Thu, 12 Mar 2015 16:04:37 +0100 |
Doug Simon |
fixed Math.pow intrinsic to make a runtime call instead of calling the original method
|
Thu, 12 Mar 2015 16:03:40 +0100 |
Doug Simon |
disallow intrinsics from calling the original method
|
Thu, 12 Mar 2015 15:25:43 +0100 |
Doug Simon |
fixed bug in managing replacement scope during bytecode parsing and improved API for querying replacement info from graph builder plugins
|
Thu, 12 Mar 2015 15:20:29 +0100 |
Doug Simon |
added test for Math.pow() intrinsic
|
Thu, 12 Mar 2015 15:19:56 +0100 |
Doug Simon |
tidied up assertions in InliningUtil
|
Thu, 12 Mar 2015 14:11:03 +0100 |
Doug Simon |
moved bytecode paring options to AbstractBytecodeParser.Options
|
Thu, 12 Mar 2015 12:45:04 +0100 |
Gilles Duboscq |
mx: Support systems where SC_ARG_MAX has not defined limit
|
Thu, 12 Mar 2015 07:29:31 +0100 |
Doug Simon |
fixed canonicalizeprojects issues
|
Thu, 12 Mar 2015 07:14:18 +0100 |
Doug Simon |
Merge.
|
Wed, 11 Mar 2015 17:45:21 -0700 |
Tom Rodriguez |
Split LeafType off from ConcreteSubtype
|
Wed, 11 Mar 2015 17:32:04 -0700 |
Tom Rodriguez |
Connect required Assumptions with answer to CHA query
|
Wed, 11 Mar 2015 16:06:35 -0700 |
Tom Rodriguez |
Simplify STATIC_PRIMITIVE_FIELD macro
|
Thu, 12 Mar 2015 07:13:05 +0100 |
Doug Simon |
moved WordTypes from c.o.g.word.phases to c.o.g.word
|
Thu, 12 Mar 2015 01:11:22 +0100 |
Doug Simon |
removed WordTypeRewriterPhase
|
Wed, 11 Mar 2015 23:36:38 +0100 |
Doug Simon |
use the 'jacoco=exclude' project property when deciding which directories to include in JaCoCo coverage reports
|
Wed, 11 Mar 2015 23:27:51 +0100 |
Doug Simon |
removed some uses of setStampFromReturnType
|
Wed, 11 Mar 2015 23:23:08 +0100 |
Doug Simon |
ensure voidness of a @NodeIntrinsic method's return kind agrees with the voidness of the stamp of the intrinsified node
|
Wed, 11 Mar 2015 21:01:10 +0100 |
Doug Simon |
workaround for JDK-8056066
|
Wed, 11 Mar 2015 20:43:52 +0100 |
Doug Simon |
Merge.
|
Wed, 11 Mar 2015 18:16:25 +0100 |
Lukas Stadler |
make sure that all split successors guard for a condition before optimizing in OptimizeGuardAnchorsPhase
|
Wed, 11 Mar 2015 16:35:26 +0100 |
Lukas Stadler |
iterative marking of loop phis in SchedulePhase
|
Wed, 11 Mar 2015 15:44:32 +0100 |
Christian Humer |
Truffle: implement recursive inlining and with a maximum depth of 4.
|
Wed, 11 Mar 2015 15:44:32 +0100 |
Christian Humer |
Fix broken windows build.
|
Wed, 11 Mar 2015 15:44:32 +0100 |
Christian Humer |
Truffle-DSL: @Cached javadoc fixes.
|
Wed, 11 Mar 2015 11:32:13 +0100 |
Christian Humer |
Truffle-DSL: make the whole specialization block atomic to avoid unnecessary node allocations.
|
Wed, 11 Mar 2015 11:30:20 +0100 |
Christian Humer |
Truffle-DSL: make specialization types mergable with other specialization types.
|
Tue, 10 Mar 2015 21:13:07 +0100 |
Christian Humer |
Truffle-DSL: removeSame does not need to run using atomic. It is always used in an atomic block already.
|
Tue, 10 Mar 2015 21:11:12 +0100 |
Christian Humer |
Truffle: add assertion to always be in atomic blocks for replaces as it now may be violated by NodeUtil#nonAtomicReplace.
|