Auto merge of #50121 - pnkfelix:revert-stabilization-of-never-type-et-al, r=alexcrichton

Revert stabilization of never_type (!) et al

Fix #49691

I *think* this correctly adopts @nikomatsakis 's desired fix of:
 * reverting stabilization of `!` and `TryFrom`, and
 * returning to the previous fallback semantics (i.e. it is once again dependent on whether the crate has opted into `#[feature(never_type)]`,
 * **without** attempting to put back in the previous future-proofing warnings regarding the change in fallback semantics.

(I'll be away from computers for a week starting now, so any updates to this PR should be either pushed into it, or someone else should adopt the task of polishing this fix and put up their own PR.)

1 files changed, 7 insertions, 0 deletions

diff --git a/src/libsyntax/feature_gate.rs b/src/libsyntax/feature_gate.rs
index 6426c9a92f2..fe2466ff005 100644
--- a/src/libsyntax/feature_gate.rs
+++ b/src/libsyntax/feature_gate.rs
@@ -272,6 +272,9 @@ declare_features! (
     // Allows cfg(target_has_atomic = "...").
     (active, cfg_target_has_atomic, "1.9.0", Some(32976), None),
 
+    // The `!` type. Does not imply exhaustive_patterns (below) any more.
+    (active, never_type, "1.13.0", Some(35121), None),
+
     // Allows exhaustive pattern matching on types that contain uninhabited types.
     (active, exhaustive_patterns, "1.13.0", None, None),
 
@@ -1644,6 +1647,10 @@ impl<'a> Visitor<'a> for PostExpansionVisitor<'a> {
             ast::TyKind::BareFn(ref bare_fn_ty) => {
                 self.check_abi(bare_fn_ty.abi, ty.span);
             }
+            ast::TyKind::Never => {
+                gate_feature_post!(&self, never_type, ty.span,
+                                   "The `!` type is experimental");
+            }
             ast::TyKind::TraitObject(_, ast::TraitObjectSyntax::Dyn) => {
                 gate_feature_post!(&self, dyn_trait, ty.span,
                                    "`dyn Trait` syntax is unstable");