Merge branch 'main' into x86-cmp-bigint-nosink

Author: RKSimon

.ci/generate_test_report_github.py

@@ -8,6 +8,7 @@
 
 import generate_test_report_lib
 
+
 def compute_platform_title() -> str:
     logo = ":window:" if platform.system() == "Windows" else ":penguin:"
     # On Linux the machine value is x86_64 on Windows it is AMD64.

.ci/generate_test_report_lib.py

@@ -100,6 +100,7 @@ def _format_ninja_failures(ninja_failures: list[tuple[str, str]]) -> list[str]:
         )
     return output
 
+
 def get_failures(junit_objects) -> dict[str, list[tuple[str, str]]]:
     failures = {}
     for results in junit_objects:

bolt/test/runtime/AArch64/inline-memcpy.s

@@ -81,14 +81,14 @@
 # CHECK-ASM: bl{{.*}}<memcpy
 
 # Register move should NOT be inlined (size unknown at compile time)
-# CHECK-ASM-LABEL: <test_register_move_negative>:
+# CHECK-ASM-LABEL: <test_register_move_unknown>:
 # CHECK-ASM: bl{{.*}}<memcpy
 
-# CHECK-ASM-LABEL: <test_x2_rewrite_negative>:
+# CHECK-ASM-LABEL: <test_x2_rewrite_unknown>:
 # CHECK-ASM: bl{{.*}}<memcpy
 
 # Live-in parameter should NOT be inlined (size unknown at compile time)
-# CHECK-ASM-LABEL: <test_live_in_negative>:
+# CHECK-ASM-LABEL: <test_live_in_unknown>:
 # CHECK-ASM: bl{{.*}}<memcpy
 
 # _memcpy8 should be inlined with end-pointer return (dest+size)
@@ -262,9 +262,9 @@ test_4_byte_add_immediate:
 	ret
 	.size	test_4_byte_add_immediate, .-test_4_byte_add_immediate
 
-	.globl	test_register_move_negative
-	.type	test_register_move_negative,@function
-test_register_move_negative:
+	.globl	test_register_move_unknown
+	.type	test_register_move_unknown,@function
+test_register_move_unknown:
 	stp	x29, x30, [sp, #-32]!
 	mov	x29, sp
 	add	x1, sp, #16
@@ -274,20 +274,20 @@ test_register_move_negative:
 	bl	memcpy
 	ldp	x29, x30, [sp], #32
 	ret
-	.size	test_register_move_negative, .-test_register_move_negative
+	.size	test_register_move_unknown, .-test_register_move_unknown
 
-	.globl  test_x2_rewrite_negative
-	.type   test_x2_rewrite_negative,@function
-test_x2_rewrite_negative:
+	.globl  test_x2_rewrite_unknown
+	.type   test_x2_rewrite_unknown,@function
+test_x2_rewrite_unknown:
 	mov     x2, #8
 	ldr     x2, [sp, #24]
 	bl      memcpy
 	ret
-	.size   test_x2_rewrite_negative, .-test_x2_rewrite_negative
+	.size   test_x2_rewrite_unknown, .-test_x2_rewrite_unknown
 
-	.globl	test_live_in_negative
-	.type	test_live_in_negative,@function
-test_live_in_negative:
+	.globl	test_live_in_unknown
+	.type	test_live_in_unknown,@function
+test_live_in_unknown:
 	# x2 comes in as parameter, no instruction sets it (should NOT inline)
 	stp	x29, x30, [sp, #-32]!
 	mov	x29, sp
@@ -297,7 +297,7 @@ test_live_in_negative:
 	bl	memcpy
 	ldp	x29, x30, [sp], #32
 	ret
-	.size	test_live_in_negative, .-test_live_in_negative
+	.size	test_live_in_unknown, .-test_live_in_unknown
 
 	.globl	test_memcpy8_4_byte
 	.type	test_memcpy8_4_byte,@function

clang/include/clang/Basic/AttrDocs.td

@@ -3450,9 +3450,9 @@ Mac, and BSD. This attribute has no effect on other targets.
 def MSABIDocs : Documentation {
   let Category = DocCatCallingConvs;
   let Content = [{
-On non-Windows x86_64 targets, this attribute changes the calling convention of
-a function to match the default convention used on Windows x86_64. This
-attribute has no effect on Windows targets or non-x86_64 targets.
+On non-Windows x86_64 and aarch64 targets, this attribute changes the calling convention of
+a function to match the default convention used on Windows. This
+attribute has no effect on Windows targets or non-x86_64, non-aarch64 targets.
   }];
 }
 

clang/include/clang/Basic/BuiltinsAMDGPU.def

@@ -180,7 +180,7 @@ BUILTIN(__builtin_amdgcn_raw_buffer_load_b128, "V4UiQbiiIi", "n")
 BUILTIN(__builtin_amdgcn_raw_ptr_buffer_atomic_add_i32, "iiQbiiIi", "")
 
