Thanks for your interest in this project.
We welcome and encourage all kinds of contribution to the project, not only code. This includes bug reports, user experience feedback, assistance in reproducing issues and more.
If you are new to working with git, you may use our Git Crash Course
to help you get started.
This project uses GitHub Issues to track ongoing development, discuss project plans, and keep track of bugs. Be sure to search for existing issues before you create another one.
Visit our Issues page on GitHub to search and submit
Our GitHub issues get labelled by committers in order to allow easier navigation, and to flag certain issues as being of interest to certain groups. A PR or Issue may have multiple labels, as many as needed to provide adequate categorization.
A subset of the labels are documented below.
-
good first issuegenerally refers to a task that would be suitable for someone new to the project with little experience with either the technology or even open-source projects and processes. They are intended for anyone who wants to gently get their feet wet on building, testing, or familiarizing themselves with part of the code base. -
bugissues are functional problems, errors or unexpected behaviour. -
build/configurelabels are used to describe issues with the build and configure system (e.g., makefiles, autotools configure). -
cilabels are used for issues and enhancements with the continuous integration system for pull request testing (e.g., Jenkins, Azure, etc.) -
cmakelabels are similar to build/configure but they apply specifically to the CMake configuration. -
compiler arch reviewlabels are used to indicate a review of this issue or pull request at the OMR Compiler Architecture meeting is requested prior to committing. -
documentationlabels are used for issues or enhancements to the documentation (either in the source code itself or stand-alone documentation files). -
epicissues are used to group together related issues and to track larger goals in the project across issues. -
GSoC Projectlabels are for potential ideas for Google Summer Of Code projects. -
help wantedissues have value to the project but no immediate human resources to undertake it. Those that are looking to complete a task that someone isn't already working on can consider these. -
licenselabels are used to annotate issues concerning the source code license. -
meetinglabels are used to annotate issues pertaining to project meeting agendas or minutes. -
toolchain buglabels are used to document issues or pull requests that describe or implement a workaround to a bug in the development toolchain (such as the compiler) used to build OMR. Toolchain workarounds should be temporary in nature, and the intention of this label is to make such workarounds easy to discover in the future lest they be lost in the code. -
toolinglabels are for issues concerning support tooling needed to support any of the code or processes within the project.
Labels are also used to classify issues and pull requests by the dominant Eclipse OMR technology component they affect. For instance,
| Label | Component | Principal Directories |
|---|---|---|
comp:compiler |
Compiler | compiler |
comp:core |
Core OMR functionality | include_code, omr |
comp:diagnostic |
Diagnostic services | ddr |
comp:doc |
OMR documentation | doc |
comp:gc |
Garbage collector | gc |
comp:glue |
Glue code | glue |
comp:jitbuilder |
JitBuilder | jitbuilder |
comp:port |
Port library | port |
comp:test |
Unit tests and testing framework | fvtest |
comp:thread |
Thread library | thread |
comp:tril |
Tril infrastructure and tests | fvtest/tril |
comp:utilities |
OMR utilities | util |
Further classification by processor architecture, operating system, and bitness can be achieved with the following labels:
You can propose contributions by sending pull requests through GitHub. Following these guidelines will help us to merge your pull requests smoothly:
-
Eclipse OMR is an Eclipse Foundation project. All contributors must sign the Eclipse Contributor Agreement (or ECA) following the instructions on that page before a contribution is accepted.
See Eclipse OMR Legal Considerations for additional ECA considerations when creating project commits as well as copyright and license requirements for files.
-
If you're not sure your contribution would be accepted, and want to validate your approach or idea before writing code, feel free to open an issue. However, not every feature or fix needs an issue. If the problem and fix are cleanly connected, and you have the fix in hand, feel free to just submit a pull request.
-
Your pull request is an opportunity to explain both what changes you'd like pulled in, but also why you'd like them added. Providing clarity on why you want changes makes it easier to accept, and provides valuable context to review.
-
Follow the coding style and format of the code you are modifying (see the coding standards). The code base is yet to be unified in style however, so if the file you are editing seems to have a diffferent style, defer to the style of the file as you found it.
-
Only use C++ language features supported by our compilers. A list of supported features can be found here.
-
Follow the commit guidelines described below.
-
Contributions from generative artificial intelligence (GenAI) coding assistants are permitted subject to the GenAI Usage Guidelines described below.
-
We encourage you to open a pull request early, and mark it as "Work In Progress" (prefix the PR title with WIP). This allows feedback to start early, and helps create a better end product. Committers will wait until after you've removed the WIP prefix to merge your changes.
