[CIR] initial support for pointer-to-data-member type

This patch adds initial support for the pointer-to-data-member type.
Specifically, this commit includes:

- New ops, types, and attributes:
  - Add a new type !cir.member_ptr
  - Add a new attribute #cir.data_member_ptr
  - Add a new operation cir.get_indirect_member
- CodeGen for pointer-to-data-member types and values
  - Lower C++ pointer-to-member type to !cir.member_ptr
  - Lower C++ expression &C::D to cir.const(#cir.data_member_ptr)
  - Lower C++ expression c.*p and c->*p to cir.get_indirect_member

Author: Lancern

clang/include/clang/CIR/Dialect/IR/CIRAttrs.td

@@ -224,6 +224,36 @@ def ConstPtrAttr : CIR_Attr<"ConstPtr", "ptr", [TypedAttrInterface]> {
   let hasCustomAssemblyFormat = 1;
 }
 
+//===----------------------------------------------------------------------===//
+// DataMemberPtrAttr
+//===----------------------------------------------------------------------===//
+
+def DataMemberPtrAttr : CIR_Attr<"DataMemberPtr", "data_member_ptr",
+                                 [TypedAttrInterface]> {
+  let summary = "Holds a constant data member pointer value";
+  let parameters = (ins AttributeSelfTypeParameter<"">:$type,
+                        "Type":$clsTy,
+                        "IntegerAttr":$memberIndex);
+  let description = [{
+    A data member pointer attribute is a literal attribute that represents a
+    constant pointer-to-data-member value.
+  }];
+  let builders = [
+    AttrBuilderWithInferredContext<(ins "Type":$type, "Type":$clsTy,
+                                        "IntegerAttr":$memberIndex), [{
+      return $_get(type.getContext(), type, clsTy, memberIndex);
+    }]>,
+    AttrBuilderWithInferredContext<(ins "Type":$type, "Type":$clsTy,
+                                        "uint64_t":$memberIndex), [{
+      return $_get(type.getContext(), type, clsTy,
+          IntegerAttr::get(IndexType::get(type.getContext()), memberIndex));
+    }]>,
+  ];
+
+  let hasCustomAssemblyFormat = 1;
+  let genVerifyDecl = 1;
+}
+
 //===----------------------------------------------------------------------===//
 // SignedOverflowBehaviorAttr
 //===----------------------------------------------------------------------===//

clang/include/clang/CIR/Dialect/IR/CIROps.td

@@ -1672,6 +1672,35 @@ def GetMemberOp : CIR_Op<"get_member"> {
   let hasVerifier = 1;
 }
 
+//===----------------------------------------------------------------------===//
+// GetIndirectMemberOp
+//===----------------------------------------------------------------------===//
+
+def GetIndirectMemberOp : CIR_Op<"get_indirect_member"> {
+  let summary = "Get the address of a member of a struct";
+  let description = [{
+    The `cir.get_indirect_member` operation gets the address of a member from
+    the input record. The target member is given by a value of type
+    `!cir.member_ptr` (i.e. a pointer-to-data-member value).
+
+    It expects a pointer to the base record as well as the pointer to the target
+    member.
+  }];
+
+  let arguments = (ins
+    Arg<CIR_PointerType, "address of the struct object", [MemRead]>:$addr,
+    Arg<CIR_MemberPtrType, "pointer to the target member">:$member);
+
+  let results = (outs Res<CIR_PointerType, "">:$result);
+
+  let assemblyFormat = [{
+    $addr `[` $member `:` qualified(type($member)) `]` attr-dict
+    `:` qualified(type($addr)) `->` qualified(type($result))
+  }];
+
+  let hasVerifier = 1;
+}
+
 //===----------------------------------------------------------------------===//
 // VecExtractOp
 //===----------------------------------------------------------------------===//

clang/include/clang/CIR/Dialect/IR/CIRTypes.td

@@ -111,6 +111,36 @@ def CIR_PointerType : CIR_Type<"Pointer", "ptr",
   let hasCustomAssemblyFormat = 1;
 }
 
