[clang][Interp] Implement __builtin_classify_type (#71972)

This adds some infrastructure for unevaluated builtin calls, and uses the implementation from ExprConstant.cpp

Author: tbaederr

clang/lib/AST/ExprConstShared.h

@@ -0,0 +1,58 @@
+//===--- ExprConstShared.h - Shared consetxpr functionality ----*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// Shared functionality between the new constant expression
+// interpreter (AST/Interp/) and the current one (ExprConstant.cpp).
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_CLANG_LIB_AST_EXPRCONSTSHARED_H
+#define LLVM_CLANG_LIB_AST_EXPRCONSTSHARED_H
+
+namespace clang {
+class QualType;
+class LangOptions;
+} // namespace clang
+using namespace clang;
+/// Values returned by __builtin_classify_type, chosen to match the values
+/// produced by GCC's builtin.
+enum class GCCTypeClass {
+  None = -1,
+  Void = 0,
+  Integer = 1,
+  // GCC reserves 2 for character types, but instead classifies them as
+  // integers.
+  Enum = 3,
+  Bool = 4,
+  Pointer = 5,
+  // GCC reserves 6 for references, but appears to never use it (because
+  // expressions never have reference type, presumably).
+  PointerToDataMember = 7,
+  RealFloat = 8,
+  Complex = 9,
+  // GCC reserves 10 for functions, but does not use it since GCC version 6 due
+  // to decay to pointer. (Prior to version 6 it was only used in C++ mode).
+  // GCC claims to reserve 11 for pointers to member functions, but *actually*
+  // uses 12 for that purpose, same as for a class or struct. Maybe it
+  // internally implements a pointer to member as a struct?  Who knows.
+  PointerToMemberFunction = 12, // Not a bug, see above.
+  ClassOrStruct = 12,
+  Union = 13,
+  // GCC reserves 14 for arrays, but does not use it since GCC version 6 due to
+  // decay to pointer. (Prior to version 6 it was only used in C++ mode).
+  // GCC reserves 15 for strings, but actually uses 5 (pointer) for string
+  // literals.
+  // Lang = 16,
+  // OpaqueType = 17,
+  BitInt = 18
+};
+
+GCCTypeClass EvaluateBuiltinClassifyType(QualType T,
+                                         const LangOptions &LangOpts);
+
+#endif

clang/lib/AST/ExprConstant.cpp

@@ -32,6 +32,7 @@
 //
 //===----------------------------------------------------------------------===//
 
+#include "ExprConstShared.h"
 #include "Interp/Context.h"
 #include "Interp/Frame.h"
 #include "Interp/State.h"
@@ -11492,43 +11493,10 @@ bool IntExprEvaluator::CheckReferencedDecl(const Expr* E, const Decl* D) {
   return false;
 }
 
-/// Values returned by __builtin_classify_type, chosen to match the values
-/// produced by GCC's builtin.
-enum class GCCTypeClass {
-  None = -1,
-  Void = 0,
-  Integer = 1,
-  // GCC reserves 2 for character types, but instead classifies them as
-  // integers.
-  Enum = 3,
-  Bool = 4,
-  Pointer = 5,
-  // GCC reserves 6 for references, but appears to never use it (because
-  // expressions never have reference type, presumably).
-  PointerToDataMember = 7,
-  RealFloat = 8,
-  Complex = 9,
-  // GCC reserves 10 for functions, but does not use it since GCC version 6 due
-  // to decay to pointer. (Prior to version 6 it was only used in C++ mode).
-  // GCC claims to reserve 11 for pointers to member functions, but *actually*
-  // uses 12 for that purpose, same as for a class or struct. Maybe it
-  // internally implements a pointer to member as a struct?  Who knows.
-  PointerToMemberFunction = 12, // Not a bug, see above.
-  ClassOrStruct = 12,
-  Union = 13,
-  // GCC reserves 14 for arrays, but does not use it since GCC version 6 due to
-  // decay to pointer. (Prior to version 6 it was only used in C++ mode).
-  // GCC reserves 15 for strings, but actually uses 5 (pointer) for string
-  // literals.
-  // Lang = 16,
-  // OpaqueType = 17,
-  BitInt = 18
-};
-
 /// EvaluateBuiltinClassifyType - Evaluate __builtin_classify_type the same way
 /// as GCC.
-static GCCTypeClass
-EvaluateBuiltinClassifyType(QualType T, const LangOptions &LangOpts) {
+GCCTypeClass EvaluateBuiltinClassifyType(QualType T,
+                                         const LangOptions &LangOpts) {
   assert(!T->isDependentType() && "unexpected dependent type");
 
