Compiler suggestions send me into an endless loop

[Lehona]

I tried this code:

fn main() {
    let data: &[u8] = &[0; 10][..];
    let points: &[i8] = data.into();
}

which outputs

error[E0277]: the trait bound `&[i8]: std::convert::From<&[u8]>` is not satisfied
 --> src/main.rs:3:25
  |
3 |     let points: &[i8] = data.into();
  |                         ^^^^^^^^^^^^^^^ the trait `std::convert::From<&[u8]>` is not implemented for `&[i8]`
  |
  = note: `std::convert::From<&[u8]>` is implemented for `&mut [i8]`, but not for `&[i8]`
  = note: required because of the requirements on the impl of `std::convert::Into<&[i8]>` for `&[u8]`
  = note: required by `std::convert::Into::into`

error: aborting due to previous error

For more information about this error, try `rustc --explain E0277`.
error: could not compile `playground`.

To learn more, run the command again with --verbose.

Note the suggestion

  = note: `std::convert::From<&[u8]>` is implemented for `&mut [i8]`, but not for `&[i8]`

Changing the type of points to &mut [i8] will simply swap the suggestion to say From<&[u8]> is not implemented for &mut [i8], but it is implemented for &[i8], ultimately making me chase my own tail. I have scoured the docs and came to the conclusion that From<&[u8]> is not implemented for either of those types, so why does the compiler suggest it is?

This behaviour occurs across stable, beta and nightly on 1.42.

[ecstatic-morse]

@estebank

rust/src/librustc_trait_selection/traits/error_reporting/suggestions.rs

Lines 691 to 704 in 0612568

	if let ty::Ref(region, t_type, mutability) = trait_ref.skip_binder().self_ty().kind {
	let trait_type = match mutability {
	hir::Mutability::Mut => self.tcx.mk_imm_ref(region, t_type),
	hir::Mutability::Not => self.tcx.mk_mut_ref(region, t_type),
	};
	let new_obligation = self.mk_obligation_for_def_id(
	trait_ref.skip_binder().def_id,
	trait_type,
	ObligationCause::dummy(),
	obligation.param_env,
	);
	if self.evaluate_obligation_no_overflow(&new_obligation).must_apply_modulo_regions()

When we call mk_obligation_for_def_id here, are we forgetting that we are trying to prove From<&[u8]> for &mut [i8] and instead looking for any impl of From<T>?