Mon, 30 Sep 2013 21:46:41 +0200 |
Andreas Woess |
gtx_ptx.cpp: fix switch default label and operator precedence in ncores().
|
Mon, 30 Sep 2013 13:03:47 -0400 |
Morris Meyer |
PTX warp limiter to available GPU processors
|
Mon, 30 Sep 2013 17:03:14 +0200 |
Gilles Duboscq |
Fix wrong local variable name in LoopPolicies.shouldFullUnroll
|
Mon, 30 Sep 2013 16:10:32 +0200 |
Gilles Duboscq |
Keep the correct Deoptimization reason and action while grouping Deoptimizations
|
Mon, 30 Sep 2013 16:09:40 +0200 |
Gilles Duboscq |
Temporarily move encodeDeoptActionAndReason to MetaAccessProvider
|
Mon, 30 Sep 2013 18:05:25 +0200 |
Gilles Duboscq |
Merge
|
Thu, 26 Sep 2013 17:33:04 +0200 |
Gilles Duboscq |
Use ControlSplitNode and ControlSinkNode in GraphBuilderPhase.isBlockEnd
|
Thu, 26 Sep 2013 17:30:19 +0200 |
Gilles Duboscq |
Add a DynamicDeoptimizeNode where the action and reason is a input node.
|
Thu, 26 Sep 2013 16:26:56 +0200 |
Gilles Duboscq |
Fix copyright date in NodeWithState
|
Thu, 26 Sep 2013 16:10:02 +0200 |
Gilles Duboscq |
Remove DeoptimizingNode.getDeoptimizationReason
|
Thu, 26 Sep 2013 16:03:28 +0200 |
Gilles Duboscq |
SpeculationLog gives back the necessary Constant when calling maySpeculate. Remove the deoptimizationReasonfrom the DebugInfo
|
Thu, 26 Sep 2013 14:41:12 +0200 |
Gilles Duboscq |
CodeCacheProvider.encodeDeoptActionAndReason now returns a Constant
|
Mon, 30 Sep 2013 16:49:08 +0200 |
Doug Simon |
Merge.
|
Mon, 30 Sep 2013 16:48:24 +0200 |
Doug Simon |
enhanced test for CipherBlockChaining substitutions to cover path where the substitutions call the original (i.e. substituted) methods on slow paths
|
Mon, 30 Sep 2013 10:03:05 +0200 |
Doug Simon |
omit canonicalization in IncrementalCanonicalizerPhase if none of the contained phases changed the graph
|
Mon, 30 Sep 2013 14:02:07 +0200 |
Lukas Stadler |
new PrintAfterCompilation option, simplify PrintCompilation output
|
Mon, 30 Sep 2013 11:36:52 +0200 |
Bernhard Urban |
graal code installer: minor style cleanup
|
Mon, 30 Sep 2013 11:16:36 +0200 |
Bernhard Urban |
AMD64HotSpot: use conditional jump for IC_MISS_HANDLER
|
Mon, 30 Sep 2013 09:32:18 +0200 |
Bernhard Urban |
AMD64HotSpot: emit jump to IC miss handler directly instead of emitting it inside verfied_entry
|
Mon, 30 Sep 2013 09:47:41 +0200 |
Doug Simon |
improved javadoc for ResolvedJavaField.readConstantValue and .readValue
|
Mon, 30 Sep 2013 09:47:32 +0200 |
Doug Simon |
improved documentation and testing for ResolvedJavaType.resolveMethod() (GRAAL-489)
|
Sun, 29 Sep 2013 14:47:12 -0400 |
Morris Meyer |
PTX: one-dimensional parallel warp invocation, ParallelOver annotation
|
Sat, 28 Sep 2013 21:06:12 -0400 |
Morris Meyer |
PTX single-threaded array store, Warp annotation
|
Fri, 27 Sep 2013 19:51:01 +0200 |
Christos Kotselidis |
Merge
|
Fri, 27 Sep 2013 19:50:35 +0200 |
Christos Kotselidis |
Readjust branch probabilities of G1 Barriers
|
Fri, 27 Sep 2013 16:39:54 +0200 |
Michael Haupt |
put Node.getEncapsulatedSourceSection() on slow path
|
Fri, 27 Sep 2013 12:36:29 +0200 |
Christos Kotselidis |
Augment Write Barrier counters
|
Thu, 26 Sep 2013 15:53:21 -0700 |
Christian Wimmer |
Provide object read/write methods similar to Pointer.readXxxx/writeXxx that do not require casts to Word first, and use them in relevant places
|