app-emulation/actool,acbuild: Remove actool and acbuild

[pothos]

Both actool and acbuild upstream are deprecated and probably not in use anymore.
Remove actool from the image and acbuild from the SDK.

Second commit:

Remove ACI image building bits

The ACI OEM container image was used with rkt for GCE. For long time
this has been migrated to systemd-nspawn and now systemd-sysext.
Remove the unused build library code.

How to use

Testing done

In the GitHub Action CI and on Jenkins.

[github-actions]

Build action triggered: https://github.com/flatcar/scripts/actions/runs/8534859383

[pothos]

Hm, that /update-sdk comment didn't work…

[dongsupark]

Hm, that /update-sdk comment didn't work…

As far as I know, once you add /update-sdk in a comment or description, you have to refresh any commit, and push to the PR again, so that CI can pick up the event. Without doing that, CI would still skip building SDK.
What I usually do is: git commit --amend --no-edit --date=now && git push --force origin <branchname>

[pothos]

Now that the nightly SDK is built I clicked "rerun" but only for the amd64 job instead of the whole Action job which means that I can't restart the arm64 job now. I guess it's still enough as this is not arch-specific.
Edit: After the first one finished it worked to restart the arm64 sub job.

[pothos]

The aci file for the GCE OEM wasn't even using this, it was using tar. I guess we can also clean up this a bit: build_library/oem_aci_util.sh, build_oem_aci and calls from image_to_vm.sh and build_library/vm_image_util.sh to oem_aci

[krnowak]

The aci file for the GCE OEM wasn't even using this, it was using tar. I guess we can also clean up this a bit: build_library/oem_aci_util.sh, build_oem_aci and calls from image_to_vm.sh and build_library/vm_image_util.sh to oem_aci

That'd be a nice cleanup. Can be done in follow-up PR if you want to merge this now.

[pothos]

Pushed the cleanup, maybe we should also do a jenkins build

[krnowak]

Pushed the cleanup, maybe we should also do a jenkins build

Yes, please.

[pothos]

Jenkins VM image building passed.

An unrelated failure in cl.verity/corruption only on amd64 is strange, though.

     --- FAIL: cl.verity/corruption (1.20s)
             verity.go:107: verity did not prevent reading from a corrupted disk (expected kernel panic)!

[pothos]

An unrelated failure in cl.verity/corruption only on amd64 is strange, though.

Since it passed on GitHub Actions this really seems to unrelated.

[krnowak]

An unrelated failure in cl.verity/corruption only on amd64 is strange, though.

Since it passed on GitHub Actions this really seems to unrelated.

Alright, let's merge it, then.

[pothos]

I wonder if the kola test could be improved to corrupt a bit more and try to flush the cashes again when the first corruption didn't directly trigger the panic.

[pothos]

I've tested the case manually and it helps to overwrite 100 MB instead of 10 MB, will file a mantle PR
Edit: flatcar/mantle#515