Chunked pools
Our size-based arena example was optimized for a single block size and specific usage patterns, but there are many other reasons to want to apply a specialized allocation strategy. In this section, we will explore the idea of a “chunked pool,” or a pool of pre-allocated raw memory of selected block sizes. This is meant as an academic example to build upon more than as something to use in production; the code that follows will be reasonably fast and can be made to become very fast, but in this book, we will focus on the general approach and leave you, dear reader, to enjoy optimizing it to your liking.
The idea in this example is that user code plans to allocate objects of similar (but not necessarily identical) sizes and of various types and supposes an upper bound on the maximal number of objects. This gives us additional knowledge; using that knowledge, we will write a ChunkSizedAllocator<N,Sz...> type where N will be the number of objects of...
