diff --git a/clang/lib/Analysis/FlowSensitive/DataflowEnvironment.cpp b/clang/lib/Analysis/FlowSensitive/DataflowEnvironment.cpp
@@ -166,17 +166,16 @@ static Value *joinDistinctValues(QualType Type, Value &Val1,
   return JoinedVal;
 }
 
-// When widening does not change `Current`, return value will equal `&Prev`.
-static Value &widenDistinctValues(QualType Type, Value &Prev,
-                                  const Environment &PrevEnv, Value &Current,
-                                  Environment &CurrentEnv,
-                                  Environment::ValueModel &Model) {
+static WidenResult widenDistinctValues(QualType Type, Value &Prev,
+                                       const Environment &PrevEnv,
+                                       Value &Current, Environment &CurrentEnv,
+                                       Environment::ValueModel &Model) {
   // Boolean-model widening.
   if (auto *PrevBool = dyn_cast<BoolValue>(&Prev)) {
-    // If previous value was already Top, re-use that to (implicitly) indicate
-    // that no change occurred.
     if (isa<TopBoolValue>(Prev))
-      return Prev;
+      // Safe to return `Prev` here, because Top is never dependent on the
+      // environment.
+      return {&Prev, LatticeEffect::Unchanged};
 
     // We may need to widen to Top, but before we do so, check whether both
     // values are implied to be either true or false in the current environment.
@@ -185,22 +184,24 @@ static Value &widenDistinctValues(QualType Type, Value &Prev,
     bool TruePrev = PrevEnv.proves(PrevBool->formula());
     bool TrueCur = CurrentEnv.proves(CurBool.formula());
     if (TruePrev && TrueCur)
-      return CurrentEnv.getBoolLiteralValue(true);
+      return {&CurrentEnv.getBoolLiteralValue(true), LatticeEffect::Unchanged};
     if (!TruePrev && !TrueCur &&
         PrevEnv.proves(PrevEnv.arena().makeNot(PrevBool->formula())) &&
         CurrentEnv.proves(CurrentEnv.arena().makeNot(CurBool.formula())))
-      return CurrentEnv.getBoolLiteralValue(false);
+      return {&CurrentEnv.getBoolLiteralValue(false), LatticeEffect::Unchanged};
 
-    return CurrentEnv.makeTopBoolValue();
+    return {&CurrentEnv.makeTopBoolValue(), LatticeEffect::Changed};
   }
 
   // FIXME: Add other built-in model widening.
 
   // Custom-model widening.
-  if (auto *W = Model.widen(Type, Prev, PrevEnv, Current, CurrentEnv))
-    return *W;
+  if (auto Result = Model.widen(Type, Prev, PrevEnv, Current, CurrentEnv))
+    return *Result;
 
-  return equateUnknownValues(Prev.getKind()) ? Prev : Current;
+  return {&Current, equateUnknownValues(Prev.getKind())
+                        ? LatticeEffect::Unchanged
+                        : LatticeEffect::Changed};
 }
 
 // Returns whether the values in `Map1` and `Map2` compare equal for those
@@ -271,7 +272,7 @@ llvm::MapVector<Key, Value *>
 widenKeyToValueMap(const llvm::MapVector<Key, Value *> &CurMap,
                    const llvm::MapVector<Key, Value *> &PrevMap,
                    Environment &CurEnv, const Environment &PrevEnv,
-                   Environment::ValueModel &Model, LatticeJoinEffect &Effect) {
+                   Environment::ValueModel &Model, LatticeEffect &Effect) {
   llvm::MapVector<Key, Value *> WidenedMap;
   for (auto &Entry : CurMap) {
     Key K = Entry.first;
@@ -290,11 +291,11 @@ widenKeyToValueMap(const llvm::MapVector<Key, Value *> &CurMap,
       continue;
     }
 
-    Value &WidenedVal = widenDistinctValues(K->getType(), *PrevIt->second,
-                                            PrevEnv, *Val, CurEnv, Model);
-    WidenedMap.insert({K, &WidenedVal});
-    if (&WidenedVal != PrevIt->second)
-      Effect = LatticeJoinEffect::Changed;
+    auto [WidenedVal, ValEffect] = widenDistinctValues(
+        K->getType(), *PrevIt->second, PrevEnv, *Val, CurEnv, Model);
+    WidenedMap.insert({K, WidenedVal});
+    if (ValEffect == LatticeEffect::Changed)
+      Effect = LatticeEffect::Changed;
   }
 
   return WidenedMap;
@@ -617,15 +618,15 @@ bool Environment::equivalentTo(const Environment &Other,
   return true;
 }
 
-LatticeJoinEffect Environment::widen(const Environment &PrevEnv,
-                                     Environment::ValueModel &Model) {
+LatticeEffect Environment::widen(const Environment &PrevEnv,
+                                 Environment::ValueModel &Model) {
   assert(DACtx == PrevEnv.DACtx);
   assert(ReturnVal == PrevEnv.ReturnVal);
   assert(ReturnLoc == PrevEnv.ReturnLoc);
   assert(ThisPointeeLoc == PrevEnv.ThisPointeeLoc);
   assert(CallStack == PrevEnv.CallStack);
 
-  auto Effect = LatticeJoinEffect::Unchanged;
+  auto Effect = LatticeEffect::Unchanged;
 
   // By the API, `PrevEnv` is a previous version of the environment for the same
   // block, so we have some guarantees about its shape. In particular, it will
@@ -646,7 +647,7 @@ LatticeJoinEffect Environment::widen(const Environment &PrevEnv,
       ExprToLoc.size() != PrevEnv.ExprToLoc.size() ||
       ExprToVal.size() != PrevEnv.ExprToVal.size() ||
       LocToVal.size() != PrevEnv.LocToVal.size())
-    Effect = LatticeJoinEffect::Changed;
+    Effect = LatticeEffect::Changed;
 
   return Effect;
 }

diff --git a/clang/lib/Basic/Targets/M68k.cpp b/clang/lib/Basic/Targets/M68k.cpp
@@ -127,16 +127,21 @@ bool M68kTargetInfo::hasFeature(StringRef Feature) const {
 
 const char *const M68kTargetInfo::GCCRegNames[] = {
     "d0", "d1", "d2", "d3", "d4", "d5", "d6", "d7",
-    "a0", "a1", "a2", "a3", "a4", "a5", "a6", "a7",
+    "a0", "a1", "a2", "a3", "a4", "a5", "a6", "sp",
     "pc"};
 
 ArrayRef<const char *> M68kTargetInfo::getGCCRegNames() const {
   return llvm::ArrayRef(GCCRegNames);
 }
 
+const TargetInfo::GCCRegAlias M68kTargetInfo::GCCRegAliases[] = {
+    {{"bp"}, "a5"},
+    {{"fp"}, "a6"},
+    {{"usp", "ssp", "isp", "a7"}, "sp"},
+};
+
 ArrayRef<TargetInfo::GCCRegAlias> M68kTargetInfo::getGCCRegAliases() const {
-  // No aliases.
-  return std::nullopt;
+  return llvm::ArrayRef(GCCRegAliases);
 }
 
 bool M68kTargetInfo::validateAsmConstraint(

diff --git a/clang/lib/Basic/Targets/M68k.h b/clang/lib/Basic/Targets/M68k.h
@@ -25,6 +25,7 @@ namespace targets {
 
 class LLVM_LIBRARY_VISIBILITY M68kTargetInfo : public TargetInfo {
   static const char *const GCCRegNames[];
+  static const TargetInfo::GCCRegAlias GCCRegAliases[];
 
   enum CPUKind {
     CK_Unknown,

diff --git a/clang/lib/Basic/Targets/Mips.h b/clang/lib/Basic/Targets/Mips.h
@@ -318,6 +318,7 @@ class LLVM_LIBRARY_VISIBILITY MipsTargetInfo : public TargetInfo {
     FPMode = isFP64Default() ? FP64 : FPXX;
     NoOddSpreg = false;
     bool OddSpregGiven = false;
+    bool StrictAlign = false;
 
     for (const auto &Feature : Features) {
       if (Feature == "+single-float")
@@ -330,6 +331,10 @@ class LLVM_LIBRARY_VISIBILITY MipsTargetInfo : public TargetInfo {
         IsMicromips = true;
       else if (Feature == "+mips32r6" || Feature == "+mips64r6")
         HasUnalignedAccess = true;
+      // We cannot be sure that the order of strict-align vs mips32r6.
+      // Thus we need an extra variable here.
+      else if (Feature == "+strict-align")
+        StrictAlign = true;
       else if (Feature == "+dsp")
         DspRev = std::max(DspRev, DSP1);
       else if (Feature == "+dspr2")
@@ -368,6 +373,9 @@ class LLVM_LIBRARY_VISIBILITY MipsTargetInfo : public TargetInfo {
     if (FPMode == FPXX && !OddSpregGiven)
       NoOddSpreg = true;
 
+    if (StrictAlign)
+      HasUnalignedAccess = false;
+
     setDataLayout();
 
     return true;

diff --git a/clang/lib/CodeGen/BackendUtil.cpp b/clang/lib/CodeGen/BackendUtil.cpp
@@ -73,18 +73,17 @@
 #include "llvm/Transforms/Instrumentation/HWAddressSanitizer.h"
 #include "llvm/Transforms/Instrumentation/InstrProfiling.h"
 #include "llvm/Transforms/Instrumentation/KCFI.h"
+#include "llvm/Transforms/Instrumentation/LowerAllowCheckPass.h"
 #include "llvm/Transforms/Instrumentation/MemProfiler.h"
 #include "llvm/Transforms/Instrumentation/MemorySanitizer.h"
 #include "llvm/Transforms/Instrumentation/PGOInstrumentation.h"
-#include "llvm/Transforms/Instrumentation/RemoveTrapsPass.h"
 #include "llvm/Transforms/Instrumentation/SanitizerBinaryMetadata.h"
 #include "llvm/Transforms/Instrumentation/SanitizerCoverage.h"
 #include "llvm/Transforms/Instrumentation/ThreadSanitizer.h"
 #include "llvm/Transforms/ObjCARC.h"
 #include "llvm/Transforms/Scalar/EarlyCSE.h"
 #include "llvm/Transforms/Scalar/GVN.h"
 #include "llvm/Transforms/Scalar/JumpThreading.h"
-#include "llvm/Transforms/Scalar/SimplifyCFG.h"
 #include "llvm/Transforms/Utils/Debugify.h"
 #include "llvm/Transforms/Utils/EntryExitInstrumenter.h"
 #include "llvm/Transforms/Utils/ModuleUtils.h"
@@ -100,21 +99,17 @@ using namespace llvm;
 namespace llvm {
 extern cl::opt<bool> PrintPipelinePasses;
 
-cl::opt<bool> ClRemoveTraps("clang-remove-traps", cl::Optional,
-                            cl::desc("Insert remove-traps pass."),
-                            cl::init(false));
-
 // Experiment to move sanitizers earlier.
 static cl::opt<bool> ClSanitizeOnOptimizerEarlyEP(
     "sanitizer-early-opt-ep", cl::Optional,
-    cl::desc("Insert sanitizers on OptimizerEarlyEP."), cl::init(false));
+    cl::desc("Insert sanitizers on OptimizerEarlyEP."));
 
 extern cl::opt<InstrProfCorrelator::ProfCorrelatorKind> ProfileCorrelate;
 
 // Re-link builtin bitcodes after optimization
 cl::opt<bool> ClRelinkBuiltinBitcodePostop(
     "relink-builtin-bitcode-postop", cl::Optional,
-    cl::desc("Re-link builtin bitcodes after optimization."), cl::init(false));
+    cl::desc("Re-link builtin bitcodes after optimization."));
 } // namespace llvm
 
 namespace {
@@ -751,18 +746,13 @@ static void addSanitizers(const Triple &TargetTriple,
     PB.registerOptimizerLastEPCallback(SanitizersCallback);
   }
 
-  if (ClRemoveTraps) {
+  if (LowerAllowCheckPass::IsRequested()) {
     // We can optimize after inliner, and PGO profile matching. The hook below
     // is called at the end `buildFunctionSimplificationPipeline`, which called
     // from `buildInlinerPipeline`, which called after profile matching.
     PB.registerScalarOptimizerLateEPCallback(
         [](FunctionPassManager &FPM, OptimizationLevel Level) {
-          // RemoveTrapsPass expects trap blocks preceded by conditional
-          // branches, which usually is not the case without SimplifyCFG.
-          // TODO: Remove `SimplifyCFGPass` after switching to dedicated
-          // intrinsic.
-          FPM.addPass(SimplifyCFGPass());
-          FPM.addPass(RemoveTrapsPass());
+          FPM.addPass(LowerAllowCheckPass());
         });
   }
 }

diff --git a/clang/lib/CodeGen/CGAtomic.cpp b/clang/lib/CodeGen/CGAtomic.cpp
@@ -197,11 +197,11 @@ namespace {
     llvm::Value *getScalarRValValueOrNull(RValue RVal) const;
 
     /// Converts an rvalue to integer value if needed.
-    llvm::Value *convertRValueToInt(RValue RVal, bool CastFP = true) const;
+    llvm::Value *convertRValueToInt(RValue RVal, bool CmpXchg = false) const;
 
     RValue ConvertToValueOrAtomic(llvm::Value *IntVal, AggValueSlot ResultSlot,
                                   SourceLocation Loc, bool AsValue,
-                                  bool CastFP = true) const;
+                                  bool CmpXchg = false) const;
 
     /// Copy an atomic r-value into atomic-layout memory.
     void emitCopyIntoMemory(RValue rvalue) const;
@@ -264,7 +264,7 @@ namespace {
                                llvm::AtomicOrdering AO, bool IsVolatile);
     /// Emits atomic load as LLVM instruction.
     llvm::Value *EmitAtomicLoadOp(llvm::AtomicOrdering AO, bool IsVolatile,
-                                  bool CastFP = true);
+                                  bool CmpXchg = false);
     /// Emits atomic compare-and-exchange op as a libcall.
     llvm::Value *EmitAtomicCompareExchangeLibcall(
         llvm::Value *ExpectedAddr, llvm::Value *DesiredAddr,
@@ -1401,13 +1401,26 @@ RValue AtomicInfo::convertAtomicTempToRValue(Address addr,
       LVal.getBaseInfo(), TBAAAccessInfo()));
 }
 
