Replacing Threads with Dispatch Queues

Replacing Threads with Dispatch Queues

To understand how you might replace threads with dispatch queues, first consider some of the ways you might be using threads in your application today: 

• Single task threads. Create a thread to perform a single task and release the thread when the task is done. 

• Worker threads. Create one or more worker threads with specific tasks in mind for each. Dispatch tasks to each thread periodically.

• Thread pools. Create a pool of generic threads and set up run loops for each one. When you have a task to perform, grab a thread from the pool and dispatch the task to it. If there are no free threads, queue the task and wait for a thread to become available. 

Although these might seem like dramatically different techniques, they are really just variants on the same principle. In each case, a thread is being used to run some task that the application has to perform. The only difference between them is the code used to manage the threads and the queueing of tasks. With dispatch queues and operation queues, you can eliminate all of your thread and thread-communication code and instead focus on just the tasks you want to perform. 

If you are using one of the above threading models, you should already have a pretty good idea of the type of tasks your application performs. Instead of submitting a task to one of your custom threads, try encapsulating that task in an operation object or a block object and dispatching it to the appropriate queue. For tasks that are not particularly contentious—that is, tasks that do not take locks—you should be able to make the following direct replacements:

• For a single task thread, encapsulate the task in a block or operation object and submit it to a concurrent queue. 

• For worker threads, you need to decide whether to use a serial queue or a concurrent queue. If you use worker threads to synchronize the execution of specific sets of tasks, use a serial queue. If you do use worker threads to execute arbitrary tasks with no interdependencies, use a concurrent queue. 

• For thread pools, encapsulate your tasks in a block or operation object and dispatch them to a concurrent queue for execution.

Of course, simple replacements like this may not work in all cases. If the tasks you are executing contend for shared resources, the ideal solution is to try to remove or minimize that contention first. If there are ways that you can refactor or rearchitect your code to eliminate mutual dependencies on shared resources, that is certainly preferable. However, if doing so is not possible or might be less efficient, there are still ways to take advantage of queues. A big advantage of queues is that they offer a more predictable way to execute your code. This predictability means that there are still ways to synchronize the execution of your code without using locks or other heavyweight synchronization mechanisms. Instead of using locks, you can use queues to perform many of the same tasks: 

• If you have tasks that must execute in a specific order, submit them to a serial dispatch queue. If you prefer to use operation queues, use operation object dependencies to ensure that those objects execute in a specific order. 

• If you are currently using locks to protect a shared resource, create a serial queue to execute any tasks that modify that resource. The serial queue then replaces your existing locks as the synchronization mechanism. For more information techniques for getting rid of locks, see Eliminating Lock-Based Code. 

• If you use thread joins to wait for background tasks to complete, consider using dispatch groups instead. You can also use an  NSBlockOperation object or operation object dependencies to achieve similar group-completion behaviors. For more information on how to track groups of executing tasks, see Replacing Thread Joins. 

• If you use a producer-consumer algorithm to manage a pool of finite resources, consider changing your implementation to the one shown in Changing Producer-Consumer Implementations. 

• If you are using threads to read and write from descriptors, or monitor file operations, use the dispatch sources as described in Dispatch Sources.

It is important to remember that queues are not a panacea for replacing threads. The asynchronous programming model offered by queues is appropriate in situations where latency is not an issue. Even though queues offer ways to configure the execution priority of tasks in the queue, higher execution priorities do not guarantee the execution of tasks at specific times. Therefore, threads are still a more appropriate choice in cases where you need minimal latency, such as in audio and video playback. 

https://developer.apple.com/library/content/documentation/General/Conceptual/ConcurrencyProgrammingGuide/ThreadMigration/ThreadMigration.html#//apple_ref/doc/uid/TP40008091-CH105-SW10