| Sat, 11 Apr 2015 00:15:55 -0700 |
Christian Wimmer |
Consistently use JNIHandles::make_local with THREAD parameter
|
| Wed, 08 Apr 2015 11:07:01 +0200 |
Gilles Duboscq |
Fix merge
|
| Tue, 07 Apr 2015 12:21:08 +0200 |
Doug Simon |
rename CompilerToVM.readUnsafeOop to readUncompressedOop and only use it for reading oops in VM data structures since Unsafe.getObject() can be used for other cases
|
| Tue, 17 Feb 2015 17:05:04 +0100 |
Thomas Wuerthinger |
Another fix in the cpp part.
|
| Tue, 17 Feb 2015 16:51:36 +0100 |
Thomas Wuerthinger |
Fix cpp code.
|
| Tue, 17 Feb 2015 16:32:47 +0100 |
Thomas Wuerthinger |
Do not eagerly resolve types in lookupType.
|
| Fri, 23 Jan 2015 10:54:21 +0100 |
Doug Simon |
removed GPU code
|
| Thu, 18 Dec 2014 15:31:12 +0100 |
Roland Schatz |
Generalize object reading in HotSpotMemoryAccessProvider.
|
| Mon, 03 Nov 2014 16:08:06 +0100 |
Doug Simon |
converted HotSpotResolvedJavaMethod to an interface
|
| Fri, 24 Oct 2014 12:45:27 -0700 |
Tom Rodriguez |
Use arrayOop subclasses in graalJavaAccess
|
| Tue, 21 Oct 2014 10:28:43 +0200 |
Gilles Duboscq |
MethodData::initialize: remove unused argument
|
| Wed, 15 Oct 2014 20:34:55 +0200 |
Gilles Duboscq |
Fix some format strings
|
| Wed, 08 Oct 2014 11:44:00 -0700 |
Tom Rodriguez |
Track code installation time
|
| Fri, 11 Jul 2014 13:47:47 +0200 |
Doug Simon |
made type resolution require an accessing class context
|
| Fri, 04 Jul 2014 16:06:44 +0200 |
Lukas Stadler |
send log output to native tty
|
| Mon, 30 Jun 2014 12:02:19 +0200 |
Stefan Anzinger |
Merge with 5cdcb94a7cf7d9782107cc582f3e4b50000d5d1f
|
| Fri, 06 Jun 2014 01:19:20 +0200 |
Stefan Anzinger |
Merge
|
| Fri, 30 May 2014 15:09:09 +0200 |
Stefan Anzinger |
Merge
|
| Thu, 24 Apr 2014 14:02:20 +0200 |
Stefan Anzinger |
Narrowing down the return value (short, char and bool) of called method on Big Endian architectures.
|
| Thu, 26 Jun 2014 11:56:21 +0200 |
Gilles Duboscq |
Respect the VM's wish to record non-safepoint debug info
|
| Tue, 24 Jun 2014 11:28:06 -0700 |
Tom Rodriguez |
use receiver type in findUniqueConcreteMethod
|
| Wed, 04 Jun 2014 04:08:02 -0700 |
Laurent Daynes |
Changed (jlong) NULL to (jlong) (address) NULL to remove warning that cause build failures on old version of gcc (4.1.2 as used on Oracle Linux Server 5.8).
|
| Tue, 03 Jun 2014 11:51:27 +0200 |
Doug Simon |
decoupled Graal runtime initialization and Graal compilation queue initialization
|
| Wed, 28 May 2014 17:41:59 +0200 |
Doug Simon |
more improvements to runtime initialization:
|
| Wed, 21 May 2014 22:22:06 -0700 |
Tom Rodriguez |
rename initializeBytecode to getBytecode and eliminate extra copy
|
| Tue, 20 May 2014 21:35:32 -0700 |
Tom Rodriguez |
don't use JNI natives to interact with VM metadata
|
| Tue, 20 May 2014 13:46:34 -0700 |
Tom Rodriguez |
initialize HotSpotVMConfig fields efficiently from C++
|
| Tue, 20 May 2014 13:13:28 +0200 |
Gilles Duboscq |
ResolvedJavaType.resolveMethod: fix javadoc, add assert in native code. update changelog
|
| Sat, 17 May 2014 12:45:46 +0200 |
Gilles Duboscq |
Rename HotSpotResolvedObjectType.metaspaceKlass to getMetaspaceKlass
|
| Sat, 17 May 2014 14:57:58 +0200 |
Gilles Duboscq |
Add resolved receiver type to ResolvedJavaMethod.isInVirtualMethodTable in order to be able to do vtable-calls for miranda and default methods
|
| Fri, 18 Apr 2014 13:50:15 +0200 |
Gilles Duboscq |
ResolvedJavaType.resolveMethod now takes a callerType that is used to check access rules. Make it work for default methods.