+/// Return true if \param ValTy is a type that should be casted to integer
+/// around the atomic memory operation. If \param CmpXchg is true, then the
+/// cast of a floating point type is made as that instruction can not have
+/// floating point operands.  TODO: Allow compare-and-exchange and FP - see
+/// comment in AtomicExpandPass.cpp.
+static bool shouldCastToInt(llvm::Type *ValTy, bool CmpXchg) {
+  if (ValTy->isFloatingPointTy())
+    return ValTy->isX86_FP80Ty() || CmpXchg;
+  return !ValTy->isIntegerTy() && !ValTy->isPointerTy();
+}
+
 RValue AtomicInfo::ConvertToValueOrAtomic(llvm::Value *Val,
                                           AggValueSlot ResultSlot,
                                           SourceLocation Loc, bool AsValue,
-                                          bool CastFP) const {
+                                          bool CmpXchg) const {
   // Try not to in some easy cases.
-  assert((Val->getType()->isIntegerTy() || Val->getType()->isIEEELikeFPTy()) &&
-         "Expected integer or floating point value");
+  assert((Val->getType()->isIntegerTy() || Val->getType()->isPointerTy() ||
+          Val->getType()->isIEEELikeFPTy()) &&
+         "Expected integer, pointer or floating point value when converting "
+         "result.");
   if (getEvaluationKind() == TEK_Scalar &&
       (((!LVal.isBitField() ||
          LVal.getBitFieldInfo().Size == ValueSizeInBits) &&
@@ -1416,13 +1429,12 @@ RValue AtomicInfo::ConvertToValueOrAtomic(llvm::Value *Val,
     auto *ValTy = AsValue
                       ? CGF.ConvertTypeForMem(ValueTy)
                       : getAtomicAddress().getElementType();
-    if (ValTy->isIntegerTy() || (!CastFP && ValTy->isIEEELikeFPTy())) {
+    if (!shouldCastToInt(ValTy, CmpXchg)) {
       assert((!ValTy->isIntegerTy() || Val->getType() == ValTy) &&
              "Different integer types.");
       return RValue::get(CGF.EmitFromMemory(Val, ValueTy));
-    } else if (ValTy->isPointerTy())
-      return RValue::get(CGF.Builder.CreateIntToPtr(Val, ValTy));
-    else if (llvm::CastInst::isBitCastable(Val->getType(), ValTy))
+    }
+    if (llvm::CastInst::isBitCastable(Val->getType(), ValTy))
       return RValue::get(CGF.Builder.CreateBitCast(Val, ValTy));
   }
 
@@ -1459,10 +1471,10 @@ void AtomicInfo::EmitAtomicLoadLibcall(llvm::Value *AddForLoaded,
 }
 
 llvm::Value *AtomicInfo::EmitAtomicLoadOp(llvm::AtomicOrdering AO,
-                                          bool IsVolatile, bool CastFP) {
+                                          bool IsVolatile, bool CmpXchg) {
   // Okay, we're doing this natively.
   Address Addr = getAtomicAddress();
-  if (!(Addr.getElementType()->isIEEELikeFPTy() && !CastFP))
+  if (shouldCastToInt(Addr.getElementType(), CmpXchg))
     Addr = castToAtomicIntPointer(Addr);
   llvm::LoadInst *Load = CGF.Builder.CreateLoad(Addr, "atomic-load");
   Load->setAtomic(AO);
@@ -1523,16 +1535,15 @@ RValue AtomicInfo::EmitAtomicLoad(AggValueSlot ResultSlot, SourceLocation Loc,
   }
 
   // Okay, we're doing this natively.
-  auto *Load = EmitAtomicLoadOp(AO, IsVolatile, /*CastFP=*/false);
+  auto *Load = EmitAtomicLoadOp(AO, IsVolatile);
 
   // If we're ignoring an aggregate return, don't do anything.
   if (getEvaluationKind() == TEK_Aggregate && ResultSlot.isIgnored())
     return RValue::getAggregate(Address::invalid(), false);
 
   // Okay, turn that back into the original value or atomic (for non-simple
   // lvalues) type.
-  return ConvertToValueOrAtomic(Load, ResultSlot, Loc, AsValue,
-                                /*CastFP=*/false);
+  return ConvertToValueOrAtomic(Load, ResultSlot, Loc, AsValue);
 }
 
 /// Emit a load from an l-value of atomic type.  Note that the r-value
@@ -1601,20 +1612,17 @@ llvm::Value *AtomicInfo::getScalarRValValueOrNull(RValue RVal) const {
   return nullptr;
 }
 
-llvm::Value *AtomicInfo::convertRValueToInt(RValue RVal, bool CastFP) const {
+llvm::Value *AtomicInfo::convertRValueToInt(RValue RVal, bool CmpXchg) const {
   // If we've got a scalar value of the right size, try to avoid going
   // through memory. Floats get casted if needed by AtomicExpandPass.
   if (llvm::Value *Value = getScalarRValValueOrNull(RVal)) {
-    if (isa<llvm::IntegerType>(Value->getType()) ||
-        (!CastFP && Value->getType()->isIEEELikeFPTy()))
+    if (!shouldCastToInt(Value->getType(), CmpXchg))
       return CGF.EmitToMemory(Value, ValueTy);
     else {
       llvm::IntegerType *InputIntTy = llvm::IntegerType::get(
           CGF.getLLVMContext(),
           LVal.isSimple() ? getValueSizeInBits() : getAtomicSizeInBits());
-      if (isa<llvm::PointerType>(Value->getType()))
-        return CGF.Builder.CreatePtrToInt(Value, InputIntTy);
-      else if (llvm::BitCastInst::isBitCastable(Value->getType(), InputIntTy))
+      if (llvm::BitCastInst::isBitCastable(Value->getType(), InputIntTy))
         return CGF.Builder.CreateBitCast(Value, InputIntTy);
     }
   }
@@ -1687,13 +1695,14 @@ std::pair<RValue, llvm::Value *> AtomicInfo::EmitAtomicCompareExchange(
 
   // If we've got a scalar value of the right size, try to avoid going
   // through memory.
-  auto *ExpectedVal = convertRValueToInt(Expected);
-  auto *DesiredVal = convertRValueToInt(Desired);
+  auto *ExpectedVal = convertRValueToInt(Expected, /*CmpXchg=*/true);
+  auto *DesiredVal = convertRValueToInt(Desired, /*CmpXchg=*/true);
   auto Res = EmitAtomicCompareExchangeOp(ExpectedVal, DesiredVal, Success,
                                          Failure, IsWeak);
   return std::make_pair(
       ConvertToValueOrAtomic(Res.first, AggValueSlot::ignored(),
-                             SourceLocation(), /*AsValue=*/false),
+                             SourceLocation(), /*AsValue=*/false,
+                             /*CmpXchg=*/true),
       Res.second);
 }
 
@@ -1787,7 +1796,7 @@ void AtomicInfo::EmitAtomicUpdateOp(
   auto Failure = llvm::AtomicCmpXchgInst::getStrongestFailureOrdering(AO);
 
   // Do the atomic load.
-  auto *OldVal = EmitAtomicLoadOp(Failure, IsVolatile);
+  auto *OldVal = EmitAtomicLoadOp(Failure, IsVolatile, /*CmpXchg=*/true);
   // For non-simple lvalues perform compare-and-swap procedure.
   auto *ContBB = CGF.createBasicBlock("atomic_cont");
   auto *ExitBB = CGF.createBasicBlock("atomic_exit");
@@ -1803,7 +1812,8 @@ void AtomicInfo::EmitAtomicUpdateOp(
     CGF.Builder.CreateStore(PHI, NewAtomicIntAddr);
   }
   auto OldRVal = ConvertToValueOrAtomic(PHI, AggValueSlot::ignored(),
-                                        SourceLocation(), /*AsValue=*/false);
+                                        SourceLocation(), /*AsValue=*/false,
+                                        /*CmpXchg=*/true);
   EmitAtomicUpdateValue(CGF, *this, OldRVal, UpdateOp, NewAtomicAddr);
   auto *DesiredVal = CGF.Builder.CreateLoad(NewAtomicIntAddr);
   // Try to write new value using cmpxchg operation.
@@ -1869,7 +1879,7 @@ void AtomicInfo::EmitAtomicUpdateOp(llvm::AtomicOrdering AO, RValue UpdateRVal,
   auto Failure = llvm::AtomicCmpXchgInst::getStrongestFailureOrdering(AO);
 
   // Do the atomic load.
-  auto *OldVal = EmitAtomicLoadOp(Failure, IsVolatile);
+  auto *OldVal = EmitAtomicLoadOp(Failure, IsVolatile, /*CmpXchg=*/true);
   // For non-simple lvalues perform compare-and-swap procedure.
   auto *ContBB = CGF.createBasicBlock("atomic_cont");
   auto *ExitBB = CGF.createBasicBlock("atomic_exit");
@@ -1969,21 +1979,16 @@ void CodeGenFunction::EmitAtomicStore(RValue rvalue, LValue dest,
     }
 
     // Okay, we're doing this natively.
-    llvm::Value *ValToStore =
-        atomics.convertRValueToInt(rvalue, /*CastFP=*/false);
+    llvm::Value *ValToStore = atomics.convertRValueToInt(rvalue);
 
     // Do the atomic store.
     Address Addr = atomics.getAtomicAddress();
-    bool ShouldCastToInt = true;
     if (llvm::Value *Value = atomics.getScalarRValValueOrNull(rvalue))
-      if (isa<llvm::IntegerType>(Value->getType()) ||
-          Value->getType()->isIEEELikeFPTy())
-        ShouldCastToInt = false;
-    if (ShouldCastToInt) {
-      Addr = atomics.castToAtomicIntPointer(Addr);
-      ValToStore = Builder.CreateIntCast(ValToStore, Addr.getElementType(),
-                                         /*isSigned=*/false);
-    }
+      if (shouldCastToInt(Value->getType(), /*CmpXchg=*/false)) {
+        Addr = atomics.castToAtomicIntPointer(Addr);
+        ValToStore = Builder.CreateIntCast(ValToStore, Addr.getElementType(),
+                                           /*isSigned=*/false);
+      }
     llvm::StoreInst *store = Builder.CreateStore(ValToStore, Addr);
 
     if (AO == llvm::AtomicOrdering::Acquire)

diff --git a/clang/lib/CodeGen/CGBlocks.cpp b/clang/lib/CodeGen/CGBlocks.cpp
@@ -962,7 +962,7 @@ llvm::Value *CodeGenFunction::EmitBlockLiteral(const CGBlockInfo &blockInfo) {
       }
 
     // If it's a reference variable, copy the reference into the block field.
-    } else if (auto refType = type->getAs<ReferenceType>()) {
+    } else if (type->getAs<ReferenceType>()) {
       Builder.CreateStore(src.emitRawPointer(*this), blockField);
 
       // If type is const-qualified, copy the value into the block field.

diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -13,6 +13,7 @@
 #include "ABIInfo.h"
 #include "CGCUDARuntime.h"
 #include "CGCXXABI.h"
+#include "CGHLSLRuntime.h"
 #include "CGObjCRuntime.h"
 #include "CGOpenCLRuntime.h"
 #include "CGRecordLayout.h"
@@ -5835,7 +5836,7 @@ RValue CodeGenFunction::EmitBuiltinExpr(const GlobalDecl GD, unsigned BuiltinID,
         EmitLifetimeEnd(TmpSize, TmpPtr);
       return Call;
     }
-    [[fallthrough]];
+    llvm_unreachable("Unexpected enqueue_kernel signature");
   }
   // OpenCL v2.0 s6.13.17.6 - Kernel query functions need bitcast of block
   // parameter.
@@ -7281,8 +7282,6 @@ static const std::pair<unsigned, unsigned> NEONEquivalentIntrinsicMap[] = {
   { NEON::BI__builtin_neon_vabdq_f16, NEON::BI__builtin_neon_vabdq_v, },
   { NEON::BI__builtin_neon_vabs_f16, NEON::BI__builtin_neon_vabs_v, },
   { NEON::BI__builtin_neon_vabsq_f16, NEON::BI__builtin_neon_vabsq_v, },
-  { NEON::BI__builtin_neon_vbsl_f16, NEON::BI__builtin_neon_vbsl_v, },
-  { NEON::BI__builtin_neon_vbslq_f16, NEON::BI__builtin_neon_vbslq_v, },
   { NEON::BI__builtin_neon_vcage_f16, NEON::BI__builtin_neon_vcage_v, },
   { NEON::BI__builtin_neon_vcageq_f16, NEON::BI__builtin_neon_vcageq_v, },
   { NEON::BI__builtin_neon_vcagt_f16, NEON::BI__builtin_neon_vcagt_v, },
@@ -7301,8 +7300,6 @@ static const std::pair<unsigned, unsigned> NEONEquivalentIntrinsicMap[] = {
   { NEON::BI__builtin_neon_vclezq_f16, NEON::BI__builtin_neon_vclezq_v, },
   { NEON::BI__builtin_neon_vcltz_f16, NEON::BI__builtin_neon_vcltz_v, },
   { NEON::BI__builtin_neon_vcltzq_f16, NEON::BI__builtin_neon_vcltzq_v, },
-  { NEON::BI__builtin_neon_vext_f16, NEON::BI__builtin_neon_vext_v, },
-  { NEON::BI__builtin_neon_vextq_f16, NEON::BI__builtin_neon_vextq_v, },
   { NEON::BI__builtin_neon_vfma_f16, NEON::BI__builtin_neon_vfma_v, },
   { NEON::BI__builtin_neon_vfma_lane_f16, NEON::BI__builtin_neon_vfma_lane_v, },
   { NEON::BI__builtin_neon_vfma_laneq_f16, NEON::BI__builtin_neon_vfma_laneq_v, },
@@ -7405,12 +7402,6 @@ static const std::pair<unsigned, unsigned> NEONEquivalentIntrinsicMap[] = {
   { NEON::BI__builtin_neon_vst4_lane_bf16, NEON::BI__builtin_neon_vst4_lane_v },
   { NEON::BI__builtin_neon_vst4q_bf16, NEON::BI__builtin_neon_vst4q_v },
   { NEON::BI__builtin_neon_vst4q_lane_bf16, NEON::BI__builtin_neon_vst4q_lane_v },
-  { NEON::BI__builtin_neon_vtrn_f16, NEON::BI__builtin_neon_vtrn_v, },
-  { NEON::BI__builtin_neon_vtrnq_f16, NEON::BI__builtin_neon_vtrnq_v, },
-  { NEON::BI__builtin_neon_vuzp_f16, NEON::BI__builtin_neon_vuzp_v, },
-  { NEON::BI__builtin_neon_vuzpq_f16, NEON::BI__builtin_neon_vuzpq_v, },
-  { NEON::BI__builtin_neon_vzip_f16, NEON::BI__builtin_neon_vzip_v, },
-  { NEON::BI__builtin_neon_vzipq_f16, NEON::BI__builtin_neon_vzipq_v, },
   // The mangling rules cause us to have one ID for each type for vldap1(q)_lane
   // and vstl1(q)_lane, but codegen is equivalent for all of them. Choose an
   // arbitrary one to be handled as tha canonical variation.
@@ -18182,6 +18173,13 @@ Value *CodeGenFunction::EmitHLSLBuiltinExpr(unsigned BuiltinID,
     return nullptr;
 
   switch (BuiltinID) {
+  case Builtin::BI__builtin_hlsl_elementwise_all: {
+    Value *Op0 = EmitScalarExpr(E->getArg(0));
+    return Builder.CreateIntrinsic(
+        /*ReturnType=*/llvm::Type::getInt1Ty(getLLVMContext()),
+        CGM.getHLSLRuntime().getAllIntrinsic(), ArrayRef<Value *>{Op0}, nullptr,
+        "hlsl.all");
+  }
   case Builtin::BI__builtin_hlsl_elementwise_any: {
     Value *Op0 = EmitScalarExpr(E->getArg(0));
     return Builder.CreateIntrinsic(

diff --git a/clang/lib/CodeGen/CGCall.cpp b/clang/lib/CodeGen/CGCall.cpp
@@ -4379,7 +4379,8 @@ void CodeGenFunction::EmitNonNullArgCheck(RValue RV, QualType ArgType,
     NNAttr = getNonNullAttr(AC.getDecl(), PVD, ArgType, ArgNo);
 
   bool CanCheckNullability = false;
-  if (SanOpts.has(SanitizerKind::NullabilityArg) && !NNAttr && PVD) {
+  if (SanOpts.has(SanitizerKind::NullabilityArg) && !NNAttr && PVD &&
+      !PVD->getType()->isRecordType()) {
     auto Nullability = PVD->getType()->getNullability();
     CanCheckNullability = Nullability &&
                           *Nullability == NullabilityKind::NonNull &&

diff --git a/clang/lib/CodeGen/CGExpr.cpp b/clang/lib/CodeGen/CGExpr.cpp
@@ -56,8 +56,7 @@ using namespace CodeGen;
 // Experiment to make sanitizers easier to debug
 static llvm::cl::opt<bool> ClSanitizeDebugDeoptimization(
     "ubsan-unique-traps", llvm::cl::Optional,
-    llvm::cl::desc("Deoptimize traps for UBSAN so there is 1 trap per check"),
-    llvm::cl::init(false));
+    llvm::cl::desc("Deoptimize traps for UBSAN so there is 1 trap per check"));
 
