[release-v1.15] Bump knative.dev/hack to 1.15 backport

[cardil]

Restoring build for release-1.15

Related to:

• [release-1.15] 🐛 Fix knative.dev/toolbox for older versions of Golang knative/hack#440
• [support/golang-1.22] magetasks: Set GOTOOLCHAIN=auto when installing Golang dependencies knative/toolbox#41

[cardil]

/cc @maschmid
/assign @maschmid

[maschmid]

/lgtm

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: cardil, maschmid

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [cardil,maschmid]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[maschmid]

go: gotest.tools/gotestsum@latest: gotest.tools/[email protected] requires go >= 1.24.0 (running go 1.22.9; GOTOOLCHAIN=local)
[MAGE] 🔧 Go install "gotest.tools/gotestsum@latest" have failed: running "go install gotest.tools/gotestsum@latest" failed with exit code 1

is that because of knative/hack#438 ?

[openshift-ci-robot]

/retest-required

Remaining retests: 0 against base HEAD eb7352e and 2 for PR HEAD 5c34b05 in total

[openshift-ci]

New changes are detected. LGTM label has been removed.

[cardil]

@maschmid: is that because of knative/hack#438 ?

Nope. This one is related: knative/toolbox#41. I just apply it with 3f4126a

[cardil]

CI said:

The following tests failed:

Test name	Commit	Details
ci/prow/419-unit	3f4126a	link
ci/prow/419-e2e	3f4126a	link

No idea why... Clearly, a proper version of magetasks was used: https://github.com/knative/toolbox/blob/support-golang-1.22/magetasks/config/deps.go#L61.

It works on my machine, ofc. Digging deeper...

[cardil]

@cardil: Digging deeper...

I manage to get a repro:

$ podman build . --layers \
  -f openshift/ci-operator/build-image/Dockerfile \
  -t localhost/kn-event/build
$ podman run --rm -v $(pwd):/root/kn-event:Z \
  -e CI=true \
  -e PROW_JOB_ID=123 \
  --workdir /root/kn-event \
  localhost/kn-event/build:latest \
  make unit
[..]
[MAGE] 🔧 › Go install "gotest.tools/gotestsum@latest"
go: downloading gotest.tools/gotestsum v1.13.0
go: downloading gotest.tools v2.2.0+incompatible
go: gotest.tools/gotestsum@latest: gotest.tools/[email protected] requires go >= 1.24.0 (running go 1.22.9; GOTOOLCHAIN=local)
[MAGE] 🔧 Go install "gotest.tools/gotestsum@latest" have failed: running "go install gotest.tools/gotestsum@latest" failed with exit code 1
Error: running "go install gotest.tools/gotestsum@latest" failed with exit code 1

Somehow, the GOTOOLCHAIN=local env variable that is set on the image level has precedence over the env set here: https://github.com/knative/toolbox/blob/support-golang-1.22/magetasks/config/deps.go#L61.

[cardil]

@cardil: Somehow, the GOTOOLCHAIN=local env variable that is set on the image level has precedence

This one is super weird:

$ podman run --rm localhost/kn-event/build:latest env | grep GOTOOLCHAIN
$ podman run --rm localhost/kn-event/build:latest go env | grep GOTOOLCHAIN
GOTOOLCHAIN='local'

$ podman run --rm -e GOTOOLCHAIN=auto localhost/kn-event/build:latest env | grep GOTOOLCHAIN
GOTOOLCHAIN=auto
$ podman run --rm -e GOTOOLCHAIN=auto localhost/kn-event/build:latest go env | grep GOTOOLCHAIN
GOTOOLCHAIN='local'

No idea why the go env GOTOOLCHAIN is still local, even after setting it via env variable?!?

[cardil]

92f91d8 is an interim fix. I'll file a PR for openshift-knative/hack to make it org-wide.

[cardil]

@cardil: 92f91d8 is an interim fix. I'll file a PR for openshift-knative/hack to make it org-wide.

Here it is: openshift-knative/hack#814

[cardil]

/test 419-unit

[cardil]

/retest

[cardil]

@maschmid The tests are failing, as they don't respect the change I'm doing to the Dockerfile here. The build log doesn't contain the sed command.

https://github.com/openshift-knative/kn-plugin-event/pull/1008/files#diff-4547208d3f6a8c17a2a8e5e1a4dc196f03d27d61f9fdbe31fc6626e6af4532f2R23

[maschmid]

I think the openshift/ci-operator/build-image/Dockerfile configuration is taken from the branch, not the PR, during the CI run, so we'll have to force-merge it to apply...

[openshift-ci]

@cardil: The following tests failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
ci/prow/419-unit	92f91d8	link	true	/test 419-unit
ci/prow/419-e2e	92f91d8	link	true	/test 419-e2e

Full PR test history. Your PR dashboard.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here.