Contributing to libinput

So you want to contribute to libinput? Great! We’d love to help you be a part of our community. Here is some important information to help you.

Code of Conduct

As a freedesktop.org project, libinput follows the freedesktop.org Contributor Covenant.

Please conduct yourself in a respectful and civilised manner when interacting with community members on mailing lists, IRC, or bug trackers. The community represents the project as a whole, and abusive or bullying behaviour is not tolerated by the project.

Contact

Questions can be asked on #wayland on freenode or on the wayland-devel@lists.freedesktop.org mailing list.

For IRC, ping user whot (Peter Hutterer, the libinput maintainer) though note that he lives on UTC+10 and thus the rest of the world is out of sync by default ;)

For anything that appears to be device specific and/or related to a new feature, just file an issue in our issue tracker. It’s usually the most efficient way to get answers.

What to work on?

If you don’t already know what you want to improve or fix with libinput, then a good way of finding something is to search for the help needed tag in our issue tracker. These are issues that have been triaged to some degree and deemed to be a possible future feature to libinput.

Note

Some of these issue may require specific hardware to reproduce.

Another good place to help out with is the documentation. For anything you find in these pages that isn’t clear enough please feel free to reword it and add what is missing.

Getting the code

The libinput build instructions have all the details but the short solution will be:

$> git clone https://gitlab.freedesktop.org/libinput/libinput
$> cd libinput
$> meson --prefix=/usr builddir/
$> ninja -C builddir/
$> sudo ninja -C builddir/ install

You can omit the last step if you only want to test locally.

Working on the code

libinput has a roughly three-parts architecture:

  • the front-end code which handles the libinput_some_function() API calls in libinput.c

  • the generic evdev interface handling which maps those API calls to the backend calls (evdev.c).

  • there are device-specific backends which do most of the actual work - evdev-mt-touchpad.c is the one for touchpads for example.

In general, things that only affect the internal workings of a device only get implemented in the device-specific backend. You only need to touch the API when you are adding configuration options. For more details, please read the libinput’s internal architecture document. There’s also a blog post describing the building blocks that may help to understand how it all fits together.

Documentation is in /doc/api for the doxygen-generated API documentation. These are extracted from the libinput source code directly. The documentation you’re reading right now is in /doc/user and generated with sphinx. Simply running ninja -C builddir will rebuild it and the final product ends up in builddir/Documentation.

Testing the code

libinput provides a bunch of Helper tools to debug any changes - without having to install libinput.

The two most useful ones are libinput debug-events and libinput debug-gui. Both tools can be run from the build directory directly and are great for quick test iterations:

$> sudo ./builddir/libinput-debug-events --verbose
$> sudo ./builddir/libinput-debug-gui --verbose

The former provides purely textual output and is useful for verifying event streams from buttons, etc. The latter is particularly useful when you are trying to debug pointer movement or placement. libinput debug-gui will also visualize the raw data from the device so you can compare pointer behavior with what comes from the kernel.

These tools create a new libinput context and will not affect your session’s behavior. Only once you’ve installed libinput and restarted your session will your changes affect the X server/Wayland compositor.

Once everything seems to be correct, it’s time to run the libinput test suite:

$> sudo ./builddir/libinput-test-suite

This test suite can take test names etc. as arguments, have a look at libinput test suite for more info. There are a bunch of other tests that are run by the CI on merge requests, you can run those locally with

$> sudo ninja -C builddir check

So it always pays to run that before submitting. This will also run the code through valgrind and pick up any memory leaks.

Submitting Code

Any patches should be sent via a Merge Request (see the GitLab docs) in the libinput GitLab instance hosted by freedesktop.org.

Below are the steps required to submit a merge request. They do not replace learning git but they should be sufficient to make some of the more confusing steps obvious.

  • Register an account in the freedesktop.org GitLab instance.

  • Fork libinput into your username’s namespace

  • Get libinput’s main repository. git will call this repository origin.

    git clone https://gitlab.freedesktop.org/libinput/libinput.git
    
  • Add the forked git repository to your remotes (replace USERNAME with your username). git will call this repository gitlab.

    cd /path/to/libinput.git
    git remote add gitlab git@gitlab.freedesktop.org:USERNAME/libinput.git
    git fetch gitlab
    
  • Create a new branch and commit your changes to that branch.

    git switch -C mynewbranch
    # edit files, make changes
    git add file1 file2
    git commit -s
    # edit commit message in the editor
    

    Replace mynewbranch (here and in the commands below) with a meaningful name. See Commit Messages for details on the commit message format.