 //===--------------------------------------------------------------------===//
 //                        Miscellaneous Helper Methods

diff --git a/clang/lib/CodeGen/CGExprComplex.cpp b/clang/lib/CodeGen/CGExprComplex.cpp
@@ -319,12 +319,12 @@ class ComplexExprEmitter
     // doubles the exponent of SmallerType.LargestFiniteVal)
     if (llvm::APFloat::semanticsMaxExponent(ElementTypeSemantics) * 2 + 1 <=
         llvm::APFloat::semanticsMaxExponent(HigherElementTypeSemantics)) {
+      FPHasBeenPromoted = true;
       return CGF.getContext().getComplexType(HigherElementType);
     } else {
-      FPHasBeenPromoted = true;
       DiagnosticsEngine &Diags = CGF.CGM.getDiags();
       Diags.Report(diag::warn_next_larger_fp_type_same_size_than_fp);
-      return CGF.getContext().getComplexType(ElementType);
+      return QualType();
     }
   }
 
@@ -1037,7 +1037,7 @@ ComplexPairTy ComplexExprEmitter::EmitBinDiv(const BinOpInfo &Op) {
       LHSi = llvm::Constant::getNullValue(RHSi->getType());
     if (Op.FPFeatures.getComplexRange() == LangOptions::CX_Improved ||
         (Op.FPFeatures.getComplexRange() == LangOptions::CX_Promoted &&
-         FPHasBeenPromoted))
+         !FPHasBeenPromoted))
       return EmitRangeReductionDiv(LHSr, LHSi, RHSr, RHSi);
     else if (Op.FPFeatures.getComplexRange() == LangOptions::CX_Basic ||
              Op.FPFeatures.getComplexRange() == LangOptions::CX_Promoted)

diff --git a/clang/lib/CodeGen/CGHLSLRuntime.cpp b/clang/lib/CodeGen/CGHLSLRuntime.cpp
@@ -17,8 +17,6 @@
 #include "CodeGenModule.h"
 #include "clang/AST/Decl.h"
 #include "clang/Basic/TargetOptions.h"
-#include "llvm/IR/IntrinsicsDirectX.h"
-#include "llvm/IR/IntrinsicsSPIRV.h"
 #include "llvm/IR/Metadata.h"
 #include "llvm/IR/Module.h"
 #include "llvm/Support/FormatVariadic.h"
@@ -117,6 +115,10 @@ GlobalVariable *replaceBuffer(CGHLSLRuntime::Buffer &Buf) {
 
 } // namespace
 
+llvm::Triple::ArchType CGHLSLRuntime::getArch() {
+  return CGM.getTarget().getTriple().getArch();
+}
+
 void CGHLSLRuntime::addConstant(VarDecl *D, Buffer &CB) {
   if (D->getStorageClass() == SC_Static) {
     // For static inside cbuffer, take as global static.
@@ -343,18 +345,8 @@ llvm::Value *CGHLSLRuntime::emitInputSemantic(IRBuilder<> &B,
     return B.CreateCall(FunctionCallee(DxGroupIndex));
   }
   if (D.hasAttr<HLSLSV_DispatchThreadIDAttr>()) {
-    llvm::Function *ThreadIDIntrinsic;
-    switch (CGM.getTarget().getTriple().getArch()) {
-    case llvm::Triple::dxil:
-      ThreadIDIntrinsic = CGM.getIntrinsic(Intrinsic::dx_thread_id);
-      break;
-    case llvm::Triple::spirv:
-      ThreadIDIntrinsic = CGM.getIntrinsic(Intrinsic::spv_thread_id);
-      break;
-    default:
-      llvm_unreachable("Input semantic not supported by target");
-      break;
-    }
+    llvm::Function *ThreadIDIntrinsic =
+        CGM.getIntrinsic(getThreadIdIntrinsic());
     return buildVectorInput(B, ThreadIDIntrinsic, Ty);
   }
   assert(false && "Unhandled parameter attribute");

diff --git a/clang/lib/CodeGen/CGHLSLRuntime.h b/clang/lib/CodeGen/CGHLSLRuntime.h
@@ -16,7 +16,11 @@
 #define LLVM_CLANG_LIB_CODEGEN_CGHLSLRUNTIME_H
 
 #include "llvm/IR/IRBuilder.h"
+#include "llvm/IR/Intrinsics.h"
+#include "llvm/IR/IntrinsicsDirectX.h"
+#include "llvm/IR/IntrinsicsSPIRV.h"
 
+#include "clang/Basic/Builtins.h"
 #include "clang/Basic/HLSLRuntime.h"
 
 #include "llvm/ADT/SmallVector.h"
@@ -26,6 +30,22 @@
 #include <optional>
 #include <vector>
 
+// A function generator macro for picking the right intrinsic
+// for the target backend
+#define GENERATE_HLSL_INTRINSIC_FUNCTION(FunctionName, IntrinsicPostfix)       \
+  llvm::Intrinsic::ID get##FunctionName##Intrinsic() {                         \
+    llvm::Triple::ArchType Arch = getArch();                                   \
+    switch (Arch) {                                                            \
+    case llvm::Triple::dxil:                                                   \
+      return llvm::Intrinsic::dx_##IntrinsicPostfix;                           \
+    case llvm::Triple::spirv:                                                  \
+      return llvm::Intrinsic::spv_##IntrinsicPostfix;                          \
+    default:                                                                   \
+      llvm_unreachable("Intrinsic " #IntrinsicPostfix                          \
+                       " not supported by target architecture");               \
+    }                                                                          \
+  }
+
 namespace llvm {
 class GlobalVariable;
 class Function;
@@ -48,6 +68,17 @@ class CodeGenModule;
 
 class CGHLSLRuntime {
 public:
+  //===----------------------------------------------------------------------===//
+  // Start of reserved area for HLSL intrinsic getters.
+  //===----------------------------------------------------------------------===//
+
+  GENERATE_HLSL_INTRINSIC_FUNCTION(All, all)
+  GENERATE_HLSL_INTRINSIC_FUNCTION(ThreadId, thread_id)
+
+  //===----------------------------------------------------------------------===//
+  // End of reserved area for HLSL intrinsic getters.
+  //===----------------------------------------------------------------------===//
+
   struct BufferResBinding {
     // The ID like 2 in register(b2, space1).
     std::optional<unsigned> Reg;
@@ -96,6 +127,7 @@ class CGHLSLRuntime {
                                    BufferResBinding &Binding);
   void addConstant(VarDecl *D, Buffer &CB);
   void addBufferDecls(const DeclContext *DC, Buffer &CB);
+  llvm::Triple::ArchType getArch();
   llvm::SmallVector<Buffer> Buffers;
 };
 

diff --git a/clang/lib/CodeGen/CGOpenMPRuntime.cpp b/clang/lib/CodeGen/CGOpenMPRuntime.cpp
@@ -2648,9 +2648,9 @@ void CGOpenMPRuntime::emitDistributeStaticInit(
 void CGOpenMPRuntime::emitForStaticFinish(CodeGenFunction &CGF,
                                           SourceLocation Loc,
                                           OpenMPDirectiveKind DKind) {
-  assert(DKind == OMPD_distribute || DKind == OMPD_for ||
-         DKind == OMPD_sections &&
-             "Expected distribute, for, or sections directive kind");
+  assert((DKind == OMPD_distribute || DKind == OMPD_for ||
+          DKind == OMPD_sections) &&
+         "Expected distribute, for, or sections directive kind");
   if (!CGF.HaveInsertPoint())
     return;
   // Call __kmpc_for_static_fini(ident_t *loc, kmp_int32 tid);

diff --git a/clang/lib/CodeGen/CGRecordLayoutBuilder.cpp b/clang/lib/CodeGen/CGRecordLayoutBuilder.cpp
@@ -185,18 +185,21 @@ struct CGRecordLowering {
   /// Lowers an ASTRecordLayout to a llvm type.
   void lower(bool NonVirtualBaseType);
   void lowerUnion(bool isNoUniqueAddress);
-  void accumulateFields();
+  void accumulateFields(bool isNonVirtualBaseType);
   RecordDecl::field_iterator
-  accumulateBitFields(RecordDecl::field_iterator Field,
+  accumulateBitFields(bool isNonVirtualBaseType,
+                      RecordDecl::field_iterator Field,
                       RecordDecl::field_iterator FieldEnd);
   void computeVolatileBitfields();
   void accumulateBases();
   void accumulateVPtrs();
   void accumulateVBases();
   /// Recursively searches all of the bases to find out if a vbase is
   /// not the primary vbase of some base class.
-  bool hasOwnStorage(const CXXRecordDecl *Decl, const CXXRecordDecl *Query);
+  bool hasOwnStorage(const CXXRecordDecl *Decl,
+                     const CXXRecordDecl *Query) const;
   void calculateZeroInit();
+  CharUnits calculateTailClippingOffset(bool isNonVirtualBaseType) const;
   /// Lowers bitfield storage types to I8 arrays for bitfields with tail
   /// padding that is or can potentially be used.
   void clipTailPadding();
@@ -287,7 +290,7 @@ void CGRecordLowering::lower(bool NVBaseType) {
     computeVolatileBitfields();
     return;
   }
-  accumulateFields();
+  accumulateFields(NVBaseType);
   // RD implies C++.
   if (RD) {
     accumulateVPtrs();
@@ -378,12 +381,12 @@ void CGRecordLowering::lowerUnion(bool isNoUniqueAddress) {
     Packed = true;
 }
 
-void CGRecordLowering::accumulateFields() {
+void CGRecordLowering::accumulateFields(bool isNonVirtualBaseType) {
   for (RecordDecl::field_iterator Field = D->field_begin(),
                                   FieldEnd = D->field_end();
        Field != FieldEnd;) {
     if (Field->isBitField()) {
-      Field = accumulateBitFields(Field, FieldEnd);
+      Field = accumulateBitFields(isNonVirtualBaseType, Field, FieldEnd);
       assert((Field == FieldEnd || !Field->isBitField()) &&
              "Failed to accumulate all the bitfields");
     } else if (Field->isZeroSize(Context)) {
@@ -404,9 +407,12 @@ void CGRecordLowering::accumulateFields() {
 }
 
 // Create members for bitfields. Field is a bitfield, and FieldEnd is the end
-// iterator of the record. Return the first non-bitfield encountered.
+// iterator of the record. Return the first non-bitfield encountered.  We need
+// to know whether this is the base or complete layout, as virtual bases could
+// affect the upper bound of bitfield access unit allocation.
 RecordDecl::field_iterator
-CGRecordLowering::accumulateBitFields(RecordDecl::field_iterator Field,
+CGRecordLowering::accumulateBitFields(bool isNonVirtualBaseType,
+                                      RecordDecl::field_iterator Field,
                                       RecordDecl::field_iterator FieldEnd) {
   if (isDiscreteBitFieldABI()) {
     // Run stores the first element of the current run of bitfields. FieldEnd is
@@ -505,6 +511,10 @@ CGRecordLowering::accumulateBitFields(RecordDecl::field_iterator Field,
       bitsToCharUnits(Context.getTargetInfo().getRegisterWidth());
   unsigned CharBits = Context.getCharWidth();
 
+  // Limit of useable tail padding at end of the record. Computed lazily and
+  // cached here.
+  CharUnits ScissorOffset = CharUnits::Zero();
+
   // Data about the start of the span we're accumulating to create an access
   // unit from. Begin is the first bitfield of the span. If Begin is FieldEnd,
   // we've not got a current span. The span starts at the BeginOffset character
@@ -630,10 +640,14 @@ CGRecordLowering::accumulateBitFields(RecordDecl::field_iterator Field,
               LimitOffset = bitsToCharUnits(getFieldBitOffset(*Probe));
               goto FoundLimit;
             }
-          // We reached the end of the fields.  We can't necessarily use tail
-          // padding in C++ structs, so the NonVirtual size is what we must
-          // use there.
-          LimitOffset = RD ? Layout.getNonVirtualSize() : Layout.getDataSize();
+          // We reached the end of the fields, determine the bounds of useable
+          // tail padding. As this can be complex for C++, we cache the result.
+          if (ScissorOffset.isZero()) {
+            ScissorOffset = calculateTailClippingOffset(isNonVirtualBaseType);
+            assert(!ScissorOffset.isZero() && "Tail clipping at zero");
+          }
+
+          LimitOffset = ScissorOffset;
         FoundLimit:;
 
           CharUnits TypeSize = getSize(Type);
@@ -838,13 +852,17 @@ void CGRecordLowering::accumulateVPtrs() {
                    llvm::PointerType::getUnqual(Types.getLLVMContext())));
 }
 
