Skip to content
Andy Arvanitis edited this page Jul 26, 2019 · 3 revisions

Note: This only apples to the C++ target

I've done some testing with the jemalloc memory allocator. All standard tests pass, and I've observed some performance improvements – one case showed a 4x runtime speed increase over using the standard system allocator, and that was just for a single-threaded case (more testing and benchmarking would be good). Simply linking with the library via LDFLAGS was all that was needed to use it.

Clone this wiki locally