Bump controller-gen to v0.9.2.

[benluddy]

With the latest controller-gen, leading and trailing whitespace is
stripped from the generated description text (kubernetes-sigs/controller-tools#517), and topology
markers (e.g. +structType, +mapType) are now being recognized when
applied to types as well as fields (kubernetes-sigs/controller-tools#692).

[benluddy]

'_output/tools/bin/controller-gen-v0.9.0' schemapatch:manifests="./operatoringress/v1" paths="./operatoringress/v1" output:dir="/tmp/tmp.1Pbg8eLCZK"
Error: unable to parse option "output:dir=/tmp/tmp.1Pbg8eLCZK": ['P' exponent requires hexadecimal mantissa (at :1:9) exponent has no digits (at :1:9)]

https://prow.ci.openshift.org/view/gs/origin-ci-test/pr-logs/pull/openshift_api/1203/pull-ci-openshift-api-master-verify/1529842193689219072#1:build-log.txt%3A109

openshift/build-machinery-go#67

/retest-required

[soltysh]

/lgtm
/approve
/label qe-approved
/label docs-approved
/label px-approved

[soltysh]

/test verify

[openshift-ci-robot]

/retest-required

Remaining retests: 2 against base HEAD 8324d65 and 8 for PR HEAD 4c8bf2e in total

[openshift-ci-robot]

/retest-required

Remaining retests: 1 against base HEAD 8324d65 and 7 for PR HEAD 4c8bf2e in total

[benluddy]

/hold

[benluddy]

Need to understand the impact of this topology marker fix first (https://kubernetes.io/docs/reference/using-api/server-side-apply/#compatibility-across-topology-changes).

/hold

[JoelSpeed]

@benluddy I was looking into making this bump as a part of some other changes I'm making, did you work out what the impact is likely to be of adding these markers?

IIUC all lists are atomic by default, so anyone using SSA who would observe a map type can't have been using it yet right? The entire list will currently have a single owner but may then in the future have multiple owners. That said, if the list is currently atomic, none of our consumers can have any logic assuming it is a map or otherwise so will end up preserving the entire list right now

Perhaps if we can enumerate the places this has changed we can determine how many consumers there are and where it is safe and where it is not safe to make the change

[JoelSpeed]

I think this is ok. The markers only affect clients using SSA. We don't currently have generated SSA clients so it wouldn't be easy for any consumers of these types to be using SSA yet anyway. Add to that that the default list type behaviour is atomic, this shouldn't be a breaking change for any of these instances.

/lgtm

[deads2k]

we need to sweep all slices before 4.12.

/approve
/assign @JoelSpeed

[benluddy]

/hold cancel

[JoelSpeed]

/lgtm

We need the updated generator so that we can continue with the introduction of tech preview fields for openshift APIs

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: benluddy, deads2k, JoelSpeed, soltysh

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 [deads2k,soltysh]

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

[openshift-ci]

@benluddy: all tests passed!

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/test-infra repository. I understand the commands that are listed here.

[Miciah]

/lgtm
/approve
/hold
for other reviewers.

[Miciah]

Sorry, wrong PR.
/hold cancel