  • Push your changes to your fork and submit a merge request

    git push gitlab mynewbranch
    

    This command will print the URL to file a merge request, you then only have to click through. Alternatively you can go to:

    Select your branch name to merge and libinput/libinput master as target branch.

  • Verify that the CI completes successfully by visiting the merge request page. A successful pipeline shows only green ticks, failure is indicated by a red cross or a yellow exclamation mark (see the GitLab Docs). For details about the failures, click on the failed jobs in the pipelines and/or click the Expand button in the box for the test summaries.

    A merge request without a successful pipeline may never be looked at by a maintainer.

  • If changes are requested by the maintainers, please amend the commit(s) and force-push the updated branch.

    # edits in file foo.c
    git add foo.c
    git commit --amend
    git push -f gitlab mynewbranch
    

    A force-push will re-trigger the CI and notify the merge request that new changes are available.

    If the branch contains more than one commit, please look at git interactive rebases to learn how to change multiple commits, or squash new changes into older commits.

Commit History

libinput strives to have a linear, ‘recipe’ style history This means that every commit should be small, digestible, stand-alone, and functional. Rather than a purely chronological commit history like this:

doc: final docs for view transforms
fix tests when disabled, redo broken doc formatting
better transformed-view iteration (thanks Hannah!)
try to catch more cases in tests
tests: add new spline test
fix compilation on splines
doc: notes on reticulating splines
compositor: add spline reticulation for view transforms

We aim to have a clean history which only reflects the final state, broken up into functional groupings:

compositor: add spline reticulation for view transforms
compositor: new iterator for view transforms
tests: add view-transform correctness tests
doc: fix Doxygen formatting for view transforms

This ensures that the final patch series only contains the final state, without the changes and missteps taken along the development process.

The first line of a commit message should contain a prefix indicating what part is affected by the patch followed by one sentence that describes the change. For example:

touchpad: add software button behavior
fallback: disable button debouncing on device foo

If in doubt what prefix to use, look at other commits that change the same file(s) as the patch being sent.

Commit Messages

Commit messages must contain a Signed-off-by line with your name and email address. An example is:

A description of this commit, and it's great work.

Signed-off-by: Claire Someone <name@domain>

If you’re not the patch’s original author, you should also gather S-o-b’s by them (and/or whomever gave the patch to you.) The significance of this is that it certifies that you created the patch, that it was created under an appropriate open source license, or provided to you under those terms. This lets us indicate a chain of responsibility for the copyright status of the code. An example is:

A description of this commit, and it's great work.

Signed-off-by: Claire Someone <name@domain>
Signed-off-by: Ferris Crab <name@domain>

When you re-send patches, revised or not, it would be very good to document the changes compared to the previous revision in the commit message and/or the merge request. If you have already received Reviewed-by or Acked-by tags, you should evaluate whether they still apply and include them in the respective commit messages. Otherwise the tags may be lost, reviewers miss the credit they deserve, and the patches may cause redundant review effort.

For further reading, please see ‘on commit messages’ as a general guideline on what commit messages should contain.

Coding Style

Please see the CODING_STYLE.md document in the source tree.

Tracking patches and follow-ups

Once submitted to GitLab, your patches will be reviewed by the libinput development team on GitLab. Review may be entirely positive and result in your code landing instantly, in which case, great! You’re done. However, we may ask you to make some revisions: fixing some bugs we’ve noticed, working to a slightly different design, or adding documentation and tests.

If you do get asked to revise the patches, please bear in mind the notes above. You should use git rebase -i to make revisions, so that your patches follow the clear linear split documented above. Following that split makes it easier for reviewers to understand your work, and to verify that the code you’re submitting is correct.

A common request is to split single large patch into multiple patches. This can happen, for example, if when adding a new feature you notice a bug in libinput’s core which you need to fix to progress. Separating these changes into separate commits will allow us to verify and land the bugfix quickly, pushing part of your work for the good of everyone, whilst revision and discussion continues on the larger feature part. It also allows us to direct you towards reviewers who best understand the different areas you are working on.

When you have made any requested changes, please rebase the commits, verify that they still individually look good, then force-push your new branch to GitLab. This will update the merge request and notify everyone subscribed to your merge request, so they can review it again.

There are also many GitLab CLI clients, if you prefer to avoid the web interface. It may be difficult to follow review comments without using the web interface though, so we do recommend using this to go through the review process, even if you use other clients to track the list of available patches.