| Tue, 09 Apr 2013 19:56:30 +0200 |
Thomas Wuerthinger |
Common base LIR instruction for call ops.
|
| Tue, 09 Apr 2013 19:51:24 +0200 |
Thomas Wuerthinger |
Common base class for LIR runtime call ops.
|
| Tue, 09 Apr 2013 19:29:26 +0200 |
Thomas Wuerthinger |
Merge.
|
| Tue, 09 Apr 2013 19:29:12 +0200 |
Thomas Wuerthinger |
Runtime calls that do not destroy the caller's registers need no spilling at the call site.
|
| Tue, 09 Apr 2013 19:28:34 +0200 |
Thomas Wuerthinger |
Change the register_finalizer stub to not overwrite any registers.
|
| Tue, 09 Apr 2013 23:06:59 +0200 |
Christos Kotselidis |
Merge
|
| Tue, 09 Apr 2013 22:34:06 +0200 |
Christos Kotselidis |
Merge
|
| Tue, 09 Apr 2013 14:53:19 +0200 |
Christos Kotselidis |
Add FixedValueAnchor to ArrayRangeWriteBarrier Snippet
|
| Tue, 09 Apr 2013 14:50:40 +0200 |
Christos Kotselidis |
Fix checkstyle errors
|
| Tue, 09 Apr 2013 11:13:35 +0200 |
Christos Kotselidis |
Creation of snippets for ArrayCopy write barriers
|
| Tue, 09 Apr 2013 22:25:45 +0200 |
Lukas Stadler |
Merge
|
| Tue, 09 Apr 2013 22:24:42 +0200 |
Lukas Stadler |
PEA: fix virtualization of CheckCastNode (check type)
|
| Tue, 09 Apr 2013 20:37:06 +0200 |
Andreas Woess |
invokedynamic: constant fold call site target with assumption; minor fixes
|
| Tue, 09 Apr 2013 19:25:20 +0200 |
Gilles Duboscq |
Remove some System.(out|err).print...
|
| Tue, 09 Apr 2013 19:24:39 +0200 |
Gilles Duboscq |
Do not skip a checkstyle test on the next run if it does not succeed in the current run
|
| Tue, 09 Apr 2013 16:28:19 +0200 |
Gilles Duboscq |
Move framestate assignement to the hir
|
| Tue, 09 Apr 2013 16:27:59 +0200 |
Gilles Duboscq |
Add a deoptState for invokes so that they can loose their stateAfter
|
| Tue, 09 Apr 2013 16:26:39 +0200 |
Gilles Duboscq |
Add missing updateUsages calls
|
| Tue, 09 Apr 2013 18:08:57 +0200 |
Doug Simon |
Merge.
|
| Tue, 09 Apr 2013 18:05:12 +0200 |
Doug Simon |
making use of negative types in the type check profile associated with an instanceof
|
| Tue, 09 Apr 2013 17:38:33 +0200 |
Doug Simon |
fixed bug in transformation of a type check profile into type check hints
|
| Tue, 09 Apr 2013 17:37:17 +0200 |
Doug Simon |
extra InstanceOf tests
|
| Tue, 09 Apr 2013 15:59:15 +0200 |
Doug Simon |
modified (some) checks in SchedulePhase to raise a SchedulingError instead of an AssertionError
|
| Tue, 09 Apr 2013 13:30:54 +0200 |
Doug Simon |
made format of toString() consistent across HotSpot subclasses of JavaField and JavaMethod
|
| Tue, 09 Apr 2013 17:25:02 +0200 |
Andreas Woess |
Merge
|
| Tue, 09 Apr 2013 17:23:32 +0200 |
Andreas Woess |
draft call site target value assumption
|
| Tue, 09 Apr 2013 17:20:29 +0200 |
Andreas Woess |
canonicalize constant call site target
|
| Tue, 09 Apr 2013 17:11:17 +0200 |
Andreas Woess |
add basic invokedynamic support
|
| Tue, 09 Apr 2013 15:26:29 +0200 |
Thomas Wuerthinger |
Move tests into *.test packages.
|
| Tue, 09 Apr 2013 14:42:31 +0200 |
Thomas Wuerthinger |
Merge.
|
| Tue, 09 Apr 2013 13:00:02 +0200 |
Thomas Wuerthinger |
Small clean up.
