Wed, 09 Sep 2015 16:43:28 +0200 |
Josef Eisl |
mx microbench: respect --suite argument.
|
Fri, 18 Sep 2015 13:12:02 +0200 |
Roland Schatz |
Update jvmci import: Use AMD64Kind instead of JavaKind in backend.
|
Fri, 18 Sep 2015 11:41:06 +0200 |
Roland Schatz |
Use dummy PlatformKind in platform independent backend test.
|
Thu, 17 Sep 2015 22:15:07 -0700 |
Tom Rodriguez |
Update jvmci import: Replace JVMCICompileWithC1Only with package based controls
|
Thu, 17 Sep 2015 16:48:47 +0200 |
Roland Schatz |
Support inline null comparisons in SPARC backend.
|
Thu, 17 Sep 2015 15:05:54 +0200 |
Roland Schatz |
Use LIRKind.Illegal instead of JavaKind.Illegal in backend.
|
Thu, 17 Sep 2015 14:29:34 +0200 |
Roland Schatz |
Use correct LIRKind for constants in phis.
|
Thu, 17 Sep 2015 11:08:55 +0200 |
Roland Schatz |
Use correct wordKind in CurrentJavaThreadNode and FrameMap.
|
Thu, 17 Sep 2015 14:10:51 +0200 |
Stefan Anzinger |
[SPARC] Remove special null handling in SPARCLIRGenerator.emitConstant
|
Wed, 16 Sep 2015 21:56:12 +0200 |
Doug Simon |
work around for Eclipse bug 477597
|
Wed, 16 Sep 2015 21:17:42 +0200 |
Doug Simon |
updated imports to be explicit and added checkstyle rule to forbid * imports
|
Wed, 16 Sep 2015 15:43:24 +0200 |
Paul Woegerer |
Allow PartialEscapePhase to be reused for PEA selective inlining
|
Wed, 16 Sep 2015 14:42:30 +0200 |
Gilles Duboscq |
Always put TRUFFLE_API on the bootclasspath
|
Wed, 16 Sep 2015 14:12:37 +0200 |
Stefan Anzinger |
Update jvmci import version
|
Wed, 16 Sep 2015 12:19:31 +0200 |
Roland Schatz |
Update jvmci import: Clean separation between backend and frontend wordKind.
|
Wed, 16 Sep 2015 12:05:27 +0200 |
Roland Schatz |
Fix verification of return kind.
|
Wed, 16 Sep 2015 11:49:07 +0200 |
Stefan Anzinger |
CheckGraalInvariants: Use deterministic class order (outer before inner) and initialize them in single thread to avoid deadlocks during initialization
|
Tue, 15 Sep 2015 22:35:17 -0700 |
Tom Rodriguez |
Update jvmci import: Add test exercising CHA with default methods
|
Wed, 16 Sep 2015 01:09:19 +0200 |
Doug Simon |
removed methods from HotSpotGraalRuntimeProvider that access statically available JVMCI functionality (GRAAL-1235)
|
Wed, 16 Sep 2015 00:48:33 +0200 |
Doug Simon |
removed HotSpotGraalRuntime singleton - it's now owned by a HotSpotGraalCompiler instance (GRAAL-1235)
|
Tue, 15 Sep 2015 21:10:57 +0200 |
Andreas Woess |
improve GraphPE error reporting
|
Tue, 15 Sep 2015 17:05:35 +0200 |
Andreas Woess |
improve TraceTrufflePerformanceWarnings output
|
Tue, 15 Sep 2015 19:03:21 +0200 |
Stefan Anzinger |
Disable unittest System_setOut temporarily
|
Tue, 15 Sep 2015 16:52:18 +0200 |
Roland Schatz |
Make narrow pointer kinds configurable by backend.
|
Tue, 15 Sep 2015 15:36:07 +0200 |
Roland Schatz |
Explicitly pass return kind to LIRGenerator.emitReturn.
|
Tue, 15 Sep 2015 15:50:36 +0200 |
Stefan Anzinger |
[SPARC] Handle loading of HotSpotMetaspaceConstant
|
Tue, 15 Sep 2015 12:01:41 +0200 |
Stefan Anzinger |
Update jvmci and truffle import version
|
Tue, 15 Sep 2015 11:18:07 +0200 |
Stefan Anzinger |
Make static use of HotSpotJVMCIRuntimeProvider.getArrayIndexScale and friends
|
Mon, 14 Sep 2015 18:31:09 +0200 |
Stefan Anzinger |
Update import version of jvmci and truffle
|
Mon, 14 Sep 2015 18:13:50 +0200 |
Stefan Anzinger |
[SPARC] Fix problems introduced on refactoring Kind.Object and Value.getKind()
|
Mon, 14 Sep 2015 18:13:28 +0200 |
Stefan Anzinger |
Set ReservedCodeCacheSize=300m for CTW (SPARC exceeds the default value)
|
Mon, 14 Sep 2015 19:48:06 +0200 |
Andreas Woess |
updated imported truffle version
|
Mon, 14 Sep 2015 19:59:52 +0200 |
Andreas Woess |
fix typo
|
Mon, 14 Sep 2015 16:34:26 +0200 |
Roland Schatz |
Properly handle metaspace constants in switch statements.
