[the8472]

The function has to be made pub so it doesn't get optimized out as unusued private function.

Godbolt is a better choice for looking at asm anyway. https://rust.godbolt.org/z/3Y9ovsoz9

[hayley-patton]

Narrator: "The code did not, in fact, auto-vectorise."

(There's only addsd/movsd instructions, which are add/move scalar-double; we want addpd/movpd which are add/move packed-double in vectorised code.)

[KineticLensman]

Ah, that worked, thanks!

Although I can now see why he didn't include the output directly.