-void CGRecordLowering::accumulateVBases() {
+CharUnits
+CGRecordLowering::calculateTailClippingOffset(bool isNonVirtualBaseType) const {
+  if (!RD)
+    return Layout.getDataSize();
+
   CharUnits ScissorOffset = Layout.getNonVirtualSize();
   // In the itanium ABI, it's possible to place a vbase at a dsize that is
   // smaller than the nvsize.  Here we check to see if such a base is placed
   // before the nvsize and set the scissor offset to that, instead of the
   // nvsize.
-  if (isOverlappingVBaseABI())
+  if (!isNonVirtualBaseType && isOverlappingVBaseABI())
     for (const auto &Base : RD->vbases()) {
       const CXXRecordDecl *BaseDecl = Base.getType()->getAsCXXRecordDecl();
       if (BaseDecl->isEmpty())
@@ -856,8 +874,13 @@ void CGRecordLowering::accumulateVBases() {
       ScissorOffset = std::min(ScissorOffset,
                                Layout.getVBaseClassOffset(BaseDecl));
     }
-  Members.push_back(MemberInfo(ScissorOffset, MemberInfo::Scissor, nullptr,
-                               RD));
+
+  return ScissorOffset;
+}
+
+void CGRecordLowering::accumulateVBases() {
+  Members.push_back(MemberInfo(calculateTailClippingOffset(false),
+                               MemberInfo::Scissor, nullptr, RD));
   for (const auto &Base : RD->vbases()) {
     const CXXRecordDecl *BaseDecl = Base.getType()->getAsCXXRecordDecl();
     if (BaseDecl->isEmpty())
@@ -882,7 +905,7 @@ void CGRecordLowering::accumulateVBases() {
 }
 
 bool CGRecordLowering::hasOwnStorage(const CXXRecordDecl *Decl,
-                                     const CXXRecordDecl *Query) {
+                                     const CXXRecordDecl *Query) const {
   const ASTRecordLayout &DeclLayout = Context.getASTRecordLayout(Decl);
   if (DeclLayout.isPrimaryBaseVirtual() && DeclLayout.getPrimaryBase() == Query)
     return false;

diff --git a/clang/lib/CodeGen/CodeGenFunction.cpp b/clang/lib/CodeGen/CodeGenFunction.cpp
@@ -990,7 +990,8 @@ void CodeGenFunction::StartFunction(GlobalDecl GD, QualType RetTy,
   // return value. Initialize the flag to 'true' and refine it in EmitParmDecl.
   if (SanOpts.has(SanitizerKind::NullabilityReturn)) {
     auto Nullability = FnRetTy->getNullability();
-    if (Nullability && *Nullability == NullabilityKind::NonNull) {
+    if (Nullability && *Nullability == NullabilityKind::NonNull &&
+        !FnRetTy->isRecordType()) {
       if (!(SanOpts.has(SanitizerKind::ReturnsNonnullAttribute) &&
             CurCodeDecl && CurCodeDecl->getAttr<ReturnsNonNullAttr>()))
         RetValNullabilityPrecondition =

diff --git a/clang/lib/CodeGen/CodeGenTBAA.cpp b/clang/lib/CodeGen/CodeGenTBAA.cpp
@@ -22,6 +22,7 @@
 #include "clang/AST/Mangle.h"
 #include "clang/AST/RecordLayout.h"
 #include "clang/Basic/CodeGenOptions.h"
+#include "clang/Basic/TargetInfo.h"
 #include "llvm/ADT/SmallSet.h"
 #include "llvm/IR/Constants.h"
 #include "llvm/IR/LLVMContext.h"
@@ -319,7 +320,13 @@ CodeGenTBAA::CollectFields(uint64_t BaseOffset,
       // base type.
       if ((*i)->isBitField()) {
         const CGBitFieldInfo &Info = CGRL.getBitFieldInfo(*i);
-        if (Info.Offset != 0)
+        // For big endian targets the first bitfield in the consecutive run is
+        // at the most-significant end; see CGRecordLowering::setBitFieldInfo
+        // for more information.
+        bool IsBE = Context.getTargetInfo().isBigEndian();
+        bool IsFirst = IsBE ? Info.StorageSize - (Info.Offset + Info.Size) == 0
+                            : Info.Offset == 0;
+        if (!IsFirst)
           continue;
         unsigned CurrentBitFieldSize = Info.StorageSize;
         uint64_t Size =

diff --git a/clang/lib/CodeGen/Targets/X86.cpp b/clang/lib/CodeGen/Targets/X86.cpp
@@ -1069,6 +1069,12 @@ Address X86_32ABIInfo::EmitVAArg(CodeGenFunction &CGF,
 
   auto TypeInfo = getContext().getTypeInfoInChars(Ty);
 
+  CCState State(*const_cast<CGFunctionInfo *>(CGF.CurFnInfo));
+  ABIArgInfo AI = classifyArgumentType(Ty, State, /*ArgIndex*/ 0);
+  // Empty records are ignored for parameter passing purposes.
+  if (AI.isIgnore())
+    return CGF.CreateMemTemp(Ty);
+
   // x86-32 changes the alignment of certain arguments on the stack.
   //
   // Just messing with TypeInfo like this works because we never pass

diff --git a/clang/lib/Driver/Driver.cpp b/clang/lib/Driver/Driver.cpp
@@ -49,6 +49,7 @@
 #include "ToolChains/WebAssembly.h"
 #include "ToolChains/XCore.h"
 #include "ToolChains/ZOS.h"
+#include "clang/Basic/DiagnosticDriver.h"
 #include "clang/Basic/TargetID.h"
 #include "clang/Basic/Version.h"
 #include "clang/Config/config.h"
@@ -5814,19 +5815,9 @@ static const char *GetModuleOutputPath(Compilation &C, const JobAction &JA,
          (C.getArgs().hasArg(options::OPT_fmodule_output) ||
           C.getArgs().hasArg(options::OPT_fmodule_output_EQ)));
 
-  if (Arg *ModuleOutputEQ =
-          C.getArgs().getLastArg(options::OPT_fmodule_output_EQ))
-    return C.addResultFile(ModuleOutputEQ->getValue(), &JA);
+  SmallString<256> OutputPath =
+      tools::getCXX20NamedModuleOutputPath(C.getArgs(), BaseInput);
 
-  SmallString<64> OutputPath;
-  Arg *FinalOutput = C.getArgs().getLastArg(options::OPT_o);
-  if (FinalOutput && C.getArgs().hasArg(options::OPT_c))
-    OutputPath = FinalOutput->getValue();
-  else
-    OutputPath = BaseInput;
-
-  const char *Extension = types::getTypeTempSuffix(JA.getType());
-  llvm::sys::path::replace_extension(OutputPath, Extension);
   return C.addResultFile(C.getArgs().MakeArgString(OutputPath.c_str()), &JA);
 }
 
@@ -5899,6 +5890,12 @@ const char *Driver::GetNamedOutputPath(Compilation &C, const JobAction &JA,
         &JA);
   }
 
+  if (JA.getType() == types::TY_API_INFO &&
+      C.getArgs().hasArg(options::OPT_emit_extension_symbol_graphs) &&
+      C.getArgs().hasArg(options::OPT_o))
+    Diag(clang::diag::err_drv_unexpected_symbol_graph_output)
+        << C.getArgs().getLastArgValue(options::OPT_o);
+
   // DXC defaults to standard out when generating assembly. We check this after
   // any DXC flags that might specify a file.
   if (AtTopLevel && JA.getType() == types::TY_PP_Asm && IsDXCMode())

diff --git a/clang/lib/Driver/ToolChains/Clang.cpp b/clang/lib/Driver/ToolChains/Clang.cpp
@@ -3839,6 +3839,24 @@ bool Driver::getDefaultModuleCachePath(SmallVectorImpl<char> &Result) {
   return false;
 }
 
+llvm::SmallString<256>
+clang::driver::tools::getCXX20NamedModuleOutputPath(const ArgList &Args,
+                                                    const char *BaseInput) {
+  if (Arg *ModuleOutputEQ = Args.getLastArg(options::OPT_fmodule_output_EQ))
+    return StringRef(ModuleOutputEQ->getValue());
+
+  SmallString<256> OutputPath;
+  if (Arg *FinalOutput = Args.getLastArg(options::OPT_o);
+      FinalOutput && Args.hasArg(options::OPT_c))
+    OutputPath = FinalOutput->getValue();
+  else
+    OutputPath = BaseInput;
+
+  const char *Extension = types::getTypeTempSuffix(types::TY_ModuleFile);
+  llvm::sys::path::replace_extension(OutputPath, Extension);
+  return OutputPath;
+}
+
 static bool RenderModulesOptions(Compilation &C, const Driver &D,
                                  const ArgList &Args, const InputInfo &Input,
                                  const InputInfo &Output, bool HaveStd20,
@@ -4027,9 +4045,18 @@ static bool RenderModulesOptions(Compilation &C, const Driver &D,
   // module fragment.
   CmdArgs.push_back("-fskip-odr-check-in-gmf");
 
-  // Claim `-fmodule-output` and `-fmodule-output=` to avoid unused warnings.
-  Args.ClaimAllArgs(options::OPT_fmodule_output);
-  Args.ClaimAllArgs(options::OPT_fmodule_output_EQ);
+  // We need to include the case the input file is a module file here.
+  // Since the default compilation model for C++ module interface unit will
+  // create temporary module file and compile the temporary module file
+  // to get the object file. Then the `-fmodule-output` flag will be
+  // brought to the second compilation process. So we have to claim it for
+  // the case too.
+  if (Input.getType() == driver::types::TY_CXXModule ||
+      Input.getType() == driver::types::TY_PP_CXXModule ||
+      Input.getType() == driver::types::TY_ModuleFile) {
+    Args.ClaimAllArgs(options::OPT_fmodule_output);
+    Args.ClaimAllArgs(options::OPT_fmodule_output_EQ);
+  }
 
   return HaveModules;
 }
@@ -5019,11 +5046,26 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
     assert(JA.getType() == types::TY_API_INFO &&
            "Extract API actions must generate a API information.");
     CmdArgs.push_back("-extract-api");
+
+    if (Arg *PrettySGFArg = Args.getLastArg(options::OPT_emit_pretty_sgf))
+      PrettySGFArg->render(Args, CmdArgs);
+
+    Arg *SymbolGraphDirArg = Args.getLastArg(options::OPT_symbol_graph_dir_EQ);
+
     if (Arg *ProductNameArg = Args.getLastArg(options::OPT_product_name_EQ))
       ProductNameArg->render(Args, CmdArgs);
     if (Arg *ExtractAPIIgnoresFileArg =
             Args.getLastArg(options::OPT_extract_api_ignores_EQ))
       ExtractAPIIgnoresFileArg->render(Args, CmdArgs);
+    if (Arg *EmitExtensionSymbolGraphs =
+            Args.getLastArg(options::OPT_emit_extension_symbol_graphs)) {
+      if (!SymbolGraphDirArg)
+        D.Diag(diag::err_drv_missing_symbol_graph_dir);
+
+      EmitExtensionSymbolGraphs->render(Args, CmdArgs);
+    }
+    if (SymbolGraphDirArg)
+      SymbolGraphDirArg->render(Args, CmdArgs);
   } else {
     assert((isa<CompileJobAction>(JA) || isa<BackendJobAction>(JA)) &&
            "Invalid action for clang tool.");
@@ -5840,7 +5882,8 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
       CM = "large";
     if (Triple.isAArch64(64)) {
       Ok = CM == "tiny" || CM == "small" || CM == "large";
-      if (CM == "large" && RelocationModel != llvm::Reloc::Static)
+      if (CM == "large" && !Triple.isOSBinFormatMachO() &&
+          RelocationModel != llvm::Reloc::Static)
         D.Diag(diag::err_drv_argument_only_allowed_with)
             << A->getAsString(Args) << "-fno-pic";
     } else if (Triple.isLoongArch()) {

diff --git a/clang/lib/Driver/ToolChains/Clang.h b/clang/lib/Driver/ToolChains/Clang.h
@@ -193,6 +193,21 @@ DwarfFissionKind getDebugFissionKind(const Driver &D,
                                      const llvm::opt::ArgList &Args,
                                      llvm::opt::Arg *&Arg);
 
+// Calculate the output path of the module file when compiling a module unit
+// with the `-fmodule-output` option or `-fmodule-output=` option specified.
+// The behavior is:
+// - If `-fmodule-output=` is specfied, then the module file is
+//   writing to the value.
+// - Otherwise if the output object file of the module unit is specified, the
+// output path
+//   of the module file should be the same with the output object file except
+//   the corresponding suffix. This requires both `-o` and `-c` are specified.
+// - Otherwise, the output path of the module file will be the same with the
+//   input with the corresponding suffix.
+llvm::SmallString<256>
+getCXX20NamedModuleOutputPath(const llvm::opt::ArgList &Args,
+                              const char *BaseInput);
+
 } // end namespace tools
 
 } // end namespace driver

diff --git a/clang/lib/Driver/ToolChains/CommonArgs.cpp b/clang/lib/Driver/ToolChains/CommonArgs.cpp
@@ -1075,53 +1075,38 @@ void tools::addLTOOptions(const ToolChain &ToolChain, const ArgList &Args,
 
 /// Adds the '-lcgpu' and '-lmgpu' libraries to the compilation to include the
 /// LLVM C library for GPUs.
-static void addOpenMPDeviceLibC(const ToolChain &TC, const ArgList &Args,
+static void addOpenMPDeviceLibC(const Compilation &C, const ArgList &Args,
                                 ArgStringList &CmdArgs) {
   if (Args.hasArg(options::OPT_nogpulib) || Args.hasArg(options::OPT_nolibc))
     return;
 
   // Check the resource directory for the LLVM libc GPU declarations. If it's
   // found we can assume that LLVM was built with support for the GPU libc.
-  SmallString<256> LibCDecls(TC.getDriver().ResourceDir);
+  SmallString<256> LibCDecls(C.getDriver().ResourceDir);
   llvm::sys::path::append(LibCDecls, "include", "llvm_libc_wrappers",
                           "llvm-libc-decls");
   bool HasLibC = llvm::sys::fs::exists(LibCDecls) &&
                  llvm::sys::fs::is_directory(LibCDecls);
   if (!Args.hasFlag(options::OPT_gpulibc, options::OPT_nogpulibc, HasLibC))
     return;
 
-  // We don't have access to the offloading toolchains here, so determine from
-  // the arguments if we have any active NVPTX or AMDGPU toolchains.
-  llvm::DenseSet<const char *> Libraries;
-  if (const Arg *Targets = Args.getLastArg(options::OPT_fopenmp_targets_EQ)) {
-    if (llvm::any_of(Targets->getValues(),
-                     [](auto S) { return llvm::Triple(S).isAMDGPU(); })) {
-      Libraries.insert("-lcgpu-amdgpu");
-      Libraries.insert("-lmgpu-amdgpu");
-    }
-    if (llvm::any_of(Targets->getValues(),
-                     [](auto S) { return llvm::Triple(S).isNVPTX(); })) {
-      Libraries.insert("-lcgpu-nvptx");
-      Libraries.insert("-lmgpu-nvptx");
-    }
-  }
+  SmallVector<const ToolChain *> ToolChains;
+  auto TCRange = C.getOffloadToolChains(Action::OFK_OpenMP);
+  for (auto TI = TCRange.first, TE = TCRange.second; TI != TE; ++TI)
+    ToolChains.push_back(TI->second);
 
