Reland [Clang] Make the SizeType, SignedSizeType and PtrdiffType be named sugar types (#149613)

The checks for the 'z' and 't' format specifiers added in the original
PR #143653 had some issues and were overly strict, causing some build
failures and were consequently reverted at
4c85bf2.

In the latest commit
27c5862,
I relaxed the checks for the 'z' and 't' format specifiers, so warnings
are now only issued when they are used with mismatched types.

The original intent of these checks was to diagnose code that assumes
the underlying type of `size_t` is `unsigned` or `unsigned long`, for
example:

```c
printf("%zu", 1ul); // Not portable, but not an error when size_t is unsigned long
```  

However, it produced a significant number of false positives. This was
partly because Clang does not treat the `typedef` `size_t` and
`__size_t` as having a common "sugar" type, and partly because a large
amount of existing code either assumes `unsigned` (or `unsigned long`)
is `size_t`, or they define the equivalent of size_t in their own way
(such as
sanitizer_internal_defs.h).https://github.com/llvm/llvm-project/blob/2e67dcfdcd023df2f06e0823eeea23990ce41534/compiler-rt/lib/sanitizer_common/sanitizer_internal_defs.h#L203

Author: YexuanXiao

clang-tools-extra/clangd/unittests/FindTargetTests.cpp

@@ -838,7 +838,7 @@ TEST_F(TargetDeclTest, OverloadExpr) {
   )cpp";
   // Sized deallocation is enabled by default in C++14 onwards.
   EXPECT_DECLS("CXXDeleteExpr",
-               "void operator delete(void *, unsigned long) noexcept");
+               "void operator delete(void *, __size_t) noexcept");
 }
 
 TEST_F(TargetDeclTest, DependentExprs) {

clang-tools-extra/clangd/unittests/HoverTests.cpp

@@ -2794,7 +2794,7 @@ TEST(Hover, All) {
           })cpp",
           [](HoverInfo &HI) {
             HI.Name = "expression";
-            HI.Type = "unsigned long";
+            HI.Type = {"__size_t", "unsigned long"};
             HI.Value = "1";
           }},
       {
@@ -2804,7 +2804,7 @@ TEST(Hover, All) {
           })cpp",
           [](HoverInfo &HI) {
             HI.Name = "expression";
-            HI.Type = "unsigned long";
+            HI.Type = {"__size_t", "unsigned long"};
             HI.Value = "1";
           }},
       {

clang/docs/ReleaseNotes.rst

@@ -46,6 +46,7 @@ Potentially Breaking Changes
   ``endbr64`` instruction at the labels named as possible branch
   destinations, so it is not safe to use a register-controlled branch
   instruction to branch to one. (In line with gcc.)
+- Added a sugar type `PredefinedSugarType` to improve diagnostic messages. (#GH143653)
 
 C/C++ Language Potentially Breaking Changes
 -------------------------------------------

clang/include/clang/AST/ASTContext.h

@@ -277,6 +277,11 @@ class ASTContext : public RefCountedBase<ASTContext> {
   mutable llvm::ContextualFoldingSet<ArrayParameterType, ASTContext &>
       ArrayParameterTypes;
 
+  /// Store the unique Type corresponding to each Kind.
+  mutable std::array<Type *,
+                     llvm::to_underlying(PredefinedSugarType::Kind::Last) + 1>
+      PredefinedSugarTypes{};
+
   /// The set of nested name specifiers.
   ///
   /// This set is managed by the NestedNameSpecifier class.
@@ -1569,6 +1574,8 @@ class ASTContext : public RefCountedBase<ASTContext> {
   /// and bit count.
   QualType getDependentBitIntType(bool Unsigned, Expr *BitsExpr) const;
 
+  QualType getPredefinedSugarType(PredefinedSugarType::Kind KD) const;
+
   /// Gets the struct used to keep track of the extended descriptor for
   /// pointer to blocks.
   QualType getBlockDescriptorExtendedType() const;
@@ -2001,11 +2008,13 @@ class ASTContext : public RefCountedBase<ASTContext> {
   /// <stddef.h>.
   ///
   /// The sizeof operator requires this (C99 6.5.3.4p4).
-  CanQualType getSizeType() const;
+  QualType getSizeType() const;
+
+  CanQualType getCanonicalSizeType() const;
 
   /// Return the unique signed counterpart of
   /// the integer type corresponding to size_t.
-  CanQualType getSignedSizeType() const;
+  QualType getSignedSizeType() const;
 
   /// Return the unique type for "intmax_t" (C99 7.18.1.5), defined in
   /// <stdint.h>.

clang/include/clang/AST/FormatString.h

@@ -489,7 +489,8 @@ class FormatSpecifier {
 
   /// For a TypedefType QT, if it is a named integer type such as size_t,
   /// assign the appropriate value to LM and return true.
-  static bool namedTypeToLengthModifier(QualType QT, LengthModifier &LM);
+  static bool namedTypeToLengthModifier(ASTContext &Ctx, QualType QT,
+                                        LengthModifier &LM);
 };
 
