RFR: 8152113: Remove _last_ditch_collection GC-cause and avoid expanding heap on Metaspace OOM (original) (raw)
Stefan Johansson stefan.johansson at oracle.com
Mon Mar 21 08:16:55 UTC 2016
- Previous message (by thread): RFR: 8152113: Remove _last_ditch_collection GC-cause and avoid expanding heap on Metaspace OOM
- Next message (by thread): RFR: JDK-8148759: G1AllocRegion::_count inconsistently used if more than one context is active
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Thanks for the review Jon, Stefan
On 2016-03-18 18:01, Jon Masamitsu wrote:
Looks good. Jon On 3/17/2016 10:03 AM, Stefan Johansson wrote: Hi,
Please review this change for RFE: https://bugs.openjdk.java.net/browse/JDK-8152113 Webrev: http://cr.openjdk.java.net/~sjohanss/8152113/hotspot.00/ Summary: Prior to this patch the GC-cause lastditchcollection has been used for 2 types of full-collections that clears soft references: 1. Last resort when out of metaspace memory 2. WhiteBox initiated full GC These have not much in common and the GC-cause is strangely named when looking at its meaning. This cause also comes with an unwanted side-effect, it will expand the heap if possible and this is strange since it is not caused by a failed heap-allocation. This change removes the lastditchcollection cause and now we instead use two separate causes, wbfullgc for the WhiteBox-case and metadataGCclearsoftrefs for the out of metaspace case. The idea is that these will work similar to lastditchcollection in every way except when it comes to growing the heap. Testing: Run several tonga test-lists in RBT with G1, CMS and Parallel without seeing any new issues. Thanks, Stefan
- Previous message (by thread): RFR: 8152113: Remove _last_ditch_collection GC-cause and avoid expanding heap on Metaspace OOM
- Next message (by thread): RFR: JDK-8148759: G1AllocRegion::_count inconsistently used if more than one context is active
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]