pydantic/monty · error

test function has an owned cell

Error message

test function has an owned cell

What it means

This panic comes from an `expect` in `FunctionMetadata::corrupt_metadata_for_tests`, a test-only fault injector that deliberately corrupts function metadata to test deserializer validation. It fires when the `CellVarLengthMismatch` fault is requested but the target function metadata has no cell param indices to pop, meaning the test fixture was built without an owned cell variable. It indicates the test setup, not the code under test, is wrong.

Source

Thrown at crates/monty/src/function.rs:322

            None
        }
    }

    /// Injects a metadata fault for dump validation tests.
    #[cfg(feature = "test-hooks")]
    pub(crate) fn corrupt_metadata_for_tests(&mut self, fault: FunctionMetadataFault) {
        match fault {
            FunctionMetadataFault::SignatureSlotsBeyondNamespace => {
                self.namespace_size = self.signature.total_slots() - 1;
            }
            FunctionMetadataFault::NamespaceTooLarge => {
                self.namespace_size = usize::from(u16::MAX) + 1;
            }
            FunctionMetadataFault::FreeVarLengthMismatch => {
                self.free_var_slots.pop().expect("test function has a free variable");
            }
            FunctionMetadataFault::CellVarLengthMismatch => {
                self.cell_param_indices.pop().expect("test function has an owned cell");
            }
            FunctionMetadataFault::FreeVarSlotOutOfRange => {
                let slot = NamespaceId::new(self.namespace_size).expect("test namespace fits in u16");
                *self
                    .free_var_slots
                    .first_mut()
                    .expect("test function has a free variable") = slot;
            }
            FunctionMetadataFault::CellVarSlotOutOfRange => {
                let slot = NamespaceId::new(self.namespace_size).expect("test namespace fits in u16");
                *self
                    .cell_var_slots
                    .first_mut()
                    .expect("test function has an owned cell") = slot;
            }
            FunctionMetadataFault::CellParamIndexOutOfRange => {
                *self
                    .cell_param_indices

View on GitHub (pinned to adc986b362)

Solutions

1. Fix the test fixture so the function metadata actually contains a cell param index before applying the fault (compile a function with a closure capturing a cell).
2. Pick a different fault variant applicable to metadata that exists (e.g. namespace-size faults).
3. Guard the fault application with an `if !self.cell_param_indices.is_empty()` check in the fault injector or skip the case in the test matrix.

Example fix

// before
fn fixture() -> FunctionMetadata { /* function without cells */ make_metadata('f') }
let mut m = fixture();
m.corrupt_metadata_for_tests(FunctionMetadataFault::CellVarLengthMismatch); // panics
// after
let mut m = make_metadata_with_cell('f'); // function capturing a cell variable
m.corrupt_metadata_for_tests(FunctionMetadataFault::CellVarLengthMismatch);

assert!(!metadata.cell_param_indices.is_empty(), 'fixture must have a cell param for CellVarLengthMismatch');

When it happens

Trigger: Calling `corrupt_metadata_for_tests(FunctionMetadataFault::CellVarLengthMismatch)` on function metadata whose `cell_param_indices` vec is empty (a function compiled without cell variables).

Common situations: Writing a new roundtrip/deserialization fuzz or fault test against a helper function that has no closures or cell variables; refactoring a test fixture so the function no longer captures cells while the fault list still assumes one.

Understand the failure class

Background: "This is a bug, please report it": internal invariant violations, unreachable panics, and SNH errors explained — this error's family across 47 libraries.

Related errors

• test namespace fits in u16
• entry just allocated
• gather commit frame id is not a GatherFuture
• gather item is not a Coroutine, ExternalFuture, or GatherFut
• Cannot get identity of Dereferenced object

AI-assisted analysis of pydantic/monty@adc986b362 (2026-09-13). Data as JSON: /api/errors/77bc99185cf48b46. Report an issue: GitHub.