   QualType CanTy = T.getCanonicalType();

clang/lib/AST/Interp/ByteCodeEmitter.cpp

@@ -14,6 +14,7 @@
 #include "Program.h"
 #include "clang/AST/ASTLambda.h"
 #include "clang/AST/DeclCXX.h"
+#include "clang/Basic/Builtins.h"
 #include <type_traits>
 
 using namespace clang;
@@ -84,10 +85,16 @@ ByteCodeEmitter::compileFunc(const FunctionDecl *FuncDecl) {
 
   // Create a handle over the emitted code.
   Function *Func = P.getFunction(FuncDecl);
-  if (!Func)
-    Func = P.createFunction(FuncDecl, ParamOffset, std::move(ParamTypes),
-                            std::move(ParamDescriptors),
-                            std::move(ParamOffsets), HasThisPointer, HasRVO);
+  if (!Func) {
+    bool IsUnevaluatedBuiltin = false;
+    if (unsigned BI = FuncDecl->getBuiltinID())
+      IsUnevaluatedBuiltin = Ctx.getASTContext().BuiltinInfo.isUnevaluated(BI);
+
+    Func =
+        P.createFunction(FuncDecl, ParamOffset, std::move(ParamTypes),
+                         std::move(ParamDescriptors), std::move(ParamOffsets),
+                         HasThisPointer, HasRVO, IsUnevaluatedBuiltin);
+  }
 
   assert(Func);
   // For not-yet-defined functions, we only create a Function instance and

clang/lib/AST/Interp/ByteCodeExprGen.cpp

@@ -2231,10 +2231,12 @@ bool ByteCodeExprGen<Emitter>::VisitBuiltinCallExpr(const CallExpr *E) {
   if (!Func)
     return false;
 
-  // Put arguments on the stack.
-  for (const auto *Arg : E->arguments()) {
-    if (!this->visit(Arg))
-      return false;
+  if (!Func->isUnevaluatedBuiltin()) {
+    // Put arguments on the stack.
+    for (const auto *Arg : E->arguments()) {
+      if (!this->visit(Arg))
+        return false;
+    }
   }
 
   if (!this->emitCallBI(Func, E, E))

clang/lib/AST/Interp/Function.cpp

@@ -20,11 +20,12 @@ Function::Function(Program &P, const FunctionDecl *F, unsigned ArgSize,
                    llvm::SmallVectorImpl<PrimType> &&ParamTypes,
                    llvm::DenseMap<unsigned, ParamDescriptor> &&Params,
                    llvm::SmallVectorImpl<unsigned> &&ParamOffsets,
-                   bool HasThisPointer, bool HasRVO)
+                   bool HasThisPointer, bool HasRVO, bool UnevaluatedBuiltin)
     : P(P), Loc(F->getBeginLoc()), F(F), ArgSize(ArgSize),
       ParamTypes(std::move(ParamTypes)), Params(std::move(Params)),
       ParamOffsets(std::move(ParamOffsets)), HasThisPointer(HasThisPointer),
-      HasRVO(HasRVO), Variadic(F->isVariadic()) {}
+      HasRVO(HasRVO), Variadic(F->isVariadic()),
+      IsUnevaluatedBuiltin(UnevaluatedBuiltin) {}
 