 } // end analyze_format_string namespace

clang/include/clang/AST/RecursiveASTVisitor.h

@@ -1208,6 +1208,8 @@ DEF_TRAVERSE_TYPE(BitIntType, {})
 DEF_TRAVERSE_TYPE(DependentBitIntType,
                   { TRY_TO(TraverseStmt(T->getNumBitsExpr())); })
 
+DEF_TRAVERSE_TYPE(PredefinedSugarType, {})
+
 #undef DEF_TRAVERSE_TYPE
 
 // ----------------- TypeLoc traversal -----------------
@@ -1524,6 +1526,8 @@ DEF_TRAVERSE_TYPELOC(DependentBitIntType, {
   TRY_TO(TraverseStmt(TL.getTypePtr()->getNumBitsExpr()));
 })
 
+DEF_TRAVERSE_TYPELOC(PredefinedSugarType, {})
+
 #undef DEF_TRAVERSE_TYPELOC
 
 // ----------------- Decl traversal -----------------

clang/include/clang/AST/Type.h

@@ -2258,6 +2258,30 @@ class alignas(TypeAlignment) Type : public ExtQualsTypeCommonBase {
     unsigned NumExpansions;
   };
 
+  enum class PredefinedSugarKind {
+    /// The "size_t" type.
+    SizeT,
+
+    /// The signed integer type corresponding to "size_t".
+    SignedSizeT,
+
+    /// The "ptrdiff_t" type.
+    PtrdiffT,
+
+    // Indicates how many items the enum has.
+    Last = PtrdiffT
+  };
+
+  class PresefinedSugarTypeBitfields {
+    friend class PredefinedSugarType;
+
+    LLVM_PREFERRED_TYPE(TypeBitfields)
+    unsigned : NumTypeBits;
+
+    LLVM_PREFERRED_TYPE(PredefinedSugarKind)
+    unsigned Kind : 8;
+  };
+
   class CountAttributedTypeBitfields {
     friend class CountAttributedType;
 
@@ -2297,6 +2321,7 @@ class alignas(TypeAlignment) Type : public ExtQualsTypeCommonBase {
       DependentTemplateSpecializationTypeBits;
     PackExpansionTypeBitfields PackExpansionTypeBits;
     CountAttributedTypeBitfields CountAttributedTypeBits;
+    PresefinedSugarTypeBitfields PredefinedSugarTypeBits;
   };
 
 private:
@@ -8038,6 +8063,37 @@ class DependentBitIntType final : public Type, public llvm::FoldingSetNode {
   }
 };
 
+class PredefinedSugarType final : public Type {
+public:
+  friend class ASTContext;
+  using Kind = PredefinedSugarKind;
+
+private:
+  PredefinedSugarType(Kind KD, const IdentifierInfo *IdentName,
+                      QualType CanonicalType)
+      : Type(PredefinedSugar, CanonicalType, TypeDependence::None),
+        Name(IdentName) {
+    PredefinedSugarTypeBits.Kind = llvm::to_underlying(KD);
+  }
+
+  static StringRef getName(Kind KD);
+
+  const IdentifierInfo *Name;
+
+public:
+  bool isSugared() const { return true; }
+
+  QualType desugar() const { return getCanonicalTypeInternal(); }
+
+  Kind getKind() const { return Kind(PredefinedSugarTypeBits.Kind); }
+
+  const IdentifierInfo *getIdentifier() const { return Name; }
+
+  static bool classof(const Type *T) {
+    return T->getTypeClass() == PredefinedSugar;
+  }
+};
+
 /// A qualifier set is used to build a set of qualifiers.
 class QualifierCollector : public Qualifiers {
 public:

clang/include/clang/AST/TypeLoc.h

@@ -2783,6 +2783,16 @@ class ObjCProtocolLoc {
   }
 };
 
+struct PredefinedSugarTypeLocInfo {}; // Nothing.
+
+class PredefinedSugarTypeLoc final
+    : public ConcreteTypeLoc<UnqualTypeLoc, PredefinedSugarTypeLoc,
+                             PredefinedSugarType, PredefinedSugarTypeLocInfo> {
+public:
+  void initializeLocal(ASTContext &Context, SourceLocation loc) {}
+  SourceRange getLocalSourceRange() const { return {}; }
+};
+
 } // namespace clang
 
 #endif // LLVM_CLANG_AST_TYPELOC_H

clang/include/clang/AST/TypeProperties.td

@@ -1028,3 +1028,12 @@ let Class = DependentBitIntType in {
     return ctx.getDependentBitIntType(isUnsigned, numBitsExpr);
   }]>;
 }
+
+let Class = PredefinedSugarType in {
+  def : Property<"kind", UInt32> {
+    let Read = [{ static_cast<uint32_t>(node->getKind()) }];
+  }
+  def : Creator<[{
+    return ctx.getPredefinedSugarType(static_cast<PredefinedSugarType::Kind>(kind));
+  }]>;
+}

clang/include/clang/Basic/TypeNodes.td

@@ -117,3 +117,4 @@ def PipeType : TypeNode<Type>;
 def AtomicType : TypeNode<Type>;
 def BitIntType : TypeNode<Type>;
 def DependentBitIntType : TypeNode<Type>, AlwaysDependent;
+def PredefinedSugarType : TypeNode<Type>, NeverCanonical;