[builtins] Introduce proper Float64Log1p operator.

src/builtins.cc

Index: src/builtins.cc
diff --git a/src/builtins.cc b/src/builtins.cc
index 79c895b74f58b28d2cab7a84d72fb91d4d7e8189..94f90efba78201948f28ffb17f721bf73be99f80 100644
--- a/src/builtins.cc
+++ b/src/builtins.cc
@@ -2474,6 +2474,18 @@ void Builtins::Generate_MathLog(CodeStubAssembler* assembler) {
   assembler->Return(result);
 }
 
+// ES6 section 20.2.2.21 Math.log1p ( x )
+void Builtins::Generate_MathLog1p(CodeStubAssembler* assembler) {
+  using compiler::Node;
+
+  Node* x = assembler->Parameter(1);
+  Node* context = assembler->Parameter(4);
+  Node* x_value = assembler->TruncateTaggedToFloat64(context, x);
+  Node* value = assembler->Float64Log1p(x_value);
+  Node* result = assembler->ChangeFloat64ToTagged(value);
+  assembler->Return(result);

[Yang, 2016/06/13 05:36:33]

If only we could reduce duplicate code for these sort of builtins (Math.log,
Math.log1p, Math.sqrt and future ones) :)

[Benedikt Meurer, 2016/06/13 05:40:17]

Acknowledged.

+}
+
 // ES6 section 20.2.2.28 Math.round ( x )
 void Builtins::Generate_MathRound(CodeStubAssembler* assembler) {
   Generate_MathRoundingOperation(assembler, &CodeStubAssembler::Float64Round);