Use patched golang1.7.1 for cross-builds targeting darwin

[ixdy]

This PR extends #32517 to use the patched go1.7.1 introduced by that PR to build all darwin targets (e.g. kubectl).

This is necessary because binaries built with earlier versions of Go regularly segfault on macOS Sierra (see #32999 and #33070).

This solution is somewhat hacky, but we intend to cherry-pick this to 1.4, and switching all of 1.4 to build with go1.7.1 is very high risk.

I haven't yet pushed the cross build image yet, so this will fail to build. Will test locally and update with results.

First step of fixing #33801.

cc @luxas @pwittrock @david-mcmahon @liggitt @smarterclayton @jfrazelle @Starefossen @gerred

---

This change is

[k8s-ci-robot]

Jenkins GCE e2e failed for commit f769b02. Full PR test history.

The magic incantation to run this job again is @k8s-bot gce e2e test this. Please help us cut down flakes by linking to an open flake issue when you hit one in your PR.

[k8s-ci-robot]

Jenkins Kubemark GCE e2e failed for commit f769b02. Full PR test history.

The magic incantation to run this job again is @k8s-bot kubemark e2e test this. Please help us cut down flakes by linking to an open flake issue when you hit one in your PR.

[k8s-ci-robot]

Jenkins GCI GCE e2e failed for commit f769b02. Full PR test history.

The magic incantation to run this job again is @k8s-bot gci gce e2e test this. Please help us cut down flakes by linking to an open flake issue when you hit one in your PR.

[gerred]

@ixdy This is fantastic. Does the gcloud CLI kubectl distrubtion use anything else beyond what's OSS and available? I'd love to ensure this change is upstreamed into that tool as well.

We'll continue providing 1.3.7 binaries at Deis until GKE makes 1.4 GA. Then I don't see a need for my workaround anymore!

[k8s-ci-robot]

Jenkins verification failed for commit f769b02. Full PR test history.

The magic incantation to run this job again is @k8s-bot verify test this. Please help us cut down flakes by linking to an open flake issue when you hit one in your PR.

[ixdy]

@gerred I think the version of kubectl distributed by gcloud follows the official release supported by GKE, so it will likely update to 1.4.0 at some point in the near future. We'll probably need to cut a 1.4.1 with the go1.7.1 darwin kubectl, but I'm not sure how long that might take to make it through the distribution channels.

The current kubectl shipped by gcloud (1.3.7) is built with go1.6.2, so I imagine it's similarly broken on macOS Sierra right now.

[ixdy]

I've built kubectl off the master branch locally using this PR and have uploaded the resulting (unofficial) binaries.

Can someone with macOS Sierra test these binaries to see if they work? Based on strings I'm pretty sure they used go1.7.1.
http://dl.k8s.io/ci/v1.5.0-alpha.0.1705+1d22e86776bc18-jgrafton/bin/darwin/386/kubectl
http://dl.k8s.io/ci/v1.5.0-alpha.0.1705+1d22e86776bc18-jgrafton/bin/darwin/amd64/kubectl

[luxas]

Can someone please manually verify it works?
@gerred maybe?

I do not have a machine to test on

Reapplying the LGTM label... @ixdy feel free to push

[gerred]

@luxas Testing now.

[gerred]

That binary works perfectly.

[ixdy]

kube-cross:1.6.3-8 pushed.

@k8s-bot test this
@k8s-bot build this

[pwittrock]

This failure:

Tag v1.6.3-8 not found in repository gcr.io/google_containers/kube-cross

[luxas]

@k8s-bot gci gke e2e test this

[ixdy]

@pwittrock which build was that from?

for some reason the GCI and Kubemark builds didn't re-trigger on my last comment.
@k8s-bot kubemark e2e test this
@k8s-bot gci gce e2e test this

[luxas]

@k8s-bot gci gce e2e test this

[luxas]

Haha, race condition between @ixdy and me :)

Let's hope everything's green now

@pwittrock When are you planning to release v1.4.1?

[pwittrock]

@luxas Original plan was a week from Monday (10/10). Need to figure out if this warrants accelerating that timeline. WDYT?

[ixdy]

We'll at least have Jenkins-built kubectl binaries on gs://kubernetes-release-dev/ci as soon as this gets cherrypicked.

[pwittrock]

Nice. I can add that to the known issues in the release notes. Anywhere else we can publish it?

[luxas]

@pwittrock If possible, I'd prefer the 5th Oct or something like that, given this and the ARM issue. As soon as some other cherrypicks that are pending like #33280, #33520 and #33806 are merged, I think you should release.

Then, already more than 20 commits have been merged since v1.4.0, and that's quite a decent amount already

WDYT about Wednesday?

[k8s-ci-robot]

Jenkins GKE smoke e2e failed for commit f769b02. Full PR test history.

The magic incantation to run this job again is @k8s-bot gke e2e test this. Please help us cut down flakes by linking to an open flake issue when you hit one in your PR.

[k8s-github-robot]

@k8s-bot test this [submit-queue is verifying that this PR is safe to merge]

[k8s-ci-robot]

Jenkins GCI GKE smoke e2e failed for commit f769b02. Full PR test history.

The magic incantation to run this job again is @k8s-bot gci gke e2e test this. Please help us cut down flakes by linking to an open flake issue when you hit one in your PR.

[k8s-ci-robot]

Jenkins unit/integration failed for commit f769b02. Full PR test history.

The magic incantation to run this job again is @k8s-bot unit test this. Please help us cut down flakes by linking to an open flake issue when you hit one in your PR.

[pwittrock]

@k8s-bot gci gke e2e test this

[pwittrock]

@k8s-bot unit test this

[k8s-github-robot]

Automatic merge from submit-queue

[k8s-cherrypick-bot]

Commit found in the "release-1.4" branch appears to be this PR. Removing the "cherrypick-candidate" label. If this is an error find help to get your PR picked.