ICE: assertion failed: !obligations.has_infer() in rustc_hir_analysis/src/impl_wf_check/min_specialization.rs with feature min_specialization

[cushionbadak]

Code

(hand-reduced)

#![feature(min_specialization)]

trait Trait {}

impl<T> Trait for T {}

trait Data {
    type Elem;
}

struct DatasetIter<'a, R: Data> {
    data: &'a R::Elem,
}

pub struct ArrayBase {}

impl<'a> Trait for DatasetIter<'a, ArrayBase> {}

fn main() {}

(original, 190+ lines)

// Because of #109628, when we compute the implied bounds from `Foo<X>`,
// we incorrectly get `X: placeholder('x)`.
// Make sure we ignore these bogus bounds and not use them for anything useful.
//
//@ revisions: fail pass
//@ [fail] check-fail
//@ [pass] check-pass

trait Trait {
    type Ty<'a> where Self: 'a;
}

impl<T> Trait for T {
    type Ty<'a> = () where Self: 'a;
}

struct Foo<T: Trait>(T)
where
    for<'x> T::Ty<'x>: Sized;

trait AnotherTrait {
    type Ty2<'a>: 'a;
}

#[cfg(fail)]
mod fail {
    use super::*;

    // implied_bound: `'lt: placeholder('x)`.
    // don't use the bound to prove `'lt: 'static`.
    fn test_lifetime<'lt, T: Trait>(_: Foo<&'lt u8>) {}
    //[fail]~^ ERROR `&'lt u8` does not fulfill the required lifetime

    // implied bound: `T::Ty2<'lt>: placeholder('x)`.
    // don't use the bound to prove `T::Ty2<'lt>: 'static`.
    fn test_alias<'lt, T: AnotherTrait>(_: Foo<T::Ty2::<'lt>>) {}
    //[fail]~^ ERROR `<T as AnotherTrait>::Ty2<'lt>` does not fulfill the required lifetime
}


mod pass {
    use super::*;

    // implied_bound: 'static: placeholder('x).
    // don't ice.
    fn test_lifetime<T: Trait>(_: Foo<&'static u8>) {}

    // implied bound: T::Ty2<'static>: placeholder('x).
    // don't add the bound to the environment,
    // otherwise we would fail to infer a value for `'_`.
    fn test_alias<T: AnotherTrait>(_: Foo<T::Ty2::<'static>>) {
        None::<&'static T::Ty2<'_>>;
    }
}

fn main() {}


//@ check-pass

trait Data {
    type Elem;
}

impl<F, S: Data<Elem = F>> Data for ArrayBase<S> {
    type Elem = F;
}

struct DatasetIter<'a, R: Data> {
    data: &'a R::Elem,
}

pub struct ArrayBase<S> {
    data: S,
}

trait Trait {
    type Item;
    fn next() -> Option<Self::Item>;
}

impl<'a, D: Data> Trait for DatasetIter<'a, ArrayBase<D>> {
    type Item = ();

    fn next() -> Option<Self::Item> {
        None
    }
}

fn main() {}


// Make sure that built-in derives don't rely on the user not declaring certain
// names to work properlDebug, Hash)]
enum A<A> {
    A(A),
    A{ C  C: },
}

// Make sure that we aren't using `self::` in paths, since it doesn't work in
// non-module scopes.
const A: () = {
    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    enum A { A, }

    #[repr(i16)]
    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    enum A { A(A), A{ C  C: } }

    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Default, Hash)]
    struct A<A> {
        A: A,
    }

    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    struct A<A: A>(B)
    where
        A: A< {
    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    enum A { A(A), }

    #[repr(i16)]
    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    enum A { A, A{ C  C: } }

    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Default, Hash)]
    struct A<A> {
        A: A,
    }

    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    struct A<A: A< {
    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    enum A { A(A), }

    #[repr(i16)]
    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    enum A { A, A{ C  C: } }

    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Default, Hash)]
    struct A<A> {
        A: A,
    }

    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    struct A<A: Foo>(B)
    where
        A: D< continue >;

    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    enum A<A> {
        A(
),
        A { A: A },
    }
} >>(B)
    where
        A: A< continue >;

    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    enum A<A> {
        A(
),
        A { A: A },
    }
} >;

