == BITSIZE(short) - 4 ? 1 : -1];
   char clz9[__builtin_clzs(0xfff) == BITSIZE(short) - 12 ? 1 : -1];
+
+  int clz10 = __builtin_clzg((unsigned char)0);
+  char clz11[__builtin_clzg((unsigned char)0, 42) == 42 ? 1 : -1];
+  char clz12[__builtin_clzg((unsigned char)0x1) == BITSIZE(char) - 1 ? 1 : -1];
+  char clz13[__builtin_clzg((unsigned char)0x1, 42) == BITSIZE(char) - 1 ? 1 : -1];
+  char clz14[__builtin_clzg((unsigned char)0xf) == BITSIZE(char) - 4 ? 1 : -1];
+  char clz15[__builtin_clzg((unsigned char)0xf, 42) == BITSIZE(char) - 4 ? 1 : -1];
+  char clz16[__builtin_clzg((unsigned char)(1 << (BITSIZE(char) - 1))) == 0 ? 1 : -1];
+  char clz17[__builtin_clzg((unsigned char)(1 << (BITSIZE(char) - 1)), 42) == 0 ? 1 : -1];
+  int clz18 = __builtin_clzg((unsigned short)0);
+  char clz19[__builtin_clzg((unsigned short)0, 42) == 42 ? 1 : -1];
+  char clz20[__builtin_clzg((unsigned short)0x1) == BITSIZE(short) - 1 ? 1 : -1];
+  char clz21[__builtin_clzg((unsigned short)0x1, 42) == BITSIZE(short) - 1 ? 1 : -1];
+  char clz22[__builtin_clzg((unsigned short)0xf) == BITSIZE(short) - 4 ? 1 : -1];
+  char clz23[__builtin_clzg((unsigned short)0xf, 42) == BITSIZE(short) - 4 ? 1 : -1];
+  char clz24[__builtin_clzg((unsigned short)(1 << (BITSIZE(short) - 1))) == 0 ? 1 : -1];
+  char clz25[__builtin_clzg((unsigned short)(1 << (BITSIZE(short) - 1)), 42) == 0 ? 1 : -1];
+  int clz26 = __builtin_clzg(0U);
+  char clz27[__builtin_clzg(0U, 42) == 42 ? 1 : -1];
+  char clz28[__builtin_clzg(0x1U) == BITSIZE(int) - 1 ? 1 : -1];
+  char clz29[__builtin_clzg(0x1U, 42) == BITSIZE(int) - 1 ? 1 : -1];
+  char clz30[__builtin_clzg(0xfU) == BITSIZE(int) - 4 ? 1 : -1];
+  char clz31[__builtin_clzg(0xfU, 42) == BITSIZE(int) - 4 ? 1 : -1];
+  char clz32[__builtin_clzg(1U << (BITSIZE(int) - 1)) == 0 ? 1 : -1];
+  char clz33[__builtin_clzg(1U << (BITSIZE(int) - 1), 42) == 0 ? 1 : -1];
+  int clz34 = __builtin_clzg(0UL);
+  char clz35[__builtin_clzg(0UL, 42) == 42 ? 1 : -1];
+  char clz36[__builtin_clzg(0x1UL) == BITSIZE(long) - 1 ? 1 : -1];
+  char clz37[__builtin_clzg(0x1UL, 42) == BITSIZE(long) - 1 ? 1 : -1];
+  char clz38[__builtin_clzg(0xfUL) == BITSIZE(long) - 4 ? 1 : -1];
+  char clz39[__builtin_clzg(0xfUL, 42) == BITSIZE(long) - 4 ? 1 : -1];
+  char clz40[__builtin_clzg(1UL << (BITSIZE(long) - 1)) == 0 ? 1 : -1];
+  char clz41[__builtin_clzg(1UL << (BITSIZE(long) - 1), 42) == 0 ? 1 : -1];
+  int clz42 = __builtin_clzg(0ULL);
+  char clz43[__builtin_clzg(0ULL, 42) == 42 ? 1 : -1];
+  char clz44[__builtin_clzg(0x1ULL) == BITSIZE(long long) - 1 ? 1 : -1];
+  char clz45[__builtin_clzg(0x1ULL, 42) == BITSIZE(long long) - 1 ? 1 : -1];
+  char clz46[__builtin_clzg(0xfULL) == BITSIZE(long long) - 4 ? 1 : -1];
+  char clz47[__builtin_clzg(0xfULL, 42) == BITSIZE(long long) - 4 ? 1 : -1];
+  char clz48[__builtin_clzg(1ULL << (BITSIZE(long long) - 1)) == 0 ? 1 : -1];
+  char clz49[__builtin_clzg(1ULL << (BITSIZE(long long) - 1), 42) == 0 ? 1 : -1];
+#ifdef __SIZEOF_INT128__
+  // int clz50 = __builtin_clzg((unsigned __int128)0);
+  char clz51[__builtin_clzg((unsigned __int128)0, 42) == 42 ? 1 : -1];
+  char clz52[__builtin_clzg((unsigned __int128)0x1) == BITSIZE(__int128) - 1 ? 1 : -1];
+  char clz53[__builtin_clzg((unsigned __int128)0x1, 42) == BITSIZE(__int128) - 1 ? 1 : -1];
+  char clz54[__builtin_clzg((unsigned __int128)0xf) == BITSIZE(__int128) - 4 ? 1 : -1];
+  char clz55[__builtin_clzg((unsigned __int128)0xf, 42) == BITSIZE(__int128) - 4 ? 1 : -1];
+#endif
+#ifndef __AVR__
+  // int clz58 = __builtin_clzg((unsigned _BitInt(128))0);
+  char clz59[__builtin_clzg((unsigned _BitInt(128))0, 42) == 42 ? 1 : -1];
+  char clz60[__builtin_clzg((unsigned _BitInt(128))0x1) == BITSIZE(_BitInt(128)) - 1 ? 1 : -1];
+  char clz61[__builtin_clzg((unsigned _BitInt(128))0x1, 42) == BITSIZE(_BitInt(128)) - 1 ? 1 : -1];
+  char clz62[__builtin_clzg((unsigned _BitInt(128))0xf) == BITSIZE(_BitInt(128)) - 4 ? 1 : -1];
+  char clz63[__builtin_clzg((unsigned _BitInt(128))0xf, 42) == BITSIZE(_BitInt(128)) - 4 ? 1 : -1];
+#endif
 }
 
 namespace ctz {
@@ -502,6 +562,66 @@ namespace ctz {
   char ctz5[__builtin_ctzl(0x10L) == 4 ? 1 : -1];
   char ctz6[__builtin_ctzll(0x100LL) == 8 ? 1 : -1];
   char ctz7[__builtin_ctzs(1 << (BITSIZE(short) - 1)) == BITSIZE(short) - 1 ? 1 : -1];
+  int ctz8 = __builtin_ctzg((unsigned char)0);
+  char ctz9[__builtin_ctzg((unsigned char)0, 42) == 42 ? 1 : -1];
+  char ctz10[__builtin_ctzg((unsigned char)0x1) == 0 ? 1 : -1];
+  char ctz11[__builtin_ctzg((unsigned char)0x1, 42) == 0 ? 1 : -1];
+  char ctz12[__builtin_ctzg((unsigned char)0x10) == 4 ? 1 : -1];
+  char ctz13[__builtin_ctzg((unsigned char)0x10, 42) == 4 ? 1 : -1];
+  char ctz14[__builtin_ctzg((unsigned char)(1 << (BITSIZE(char) - 1))) == BITSIZE(char) - 1 ? 1 : -1];
+  char ctz15[__builtin_ctzg((unsigned char)(1 << (BITSIZE(char) - 1)), 42) == BITSIZE(char) - 1 ? 1 : -1];
+  int ctz16 = __builtin_ctzg((unsigned short)0);
+  char ctz17[__builtin_ctzg((unsigned short)0, 42) == 42 ? 1 : -1];
+  char ctz18[__builtin_ctzg((unsigned short)0x1) == 0 ? 1 : -1];
+  char ctz19[__builtin_ctzg((unsigned short)0x1, 42) == 0 ? 1 : -1];
+  char ctz20[__builtin_ctzg((unsigned short)0x10) == 4 ? 1 : -1];
+  char ctz21[__builtin_ctzg((unsigned short)0x10, 42) == 4 ? 1 : -1];
+  char ctz22[__builtin_ctzg((unsigned short)(1 << (BITSIZE(short) - 1))) == BITSIZE(short) - 1 ? 1 : -1];
+  char ctz23[__builtin_ctzg((unsigned short)(1 << (BITSIZE(short) - 1)), 42) == BITSIZE(short) - 1 ? 1 : -1];
+  int ctz24 = __builtin_ctzg(0U);
+  char ctz25[__builtin_ctzg(0U, 42) == 42 ? 1 : -1];
+  char ctz26[__builtin_ctzg(0x1U) == 0 ? 1 : -1];
+  char ctz27[__builtin_ctzg(0x1U, 42) == 0 ? 1 : -1];
+  char ctz28[__builtin_ctzg(0x10U) == 4 ? 1 : -1];
+  char ctz29[__builtin_ctzg(0x10U, 42) == 4 ? 1 : -1];
+  char ctz30[__builtin_ctzg(1U << (BITSIZE(int) - 1)) == BITSIZE(int) - 1 ? 1 : -1];
+  char ctz31[__builtin_ctzg(1U << (BITSIZE(int) - 1), 42) == BITSIZE(int) - 1 ? 1 : -1];
+  int ctz32 = __builtin_ctzg(0UL);
+  char ctz33[__builtin_ctzg(0UL, 42) == 42 ? 1 : -1];
+  char ctz34[__builtin_ctzg(0x1UL) == 0 ? 1 : -1];
+  char ctz35[__builtin_ctzg(0x1UL, 42) == 0 ? 1 : -1];
+  char ctz36[__builtin_ctzg(0x10UL) == 4 ? 1 : -1];
+  char ctz37[__builtin_ctzg(0x10UL, 42) == 4 ? 1 : -1];
+  char ctz38[__builtin_ctzg(1UL << (BITSIZE(long) - 1)) == BITSIZE(long) - 1 ? 1 : -1];
+  char ctz39[__builtin_ctzg(1UL << (BITSIZE(long) - 1), 42) == BITSIZE(long) - 1 ? 1 : -1];
+  int ctz40 = __builtin_ctzg(0ULL);
+  char ctz41[__builtin_ctzg(0ULL, 42) == 42 ? 1 : -1];
+  char ctz42[__builtin_ctzg(0x1ULL) == 0 ? 1 : -1];
+  char ctz43[__builtin_ctzg(0x1ULL, 42) == 0 ? 1 : -1];
+  char ctz44[__builtin_ctzg(0x10ULL) == 4 ? 1 : -1];
+  char ctz45[__builtin_ctzg(0x10ULL, 42) == 4 ? 1 : -1];
+  char ctz46[__builtin_ctzg(1ULL << (BITSIZE(long long) - 1)) == BITSIZE(long long) - 1 ? 1 : -1];
+  char ctz47[__builtin_ctzg(1ULL << (BITSIZE(long long) - 1), 42) == BITSIZE(long long) - 1 ? 1 : -1];
+#ifdef __SIZEOF_INT128__
+  // int ctz48 = __builtin_ctzg((unsigned __int128)0);
+  char ctz49[__builtin_ctzg((unsigned __int128)0, 42) == 42 ? 1 : -1];
+  char ctz50[__builtin_ctzg((unsigned __int128)0x1) == 0 ? 1 : -1];
+  char ctz51[__builtin_ctzg((unsigned __int128)0x1, 42) == 0 ? 1 : -1];
+  char ctz52[__builtin_ctzg((unsigned __int128)0x10) == 4 ? 1 : -1];
+  char ctz53[__builtin_ctzg((unsigned __int128)0x10, 42) == 4 ? 1 : -1];
+  char ctz54[__builtin_ctzg((unsigned __int128)1 << (BITSIZE(__int128) - 1)) == BITSIZE(__int128) - 1 ? 1 : -1];
+  char ctz55[__builtin_ctzg((unsigned __int128)1 << (BITSIZE(__int128) - 1), 42) == BITSIZE(__int128) - 1 ? 1 : -1];
+#endif
+#ifndef __AVR__
+  // int ctz56 = __builtin_ctzg((unsigned _BitInt(128))0);
+  char ctz57[__builtin_ctzg((unsigned _BitInt(128))0, 42) == 42 ? 1 : -1];
+  char ctz58[__builtin_ctzg((unsigned _BitInt(128))0x1) == 0 ? 1 : -1];
+  char ctz59[__builtin_ctzg((unsigned _BitInt(128))0x1, 42) == 0 ? 1 : -1];
+  char ctz60[__builtin_ctzg((unsigned _BitInt(128))0x10) == 4 ? 1 : -1];
+  char ctz61[__builtin_ctzg((unsigned _BitInt(128))0x10, 42) == 4 ? 1 : -1];
+  char ctz62[__builtin_ctzg((unsigned _BitInt(128))1 << (BITSIZE(_BitInt(128)) - 1)) == BITSIZE(_BitInt(128)) - 1 ? 1 : -1];
+  char ctz63[__builtin_ctzg((unsigned _BitInt(128))1 << (BITSIZE(_BitInt(128)) - 1), 42) == BITSIZE(_BitInt(128)) - 1 ? 1 : -1];
+#endif
 }
 