-  for (StringRef Arch : Args.getAllArgValues(options::OPT_offload_arch_EQ)) {
-    if (llvm::any_of(llvm::split(Arch, ","), [](StringRef Str) {
-          return IsAMDGpuArch(StringToCudaArch(Str));
-        })) {
-      Libraries.insert("-lcgpu-amdgpu");
-      Libraries.insert("-lmgpu-amdgpu");
-    }
-    if (llvm::any_of(llvm::split(Arch, ","), [](StringRef Str) {
-          return IsNVIDIAGpuArch(StringToCudaArch(Str));
-        })) {
-      Libraries.insert("-lcgpu-nvptx");
-      Libraries.insert("-lmgpu-nvptx");
-    }
+  if (llvm::any_of(ToolChains, [](const ToolChain *TC) {
+        return TC->getTriple().isAMDGPU();
+      })) {
+    CmdArgs.push_back("-lcgpu-amdgpu");
+    CmdArgs.push_back("-lmgpu-amdgpu");
+  }
+  if (llvm::any_of(ToolChains, [](const ToolChain *TC) {
+        return TC->getTriple().isNVPTX();
+      })) {
+    CmdArgs.push_back("-lcgpu-nvptx");
+    CmdArgs.push_back("-lmgpu-nvptx");
   }
-
-  llvm::append_range(CmdArgs, Libraries);
 }
 
 void tools::addOpenMPRuntimeLibraryPath(const ToolChain &TC,
@@ -1153,9 +1138,10 @@ void tools::addArchSpecificRPath(const ToolChain &TC, const ArgList &Args,
   }
 }
 
-bool tools::addOpenMPRuntime(ArgStringList &CmdArgs, const ToolChain &TC,
-                             const ArgList &Args, bool ForceStaticHostRuntime,
-                             bool IsOffloadingHost, bool GompNeedsRT) {
+bool tools::addOpenMPRuntime(const Compilation &C, ArgStringList &CmdArgs,
+                             const ToolChain &TC, const ArgList &Args,
+                             bool ForceStaticHostRuntime, bool IsOffloadingHost,
+                             bool GompNeedsRT) {
   if (!Args.hasFlag(options::OPT_fopenmp, options::OPT_fopenmp_EQ,
                     options::OPT_fno_openmp, false))
     return false;
@@ -1196,7 +1182,7 @@ bool tools::addOpenMPRuntime(ArgStringList &CmdArgs, const ToolChain &TC,
     CmdArgs.push_back("-lomptarget.devicertl");
 
   if (IsOffloadingHost)
-    addOpenMPDeviceLibC(TC, Args, CmdArgs);
+    addOpenMPDeviceLibC(C, Args, CmdArgs);
 
   addArchSpecificRPath(TC, Args, CmdArgs);
   addOpenMPRuntimeLibraryPath(TC, Args, CmdArgs);

diff --git a/clang/lib/Driver/ToolChains/CommonArgs.h b/clang/lib/Driver/ToolChains/CommonArgs.h
@@ -111,8 +111,8 @@ void addOpenMPRuntimeLibraryPath(const ToolChain &TC,
                                  const llvm::opt::ArgList &Args,
                                  llvm::opt::ArgStringList &CmdArgs);
 /// Returns true, if an OpenMP runtime has been added.
-bool addOpenMPRuntime(llvm::opt::ArgStringList &CmdArgs, const ToolChain &TC,
-                      const llvm::opt::ArgList &Args,
+bool addOpenMPRuntime(const Compilation &C, llvm::opt::ArgStringList &CmdArgs,
+                      const ToolChain &TC, const llvm::opt::ArgList &Args,
                       bool ForceStaticHostRuntime = false,
                       bool IsOffloadingHost = false, bool GompNeedsRT = false);
 

diff --git a/clang/lib/Driver/ToolChains/Darwin.cpp b/clang/lib/Driver/ToolChains/Darwin.cpp
@@ -686,7 +686,7 @@ void darwin::Linker::ConstructJob(Compilation &C, const JobAction &JA,
   }
 
   if (!Args.hasArg(options::OPT_nostdlib, options::OPT_nodefaultlibs))
-    addOpenMPRuntime(CmdArgs, getToolChain(), Args);
+    addOpenMPRuntime(C, CmdArgs, getToolChain(), Args);
 
   if (isObjCRuntimeLinked(Args) &&
       !Args.hasArg(options::OPT_nostdlib, options::OPT_nodefaultlibs)) {

diff --git a/clang/lib/Driver/ToolChains/DragonFly.cpp b/clang/lib/Driver/ToolChains/DragonFly.cpp
@@ -136,7 +136,7 @@ void dragonfly::Linker::ConstructJob(Compilation &C, const JobAction &JA,
 
     // Use the static OpenMP runtime with -static-openmp
     bool StaticOpenMP = Args.hasArg(options::OPT_static_openmp) && !Static;
-    addOpenMPRuntime(CmdArgs, ToolChain, Args, StaticOpenMP);
+    addOpenMPRuntime(C, CmdArgs, ToolChain, Args, StaticOpenMP);
 
     if (D.CCCIsCXX()) {
       if (ToolChain.ShouldLinkCXXStdlib(Args))

diff --git a/clang/lib/Driver/ToolChains/Flang.cpp b/clang/lib/Driver/ToolChains/Flang.cpp
@@ -35,27 +35,18 @@ static void addDashXForInput(const ArgList &Args, const InputInfo &Input,
 
 void Flang::addFortranDialectOptions(const ArgList &Args,
                                      ArgStringList &CmdArgs) const {
-  Args.addAllArgs(CmdArgs, {options::OPT_ffixed_form,
-                            options::OPT_ffree_form,
-                            options::OPT_ffixed_line_length_EQ,
-                            options::OPT_fopenmp,
-                            options::OPT_fopenmp_version_EQ,
-                            options::OPT_fopenacc,
-                            options::OPT_finput_charset_EQ,
-                            options::OPT_fimplicit_none,
-                            options::OPT_fno_implicit_none,
-                            options::OPT_fbackslash,
-                            options::OPT_fno_backslash,
-                            options::OPT_flogical_abbreviations,
-                            options::OPT_fno_logical_abbreviations,
-                            options::OPT_fxor_operator,
-                            options::OPT_fno_xor_operator,
-                            options::OPT_falternative_parameter_statement,
-                            options::OPT_fdefault_real_8,
-                            options::OPT_fdefault_integer_8,
-                            options::OPT_fdefault_double_8,
-                            options::OPT_flarge_sizes,
-                            options::OPT_fno_automatic});
+  Args.addAllArgs(
+      CmdArgs, {options::OPT_ffixed_form, options::OPT_ffree_form,
+                options::OPT_ffixed_line_length_EQ, options::OPT_fopenacc,
+                options::OPT_finput_charset_EQ, options::OPT_fimplicit_none,
+                options::OPT_fno_implicit_none, options::OPT_fbackslash,
+                options::OPT_fno_backslash, options::OPT_flogical_abbreviations,
+                options::OPT_fno_logical_abbreviations,
+                options::OPT_fxor_operator, options::OPT_fno_xor_operator,
+                options::OPT_falternative_parameter_statement,
+                options::OPT_fdefault_real_8, options::OPT_fdefault_integer_8,
+                options::OPT_fdefault_double_8, options::OPT_flarge_sizes,
+                options::OPT_fno_automatic});
 }
 
 void Flang::addPreprocessingOptions(const ArgList &Args,
@@ -763,6 +754,35 @@ void Flang::ConstructJob(Compilation &C, const JobAction &JA,
   // Add other compile options
   addOtherOptions(Args, CmdArgs);
 
+  // Forward flags for OpenMP. We don't do this if the current action is an
+  // device offloading action other than OpenMP.
+  if (Args.hasFlag(options::OPT_fopenmp, options::OPT_fopenmp_EQ,
+                   options::OPT_fno_openmp, false) &&
+      (JA.isDeviceOffloading(Action::OFK_None) ||
+       JA.isDeviceOffloading(Action::OFK_OpenMP))) {
+    switch (D.getOpenMPRuntime(Args)) {
+    case Driver::OMPRT_OMP:
+    case Driver::OMPRT_IOMP5:
+      // Clang can generate useful OpenMP code for these two runtime libraries.
+      CmdArgs.push_back("-fopenmp");
+      Args.AddAllArgs(CmdArgs, options::OPT_fopenmp_version_EQ);
+
+      // FIXME: Clang supports a whole bunch more flags here.
+      break;
+    default:
+      // By default, if Clang doesn't know how to generate useful OpenMP code
+      // for a specific runtime library, we just don't pass the '-fopenmp' flag
+      // down to the actual compilation.
+      // FIXME: It would be better to have a mode which *only* omits IR
+      // generation based on the OpenMP support so that we get consistent
+      // semantic analysis, etc.
+      const Arg *A = Args.getLastArg(options::OPT_fopenmp_EQ);
+      D.Diag(diag::warn_drv_unsupported_openmp_library)
+          << A->getSpelling() << A->getValue();
+      break;
+    }
+  }
+
   // Offloading related options
   addOffloadOptions(C, Inputs, JA, Args, CmdArgs);
 

diff --git a/clang/lib/Driver/ToolChains/FreeBSD.cpp b/clang/lib/Driver/ToolChains/FreeBSD.cpp
@@ -295,7 +295,7 @@ void freebsd::Linker::ConstructJob(Compilation &C, const JobAction &JA,
     // Use the static OpenMP runtime with -static-openmp
     bool StaticOpenMP = Args.hasArg(options::OPT_static_openmp) &&
                         !Args.hasArg(options::OPT_static);
-    addOpenMPRuntime(CmdArgs, ToolChain, Args, StaticOpenMP);
+    addOpenMPRuntime(C, CmdArgs, ToolChain, Args, StaticOpenMP);
 
     if (D.CCCIsCXX()) {
       if (ToolChain.ShouldLinkCXXStdlib(Args))

diff --git a/clang/lib/Driver/ToolChains/Gnu.cpp b/clang/lib/Driver/ToolChains/Gnu.cpp
@@ -598,7 +598,7 @@ void tools::gnutools::Linker::ConstructJob(Compilation &C, const JobAction &JA,
 
       // FIXME: Only pass GompNeedsRT = true for platforms with libgomp that
       // require librt. Most modern Linux platforms do, but some may not.
-      if (addOpenMPRuntime(CmdArgs, ToolChain, Args, StaticOpenMP,
+      if (addOpenMPRuntime(C, CmdArgs, ToolChain, Args, StaticOpenMP,
                            JA.isHostOffloading(Action::OFK_OpenMP),
                            /* GompNeedsRT= */ true))
         // OpenMP runtimes implies pthreads when using the GNU toolchain.

diff --git a/clang/lib/Driver/ToolChains/Haiku.cpp b/clang/lib/Driver/ToolChains/Haiku.cpp
@@ -107,7 +107,7 @@ void haiku::Linker::ConstructJob(Compilation &C, const JobAction &JA,
                    options::OPT_r)) {
     // Use the static OpenMP runtime with -static-openmp
     bool StaticOpenMP = Args.hasArg(options::OPT_static_openmp) && !Static;
-    addOpenMPRuntime(CmdArgs, ToolChain, Args, StaticOpenMP);
+    addOpenMPRuntime(C, CmdArgs, ToolChain, Args, StaticOpenMP);
 
     if (D.CCCIsCXX() && ToolChain.ShouldLinkCXXStdlib(Args))
       ToolChain.AddCXXStdlibLibArgs(Args, CmdArgs);

diff --git a/clang/lib/Driver/ToolChains/NetBSD.cpp b/clang/lib/Driver/ToolChains/NetBSD.cpp
@@ -311,7 +311,7 @@ void netbsd::Linker::ConstructJob(Compilation &C, const JobAction &JA,
                    options::OPT_r)) {
     // Use the static OpenMP runtime with -static-openmp
     bool StaticOpenMP = Args.hasArg(options::OPT_static_openmp) && !Static;
-    addOpenMPRuntime(CmdArgs, ToolChain, Args, StaticOpenMP);
+    addOpenMPRuntime(C, CmdArgs, ToolChain, Args, StaticOpenMP);
 
     if (D.CCCIsCXX()) {
       if (ToolChain.ShouldLinkCXXStdlib(Args))

diff --git a/clang/lib/Driver/ToolChains/OpenBSD.cpp b/clang/lib/Driver/ToolChains/OpenBSD.cpp
@@ -221,7 +221,7 @@ void openbsd::Linker::ConstructJob(Compilation &C, const JobAction &JA,
                    options::OPT_r)) {
     // Use the static OpenMP runtime with -static-openmp
     bool StaticOpenMP = Args.hasArg(options::OPT_static_openmp) && !Static;
-    addOpenMPRuntime(CmdArgs, ToolChain, Args, StaticOpenMP);
+    addOpenMPRuntime(C, CmdArgs, ToolChain, Args, StaticOpenMP);
 
     if (D.CCCIsCXX()) {
       if (ToolChain.ShouldLinkCXXStdlib(Args))

diff --git a/clang/lib/Driver/ToolChains/Solaris.cpp b/clang/lib/Driver/ToolChains/Solaris.cpp
@@ -211,7 +211,7 @@ void solaris::Linker::ConstructJob(Compilation &C, const JobAction &JA,
     // Use the static OpenMP runtime with -static-openmp
     bool StaticOpenMP = Args.hasArg(options::OPT_static_openmp) &&
                         !Args.hasArg(options::OPT_static);
-    addOpenMPRuntime(CmdArgs, ToolChain, Args, StaticOpenMP);
+    addOpenMPRuntime(C, CmdArgs, ToolChain, Args, StaticOpenMP);
 
     if (D.CCCIsCXX()) {
       if (ToolChain.ShouldLinkCXXStdlib(Args))

diff --git a/clang/lib/ExtractAPI/API.cpp b/clang/lib/ExtractAPI/API.cpp
diff --git a/clang/lib/ExtractAPI/DeclarationFragments.cpp b/clang/lib/ExtractAPI/DeclarationFragments.cpp
@@ -14,14 +14,11 @@
 #include "clang/ExtractAPI/DeclarationFragments.h"
 #include "clang/AST/Decl.h"
 #include "clang/AST/DeclCXX.h"
-#include "clang/AST/QualTypeNames.h"
 #include "clang/AST/Type.h"
 #include "clang/AST/TypeLoc.h"
-#include "clang/Basic/OperatorKinds.h"
 #include "clang/ExtractAPI/TypedefUnderlyingTypeResolver.h"
 #include "clang/Index/USRGeneration.h"
 #include "llvm/ADT/StringSwitch.h"
-#include <typeinfo>
 
 using namespace clang::extractapi;
 using namespace llvm;
@@ -60,23 +57,44 @@ void findTypeLocForBlockDecl(const clang::TypeSourceInfo *TSInfo,
 