 TARGET_BUILTIN(__builtin_amdgcn_raw_ptr_buffer_atomic_fadd_f32, "ffQbiiIi", "", "atomic-fadd-rtn-insts")
-TARGET_BUILTIN(__builtin_amdgcn_raw_ptr_buffer_atomic_fadd_v2f16, "V2hV2hQbiiIi", "t", "atomic-buffer-global-pk-add-f16-insts")
+TARGET_BUILTIN(__builtin_amdgcn_raw_ptr_buffer_atomic_fadd_v2f16, "V2hV2hQbiiIi", "", "atomic-buffer-global-pk-add-f16-insts")
 
 TARGET_BUILTIN(__builtin_amdgcn_raw_ptr_buffer_atomic_fmin_f32, "ffQbiiIi", "", "atomic-fmin-fmax-global-f32")
 TARGET_BUILTIN(__builtin_amdgcn_raw_ptr_buffer_atomic_fmax_f32, "ffQbiiIi", "", "atomic-fmin-fmax-global-f32")

clang/include/clang/Sema/Sema.h

@@ -11668,7 +11668,7 @@ class Sema final : public SemaBase {
       ASTTemplateArgsPtr TemplateArgsIn, SourceLocation RAngleLoc);
 
   DeclResult ActOnVarTemplateSpecialization(
-      Scope *S, Declarator &D, TypeSourceInfo *DI, LookupResult &Previous,
+      Scope *S, Declarator &D, TypeSourceInfo *TSI, LookupResult &Previous,
       SourceLocation TemplateKWLoc, TemplateParameterList *TemplateParams,
       StorageClass SC, bool IsPartialSpecialization);
 

clang/lib/AST/ASTContext.cpp

@@ -3108,9 +3108,9 @@ TypeSourceInfo *ASTContext::CreateTypeSourceInfo(QualType T,
 
 TypeSourceInfo *ASTContext::getTrivialTypeSourceInfo(QualType T,
                                                      SourceLocation L) const {
-  TypeSourceInfo *DI = CreateTypeSourceInfo(T);
-  DI->getTypeLoc().initialize(const_cast<ASTContext &>(*this), L);
-  return DI;
+  TypeSourceInfo *TSI = CreateTypeSourceInfo(T);
+  TSI->getTypeLoc().initialize(const_cast<ASTContext &>(*this), L);
+  return TSI;
 }
 
 const ASTRecordLayout &
@@ -5891,11 +5891,11 @@ TypeSourceInfo *ASTContext::getTemplateSpecializationTypeInfo(
   QualType TST = getTemplateSpecializationType(
       Keyword, Name, SpecifiedArgs.arguments(), CanonicalArgs, Underlying);
 
-  TypeSourceInfo *DI = CreateTypeSourceInfo(TST);
-  DI->getTypeLoc().castAs<TemplateSpecializationTypeLoc>().set(
+  TypeSourceInfo *TSI = CreateTypeSourceInfo(TST);
+  TSI->getTypeLoc().castAs<TemplateSpecializationTypeLoc>().set(
       ElaboratedKeywordLoc, QualifierLoc, TemplateKeywordLoc, NameLoc,
       SpecifiedArgs);
-  return DI;
+  return TSI;
 }
 
 QualType ASTContext::getTemplateSpecializationType(

clang/lib/AST/ByteCode/Compiler.cpp

@@ -208,6 +208,19 @@ template <class Emitter> class LocOverrideScope final {
 } // namespace interp
 } // namespace clang
 