+//===----------------------------------------------------------------------===//
+// MemberPtrType
+//===----------------------------------------------------------------------===//
+
+def CIR_MemberPtrType : CIR_Type<"MemberPtr", "member_ptr",
+    [DeclareTypeInterfaceMethods<DataLayoutTypeInterface>]> {
+
+  let summary = "CIR pointer to data member type";
+  let description = [{
+    `cir.member_ptr` models the pointer-to-member type in C++.
+  }];
+
+  let parameters = (ins "mlir::Type":$memberTy, "mlir::Type":$clsTy);
+
+  let assemblyFormat = [{
+    `<` $memberTy `in` $clsTy `>`
+  }];
+
+  let extraClassDeclaration = [{
+    /// Determine whether this type represents a pointer-to-member-data type.
+    bool isMemberDataPtr() const;
+
+    /// Determine whether this type represents a pointer-to-member-function
+    /// type.
+    bool isMemberFunctionPtr() const;
+  }];
+
+  let genVerifyDecl = 1;
+}
+
 //===----------------------------------------------------------------------===//
 // BoolType
 //
@@ -262,8 +292,8 @@ def CIR_StructType : Type<CPred<"$_self.isa<::mlir::cir::StructType>()">,
                           "CIR struct type">;
 
 def CIR_AnyType : AnyTypeOf<[
-  CIR_IntType, CIR_PointerType, CIR_BoolType, CIR_ArrayType, CIR_VectorType,
-  CIR_FuncType, CIR_VoidType, CIR_StructType, AnyFloat,
+  CIR_IntType, CIR_PointerType, CIR_MemberPtrType, CIR_BoolType, CIR_ArrayType,
+  CIR_VectorType, CIR_FuncType, CIR_VoidType, CIR_StructType, AnyFloat,
 ]>;
 
 

clang/lib/CIR/CodeGen/CIRGenClass.cpp

@@ -1476,3 +1476,63 @@ mlir::Value CIRGenFunction::getVTablePtr(SourceLocation Loc, Address This,
 
   return VTable;
 }
+
+Address CIRGenFunction::buildCXXMemberDataPointerAddress(
+    const Expr *E, Address base, mlir::Value memberPtr,
+    const MemberPointerType *memberPtrType, LValueBaseInfo *baseInfo) {
+  auto memberPtrTy = memberPtr.getType().cast<mlir::cir::MemberPtrType>();
+  // TODO(cir): consider address space.
+  auto resultTy = builder.getPointerTo(memberPtrTy.getMemberTy());
+
+  auto op = builder.create<mlir::cir::GetIndirectMemberOp>(
+      getLoc(E->getSourceRange()), resultTy, base.getPointer(), memberPtr);
+
+  QualType memberType = memberPtrType->getPointeeType();
+  CharUnits memberAlign = CGM.getNaturalTypeAlignment(memberType, baseInfo);
+  memberAlign = CGM.getDynamicOffsetAlignment(
+      base.getAlignment(), memberPtrType->getClass()->getAsCXXRecordDecl(),
+      memberAlign);
+
+  return Address(op, convertTypeForMem(memberPtrType->getPointeeType()),
+                 memberAlign);
+}
+
+clang::CharUnits
+CIRGenModule::getDynamicOffsetAlignment(clang::CharUnits actualBaseAlign,
+                                        const clang::CXXRecordDecl *baseDecl,
+                                        clang::CharUnits expectedTargetAlign) {
+  // If the base is an incomplete type (which is, alas, possible with
+  // member pointers), be pessimistic.
+  if (!baseDecl->isCompleteDefinition())
+    return std::min(actualBaseAlign, expectedTargetAlign);
+
+  auto &baseLayout = getASTContext().getASTRecordLayout(baseDecl);
+  CharUnits expectedBaseAlign = baseLayout.getNonVirtualAlignment();
+
+  // If the class is properly aligned, assume the target offset is, too.
+  //
+  // This actually isn't necessarily the right thing to do --- if the
+  // class is a complete object, but it's only properly aligned for a
+  // base subobject, then the alignments of things relative to it are
+  // probably off as well.  (Note that this requires the alignment of
+  // the target to be greater than the NV alignment of the derived
+  // class.)
+  //
+  // However, our approach to this kind of under-alignment can only
+  // ever be best effort; after all, we're never going to propagate
+  // alignments through variables or parameters.  Note, in particular,
+  // that constructing a polymorphic type in an address that's less
+  // than pointer-aligned will generally trap in the constructor,
+  // unless we someday add some sort of attribute to change the
+  // assumed alignment of 'this'.  So our goal here is pretty much
+  // just to allow the user to explicitly say that a pointer is
+  // under-aligned and then safely access its fields and vtables.
+  if (actualBaseAlign >= expectedBaseAlign) {
+    return expectedTargetAlign;
+  }
+
+  // Otherwise, we might be offset by an arbitrary multiple of the
+  // actual alignment.  The correct adjustment is to take the min of
+  // the two alignments.
+  return std::min(actualBaseAlign, expectedTargetAlign);
+}

clang/lib/CIR/CodeGen/CIRGenExpr.cpp

@@ -865,6 +865,27 @@ LValue CIRGenFunction::buildDeclRefLValue(const DeclRefExpr *E) {
   llvm_unreachable("Unhandled DeclRefExpr");
 }
 