 } // namespace
 
-DeclarationFragments &DeclarationFragments::appendSpace() {
+DeclarationFragments &
+DeclarationFragments::appendUnduplicatedTextCharacter(char Character) {
   if (!Fragments.empty()) {
     Fragment &Last = Fragments.back();
     if (Last.Kind == FragmentKind::Text) {
       // Merge the extra space into the last fragment if the last fragment is
       // also text.
-      if (Last.Spelling.back() != ' ') { // avoid extra trailing spaces.
-        Last.Spelling.push_back(' ');
+      if (Last.Spelling.back() != Character) { // avoid duplicates at end
+        Last.Spelling.push_back(Character);
       }
     } else {
-      append(" ", FragmentKind::Text);
+      append("", FragmentKind::Text);
+      Fragments.back().Spelling.push_back(Character);
     }
   }
 
   return *this;
 }
 
+DeclarationFragments &DeclarationFragments::appendSpace() {
+  return appendUnduplicatedTextCharacter(' ');
+}
+
+DeclarationFragments &DeclarationFragments::appendSemicolon() {
+  return appendUnduplicatedTextCharacter(';');
+}
+
+DeclarationFragments &DeclarationFragments::removeTrailingSemicolon() {
+  if (Fragments.empty())
+    return *this;
+
+  Fragment &Last = Fragments.back();
+  if (Last.Kind == FragmentKind::Text && Last.Spelling.back() == ';')
+    Last.Spelling.pop_back();
+
+  return *this;
+}
+
 StringRef DeclarationFragments::getFragmentKindString(
     DeclarationFragments::FragmentKind Kind) {
   switch (Kind) {
@@ -469,7 +487,7 @@ DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForNamespace(
   if (!Decl->isAnonymousNamespace())
     Fragments.appendSpace().append(
         Decl->getName(), DeclarationFragments::FragmentKind::Identifier);
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 DeclarationFragments
@@ -511,7 +529,7 @@ DeclarationFragmentsBuilder::getFragmentsForVar(const VarDecl *Var) {
   return Fragments
       .append(Var->getName(), DeclarationFragments::FragmentKind::Identifier)
       .append(std::move(After))
-      .append(";", DeclarationFragments::FragmentKind::Text);
+      .appendSemicolon();
 }
 
 DeclarationFragments
@@ -535,15 +553,13 @@ DeclarationFragmentsBuilder::getFragmentsForVarTemplate(const VarDecl *Var) {
       getFragmentsForType(T, Var->getASTContext(), After);
   if (StringRef(ArgumentFragment.begin()->Spelling)
           .starts_with("type-parameter")) {
-    std::string ProperArgName = getNameForTemplateArgument(
-        Var->getDescribedVarTemplate()->getTemplateParameters()->asArray(),
-        ArgumentFragment.begin()->Spelling);
+    std::string ProperArgName = T.getAsString();
     ArgumentFragment.begin()->Spelling.swap(ProperArgName);
   }
   Fragments.append(std::move(ArgumentFragment))
       .appendSpace()
       .append(Var->getName(), DeclarationFragments::FragmentKind::Identifier)
-      .append(";", DeclarationFragments::FragmentKind::Text);
+      .appendSemicolon();
   return Fragments;
 }
 
@@ -570,12 +586,7 @@ DeclarationFragmentsBuilder::getFragmentsForParam(const ParmVarDecl *Param) {
 
   if (StringRef(TypeFragments.begin()->Spelling)
           .starts_with("type-parameter")) {
-    std::string ProperArgName = getNameForTemplateArgument(
-        dyn_cast<FunctionDecl>(Param->getDeclContext())
-            ->getDescribedFunctionTemplate()
-            ->getTemplateParameters()
-            ->asArray(),
-        TypeFragments.begin()->Spelling);
+    std::string ProperArgName = Param->getOriginalType().getAsString();
     TypeFragments.begin()->Spelling.swap(ProperArgName);
   }
 
@@ -668,11 +679,7 @@ DeclarationFragmentsBuilder::getFragmentsForFunction(const FunctionDecl *Func) {
       getFragmentsForType(Func->getReturnType(), Func->getASTContext(), After);
   if (StringRef(ReturnValueFragment.begin()->Spelling)
           .starts_with("type-parameter")) {
-    std::string ProperArgName =
-        getNameForTemplateArgument(Func->getDescribedFunctionTemplate()
-                                       ->getTemplateParameters()
-                                       ->asArray(),
-                                   ReturnValueFragment.begin()->Spelling);
+    std::string ProperArgName = Func->getReturnType().getAsString();
     ReturnValueFragment.begin()->Spelling.swap(ProperArgName);
   }
 
@@ -712,7 +719,7 @@ DeclarationFragmentsBuilder::getFragmentsForFunction(const FunctionDecl *Func) {
   Fragments.append(DeclarationFragments::getExceptionSpecificationString(
       Func->getExceptionSpecType()));
 
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForEnumConstant(
@@ -741,7 +748,7 @@ DeclarationFragmentsBuilder::getFragmentsForEnum(const EnumDecl *EnumDecl) {
             getFragmentsForType(IntegerType, EnumDecl->getASTContext(), After))
         .append(std::move(After));
 
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 DeclarationFragments
@@ -757,7 +764,7 @@ DeclarationFragmentsBuilder::getFragmentsForField(const FieldDecl *Field) {
       .appendSpace()
       .append(Field->getName(), DeclarationFragments::FragmentKind::Identifier)
       .append(std::move(After))
-      .append(";", DeclarationFragments::FragmentKind::Text);
+      .appendSemicolon();
 }
 
 DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForRecordDecl(
@@ -775,7 +782,7 @@ DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForRecordDecl(
     Fragments.appendSpace().append(
         Record->getName(), DeclarationFragments::FragmentKind::Identifier);
 
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForCXXClass(
@@ -790,7 +797,7 @@ DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForCXXClass(
     Fragments.appendSpace().append(
         Record->getName(), DeclarationFragments::FragmentKind::Identifier);
 
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 DeclarationFragments
@@ -820,7 +827,7 @@ DeclarationFragmentsBuilder::getFragmentsForSpecialCXXMethod(
   Fragments.append(DeclarationFragments::getExceptionSpecificationString(
       Method->getExceptionSpecType()));
 
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForCXXMethod(
@@ -860,7 +867,7 @@ DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForCXXMethod(
   Fragments.append(DeclarationFragments::getExceptionSpecificationString(
       Method->getExceptionSpecType()));
 
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 DeclarationFragments
@@ -891,7 +898,7 @@ DeclarationFragmentsBuilder::getFragmentsForConversionFunction(
     Fragments.appendSpace().append("const",
                                    DeclarationFragments::FragmentKind::Keyword);
 
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 DeclarationFragments
@@ -923,7 +930,7 @@ DeclarationFragmentsBuilder::getFragmentsForOverloadedOperator(
   Fragments.append(DeclarationFragments::getExceptionSpecificationString(
       Method->getExceptionSpecType()));
 
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 // Get fragments for template parameters, e.g. T in tempalte<typename T> ...
@@ -961,25 +968,6 @@ DeclarationFragmentsBuilder::getFragmentsForTemplateParameters(
   return Fragments;
 }
 
-// Find the name of a template argument from the template's parameters.
-std::string DeclarationFragmentsBuilder::getNameForTemplateArgument(
-    const ArrayRef<NamedDecl *> TemplateParameters, std::string TypeParameter) {
-  // The arg is a generic parameter from a partial spec, e.g.
-  // T in template<typename T> Foo<T, int>.
-  //
-  // Those names appear as "type-parameter-<index>-<depth>", so we must find its
-  // name from the template's parameter list.
-  for (unsigned i = 0; i < TemplateParameters.size(); ++i) {
-    const auto *Parameter =
-        dyn_cast<TemplateTypeParmDecl>(TemplateParameters[i]);
-    if (TypeParameter.compare("type-parameter-" +
-                              std::to_string(Parameter->getDepth()) + "-" +
-                              std::to_string(Parameter->getIndex())) == 0)
-      return std::string(TemplateParameters[i]->getName());
-  }
-  llvm_unreachable("Could not find the name of a template argument.");
-}
-
 // Get fragments for template arguments, e.g. int in template<typename T>
 // Foo<int>;
 //
@@ -989,7 +977,7 @@ std::string DeclarationFragmentsBuilder::getNameForTemplateArgument(
 DeclarationFragments
 DeclarationFragmentsBuilder::getFragmentsForTemplateArguments(
     const ArrayRef<TemplateArgument> TemplateArguments, ASTContext &Context,
-    const std::optional<ArrayRef<NamedDecl *>> TemplateParameters) {
+    const std::optional<ArrayRef<TemplateArgumentLoc>> TemplateArgumentLocs) {
   DeclarationFragments Fragments;
   for (unsigned i = 0, end = TemplateArguments.size(); i != end; ++i) {
     if (i)
@@ -1003,8 +991,10 @@ DeclarationFragmentsBuilder::getFragmentsForTemplateArguments(
 
     if (StringRef(ArgumentFragment.begin()->Spelling)
             .starts_with("type-parameter")) {
-      std::string ProperArgName = getNameForTemplateArgument(
-          TemplateParameters.value(), ArgumentFragment.begin()->Spelling);
+      std::string ProperArgName = TemplateArgumentLocs.value()[i]
+                                      .getTypeSourceInfo()
+                                      ->getType()
+                                      .getAsString();
       ArgumentFragment.begin()->Spelling.swap(ProperArgName);
     }
     Fragments.append(std::move(ArgumentFragment));
@@ -1028,7 +1018,7 @@ DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForConcept(
       .appendSpace()
       .append(Concept->getName().str(),
               DeclarationFragments::FragmentKind::Identifier)
-      .append(";", DeclarationFragments::FragmentKind::Text);
+      .appendSemicolon();
 }
 
 DeclarationFragments
@@ -1069,7 +1059,7 @@ DeclarationFragmentsBuilder::getFragmentsForClassTemplateSpecialization(
           getFragmentsForTemplateArguments(Decl->getTemplateArgs().asArray(),
                                            Decl->getASTContext(), std::nullopt))
       .append(">", DeclarationFragments::FragmentKind::Text)
-      .append(";", DeclarationFragments::FragmentKind::Text);
+      .appendSemicolon();
 }
 
 DeclarationFragments
@@ -1089,9 +1079,9 @@ DeclarationFragmentsBuilder::getFragmentsForClassTemplatePartialSpecialization(
       .append("<", DeclarationFragments::FragmentKind::Text)
       .append(getFragmentsForTemplateArguments(
           Decl->getTemplateArgs().asArray(), Decl->getASTContext(),
-          Decl->getTemplateParameters()->asArray()))
+          Decl->getTemplateArgsAsWritten()->arguments()))
       .append(">", DeclarationFragments::FragmentKind::Text)
-      .append(";", DeclarationFragments::FragmentKind::Text);
+      .appendSemicolon();
 }
 
 DeclarationFragments
@@ -1110,7 +1100,7 @@ DeclarationFragmentsBuilder::getFragmentsForVarTemplateSpecialization(
           getFragmentsForTemplateArguments(Decl->getTemplateArgs().asArray(),
                                            Decl->getASTContext(), std::nullopt))
       .append(">", DeclarationFragments::FragmentKind::Text)
-      .append(";", DeclarationFragments::FragmentKind::Text);
+      .appendSemicolon();
 }
 
 DeclarationFragments
@@ -1130,9 +1120,9 @@ DeclarationFragmentsBuilder::getFragmentsForVarTemplatePartialSpecialization(
       .append("<", DeclarationFragments::FragmentKind::Text)
       .append(getFragmentsForTemplateArguments(
           Decl->getTemplateArgs().asArray(), Decl->getASTContext(),
-          Decl->getTemplateParameters()->asArray()))
+          Decl->getTemplateArgsAsWritten()->arguments()))
       .append(">", DeclarationFragments::FragmentKind::Text)
-      .append(";", DeclarationFragments::FragmentKind::Text);
+      .appendSemicolon();
 }
 
 DeclarationFragments
@@ -1203,7 +1193,7 @@ DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForObjCCategory(
 
   Fragments.append("@interface", DeclarationFragments::FragmentKind::Keyword)
       .appendSpace()
-      .append(Category->getClassInterface()->getName(),
+      .append(Interface->getName(),
               DeclarationFragments::FragmentKind::TypeIdentifier, InterfaceUSR,
               Interface)
       .append(" (", DeclarationFragments::FragmentKind::Text)
@@ -1277,7 +1267,7 @@ DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForObjCMethod(
     Fragments.append(getFragmentsForParam(Param));
   }
 
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForObjCProperty(
@@ -1378,7 +1368,7 @@ DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForObjCProperty(
       .append(Property->getName(),
               DeclarationFragments::FragmentKind::Identifier)
       .append(std::move(After))
-      .append(";", DeclarationFragments::FragmentKind::Text);
+      .appendSemicolon();
 }
 
 DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForObjCProtocol(
@@ -1422,7 +1412,7 @@ DeclarationFragments DeclarationFragmentsBuilder::getFragmentsForTypedef(
       .appendSpace()
       .append(Decl->getName(), DeclarationFragments::FragmentKind::Identifier);
 
-  return Fragments.append(";", DeclarationFragments::FragmentKind::Text);
+  return Fragments.appendSemicolon();
 }
 
 // Instantiate template for FunctionDecl.

diff --git a/clang/lib/ExtractAPI/ExtractAPIConsumer.cpp b/clang/lib/ExtractAPI/ExtractAPIConsumer.cpp
@@ -30,6 +30,7 @@
 #include "clang/Frontend/CompilerInstance.h"
 #include "clang/Frontend/FrontendOptions.h"
 #include "clang/Frontend/MultiplexConsumer.h"
+#include "clang/Index/USRGeneration.h"
 #include "clang/InstallAPI/HeaderFile.h"
 #include "clang/Lex/MacroInfo.h"
 #include "clang/Lex/PPCallbacks.h"
@@ -39,6 +40,7 @@
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/SmallVector.h"
+#include "llvm/ADT/StringRef.h"
 #include "llvm/Support/Casting.h"
 #include "llvm/Support/Error.h"
 #include "llvm/Support/FileSystem.h"
@@ -327,11 +329,12 @@ class MacroCallback : public PPCallbacks {
 
       StringRef Name = PM.MacroNameToken.getIdentifierInfo()->getName();
       PresumedLoc Loc = SM.getPresumedLoc(PM.MacroNameToken.getLocation());
-      StringRef USR =
-          API.recordUSRForMacro(Name, PM.MacroNameToken.getLocation(), SM);
+      SmallString<128> USR;
+      index::generateUSRForMacro(Name, PM.MacroNameToken.getLocation(), SM,
+                                 USR);
 
-      API.addMacroDefinition(
-          Name, USR, Loc,
+      API.createRecord<extractapi::MacroDefinitionRecord>(
+          USR, Name, SymbolReference(), Loc,
           DeclarationFragmentsBuilder::getFragmentsForMacro(Name, PM.MD),
           DeclarationFragmentsBuilder::getSubHeadingForMacro(Name),
           SM.isInSystemHeader(PM.MacroNameToken.getLocation()));
@@ -372,40 +375,57 @@ class APIMacroCallback : public MacroCallback {
   LocationFileChecker &LCF;
 };
 