-
If you are contributing a change to the compiler technology that involves modifications to the Testarossa intermediate language (IL) (including, but not limited to, adding a new IL opcode, changing the properties of an opcode, or adding a new datatype) or, in the opinion of a committer, a fundamental element of compiler infrastructure, a committer will request that this pull request be presented at an upcoming OMR Compiler Architecture meeting to invite community discussion prior to merging. Issues of the same nature may also be asked to be discussed before the same architecture meeting prior to producing a pull request.
If a downstream project consuming OMR has a dependency on the contents of your pull request, i.e. the pull request should not be merged until all downstream preparations have been complete, you as the contributor of the pull request may use the WIP prefix in addition to comment in the said pull request to inform the committer of your particular circumstances for merging. That is, you may request a review of the pull request even though it's marked WIP, informing the committer of your desire to coordinate when the PR is merged. Once all downstream preparations have been complete it is your responsibility as the contributor to remove the WIP prefix and inform the committers that the pull request is ready for merging, pending reviews have been completed.
The first line describes the change made. It is written in the imperative mood, and should say what happens when the patch is applied. Keep it short and simple. The first line should be less than 70 characters, where reasonable, and should be written in sentence case preferably not ending in a period. Leave a blank line between the first line and the message body.
The body should be wrapped at 72 characters, where reasonable.
Include as much information in your commit as possible. You may want to include designs and rationale, examples and code, or issues and next steps. Prefer copying resources into the body of the commit over providing external links. Structure large commit messages with headers, references etc. Remember however that the commit message is always going to be rendered in plain text.
Please add [skip ci] to the commit message when the change doesn't require a
compilation, such as documentation only changes, to avoid unnecessarily wasting
the project's build resources.
Use the commit footer to place commit metadata. The footer is the last block of contiguous text in the message. It is separated from the body by one or more blank lines, and as such cannot contain any blank lines. Lines in the footer are of the form:
Key: Value
When a commit has related issues or commits, explain the relation in the message
body. You should also leave an Issue tag in the footer. However, if this is the
final commit that fixes an issue you should leave a Closes tag (or one of the
tags described here
in the footer instead which will automatically close the referenced issue when
this pull request is merged.
For example, if this is just one of the commits necessary to address Issue 1234 then the following is a valid commit message:
Correct race in frobnicator
This patch eliminates the race condition in issue #1234.
Issue: #1234
However, if this is the final commit that addresses the issue then the following is an acceptable commit message:
Correct race in frobnicator
This patch eliminates the race condition in issue #1234.
Closes: #1234
Be sure to author commits with the same email address that you used to sign the Eclipse Contributor Agreement.
If there are multiple authors of a commit they should all be credited
with a Co-authored-by: Full Name <email address> line at the footer
of each commit. Each co-author must sign the ECA.
Each time a commit is pushed, a GitHub CI check will be run automatically that verifies that the commit author(s) has signed the Eclipse Contributor Agreement. If not, the check will fail and the result will be visible in the pull request.
Here is an example of a good commit:
Update and expand the commit guidelines
Elaborate on the style guidelines for commit messages. These new
style guidelines reflect the conversation found in #124.
The guidelines are changed to:
- Provide guidance on how to write a good first line.
- Elaborate on formatting requirements.
- Relax the advice on using issues for nontrivial commits.
- Move issue references from the first line to the message footer.
- Encourage contributors to put more information into the commit
message.
Issue: #124
The first line is meaningful and imperative. The body contains enough information that the reader understands the why and how of the commit, and its relation to any issues. The issue is properly tagged.
The following is an example of a bad commit:
FIX #124: Changing a couple random things in CONTRIBUTING.md.
Also, there are some bug fixes in the thread library.
The commit rolls unrelated changes together in a very bad way. There is not enough information for the commit message to be useful. The first line is not meaningful or imperative. The message is not formatted correctly and the issue is improperly referenced.
We have a script that attempts to enforce some of these commit guidelines. You can install it by following instructions similar to the ones below.
cd omr_repo_dir
cp scripts/commit-msg .git/hooks
chmod +x .git/hooks/commit-msg
If the hook declines your commit, the message will remain in
omr_repo_dir/.git/COMMIT_EDITMSG.
Be sure to update your version of the script occasionally as it may evolve as our commit guidelines evolve.
The Eclipse OMR project acknowledges that some contributions may be authored in whole or in part from generative artificial intelligence coding assistants. Such contributions are permitted per the Eclipse Foundation Generative AI Usage Guidelines which describes contributor and committer responsibilities, copyright and licensing considerations, and attribution requirements for GenAI assisted contributions.
All contributors are encouraged to read and understand the advice and guidelines therein before submitting a contribution to Eclipse OMR even partially generated by an AI coding assistant.
The Eclipse OMR project is subject to those policies and provides additional guidelines described below for contributors and committers governing GenAI-assisted contributions.