|
| Tue, 09 Apr 2013 12:48:40 +0200 |
Christian Wirth |
Merged
|
| Tue, 09 Apr 2013 11:54:28 +0200 |
Christian Wirth |
commented out a check for the existence of "cl.exe" for the windows build. The command returns something invalid, so the build fails BECAUSE of this test
|
| Tue, 09 Apr 2013 11:55:19 +0200 |
Christos Kotselidis |
Make static the private calls of the WriteBarrierAdditionPhase
|
| Tue, 09 Apr 2013 11:38:46 +0200 |
Christos Kotselidis |
Remove unused packages
|
| Tue, 09 Apr 2013 11:35:01 +0200 |
Christos Kotselidis |
Merge
|
| Tue, 09 Apr 2013 11:32:42 +0200 |
Christos Kotselidis |
Merge
|
| Tue, 09 Apr 2013 09:45:35 +0200 |
Christos Kotselidis |
Move write barriers' addition to a separate stage after lowering
|
| Tue, 09 Apr 2013 10:48:14 +0200 |
Lukas Stadler |
Merge
|
| Tue, 09 Apr 2013 10:29:44 +0200 |
Lukas Stadler |
jacoco-exclude StampFactory (fixes problems with BeginNode.anchor)
|
| Tue, 09 Apr 2013 09:37:23 +0200 |
Lukas Stadler |
Merge
|
| Mon, 08 Apr 2013 20:15:44 +0200 |
Lukas Stadler |
Merge
|
| Mon, 08 Apr 2013 19:37:23 +0200 |
Lukas Stadler |
Merge
|
| Mon, 08 Apr 2013 19:18:41 +0200 |
Lukas Stadler |
add test for read elimination
|
| Mon, 08 Apr 2013 19:18:18 +0200 |
Lukas Stadler |
PEA: fix PartialEscapeAnalysisTest
|
| Mon, 08 Apr 2013 19:17:29 +0200 |
Lukas Stadler |
add more tests to BoxingEliminationTest
|
| Mon, 08 Apr 2013 19:16:48 +0200 |
Lukas Stadler |
PEA: update EscapeAnalysisTest
|
| Mon, 08 Apr 2013 19:15:00 +0200 |
Lukas Stadler |
PEA: handle identity-less objects in IntegerEqualsNode
|
| Mon, 08 Apr 2013 19:14:32 +0200 |
Lukas Stadler |
PEA: unify merge and loop logic, implement handling of identity-less virtual objects
|
| Mon, 08 Apr 2013 19:12:19 +0200 |
Lukas Stadler |
PEA: remove level from EffectList
|
| Mon, 08 Apr 2013 19:11:30 +0200 |
Lukas Stadler |
small change to MaterializeObjectNode interface
|
| Mon, 08 Apr 2013 19:07:15 +0200 |
Lukas Stadler |
allow GraalCompilerTests to ignore virtual and floating nodes in graph comparison
|
| Mon, 08 Apr 2013 19:05:54 +0200 |
Lukas Stadler |
small change to ReentrantBlockIterator interface
|
| Mon, 08 Apr 2013 19:05:13 +0200 |
Lukas Stadler |
missing change for 7828409c364c (pull materialization logic into VirtualObjectNode classes)
|
| Mon, 08 Apr 2013 19:03:50 +0200 |
Lukas Stadler |
only eliminate locks for instances
|
| Mon, 08 Apr 2013 19:03:25 +0200 |
Lukas Stadler |
move canonicalization of BoxNode to lowering
|
| Mon, 08 Apr 2013 19:01:07 +0200 |
Lukas Stadler |
UnboxNode is not a VirtualizableAllocation
|
| Mon, 08 Apr 2013 19:00:22 +0200 |
Lukas Stadler |
allow MethodSubstitutions and MacroSubstitutions to be forced (instead of depending on inlining)
|
| Mon, 08 Apr 2013 17:31:50 +0200 |
Lukas Stadler |
pull materialization logic into VirtualObjectNode classes
|
| Mon, 08 Apr 2013 17:30:05 +0200 |
Lukas Stadler |
refactored boxing identification and lowering, removed BoxingMethodPool and explicit boxing phases
|
| Mon, 08 Apr 2013 14:19:52 +0200 |
Lukas Stadler |
new GraalOption.BootstrapReplacements for debugging purposes
|
| Thu, 04 Apr 2013 18:41:39 +0200 |
Lukas Stadler |
small PEA refactoring
|
| Tue, 09 Apr 2013 10:11:52 +0200 |
Bernhard Urban |
ReadNode: nullcheck stamptype
|
| Tue, 09 Apr 2013 09:58:09 +0200 |
Doug Simon |
expanded type check hints to support use of negative hints (i.e., profiled types that failed the corresponding type check)
|
| Tue, 09 Apr 2013 09:35:07 +0200 |
Doug Simon |
convert non-int boxed primitives whose stack kind is int to Constants of the right kind
|
| Mon, 08 Apr 2013 21:33:22 +0200 |
Christos Kotselidis |
Remove unused import
|
| Mon, 08 Apr 2013 21:30:56 +0200 |
Christos Kotselidis |
Merge
|
| Mon, 08 Apr 2013 21:17:34 +0200 |
Christos Kotselidis |
Rename ArrayWriteBarrier to SerialWriteBarrier and move it to graal.nodes
|
| Mon, 08 Apr 2013 20:58:36 +0200 |
Christos Kotselidis |
Remove FieldWriteBarrier
|
| Mon, 08 Apr 2013 20:52:04 +0200 |
Christos Kotselidis |
Merge FieldWriteBarrier into ArrayWriteBarrier
|
| Mon, 08 Apr 2013 19:34:39 +0200 |
Doug Simon |
made method (and macro) substitution optional to support different JDK versions
|
| Mon, 08 Apr 2013 18:47:06 +0200 |
Gilles Duboscq |
Be more careful while updating jvm.cfg (fixes a bug where the -server configuration would overwrite the -server0 one)
|
| Mon, 08 Apr 2013 18:40:18 +0200 |
Gilles Duboscq |
Make a copy of the environ in mx clean to avoid forcing things like HOTSPOT_BUILD_JOBS=16 for other commands
|
| Mon, 08 Apr 2013 15:13:05 +0200 |
Gilles Duboscq |
Use the DeoptimizingNode interface of the Invokes in the LIRGenerator
|
| Mon, 08 Apr 2013 17:27:16 +0200 |
Bernhard Urban |
unsafeArrayCast: anchor cast after initialization of an array
|
| Tue, 02 Apr 2013 12:14:15 +0200 |
Bernhard Urban |
unsafeCast: anchor cast after initialization of an object
|
| Mon, 08 Apr 2013 17:09:10 +0200 |
Bernhard Urban |
rename: s/PushNodesThroughPi/PushThroughPiPhase/
|
| Mon, 08 Apr 2013 16:43:38 +0200 |
Christos Kotselidis |
Remove unused imports
|
| Mon, 08 Apr 2013 16:30:51 +0200 |
Christos Kotselidis |
Merge
|
| Mon, 08 Apr 2013 16:26:17 +0200 |
Christos Kotselidis |
Remove references to G1
|
| Mon, 08 Apr 2013 09:26:06 +0200 |
Gilles Duboscq |
Introduce DeoptimizingNode interface
|
| Wed, 27 Mar 2013 15:31:23 +0100 |
Gilles Duboscq |
mx: bench command should not run dacapos with -n 0 when using a specification such as scaladacapo:specs
|
| Mon, 08 Apr 2013 11:20:19 +0200 |
Doug Simon |
delete build/linux/shared/graal before compile in build-graal.xml to handle stale class files left behind after refactoring in Eclipse
|
| Sun, 07 Apr 2013 22:27:16 +0200 |
Gilles Duboscq |
Update the -client line in jvm.cfg when it already exists (ie IGNORE)
|
| Sun, 07 Apr 2013 19:37:38 +0200 |
Gilles Duboscq |
Merge
|
| Sun, 07 Apr 2013 17:04:59 +0200 |
Gilles Duboscq |
Fix client build in mx
|
| Sun, 07 Apr 2013 14:53:18 +0200 |
Gilles Duboscq |
Fix merge
|
| Sun, 07 Apr 2013 13:26:16 +0200 |
Gilles Duboscq |
Fix jdk creation check if db an man exist in source jdk not in the one we are creating
|
| Sun, 07 Apr 2013 13:30:37 +0200 |
Gilles Duboscq |
Adapt to changes in hotspot makefiles : pass our JAVA_HOME as ALT_BOOTDIR (so that things are built with a 'safe' jdk) and our jdk copy as JAVA_HOME (so that the jvm gets installed there)
|
| Sat, 06 Apr 2013 20:04:06 +0200 |
Gilles Duboscq |
Merge with hsx25
|
| Thu, 04 Apr 2013 21:06:39 -0700 |
amurillo |
Added tag hs25-b26 for changeset a947f40fb536
jdk8-b85
|
| Thu, 04 Apr 2013 21:06:38 -0700 |
amurillo |
Merge
hs25-b26
|
| Thu, 04 Apr 2013 17:01:34 -0700 |
bharadwaj |
Merge
|
| Thu, 04 Apr 2013 12:18:46 -0700 |
kvn |
8011198: LP64 setting is not preserved on Solaris after 8006965
|
| Thu, 04 Apr 2013 09:30:06 +0200 |
neliasso |
8006014: Memory leak in hotspot/src/share/vm/adlc/dfa.cpp
|
| Thu, 04 Apr 2013 02:48:21 -0700 |
roland |
Merge
|
| Thu, 04 Apr 2013 09:33:24 +0200 |
roland |
8010399: Test8009761.java "Failed: init recursive calls: 5498. After deopt 5494".
