[builtins] Introduce proper Float64Exp operator.

src/arm64/codegen-arm64.h

 // Copyright 2013 the V8 project authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #ifndef V8_ARM64_CODEGEN_ARM64_H_
 #define V8_ARM64_CODEGEN_ARM64_H_
 
 #include "src/ast/ast.h"
 #include "src/macro-assembler.h"
 
 namespace v8 {
 namespace internal {
 
 class StringCharLoadGenerator : public AllStatic {
  public:
   // Generates the code for handling different string types and loading the
   // indexed character into |result|.  We expect |index| as untagged input and
   // |result| as untagged output. Register index is asserted to be a 32-bit W
   // register.
   static void Generate(MacroAssembler* masm,
                        Register string,
                        Register index,
                        Register result,
                        Label* call_runtime);
 
  private:
   DISALLOW_COPY_AND_ASSIGN(StringCharLoadGenerator);
 };
 
-
-class MathExpGenerator : public AllStatic {
- public:
-  static void EmitMathExp(MacroAssembler* masm,
-                          DoubleRegister input,
-                          DoubleRegister result,
-                          DoubleRegister double_scratch1,
-                          DoubleRegister double_scratch2,
-                          Register temp1,
-                          Register temp2,
-                          Register temp3);
-
- private:
-  DISALLOW_COPY_AND_ASSIGN(MathExpGenerator);
-};
-
 }  // namespace internal
 }  // namespace v8
 
 #endif  // V8_ARM64_CODEGEN_ARM64_H_