Fri, 10 Dec 2010 17:59:46 -0800 Merge
trims [Fri, 10 Dec 2010 17:59:46 -0800] rev 7422
Merge
Thu, 09 Dec 2010 21:47:58 -0800 Merge
ysr [Thu, 09 Dec 2010 21:47:58 -0800] rev 7421
Merge
Thu, 09 Dec 2010 09:22:57 -0800 7005259: CMS: BubbleUpRef asserts referent(obj)->is_oop() failed: Enqueued a bad referent
ysr [Thu, 09 Dec 2010 09:22:57 -0800] rev 7420
7005259: CMS: BubbleUpRef asserts referent(obj)->is_oop() failed: Enqueued a bad referent Summary: Relaxed the assert by allowing NULL referents when discovery may be concurrent. Reviewed-by: johnc, jcoomes
Tue, 07 Dec 2010 21:55:53 -0800 7001033: assert(gch->gc_cause() == GCCause::_scavenge_alot || !gch->incremental_collection_failed())
ysr [Tue, 07 Dec 2010 21:55:53 -0800] rev 7419
7001033: assert(gch->gc_cause() == GCCause::_scavenge_alot || !gch->incremental_collection_failed()) 7002546: regression on SpecJbb2005 on 7b118 comparing to 7b117 on small heaps Summary: Relaxed assertion checking related to incremental_collection_failed flag to allow for ExplicitGCInvokesConcurrent behaviour where we do not want a failing scavenge to bail to a stop-world collection. Parameterized incremental_collection_will_fail() so we can selectively use, or not use, as appropriate, the statistical prediction at specific use sites. This essentially reverts the scavenge bail-out logic to what it was prior to some recent changes that had inadvertently started using the statistical prediction which can be noisy in the presence of bursty loads. Added some associated verbose non-product debugging messages. Reviewed-by: johnc, tonyp
(0) -3000 -1000 -300 -100 -30 -10 -4 +4 +10 +30 +100 +300 +1000 +3000 +10000 +30000 tip