Contributions that include GenAI-assisted content must disclose that authorship in each file where GenAI content is present. Contributors must include a statement like the following just below the copyright header at the beginning of the file that acknowledges their AI assistant (assuming such a statement does not already exist).
// Some sections generated by <YOUR AI ASSISTANT>
In addition, individual commits that include GenAI authored content must attribute that at the end of the commit message (before any human authorship signatures).
Co-authored by <YOUR AI ASSISTANT>
Some contributors' employers may require their AI-generated code and documentation contributions to be annotated with inline tags providing additional metadata about the generated content for legal and license attribution reasons. Eclipse OMR will accept contributions that include such metadata with some guidance below for contributors and committers.
Inline tagging of code or documentation generated by coding assistants is not mandatory.
If tags are used, they must clearly demarcate the boundaries of GenAI-assisted content (e.g., via paired "begin" and "end" code comments). These boundaries may span consecutive lines, functions, paragraphs, or entire files. The term "assisted" is an important qualifier to describe the enclosed content because it may contain a blend of AI-generated content and human contributions. The "assisted" distinction is necessary because it may not be possible to clearly identify just the AI-generated content in the contribution, and it allows for human changes to the demarcated contribution over time while preserving the meaning of the tags.
At this time, Eclipse OMR does not specify the format of the tags, except that they must not contain any corporate-specific information or other identifiers that have no meaning beyond the original contributor (e.g., internal issue numbers or URLs). Information about the model used to generate the code or documentation is permitted, however.
Tags must be concise and as minimally-invasive as possible. Tags that are excessively verbose or impact the readability of the code or documentation may be rejected by committers at their discretion.
For consistency, tags produced for the same coding assistant must use the same format regardless of the contributor or pull request.
If the contents of a tagged region are modified by a human then any enclosing tags should remain in place.
If code or documentation within a tagged region is copied in whole or in part elsewhere in the project, the original surrounding tags should accompany the copy to its destination. If such code or documentation is copied to a file that does not already have the Eclipse Foundation GenAI attribution after the copyright header, such attribution must be added.
Eclipse OMR implements a good-faith policy for contributors and committers to maintain GenAI tags in the code and documentation for as long as possible. Existing tags and their boundaries will be maintained on a best-effort basis, and they should not be removed simply for the sake of removing them.
However, there are legitimate situations where GenAI tagging may be considered for removal, including if the tags:
- interfere with the readability or understandability of the code or documentation,
- become difficult or unwieldy to maintain (e.g., through code refactoring or code sharing),
- become obsolete (e.g., information within a tagged section is removed entirely),
- become redundant (e.g., an inner set of tags is completely consumed by an outer set)
This list is not exhaustive, and the disposition of GenAI tags in a pull request will ultimately be at the discretion of Eclipse OMR committers who will balance the spirit of these tagging guidelines with the best interests of the Eclipse OMR project.
When a pull request is created (or when new commits are pushed to an existing pull request), a webservice will automatically run that will verify that each commit author's email address matches an email address that signed the ECA. This service is case-sensitive, so the case of the email addresses must match exactly. The success or failure of this webservice will be reported in the GitHub pull request. Pull requests with a failing ECA check will not be merged.
Eclipse OMR pull requests do not need to be signed, but most contributors do so as a best practice.
It is the responsibility of each contributor to obtain legal advice, and to ensure that their contributions fulfill the legal requirements of their organization. This document is not legal advice.
OMR is dual-licensed under the Eclipse Public License 2.0 and the Apache License v2.0. Any previously unlicensed contribution should be released under the same license.
- If you wish to contribute code under a different license, you must consult with a committer before contributing.
- For any scenario not covered by this document, please discuss the copyright notice and licensing requirements with a committer before contributing.
The template for the copyright notice and dual-license is as follows:
/*******************************************************************************
* Copyright ${author} and others 2017
*
* This program and the accompanying materials are made available under
* the terms of the Eclipse Public License 2.0 which accompanies this
* distribution and is available at https://www.eclipse.org/legal/epl-2.0/
* or the Apache License, Version 2.0 which accompanies this distribution and
* is available at https://www.apache.org/licenses/LICENSE-2.0.
*
* This Source Code may also be made available under the following
* Secondary Licenses when the conditions for such availability set
* forth in the Eclipse Public License, v. 2.0 are satisfied: GNU
* General Public License, version 2 with the GNU Classpath
* Exception [1] and GNU General Public License, version 2 with the
* OpenJDK Assembly Exception [2].
*
* [1] https://www.gnu.org/software/classpath/license.html
* [2] https://openjdk.org/legal/assembly-exception.html
*
* SPDX-License-Identifier: EPL-2.0 OR Apache-2.0 OR GPL-2.0-only WITH Classpath-exception-2.0 OR GPL-2.0-only WITH OpenJDK-assembly-exception-1.0
*******************************************************************************/