| Thu, 02 May 2013 12:12:51 -0700 |
Christian Wimmer |
Use a LocationIdentity interface instead of just Object for the location identity of a LocationNode
|
| Thu, 02 May 2013 10:38:48 -0700 |
Christian Wimmer |
Consistent naming of accessor methods in the LocationNode class hierarchy
|
| Thu, 02 May 2013 10:36:09 -0700 |
Christian Wimmer |
Refactor the LocationNode class hierarchy to allow a SnippetLocationNode, i.e., a LocationNode that can be used in snippets without requiring the individual components to be constants at the time the snippet is prepared
|
| Thu, 02 May 2013 10:26:30 -0700 |
Christian Wimmer |
Rename field to be consistent with name of accessor method
|
| Thu, 02 May 2013 01:06:50 +0200 |
Andreas Woess |
Preliminary Nashorn support in IGV.
|
| Thu, 02 May 2013 18:37:53 +0200 |
Andreas Woess |
FrameDescriptor: fix default frame slot kind.
|
| Thu, 02 May 2013 17:35:36 +0200 |
Doug Simon |
Merge.
|
| Thu, 02 May 2013 17:17:11 +0200 |
Doug Simon |
exposed whether ASSERT is defined to Java code and use it to enable checks in ExceptionHandlerStub
|
| Thu, 02 May 2013 17:16:00 +0200 |
Doug Simon |
exception PC stored in thread is cleared before entering ExceptionHandlerStub
|
| Thu, 02 May 2013 15:12:52 +0200 |
Doug Simon |
added comments clarifying the difference between _exception_oop and _pending_exception in the HotSpot Thread class
|
| Thu, 02 May 2013 14:12:24 +0200 |
Doug Simon |
Merge.
|
| Thu, 02 May 2013 12:19:41 +0200 |
Doug Simon |
Merge.
|
| Thu, 02 May 2013 11:40:07 +0200 |
Doug Simon |
rename: VERY_FAST_DEOPT_PATH_PROBABILITY -> VERY_FAST_PATH_PROBABILITY
|
| Thu, 02 May 2013 11:39:07 +0200 |
Doug Simon |
rename: HotSpotSnippetUtils -> HotSpotReplacementsUtil
|
| Thu, 02 May 2013 14:37:27 +0200 |
Gilles Duboscq |
Remove dead IsNull nodes in Guard lowering phase
|
| Thu, 02 May 2013 14:35:48 +0200 |
Gilles Duboscq |
Enable the loop safepoint elimination.
|
| Thu, 02 May 2013 16:32:39 +0200 |
Roland Schatz |
Pass fixed node into getMaterializedRepresentation.
|
| Thu, 02 May 2013 16:31:44 +0200 |
Roland Schatz |
Remove StackSlot and RegisterValue cache.
|
| Thu, 02 May 2013 16:30:31 +0200 |
Roland Schatz |
Merge.
|
| Thu, 02 May 2013 14:13:54 +0200 |
Roland Schatz |
Make register saving extensible.
|
| Thu, 02 May 2013 14:12:10 +0200 |
Roland Schatz |
Restructure register zapping.
|
| Thu, 02 May 2013 11:53:04 +0200 |
Roland Schatz |
Backed out changeset: 45a73cd19baa
|
| Tue, 30 Apr 2013 17:29:33 +0200 |
Lukas Stadler |
refactored monitor handling in EA: MonitorEnter/Exit is removed, not eliminated
|
| Tue, 30 Apr 2013 12:14:59 +0200 |
Lukas Stadler |
handle "~" in EscapeAnalyzeOnly
|
| Tue, 30 Apr 2013 12:11:52 +0200 |
Lukas Stadler |
remove debug code
|
| Mon, 29 Apr 2013 14:53:08 +0200 |
Lukas Stadler |
coalesce allocations during escape analysis
|
| Mon, 29 Apr 2013 18:38:16 +0200 |
Lukas Stadler |
factor out createFieldLocation in HotSpotRuntime
|
| Mon, 29 Apr 2013 10:12:37 +0200 |
Lukas Stadler |
only virtualize StoreIndexedNode if no store check is required
|
| Tue, 30 Apr 2013 13:25:05 +0200 |
Lukas Stadler |
lower FixedGuardNode to if-condition-deopt in after-guard lowering
|
| Thu, 25 Apr 2013 17:19:41 +0200 |
Lukas Stadler |
late lowering of allocations (NewInstanceNode, NewArrayNode, NewMultiArrayNode)
|
| Thu, 25 Apr 2013 17:10:08 +0200 |
Lukas Stadler |
don't verify ControlFlowGraph when connectBlocks == false
|
| Thu, 25 Apr 2013 17:09:40 +0200 |
Lukas Stadler |
make Node.getDebugProperties() final
|
| Thu, 02 May 2013 10:55:11 +0200 |
Bernhard Urban |
CheckCastNode: check if input of next CheckCastNode is the node itself
|
| Thu, 02 May 2013 10:55:07 +0200 |
Bernhard Urban |
CheckCastNode: better attempt to combine checkcasts
|
| Thu, 02 May 2013 10:55:04 +0200 |
Bernhard Urban |
nested checkcasts: add failing testcase
|
| Thu, 02 May 2013 06:08:02 +0200 |
Doug Simon |
a compiled stub can now specify whether it needs to preserve registers. If a stub does not preserve registers and assertions are enabled, then all non-temporary registers are zapped after a C runtime call from the stub.
