actions/checkout@v4 SIGSEGV (exit code 139) in ubuntu-24.04-arm

[MikeMcC399]

Issue

The step

- uses: actions/checkout@v4

sporadically fails in ubuntu-24.04-arm.

##[debug]Node Action run completed with exit code 139

This indicates SIGSEGV.

If the job is re-run, the step is often successful.

This is in a workflow with parallel runs with the matrix: construct. If I strip out the workflow to only include the checkout step and then run it, I got no failure in five attempts.

I have also observed it using matrix: with only one combination:

    strategy:
      fail-fast: false
      matrix:
        os: [ubuntu-24.04-arm]
    runs-on: ${{ matrix.os }}

Logs

See for example https://github.com/cypress-io/cypress-docker-images/actions/runs/13159700085/job/36725026693

Related

• Also mentioned in ubuntu-24.04-arm - consistent failures of 1-3 jobs at steps with actions in a matrix jobs #37

[MikeMcC399]

If the workflow is successful in actions/checkout, then it sporadically fails executing npm ci.

These are not hard errors.

The same workflow has no reliability issue on ubuntu-24.04. The reliability issues only occur on ubuntu-24.04-arm.

[MikeMcC399]

I have simplified a test workflow and put it into

https://github.com/MikeMcC399/cypress-docker-images/blob/test/arm/.github/workflows/example-tests.yml

[MikeMcC399]

This is probably a duplicate of other issues mentioning segmentation faults in ubuntu-24.04-arm

• Segmentation faults when using vcpkg on ubuntu-24.04-arm64 runner #46
• ubuntu-24.04-arm - inconsistent rustc failures when compile CRuby #40
• Intermittent segmentation faults from rustc on github_arm_linux_runner_2404 runner-images#11533

I don't currently see anybody assigned to any of these issues. Is there somewhere to subscribe to some consolidated place where related issues in ubuntu-24.04-arm are being worked on?

@tedchamb

Are you officially / unofficially working on this by any chance?

[MikeMcC399]

Also seeing sporadic error, not running parallel tests:

docker: Error response from daemon: failed to create task for container: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: unable to apply cgroup configuration: unable to start unit "docker-af2bf5da54cb91fa916e89d0291d992a17a368f9c219835e8f04e6458e435be4.scope" (properties [{Name:Description Value:"libcontainer container af2bf5da54cb91fa916e89d0291d992a17a368f9c219835e8f04e6458e435be4"} {Name:Slice Value:"system.slice"} {Name:Delegate Value:true} {Name:PIDs Value:@au [3153]} {Name:MemoryAccounting Value:true} {Name:CPUAccounting Value:true} {Name:IOAccounting Value:true} {Name:TasksAccounting Value:true} {Name:DefaultDependencies Value:false}]): Failed to activate service 'org.freedesktop.systemd1': timed out (service_start_timeout=25000ms): unknown.

[MikeMcC399]

We've removed the use of ubuntu-24.04-arm in the production repo https://github.com/cypress-io/cypress-docker-images/ until these issues are resolved.

The issues are still reproducible however in a branch of my fork
https://github.com/MikeMcC399/cypress-docker-images/blob/test/arm/.github/workflows/example-tests.yml

Please let me know if there is anything additional I can do to aid troubleshooting.

[pooley182]

I am also seeing this sporadic SIGSEGV error on the checkout action.
However, I'm only seeing issues on the public runners.

Within our orginisation we added new Github Hosted Runners, using the ubuntu-24.04-arm image (as the public runners aren't available to orgs). We have not seen a single instance of either the SIGSEGV for actions/checkout@v4 or any of the other transient issues, like the docker daemon not running.

[MrCroxx]

Hi, I met the same problem in my repo.

I opened the issue in the wrong repo: actions/runner-images#11558

There were debug logs and log archives there that might help debug.

[bbasata]

I have reproduced this issue again today: https://github.com/bbasata/setup-terraform-issue-459/actions/runs/13310867224.

[MikeMcC399]

@bbasata

• We're really just waiting for a maintainer response now. I also asked Linux arm64 runner - Public Preview status? #50 for a general status update.

[bbasata]

We're really just waiting for a maintainer response now. I also asked Linux arm64 runner - Public Preview status? #50 for a general status update.

Thank you. I will contact GitHub Support.