6.4. Summary

This chapter started by looking at what it means to design a data structure for concurrency and providing some guidelines for doing so. We then worked through several common data structures (stack, queue, hash map, and linked list), looking at how to apply those guidelines to implement them in a way designed for concurrent access, using locks to protect the data and prevent data races. You should now be able to look at the design of your own data structures to see where the opportunities for concurrency lie and where there’s potential for race conditions.

In chapter 7 we’ll look at ways of avoiding locks entirely, using the low-level atomic operations to provide the necessary ordering constraints, while sticking to the same set of guidelines.