|
| Wed, 30 Apr 2014 15:41:44 -1000 |
twisti |
AMD64: implemented DeoptimizationStub.deoptimizationHandler
|
| Thu, 17 Apr 2014 17:21:55 +0200 |
Gilles Duboscq |
Add eager resolving for indy call sites
|
| Thu, 17 Apr 2014 00:44:32 +0200 |
Doug Simon |
HSAIL: added safepoint support
|
| Wed, 16 Apr 2014 19:00:14 +0200 |
Thomas Wuerthinger |
Merge.
|
| Wed, 16 Apr 2014 18:33:10 +0200 |
Thomas Wuerthinger |
Introduce versioning of installed code.
|
| Tue, 15 Apr 2014 15:01:40 +0200 |
Thomas Wuerthinger |
Convert InstalledCode from an interface into an abstract class.
|
| Tue, 15 Apr 2014 13:34:29 +0200 |
Doug Simon |
create NodeClass for each Node subclass during class initialization of the latter
|
| Mon, 14 Apr 2014 17:21:49 -1000 |
twisti |
add DeoptimizationStub
|
| Mon, 14 Apr 2014 19:11:47 +0200 |
Lukas Stadler |
add support for skipping stack frames in StackIntrospection.getStackTrace
|
| Mon, 14 Apr 2014 18:20:09 +0200 |
Lukas Stadler |
change to StackIntrospection and Truffle getStackTrace implementation
|
| Fri, 11 Apr 2014 11:52:19 +0200 |
Lukas Stadler |
new StackIntrospection interface to allow access to stack contents
|
| Thu, 10 Apr 2014 17:48:23 +0200 |
Doug Simon |
fixed performance regression in hosted mode execution caused by 5e6f29f287d6
|
| Thu, 10 Apr 2014 09:14:35 +0200 |
Doug Simon |
Merge.
|
| Thu, 10 Apr 2014 00:01:30 +0200 |
Doug Simon |
added InstanceKlass::_graal_node_class field to accelerate Node.getNodeClass()
|
| Wed, 09 Apr 2014 16:33:32 -0700 |
Tom Rodriguez |
fix PrintCompilation formatting and use same time source
|
| Fri, 21 Mar 2014 14:26:58 -0700 |
twisti |
make HotSpotResolvedJavaMethod leaner
|
| Fri, 14 Mar 2014 22:56:59 +0100 |
Doug Simon |
removed ResolvedJavaMethod.getCompiledCodeSize()
|
| Fri, 14 Mar 2014 22:09:46 +0100 |
Doug Simon |
modified HotSpotResolvedJavaMethod.getCompiledCodeSize() to only return Graal compiled code size
|
| Wed, 12 Mar 2014 17:33:51 -0700 |
twisti |
moved CompilerToVM.loadReferencedTypeInPool logic into Java
|
| Wed, 12 Mar 2014 15:12:28 -0700 |
twisti |
added HotSpotConstantPool.JVM_CONSTANT enum
|
| Thu, 13 Mar 2014 15:36:30 +0100 |
Doug Simon |
Made graph caching compilation local
|
| Sat, 08 Mar 2014 10:04:44 +0100 |
Doug Simon |
removed unnecessary lookupKlassByName method
|
| Thu, 06 Mar 2014 17:11:39 -0800 |
Tom Rodriguez |
enable C1 + Graal tiered
|
| Thu, 06 Mar 2014 22:45:25 -0800 |
twisti |
reorganized graalCompilerToVM.cpp
|
| Thu, 06 Mar 2014 22:40:48 -0800 |
twisti |
make HotSpotResolvedJavaMethod leaner
|
| Thu, 06 Mar 2014 21:03:59 -0800 |
twisti |
pass metaspace pointers instead of object to VM
|
| Thu, 06 Mar 2014 15:18:47 -0800 |
twisti |
refactor CompilerToVM interface
|
| Wed, 05 Mar 2014 19:07:29 +0100 |
Doug Simon |
removed remaining VMToCompiler calls for allocating Graal API objects
|
| Tue, 04 Mar 2014 22:23:24 -0800 |
twisti |
moved to-constant-pool-index conversion up to Java
|
| Mon, 24 Feb 2014 15:08:02 -0800 |
twisti |
remove CompilerToVM.getInstanceFields
|
| Tue, 18 Feb 2014 10:47:13 -0800 |
Tom Rodriguez |
Acquire proper locks before calling assign_compile_id
|
| Thu, 13 Feb 2014 18:46:15 +0100 |
Gilles Duboscq |
Add shouldBeInlined method to ResolvedJavaMethod, implement it for HotSpot and use it in the inlining phase
|
| Thu, 06 Feb 2014 18:44:14 +0100 |
Doug Simon |
consolidated GNFI code into graal.hotspot project and cleaned up the documentation and code
|
| Wed, 05 Feb 2014 10:37:07 +0100 |
Matthias Grimmer |
GNFI: set invalid rtld_default in HotSpotVMConfig
|
| Wed, 05 Feb 2014 09:32:30 +0100 |
