[compiler] Drop the CompareNilIC.

src/full-codegen/x64/full-codegen-x64.cc

Index: src/full-codegen/x64/full-codegen-x64.cc
diff --git a/src/full-codegen/x64/full-codegen-x64.cc b/src/full-codegen/x64/full-codegen-x64.cc
index 910b2cf9f062b578ba7d26d74943e85c3a1856d9..39862bc4d39cea38c63adf8a83ae7a218ce61909 100644
--- a/src/full-codegen/x64/full-codegen-x64.cc
+++ b/src/full-codegen/x64/full-codegen-x64.cc
@@ -3919,10 +3919,11 @@ void FullCodeGenerator::EmitLiteralCompareNil(CompareOperation* expr,
     __ CompareRoot(rax, nil_value);
     Split(equal, if_true, if_false, fall_through);
   } else {
-    Handle<Code> ic = CompareNilICStub::GetUninitialized(isolate(), nil);
-    CallIC(ic, expr->CompareOperationFeedbackId());
-    __ CompareRoot(rax, Heap::kTrueValueRootIndex);
-    Split(equal, if_true, if_false, fall_through);
+    __ JumpIfSmi(rax, if_false);
+    __ movp(rax, FieldOperand(rax, HeapObject::kMapOffset));
+    __ testb(FieldOperand(rax, Map::kBitFieldOffset),
+             Immediate(1 << Map::kIsUndetectable));

[mvstanton, 2016/02/25 23:08:57]

Adam and I were speculating that this allows a Map (internal object) to fall out
into a later ToBooleanStub input through rax, per today's Canary. Could it be
so?

[adamk, 2016/02/25 23:38:38]

If there's something wrong with this patch, I don't think this is it. I can't
figure a way that rax leaks into anything else past this block (certainly
haven't been able to generate a test case where we do this compare and then try
to use the ToBoolean stub).

+    Split(not_zero, if_true, if_false, fall_through);
   }
   context()->Plug(if_true, if_false);
 }