Wed, 23 Apr 2014 17:05:42 +0200 |
Bernhard Urban |
mx: fix imports
|
Wed, 23 Apr 2014 17:02:12 +0200 |
Bernhard Urban |
mx: use generators
|
Wed, 23 Apr 2014 14:59:05 +0200 |
Bernhard Urban |
mx: support simple filename globbing in whitelist
|
Wed, 23 Apr 2014 14:44:24 +0200 |
Bernhard Urban |
mx: support comments in whitelist
|
Wed, 23 Apr 2014 13:24:09 +0200 |
Bernhard Urban |
gate: add baseline compiler unittest
|
Wed, 23 Apr 2014 13:02:07 +0200 |
Bernhard Urban |
mx: add whitelist for shortunittest
|
Wed, 23 Apr 2014 11:48:38 +0200 |
Bernhard Urban |
mx: allow to specify a path for the whitelist in the unittest command
|
Wed, 23 Apr 2014 11:42:58 +0200 |
Bernhard Urban |
mx: remove longunittest
|
Wed, 23 Apr 2014 11:48:54 +0200 |
Bernhard Urban |
unittest: remove @LongTest
|
Wed, 23 Apr 2014 16:22:17 +0200 |
Gilles Duboscq |
Use association type for AddLocationNode inputs while waiting for a better solution.
|
Wed, 23 Apr 2014 13:29:55 +0200 |
Gilles Duboscq |
InliningUtil: split framestate processing to a separate method
|
Wed, 23 Apr 2014 16:16:48 +0200 |
Gilles Duboscq |
ininling: while fixing framestates, modify callee framestates rather than caller ones.
|
Wed, 23 Apr 2014 11:49:51 +0200 |
Gilles Duboscq |
When replacing sate splits with deopts, be careful about BeginNodes
|
Wed, 23 Apr 2014 15:48:38 +0200 |
Roland Schatz |
Merge.
|
Wed, 23 Apr 2014 15:22:20 +0200 |
Roland Schatz |
Fix bug in InputType specific replaceAtUsages.
|
Tue, 22 Apr 2014 17:58:49 +0200 |
Roland Schatz |
Make ValueAnchorNode implement AnchoringNode interface.
|
Tue, 22 Apr 2014 16:21:37 +0200 |
Roland Schatz |
Override asNode() in FloatingNode and FixedWithNextNode.
|
Tue, 15 Apr 2014 18:31:23 +0200 |
Christian Humer |
Truffle: fixes to the inlining heuristic.
|
Wed, 23 Apr 2014 12:38:05 +0200 |
Doug Simon |
added ModifiersProvider as superinterface for ResolvedJava[Method|Field|Type] and implemented all checks against modifiers as default methods
|
Wed, 23 Apr 2014 12:36:11 +0200 |
Doug Simon |
mx: ensure newest version of c1visualizer is being used
|
Wed, 23 Apr 2014 11:38:05 +0200 |
Josef Eisl |
Canonicalize projects.
|
Wed, 23 Apr 2014 10:08:22 +0200 |
Josef Eisl |
Rename LIRTypeTool to PlatformKindTool.
|
Wed, 23 Apr 2014 10:00:29 +0200 |
Josef Eisl |
Rename LIRLoop to BaselineLoop.
|
Tue, 22 Apr 2014 20:16:38 +0200 |
Josef Eisl |
Rename LIRFrameStateBuilder to BaselineFrameStateBuilder.
|
Tue, 22 Apr 2014 20:14:37 +0200 |
Josef Eisl |
Rename LIRControlFlowGraph to BaselineControlFlowGraph.
|
Tue, 22 Apr 2014 20:13:24 +0200 |
Josef Eisl |
Rename LIRProviders to CodeGenProviders.
|
Tue, 22 Apr 2014 20:12:21 +0200 |
Josef Eisl |
Remove (unused) LIRBlock.
|
Tue, 22 Apr 2014 19:45:52 +0200 |
Josef Eisl |
Remove deprecated isCompressed* checks from *HotSpotLIRGenerator.
|
Tue, 22 Apr 2014 18:41:00 +0200 |
Josef Eisl |
Change DeoptimizingNode and Access to LIRFrameState in *LIRGenerator*.
|
Tue, 22 Apr 2014 16:45:47 +0200 |
Josef Eisl |
Change graal.lir->graal.nodes project dependency to graal.nodes->graal.lir (errors).
|
Tue, 22 Apr 2014 16:37:53 +0200 |
Josef Eisl |
Move ArithmeticLIRGenerator to graal.lir (errors).
