Thu, 20 Aug 2015 10:58:57 -0700 |
dcubed |
8130910: hsperfdata file is created in wrong directory and not cleaned up if /tmp/hsperfdata_<username> has wrong permissions
|
Thu, 20 Aug 2015 17:39:43 +0200 |
coleenp |
Merge
|
Thu, 20 Aug 2015 17:36:00 +0200 |
coleenp |
Merge
|
Thu, 20 Aug 2015 11:18:51 -0400 |
coleenp |
8133951: Zero interpreter asserts in stubRoutines.cpp
|
Thu, 20 Aug 2015 15:17:43 +0200 |
tschatzl |
8067339: PLAB reallocation might result in failure to allocate object in that recently allocated PLAB
|
Thu, 20 Aug 2015 15:17:43 +0200 |
tschatzl |
8073146: G1 merges thread local age tables too early with global age table
|
Thu, 20 Aug 2015 15:17:43 +0200 |
tschatzl |
8040162: Avoid reallocating PLABs between GC phases in G1
|
Thu, 20 Aug 2015 15:17:41 +0200 |
tschatzl |
8133530: Add JFR event for evacuation statistics
|
Wed, 19 Aug 2015 13:59:39 +0200 |
tschatzl |
8073013: Add detailed information about PLAB memory usage
|
Wed, 19 Aug 2015 13:50:50 +0200 |
tschatzl |
8133470: Uses of Atomic methods in plab.hpp should be moved to .inline.hpp file
|
Wed, 19 Aug 2015 13:47:40 +0200 |
tschatzl |
8003237: G1: Reduce unnecessary (and failing) allocation attempts when handling an evacuation failure
|
Thu, 20 Aug 2015 10:27:04 +0200 |
ehelin |
8133825: Remove the class G1CollectorPolicyExt
|
Thu, 20 Aug 2015 08:42:25 +0200 |
ehelin |
8133821: Refactor initialization of the heap and the collector policy
|
Wed, 19 Aug 2015 11:11:06 -0700 |
dcubed |
8133537: clarify position of unlock options in error messages
|
Thu, 20 Aug 2015 09:31:28 +0200 |
roland |
Merge
|
Wed, 29 Jul 2015 17:25:04 +0200 |
roland |
8080289: Intermediate writes in a loop not eliminated by optimizer
|
Fri, 14 Aug 2015 00:28:45 +0200 |
roland |
8133599: Unsafe.getAndSetObject() is no longer intrinsified by c2
|
Sat, 15 Aug 2015 02:54:18 +0200 |
roland |
8130847: Cloned object's fields observed as null after C2 escape analysis
|
Wed, 12 Aug 2015 09:58:39 +0300 |
shade |
8019968: Reference CAS induces GC store barrier even on failure
|
Tue, 18 Aug 2015 14:04:24 -0400 |
ecaspole |
8133349: CMS: Assert failed: Ctl pt invariant
|
Tue, 18 Aug 2015 17:32:10 +0200 |
jwilhelm |
8130308: Too low memory usage in TestPromotionFromSurvivorToTenuredAfterMinorGC.java
|
Mon, 17 Aug 2015 13:55:02 +0200 |
jwilhelm |
8133669: Remove unused code in Arguments
|
Sun, 16 Aug 2015 09:33:52 -0500 |
gziemski |
8133678: test fails due to 'CICompilerCount=0 must be at least 1' missing from stdout/stderr
|
Mon, 17 Aug 2015 10:53:11 +0000 |
dsamersoff |
Merge
|
Mon, 17 Aug 2015 12:43:45 +0300 |
dsamersoff |
Merge
|
Mon, 17 Aug 2015 10:43:11 +0200 |
david |
8133553: Running with -XX:+UseParallelGC -XX:OldSize=30k crashes jvm
|
Mon, 29 Jun 2015 11:11:12 +0200 |
stefank |
8087324: Use semaphores when starting and stopping GC task threads
|
Mon, 29 Jun 2015 11:09:39 +0200 |
stefank |
8087323: Unify and split the work gang classes
|
Fri, 14 Aug 2015 22:20:21 +0000 |
minqi |
Merge
|
Fri, 14 Aug 2015 20:40:48 +0000 |
minqi |
Merge
|
Fri, 14 Aug 2015 10:10:35 -0700 |
minqi |
8130115: REDO - Reduce Symbol::_identity_hash to 2 bytes
|
Fri, 14 Aug 2015 19:23:27 +0000 |
coleenp |
Merge
|
Fri, 14 Aug 2015 15:08:55 -0400 |
coleenp |
8133557: Missing test before a branch when checking for MethodCounters in TemplateTable::branch() on x86
