Fri, 22 Mar 2013 12:32:09 -0700 Merge
stefank [Fri, 22 Mar 2013 12:32:09 -0700] rev 16393
Merge
Fri, 22 Mar 2013 10:32:21 +0100 Merge
stefank [Fri, 22 Mar 2013 10:32:21 +0100] rev 16392
Merge
Thu, 21 Mar 2013 09:07:43 +0100 8004241: NPG: Metaspace occupies more memory than specified by -XX:MaxMetaspaceSize option
mgerdin [Thu, 21 Mar 2013 09:07:43 +0100] rev 16391
8004241: NPG: Metaspace occupies more memory than specified by -XX:MaxMetaspaceSize option Summary: Enforce MaxMetaspaceSize for both metaspace parts, check MaxMetaspaceSize against "reserved", not "capacity" Reviewed-by: jmasa, johnc
Tue, 19 Mar 2013 09:38:37 -0700 8008301: G1: guarantee(satb_mq_set.completed_buffers_num() == 0) failure
johnc [Tue, 19 Mar 2013 09:38:37 -0700] rev 16390
8008301: G1: guarantee(satb_mq_set.completed_buffers_num() == 0) failure Summary: If the marking stack overflows while the marking tasks are draining the SATB buffers, remark will exit with some SATB buffers left unprocessed. Relax the guarantee to allow for overflow. Reviewed-by: jmasa, brutisso
Tue, 19 Mar 2013 00:57:39 -0700 8009940: G1: assert(_finger == _heap_end) failed, concurrentMark.cpp:809
johnc [Tue, 19 Mar 2013 00:57:39 -0700] rev 16389
8009940: G1: assert(_finger == _heap_end) failed, concurrentMark.cpp:809 Summary: Skip reference processing if the global marking stack overflows during remark. Refactor and rename set_phase(); move code that sets the concurrency level into its own routine. Do not call set_phase() from within parallel reference processing; use the concurrency level routine instead. The marking state should only set reset by CMTask[0] during the concurrent phase of the marking cycle; if an overflow occurs at any stage during the remark, the marking state will be reset after reference processing. Reviewed-by: brutisso, jmasa
Mon, 18 Mar 2013 11:05:27 -0700 8009536: G1: Apache Lucene hang during reference processing
johnc [Mon, 18 Mar 2013 11:05:27 -0700] rev 16388
8009536: G1: Apache Lucene hang during reference processing Summary: In CMTask::do_marking_step(), Skip offering termination and entering the first and second synchronization barriers if called from a serial context, i.e. the VM thread. Reviewed-by: brutisso, tschatzl
Mon, 18 Mar 2013 09:34:58 +0100 Merge
stefank [Mon, 18 Mar 2013 09:34:58 +0100] rev 16387
Merge
(0) -10000 -3000 -1000 -300 -100 -30 -10 -7 +7 +10 +30 +100 +300 +1000 +3000 +10000 +30000 tip