Add a condition disk clean-up step to reclaim some disk space in pr-comment-trigger & provider-ci reusable workflows

[ulucinar]

Description of your changes

We have failing CI workflows in the release-0.47 branch of upbound/provider-aws with insufficient disk space. An example run is here. We would like to give the jlumbroso/free-disk-space action a try in that branch to see whether it resolves these issues. Previously we had relied on manual running of the failing jobs for that branch. This action has already helped us for running uptest on the free hosted runners.

This PR also makes them conditional. An observation from @sergenyalcin after we introduced the jlumbroso/free-disk-space is that it takes around 7m on a larger hosted runner with more disk space. We don't need to run the disk cleanup action for those larger runners and thus a switch to control the enablement of the step will help. By default, we won't running the disk cleanup step and we will only enable it when needed.

In this PR, we don't implement a fine granular switch specific to each job which controls the disk cleanup step but rather a global switch for all the jobs defined in a workflow. Also we did not include the disk cleanup step for each and every job in a workflow. I don't think we currently need a granular control specific to each job atm. We can keep adding the clean-up step for the remaining jobs as needed.

I have:

• Run make reviewable test to ensure this PR is ready for review.

How has this code been tested

Tested with the previously failing check-diff job here: https://github.com/upbound/provider-aws/actions/runs/8186204677/job/22384138959?pr=1197. Please also see crossplane-contrib/provider-upjet-aws#1197.

[sergenyalcin]

Thanks @ulucinar LGTM!