+std::unique_ptr<llvm::raw_pwrite_stream>
+createAdditionalSymbolGraphFile(CompilerInstance &CI, Twine BaseName) {
+  auto OutputDirectory = CI.getFrontendOpts().SymbolGraphOutputDir;
+
+  SmallString<256> FileName;
+  llvm::sys::path::append(FileName, OutputDirectory,
+                          BaseName + ".symbols.json");
+  return CI.createOutputFile(
+      FileName, /*Binary*/ false, /*RemoveFileOnSignal*/ false,
+      /*UseTemporary*/ true, /*CreateMissingDirectories*/ true);
+}
+
 } // namespace
 
-void ExtractAPIActionBase::ImplEndSourceFileAction() {
-  if (!OS)
-    return;
+void ExtractAPIActionBase::ImplEndSourceFileAction(CompilerInstance &CI) {
+  SymbolGraphSerializerOption SerializationOptions;
+  SerializationOptions.Compact = !CI.getFrontendOpts().EmitPrettySymbolGraphs;
+  SerializationOptions.EmitSymbolLabelsForTesting =
+      CI.getFrontendOpts().EmitSymbolGraphSymbolLabelsForTesting;
+
+  if (CI.getFrontendOpts().EmitExtensionSymbolGraphs) {
+    auto ConstructOutputFile = [&CI](Twine BaseName) {
+      return createAdditionalSymbolGraphFile(CI, BaseName);
+    };
+
+    SymbolGraphSerializer::serializeWithExtensionGraphs(
+        *OS, *API, IgnoresList, ConstructOutputFile, SerializationOptions);
+  } else {
+    SymbolGraphSerializer::serializeMainSymbolGraph(*OS, *API, IgnoresList,
+                                                    SerializationOptions);
+  }
 
-  // Setup a SymbolGraphSerializer to write out collected API information in
-  // the Symbol Graph format.
-  // FIXME: Make the kind of APISerializer configurable.
-  SymbolGraphSerializer SGSerializer(*API, IgnoresList);
-  SGSerializer.serialize(*OS);
+  // Flush the stream and close the main output stream.
   OS.reset();
 }
 
-std::unique_ptr<raw_pwrite_stream>
-ExtractAPIAction::CreateOutputFile(CompilerInstance &CI, StringRef InFile) {
-  std::unique_ptr<raw_pwrite_stream> OS;
-  OS = CI.createDefaultOutputFile(/*Binary=*/false, InFile,
-                                  /*Extension=*/"json",
-                                  /*RemoveFileOnSignal=*/false);
-  if (!OS)
-    return nullptr;
-  return OS;
-}
-
 std::unique_ptr<ASTConsumer>
 ExtractAPIAction::CreateASTConsumer(CompilerInstance &CI, StringRef InFile) {
-  OS = CreateOutputFile(CI, InFile);
+  auto ProductName = CI.getFrontendOpts().ProductName;
+
+  if (CI.getFrontendOpts().SymbolGraphOutputDir.empty())
+    OS = CI.createDefaultOutputFile(/*Binary*/ false, InFile,
+                                    /*Extension*/ "symbols.json",
+                                    /*RemoveFileOnSignal*/ false,
+                                    /*CreateMissingDirectories*/ true);
+  else
+    OS = createAdditionalSymbolGraphFile(CI, ProductName);
 
   if (!OS)
     return nullptr;
 
-  auto ProductName = CI.getFrontendOpts().ProductName;
-
   // Now that we have enough information about the language options and the
   // target triple, let's create the APISet before anyone uses it.
   API = std::make_unique<APISet>(
@@ -495,7 +515,9 @@ bool ExtractAPIAction::PrepareToExecuteAction(CompilerInstance &CI) {
   return true;
 }
 
-void ExtractAPIAction::EndSourceFileAction() { ImplEndSourceFileAction(); }
+void ExtractAPIAction::EndSourceFileAction() {
+  ImplEndSourceFileAction(getCompilerInstance());
+}
 
 std::unique_ptr<ASTConsumer>
 WrappingExtractAPIAction::CreateASTConsumer(CompilerInstance &CI,
@@ -506,11 +528,9 @@ WrappingExtractAPIAction::CreateASTConsumer(CompilerInstance &CI,
 
   CreatedASTConsumer = true;
 
-  OS = CreateOutputFile(CI, InFile);
-  if (!OS)
-    return nullptr;
-
-  auto ProductName = CI.getFrontendOpts().ProductName;
+  ProductName = CI.getFrontendOpts().ProductName;
+  auto InputFilename = llvm::sys::path::filename(InFile);
+  OS = createAdditionalSymbolGraphFile(CI, InputFilename);
 
   // Now that we have enough information about the language options and the
   // target triple, let's create the APISet before anyone uses it.
@@ -552,32 +572,6 @@ void WrappingExtractAPIAction::EndSourceFileAction() {
   WrapperFrontendAction::EndSourceFileAction();
 
   if (CreatedASTConsumer) {
-    ImplEndSourceFileAction();
+    ImplEndSourceFileAction(getCompilerInstance());
   }
 }
-
-std::unique_ptr<raw_pwrite_stream>
-WrappingExtractAPIAction::CreateOutputFile(CompilerInstance &CI,
-                                           StringRef InFile) {
-  std::unique_ptr<raw_pwrite_stream> OS;
-  std::string OutputDir = CI.getFrontendOpts().SymbolGraphOutputDir;
-
-  // The symbol graphs need to be generated as a side effect of regular
-  // compilation so the output should be dumped in the directory provided with
-  // the command line option.
-  llvm::SmallString<128> OutFilePath(OutputDir);
-  auto Seperator = llvm::sys::path::get_separator();
-  auto Infilename = llvm::sys::path::filename(InFile);
-  OutFilePath.append({Seperator, Infilename});
-  llvm::sys::path::replace_extension(OutFilePath, "json");
-  // StringRef outputFilePathref = *OutFilePath;
-
-  // don't use the default output file
-  OS = CI.createOutputFile(/*OutputPath=*/OutFilePath, /*Binary=*/false,
-                           /*RemoveFileOnSignal=*/true,
-                           /*UseTemporary=*/true,
-                           /*CreateMissingDirectories=*/true);
-  if (!OS)
-    return nullptr;
-  return OS;
-}
diff --git a/clang/lib/ExtractAPI/Serialization/SymbolGraphSerializer.cpp b/clang/lib/ExtractAPI/Serialization/SymbolGraphSerializer.cpp
diff --git a/clang/lib/ExtractAPI/TypedefUnderlyingTypeResolver.cpp b/clang/lib/ExtractAPI/TypedefUnderlyingTypeResolver.cpp
@@ -12,6 +12,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "clang/ExtractAPI/TypedefUnderlyingTypeResolver.h"
+#include "clang/Basic/Module.h"
 #include "clang/Index/USRGeneration.h"
 
 using namespace clang;
@@ -50,17 +51,20 @@ TypedefUnderlyingTypeResolver::getSymbolReferenceForType(QualType Type,
   SmallString<128> TypeUSR;
   const NamedDecl *TypeDecl = getUnderlyingTypeDecl(Type);
   const TypedefType *TypedefTy = Type->getAs<TypedefType>();
+  StringRef OwningModuleName;
 
   if (TypeDecl) {
     if (!TypedefTy)
       TypeName = TypeDecl->getName().str();
 
     clang::index::generateUSRForDecl(TypeDecl, TypeUSR);
+    if (auto *OwningModule = TypeDecl->getImportedOwningModule())
+      OwningModuleName = OwningModule->Name;
   } else {
     clang::index::generateUSRForType(Type, Context, TypeUSR);
   }
 
-  return {API.copyString(TypeName), API.copyString(TypeUSR)};
+  return API.createSymbolReference(TypeName, TypeUSR, OwningModuleName);
 }
 
 std::string TypedefUnderlyingTypeResolver::getUSRForType(QualType Type) const {

diff --git a/clang/lib/Format/Format.cpp b/clang/lib/Format/Format.cpp
@@ -955,6 +955,8 @@ template <> struct MappingTraits<FormatStyle> {
                    Style.BreakBeforeTernaryOperators);
     IO.mapOptional("BreakConstructorInitializers",
                    Style.BreakConstructorInitializers);
+    IO.mapOptional("BreakFunctionDefinitionParameters",
+                   Style.BreakFunctionDefinitionParameters);
     IO.mapOptional("BreakInheritanceList", Style.BreakInheritanceList);
     IO.mapOptional("BreakStringLiterals", Style.BreakStringLiterals);
     IO.mapOptional("BreakTemplateDeclarations",
@@ -1465,6 +1467,7 @@ FormatStyle getLLVMStyle(FormatStyle::LanguageKind Language) {
   LLVMStyle.BreakBeforeInlineASMColon = FormatStyle::BBIAS_OnlyMultiline;
   LLVMStyle.BreakBeforeTernaryOperators = true;
   LLVMStyle.BreakConstructorInitializers = FormatStyle::BCIS_BeforeColon;
+  LLVMStyle.BreakFunctionDefinitionParameters = false;
   LLVMStyle.BreakInheritanceList = FormatStyle::BILS_BeforeColon;
   LLVMStyle.BreakStringLiterals = true;
   LLVMStyle.BreakTemplateDeclarations = FormatStyle::BTDS_MultiLine;
@@ -3578,7 +3581,7 @@ cleanupAroundReplacements(StringRef Code, const tooling::Replacements &Replaces,
   // We need to use lambda function here since there are two versions of
   // `cleanup`.
   auto Cleanup = [](const FormatStyle &Style, StringRef Code,
-                    std::vector<tooling::Range> Ranges,
+                    ArrayRef<tooling::Range> Ranges,
                     StringRef FileName) -> tooling::Replacements {
     return cleanup(Style, Code, Ranges, FileName);
   };

diff --git a/clang/lib/Format/FormatToken.h b/clang/lib/Format/FormatToken.h
@@ -574,6 +574,9 @@ struct FormatToken {
   /// Is optional and can be removed.
   bool Optional = false;
 
+  /// Might be function declaration open/closing paren.
+  bool MightBeFunctionDeclParen = false;
+
   /// Number of optional braces to be inserted after this token:
   ///   -1: a single left brace
   ///    0: no braces

diff --git a/clang/lib/Format/TokenAnnotator.cpp b/clang/lib/Format/TokenAnnotator.cpp
@@ -1550,6 +1550,7 @@ class AnnotatingParser {
             (!Previous->isAttribute() &&
              !Previous->isOneOf(TT_RequiresClause, TT_LeadingJavaAnnotation))) {
           Line.MightBeFunctionDecl = true;
+          Tok->MightBeFunctionDeclParen = true;
         }
       }
       break;
@@ -3888,6 +3889,8 @@ void TokenAnnotator::calculateFormattingInformation(AnnotatedLine &Line) const {
       }
     } else if (ClosingParen) {
       for (auto *Tok = ClosingParen->Next; Tok; Tok = Tok->Next) {
+        if (Tok->is(TT_CtorInitializerColon))
+          break;
         if (Tok->is(tok::arrow)) {
           Tok->setType(TT_TrailingReturnArrow);
           break;
@@ -5392,6 +5395,12 @@ bool TokenAnnotator::mustBreakBefore(const AnnotatedLine &Line,
   if (Right.NewlinesBefore > 1 && Style.MaxEmptyLinesToKeep > 0)
     return true;
 
+  if (Style.BreakFunctionDefinitionParameters && Line.MightBeFunctionDecl &&
+      Line.mightBeFunctionDefinition() && Left.MightBeFunctionDeclParen &&
+      Left.ParameterCount > 0) {
+    return true;
+  }
+
   if (Style.isCSharp()) {
     if (Left.is(TT_FatArrow) && Right.is(tok::l_brace) &&
         Style.BraceWrapping.AfterFunction) {

diff --git a/clang/lib/Format/UnwrappedLineParser.cpp b/clang/lib/Format/UnwrappedLineParser.cpp
@@ -495,20 +495,22 @@ void UnwrappedLineParser::calculateBraceTypes(bool ExpectClassBody) {
   };
   SmallVector<StackEntry, 8> LBraceStack;
   assert(Tok->is(tok::l_brace));
+
   do {
-    // Get next non-comment, non-preprocessor token.
     FormatToken *NextTok;
     do {
       NextTok = Tokens->getNextToken();
     } while (NextTok->is(tok::comment));
-    if (!Style.isTableGen()) {
-      // InTableGen, '#' is like binary operator. Not a preprocessor directive.
-      while (NextTok->is(tok::hash) && !Line->InMacroBody) {
-        NextTok = Tokens->getNextToken();
+
+    if (!Line->InMacroBody && !Style.isTableGen()) {
+      // Skip PPDirective lines and comments.
+      while (NextTok->is(tok::hash)) {
         do {
           NextTok = Tokens->getNextToken();
-        } while (NextTok->is(tok::comment) ||
-                 (NextTok->NewlinesBefore == 0 && NextTok->isNot(tok::eof)));
+        } while (NextTok->NewlinesBefore == 0 && NextTok->isNot(tok::eof));
+
+        while (NextTok->is(tok::comment))
+          NextTok = Tokens->getNextToken();
       }
     }
 
@@ -640,6 +642,7 @@ void UnwrappedLineParser::calculateBraceTypes(bool ExpectClassBody) {
     default:
       break;
     }
+
     PrevTok = Tok;
     Tok = NextTok;
   } while (Tok->isNot(tok::eof) && !LBraceStack.empty());

diff --git a/clang/lib/Frontend/CompilerInvocation.cpp b/clang/lib/Frontend/CompilerInvocation.cpp
@@ -533,10 +533,10 @@ static T extractMaskValue(T KeyPath) {
 
