[builtins] Introduce proper Float64Tan operator.

src/crankshaft/x64/lithium-codegen-x64.cc

Index: src/crankshaft/x64/lithium-codegen-x64.cc
diff --git a/src/crankshaft/x64/lithium-codegen-x64.cc b/src/crankshaft/x64/lithium-codegen-x64.cc
index aa56ea166a025d101f4162dc9b0cd532ddd9a686..901e928c59e7a62116a485f3970d85ffe168cb77 100644
--- a/src/crankshaft/x64/lithium-codegen-x64.cc
+++ b/src/crankshaft/x64/lithium-codegen-x64.cc
@@ -5096,17 +5096,6 @@ void LCodeGen::DoDoubleBits(LDoubleBits* instr) {
 }
 
 
-void LCodeGen::DoConstructDouble(LConstructDouble* instr) {
-  Register hi_reg = ToRegister(instr->hi());
-  Register lo_reg = ToRegister(instr->lo());
-  XMMRegister result_reg = ToDoubleRegister(instr->result());
-  __ movl(kScratchRegister, hi_reg);
-  __ shlq(kScratchRegister, Immediate(32));
-  __ orq(kScratchRegister, lo_reg);
-  __ Movq(result_reg, kScratchRegister);
-}
-
-
 void LCodeGen::DoAllocate(LAllocate* instr) {
   class DeferredAllocate final : public LDeferredCode {
    public: