KIC-KGO integration

[mlavacca]

What this PR does / why we need it:
Fixes #2913

Which issue this PR fixes:

Special notes for your reviewer:

PR Readiness Checklist:

Complete these before marking the PR as ready to review:

• the CHANGELOG.md release notes have been updated to reflect any significant (and particularly user-facing) changes introduced by this PR

[shaneutt]

Looks like we're on the right path. We updated the examples too, which is important. I'm worried about testing: are there some extra test cases that might be prudent to add? Have we manually tested this? I would like to hear @rainest's thoughts as well.

[mlavacca]

Looks like we're on the right path. We updated the examples too, which is important. I'm worried about testing: are there some extra test cases that might be prudent to add? Have we manually tested this? I would like to hear @rainest's thoughts as well.

Which scenario do you feel we are not testing properly? I think we can add a test case in which a managed GatewayClass is used by the Gateway, and we check that it's not eventually reconciled. WDYT?

[shaneutt]

Looks like we're on the right path. We updated the examples too, which is important. I'm worried about testing: are there some extra test cases that might be prudent to add? Have we manually tested this? I would like to hear @rainest's thoughts as well.

Which scenario do you feel we are not testing properly? I think we can add a test case in which a managed GatewayClass is used by the Gateway, and we check that it's not eventually reconciled. WDYT?

Sounds good 👍

[randmonkey]

Generally OK, but I have a question: ExtractUnmanagedGatewayClassMode could cause panic when annotations is nil. If we have some mechanism elsewhere to guarantee that gatewayClass.Annotations is always non-nil, could we add some comments somewhere to notice that? If not guaranteed, we should judge whether annotation is nil in that function.

[mlavacca]

waiting for @shaneutt or @rainest to get the final approval as a double check.

[shaneutt]

I'm approving, but I would like us to wait until we have @rainest's explicit review before we merge as he's spent more time than I in this area of the code recently.

[rainest]

I have? I don't remember anything other than 3.x compatibility issues at this point. I touched status handling stuff but that's only after we've decided to interact with a Gateway.

Semantically we probably should have made separate class/unmanaged checks and called them separately in an umbrella process/don't function, but whatever, that's easy to change later. The tests look appropriate and nothing in the code looks that unusual,