13 libc++ uses Buildkite for running its CI. Setting up new CI jobs is easy, and
14 these jobs can run either on our existing infrastructure, or on your own.
16 If you need to run the job on your own machines, please follow the
17 `Buildkite guide <https://buildkite.com/docs/agent/v3>`_ to setup your
18 own agents. Make sure you tag your agents in a way that you'll be able
19 to recognize them when defining your job below. Finally, in order for the
20 agent to register itself to Buildkite, it will need a ``BUILDKITE_AGENT_TOKEN``.
21 Please contact a maintainer to get your token.
23 Then, simply add a job to the Buildkite pipeline by editing ``libcxx/utils/ci/buildkite-pipeline.yml``.
24 Take a look at how the surrounding jobs are defined and do something similar.
25 An example of a job definition is:
30 command: "libcxx/utils/ci/run-buildbot generic-cxx11"
32 - "**/test-results.xml"
34 queue: "libcxx-builders"
37 - exit_status: -1 # Agent was lost
40 If you've created your own agents, you should provide the tag that you used
41 when creating them in the ``queue`` entry -- this will instruct Buildkite to
42 run that job only on agents that have that tag.
44 We try to keep the pipeline definition file as simple as possible, and to
45 keep any script used for CI inside ``libcxx/utils/ci``. This ensures that
46 it's possible to reproduce CI issues locally with ease, understanding of
47 course that some setups may require access to special hardware that is not
53 Testing your new job is easy -- once your agent is set up (if any), just open
54 a code review and the libc++ CI pipeline will run, including any changes you
55 might have made to the pipeline definition itself.
57 Service Level Agreement
58 =======================
60 To keep the libc++ CI useful for everyone, we aim for a quick turnaround time
61 for all CI jobs. This allows the overall pipeline to finish in a reasonable
62 amount of time, which is important because it directly affects our development
63 velocity. We also try to make sure that jobs run on reliable infrastructure in
64 order to avoid flaky failures, which reduce the value of CI for everyone.
66 We may be reluctant to add and support CI jobs that take a long time to finish
67 or that are too flaky.