Flake: CI Validation: Golangci-lint v1.54.2: fatal error: concurrent map writes

[cevich]

Issue Description

While running the Validate fedora-38 Build CI job in an image-update PR (c20230824t202814z-f38f37d13) golangci-lint is barfing with a huge traceback (log).

Steps to reproduce the issue

Steps to reproduce the issue

1. Open a PR (with the CI VM image suffix mentioned above)
2. Run CI

Describe the results you received

Validate fedora-38 Build job fails, but passes on re-run.

Describe the results you expected

CI should always pass

podman info output

N/A

Podman in a container

No

Privileged Or Rootless

None

Upstream Latest Release

Yes

Additional environment details

N/A

Additional information

No response

[cevich]

cc @edsantiago - I didn't see this on your flake list. Just FYI. Also, renovate is already configured to update golangci-lint, there's no update available ATM.

[Luap99]

I recommend you file this against golgangci-lint, this error is pretty much always a logic bug in the code, parallel map access is not allowed in go and must be guarded with a mutex.

[cevich]

Yeah that's probably a good idea, though being a flake it's probably hard to reproduce 😞 In anycase I wanted to report it here in case anybody else hits it and as a reminder that we probably need to update golangci-lint when one is available.

[Luap99]

A reproducer is helpful but not necessary because you captured the full stack trace we can see where the concurrent write happens so it should be easy enough for upstream developers to understand and fix IMO.

Looking at the upstream repo I see https://gitlab.com/bosi/decorder/-/merge_requests/91 and golangci/golangci-lint#4056 so I expect the fix for this to be included in the next golangci-lint version.

[github-actions]

A friendly reminder that this issue had no activity for 30 days.