Core Data `context.performAndWait` in an actor

Hello. I am re-writing our way of storing data into Core Data in our app, so it can be done concurrently.

The solution I opted for is to have a singleton actor that takes an API model, and maps it to a Core Data object and saves it.

For example, to store an API order model, I have something like this:

func store(
  order apiOrder: APIOrder,
  currentContext: NSManagedObjectContext?
) -> NSManagedObjectID? {
  let context = currentContext ?? self.persistentContainer.newBackgroundContext()

  // …
}

In the arguments, there is a context you can pass, in case you need to create additional models and relate them to each other. I am not sure this is how you're supposed to do it, but it seemed to work.

From what I've understood of Core Data and using multiple contexts, the appropriate way use them is with context.perform or context.performAndWait.

However, since my storage helper is an actor, @globalActor actor Storage2 { … }, my storage's methods are actor-isolated.

This gives me warnings / errors in Swift 6 when I try to pass the context for to another of my actor's methods.

let context = …
return context.performAndWait {
  // …

  if let apiBooking = apiOrder.booking {
    self.store(booking: apiBooking, context: context)
    /* causes warning:
     Sending 'context' risks causing data races; this is an error in the Swift 6 language mode
        'self'-isolated 'context' is captured by a actor-isolated closure. actor-isolated uses in closure may race against later nonisolated uses
        Access can happen concurrently
    */
  }

  // …
}

From what I understand this is because my methods are actor-isolated, but the closure of performAndWait does not execute in a thread safe environment.

With all this, what are my options? I've extracted the store(departure:context:) into its own method to avoid duplicated code, but since I can't call it from within performAndWait I am not sure what to do.

Can I ditch the performAndWait? Removing that makes the warning "go away", but I don't feel confident enough with Core Data to know the answer.

I would love to get any feedback on this, hoping to learn!