We're hiring!
*

Run your own CI pipeline with GStreamer's new monorepo

Xavier Claessens avatar

Xavier Claessens
October 26, 2021

Share this post:

Reading time:

Recently, the GStreamer project merged all its git repositories into a single, unified repository, often called monorepo. You can read more about this change here.

One benefit is it greatly simplifies maintaining custom, project specific, GStreamer patches. Previously, projects that needed to develop, or backport, some patches had to go through multiple steps to complete the task. Projects had to fork various git repositories (e.g. gst-plugins-good, gst-plugins-bad, etc), each repository would then have a new branch with the extra commits, and often, gst-build was used to pull all of these repositories together (and gst-build itself had to be patched beforehand to download forked repositories). Thankfully, all that will be a thing of the past.

When maintaining a non-trivial set of patches, it's always a good idea to also run a continuous integration (CI) pipeline. Prior to moving to a single repository, GStreamer's public CI had a complicated manifest mechanism to be able to pull branches from different repositories, and test them together. Replicating that system with a private GitLab instance was close to impossible.

With GStreamer monorepo, all those patches can now be maintained in a single git repository, in a single branch. For example, if you are forking GStreamer's main branch into a private GitLab instance, the CI will now work almost out of the box! Your team can create Merge Requests in your private GitLab instance, and develop features while relying on your own CI. Of course, once the feature is ready, it often makes sense to upstream. A single Merge Request in the public repository and you're good to go.

How does it work?

1. Clone the GStreamer repository:

git clone https://gitlab.freedesktop.org/gstreamer/gstreamer.git

2. Push the main branch to your private GitLab instance:

git remote add private https://gitlab.example.com/my-project/gstreamer.git
git push private main

3. Setup your new GStreamer repository:

  • Navigate to https://gitlab.example.com/my-project/gstreamer/edit
  • Under Visibility, project features, permissions make sure to enable CI/CD and Container registry.

4. Register your GitLab runners:

  • Navigate to https://gitlab.example.com/my-project/gstreamer/-/settings/ci_cd
  • Under Runners you'll find the registration token to register your gitlab-runner.

That's it! You can now run the same CI pipeline as on the public GitLab instance. Every member of your team can create their fork of that repositry into their own account and create internal Merge Requests.

Also, you'll notice that your CI will rebuild the needed Linux and Windows docker images thanks to Freedesktop's ci-template.

Known issues

Not everything is that easy however...

Cerbero trigger job fails

GitLab will try to trigger Cerbero's CI into the public repository, which will fail because your access token from your private GitLab instance won't be accepted by another GitLab instance.

You can either edit .gitlab-ci.yml and delete the cerbero trigger job, or you can also fork cerbero into our private GitLab instance. Note that if you don't fork cerbero under gstreamer/cerbero, you will have to edit ci/gitlab/trigger_cerbero_pipeline.py and change to CERBERO_PROJECT = 'my-project/cerbero'.

I need a Windows runner

You can delete from .gitlab-ci.yml jobs that require Windows, or you can setup your own runner.

  1. Install Windows on a powerful machine Either latest Windows 10 (not a VM), or Windows Server 1809 VM.
  2. Install Docker Desktop https://docs.docker.com/desktop/windows/install/ Make sure to switch it to run Windows Containers instead of Linux Containers.
  3. Install gitlab-runner https://docs.gitlab.com/runner/install/windows.html When registering your runner, make sure to set tags: windows,1809,shell,docker.

Conclusion

This unification of Git repositories will greatly simply the workflow for all contributors and distributors. We are really looking forward to seeing it in action in the next major GStreamer release. With the feature freeze having begun on October 18th, we should hopefully see a final release of GStreamer 1.20 before the end of the year. As usual, if you need any assistance with GStreamer, please don't hesitate to contact us, we're here to help!

Comments (0)


Add a Comment






Allowed tags: <b><i><br>Add a new comment:


Search the newsroom

Latest Blog Posts

Faster inference: torch.compile vs TensorRT

19/12/2024

In the world of deep learning optimization, two powerful tools stand out: torch.compile, PyTorch’s just-in-time (JIT) compiler, and NVIDIA’s…

Mesa CI and the power of pre-merge testing

08/10/2024

Having multiple developers work on pre-merge testing distributes the process and ensures that every contribution is rigorously tested before…

A shifty tale about unit testing with Maxwell, NVK's backend compiler

15/08/2024

After rigorous debugging, a new unit testing framework was added to the backend compiler for NVK. This is a walkthrough of the steps taken…

A journey towards reliable testing in the Linux Kernel

01/08/2024

We're reflecting on the steps taken as we continually seek to improve Linux kernel integration. This will include more detail about the…

Building a Board Farm for Embedded World

27/06/2024

With each board running a mainline-first Linux software stack and tested in a CI loop with the LAVA test framework, the Farm showcased Collabora's…

Smart audio filters with WirePlumber 0.5

26/06/2024

WirePlumber 0.5 arrived recently with many new and essential features including the Smart Filter Policy, enabling audio filters to automatically…

Open Since 2005 logo

Our website only uses a strictly necessary session cookie provided by our CMS system. To find out more please follow this link.

Collabora Limited © 2005-2024. All rights reserved. Privacy Notice. Sitemap.