Mercurial > hg > graal-jvmci-8
graph
-
Truffle: further fixes to the new CallNode.Wed, 29 Jan 2014 20:49:09 +0100, by Christian Humer
-
Merge.Wed, 29 Jan 2014 12:19:03 +0100, by Christian Humer
-
SL: added junit integration for external tests.Tue, 28 Jan 2014 19:37:26 +0100, by Christian Humer
-
SL: minor fix in direct dispatch node.Tue, 28 Jan 2014 19:36:54 +0100, by Christian Humer
-
SL: migration to Truffle CallNode.Tue, 28 Jan 2014 13:37:07 +0100, by Christian Humer
-
Merge fixes.Tue, 28 Jan 2014 13:36:16 +0100, by Christian Humer
-
Merge.Tue, 28 Jan 2014 12:55:06 +0100, by Christian Humer
-
Merge.Tue, 28 Jan 2014 12:52:57 +0100, by Christian Humer
-
Truffle: refactored inlining interfaces to a more compact CallNode.Fri, 24 Jan 2014 15:55:41 +0100, by Christian Humer
-
later lowering of MonitorEnterNode and MonitorExitNodeWed, 29 Jan 2014 18:30:42 +0100, by Lukas Stadler
-
small cosmetic fixWed, 29 Jan 2014 18:32:19 +0100, by Lukas Stadler
-
In InstanceOfSnippets selection, use the exact (non-deoptimizing) snippet when we have an exact type even when the hint probabiliy is high enough for the instanceofWithProfile snippet.Wed, 29 Jan 2014 16:53:35 +0100, by Gilles Duboscq
-
FixedGuardNode simplification should conserve the deoptimization actionWed, 29 Jan 2014 15:46:15 +0100, by Gilles Duboscq
-
GraphBuilderPhase: remove unused constant added by createExceptionDispatchWed, 29 Jan 2014 13:36:20 +0100, by Andreas Woess
-
Truffle: disable UseExceptionProbability optimistic optimizationWed, 29 Jan 2014 13:10:02 +0100, by Andreas Woess
-
rework of benchmark counters: fix size at VM entry, not at build timeWed, 29 Jan 2014 09:21:50 +0100, by Lukas Stadler
-
make MonitorEnterNode iterableWed, 29 Jan 2014 10:45:43 +0100, by Lukas Stadler
-
add GraphUtil.removeFixedWithUnusedInputsWed, 29 Jan 2014 10:45:43 +0100, by Lukas Stadler
-
Fix incorrect code generation for conditional branch instructions. Reenable ControlPTXTest.Tue, 28 Jan 2014 21:37:29 -0500, by S.Bharadwaj Yadavalli
-
kill unused frame states when loop exits are removedTue, 28 Jan 2014 23:38:22 +0100, by Andreas Woess
-
fix handling of skipped exceptions (must not cut off other exception types or unwind)Tue, 28 Jan 2014 22:55:50 +0100, by Andreas Woess
-
add partial evaluation test for a ControlFlowException catch block that follows a SlowPathException catch blockTue, 28 Jan 2014 21:48:50 +0100, by Andreas Woess
-
PartialEvaluationTest should use the same GraphBuilderConfiguration as the TruffleCompiler; code cleanupTue, 28 Jan 2014 21:42:50 +0100, by Andreas Woess
-
guard again null receiver object when reading a constant fieldTue, 28 Jan 2014 21:24:24 +0100, by Doug Simon
-
added documentationTue, 28 Jan 2014 18:46:43 +0100, by Doug Simon
-
ignore StaticMethod16InArraysTest until multi-GPU support is completeTue, 28 Jan 2014 18:34:04 +0100, by Doug Simon
-
Merge.Tue, 28 Jan 2014 18:32:30 +0100, by Doug Simon
-
moved HSAIL compilation harness logic into HSAILHotSpotBackend and deleted HSAILCompilationResultMon, 27 Jan 2014 20:24:12 +0100, by Doug Simon
-
better assertion message in GraphOrderTue, 28 Jan 2014 14:46:17 +0100, by Lukas Stadler
-
remember InvokeKind in MacroNodeTue, 28 Jan 2014 14:45:56 +0100, by Lukas Stadler
