[builtins] Introduce proper Float64Tan operator.

src/compiler/js-builtin-reducer.h

 // Copyright 2014 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_COMPILER_JS_BUILTIN_REDUCER_H_
 #define V8_COMPILER_JS_BUILTIN_REDUCER_H_
 
 #include "src/compiler/graph-reducer.h"
 
 namespace v8 {
@@ skipping 32 matching lines @@
   Reduction ReduceMathLog1p(Node* node);
   Reduction ReduceMathLog10(Node* node);
   Reduction ReduceMathLog2(Node* node);
   Reduction ReduceMathMax(Node* node);
   Reduction ReduceMathMin(Node* node);
   Reduction ReduceMathCbrt(Node* node);
   Reduction ReduceMathExpm1(Node* node);
   Reduction ReduceMathRound(Node* node);
   Reduction ReduceMathSin(Node* node);
   Reduction ReduceMathSqrt(Node* node);
+  Reduction ReduceMathTan(Node* node);
   Reduction ReduceMathTrunc(Node* node);
   Reduction ReduceStringFromCharCode(Node* node);
 
   Node* ToNumber(Node* value);
   Node* ToUint32(Node* value);
 
   Graph* graph() const;
   JSGraph* jsgraph() const { return jsgraph_; }
   Isolate* isolate() const;
   CommonOperatorBuilder* common() const;
   SimplifiedOperatorBuilder* simplified() const;
 
   JSGraph* const jsgraph_;
   TypeCache const& type_cache_;
 };
 
 }  // namespace compiler
 }  // namespace internal
 }  // namespace v8
 
 #endif  // V8_COMPILER_JS_BUILTIN_REDUCER_H_