|
Fri, 14 Aug 2015 04:21:37 +0200 |
dholmes |
Merge
|
Thu, 13 Aug 2015 22:08:40 -0400 |
dholmes |
8029453: java/util/concurrent/locks/ReentrantLock/TimeoutLockLoops.java failed by timeout
|
Thu, 13 Aug 2015 23:35:32 +0000 |
coleenp |
Merge
|
Thu, 13 Aug 2015 17:17:56 -0500 |
gziemski |
8112746: Followup to JDK-8059557 (JEP 245)
|
Thu, 13 Aug 2015 09:32:01 +0200 |
asiebenborn |
8133121: Move implementation of process_grey_object to concurrentMark.inline.hpp
|
Wed, 12 Aug 2015 14:18:12 -0400 |
mockner |
8098791: Remove PrintClassStatistics and PrintMethodStatistics
|
Wed, 12 Aug 2015 16:32:29 +0200 |
david |
8133193: Memory leak in G1 because G1RootProcessor doesn't have desctructor
|
Tue, 29 Sep 2015 16:51:31 +0200 |
rkennke |
Reworked C2 Shenandoah optimizer. Fixes some methods that couldn't be compiled, a SEGV caused by implicit null checks, and improves performance.
|
Fri, 25 Sep 2015 16:45:39 +0200 |
rkennke |
Resolve and update the referent field in Reference object during mark-compact.
|
Thu, 24 Sep 2015 12:49:23 +0200 |
rkennke |
Fix Shenandoah-improved CAS-obj intrinsic in C2.
|
Tue, 22 Sep 2015 23:47:24 +0200 |
rkennke |
Several small fixes to mark-compact. Most importantly, compact regions even if they are empty. That's important for properly setting their top after compaction.
|
Fri, 18 Sep 2015 15:44:29 +0200 |
rkennke |
Several smaller fixes to mark-compact: Fix derived pointers, more efficient next-compaction-region, initialize freed humongous regions correctly.
|
Wed, 16 Sep 2015 17:22:25 +0200 |
rkennke |
Use StoreLoad barrier after setting _do_full_gc flag. Makes sure that other threads see it when they ought to.
|
Wed, 16 Sep 2015 17:21:59 +0200 |
rkennke |
Implement barriers in new C2 intrinsics: multiplyToLen, squareToLen, mulAdd, montgomeryMultiply, montgomerySquare.
|
Wed, 16 Sep 2015 17:21:25 +0200 |
rkennke |
Update references in weak roots too, in mark-compact.
|
Wed, 16 Sep 2015 17:21:00 +0200 |
rkennke |
Don't do repeated write barrier on same object. Re-use the first that comes.
|
Wed, 16 Sep 2015 17:20:28 +0200 |
rkennke |
Eliminate barriers at parse time when possible.
|
Wed, 16 Sep 2015 17:19:55 +0200 |
rkennke |
Enable implicit null checks for write barriers.
|
Fri, 11 Sep 2015 18:36:30 +0200 |
rkennke |
Assert heap size consistency withing lock, to avoid race.
|
Fri, 11 Sep 2015 18:36:05 +0200 |
rkennke |
Fix small miscalculation when freeing humongous regions.
|
Fri, 11 Sep 2015 18:35:26 +0200 |
rkennke |
Only mark new objects during evacuation when updating refs early (i.e. when needed).
|
Tue, 08 Sep 2015 21:00:24 +0200 |
rkennke |
More efficient test for clear-bitmap. Print reset-bitmap trace in ShenandoahTracePhases.
|
Tue, 08 Sep 2015 20:59:48 +0200 |
rkennke |
Use release-store and load-acquire to read/write _do_full_gc field, to ensure consistency between GC and Java threads.
|
Tue, 08 Sep 2015 12:15:42 +0200 |
rkennke |
Don't pass timer to weak ref processor. We don't use those timings and it triggers asserts.
|
Tue, 08 Sep 2015 12:14:22 +0200 |
rkennke |
Only compact non-empty regions in ShenandoahMarkCompact.
|
Mon, 07 Sep 2015 23:43:31 +0200 |
rkennke |
Reset mark bitmap concurrently with Java threads and using all concurrent GC threads.
|
Mon, 07 Sep 2015 17:24:25 +0200 |
rkennke |
Merge
|