 #define PARSE_OPTION_WITH_MARSHALLING(                                         \
     ARGS, DIAGS, PREFIX_TYPE, SPELLING, ID, KIND, GROUP, ALIAS, ALIASARGS,     \
-    FLAGS, VISIBILITY, PARAM, HELPTEXT, METAVAR, VALUES, SHOULD_PARSE,         \
-    ALWAYS_EMIT, KEYPATH, DEFAULT_VALUE, IMPLIED_CHECK, IMPLIED_VALUE,         \
-    NORMALIZER, DENORMALIZER, MERGER, EXTRACTOR, TABLE_INDEX)                  \
-  if ((VISIBILITY)&options::CC1Option) {                                       \
+    FLAGS, VISIBILITY, PARAM, HELPTEXT, HELPTEXTSFORVARIANTS, METAVAR, VALUES, \
+    SHOULD_PARSE, ALWAYS_EMIT, KEYPATH, DEFAULT_VALUE, IMPLIED_CHECK,          \
+    IMPLIED_VALUE, NORMALIZER, DENORMALIZER, MERGER, EXTRACTOR, TABLE_INDEX)   \
+  if ((VISIBILITY) & options::CC1Option) {                                     \
     KEYPATH = MERGER(KEYPATH, DEFAULT_VALUE);                                  \
     if (IMPLIED_CHECK)                                                         \
       KEYPATH = MERGER(KEYPATH, IMPLIED_VALUE);                                \
@@ -550,10 +550,10 @@ static T extractMaskValue(T KeyPath) {
 // with lifetime extension of the reference.
 #define GENERATE_OPTION_WITH_MARSHALLING(                                      \
     CONSUMER, PREFIX_TYPE, SPELLING, ID, KIND, GROUP, ALIAS, ALIASARGS, FLAGS, \
-    VISIBILITY, PARAM, HELPTEXT, METAVAR, VALUES, SHOULD_PARSE, ALWAYS_EMIT,   \
-    KEYPATH, DEFAULT_VALUE, IMPLIED_CHECK, IMPLIED_VALUE, NORMALIZER,          \
-    DENORMALIZER, MERGER, EXTRACTOR, TABLE_INDEX)                              \
-  if ((VISIBILITY)&options::CC1Option) {                                       \
+    VISIBILITY, PARAM, HELPTEXT, HELPTEXTSFORVARIANTS, METAVAR, VALUES,        \
+    SHOULD_PARSE, ALWAYS_EMIT, KEYPATH, DEFAULT_VALUE, IMPLIED_CHECK,          \
+    IMPLIED_VALUE, NORMALIZER, DENORMALIZER, MERGER, EXTRACTOR, TABLE_INDEX)   \
+  if ((VISIBILITY) & options::CC1Option) {                                     \
     [&](const auto &Extracted) {                                               \
       if (ALWAYS_EMIT ||                                                       \
           (Extracted !=                                                        \
@@ -3516,7 +3516,8 @@ void CompilerInvocationBase::GenerateLangArgs(const LangOptions &Opts,
     GenerateArg(Consumer, OPT_fblocks);
 
   if (Opts.ConvergentFunctions &&
-      !(Opts.OpenCL || (Opts.CUDA && Opts.CUDAIsDevice) || Opts.SYCLIsDevice))
+      !(Opts.OpenCL || (Opts.CUDA && Opts.CUDAIsDevice) || Opts.SYCLIsDevice ||
+        Opts.HLSL))
     GenerateArg(Consumer, OPT_fconvergent_functions);
 
   if (Opts.NoBuiltin && !Opts.Freestanding)
@@ -3914,7 +3915,7 @@ bool CompilerInvocation::ParseLangArgs(LangOptions &Opts, ArgList &Args,
 
   Opts.ConvergentFunctions = Args.hasArg(OPT_fconvergent_functions) ||
                              Opts.OpenCL || (Opts.CUDA && Opts.CUDAIsDevice) ||
-                             Opts.SYCLIsDevice;
+                             Opts.SYCLIsDevice || Opts.HLSL;
 
   Opts.NoBuiltin = Args.hasArg(OPT_fno_builtin) || Opts.Freestanding;
   if (!Opts.NoBuiltin)

diff --git a/clang/lib/Frontend/FrontendActions.cpp b/clang/lib/Frontend/FrontendActions.cpp
@@ -453,6 +453,8 @@ class DefaultTemplateInstCallback : public TemplateInstantiationCallback {
       return "BuildingBuiltinDumpStructCall";
     case CodeSynthesisContext::BuildingDeductionGuides:
       return "BuildingDeductionGuides";
+    case Sema::CodeSynthesisContext::TypeAliasTemplateInstantiation:
+      return "TypeAliasTemplateInstantiation";
     }
     return "";
   }

diff --git a/clang/lib/Frontend/PrecompiledPreamble.cpp b/clang/lib/Frontend/PrecompiledPreamble.cpp
@@ -290,8 +290,7 @@ class PrecompilePreambleAction : public ASTFrontendAction {
 
 class PrecompilePreambleConsumer : public PCHGenerator {
 public:
-  PrecompilePreambleConsumer(PrecompilePreambleAction &Action,
-                             const Preprocessor &PP,
+  PrecompilePreambleConsumer(PrecompilePreambleAction &Action, Preprocessor &PP,
                              InMemoryModuleCache &ModuleCache,
                              StringRef isysroot,
                              std::shared_ptr<PCHBuffer> Buffer)

diff --git a/clang/lib/FrontendTool/ExecuteCompilerInvocation.cpp b/clang/lib/FrontendTool/ExecuteCompilerInvocation.cpp
@@ -181,9 +181,13 @@ CreateFrontendAction(CompilerInstance &CI) {
 #endif
 
   // Wrap the base FE action in an extract api action to generate
-  // symbol graph as a biproduct of compilation ( enabled with
-  // --emit-symbol-graph option )
-  if (!FEOpts.SymbolGraphOutputDir.empty()) {
+  // symbol graph as a biproduct of compilation (enabled with
+  // --emit-symbol-graph option)
+  if (FEOpts.EmitSymbolGraph) {
+    if (FEOpts.SymbolGraphOutputDir.empty()) {
+      CI.getDiagnostics().Report(diag::warn_missing_symbol_graph_dir);
+      CI.getFrontendOpts().SymbolGraphOutputDir = ".";
+    }
     CI.getCodeGenOpts().ClearASTBeforeBackend = false;
     Act = std::make_unique<WrappingExtractAPIAction>(std::move(Act));
   }

diff --git a/clang/lib/Headers/__stddef_unreachable.h b/clang/lib/Headers/__stddef_unreachable.h
@@ -7,6 +7,8 @@
  *===-----------------------------------------------------------------------===
  */
 
+#ifndef __cplusplus
+
 /*
  * When -fbuiltin-headers-in-system-modules is set this is a non-modular header
  * and needs to behave as if it was textual.
@@ -15,3 +17,5 @@
     (__has_feature(modules) && !__building_module(_Builtin_stddef))
 #define unreachable() __builtin_unreachable()
 #endif
+
+#endif
diff --git a/clang/lib/Headers/hlsl/hlsl_intrinsics.h b/clang/lib/Headers/hlsl/hlsl_intrinsics.h
@@ -100,6 +100,118 @@ double3 abs(double3);
 _HLSL_BUILTIN_ALIAS(__builtin_elementwise_abs)
 double4 abs(double4);
 
+//===----------------------------------------------------------------------===//
+// all builtins
+//===----------------------------------------------------------------------===//
+
+/// \fn bool all(T x)
+/// \brief Returns True if all components of the \a x parameter are non-zero;
+/// otherwise, false. \param x The input value.
+
+#ifdef __HLSL_ENABLE_16_BIT
+_HLSL_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int16_t);
+_HLSL_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int16_t2);
+_HLSL_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int16_t3);
+_HLSL_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int16_t4);
+_HLSL_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint16_t);
+_HLSL_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint16_t2);
+_HLSL_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint16_t3);
+_HLSL_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint16_t4);
+#endif
+
+_HLSL_16BIT_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(half);
+_HLSL_16BIT_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(half2);
+_HLSL_16BIT_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(half3);
+_HLSL_16BIT_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(half4);
+
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(bool);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(bool2);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(bool3);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(bool4);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int2);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int3);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int4);
+
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint2);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint3);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint4);
+
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(float);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(float2);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(float3);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(float4);
+
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int64_t);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int64_t2);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int64_t3);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(int64_t4);
+
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint64_t);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint64_t2);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint64_t3);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(uint64_t4);
+
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(double);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(double2);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(double3);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_all)
+bool all(double4);
+
 //===----------------------------------------------------------------------===//
 // any builtins
 //===----------------------------------------------------------------------===//

diff --git a/clang/lib/InstallAPI/Visitor.cpp b/clang/lib/InstallAPI/Visitor.cpp
@@ -205,10 +205,10 @@ bool InstallAPIVisitor::VisitObjCCategoryDecl(const ObjCCategoryDecl *D) {
   const ObjCInterfaceDecl *InterfaceD = D->getClassInterface();
   const StringRef InterfaceName = InterfaceD->getName();
 
-  std::pair<ObjCCategoryRecord *, FrontendAttrs *> Category =
-      Ctx.Slice->addObjCCategory(InterfaceName, CategoryName, Avail, D,
-                                 *Access);
-  recordObjCInstanceVariables(D->getASTContext(), Category.first, InterfaceName,
+  ObjCCategoryRecord *CategoryRecord =
+      Ctx.Slice->addObjCCategory(InterfaceName, CategoryName, Avail, D, *Access)
+          .first;
+  recordObjCInstanceVariables(D->getASTContext(), CategoryRecord, InterfaceName,
                               D->ivars());
   return true;
 }

diff --git a/clang/lib/Parse/ParseDeclCXX.cpp b/clang/lib/Parse/ParseDeclCXX.cpp
@@ -1502,6 +1502,15 @@ void Parser::ParseMicrosoftInheritanceClassAttributes(ParsedAttributes &attrs) {
   }
 }
 
+void Parser::ParseNullabilityClassAttributes(ParsedAttributes &attrs) {
+  while (Tok.is(tok::kw__Nullable)) {
+    IdentifierInfo *AttrName = Tok.getIdentifierInfo();
+    auto Kind = Tok.getKind();
+    SourceLocation AttrNameLoc = ConsumeToken();
+    attrs.addNew(AttrName, AttrNameLoc, nullptr, AttrNameLoc, nullptr, 0, Kind);
+  }
+}
+
 /// Determine whether the following tokens are valid after a type-specifier
 /// which could be a standalone declaration. This will conservatively return
 /// true if there's any doubt, and is appropriate for insert-';' fixits.
@@ -1683,15 +1692,21 @@ void Parser::ParseClassSpecifier(tok::TokenKind TagTokKind,
 
   ParsedAttributes attrs(AttrFactory);
   // If attributes exist after tag, parse them.
-  MaybeParseAttributes(PAKM_CXX11 | PAKM_Declspec | PAKM_GNU, attrs);
-
-  // Parse inheritance specifiers.
-  if (Tok.isOneOf(tok::kw___single_inheritance, tok::kw___multiple_inheritance,
-                  tok::kw___virtual_inheritance))
-    ParseMicrosoftInheritanceClassAttributes(attrs);
-
-  // Allow attributes to precede or succeed the inheritance specifiers.
-  MaybeParseAttributes(PAKM_CXX11 | PAKM_Declspec | PAKM_GNU, attrs);
+  for (;;) {
+    MaybeParseAttributes(PAKM_CXX11 | PAKM_Declspec | PAKM_GNU, attrs);
+    // Parse inheritance specifiers.
+    if (Tok.isOneOf(tok::kw___single_inheritance,
+                    tok::kw___multiple_inheritance,
+                    tok::kw___virtual_inheritance)) {
+      ParseMicrosoftInheritanceClassAttributes(attrs);
+      continue;
+    }
+    if (Tok.is(tok::kw__Nullable)) {
+      ParseNullabilityClassAttributes(attrs);
+      continue;
+    }
+    break;
+  }
 
   // Source location used by FIXIT to insert misplaced
   // C++11 attributes

diff --git a/clang/lib/Parse/ParseOpenACC.cpp b/clang/lib/Parse/ParseOpenACC.cpp
@@ -14,6 +14,7 @@
 #include "clang/Parse/ParseDiagnostic.h"
 #include "clang/Parse/Parser.h"
 #include "clang/Parse/RAIIObjectsForParser.h"
+#include "clang/Sema/SemaOpenACC.h"
 #include "llvm/ADT/StringRef.h"
 #include "llvm/ADT/StringSwitch.h"
 
@@ -777,7 +778,7 @@ bool Parser::ParseOpenACCClause(OpenACCDirectiveKind DirKind) {
   SourceLocation ClauseLoc = ConsumeToken();
 
   bool Result = ParseOpenACCClauseParams(DirKind, Kind);
-  getActions().ActOnOpenACCClause(Kind, ClauseLoc);
+  getActions().OpenACC().ActOnClause(Kind, ClauseLoc);
   return Result;
 }
 
@@ -1151,7 +1152,7 @@ Parser::OpenACCDirectiveParseInfo Parser::ParseOpenACCDirective() {
   SourceLocation StartLoc = getCurToken().getLocation();
   OpenACCDirectiveKind DirKind = ParseOpenACCDirectiveKind(*this);
 
-  getActions().ActOnOpenACCConstruct(DirKind, StartLoc);
+  getActions().OpenACC().ActOnConstruct(DirKind, StartLoc);
 
   // Once we've parsed the construct/directive name, some have additional
   // specifiers that need to be taken care of. Atomic has an 'atomic-clause'
@@ -1223,12 +1224,12 @@ Parser::DeclGroupPtrTy Parser::ParseOpenACCDirectiveDecl() {
 
   OpenACCDirectiveParseInfo DirInfo = ParseOpenACCDirective();
 
-  if (getActions().ActOnStartOpenACCDeclDirective(DirInfo.DirKind,
-                                                  DirInfo.StartLoc))
+  if (getActions().OpenACC().ActOnStartDeclDirective(DirInfo.DirKind,
+                                                     DirInfo.StartLoc))
     return nullptr;
 
   // TODO OpenACC: Do whatever decl parsing is required here.
-  return DeclGroupPtrTy::make(getActions().ActOnEndOpenACCDeclDirective());
+  return DeclGroupPtrTy::make(getActions().OpenACC().ActOnEndDeclDirective());
 }
 
 // Parse OpenACC Directive on a Statement.
@@ -1239,8 +1240,8 @@ StmtResult Parser::ParseOpenACCDirectiveStmt() {
   ConsumeAnnotationToken();
 
   OpenACCDirectiveParseInfo DirInfo = ParseOpenACCDirective();
-  if (getActions().ActOnStartOpenACCStmtDirective(DirInfo.DirKind,
-                                                  DirInfo.StartLoc))
+  if (getActions().OpenACC().ActOnStartStmtDirective(DirInfo.DirKind,
+                                                     DirInfo.StartLoc))
     return StmtError();
 
   StmtResult AssocStmt;
@@ -1249,10 +1250,10 @@ StmtResult Parser::ParseOpenACCDirectiveStmt() {
     ParsingOpenACCDirectiveRAII DirScope(*this, /*Value=*/false);
     ParseScope ACCScope(this, getOpenACCScopeFlags(DirInfo.DirKind));
 
-    AssocStmt = getActions().ActOnOpenACCAssociatedStmt(DirInfo.DirKind,
-                                                        ParseStatement());
+    AssocStmt = getActions().OpenACC().ActOnAssociatedStmt(DirInfo.DirKind,
+                                                           ParseStatement());
   }
 
-  return getActions().ActOnEndOpenACCStmtDirective(
+  return getActions().OpenACC().ActOnEndStmtDirective(
       DirInfo.DirKind, DirInfo.StartLoc, DirInfo.EndLoc, AssocStmt);
 }
diff --git a/clang/lib/Sema/JumpDiagnostics.cpp b/clang/lib/Sema/JumpDiagnostics.cpp
@@ -16,6 +16,7 @@
 #include "clang/AST/ExprCXX.h"
 #include "clang/AST/StmtCXX.h"
 #include "clang/AST/StmtObjC.h"
+#include "clang/AST/StmtOpenACC.h"
 #include "clang/AST/StmtOpenMP.h"
 #include "clang/Basic/SourceLocation.h"
 #include "clang/Sema/SemaInternal.h"