Auto merge of #73446 - ecstatic-morse:issue-73431, r=pnkfelix

Make novel structural match violations not a `bug`

Fixes (on master) #73431.

Ideally, `CustomEq` would emit a strict subset of the structural match errors that are found by `search_for_structural_match_violation`, since it allows more cases due to value-based reasoning. However, const qualification is more conservative than `search_for_structural_match_violation` around associated constants, since qualification does not try to substitute type parameters.

In the long term, we should probably make const qualification work for generic associated constants, but I don't like extending its capabilities even further.

r? @pnkfelix
This commit is contained in:
bors 2020-06-18 21:50:45 +00:00
commit 036b5fec49
2 changed files with 37 additions and 1 deletions

View File

@ -107,8 +107,15 @@ fn to_pat(
cv.ty, structural
);
// This can occur because const qualification treats all associated constants as
// opaque, whereas `search_for_structural_match_violation` tries to monomorphize them
// before it runs.
//
// FIXME(#73448): Find a way to bring const qualification into parity with
// `search_for_structural_match_violation`.
if structural.is_none() && mir_structural_match_violation {
bug!("MIR const-checker found novel structural match violation");
warn!("MIR const-checker found novel structural match violation. See #73448.");
return inlined_const_as_pat;
}
if let Some(non_sm_ty) = structural {

View File

@ -0,0 +1,29 @@
// run-pass
// Regression test for https://github.com/rust-lang/rust/issues/73431.
pub trait Zero {
const ZERO: Self;
}
impl Zero for usize {
const ZERO: Self = 0;
}
impl<T: Zero> Zero for Wrapper<T> {
const ZERO: Self = Wrapper(T::ZERO);
}
#[derive(Debug, PartialEq, Eq)]
pub struct Wrapper<T>(T);
fn is_zero(x: Wrapper<usize>) -> bool {
match x {
Zero::ZERO => true,
_ => false,
}
}
fn main() {
let _ = is_zero(Wrapper(42));
}