Rust has #![no_core] which is perma-unstable.
It's used by core itself and I suppose could be useful if core wants to use crates on crates.io, then the crate will need to mark itself #![no_core]
When your crate is #![no_core], you don't get anything provided by Rust by default such as Copy so the compiler crashes, expecting language items to exist but they don't. You have to manually define necessary language item with #[lang_item] if you want to use them
#![no_core] can be useful to implement your own core. But why would anyone want to do this?
Are there any reasons why someone would want to mark their crate as #![no_core] aside from implementation of core itself?