 namespace bswap {

diff --git a/clang/test/AST/Interp/c.c b/clang/test/AST/Interp/c.c
@@ -209,3 +209,21 @@ const struct StrA * const sb = &sa;
 const struct StrA sc = *sb;
 _Static_assert(sc.a == 12, ""); // pedantic-ref-warning {{GNU extension}} \
                                 // pedantic-expected-warning {{GNU extension}}
+
+_Static_assert(((void*)0 + 1) != (void*)0, ""); // pedantic-expected-warning {{arithmetic on a pointer to void is a GNU extension}} \
+                                                // pedantic-expected-warning {{not an integer constant expression}} \
+                                                // pedantic-expected-note {{cannot perform pointer arithmetic on null pointer}} \
+                                                // pedantic-ref-warning {{arithmetic on a pointer to void is a GNU extension}} \
+                                                // pedantic-ref-warning {{not an integer constant expression}} \
+                                                // pedantic-ref-note {{cannot perform pointer arithmetic on null pointer}}
+
+typedef __INTPTR_TYPE__ intptr_t;
+int array[(intptr_t)(int*)1]; // ref-warning {{variable length array folded to constant array}} \
+                              // pedantic-ref-warning {{variable length array folded to constant array}} \
+                              // expected-warning {{variable length array folded to constant array}} \
+                              // pedantic-expected-warning {{variable length array folded to constant array}}
+
+int castViaInt[*(int*)(unsigned long)"test"]; // ref-error {{variable length array}} \
+                                              // pedantic-ref-error {{variable length array}} \
+                                              // expected-error {{variable length array}} \
+                                              // pedantic-expected-error {{variable length array}}
diff --git a/clang/test/AST/Interp/const-eval.c b/clang/test/AST/Interp/const-eval.c
@@ -0,0 +1,192 @@
+// RUN: %clang_cc1 -fsyntax-only -verify=both,ref      -triple x86_64-linux %s -Wno-tautological-pointer-compare -Wno-pointer-to-int-cast
+// RUN: %clang_cc1 -fsyntax-only -verify=both,expected -triple x86_64-linux %s -Wno-tautological-pointer-compare -Wno-pointer-to-int-cast -fexperimental-new-constant-interpreter -DNEW_INTERP
+// RUN: %clang_cc1 -fsyntax-only -verify=both,ref      -triple powerpc64-ibm-aix-xcoff %s -Wno-tautological-pointer-compare -Wno-pointer-to-int-cast
+// RUN: %clang_cc1 -fsyntax-only -verify=both,expected -triple powerpc64-ibm-aix-xcoff %s -Wno-tautological-pointer-compare -Wno-pointer-to-int-cast -fexperimental-new-constant-interpreter -DNEW_INTERP
+
+/// This is a version of test/Sema/const-eval.c with the
+/// tests commented out that the new constant expression interpreter does
+/// not support yet. They are all marked with the NEW_INTERP define:
+///
+///   - builtin_constant_p
+///   - unions
+
+
+#define EVAL_EXPR(testno, expr) enum { test##testno = (expr) }; struct check_positive##testno { int a[test##testno]; };
+int x;
+EVAL_EXPR(1, (_Bool)&x)
+EVAL_EXPR(2, (int)(1.0+(double)4))
+EVAL_EXPR(3, (int)(1.0+(float)4.0))
+EVAL_EXPR(4, (_Bool)(1 ? (void*)&x : 0))
+EVAL_EXPR(5, (_Bool)(int[]){0})
+struct y {int x,y;};
+EVAL_EXPR(6, (int)(1+(struct y*)0))
+_Static_assert((long)&((struct y*)0)->y > 0, "");
+EVAL_EXPR(7, (int)&((struct y*)0)->y)
+EVAL_EXPR(8, (_Bool)"asdf")
+EVAL_EXPR(9, !!&x)
+EVAL_EXPR(10, ((void)1, 12))
+void g0(void);
+EVAL_EXPR(11, (g0(), 12)) // both-error {{not an integer constant expression}}
+EVAL_EXPR(12, 1.0&&2.0)
+EVAL_EXPR(13, x || 3.0) // both-error {{not an integer constant expression}}
+
+unsigned int l_19 = 1;
+EVAL_EXPR(14, (1 ^ l_19) && 1); // both-error {{not an integer constant expression}}
+
+void f(void)
+{
+  int a;
+  EVAL_EXPR(15, (_Bool)&a);
+}
+
+_Complex float g16 = (1.0f + 1.0fi);
+
+// ?: in constant expressions.
+int g17[(3?:1) - 2];
+
+EVAL_EXPR(18, ((int)((void*)10 + 10)) == 20 ? 1 : -1);
+
+struct s {
+  int a[(int)-1.0f]; // both-error {{array size is negative}}
+};
+
+EVAL_EXPR(19, ((int)&*(char*)10 == 10 ? 1 : -1));
+
+#ifndef NEW_INTERP
+EVAL_EXPR(20, __builtin_constant_p(*((int*) 10)));
+#endif
+
+EVAL_EXPR(21, (__imag__ 2i) == 2 ? 1 : -1);
+
+EVAL_EXPR(22, (__real__ (2i+3)) == 3 ? 1 : -1);
+
+int g23[(int)(1.0 / 1.0)] = { 1 }; // both-warning {{folded to constant array}}
+int g24[(int)(1.0 / 1.0)] = { 1 , 2 }; // both-warning {{folded to constant array}} \
+                                       // both-warning {{excess elements in array initializer}}
+int g25[(int)(1.0 + 1.0)], g26 = sizeof(g25); // both-warning {{folded to constant array}}
+
+EVAL_EXPR(26, (_Complex double)0 ? -1 : 1)
+EVAL_EXPR(27, (_Complex int)0 ? -1 : 1)
+EVAL_EXPR(28, (_Complex double)1 ? 1 : -1)
+EVAL_EXPR(29, (_Complex int)1 ? 1 : -1)
+
+// PR4027
+struct a { int x, y; };
+static struct a V2 = (struct a)(struct a){ 1, 2};
+static const struct a V1 = (struct a){ 1, 2};
+
+EVAL_EXPR(30, (int)(_Complex float)((1<<30)-1) == (1<<30) ? 1 : -1)
+EVAL_EXPR(31, (int*)0 == (int*)0 ? 1 : -1)
+EVAL_EXPR(32, (int*)0 != (int*)0 ? -1 : 1)
+EVAL_EXPR(33, (void*)0 - (void*)0 == 0 ? 1 : -1)
+
+void foo(void) {}
+EVAL_EXPR(34, (foo == (void *)0) ? -1 : 1)
+
+// No PR. Mismatched bitwidths lead to a crash on second evaluation.
+const _Bool constbool = 0;
+EVAL_EXPR(35, constbool)
+EVAL_EXPR(36, constbool)
+
+EVAL_EXPR(37, ((void)1,2.0) == 2.0 ? 1 : -1)
+EVAL_EXPR(38, __builtin_expect(1,1) == 1 ? 1 : -1)
+
+// PR7884
+EVAL_EXPR(39, __real__(1.f) == 1 ? 1 : -1)
+EVAL_EXPR(40, __imag__(1.f) == 0 ? 1 : -1)
+
+// From gcc testsuite
+EVAL_EXPR(41, (int)(1+(_Complex unsigned)2))
+
+void rdar8875946(void) {
+  double _Complex  P;
+  float _Complex  P2 = 3.3f + P;
+}
+
+double d = (d = 0.0); // both-error {{not a compile-time constant}}
+double d2 = ++d; // both-error {{not a compile-time constant}}
+
+int n = 2;
+int intLvalue[*(int*)((long)&n ?: 1)] = { 1, 2 }; // both-error {{variable length array}}
+
+union u { int a; char b[4]; };
+char c = ((union u)(123456)).b[0]; // both-error {{not a compile-time constant}}
+
+#ifndef NEW_INTERP
+extern const int weak_int __attribute__((weak));
+const int weak_int = 42;
+int weak_int_test = weak_int; // both-error {{not a compile-time constant}}
+#endif
+
+int literalVsNull1 = "foo" == 0;
+int literalVsNull2 = 0 == "foo";
+
+// PR11385.
+int castViaInt[*(int*)(unsigned long)"test"]; // both-error {{variable length array}}
+
+// PR11391.
+#ifndef NEW_INTERP
+struct PR11391 { _Complex float f; } pr11391;
+EVAL_EXPR(42, __builtin_constant_p(pr11391.f = 1))
+#endif
+
+// PR12043
+float varfloat;
+const float constfloat = 0;
+EVAL_EXPR(43, varfloat && constfloat) // both-error {{not an integer constant expression}}
+EVAL_EXPR(45, ((char*)-1) + 1 == 0 ? 1 : -1)
+EVAL_EXPR(46, ((char*)-1) + 1 < (char*) -1 ? 1 : -1)
+EVAL_EXPR(47, &x < &x + 1 ? 1 : -1)
+EVAL_EXPR(48, &x != &x - 1 ? 1 : -1)
+EVAL_EXPR(49, &x < &x - 100 ? 1 : -1) // ref-error {{not an integer constant expression}}
+
+/// FIXME: Rejecting this is correct, BUT when converting the innermost pointer
+/// to an integer, we do not preserve the information where it came from. So when we later
+/// create a pointer from it, it also doesn't have that information, which means
+/// hasSameBase() for those two pointers will return false. And in those cases, we emit
+/// the diagnostic:
+///   comparison between '&Test50' and '&(631578)' has unspecified value
+extern struct Test50S Test50;
+EVAL_EXPR(50, &Test50 < (struct Test50S*)((unsigned long)&Test50 + 10)) // both-error {{not an integer constant expression}} \
+                                                                        // expected-note {{comparison between}}
+
+EVAL_EXPR(51, 0 != (float)1e99)
+
+// PR21945
+void PR21945(void) { int i = (({}), 0l); }
+
+void PR24622(void);
+struct PR24622 {} pr24622;
+EVAL_EXPR(52, &pr24622 == (void *)&PR24622);
+
+// We evaluate these by providing 2s' complement semantics in constant
+// expressions, like we do for integers.
+void *PR28739a = (__int128)(unsigned long)-1 + &PR28739a;                  // both-warning {{the pointer incremented by 18446744073709551615 refers past the last possible element for an array in 64-bit address space containing 64-bit (8-byte) elements (max possible 2305843009213693952 elements)}}
+
+void *PR28739b = &PR28739b + (__int128)(unsigned long)-1;                  // both-warning {{refers past the last possible element}}
+__int128 PR28739c = (&PR28739c + (__int128)(unsigned long)-1) - &PR28739c; // both-warning {{refers past the last possible element}}
+void *PR28739d = &(&PR28739d)[(__int128)(unsigned long)-1];                // both-warning {{refers past the last possible element}}
+
+struct PR35214_X {
+  int k;
+  int arr[];
+};
+int PR35214_x;
+int PR35214_y = ((struct PR35214_X *)&PR35214_x)->arr[1]; // both-error {{not a compile-time constant}}
+#ifndef NEW_INTERP
+int *PR35214_z = &((struct PR35214_X *)&PR35214_x)->arr[1]; // ok, &PR35214_x + 2
+#endif
+
+/// From const-eval-64.c
+EVAL_EXPR(53, ((char*)-1LL) + 1 == 0 ? 1 : -1)
+EVAL_EXPR(54, ((char*)-1LL) + 1 < (char*) -1 ? 1 : -1)
+
+/// === Additions ===
+#if __SIZEOF_INT__ == 4
+typedef __INTPTR_TYPE__ intptr_t;
+const intptr_t A = (intptr_t)(((int*) 0) + 1);
+const intptr_t B = (intptr_t)(((char*)0) + 3);
+_Static_assert(A > B, "");
+#else
+#error :(
+#endif
diff --git a/clang/test/AST/Interp/functions.cpp b/clang/test/AST/Interp/functions.cpp
@@ -185,6 +185,21 @@ namespace FunctionReturnType {
   constexpr int (*invalidFnPtr)() = m;
   static_assert(invalidFnPtr() == 5, ""); // both-error {{not an integral constant expression}} \
                                           // both-note {{non-constexpr function 'm'}}
+
+
+namespace ToBool {
+  void mismatched(int x) {}
+  typedef void (*callback_t)(int);
+  void foo() {
+    callback_t callback = (callback_t)mismatched; // warns
+    /// Casts a function pointer to a boolean and then back to a function pointer.
+    /// This is extracted from test/Sema/callingconv-cast.c
+    callback = (callback_t)!mismatched; // both-warning {{address of function 'mismatched' will always evaluate to 'true'}} \
+                                        // both-note {{prefix with the address-of operator to silence this warning}}
+  }
+}
+
+
 }
 