+LValue
+CIRGenFunction::buildPointerToDataMemberBinaryExpr(const BinaryOperator *E) {
+  auto baseAddr = Address::invalid();
+  if (E->getOpcode() == BO_PtrMemD)
+    baseAddr = buildLValue(E->getLHS()).getAddress();
+  else if (E->getOpcode() == BO_PtrMemI)
+    baseAddr = buildPointerWithAlignment(E->getLHS());
+  else
+    llvm_unreachable("unexpected binary operator opcode");
+
+  const auto *memberPtrTy = E->getRHS()->getType()->castAs<MemberPointerType>();
+
+  auto memberPtr = buildScalarExpr(E->getRHS());
+
+  LValueBaseInfo baseInfo;
+  auto memberAddr = buildCXXMemberDataPointerAddress(E, baseAddr, memberPtr,
+                                                     memberPtrTy, &baseInfo);
+
+  return makeAddrLValue(memberAddr, memberPtrTy->getPointeeType(), baseInfo);
+}
+
 LValue CIRGenFunction::buildBinaryOperatorLValue(const BinaryOperator *E) {
   // Comma expressions just emit their LHS then their RHS as an l-value.
   if (E->getOpcode() == BO_Comma) {
@@ -873,7 +894,7 @@ LValue CIRGenFunction::buildBinaryOperatorLValue(const BinaryOperator *E) {
   }
 
   if (E->getOpcode() == BO_PtrMemD || E->getOpcode() == BO_PtrMemI)
-    assert(0 && "not implemented");
+    return buildPointerToDataMemberBinaryExpr(E);
 
   assert(E->getOpcode() == BO_Assign && "unexpected binary l-value");
 

clang/lib/CIR/CodeGen/CIRGenExprConst.cpp

@@ -1532,6 +1532,20 @@ mlir::Attribute ConstantEmitter::tryEmitPrivate(const APValue &Value,
     return buildArrayConstant(CGM, Desired, CommonElementType, NumElements,
                               Elts, typedFiller);
   }
+  case APValue::MemberPointer: {
+    const ValueDecl *memberDecl = Value.getMemberPointerDecl();
+    assert(!Value.isMemberPointerToDerivedMember() && "NYI");
+
+    auto cirTy =
+        CGM.getTypes().ConvertType(DestType).cast<mlir::cir::MemberPtrType>();
+
+    if (const auto *memberFuncDecl = dyn_cast<CXXMethodDecl>(memberDecl))
+      assert(0 && "not implemented");
+
+    const auto *fieldDecl = cast<FieldDecl>(memberDecl);
+    return mlir::cir::DataMemberPtrAttr::get(cirTy, cirTy.getClsTy(),
+                                             fieldDecl->getFieldIndex());
+  }
   case APValue::LValue:
     return ConstantLValueEmitter(*this, Value, DestType).tryEmit();
   case APValue::Struct:
@@ -1542,7 +1556,6 @@ mlir::Attribute ConstantEmitter::tryEmitPrivate(const APValue &Value,
   case APValue::ComplexFloat:
   case APValue::Vector:
   case APValue::AddrLabelDiff:
-  case APValue::MemberPointer:
     assert(0 && "not implemented");
   }
   llvm_unreachable("Unknown APValue kind");

clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp

@@ -495,7 +495,11 @@ class ScalarExprEmitter : public StmtVisitor<ScalarExprEmitter, mlir::Value> {
   }
 
   mlir::Value VisitUnaryAddrOf(const UnaryOperator *E) {
-    assert(!llvm::isa<MemberPointerType>(E->getType()) && "not implemented");
+    if (llvm::isa<MemberPointerType>(E->getType())) {
+      const DeclRefExpr *DRE = cast<DeclRefExpr>(E->getSubExpr());
+      return CGF.buildMemberPtrConstant(DRE, E->getType());
+    }
+
     return CGF.buildLValue(E->getSubExpr()).getPointer();
   }
 
@@ -647,8 +651,13 @@ class ScalarExprEmitter : public StmtVisitor<ScalarExprEmitter, mlir::Value> {
     return Visit(E->getRHS());
   }
 
