Make FakeExec safe for concurrency

[artemvmin]

What type of PR is this?
/kind feature

What this PR does / why we need it:
This PR adds a mutex to the FakeExec struct.

The real executor shares no state between osexec.Command() calls. Meanwhile, FakeExec shares all the state stored on the struct.

I am working with a CSI driver that uses a single object for all provisioning calls. This change will allow me to test its behavior when calls are made concurrently.

This change is backwards compatible, since the existing behavior is undefined (and nondeterministic) for concurrent calls.

Release note:

The FakeExec test double is now safe for concurrency.

[k8s-ci-robot]

Welcome @artemvmin!

It looks like this is your first PR to kubernetes/utils 🎉. Please refer to our pull request process documentation to help your PR have a smooth ride to approval.

You will be prompted by a bot to use commands during the review process. Do not be afraid to follow the prompts! It is okay to experiment. Here is the bot commands documentation.

You can also check if kubernetes/utils has its own contribution guidelines.

You may want to refer to our testing guide if you run into trouble with your tests not passing.

If you are having difficulty getting your pull request seen, please follow the recommended escalation practices. Also, for tips and tricks in the contribution process you may want to read the Kubernetes contributor cheat sheet. We want to make sure your contribution gets all the attention it needs!

Thank you, and welcome to Kubernetes. 😃

[artemvmin]

/assign @msau42

[msau42]

/lgtm

[msau42]

/assign @dims

[apelisse]

Can this lead to a lock ordering inversion or even just a recursive lock?

[artemvmin]

Command() and LookPath() don't explicitly call each other. This means that once either function acquires a lock, it will be able to finish and unlock. Still, the caller could embed one method inside the other, which would cause deadlock.

I went ahead and tightened the lock to only the section violating the invariant, i.e. when CommandCalls is incremented and CommandScript accessed non-atomically.

There's still the implication that the fields in the fake object shouldn't be modified while it's being used in a concurrent context, but there are easier ways to make a brittle test.

[apelisse]

Thanks, I would also at least mention this in the documentation for the function that the risk exists.

[artemvmin]

Good idea. Done.

[apelisse]

/approve

@msau42 do you mind re-applying lgtm?

Thanks!

[msau42]

/lgtm

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: apelisse, artemvmin, msau42

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 [apelisse]

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