 namespace Comparison {

diff --git a/clang/test/AST/Interp/vectors.cpp b/clang/test/AST/Interp/vectors.cpp
@@ -0,0 +1,22 @@
+// RUN: %clang_cc1 -fexperimental-new-constant-interpreter -verify=expected,both %s
+// RUN: %clang_cc1 -verify=ref,both %s
+
+// both-no-diagnostics
+
+typedef int __attribute__((vector_size(16))) VI4;
+constexpr VI4 A = {1,2,3,4};
+
+/// From constant-expression-cxx11.cpp
+namespace Vector {
+  typedef int __attribute__((vector_size(16))) VI4;
+  constexpr VI4 f(int n) {
+    return VI4 { n * 3, n + 4, n - 5, n / 6 };
+  }
+  constexpr auto v1 = f(10);
+
+  typedef double __attribute__((vector_size(32))) VD4;
+  constexpr VD4 g(int n) {
+    return (VD4) { n / 2.0, n + 1.5, n - 5.4, n * 0.9 };
+  }
+  constexpr auto v2 = g(4);
+}
diff --git a/clang/test/CodeGen/X86/x86_64-arguments.c b/clang/test/CodeGen/X86/x86_64-arguments.c
@@ -533,6 +533,24 @@ typedef float t66 __attribute__((__vector_size__(128), __aligned__(128)));
 void f66(t66 a0) {
 }
 
+typedef long long t67 __attribute__((aligned (4)));
+struct s67 {
+  int a;
+  t67 b;
+};
+// CHECK-LABEL: define{{.*}} void @f67(ptr noundef byval(%struct.s67) align 8 %x)
+void f67(struct s67 x) {
+}
+
+typedef double t68 __attribute__((aligned (4)));
+struct s68 {
+  int a;
+  t68 b;
+};
+// CHECK-LABEL: define{{.*}} void @f68(ptr noundef byval(%struct.s68) align 8 %x)
+void f68(struct s68 x) {
+}
+
 /// The synthesized __va_list_tag does not have file/line fields.
 // CHECK:      = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "__va_list_tag",
 // CHECK-NOT:  file:

diff --git a/clang/test/CodeGenCXX/arm64ec-vectorcall.cpp b/clang/test/CodeGenCXX/arm64ec-vectorcall.cpp
@@ -0,0 +1,14 @@
+// RUN: %clang_cc1 -triple arm64ec-windows-msvc -emit-llvm -o - %s -verify
+
+// ARM64EC doesn't support generating __vectorcall calls... but __vectorcall
+// function types need to be distinct from __cdecl function types to support
+// compiling the STL. Make sure we only diagnose constructs that actually
+// require generating code.
+void __vectorcall f1();
+void f2(void __vectorcall p()) {}
+void f2(void p()) {}
+void __vectorcall (*f3)();
+void __vectorcall f4(); // expected-error {{__vectorcall}}
+void __vectorcall f5() { // expected-error {{__vectorcall}}
+  f4(); // expected-error{{__vectorcall}}
+}
diff --git a/clang/test/CodeGenCXX/control-flow-in-stmt-expr.cpp b/clang/test/CodeGenCXX/control-flow-in-stmt-expr.cpp
diff --git a/clang/test/CodeGenCoroutines/coro-suspend-cleanups.cpp b/clang/test/CodeGenCoroutines/coro-suspend-cleanups.cpp
@@ -0,0 +1,93 @@
+// RUN: %clang_cc1 --std=c++20 -triple x86_64-linux-gnu -emit-llvm %s -o - | FileCheck %s
+
+#include "Inputs/coroutine.h"
+
+struct Printy {
+  Printy(const char *name) : name(name) {}
+  ~Printy() {}
+  const char *name;
+};
+
+struct coroutine {
+  struct promise_type;
+  std::coroutine_handle<promise_type> handle;
+  ~coroutine() {
+    if (handle) handle.destroy();
+  }
+};
+
+struct coroutine::promise_type {
+  coroutine get_return_object() {
+    return {std::coroutine_handle<promise_type>::from_promise(*this)};
+  }
+  std::suspend_never initial_suspend() noexcept { return {}; }
+  std::suspend_always final_suspend() noexcept { return {}; }
+  void return_void() {}
+  void unhandled_exception() {}
+};
+
+struct Awaiter : std::suspend_always {
+  Printy await_resume() { return {"awaited"}; }
+};
+
+int foo() { return 2; }
+
+coroutine ArrayInitCoro() {
+  // Verify that:
+  //  - We do the necessary stores for array cleanups.
+  //  - Array cleanups are called by await.cleanup.
+  //  - We activate the cleanup after the first element and deactivate it in await.ready (see cleanup.isactive).
+
+  // CHECK-LABEL: define dso_local void @_Z13ArrayInitCorov
+  // CHECK: %arrayinit.endOfInit = alloca ptr, align 8
+  // CHECK: %cleanup.isactive = alloca i1, align 1
+  Printy arr[2] = {
+    Printy("a"),
+    // CHECK:       %arrayinit.begin = getelementptr inbounds [2 x %struct.Printy], ptr %arr.reload.addr, i64 0, i64 0
+    // CHECK-NEXT:  %arrayinit.begin.spill.addr = getelementptr inbounds %_Z13ArrayInitCorov.Frame, ptr %0, i32 0, i32 10
+    // CHECK-NEXT:  store ptr %arrayinit.begin, ptr %arrayinit.begin.spill.addr, align 8
+    // CHECK-NEXT:  store i1 true, ptr %cleanup.isactive.reload.addr, align 1
+    // CHECK-NEXT:  store ptr %arrayinit.begin, ptr %arrayinit.endOfInit.reload.addr, align 8
+    // CHECK-NEXT:  call void @_ZN6PrintyC1EPKc(ptr noundef nonnull align 8 dereferenceable(8) %arrayinit.begin, ptr noundef @.str)
+    // CHECK-NEXT:  %arrayinit.element = getelementptr inbounds %struct.Printy, ptr %arrayinit.begin, i64 1
+    // CHECK-NEXT:  %arrayinit.element.spill.addr = getelementptr inbounds %_Z13ArrayInitCorov.Frame, ptr %0, i32 0, i32 11
+    // CHECK-NEXT:  store ptr %arrayinit.element, ptr %arrayinit.element.spill.addr, align 8
+    // CHECK-NEXT:  store ptr %arrayinit.element, ptr %arrayinit.endOfInit.reload.addr, align 8
+    co_await Awaiter{}
+    // CHECK-NEXT:  @_ZNSt14suspend_always11await_readyEv
+    // CHECK-NEXT:  br i1 %{{.+}}, label %await.ready, label %CoroSave30
+  };
+  // CHECK:       await.cleanup:                                    ; preds = %AfterCoroSuspend{{.*}}
+  // CHECK-NEXT:    br label %cleanup{{.*}}.from.await.cleanup
+
+  // CHECK:       cleanup{{.*}}.from.await.cleanup:                      ; preds = %await.cleanup
+  // CHECK:         br label %cleanup{{.*}}
+
+  // CHECK:       await.ready:
+  // CHECK-NEXT:    %arrayinit.element.reload.addr = getelementptr inbounds %_Z13ArrayInitCorov.Frame, ptr %0, i32 0, i32 11
+  // CHECK-NEXT:    %arrayinit.element.reload = load ptr, ptr %arrayinit.element.reload.addr, align 8
+  // CHECK-NEXT:    call void @_ZN7Awaiter12await_resumeEv
+  // CHECK-NEXT:    store i1 false, ptr %cleanup.isactive.reload.addr, align 1
+  // CHECK-NEXT:    br label %cleanup{{.*}}.from.await.ready
+
+  // CHECK:       cleanup{{.*}}:                                         ; preds = %cleanup{{.*}}.from.await.ready, %cleanup{{.*}}.from.await.cleanup
+  // CHECK:         %cleanup.is_active = load i1, ptr %cleanup.isactive.reload.addr, align 1
+  // CHECK-NEXT:    br i1 %cleanup.is_active, label %cleanup.action, label %cleanup.done
+
+  // CHECK:       cleanup.action:
+  // CHECK:         %arraydestroy.isempty = icmp eq ptr %arrayinit.begin.reload{{.*}}, %{{.*}}
+  // CHECK-NEXT:    br i1 %arraydestroy.isempty, label %arraydestroy.done{{.*}}, label %arraydestroy.body.from.cleanup.action
+  // Ignore rest of the array cleanup.
+}
+
+coroutine ArrayInitWithCoReturn() {
+  // CHECK-LABEL: define dso_local void @_Z21ArrayInitWithCoReturnv
+  // Verify that we start to emit the array destructor.
+  // CHECK: %arrayinit.endOfInit = alloca ptr, align 8
+  Printy arr[2] = {"a", ({
+                      if (foo()) {
+                        co_return;
+                      }
+                      "b";
+                    })};
+}
diff --git a/clang/test/Driver/amdgpu-features.c b/clang/test/Driver/amdgpu-features.c
@@ -32,3 +32,9 @@
 
 // RUN: %clang -### -target amdgcn -mcpu=gfx1010 -mno-cumode %s 2>&1 | FileCheck --check-prefix=NO-CUMODE %s
 // NO-CUMODE: "-target-feature" "-cumode"
+
+// RUN: %clang -### -target amdgcn -mcpu=gfx1010 -mamdgpu-precise-memory-op %s 2>&1 | FileCheck --check-prefix=PREC-MEM %s
+// PREC-MEM: "-target-feature" "+precise-memory"
+
+// RUN: %clang -### -target amdgcn -mcpu=gfx1010 -mno-amdgpu-precise-memory-op %s 2>&1 | FileCheck --check-prefix=NO-PREC-MEM %s
+// NO-PREC-MEM-NOT: {{".*precise-memory"}}
diff --git a/clang/test/Driver/linker-wrapper-libs.c b/clang/test/Driver/linker-wrapper-libs.c
@@ -44,6 +44,8 @@ int bar() { return weak; }
 // RUN: %clang -cc1 %s -triple x86_64-unknown-linux-gnu -emit-obj -o %t.o -fembed-offload-object=%t.out
 // RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run \
 // RUN:   --linker-path=/usr/bin/ld %t.o %t.a -o a.out 2>&1 \
+// RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run \
+// RUN:   --linker-path=/usr/bin/ld %t.a %t.o -o a.out 2>&1 \
 // RUN: | FileCheck %s --check-prefix=LIBRARY-RESOLVES
 
 // LIBRARY-RESOLVES: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx1030 {{.*}}.o {{.*}}.o
@@ -66,6 +68,8 @@ int bar() { return weak; }
 // RUN: %clang -cc1 %s -triple x86_64-unknown-linux-gnu -emit-obj -o %t.o -fembed-offload-object=%t.out
 // RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run \
 // RUN:   --linker-path=/usr/bin/ld %t.o %t.a -o a.out 2>&1 \
+// RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run \
+// RUN:   --linker-path=/usr/bin/ld %t.a %t.o -o a.out 2>&1 \
 // RUN: | FileCheck %s --check-prefix=LIBRARY-GLOBAL
 
 // LIBRARY-GLOBAL: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx1030 {{.*}}.o {{.*}}.o

diff --git a/clang/test/Driver/lld-repro.c b/clang/test/Driver/lld-repro.c
@@ -4,12 +4,12 @@
 // RUN: echo "-nostartfiles -nostdlib -fuse-ld=lld -gen-reproducer=error -fcrash-diagnostics-dir=%t" \
 // RUN:   | sed -e 's/\\/\\\\/g' > %t.rsp
 
-// RUN: not %clang %s @%t.rsp -fcrash-diagnostics=all 2>&1 \
+// RUN: not %clang %s @%t.rsp -fcrash-diagnostics=all -o /dev/null 2>&1 \
 // RUN:   | FileCheck %s
 
 // Test that the reproducer can still be created even when the input source cannot be preprocessed
 // again, like when reading from stdin.
-// RUN: not %clang -x c - @%t.rsp -fcrash-diagnostics=all 2>&1 < %s \
+// RUN: not %clang -x c - @%t.rsp -fcrash-diagnostics=all -o /dev/null 2>&1 < %s \
 // RUN:   | FileCheck %s
 
 // check that we still get lld's output
@@ -20,9 +20,9 @@
 // CHECK-NEXT: note: diagnostic msg:
 // CHECK: ********************
 
-// RUN: not %clang %s @%t.rsp -fcrash-diagnostics=compiler 2>&1 \
+// RUN: not %clang %s @%t.rsp -fcrash-diagnostics=compiler -o /dev/null 2>&1 \
 // RUN:   | FileCheck %s --check-prefix=NO-LINKER
-// RUN: not %clang %s @%t.rsp 2>&1 \
+// RUN: not %clang %s @%t.rsp -o /dev/null 2>&1 \
 // RUN:   | FileCheck %s --check-prefix=NO-LINKER
 
 // NO-LINKER-NOT: Preprocessed source(s) and associated run script(s) are located at:

diff --git a/clang/test/Modules/home-is-cwd-search-paths.c b/clang/test/Modules/home-is-cwd-search-paths.c
@@ -0,0 +1,34 @@
+// This test demonstrates how -fmodule-map-file-home-is-cwd with -fmodules-embed-all-files
+// extend the importer search paths by relying on the side effects of pragma diagnostic
+// mappings deserialization.
+
+// RUN: rm -rf %t
+// RUN: split-file %s %t
+
+//--- dir1/a.modulemap
+module a { header "a.h" }
+//--- dir1/a.h
+#include "search.h"
+// The first compilation is configured such that -I search does contain the search.h header.
+//--- dir1/search/search.h
+#pragma clang diagnostic push
+#pragma clang diagnostic ignored "-Wparentheses"
+#pragma clang diagnostic pop
+// RUN: cd %t/dir1 && %clang_cc1 -fmodules -I search \
+// RUN:   -emit-module -fmodule-name=a a.modulemap -o %t/a.pcm \
+// RUN:   -fmodules-embed-all-files -fmodule-map-file-home-is-cwd
+
+//--- dir2/b.modulemap
+module b { header "b.h" }
+//--- dir2/b.h
+#include "search.h" // expected-error{{'search.h' file not found}}
+// The second compilation is configured such that -I search is an empty directory.
+// However, since b.pcm simply embeds the headers as "search/search.h", this compilation
+// ends up seeing it too. This relies solely on ASTReader::ReadPragmaDiagnosticMappings()
+// eagerly reading the corresponding INPUT_FILE record before header search happens.
+// Removing the eager deserialization makes this header invisible and so does removing
+// the pragma directives.
+// RUN: mkdir %t/dir2/search
+// RUN: cd %t/dir2 && %clang_cc1 -fmodules -I search \
+// RUN:   -emit-module -fmodule-name=b b.modulemap -o %t/b.pcm \
+// RUN:   -fmodule-file=%t/a.pcm -verify
diff --git a/clang/test/OpenMP/nvptx_target_teams_generic_loop_codegen.cpp b/clang/test/OpenMP/nvptx_target_teams_generic_loop_codegen.cpp
@@ -320,35 +320,44 @@ int bar(int n){
 //
 //
 // CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33
-// CHECK1-SAME: (ptr noalias noundef [[DYN_PTR:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]]) #[[ATTR4:[0-9]+]] {
+// CHECK1-SAME: (ptr noalias noundef [[DYN_PTR:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR4:[0-9]+]] {
 // CHECK1-NEXT:  entry:
 // CHECK1-NEXT:    [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 8
 // CHECK1-NEXT:    [[B_ADDR:%.*]] = alloca ptr, align 8
+// CHECK1-NEXT:    [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
+// CHECK1-NEXT:    [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i64, align 8
 // CHECK1-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
 // CHECK1-NEXT:    [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
 // CHECK1-NEXT:    store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 8
 // CHECK1-NEXT:    store ptr [[B]], ptr [[B_ADDR]], align 8
+// CHECK1-NEXT:    store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
 // CHECK1-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK1-NEXT:    [[TMP1:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_kernel_environment, ptr [[DYN_PTR]])
 // CHECK1-NEXT:    [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP1]], -1
 // CHECK1-NEXT:    br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
 // CHECK1:       user_code.entry:
 // CHECK1-NEXT:    [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
+// CHECK1-NEXT:    [[TMP3:%.*]] = load i8, ptr [[DOTCAPTURE_EXPR__ADDR]], align 1
+// CHECK1-NEXT:    [[TOBOOL:%.*]] = trunc i8 [[TMP3]] to i1
+// CHECK1-NEXT:    [[FROMBOOL:%.*]] = zext i1 [[TOBOOL]] to i8
+// CHECK1-NEXT:    store i8 [[FROMBOOL]], ptr [[DOTCAPTURE_EXPR__CASTED]], align 1
+// CHECK1-NEXT:    [[TMP4:%.*]] = load i64, ptr [[DOTCAPTURE_EXPR__CASTED]], align 8
 // CHECK1-NEXT:    store i32 0, ptr [[DOTZERO_ADDR]], align 4
 // CHECK1-NEXT:    store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
-// CHECK1-NEXT:    call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[TMP0]]) #[[ATTR2]]
+// CHECK1-NEXT:    call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[TMP0]], i64 [[TMP4]]) #[[ATTR2]]
 // CHECK1-NEXT:    call void @__kmpc_target_deinit()
 // CHECK1-NEXT:    ret void
 // CHECK1:       worker.exit:
 // CHECK1-NEXT:    ret void
 //
 //
 // CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_omp_outlined
-// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]]) #[[ATTR1]] {
+// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1]] {
 // CHECK1-NEXT:  entry:
 // CHECK1-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
 // CHECK1-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
 // CHECK1-NEXT:    [[B_ADDR:%.*]] = alloca ptr, align 8
+// CHECK1-NEXT:    [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
 // CHECK1-NEXT:    [[DOTOMP_IV:%.*]] = alloca i32, align 4
 // CHECK1-NEXT:    [[TMP:%.*]] = alloca i32, align 4
 // CHECK1-NEXT:    [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4
@@ -360,6 +369,7 @@ int bar(int n){
 // CHECK1-NEXT:    store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
 // CHECK1-NEXT:    store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
 // CHECK1-NEXT:    store ptr [[B]], ptr [[B_ADDR]], align 8
+// CHECK1-NEXT:    store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
 // CHECK1-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK1-NEXT:    store i32 0, ptr [[DOTOMP_COMB_LB]], align 4
 // CHECK1-NEXT:    store i32 9, ptr [[DOTOMP_COMB_UB]], align 4
@@ -1566,35 +1576,44 @@ int bar(int n){
 //
 //
 // CHECK2-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33
-// CHECK2-SAME: (ptr noalias noundef [[DYN_PTR:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]]) #[[ATTR4:[0-9]+]] {
+// CHECK2-SAME: (ptr noalias noundef [[DYN_PTR:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR4:[0-9]+]] {
 // CHECK2-NEXT:  entry:
 // CHECK2-NEXT:    [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 8
 // CHECK2-NEXT:    [[B_ADDR:%.*]] = alloca ptr, align 8
+// CHECK2-NEXT:    [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
+// CHECK2-NEXT:    [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i64, align 8
 // CHECK2-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
 // CHECK2-NEXT:    [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
 // CHECK2-NEXT:    store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 8
 // CHECK2-NEXT:    store ptr [[B]], ptr [[B_ADDR]], align 8
+// CHECK2-NEXT:    store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
 // CHECK2-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK2-NEXT:    [[TMP1:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_kernel_environment, ptr [[DYN_PTR]])
 // CHECK2-NEXT:    [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP1]], -1
 // CHECK2-NEXT:    br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
 // CHECK2:       user_code.entry:
 // CHECK2-NEXT:    [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
+// CHECK2-NEXT:    [[TMP3:%.*]] = load i8, ptr [[DOTCAPTURE_EXPR__ADDR]], align 1
+// CHECK2-NEXT:    [[TOBOOL:%.*]] = trunc i8 [[TMP3]] to i1
+// CHECK2-NEXT:    [[FROMBOOL:%.*]] = zext i1 [[TOBOOL]] to i8
+// CHECK2-NEXT:    store i8 [[FROMBOOL]], ptr [[DOTCAPTURE_EXPR__CASTED]], align 1
+// CHECK2-NEXT:    [[TMP4:%.*]] = load i64, ptr [[DOTCAPTURE_EXPR__CASTED]], align 8
 // CHECK2-NEXT:    store i32 0, ptr [[DOTZERO_ADDR]], align 4
 // CHECK2-NEXT:    store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
-// CHECK2-NEXT:    call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[TMP0]]) #[[ATTR2]]
+// CHECK2-NEXT:    call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[TMP0]], i64 [[TMP4]]) #[[ATTR2]]
 // CHECK2-NEXT:    call void @__kmpc_target_deinit()
 // CHECK2-NEXT:    ret void
 // CHECK2:       worker.exit:
 // CHECK2-NEXT:    ret void
 //
 //
 // CHECK2-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_omp_outlined
-// CHECK2-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]]) #[[ATTR1]] {
+// CHECK2-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1]] {
 // CHECK2-NEXT:  entry:
 // CHECK2-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
 // CHECK2-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
 // CHECK2-NEXT:    [[B_ADDR:%.*]] = alloca ptr, align 8
+// CHECK2-NEXT:    [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
 // CHECK2-NEXT:    [[DOTOMP_IV:%.*]] = alloca i32, align 4
 // CHECK2-NEXT:    [[TMP:%.*]] = alloca i32, align 4
 // CHECK2-NEXT:    [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4
@@ -1606,6 +1625,7 @@ int bar(int n){
 // CHECK2-NEXT:    store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
 // CHECK2-NEXT:    store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
 // CHECK2-NEXT:    store ptr [[B]], ptr [[B_ADDR]], align 8
+// CHECK2-NEXT:    store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
 // CHECK2-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[B_ADDR]], align 8
 // CHECK2-NEXT:    store i32 0, ptr [[DOTOMP_COMB_LB]], align 4
 // CHECK2-NEXT:    store i32 9, ptr [[DOTOMP_COMB_UB]], align 4
@@ -2801,35 +2821,44 @@ int bar(int n){
 //
 //
 // CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33
-// CHECK3-SAME: (ptr noalias noundef [[DYN_PTR:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]]) #[[ATTR4:[0-9]+]] {
+// CHECK3-SAME: (ptr noalias noundef [[DYN_PTR:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]], i32 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR4:[0-9]+]] {
 // CHECK3-NEXT:  entry:
 // CHECK3-NEXT:    [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 4
 // CHECK3-NEXT:    [[B_ADDR:%.*]] = alloca ptr, align 4
+// CHECK3-NEXT:    [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i32, align 4
+// CHECK3-NEXT:    [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i32, align 4
 // CHECK3-NEXT:    [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
 // CHECK3-NEXT:    [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
 // CHECK3-NEXT:    store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 4
 // CHECK3-NEXT:    store ptr [[B]], ptr [[B_ADDR]], align 4
+// CHECK3-NEXT:    store i32 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
 // CHECK3-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[B_ADDR]], align 4
 // CHECK3-NEXT:    [[TMP1:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_kernel_environment, ptr [[DYN_PTR]])
 // CHECK3-NEXT:    [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP1]], -1
 // CHECK3-NEXT:    br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
 // CHECK3:       user_code.entry:
 // CHECK3-NEXT:    [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
+// CHECK3-NEXT:    [[TMP3:%.*]] = load i8, ptr [[DOTCAPTURE_EXPR__ADDR]], align 1
+// CHECK3-NEXT:    [[TOBOOL:%.*]] = trunc i8 [[TMP3]] to i1
+// CHECK3-NEXT:    [[FROMBOOL:%.*]] = zext i1 [[TOBOOL]] to i8
+// CHECK3-NEXT:    store i8 [[FROMBOOL]], ptr [[DOTCAPTURE_EXPR__CASTED]], align 1
+// CHECK3-NEXT:    [[TMP4:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__CASTED]], align 4
 // CHECK3-NEXT:    store i32 0, ptr [[DOTZERO_ADDR]], align 4
 // CHECK3-NEXT:    store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
-// CHECK3-NEXT:    call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[TMP0]]) #[[ATTR2]]
+// CHECK3-NEXT:    call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[TMP0]], i32 [[TMP4]]) #[[ATTR2]]
 // CHECK3-NEXT:    call void @__kmpc_target_deinit()
 // CHECK3-NEXT:    ret void
 // CHECK3:       worker.exit:
 // CHECK3-NEXT:    ret void
 //
 //
 // CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l33_omp_outlined
-// CHECK3-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]]) #[[ATTR1]] {
+// CHECK3-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(40) [[B:%.*]], i32 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1]] {
 // CHECK3-NEXT:  entry:
 // CHECK3-NEXT:    [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
 // CHECK3-NEXT:    [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
 // CHECK3-NEXT:    [[B_ADDR:%.*]] = alloca ptr, align 4
+// CHECK3-NEXT:    [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i32, align 4
 // CHECK3-NEXT:    [[DOTOMP_IV:%.*]] = alloca i32, align 4
 // CHECK3-NEXT:    [[TMP:%.*]] = alloca i32, align 4
 // CHECK3-NEXT:    [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4
@@ -2841,6 +2870,7 @@ int bar(int n){
 // CHECK3-NEXT:    store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
 // CHECK3-NEXT:    store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
 // CHECK3-NEXT:    store ptr [[B]], ptr [[B_ADDR]], align 4
+// CHECK3-NEXT:    store i32 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
 // CHECK3-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[B_ADDR]], align 4
 // CHECK3-NEXT:    store i32 0, ptr [[DOTOMP_COMB_LB]], align 4
 // CHECK3-NEXT:    store i32 9, ptr [[DOTOMP_COMB_UB]], align 4

