Rollup merge of #102472 - lcnr:static-in-eval, r=jackh726

stop special-casing `'static` in evaluation

fixes #102360

I have no idea whether this actually removed all places where `'static` matters. Without canonicalization it's very easy to accidentally rely on `'static` again. Blocked on changing the `order_dependent_trait_objects` future-compat lint to a hard error

r? `@nikomatsakis`

tests/ui/marker_trait_attr/overlap-marker-trait-with-static-lifetime.rs

@@ -1,4 +1,8 @@
-// check-pass
+// known-bug: #89515
+//
+// The trait solver cannot deal with ambiguous marker trait impls
+// if there are lifetimes involved. As we must not special-case any
+// regions this does not work, even with 'static
 #![feature(marker_trait_attr)]
 
 #[marker]

tests/ui/marker_trait_attr/overlap-marker-trait-with-static-lifetime.stderr

@@ -0,0 +1,31 @@
+error[E0283]: type annotations needed: cannot satisfy `&'static (): Marker`
+  --> $DIR/overlap-marker-trait-with-static-lifetime.rs:11:17
+   |
+LL | impl Marker for &'static () {}
+   |                 ^^^^^^^^^^^
+   |
+note: multiple `impl`s satisfying `&'static (): Marker` found
+  --> $DIR/overlap-marker-trait-with-static-lifetime.rs:11:1
+   |
+LL | impl Marker for &'static () {}
+   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^
+LL | impl Marker for &'static () {}
+   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+error[E0283]: type annotations needed: cannot satisfy `&'static (): Marker`
+  --> $DIR/overlap-marker-trait-with-static-lifetime.rs:12:17
+   |
+LL | impl Marker for &'static () {}
+   |                 ^^^^^^^^^^^
+   |
+note: multiple `impl`s satisfying `&'static (): Marker` found
+  --> $DIR/overlap-marker-trait-with-static-lifetime.rs:11:1
+   |
+LL | impl Marker for &'static () {}
+   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^
+LL | impl Marker for &'static () {}
+   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+error: aborting due to 2 previous errors
+
+For more information about this error, try `rustc --explain E0283`.

tests/ui/marker_trait_attr/overlapping-impl-1-modulo-regions.rs

@@ -1,9 +1,17 @@
-// check-pass
+// known-bug: #109481
+//
+// While the `T: Copy` is always applicable when checking
+// that the impl `impl<T: Copy> F for T {}` is well formed,
+// the old trait solver can only approximate this by checking
+// that there are no inference variables in the obligation and
+// no region constraints in the evaluation result.
+//
+// Because of this we end up with ambiguity here.
 #![feature(marker_trait_attr)]
 
 #[marker]
 pub trait F {}
-impl<T> F for T where T: Copy {}
-impl<T> F for T where T: 'static {}
+impl<T: Copy> F for T {}
+impl<T: 'static> F for T {}
 
 fn main() {}

tests/ui/marker_trait_attr/overlapping-impl-1-modulo-regions.stderr

@@ -0,0 +1,14 @@
+error[E0310]: the parameter type `T` may not live long enough
+  --> $DIR/overlapping-impl-1-modulo-regions.rs:14:21
+   |
+LL | impl<T: Copy> F for T {}
+   |                     ^ ...so that the type `T` will meet its required lifetime bounds
+   |
+help: consider adding an explicit lifetime bound...
+   |
+LL | impl<T: Copy + 'static> F for T {}
+   |              +++++++++
+
+error: aborting due to previous error
+
+For more information about this error, try `rustc --explain E0310`.