My naive implementation (as shown in the linked question) in Rust was mind boggling slow. C could do around 150MiB/s, while Rust could only do 10MiB/s. A simple `objdump -d` shows Rust is generating a lot more code. I'm not sure how much of that is relevant though.
At that point my coffee time ran out. I wish I had more time to figure out why. :-(
At that point my coffee time ran out. I wish I had more time to figure out why. :-(