[builtins] Introduce proper Float64Tan operator.

src/crankshaft/x64/lithium-x64.cc

Index: src/crankshaft/x64/lithium-x64.cc
diff --git a/src/crankshaft/x64/lithium-x64.cc b/src/crankshaft/x64/lithium-x64.cc
index cf0ebc512d9e32018ebed40e4898110bd3df9e18..3302b69de965e1104a13cf34e6fa12978e7c5513 100644
--- a/src/crankshaft/x64/lithium-x64.cc
+++ b/src/crankshaft/x64/lithium-x64.cc
@@ -1974,13 +1974,6 @@ LInstruction* LChunkBuilder::DoDoubleBits(HDoubleBits* instr) {
 }
 
 
-LInstruction* LChunkBuilder::DoConstructDouble(HConstructDouble* instr) {
-  LOperand* lo = UseRegister(instr->lo());
-  LOperand* hi = UseRegister(instr->hi());
-  return DefineAsRegister(new(zone()) LConstructDouble(hi, lo));
-}
-
-
 LInstruction* LChunkBuilder::DoReturn(HReturn* instr) {
   LOperand* context = info()->IsStub() ? UseFixed(instr->context(), rsi) : NULL;
   LOperand* parameter_count = UseRegisterOrConstant(instr->parameter_count());