|
Sun, 13 Sep 2015 21:03:52 +0200 |
Doug Simon |
removed unused "foldable" element from @NodeIntrinsic
|
Sun, 13 Sep 2015 21:03:05 +0200 |
Doug Simon |
updated imported jvmci version
|
Fri, 11 Sep 2015 08:59:29 -0700 |
Tom Rodriguez |
Update jvmci import: Move CTW command into Graal
|
Thu, 10 Sep 2015 10:13:16 -0700 |
Tom Rodriguez |
Only linked methods can be inlined
|
Wed, 09 Sep 2015 17:20:13 -0700 |
Tom Rodriguez |
Make some service classes public
|
Wed, 09 Sep 2015 22:12:46 +0200 |
Doug Simon |
updated imported jvmci version
|
Wed, 09 Sep 2015 21:31:22 +0200 |
Doug Simon |
cleanups in preparation for CompilerToVM becoming a concrete class (GRAAL-1255)
|
Wed, 09 Sep 2015 11:50:28 -0700 |
Tom Rodriguez |
Updated imported jvmci version
|
Wed, 09 Sep 2015 16:12:29 +0200 |
Doug Simon |
restricted Unsafe capability access to package or class scope (GRAAL-1257)
|
Wed, 09 Sep 2015 13:38:21 +0200 |
Josef Eisl |
c.o.g.microbenchmarks: fix checkstyle issues.
|
Wed, 09 Sep 2015 14:27:14 +0200 |
Josef Eisl |
c.o.g.microbenchmarks: fix warnings.
|
Wed, 09 Sep 2015 13:02:41 +0200 |
Josef Eisl |
c.o.g.microbenchmarks: add copyright header.
|
Wed, 09 Sep 2015 12:54:12 +0200 |
Josef Eisl |
c.o.g.microbenchmarks: add graal jmh benchmarks.
|
Wed, 09 Sep 2015 13:29:51 +0200 |
Josef Eisl |
mx microbench: disable JVMCI classloader.
|
Wed, 09 Sep 2015 14:24:10 +0200 |
Christian Humer |
Update truffle version.
|
Tue, 08 Sep 2015 18:22:02 -0700 |
Tom Rodriguez |
Disable getComponentType intrinsic for 1.9 and later
|
Tue, 08 Sep 2015 18:21:11 -0700 |
Tom Rodriguez |
Update jvmci import: Renames to be more consistent with jvmci-9
|
Tue, 08 Sep 2015 17:51:49 -0700 |
Tom Rodriguez |
Updated imported jvmci version
|
Tue, 08 Sep 2015 19:57:39 +0200 |
Roland Schatz |
Update jvmci import: Refactoring: Rename Kind to JavaKind.
|
Tue, 08 Sep 2015 18:53:24 +0200 |
Roland Schatz |
Update jvmci import: Use wordKind instead of Kind.Object in backend.
|
Tue, 08 Sep 2015 17:32:21 +0200 |
Josef Eisl |
TraceRA: reactivate assertion.
|
Tue, 08 Sep 2015 17:46:39 +0200 |
Josef Eisl |
TraceRA: TraceLinearScanLifetimeAnalysisPhase: remove unused code.
|
Tue, 08 Sep 2015 16:38:30 +0200 |
Josef Eisl |
TraceRA: make TraceLinearScanAllocationPhases static.
|
Tue, 08 Sep 2015 18:53:43 +0200 |
Josef Eisl |
TraceRA: introduce TraceAllocationPhase and TraceLinearScanAllocationPhase.
|
Tue, 08 Sep 2015 10:50:21 +0200 |
Josef Eisl |
TraceRA: always insert spill-moves and use TraceLinearScanResolveDataFlowPhase only for SSI resolution.
|
Mon, 07 Sep 2015 17:38:39 +0200 |
Josef Eisl |
TraceRA: TraceLinearScanLifetimeAnalysisPhase: higher dump level for intermediate interval dumps.
|