Hadrien G.

Ah, after looking through the source code, it's actually there, but incorrectly named shuffle_av_**i32**_all_m256. So another candidate for [naming convention cleanup](https://github.com/Lokathor/safe_arch/issues/105).

Thanks for opening this issue! Since I went through this little porting exercise, I've had a bunch of language/lib ideas like this written somewhere on a TODO list, but haven't...

@dependabot rebase

This is actually not fixed for \`inline code blocks\`.

Thank you !

In general, I would hesitate to use map_or_else when either branch is side-effectful (assertions, I/O, etc), because I that feel that obscures the control flow too much and control flow...

Out of curiosity, what issues did you encounter with `inline(always)`? I tend to use it in "not inlining would be a total performance disaster" situations, so if it can misbehave...

Mmm... sounds bug-report-worthy to me if you reproduce it again someday. Searching a bit through rustc issues, there seems to have been a number of issues concerning funky soundness issues...

I have spent some time thinking about this "one element at a time" design before, and [think it would not be workable with good API ergonomics and reasonable compile-time overhead](https://github.com/rust-lang/rust/issues/80094#issuecomment-1464853105)....

Re. phf, I was trying to come up with an example of a container... - Whose size is known at compile time - Whose backing store is more complex than...