-  mlir::Value VisitBinPtrMemD(const Expr *E) { llvm_unreachable("NYI"); }
-  mlir::Value VisitBinPtrMemI(const Expr *E) { llvm_unreachable("NYI"); }
+  mlir::Value VisitBinPtrMemD(const BinaryOperator *E) {
+    return buildLoadOfLValue(E);
+  }
+
+  mlir::Value VisitBinPtrMemI(const BinaryOperator *E) {
+    return buildLoadOfLValue(E);
+  }
 
   mlir::Value VisitCXXRewrittenBinaryOperator(CXXRewrittenBinaryOperator *E) {
     llvm_unreachable("NYI");
@@ -2289,6 +2298,29 @@ mlir::Value ScalarExprEmitter::VisitUnaryExprOrTypeTraitExpr(
                              E->EvaluateKnownConstInt(CGF.getContext()));
 }
 
+mlir::Value CIRGenFunction::buildMemberPtrConstant(const clang::DeclRefExpr *E,
+                                                   QualType Ty) {
+  auto loc = getLoc(E->getSourceRange());
+
+  const auto *memberPtrTy = cast<MemberPointerType>(Ty);
+  auto cirTy = getCIRType(Ty);
+  auto cirClsTy = getCIRType(QualType(memberPtrTy->getClass(), 0));
+
+  const auto *decl = E->getDecl();
+
+  // A member function pointer.
+  // Member function pointer is not supported yet.
+  if (const auto *methodDecl = dyn_cast<CXXMethodDecl>(decl))
+    assert(0 && "not implemented");
+
+  // Otherwise, a member data pointer.
+  const auto *fieldDecl = cast<FieldDecl>(decl);
+  return builder.create<mlir::cir::ConstantOp>(
+      loc, cirTy,
+      mlir::cir::DataMemberPtrAttr::get(cirTy, cirClsTy,
+                                        fieldDecl->getFieldIndex()));
+}
+
 mlir::Value CIRGenFunction::buildCheckedInBoundsGEP(
     mlir::Type ElemTy, mlir::Value Ptr, ArrayRef<mlir::Value> IdxList,
     bool SignedIndices, bool IsSubtraction, SourceLocation Loc) {

clang/lib/CIR/CodeGen/CIRGenFunction.h

@@ -746,6 +746,13 @@ class CIRGenFunction : public CIRGenTypeCache {
 
   LValue buildStmtExprLValue(const StmtExpr *E);
 
+  LValue buildPointerToDataMemberBinaryExpr(const BinaryOperator *E);
+
+  /// TODO: Add TBAAAccessInfo
+  Address buildCXXMemberDataPointerAddress(
+      const Expr *E, Address base, mlir::Value memberPtr,
+      const MemberPointerType *memberPtrType, LValueBaseInfo *baseInfo);
+
   /// Generate a call of the given function, expecting the given
   /// result type, and using the given argument list which specifies both the
   /// LLVM arguments and the types they were derived from.
@@ -1137,6 +1144,8 @@ class CIRGenFunction : public CIRGenTypeCache {
                             LValueBaseInfo *BaseInfo = nullptr,
                             KnownNonNull_t IsKnownNonNull = NotKnownNonNull);
 
+  mlir::Value buildMemberPtrConstant(const clang::DeclRefExpr *E, QualType Ty);
+
   LValue
   buildConditionalOperatorLValue(const AbstractConditionalOperator *expr);
 

clang/lib/CIR/CodeGen/CIRGenModule.h

@@ -374,6 +374,12 @@ class CIRGenModule : public CIRGenTypeCache {
                                            LValueBaseInfo *BaseInfo = nullptr,
                                            bool forPointeeType = false);
 
+  /// TODO: Add TBAAAccessInfo
+  clang::CharUnits
+  getDynamicOffsetAlignment(clang::CharUnits actualBaseAlign,
+                            const clang::CXXRecordDecl *baseDecl,
+                            clang::CharUnits expectedTargetAlign);
+
   mlir::cir::FuncOp getAddrOfCXXStructor(
       clang::GlobalDecl GD, const CIRGenFunctionInfo *FnInfo = nullptr,
       mlir::cir::FuncType FnType = nullptr, bool DontDefer = false,

clang/lib/CIR/CodeGen/CIRGenTypes.cpp

@@ -692,7 +692,11 @@ mlir::Type CIRGenTypes::ConvertType(QualType T) {
   }
 
   case Type::MemberPointer: {
-    assert(0 && "not implemented");
+    const auto *MPT = cast<MemberPointerType>(Ty);
+    auto memberTy = ConvertType(MPT->getPointeeType());
+    auto clsTy = ConvertType(QualType(MPT->getClass(), 0));
+    ResultType =
+        mlir::cir::MemberPtrType::get(Builder.getContext(), memberTy, clsTy);
     break;
   }