| Port printing of stack traces from SOMns |
|
More...
|
over 4 years ago
|
| Added config and CI task to quickly run AWFY to check compatibility |
|
More...
|
over 4 years ago
|
| We don’t build the libgraal-jdk so, need to force interpreter here |
|
More...
|
over 4 years ago
|
| Merge PR #59: Native image >21.1 black lists many operations from compilation |
|
More...
|
over 4 years ago
|
| Native image >21.1 black lists many operations from compilation |
|
More...
|
over 4 years ago
|
| Reduce code duplication for inlining optimizations and add whileFalse inlining |
|
More...
|
over 4 years ago
|
| Added loop count reporting |
|
More...
|
over 4 years ago
|
| Introduce the JUMP2 bytecodes, which use the second byte to encode longer jumps |
|
More...
|
over 4 years ago
|
| Reduce code duplication in inlineInto |
|
More...
|
over 4 years ago
|
| Give BytecodeLoopNode a toString() |
|
More...
|
over 4 years ago
|
| The inlining guard should be set longer, because POP otherwise optimizes away jump targets |
|
More...
|
over 4 years ago
|
| Inline #whileTrue: in bytecodes, when it has two literal blocks |
|
More...
|
over 4 years ago
|
| Be more explicit about bytecodes and put safeguard in |
|
More...
|
over 4 years ago
|
| Merge PR #58: Add bytecode-level inlining of ifTrue/ifFalse* messages |
|
More...
|
over 4 years ago
|
| Set holder for new block methods |
|
More...
|
over 4 years ago
|
| Make sure we only adapt the context level, when there was inlining |
|
More...
|
over 4 years ago
|
| Exclude all blocks from “trivial” optimization to avoid interference with block inlining |
|
More...
|
over 4 years ago
|
| Merge PR #57: Optimize access, and other trivial methods |
|
More...
|
over 4 years ago
|
| Implement optimizations for trivial methods in the bytecode method gen |
|
More...
|
over 4 years ago
|
| Move the WriteAndReturnSelf node into FieldNode |
|
More...
|
over 4 years ago
|
| Change GlobalNode to initialize source section like other nodes |
|
More...
|
over 4 years ago
|
| [WIP] Start bytecode support |
|
More...
|
over 4 years ago
|
| Generalize the dispatch chain embedding to trivial methods |
|
More...
|
over 4 years ago
|
| Turn assertion about jump length into proper ParseError |
|
More...
|
over 4 years ago
|
| Update core-lib: small SomSom improvement |
|
More...
|
over 4 years ago
|
| Merge PR #56: Quicken Super Sends and optimize Object Instantiation |
|
More...
|
over 4 years ago
|
| Added a toString to uninitialized message send nodes |
|
More...
|
over 4 years ago
|
| Use dispatch chain to get lexical specialization of new |
|
More...
|
over 4 years ago
|
| Construct methods/primitives directly instead of going through the universe |
|
More...
|
over 4 years ago
|
| Cache object layout to optimize creation of new object instances |
|
More...
|
over 4 years ago
|