Matthias Grimmer |
GNFI: add comments
|
| Thu, 30 Jan 2014 15:50:46 +0100 |
Matthias Grimmer |
GNFI fixes
|
| Thu, 30 Jan 2014 09:19:45 +0100 |
Matthias Grimmer |
Add the Graal Native Function Interface
|
| Thu, 30 Jan 2014 00:52:33 +0100 |
Doug Simon |
added support for co-existing GPU backends (JBS:GRAAL-1)
|
| Wed, 29 Jan 2014 09:21:50 +0100 |
Lukas Stadler |
rework of benchmark counters: fix size at VM entry, not at build time
|
| Mon, 27 Jan 2014 15:39:27 +0100 |
Gilles Duboscq |
Fix: HotSpotDisassemblerProvider should return null when the code could not be disassembled.
|
| Wed, 22 Jan 2014 11:10:09 -0800 |
Tom Rodriguez |
Only use mature method datas
|
| Mon, 20 Jan 2014 16:27:24 +0100 |
Doug Simon |
added check for invalidated code in executeCompiledMethodVarargs
|
| Thu, 12 Dec 2013 15:13:02 +0100 |
Gilles Duboscq |
Add a speculation oop for uncommon trap deoptimization. Save it in the SpeculationLog during deoptimization.
|
| Tue, 14 Jan 2014 10:14:31 -0800 |
Tom Rodriguez |
Use compile ids assigned by hotspot
|
| Thu, 09 Jan 2014 11:29:56 +0100 |
Gilles Duboscq |
Add DynamicNewInstanceNode and use it to intrinsify Unsafe.allocateInstance
|
| Tue, 07 Jan 2014 14:50:02 -0800 |
twisti |
removed unused variable in lookupFieldInPool
|
| Thu, 02 Jan 2014 18:02:01 +0100 |
Doug Simon |
made CodeInstaller subclassable and virtualized the methods for creating ScopeValues
|
| Thu, 26 Dec 2013 12:37:28 -0800 |
twisti |
moved JavaType creation in CompilerToVM.lookupType into Java
|
| Wed, 25 Dec 2013 20:27:59 -0800 |
twisti |
remove Graal mirror from Class
|
| Wed, 25 Dec 2013 12:14:29 -0800 |
twisti |
replaced CompilerToVM.getUniqueImplementor with getKlassImplementor and moved the logic into Java
|
| Wed, 18 Dec 2013 17:33:00 +0100 |
Christian Wirth |
JS: fix, evaluating jboolean directly gives a warning (and thus error) on Windows
|
| Tue, 17 Dec 2013 20:58:58 -0800 |
twisti |
cleaned up VMToCompiler.createResolvedJavaType
|
| Tue, 17 Dec 2013 21:25:27 -0800 |
twisti |
make CompilerToVM.resolveMethod return a metadata method instead of a Java type
|
| Tue, 17 Dec 2013 19:09:29 -0800 |
twisti |
moved most CompilerToVM.getUniqueConcreteMethod logic up to Java and replace getUniqueConcreteMethod with findUniqueConcreteMethod
|
| Tue, 17 Dec 2013 16:45:02 +0100 |
Doug Simon |
added -G:PrintCompRate option for periodically printing out the current compilation rate
|
| Thu, 12 Dec 2013 14:56:52 -0800 |
twisti |
moved most CompilerToVM.getLocalVariableTable to Java
|
| Wed, 11 Dec 2013 20:42:50 -0800 |
twisti |
moved most HotSpotResolvedJavaMethod.getExceptionHandlers logic to Java
|
| Wed, 11 Dec 2013 20:46:25 +0100 |
Doug Simon |
simplified implementation of HotSpotResolvedObjectType.getClassInitializer()
|
| Sun, 08 Dec 2013 21:55:09 -0800 |
twisti |
merged CompilerToVM.getMetaspaceConstructor into CompilerToVM.getMetaspaceMethod
|
| Sun, 08 Dec 2013 18:06:28 -0800 |
twisti |
removed CompilerToVM.getJavaField
|
| Sun, 08 Dec 2013 13:27:52 -0800 |
twisti |
don't go through VM to create HotSpotResolvedObjectType (part 1)
|
| Sun, 08 Dec 2013 11:21:49 -0800 |
twisti |
use Unsafe.ensureClassInitialized in HotSpotResolvedObjectType.initialize
|
| Sun, 08 Dec 2013 11:13:26 -0800 |
twisti |
every HotSpotResolvedJavaMethod needs its own constant pool
|
| Sat, 07 Dec 2013 17:11:49 -0800 |
twisti |
move HotSpotResolvedPrimitiveType's from VMToCompilerImpl to HotSpotGraalRuntime
|
| Thu, 05 Dec 2013 18:13:04 -0800 |
twisti |
Teach Graal about Symbol and ConstantPool so we can move more logic into Java. We'll see how that ends...
