In the previous section on parallel programming we briefly covered using an Interlocked.Add to avoid errors when writing to the same location in memory from multiple threads. Concurrency and locking are the hardest things to get right when it comes to multithreaded programming. Getting it wrong can hurt performance and even cause incorrect calculations.
As locking it is so often an area where mistakes are made, we will include a short primer. This is by no means extensive but will help you identify areas where you may need to be careful.
Locking is a way of ensuring that a resource can only be used by a single thread at one time. Every thread that wishes to use the locked item must wait its turn, so that operations are performed sequentially. This series processing ensures consistency but can also slow a system down if there is significant waiting for locks to be released.
Locking is also applicable to databases and you may have used it in SQL, but here...