+template <class Emitter>
+bool Compiler<Emitter>::isValidBitCast(const CastExpr *E) {
+  QualType FromTy = E->getSubExpr()->getType()->getPointeeType();
+  QualType ToTy = E->getType()->getPointeeType();
+
+  if (classify(FromTy) == classify(ToTy))
+    return true;
+
+  if (FromTy->isVoidType() || ToTy->isVoidType())
+    return true;
+  return false;
+}
+
 template <class Emitter>
 bool Compiler<Emitter>::VisitCastExpr(const CastExpr *CE) {
   const Expr *SubExpr = CE->getSubExpr();
@@ -476,8 +489,9 @@ bool Compiler<Emitter>::VisitCastExpr(const CastExpr *CE) {
     return this->delegate(SubExpr);
 
   case CK_BitCast: {
+    QualType CETy = CE->getType();
     // Reject bitcasts to atomic types.
-    if (CE->getType()->isAtomicType()) {
+    if (CETy->isAtomicType()) {
       if (!this->discard(SubExpr))
         return false;
       return this->emitInvalidCast(CastKind::Reinterpret, /*Fatal=*/true, CE);
@@ -492,6 +506,10 @@ bool Compiler<Emitter>::VisitCastExpr(const CastExpr *CE) {
     if (!FromT || !ToT)
       return false;
 
+    if (!this->isValidBitCast(CE) &&
+        !this->emitInvalidCast(CastKind::ReinterpretLike, /*Fatal=*/false, CE))
+      return false;
+
     assert(isPtrType(*FromT));
     assert(isPtrType(*ToT));
     if (FromT == ToT) {

clang/lib/AST/ByteCode/Compiler.h

@@ -425,6 +425,8 @@ class Compiler : public ConstStmtVisitor<Compiler<Emitter>, bool>,
 
   bool refersToUnion(const Expr *E);
 
+  bool isValidBitCast(const CastExpr *E);
+
 protected:
   /// Variable to storage mapping.
   llvm::DenseMap<const ValueDecl *, Scope::Local> Locals;

clang/lib/AST/ByteCode/Interp.h

@@ -1915,6 +1915,9 @@ bool Load(InterpState &S, CodePtr OpPC) {
     return false;
   if (!Ptr.isBlockPointer())
     return false;
+  if (const Descriptor *D = Ptr.getFieldDesc();
+      !(D->isPrimitive() || D->isPrimitiveArray()) || D->getPrimType() != Name)
+    return false;
   S.Stk.push<T>(Ptr.deref<T>());
   return true;
 }
@@ -1926,6 +1929,9 @@ bool LoadPop(InterpState &S, CodePtr OpPC) {
     return false;
   if (!Ptr.isBlockPointer())
     return false;
+  if (const Descriptor *D = Ptr.getFieldDesc();
+      !(D->isPrimitive() || D->isPrimitiveArray()) || D->getPrimType() != Name)
+    return false;
   S.Stk.push<T>(Ptr.deref<T>());
   return true;
 }
@@ -3288,12 +3294,18 @@ inline bool InvalidCast(InterpState &S, CodePtr OpPC, CastKind Kind,
                         bool Fatal) {
   const SourceLocation &Loc = S.Current->getLocation(OpPC);
 
-  if (Kind == CastKind::Reinterpret) {
+  switch (Kind) {
+  case CastKind::Reinterpret:
     S.CCEDiag(Loc, diag::note_constexpr_invalid_cast)
-        << static_cast<unsigned>(Kind) << S.Current->getRange(OpPC);
+        << diag::ConstexprInvalidCastKind::Reinterpret
+        << S.Current->getRange(OpPC);
     return !Fatal;
-  }
-  if (Kind == CastKind::Volatile) {
+  case CastKind::ReinterpretLike:
+    S.CCEDiag(Loc, diag::note_constexpr_invalid_cast)
+        << diag::ConstexprInvalidCastKind::ThisConversionOrReinterpret
+        << S.getLangOpts().CPlusPlus << S.Current->getRange(OpPC);
+    return !Fatal;
+  case CastKind::Volatile:
     if (!S.checkingPotentialConstantExpression()) {
       const auto *E = cast<CastExpr>(S.Current->getExpr(OpPC));
       if (S.getLangOpts().CPlusPlus)
@@ -3304,14 +3316,13 @@ inline bool InvalidCast(InterpState &S, CodePtr OpPC, CastKind Kind,
     }
 
     return false;
-  }
-  if (Kind == CastKind::Dynamic) {
+  case CastKind::Dynamic:
     assert(!S.getLangOpts().CPlusPlus20);
-    S.CCEDiag(S.Current->getSource(OpPC), diag::note_constexpr_invalid_cast)
+    S.CCEDiag(Loc, diag::note_constexpr_invalid_cast)
         << diag::ConstexprInvalidCastKind::Dynamic;
     return true;
   }
-
+  llvm_unreachable("Unhandled CastKind");
   return false;
 }