|
Tue, 22 Apr 2014 16:36:16 +0200 |
Josef Eisl |
Move LIRGeneratorTool to graal.lir (errors).
|
Tue, 22 Apr 2014 15:55:54 +0200 |
Josef Eisl |
Move LIRGenerator to graal.lir (errors).
|
Tue, 22 Apr 2014 15:40:13 +0200 |
Josef Eisl |
Move state(DeoptimizingNode) from LIRGenerator to NodeLIRBuilder (errors).
|
Tue, 22 Apr 2014 17:15:55 +0200 |
Josef Eisl |
Move FieldIntrospection to graal.compiler.common.
|
Wed, 23 Apr 2014 11:50:59 +0200 |
Josef Eisl |
Move FloatConvertNode.FloatConvert to compiler.common.
|
Tue, 22 Apr 2014 16:22:21 +0200 |
Josef Eisl |
Deprecate LIRGenerator.getDebugInfoBuilder().
|
Tue, 22 Apr 2014 15:45:19 +0200 |
Josef Eisl |
Move LIRGenerationResult* to graal.lir.
|
Tue, 22 Apr 2014 15:24:14 +0200 |
Josef Eisl |
Move UnsafeAccess to graal.compiler.common.
|
Wed, 23 Apr 2014 10:26:19 +0200 |
Lukas Stadler |
simplify usage of IncrementalCanonicalizerPhase, use incremental canonicalizer for GuardLowering
|
Wed, 23 Apr 2014 10:26:19 +0200 |
Lukas Stadler |
additional scope in inlining to get caller graph upon exceptions
|
Wed, 23 Apr 2014 10:26:19 +0200 |
Lukas Stadler |
loosen restrictions a bit in GraphOrder
|
Wed, 23 Apr 2014 10:26:19 +0200 |
Lukas Stadler |
pass CallTarget name to Truffle-generated StructuredGraphs
|
Wed, 23 Apr 2014 10:26:19 +0200 |
Lukas Stadler |
handle array length in PEAReadElimination
|
Wed, 23 Apr 2014 10:26:19 +0200 |
Lukas Stadler |
deal with StateSplits in GraphUtil.removeFixedWithUnusedInputs
|
Wed, 23 Apr 2014 10:26:19 +0200 |
Lukas Stadler |
correctly schedule loop proxies
|
Wed, 23 Apr 2014 10:26:18 +0200 |
Lukas Stadler |
warning in AbstractFrameStateBuilder
|
Wed, 23 Apr 2014 10:26:18 +0200 |
Lukas Stadler |
lowering for GuardProxyNode
|
Tue, 22 Apr 2014 18:32:08 -0700 |
Michael Van De Vanter |
Merge
|
Tue, 22 Apr 2014 18:22:40 -0700 |
Michael Van De Vanter |
Merge
|
Sun, 20 Apr 2014 20:37:27 -0700 |
Michael Van De Vanter |
Truffle: Major cleanup and extension of the Truffle Instrumentation framework in com.oracle.truffle.api
|
Tue, 22 Apr 2014 17:25:31 -0700 |
Tom Rodriguez |
use movl instead of andl for L2I
|
Tue, 22 Apr 2014 17:25:27 -0700 |
Tom Rodriguez |
add rotate instructions and make the shift encodings more uniform
|
Tue, 22 Apr 2014 17:25:24 -0700 |
Tom Rodriguez |
fix code to improve unsafe array location identity
|
Tue, 22 Apr 2014 16:41:46 -0700 |
Christian Wimmer |
Update c1visualizer to version compatible with Java 8
|
Wed, 23 Apr 2014 00:54:41 +0200 |
Doug Simon |
mx: generalized workaround for system command line length limits and used it to fix issue with eclipseformat command
|
Tue, 22 Apr 2014 20:35:55 +0200 |
Gilles Duboscq |
Exceptions reaching the unwind are already null-checked
|
Tue, 22 Apr 2014 20:18:45 +0200 |
Gilles Duboscq |
Dumping: translate special bcis into a string in the debug properties map.
|
Tue, 22 Apr 2014 19:52:05 +0200 |
Gilles Duboscq |
Move HotSpot-specific assertion from DebugInfoBuilder.computeFrameForState to HotSpotDebugInfoBuilder
|
Tue, 22 Apr 2014 18:54:03 +0200 |
Gilles Duboscq |
Remove special handling of monitor exit with AFTER_EXCEPTION_BCI in inlining. Anything using a framestate that has no equivalent after inlining is replaced by a deopt.
|