|
| Wed, 01 May 2013 18:08:07 -0700 |
Christian Wimmer |
Merge
|
| Wed, 01 May 2013 18:07:00 -0700 |
Christian Wimmer |
Use MetaAccessProvider instead of CodeCacheProvider
|
| Wed, 01 May 2013 18:06:01 -0700 |
Christian Wimmer |
Add generic type to avoid Java compiler warnings
|
| Wed, 01 May 2013 23:42:23 +0200 |
Andreas Woess |
NodeUtil.findFirstNodeInstance bug fix
|
| Wed, 01 May 2013 19:21:32 +0200 |
Andreas Woess |
NodeUtil.cloneNode: cloned children array should preserve component type.
|
| Wed, 01 May 2013 23:02:46 +0200 |
Andreas Woess |
Windows build fix for PTX code
|
| Wed, 01 May 2013 13:33:07 -0700 |
Christian Wimmer |
Allow rounding errors in probabilities
|
| Wed, 01 May 2013 12:06:01 -0700 |
Christian Wimmer |
Pass Assumptions to CanonicalizerPhase to avoid NullPointerException
|
| Wed, 01 May 2013 11:29:19 -0700 |
Christian Wimmer |
Better handling of array access nodes in Word type rewriter
|
| Wed, 01 May 2013 11:27:45 -0700 |
Christian Wimmer |
Relax the type requirements for CheckCastDynamicNode
|
| Wed, 01 May 2013 10:01:17 -0700 |
Christian Wimmer |
Merge
|
| Wed, 01 May 2013 09:22:20 -0700 |
Christian Wimmer |
Merge
|
| Wed, 01 May 2013 09:21:35 -0700 |
Christian Wimmer |
Allow distinction between ClassCastException and ArrayStoreException. Add more canonicalizations for check casts.
|
| Wed, 01 May 2013 09:13:34 -0700 |
Christian Wimmer |
Bugfix: loop transformation needs the Assumptions object since it internally calls the Canonicalizer
|
| Wed, 01 May 2013 09:12:45 -0700 |
Christian Wimmer |
Inline method in snippet when replacement is available. Replacements are frequently used for native methods, which are per default excluded from inlining. But if there is a non-native replacement, we want it inlined.
|
| Wed, 01 May 2013 09:11:13 -0700 |
Christian Wimmer |
Infer stamp for LoadIndexedNode, so that we have more precise type information for the result of the array load. This is important when the array is a originally a phi function (whithout any type information) and the phi function gets eliminated later on.
|
| Wed, 01 May 2013 09:09:33 -0700 |
Christian Wimmer |
Fix parameter type of NodeIntrinsic
|
| Wed, 01 May 2013 09:09:08 -0700 |
Christian Wimmer |
Add JavaType-to-String that produces same result as Class.getName()
|
| Wed, 01 May 2013 18:04:28 +0200 |
Doug Simon |
Merge.
|
| Wed, 01 May 2013 17:46:12 +0200 |
Doug Simon |
replaced exception_handler_nofpu assembler stub with a compiled stub (GRAAL-81)
|
| Wed, 01 May 2013 17:28:04 +0200 |
Thomas Wuerthinger |
Make sure if trueSuccessor probability is always <= 1.0.
|
| Wed, 01 May 2013 17:05:14 +0200 |
Thomas Wuerthinger |
Fixes for probability calculation.
|
| Wed, 01 May 2013 16:49:14 +0200 |
Thomas Wuerthinger |
Remove dead code.
|
| Wed, 01 May 2013 16:30:52 +0200 |
Thomas Wuerthinger |
Ensure probabilities are never negative. Add additional assertions.
|
