Rename some things around incremental marking triggers

src/heap/heap.h

Index: src/heap/heap.h
diff --git a/src/heap/heap.h b/src/heap/heap.h
index 5dfae06c3ab1bd01dc673e3005fb7f923207fb49..4bcfcbb80a4f71b70c59c1455b45c55ced223eca 100644
--- a/src/heap/heap.h
+++ b/src/heap/heap.h
@@ -1241,9 +1241,7 @@ class Heap {
     survived_since_last_expansion_ += survived;
   }
 
-  inline bool NextGCIsLikelyToBeFull(intptr_t limit) {
-    if (FLAG_gc_global) return true;

[Erik Corry Chromium.org, 2015/04/17 10:40:41]

This is the only functional change. Previously, in --gc-global mode we would
always start incremental marking immediately after the previous GC.

This isn't really how I understand --gc-global which is described as "always
perform global GCs", ie it should just avoid new-space-only GCs.

I left in the bit about early incremental start with --stress-compaction,
because it seems like a stressy thing to do.

[Hannes Payer (out of office), 2015/04/17 10:48:17]

Acknowledged.

[titzer, 2015/04/20 12:47:24]

Your CL description claimed this was a refactoring CL. The subsequent functional
change and crashing on the waterfall on amd64 determined that was a lie.

-
+  inline bool HeapIsFullEnoughToStartIncrementalMarking(intptr_t limit) {

[Hannes Payer (out of office), 2015/04/17 10:48:17]

I don't feel strong about the function name.

     if (FLAG_stress_compaction && (gc_count_ & 1) != 0) return true;
 
     intptr_t adjusted_allocation_limit = limit - new_space_.Capacity();
@@ -2083,8 +2081,6 @@ class Heap {
       double idle_time_in_ms, size_t size_of_objects,
       size_t mark_compact_speed_in_bytes_per_ms);
 
-  bool WorthActivatingIncrementalMarking();
-
   void ClearObjectStats(bool clear_last_time_stats = false);
 
   inline void UpdateAllocationsHash(HeapObject* object);