-
test for ShortCiruitNode/guard handlingTue, 28 Jan 2014 14:39:52 +0100, by Lukas Stadler
-
canonicalize a LoadField on a null constant object to deoptTue, 28 Jan 2014 14:39:48 +0100, by Lukas Stadler
-
fixed pylint warningsTue, 28 Jan 2014 12:38:08 +0100, by Doug Simon
-
PTXHotSpotBackend$1.getGraphFor() checks GPUOffload VM option (JBS:GRAAL-6)Tue, 28 Jan 2014 12:31:31 +0100, by Doug Simon
-
fixed HotSpotResolvedJavaMethod.isSynthetic so that it doesn't do any class loading (JBS:GRAAL-5)Tue, 28 Jan 2014 12:27:39 +0100, by Doug Simon
-
an empty architecture specification implies all architecturesTue, 28 Jan 2014 12:19:30 +0100, by Doug Simon
-
use lowest version JRE specified by any project for source path in Eclipse attach/launch configurationTue, 28 Jan 2014 12:18:38 +0100, by Doug Simon
-
Use a OptimizedCallTarget-bound SpeculationLog for Truffle compilationsMon, 27 Jan 2014 16:39:17 +0100, by Gilles Duboscq
-
Throw-away speculation logs should not be used. Just use null insteadMon, 27 Jan 2014 16:05:10 +0100, by Gilles Duboscq
-
Fix: HotSpotDisassemblerProvider should return null when the code could not be disassembled.Mon, 27 Jan 2014 15:39:27 +0100, by Gilles Duboscq
-
MergeMon, 27 Jan 2014 15:28:50 -0800, by Tom Rodriguez
-
Add duplicate detection to IGV binary graphsMon, 27 Jan 2014 15:26:26 -0800, by Tom Rodriguez
-
Only use mature method datasWed, 22 Jan 2014 11:10:09 -0800, by Tom Rodriguez
-
print name and sig in TraceDeoptimization outputWed, 22 Jan 2014 10:54:38 -0800, by Tom Rodriguez
-
added getCodeString() to ExternalCompilationResultMon, 27 Jan 2014 15:50:19 +0100, by Doug Simon
-
removed unused PTX codeMon, 27 Jan 2014 15:49:33 +0100, by Doug Simon
-
updated mx/projects for new Okra versionSat, 25 Jan 2014 20:02:07 +0100, by Doug Simon
-
Fix SL unit test directoryFri, 24 Jan 2014 19:19:16 -0800, by Christian Wimmer
-
Fix Eclipse formatter issuesFri, 24 Jan 2014 18:42:52 -0800, by Christian Wimmer
-
Use source and expected output files to test Simple Language, instead of individual JUnit tests with the source and expected output as stringsFri, 24 Jan 2014 18:18:49 -0800, by Christian Wimmer
-
Refactoring and cleanup of Simple Language (more to come soon)Fri, 24 Jan 2014 18:16:24 -0800, by Christian Wimmer
-
Merge.Fri, 24 Jan 2014 18:34:18 +0100, by Doug Simon
-
Fix CollapseFrameForSingleSideEffectPhase.Fri, 24 Jan 2014 17:43:14 +0100, by Roland Schatz
-
refactored phase for stamping parameters as non-null that was duplicated in the GPU backendsFri, 24 Jan 2014 18:06:03 +0100, by Doug Simon
-
added ObjectPTXTest to test object parameter passing - all tests are disabled until GC issues are resolvedFri, 24 Jan 2014 18:04:49 +0100, by Doug Simon
-
enabled some tests in FloatPTXTest that now workFri, 24 Jan 2014 18:04:11 +0100, by Doug Simon
-
disabled ControlPTXTest until issues are resolvedFri, 24 Jan 2014 18:03:55 +0100, by Doug Simon
-
added work-around to handle CUDA context management in PTX C++ codeFri, 24 Jan 2014 18:03:09 +0100, by Doug Simon
-
partially fixed passing of object parameters to PTX kernels; use a C++ object for managing resource allocation and cleanup around a PTX kernel executionFri, 24 Jan 2014 18:01:36 +0100, by Doug Simon