Performance measurements
The library needs micro-and-macro benchmarking to evaluate the performance of various allocators. Of particular interest is replaying a trace of real applications allocation behavior on a vspace and measuring external fragmentation and average, worst, and best-case times for each operation.