[NO TESTS NEEDED] Fix (experimental) Volume Plugin in rootless mode

[thephoenixofthevoid]

According to #9650 there are few issues when using volume plugins with rootless containers.

The reason behind these is local-driver-specific code being executed for all cases. This is a minimal fix to address these issues.

Please also note that:

1. In fact, there is a need to clearly separate local-driver-specific code from plugin-related, because there are little to none logic that should be executed in both cases.
2. The removed check is actually already correctly implemented here
   

   podman/libpod/runtime_volume_linux.go

   Lines 66 to 77 in d92b946

   	if volume.config.Driver == define.VolumeDriverLocal {
   	logrus.Debugf("Validating options for local driver")
   	// Validate options
   	for key := range volume.config.Options {
   	switch key {
   	case "device", "o", "type", "UID", "GID":
   	// Do nothing, valid keys
   	default:
   	return nil, errors.Wrapf(define.ErrInvalidArg, "invalid mount option %s for driver 'local'", key)
   	}
   	}
   	}

3. These kind of checks is actually implementation specific, so it seems natural to keep them close to the implementation.

[mheon]

Git validation is complaining that your commits don't have a signoff - git commit --amend -s should add one.

I think it will also complain that you don't have tests. You can add a test (a slightly-changed version (option added) of https://github.com/containers/podman/blob/master/test/e2e/volume_plugin_test.go#L53-L85 would work) or add [NO TESTS NEEDED] to your PR title before you re-push.

/approve

[thephoenixofthevoid]

At this point no new tests are needed, because there is no reliable and stable volume plugin yet. Using unstable volume plugin is rather harmful, as such test is much more likely to break due to unrelated issue, than due to regression of this fix.

[mheon]

Sorry - it's now complaining about your commit subject length:
- FAIL - commit subject exceeds 90 characters

[mheon]

This test failure doesn't make any sense. The commit subject it's complaining about doesn't seem to be present in the branch.
@cevich @edsantiago Any ideas?

[edsantiago]

The second line of a git commit must be empty. @thephoenixofthevoid please insert an empty line between Fix rootless... and In a case of.

[thephoenixofthevoid]

Oh, no, it failed again.

[mheon]

Ah, this one's simple. Just need to rebase against latest master - it's a known test failure due to timezone change that we've fixed already.

[openshift-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: mheon, thephoenixofthevoid

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

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

[mheon]

@thephoenixofthevoid Did you mean to close?

[thephoenixofthevoid]

No. It happened as a result of rebase automatically.

[mheon]

Hm. Git is saying that there are 0 new commits - I think your new commit got lost somehow?

[thephoenixofthevoid]

Luckly, I have patch files, and will recreate everything.

[thephoenixofthevoid]

#9808

[thephoenixofthevoid]

That was completely unexpected, that github closes PR if by chance/by mistake commits get excluded from branch.