diff --git a/clang/test/OpenMP/nvptx_target_teams_generic_loop_generic_mode_codegen.cpp b/clang/test/OpenMP/nvptx_target_teams_generic_loop_generic_mode_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_generic_loop_codegen.cpp b/clang/test/OpenMP/target_teams_generic_loop_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_generic_loop_codegen_as_distribute.cpp b/clang/test/OpenMP/target_teams_generic_loop_codegen_as_distribute.cpp
diff --git a/clang/test/OpenMP/target_teams_generic_loop_codegen_as_parallel_for.cpp b/clang/test/OpenMP/target_teams_generic_loop_codegen_as_parallel_for.cpp
diff --git a/clang/test/OpenMP/target_teams_generic_loop_if_codegen.cpp b/clang/test/OpenMP/target_teams_generic_loop_if_codegen.cpp
diff --git a/clang/test/OpenMP/target_teams_generic_loop_private_codegen.cpp b/clang/test/OpenMP/target_teams_generic_loop_private_codegen.cpp
diff --git a/clang/test/OpenMP/teams_generic_loop_codegen-1.cpp b/clang/test/OpenMP/teams_generic_loop_codegen-1.cpp
diff --git a/clang/test/OpenMP/teams_generic_loop_codegen.cpp b/clang/test/OpenMP/teams_generic_loop_codegen.cpp
diff --git a/clang/test/OpenMP/teams_generic_loop_collapse_codegen.cpp b/clang/test/OpenMP/teams_generic_loop_collapse_codegen.cpp
diff --git a/clang/test/OpenMP/teams_generic_loop_private_codegen.cpp b/clang/test/OpenMP/teams_generic_loop_private_codegen.cpp
diff --git a/clang/test/OpenMP/teams_generic_loop_reduction_codegen.cpp b/clang/test/OpenMP/teams_generic_loop_reduction_codegen.cpp
diff --git a/clang/test/ParserOpenACC/parse-clauses.c b/clang/test/ParserOpenACC/parse-clauses.c
@@ -173,45 +173,37 @@ void DefaultClause() {
 #pragma acc serial default), seq
   for(;;){}
 
-  // expected-error@+2{{expected identifier}}
-  // expected-warning@+1{{OpenACC clause 'default' not yet implemented, clause ignored}}
+  // expected-error@+1{{expected identifier}}
 #pragma acc serial default()
   for(;;){}
 
-  // expected-error@+3{{expected identifier}}
-  // expected-warning@+2{{OpenACC clause 'default' not yet implemented, clause ignored}}
+  // expected-error@+2{{expected identifier}}
   // expected-warning@+1{{OpenACC clause 'seq' not yet implemented, clause ignored}}
 #pragma acc serial default() seq
   for(;;){}
 
-  // expected-error@+3{{expected identifier}}
-  // expected-warning@+2{{OpenACC clause 'default' not yet implemented, clause ignored}}
+  // expected-error@+2{{expected identifier}}
   // expected-warning@+1{{OpenACC clause 'seq' not yet implemented, clause ignored}}
 #pragma acc serial default(), seq
   for(;;){}
 
-  // expected-error@+2{{invalid value for 'default' clause; expected 'present' or 'none'}}
-  // expected-warning@+1{{OpenACC clause 'default' not yet implemented, clause ignored}}
+  // expected-error@+1{{invalid value for 'default' clause; expected 'present' or 'none'}}
 #pragma acc serial default(invalid)
   for(;;){}
 
-  // expected-error@+3{{invalid value for 'default' clause; expected 'present' or 'none'}}
-  // expected-warning@+2{{OpenACC clause 'default' not yet implemented, clause ignored}}
+  // expected-error@+2{{invalid value for 'default' clause; expected 'present' or 'none'}}
   // expected-warning@+1{{OpenACC clause 'seq' not yet implemented, clause ignored}}
 #pragma acc serial default(auto) seq
   for(;;){}
 
-  // expected-error@+3{{invalid value for 'default' clause; expected 'present' or 'none'}}
-  // expected-warning@+2{{OpenACC clause 'default' not yet implemented, clause ignored}}
+  // expected-error@+2{{invalid value for 'default' clause; expected 'present' or 'none'}}
   // expected-warning@+1{{OpenACC clause 'seq' not yet implemented, clause ignored}}
 #pragma acc serial default(invalid), seq
   for(;;){}
 
-  // expected-warning@+1{{OpenACC clause 'default' not yet implemented, clause ignored}}
 #pragma acc serial default(none)
   for(;;){}
 
-  // expected-warning@+2{{OpenACC clause 'default' not yet implemented, clause ignored}}
   // expected-warning@+1{{OpenACC clause 'seq' not yet implemented, clause ignored}}
 #pragma acc serial default(present), seq
   for(;;){}

diff --git a/clang/test/Sema/aarch64-incompat-sm-builtin-calls.c b/clang/test/Sema/aarch64-incompat-sm-builtin-calls.c
@@ -1,6 +1,6 @@
 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
 // RUN: %clang_cc1  -triple aarch64-none-linux-gnu -target-feature +sve \
-// RUN:   -target-feature +sme2 -target-feature +sve2 -target-feature +neon -fsyntax-only -verify %s
+// RUN:   -target-feature +sme2 -target-feature +sve2 -target-feature +neon -Waarch64-sme-attributes -fsyntax-only -verify %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -33,6 +33,7 @@ svuint32_t incompat_sve_sm(svbool_t pg, svuint32_t a, int16_t b) __arm_streaming
   return __builtin_sve_svld1_gather_u32base_index_u32(pg, a, b);
 }
 
+// expected-warning@+1 {{passing/returning a VL-dependent argument to/from a __arm_locally_streaming function. The streaming and non-streaming vector lengths may be different}}
 __arm_locally_streaming svuint32_t incompat_sve_ls(svbool_t pg, svuint32_t a, int64_t b) {
   // expected-warning@+1 {{builtin call has undefined behaviour when called from a streaming function}}
   return __builtin_sve_svld1_gather_u32base_index_u32(pg, a, b);
@@ -48,6 +49,7 @@ svuint32_t incompat_sve2_sm(svbool_t pg, svuint32_t a, int64_t b) __arm_streamin
   return __builtin_sve_svldnt1_gather_u32base_index_u32(pg, a, b);
 }
 
+// expected-warning@+1 {{passing/returning a VL-dependent argument to/from a __arm_locally_streaming function. The streaming and non-streaming vector lengths may be different}}
 __arm_locally_streaming svuint32_t incompat_sve2_ls(svbool_t pg, svuint32_t a, int64_t b) {
   // expected-warning@+1 {{builtin call has undefined behaviour when called from a streaming function}}
   return __builtin_sve_svldnt1_gather_u32base_index_u32(pg, a, b);
@@ -68,6 +70,7 @@ svfloat64_t streaming_caller_sve(svbool_t pg, svfloat64_t a, float64_t b) __arm_
   return svadd_n_f64_m(pg, a, b);
 }
 
+// expected-warning@+1 {{passing/returning a VL-dependent argument to/from a __arm_locally_streaming function. The streaming and non-streaming vector lengths may be different}}
 __arm_locally_streaming svfloat64_t locally_streaming_caller_sve(svbool_t pg, svfloat64_t a, float64_t b) {
   // expected-no-warning
   return svadd_n_f64_m(pg, a, b);
@@ -83,6 +86,7 @@ svint16_t streaming_caller_sve2(svint16_t op1, svint16_t op2) __arm_streaming {
   return svmul_lane_s16(op1, op2, 0);
 }
 
+// expected-warning@+1 {{passing/returning a VL-dependent argument to/from a __arm_locally_streaming function. The streaming and non-streaming vector lengths may be different}}
 __arm_locally_streaming svint16_t locally_streaming_caller_sve2(svint16_t op1, svint16_t op2) {
   // expected-no-warning
   return svmul_lane_s16(op1, op2, 0);

diff --git a/clang/test/Sema/aarch64-sme-func-attrs.c b/clang/test/Sema/aarch64-sme-func-attrs.c
@@ -1,5 +1,5 @@
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -fsyntax-only -verify %s
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -fsyntax-only -verify=expected-cpp -x c++ %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -target-feature +sve -Waarch64-sme-attributes -fsyntax-only -verify %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -target-feature +sve -Waarch64-sme-attributes -fsyntax-only -verify=expected-cpp -x c++ %s
 
