Why is F16C still unstable?

[netizen-a]

I noticed that most x86_64 intrinsics were stable, yet F16C isn't. I'd like to know what is keeping it from hitting stable, and what can be done to make it stable. According to another tracking issue comment there doesn't seem to be anything blocking it.

[Amanieu]

There is no specific reason, just that nobody has pushed for its stabilization yet. In this case all that would be needed is a PR to add stability attributes to these function to stabilize them.

[aloucks]

It's odd that the runtime check seems to think it was stablized in 1.27.0:

stdarch/crates/std_detect/src/detect/arch/x86.rs

Line 166 in 91c64cd

@FEATURE: #[stable(feature = "simd_x86", since = "1.27.0")] f16c: "f16c";

It looks like this was missed:
https://github.com/rust-lang/rust/blob/887999d163bace7e79370b952bdd1f930ff4cdd5/compiler/rustc_feature/src/active.rs#L241

Does it just need to be moved here?:
https://github.com/rust-lang/rust/blob/887999d163bace7e79370b952bdd1f930ff4cdd5/compiler/rustc_feature/src/accepted.rs

[Amanieu]

The feature detection was stabilized separately from the intrinsics.

[Amanieu]

The F16C intrinsics aren't stable yet, this is still waiting on FCP in #1366.