|
| Tue, 03 Dec 2013 14:48:09 -0800 |
twisti |
remove CompilerToVM.initializeMethodData
|
| Tue, 03 Dec 2013 18:19:56 +0100 |
Doug Simon |
Merge.
|
| Tue, 03 Dec 2013 18:03:39 +0100 |
Doug Simon |
added extra space to disassembler buffer to account for disassembly header
|
| Mon, 02 Dec 2013 20:05:24 -0800 |
twisti |
moved CompilerToVM.isTypeInitialized and isTypeLinked to Java
|
| Wed, 27 Nov 2013 02:31:03 +0100 |
Doug Simon |
fixed bug in MethodData and added assertion that precomputed size == initialized size
|
| Thu, 21 Nov 2013 22:08:04 +0100 |
Doug Simon |
moved notification of Graal compilation statistics from VMToCompiler to CompilerToVM
|
| Tue, 19 Nov 2013 01:31:19 +0100 |
Doug Simon |
made Graal report its compilation info under -XX:+CITime in the same format as c1 and c2
|
| Wed, 23 Oct 2013 19:50:14 -0700 |
twisti |
read HotSpotVMConfig fields from HotSpot's vmStructs via annotations
|
| Wed, 23 Oct 2013 13:40:56 +0200 |
Matthias Grimmer |
rename CompilerToVm.dontInline to doNotInlineOrCompile
|
| Wed, 23 Oct 2013 12:54:21 +0200 |
Matthias Grimmer |
Remove unused parameter from CompilerToVm.dontInline
|
| Wed, 23 Oct 2013 12:48:06 +0200 |
Matthias Grimmer |
Extend the CompilerToVM interface to explicitly avoid compilation and inlining of methods
|
| Tue, 15 Oct 2013 01:11:02 +0200 |
Thomas Wuerthinger |
Make sure HotSpot installed code link is cut when code is explicitely invalidated.
|
| Fri, 11 Oct 2013 15:41:33 +0200 |
Gilles Duboscq |
Reduce HotSpot diff and fix previous merge
|
| Mon, 30 Sep 2013 18:32:26 +0200 |
Lukas Stadler |
move benchmark counters into separate class and make them correct for multithreaded applications
|
| Tue, 03 Sep 2013 18:09:02 +0200 |
Gilles Duboscq |
Add isLinked method to ResolvedJavaType
|
| Tue, 03 Sep 2013 18:02:29 +0200 |
Gilles Duboscq |
Make is_rewritten a guarantee in initializeBytecode
|
| Tue, 03 Sep 2013 14:14:09 +0200 |
Gilles Duboscq |
Fix graalCompilerToVM::initializeBytecode regarding bytecode rewritting for methods that are not rewritten.
|
| Tue, 20 Aug 2013 12:48:28 -0700 |
twisti |
HotSpotInstalledCode: moved getCode to Java and added the ability to get the whole blob
|
| Tue, 06 Aug 2013 23:36:36 +0200 |
Doug Simon |
add intrinsics for java.util.zip.CRC32 (GRAAL-385)
|
| Fri, 02 Aug 2013 19:03:05 +0200 |
Christos Kotselidis |
Add heap sanity checker with premature hard crash for debugging write barriers
|
| Tue, 30 Jul 2013 16:42:51 -0700 |
twisti |
adding back setting safepointPollingAddress
|
| Fri, 26 Jul 2013 14:03:07 -0700 |
twisti |
read HotSpot VM options via HotSpotDiagnosticMXBean
|
| Thu, 25 Jul 2013 22:15:30 -0400 |
Morris Meyer |
PTX support for Linux
|
| Fri, 12 Jul 2013 10:12:06 -0700 |
twisti |
Merge
|