 // Valid attributes
 
@@ -496,3 +496,135 @@ void fmv_caller() {
     just_fine();
     incompatible_locally_streaming();
 }
+
+void sme_streaming_with_vl_arg(__SVInt8_t a) __arm_streaming { }
+
+__SVInt8_t sme_streaming_returns_vl(void) __arm_streaming { __SVInt8_t r; return r; }
+
+void sme_streaming_compatible_with_vl_arg(__SVInt8_t a) __arm_streaming_compatible { }
+
+__SVInt8_t sme_streaming_compatible_returns_vl(void) __arm_streaming_compatible { __SVInt8_t r; return r; }
+
+void sme_no_streaming_with_vl_arg(__SVInt8_t a) { }
+
+__SVInt8_t sme_no_streaming_returns_vl(void) { __SVInt8_t r; return r; }
+
+// expected-warning@+2 {{passing/returning a VL-dependent argument to/from a __arm_locally_streaming function. The streaming and non-streaming vector lengths may be different}}
+// expected-cpp-warning@+1 {{passing/returning a VL-dependent argument to/from a __arm_locally_streaming function. The streaming and non-streaming vector lengths may be different}}
+__arm_locally_streaming void sme_locally_streaming_with_vl_arg(__SVInt8_t a) { }
+
+// expected-warning@+2 {{passing/returning a VL-dependent argument to/from a __arm_locally_streaming function. The streaming and non-streaming vector lengths may be different}}
+// expected-cpp-warning@+1 {{passing/returning a VL-dependent argument to/from a __arm_locally_streaming function. The streaming and non-streaming vector lengths may be different}}
+__arm_locally_streaming __SVInt8_t sme_locally_streaming_returns_vl(void) { __SVInt8_t r; return r; }
+
+void sme_no_streaming_calling_streaming_with_vl_args() {
+  __SVInt8_t a;
+  // expected-warning@+2 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  // expected-cpp-warning@+1 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  sme_streaming_with_vl_arg(a);
+}
+
+void sme_no_streaming_calling_streaming_with_return_vl() {
+  // expected-warning@+2 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  // expected-cpp-warning@+1 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  __SVInt8_t r = sme_streaming_returns_vl();
+}
+
+void sme_streaming_calling_non_streaming_with_vl_args(void) __arm_streaming {
+  __SVInt8_t a;
+  // expected-warning@+2 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  // expected-cpp-warning@+1 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  sme_no_streaming_with_vl_arg(a);
+}
+
+void sme_streaming_calling_non_streaming_with_return_vl(void) __arm_streaming {
+  // expected-warning@+2 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  // expected-cpp-warning@+1 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  __SVInt8_t r = sme_no_streaming_returns_vl();
+}
+
+void sme_no_streaming_calling_streaming_with_vl_args_param(__SVInt8_t arg, void (*sc)( __SVInt8_t arg) __arm_streaming) {
+  // expected-warning@+2 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  // expected-cpp-warning@+1 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  sc(arg);
+}
+
+__SVInt8_t sme_no_streaming_calling_streaming_return_vl_param(__SVInt8_t (*s)(void) __arm_streaming) {
+  // expected-warning@+2 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  // expected-cpp-warning@+1 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  return s();
+}
+
+void sme_streaming_compatible_calling_streaming_with_vl_args(__SVInt8_t arg) __arm_streaming_compatible {
+  // expected-warning@+2 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  // expected-cpp-warning@+1 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  sme_streaming_with_vl_arg(arg);
+}
+
+void sme_streaming_compatible_calling_sme_streaming_return_vl(void) __arm_streaming_compatible {
+  // expected-warning@+2 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  // expected-cpp-warning@+1 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  __SVInt8_t r = sme_streaming_returns_vl();
+}
+
+void sme_streaming_compatible_calling_no_streaming_with_vl_args(__SVInt8_t arg) __arm_streaming_compatible {
+  // expected-warning@+2 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  // expected-cpp-warning@+1 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  sme_no_streaming_with_vl_arg(arg);
+}
+
+void sme_streaming_compatible_calling_no_sme_streaming_return_vl(void) __arm_streaming_compatible {
+  // expected-warning@+2 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  // expected-cpp-warning@+1 {{passing a VL-dependent argument to/from a function that has a different streaming-mode. The streaming and non-streaming vector lengths may be different}}
+  __SVInt8_t r = sme_no_streaming_returns_vl();
+}
+
+void sme_streaming_calling_streaming(__SVInt8_t arg, void (*s)( __SVInt8_t arg) __arm_streaming) __arm_streaming {
+  s(arg);
+}
+
+__SVInt8_t sme_streaming_calling_streaming_return_vl(__SVInt8_t (*s)(void) __arm_streaming) __arm_streaming {
+  return s();
+}
+
+void sme_streaming_calling_streaming_with_vl_args(__SVInt8_t a) __arm_streaming {
+  sme_streaming_with_vl_arg(a);
+}
+
+void sme_streaming_calling_streaming_with_return_vl(void) __arm_streaming {
+  __SVInt8_t r = sme_streaming_returns_vl();
+}
+
+void sme_streaming_calling_streaming_compatible_with_vl_args(__SVInt8_t a) __arm_streaming {
+  sme_streaming_compatible_with_vl_arg(a);
+}
+
+void sme_streaming_calling_streaming_compatible_with_return_vl(void) __arm_streaming {
+  __SVInt8_t r = sme_streaming_compatible_returns_vl();
+}
+
+void sme_no_streaming_calling_streaming_compatible_with_vl_args() {
+  __SVInt8_t a;
+  sme_streaming_compatible_with_vl_arg(a);
+}
+
+void sme_no_streaming_calling_streaming_compatible_with_return_vl() {
+  __SVInt8_t r = sme_streaming_compatible_returns_vl();
+}
+
+void sme_no_streaming_calling_non_streaming_compatible_with_vl_args() {
+  __SVInt8_t a;
+  sme_no_streaming_with_vl_arg(a);
+}
+
+void sme_no_streaming_calling_non_streaming_compatible_with_return_vl() {
+  __SVInt8_t r = sme_no_streaming_returns_vl();
+}
+
+void sme_streaming_compatible_calling_streaming_compatible_with_vl_args(__SVInt8_t arg) __arm_streaming_compatible {
+  sme_streaming_compatible_with_vl_arg(arg);
+}
+
+void sme_streaming_compatible_calling_streaming_compatible_with_return_vl(void) __arm_streaming_compatible {
+  __SVInt8_t r = sme_streaming_compatible_returns_vl();
+}
diff --git a/clang/test/SemaCXX/type-traits.cpp b/clang/test/SemaCXX/type-traits.cpp
@@ -1759,6 +1759,7 @@ void is_layout_compatible(int n)
   // expected-error@-1 {{variable length arrays are not supported in '__is_layout_compatible'}}
   static_assert(!__is_layout_compatible(int[n], int[n]));
   // expected-error@-1 {{variable length arrays are not supported in '__is_layout_compatible'}}
+  // expected-error@-2 {{variable length arrays are not supported in '__is_layout_compatible'}}
   static_assert(__is_layout_compatible(int&, int&));
   static_assert(!__is_layout_compatible(int&, char&));
   static_assert(__is_layout_compatible(void(int), void(int)));

diff --git a/clang/test/SemaOpenACC/compute-construct-ast.cpp b/clang/test/SemaOpenACC/compute-construct-ast.cpp
@@ -12,14 +12,16 @@ void NormalFunc() {
   // CHECK-LABEL: NormalFunc
   // CHECK-NEXT: CompoundStmt
   // CHECK-NEXT: OpenACCComputeConstruct {{.*}}parallel
+  // CHECK-NEXT: default(none)
   // CHECK-NEXT: CompoundStmt
-#pragma acc parallel
+#pragma acc parallel default(none)
   {
 #pragma acc parallel
   // CHECK-NEXT: OpenACCComputeConstruct {{.*}}parallel
   // CHECK-NEXT: OpenACCComputeConstruct {{.*}}parallel
+  // CHECK-NEXT: default(present)
   // CHECK-NEXT: CompoundStmt
-#pragma acc parallel
+#pragma acc parallel default(present)
     {}
   }
   // FIXME: Add a test once we have clauses for this.
@@ -50,12 +52,12 @@ void NormalFunc() {
 
 template<typename T>
 void TemplFunc() {
-#pragma acc parallel
+#pragma acc parallel default(none)
   {
     typename T::type I;
   }
 
-#pragma acc serial
+#pragma acc serial default(present)
   {
     typename T::type I;
   }
@@ -72,10 +74,12 @@ void TemplFunc() {
   // CHECK-NEXT: FunctionDecl
   // CHECK-NEXT: CompoundStmt
   // CHECK-NEXT: OpenACCComputeConstruct {{.*}}parallel
+  // CHECK-NEXT: default(none)
   // CHECK-NEXT: CompoundStmt
   // CHECK-NEXT: DeclStmt
   // CHECK-NEXT: VarDecl{{.*}} I 'typename T::type'
   // CHECK-NEXT: OpenACCComputeConstruct {{.*}}serial
+  // CHECK-NEXT: default(present)
   // CHECK-NEXT: CompoundStmt
   // CHECK-NEXT: DeclStmt
   // CHECK-NEXT: VarDecl{{.*}} I 'typename T::type'
@@ -91,10 +95,12 @@ void TemplFunc() {
   // CHECK-NEXT: CXXRecord
   // CHECK-NEXT: CompoundStmt
   // CHECK-NEXT: OpenACCComputeConstruct {{.*}}parallel
+  // CHECK-NEXT: default(none)
   // CHECK-NEXT: CompoundStmt
   // CHECK-NEXT: DeclStmt
   // CHECK-NEXT: VarDecl{{.*}} I 'typename S::type':'int'
   // CHECK-NEXT: OpenACCComputeConstruct {{.*}}serial
+  // CHECK-NEXT: default(present)
   // CHECK-NEXT: CompoundStmt
   // CHECK-NEXT: DeclStmt
   // CHECK-NEXT: VarDecl{{.*}} I 'typename S::type':'int'

