Reducing MD5 (and SHA) overheads


Following up on a recent post - (Investigating MD5 Overheads - where I detailed some of the analysis I did around improving java.util.UUID::nameUUIDFromBytes. As of writing that PR is still open, but we have integrated a few optimizations that improve things substantially in the vicinity…

More at https://cl4es.github.io/2021/02/14/Reducing-MD5-and-SHA-Overheads.html