|
| Thu, 04 Apr 2013 09:18:47 +0200 |
neliasso |
8006008: Memory leak in hotspot/src/share/vm/adlc/archDesc.cpp
|
| Wed, 03 Apr 2013 15:00:55 -0700 |
rasbold |
8010437: guarantee(this->is8bit(imm8)) failed: Short forward jump exceeds 8-bit offset
|
| Wed, 03 Apr 2013 11:12:57 -0700 |
kvn |
8011102: Clear AVX registers after return from JNI call
|
| Tue, 02 Apr 2013 09:30:07 +0200 |
neliasso |
7034299: Faulty winsock initialization code
|
| Sat, 30 Mar 2013 08:01:05 -0700 |
kmo |
8011009: Use do-while(0) instead of while(0) in EC_TRACE and RC_TRACE* macros
|
| Fri, 29 Mar 2013 20:52:01 -0700 |
bharadwaj |
Merge
|
| Fri, 29 Mar 2013 17:25:27 +0100 |
roland |
8010934: assert failure in c1_LinearScan.cpp: "asumption: non-Constant instructions have only virtual operands"
|
| Tue, 26 Mar 2013 12:55:26 -0700 |
kvn |
8004640: C2 assert failure in memnode.cpp: NULL+offs not RAW address
|
| Tue, 26 Mar 2013 10:05:33 +0100 |
neliasso |
8010281: Remove code that is never executed
|
| Mon, 25 Mar 2013 17:13:26 -0700 |
twisti |
7198429: need checked categorization of caller-sensitive methods in the JDK
|
| Sun, 24 Mar 2013 12:43:13 -0700 |
morris |
Merge
|
| Sat, 23 Mar 2013 06:22:07 -0700 |
morris |
8009026: [parfait] Null pointer deference in hotspot/src/share/vm/code/nmethod.cpp
|
| Fri, 22 Mar 2013 13:35:31 -0400 |
bharadwaj |
Merge
|
| Fri, 22 Mar 2013 07:58:55 -0700 |
bharadwaj |
8009539: JVM crash when run lambda testng tests
|
| Thu, 21 Mar 2013 22:00:14 -0700 |
roland |
Merge
|
| Thu, 21 Mar 2013 09:27:54 +0100 |
roland |
7153771: array bound check elimination for c1
|
| Thu, 04 Apr 2013 19:07:32 +0200 |
mgerdin |
Merge
|
| Wed, 03 Apr 2013 09:19:02 +0200 |
mgerdin |
Merge
|
| Tue, 02 Apr 2013 10:03:02 +0200 |
tschatzl |
8005857: assert in GC_locker from PSOldGen::expand with -XX:+PrintGCDetails and Verbose
|
| Sat, 23 Mar 2013 09:16:37 +0100 |
ehelin |
8009408: gc/metaspace/ClassMetaspaceSizeInJmapHeap.java fails with "exit code 1"
|
| Fri, 29 Mar 2013 13:49:37 -0700 |
johnc |
8010463: G1: Crashes with -UseTLAB and heap verification
|
| Wed, 27 Mar 2013 19:21:18 +0100 |
tschatzl |
7112912: Message "Error occurred during initialization of VM" on boxes with lots of RAM
|
| Thu, 28 Mar 2013 10:27:28 +0100 |
mgerdin |
7014552: gc/lock/jni/jnilockXXX works too slow on 1-processor machine
|