diff --git a/clang/test/SemaOpenACC/compute-construct-clause-ast.cpp b/clang/test/SemaOpenACC/compute-construct-clause-ast.cpp
@@ -0,0 +1,74 @@
+// RUN: %clang_cc1 %s -fopenacc -ast-dump | FileCheck %s
+
+// Test this with PCH.
+// RUN: %clang_cc1 %s -fopenacc -emit-pch -o %t %s
+// RUN: %clang_cc1 %s -fopenacc -include-pch %t -ast-dump-all | FileCheck %s
+
+#ifndef PCH_HELPER
+#define PCH_HELPER
+void NormalFunc() {
+  // CHECK: FunctionDecl{{.*}}NormalFunc
+  // CHECK-NEXT: CompoundStmt
+#pragma acc parallel default(none)
+  while(true);
+  // CHECK-NEXT: OpenACCComputeConstruct{{.*}}parallel
+  // CHECK-NEXT: default(none)
+  // CHECK-NEXT: WhileStmt
+  // CHECK-NEXT: CXXBoolLiteralExpr
+  // CHECK-NEXT: NullStmt
+
+#pragma acc serial default(present)
+  while(true);
+  // CHECK-NEXT: OpenACCComputeConstruct{{.*}}serial
+  // CHECK-NEXT: default(present)
+  // CHECK-NEXT: WhileStmt
+  // CHECK-NEXT: CXXBoolLiteralExpr
+  // CHECK-NEXT: NullStmt
+}
+
+template<typename T>
+void TemplFunc() {
+  // CHECK: FunctionTemplateDecl{{.*}}TemplFunc
+  // CHECK-NEXT: TemplateTypeParmDecl
+
+  // Match the prototype:
+  // CHECK-NEXT: FunctionDecl{{.*}}TemplFunc
+  // CHECK-NEXT: CompoundStmt
+
+#pragma acc kernels default(none)
+  while(true);
+  // CHECK-NEXT: OpenACCComputeConstruct{{.*}}kernels
+  // CHECK-NEXT: default(none)
+  // CHECK-NEXT: WhileStmt
+  // CHECK-NEXT: CXXBoolLiteralExpr
+  // CHECK-NEXT: NullStmt
+
+#pragma acc parallel default(present)
+  while(true);
+  // CHECK-NEXT: OpenACCComputeConstruct{{.*}}parallel
+  // CHECK-NEXT: default(present)
+  // CHECK-NEXT: WhileStmt
+  // CHECK-NEXT: CXXBoolLiteralExpr
+  // CHECK-NEXT: NullStmt
+
+  // Match the instantiation:
+  // CHECK: FunctionDecl{{.*}}TemplFunc{{.*}}implicit_instantiation
+  // CHECK-NEXT: TemplateArgument type 'int'
+  // CHECK-NEXT: BuiltinType
+  // CHECK-NEXT: CompoundStmt
+  // CHECK-NEXT: OpenACCComputeConstruct{{.*}}kernels
+  // CHECK-NEXT: default(none)
+  // CHECK-NEXT: WhileStmt
+  // CHECK-NEXT: CXXBoolLiteralExpr
+  // CHECK-NEXT: NullStmt
+  // CHECK-NEXT: OpenACCComputeConstruct{{.*}}parallel
+  // CHECK-NEXT: default(present)
+  // CHECK-NEXT: WhileStmt
+  // CHECK-NEXT: CXXBoolLiteralExpr
+  // CHECK-NEXT: NullStmt
+}
+
+void Instantiate() {
+  TemplFunc<int>();
+}
+#endif
diff --git a/clang/test/SemaOpenACC/compute-construct-default-clause.c b/clang/test/SemaOpenACC/compute-construct-default-clause.c
@@ -0,0 +1,55 @@
+// RUN: %clang_cc1 %s -fopenacc -verify
+
+void SingleOnly() {
+  #pragma acc parallel default(none)
+  while(0);
+
+  // expected-warning@+3{{OpenACC clause 'seq' not yet implemented}}
+  // expected-error@+2{{OpenACC 'default' clause cannot appear more than once on a 'serial' directive}}
+  // expected-note@+1{{previous clause is here}}
+  #pragma acc serial default(present) seq default(none)
+  while(0);
+
+  // expected-warning@+5{{OpenACC clause 'seq' not yet implemented}}
+  // expected-warning@+4{{OpenACC clause 'seq' not yet implemented}}
+  // expected-warning@+3{{OpenACC clause 'seq' not yet implemented}}
+  // expected-error@+2{{OpenACC 'default' clause cannot appear more than once on a 'kernels' directive}}
+  // expected-note@+1{{previous clause is here}}
+  #pragma acc kernels seq default(present) seq default(none) seq
+  while(0);
+
+  // expected-warning@+6{{OpenACC construct 'parallel loop' not yet implemented}}
+  // expected-warning@+5{{OpenACC clause 'seq' not yet implemented}}
+  // expected-warning@+4{{OpenACC clause 'seq' not yet implemented}}
+  // expected-warning@+3{{OpenACC clause 'seq' not yet implemented}}
+  // expected-warning@+2{{OpenACC clause 'default' not yet implemented}}
+  // expected-warning@+1{{OpenACC clause 'default' not yet implemented}}
+  #pragma acc parallel loop seq default(present) seq default(none) seq
+  while(0);
+
+  // expected-warning@+3{{OpenACC construct 'serial loop' not yet implemented}}
+  // expected-warning@+2{{OpenACC clause 'seq' not yet implemented}}
+  // expected-error@+1{{expected '('}}
+  #pragma acc serial loop seq default seq default(none) seq
+  while(0);
+
+  // expected-warning@+2{{OpenACC construct 'kernels loop' not yet implemented}}
+  // expected-warning@+1{{OpenACC clause 'default' not yet implemented}}
+  #pragma acc kernels loop default(none)
+  while(0);
+
+  // expected-warning@+2{{OpenACC construct 'data' not yet implemented}}
+  // expected-warning@+1{{OpenACC clause 'default' not yet implemented}}
+  #pragma acc data default(none)
+  while(0);
+
+  // expected-warning@+2{{OpenACC construct 'loop' not yet implemented}}
+  // expected-error@+1{{OpenACC 'default' clause is not valid on 'loop' directive}}
+  #pragma acc loop default(none)
+  while(0);
+
+  // expected-warning@+2{{OpenACC construct 'wait' not yet implemented}}
+  // expected-error@+1{{OpenACC 'default' clause is not valid on 'wait' directive}}
+  #pragma acc wait default(none)
+  while(0);
+}
diff --git a/clang/test/SemaOpenACC/compute-construct-default-clause.cpp b/clang/test/SemaOpenACC/compute-construct-default-clause.cpp
@@ -0,0 +1,39 @@
+// RUN: %clang_cc1 %s -fopenacc -verify
+
+template<typename T>
+void SingleOnly() {
+  #pragma acc parallel default(none)
+  while(false);
+
+  // expected-warning@+3{{OpenACC clause 'seq' not yet implemented}}
+  // expected-error@+2{{OpenACC 'default' clause cannot appear more than once on a 'parallel' directive}}
+  // expected-note@+1{{previous clause is here}}
+  #pragma acc parallel default(present) seq default(none)
+  while(false);
+
+  // expected-warning@+5{{OpenACC clause 'seq' not yet implemented}}
+  // expected-warning@+4{{OpenACC clause 'seq' not yet implemented}}
+  // expected-warning@+3{{OpenACC clause 'seq' not yet implemented}}
+  // expected-error@+2{{OpenACC 'default' clause cannot appear more than once on a 'serial' directive}}
+  // expected-note@+1{{previous clause is here}}
+  #pragma acc serial seq default(present) seq default(none) seq
+  while(false);
+
+  // expected-warning@+5{{OpenACC clause 'seq' not yet implemented}}
+  // expected-warning@+4{{OpenACC clause 'seq' not yet implemented}}
+  // expected-warning@+3{{OpenACC clause 'seq' not yet implemented}}
+  // expected-error@+2{{OpenACC 'default' clause cannot appear more than once on a 'kernels' directive}}
+  // expected-note@+1{{previous clause is here}}
+  #pragma acc kernels seq default(present) seq default(none) seq
+  while(false);
+
+  // expected-warning@+3{{OpenACC clause 'seq' not yet implemented}}
+  // expected-warning@+2{{OpenACC clause 'seq' not yet implemented}}
+  // expected-error@+1{{expected '('}}
+  #pragma acc parallel seq default(none) seq default seq
+  while(false);
+}
+
+void Instantiate() {
+  SingleOnly<int>();
+}
diff --git a/clang/test/SemaTemplate/alias-templates.cpp b/clang/test/SemaTemplate/alias-templates.cpp
@@ -236,6 +236,29 @@ namespace PR14858 {
   void test_q(int (&a)[5]) { Q<B, B, B>().f<B, B>(&a); }
 }
 
+namespace PR84220 {
+
+template <class...> class list {};
+
+template <int> struct foo_impl {
+  template <class> using f = int;
+};
+
+template <class... xs>
+using foo = typename foo_impl<sizeof...(xs)>::template f<xs...>;
+
+// We call getFullyPackExpandedSize at the annotation stage
+// before parsing the ellipsis next to the foo<xs>. This happens before
+// a PackExpansionType is formed for foo<xs>.
+// getFullyPackExpandedSize shouldn't determine the value here. Otherwise,
+// foo_impl<sizeof...(xs)> would lose its dependency despite the template
+// arguments being unsubstituted.
+template <class... xs> using test = list<foo<xs>...>;
+
+test<int> a;
+
+}
+
 namespace redecl {
   template<typename> using A = int;
   template<typename = void> using A = int;

diff --git a/clang/test/SemaTemplate/ms-function-specialization-class-scope.cpp b/clang/test/SemaTemplate/ms-function-specialization-class-scope.cpp
@@ -1,6 +1,7 @@
-// RUN: %clang_cc1 -fms-extensions -fsyntax-only -Wno-unused-value -verify %s
-// RUN: %clang_cc1 -fms-extensions -fdelayed-template-parsing -fsyntax-only -Wno-unused-value -verify %s
+// RUN: %clang_cc1 -fms-extensions -fsyntax-only -verify %s
+// RUN: %clang_cc1 -fms-extensions -fdelayed-template-parsing -fsyntax-only -verify %s
 
+// expected-no-diagnostics
 class A {
 public:
   template<class U> A(U p) {}
@@ -75,42 +76,3 @@ struct S {
   int f<0>(int);
 };
 }
-
-namespace UsesThis {
-  template<typename T>
-  struct A {
-    int x;
-
-    template<typename U>
-    static void f();
-
-    template<>
-    void f<int>() {
-      this->x; // expected-error {{invalid use of 'this' outside of a non-static member function}}
-      x; // expected-error {{invalid use of member 'x' in static member function}}
-      A::x; // expected-error {{invalid use of member 'x' in static member function}}
-      +x; // expected-error {{invalid use of member 'x' in static member function}}
-      +A::x; // expected-error {{invalid use of member 'x' in static member function}}
-    }
-
-    template<typename U>
-    void g();
-
-    template<>
-    void g<int>() {
-      this->x;
-      x;
-      A::x;
-      +x;
-      +A::x;
-    }
-
-    template<typename U>
-    static auto h() -> A*;
-
-    template<>
-    auto h<int>() -> decltype(this); // expected-error {{'this' cannot be used in a static member function declaration}}
-  };
-
-  template struct A<int>; // expected-note 2{{in instantiation of}}
-}