Mercurial > hg > truffle
comparison src/share/vm/oops/instanceKlass.cpp @ 13438:ad72068ac41e
8028993: Full collections with ParallelScavenge slower in JDK 8 compared to 7u40
Summary: Reducing the number of calls to follow_class_loader to speed up the marking phase. Also removed some unnecessary calls to adjust_klass.
Reviewed-by: stefank, jmasa, mgerdin
| author | sjohanss |
|---|---|
| date | Tue, 10 Dec 2013 10:31:00 +0100 |
| parents | 816c89d5957d |
| children | 02f27ecb4f3a 2353011244bd 5da8bb64b370 |
comparison
equal
deleted
inserted
replaced
| 13437:1de8e5356754 | 13438:ad72068ac41e |
|---|---|
| 2197 int size = size_helper(); | 2197 int size = size_helper(); |
| 2198 InstanceKlass_OOP_MAP_ITERATE( \ | 2198 InstanceKlass_OOP_MAP_ITERATE( \ |
| 2199 obj, \ | 2199 obj, \ |
| 2200 PSParallelCompact::adjust_pointer(p), \ | 2200 PSParallelCompact::adjust_pointer(p), \ |
| 2201 assert_is_in) | 2201 assert_is_in) |
| 2202 obj->update_header(cm); | |
| 2203 return size; | 2202 return size; |
| 2204 } | 2203 } |
| 2205 | 2204 |
| 2206 #endif // INCLUDE_ALL_GCS | 2205 #endif // INCLUDE_ALL_GCS |
| 2207 | 2206 |