 Function::ParamDescriptor Function::getParamDescriptor(unsigned Offset) const {
   auto It = Params.find(Offset);

clang/lib/AST/Interp/Function.h

@@ -179,6 +179,8 @@ class Function final {
 
   bool isBuiltin() const { return F->getBuiltinID() != 0; }
 
+  bool isUnevaluatedBuiltin() const { return IsUnevaluatedBuiltin; }
+
   unsigned getNumParams() const { return ParamTypes.size(); }
 
   unsigned getParamOffset(unsigned ParamIndex) const {
@@ -191,7 +193,7 @@ class Function final {
            llvm::SmallVectorImpl<PrimType> &&ParamTypes,
            llvm::DenseMap<unsigned, ParamDescriptor> &&Params,
            llvm::SmallVectorImpl<unsigned> &&ParamOffsets, bool HasThisPointer,
-           bool HasRVO);
+           bool HasRVO, bool UnevaluatedBuiltin);
 
   /// Sets the code of a function.
   void setCode(unsigned NewFrameSize, std::vector<std::byte> &&NewCode,
@@ -250,6 +252,7 @@ class Function final {
   bool HasBody = false;
   bool Defined = false;
   bool Variadic = false;
+  bool IsUnevaluatedBuiltin = false;
 
 public:
   /// Dumps the disassembled bytecode to \c llvm::errs().

clang/lib/AST/Interp/Interp.cpp

@@ -131,6 +131,9 @@ void cleanupAfterFunctionCall(InterpState &S, CodePtr OpPC) {
   const Function *CurFunc = S.Current->getFunction();
   assert(CurFunc);
 
+  if (CurFunc->isUnevaluatedBuiltin())
+    return;
+
   if (S.Current->Caller && CurFunc->isVariadic()) {
     // CallExpr we're look for is at the return PC of the current function, i.e.
     // in the caller.

clang/lib/AST/Interp/InterpBuiltin.cpp

@@ -5,6 +5,7 @@
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
+#include "../ExprConstShared.h"
 #include "Boolean.h"
 #include "Interp.h"
 #include "PrimType.h"
@@ -517,6 +518,21 @@ static bool interp__builtin_bitreverse(InterpState &S, CodePtr OpPC,
   return true;
 }
 
+static bool interp__builtin_classify_type(InterpState &S, CodePtr OpPC,
+                                          const InterpFrame *Frame,
+                                          const Function *Func,
+                                          const CallExpr *Call) {
+  // This is an unevaluated call, so there are no arguments on the stack.
+  assert(Call->getNumArgs() == 1);
+  const Expr *Arg = Call->getArg(0);
+
+  GCCTypeClass ResultClass =
+      EvaluateBuiltinClassifyType(Arg->getType(), S.getLangOpts());
+  int32_t ReturnVal = static_cast<int32_t>(ResultClass);
+  pushInt(S, ReturnVal);
+  return true;
+}
+
 bool InterpretBuiltin(InterpState &S, CodePtr OpPC, const Function *F,
                       const CallExpr *Call) {
   InterpFrame *Frame = S.Current;
@@ -681,6 +697,11 @@ bool InterpretBuiltin(InterpState &S, CodePtr OpPC, const Function *F,
       return false;
     break;
 
+  case Builtin::BI__builtin_classify_type:
+    if (!interp__builtin_classify_type(S, OpPC, Frame, F, Call))
+      return false;
+    break;
+
   default:
     return false;
   }

clang/test/Sema/builtin-classify-type.c

@@ -1,4 +1,5 @@
 // RUN: %clang_cc1 -fsyntax-only -verify %s -fblocks
+// RUN: %clang_cc1 -fsyntax-only -verify %s -fblocks -fexperimental-new-constant-interpreter
 
 // expected-no-diagnostics
 

clang/test/SemaCXX/builtin-classify-type.cpp

@@ -1,4 +1,5 @@
 // RUN: %clang_cc1 -fsyntax-only -verify -fblocks %s
+// RUN: %clang_cc1 -fsyntax-only -verify -fblocks %s -fexperimental-new-constant-interpreter
 
 // expected-no-diagnostics