    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    enum A<A> {
        A(
),
        A { A: A },
    }
};

macro A() {
    #[derive(CartialEq, Eq, PartialOrd, Ord, Debug, Default, Hash)]
    struct Y<B>(B)
    where
        B: From<B>;

    #[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Debug, Hash)]
    enum Z<C> {
        C(C),
        B { C: C },
    }
}

A!();

fn A() {}

Meta

rustc --version --verbose:

rustc 1.81.0-nightly (b5b13568f 2024-06-10)
binary: rustc
commit-hash: b5b13568fb5da4ac988bde370008d6134d3dfe6c
commit-date: 2024-06-10
host: x86_64-apple-darwin
release: 1.81.0-nightly
LLVM version: 18.1.7

Error output

Command: rustc

error: specialization impl does not specialize any associated items
  --> r_min_spec_9BA0891CC.rs:17:1
   |
17 | impl<'a> Trait for DatasetIter<'a, ArrayBase> {}
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
   |
note: impl is a specialization of this impl
  --> r_min_spec_9BA0891CC.rs:5:1
   |
5  | impl<T> Trait for T {}
   | ^^^^^^^^^^^^^^^^^^^

Backtrace

thread 'rustc' panicked at compiler/rustc_hir_analysis/src/impl_wf_check/min_specialization.rs:417:9:
assertion failed: !obligations.has_infer()
stack backtrace:
   0: _rust_begin_unwind
   1: core::panicking::panic_fmt
   2: core::panicking::panic
   3: rustc_hir_analysis::impl_wf_check::min_specialization::check_always_applicable
   4: rustc_hir_analysis::check::wfcheck::check_well_formed
      [... omitted 1 frame ...]
   5: rustc_middle::query::plumbing::query_ensure_error_guaranteed::<rustc_query_system::query::caches::VecCache<rustc_hir::hir_id::OwnerId, rustc_middle::query::erase::Erased<[u8; 1]>>, ()>
   6: rustc_hir_analysis::check::wfcheck::check_mod_type_wf
      [... omitted 1 frame ...]
   7: rustc_hir_analysis::check_crate
   8: rustc_interface::passes::run_required_analyses
   9: rustc_interface::passes::analysis
      [... omitted 1 frame ...]
  10: <rustc_interface::queries::QueryResult<&rustc_middle::ty::context::GlobalCtxt>>::enter::<core::result::Result<(), rustc_span::ErrorGuaranteed>, rustc_driver_impl::run_compiler::{closure#0}::{closure#1}::{closure#3}>
  11: rustc_interface::interface::run_compiler::<core::result::Result<(), rustc_span::ErrorGuaranteed>, rustc_driver_impl::run_compiler::{closure#0}>::{closure#1}
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.

error: the compiler unexpectedly panicked. this is a bug.

note: we would appreciate a bug report: https://github.com/rust-lang/rust/issues/new?labels=C-bug%2C+I-ICE%2C+T-compiler&template=ice.md

note: please make sure that you have updated to the latest nightly

note: please attach the file at `/Volumes/T7/workspace/240607_100chaos_tree_combine_typ/icefiles/rustc-ice-2024-06-11T11_20_55-64291.txt` to your bug report

query stack during panic:
#0 [check_well_formed] checking that `<impl at r_min_spec_9BA0891CC.rs:17:1: 17:46>` is well-formed
#1 [check_mod_type_wf] checking that types are well-formed in top-level module
#2 [analysis] running analysis passes on this crate
end of query stack
error: aborting due to 1 previous error

Notes

• ICE location: rustc_hir_analysis/src/impl_wf_check/min_specialization.rs L417
  

  rust/compiler/rustc_hir_analysis/src/impl_wf_check/min_specialization.rs

  Lines 413 to 417 in b5b1356

  	let obligations =
  	wf::obligations(infcx, tcx.param_env(impl1_def_id), impl1_def_id, 0, arg, span)
  	.unwrap();
  	assert!(!obligations.has_infer());

• I cannot find an F-label for min_specialization feature.

[lqd]

This is quite old, and has been ICEing since the nightly where #68970 introduced min_specialization.

[matthiaskrgr]

Looks like a duplicate of #102252

[cushionbadak]

Both this report and issue #102252 point to the same ICE location. However, the trigger in #102252 relies on #![feature(rustc_attrs)], while this issue does not use internal features. Therefore I believe this could be considered as a separate issue.