Add new workspace condition "stoppedByRequest"

[jankeromnes]

Bug description

When we call stopWorkspace with a StopWorkspaceRequest on some workspace, the stopped workspace should come back to the bridge with a new type of condition attached: for example called conditions.stoppedByRequest (in addition to the existing conditions timeout, failed and headlessTaskFailed).

This would allow cancelling prebuilds (because when a StopWorkspaceRequest is called on a prebuild a.k.a. headless workspace, we know that this is a manual cancellation, as the prebuild workspace wasn't left to run until completion or timeout).

This would help unblock the cancellable prebuilds PR: #5865 where we currently:

• Stop the workspace (but just doing that currently leads to a "successful" Prebuild, which shows up in green in the UI and not in red as expected)
• Set prebuild.state = 'aborted' immediately from the server, which seems to work on the surface, but is risky (because prebuild.state could get modified again by the bridge later when the workspace eventually stops)
• Once the new stoppedByRequest condition is implemented, we'd be able to look out for that on prebuild workspaces in the bridge and set the correct prebuild.state.

Steps to reproduce

N/A

Workspace affected

No response

Expected behavior

No response

Example repository

No response

Anything else?

No response

[AlexTugarev]

Just for completeness, let's not call this "manual cancellation" but just generic "cancellation" in code, or prepare a field for a reason. Any further improvements towards of prebuild queues would also require to "skip" even potentially running prebuilds, thus cancellation might be requested.

[csweichel]

/schedule

[roboquat]

@csweichel: Issue scheduled in the workspace team (WIP: 0)

In response to this:

/schedule

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.

[csweichel]

/assign