The Eclipse OMR project is a set of open source C and C++ components that can be used to build robust language runtimes that support many different hardware and operating system platforms.
Our current components are:
gc: Garbage collection framework for managed heapscompiler: Components for builing compiler technology, such as JIT compilers.jitbuilder: An easy to use high level abstraction on top of the compiler technology.port: Platform porting librarythread: A cross platform pthread-like threading libraryutil: general utilities useful for building cross platform runtimesomrsigcompat: Signal handling compatibility libraryomrtrace: Tracing library for communication with IBM Health Center monitoring toolstool: Code generation tools for the build systemvm: APIs to manage per-interpreter and per-thread contextsexample: Demonstration code to show how a language runtime might consume some Eclipse OMR componentsfvtest: A language-independent test framework so that Eclipse OMR components can be tested outside of a language runtime
The long term goal for the Eclipse OMR project is to foster an open ecosystem of language runtime developers to collaborate and collectively innovate with hardware platform designers, operating system developers, as well as tool and framework developers and to provide a robust runtime technology platform so that language implementers can much more quickly and easily create more fully featured languages to enrich the options available to programmers.
It is our community's fervent goal to be one of active contribution, improvement, and continual consumption.
We will be operating under the Eclipse Code of Conduct to promote fairness, openness, and inclusion.
- The most comprehensive consumer of the Eclipse OMR technology is the IBM J9 Virtual Machine: a high performance, scalable, enterprise class Java Virtual Machine implementation representing hundreds of person years of effort, built using the core implementations provided by Eclipse OMR. IBM is working actively to open source J9.
- The Ruby+OMR Technology Preview has used Eclipse OMR components to add a JIT compiler to the CRuby implementation, and to experiment with replacing the garbage collector in CRuby.
- A SOM++ Smalltalk runtime has also been modified to use Eclipse OMR componentry.
- An experimental version of CPython using Eclipse OMR components has also been created but is not yet available in the open. (Our focus has been dominated by getting this code out into the open!)
All Eclipse OMR project materials are made available under the Eclipse Public License V1.0 and the Apache 2.0 license. You can choose which license you wish to follow. Please see our LICENSE file for more details.
There are some active contribution projects underway right now:
- Documentation: code comments are great, but we need more overview documentation so we're writing that
- FAQ: Frequently Asked Questions from real people's questions (request: ask questions!)
- Beginner issues: relatively simple but useful work items meant for people new to the project.
diag: more diagnostic support for language runtimes to aid developers and usershcagent: the core code for the IBM Health Center agent for interfacing to a runtimegc: adding generational and other GC policies
The best way to get an initial understanding of the Eclipse OMR technology is to look at a 'standalone' build, which hooks Eclipse OMR up to the its testing system only.
To build standalone Eclipse OMR, run the following commands from the top of the source tree. The top of the Eclipse OMR source tree is the directory that contains run_configure.mk.
# Generate autotools makefiles with SPEC-specific presets
make -f run_configure.mk SPEC=linux_x86-64 OMRGLUE=./example/glue
# Build
make
# Run tests (note that no contribution should cause new test failures in "make test")
make test
Run make -f run_configure.mk help for a list of configure makefile targets.
Run make help for a list of build targets.
A shell script interpreter, such as bash, is required to run configure.
Run ./configure --help to see the full list of configure command-line
options.
To run configure using both SPEC presets and custom options, pass the
EXTRA_CONFIGURE_ARGS option to run_configure.mk.
For example, to disable optimizations, run configure like this:
# Example configure
make -f run_configure.mk SPEC=linux_x86-64 OMRGLUE=./example/glue 'EXTRA_CONFIGURE_ARGS=--disable-optimized' clean all
To disable building fvtests, run configure like this:
# Example configure disabling fvtests
make -f run_configure.mk SPEC=linux_x86-64 OMRGLUE=./example/glue 'EXTRA_CONFIGURE_ARGS=--disable-fvtest' clean all
Note that the clean target of run_configure.mk deletes the header files and
makefiles generated by configure. Invoking the clean all targets ensures that
the header files and makefiles are regenerated using the custom options.
The minimal invocation of configure is:
# Basic configure example
$ ./configure OMRGLUE=./example/glue
-
Mark Stoodley's talk at the JVM Languages Summit in August, 2015: A VM is a VM is a VM: The Secret Path to High Performance Multi-Language Runtimes
-
Daryl Maier's slides from Java One in October, 2015: Beyond the Coffee Cup: Leveraging Java Runtime Technologies for the Polyglot
-
Charlie Gracie's slides from Java One in October, 2015: What's in an Object? Java Garbage Collection for the Polyglot
-
Angela Lin, Robert Young, Craig Lehmann and Xiaoli Liang CASCON Workshop in November, 2015 Building Your Own Runtime
- Note: these slides have been modified since the original presentation to use the Eclipse OMR project name
-
Charlie Gracie's talk from FOSDEM in February, 2016: Ruby and OMR: Experiments in utilizing OMR technologies in MRI
-
Charlie Gracie's slides from jFokus in February, 2016 A JVMs Journey into Polyglot Runtimes
-
Mark Stoodley's slides from EclipseCON in March, 2016 Eclipse OMR: a modern toolkit for building language runtimes