Backtestspydanticrun-20260930-144322#13419
"dictionary changed size during iteration" during model_rebuild
The merged fix was judged better than our patch.
In this judgment, Patch A = our agent, Patch B = the merged fix.
Root cause: `model_rebuild()` is not thread-safe. Several threads rebuild the same incomplete class at once. One mutates the class `__dict__` (delattr/setattr of schema, validator and serializer) while another iterates that live `vars(cls)` in `LazyLocalNamespace.data`, or races on the check-then-delete. Patch A treats the two symptoms separately. It swaps the dict comprehension for `dict.update()`, relying on C-level iteration not releasing the GIL; the comment itself admits this holds only on GIL-enabled builds. It also swallows `AttributeError` from `delattr`. The rebuild still runs concurrently, so multiple threads redo schema generation, can call `__pydantic_on_complete__` more than once, and can briefly observe missing or parent attributes. On free-threaded Python (the issue is on 3.14) the dict race can still occur. Its tests are clever and deterministic (trace-function injection) and would fail before the fix, but they check the narrow symptom only. Patch B fixes the cause by serializing rebuilds with a global reentrant lock. It keeps an unlocked fast path and re-checks completeness inside the lock, so a thread that waited returns `None` instead of rebuilding again. The class-dict mutations and the namespace iteration during another rebuild both happen under the lock, which removes both the RuntimeError and the AttributeError without depending on GIL atomicity. It also explains its design choices: an RLock for the nested rebuild triggered via generic `__class_getitem__`, and a global rather than per-class lock to avoid deadlocks. It tightens the mock check to include `MockCoreSchema` and applies the same check in `rebuild_dataclass`. B's tests cover reentrancy during parametrization and a deterministic two-thread scenario. That scenario asserts the second thread returns `None` and the completion hook fires once; it would fail without the lock. B carries some risks: a global lock could deadlock if user schema hooks wait on other threads, and class creation and dataclass rebuild are not locked. Even so, B addresses the actual concurrency problem, while A mostly masks symptoms under GIL assumptions.
The judge saw only the two patches, labelled A and B in random order, with issue links and version numbers removed. The reasoning above is shown verbatim.
Solve
4m 43s14 turns
Wall-clock time in the container
Full test suite
No new failures
12,261 passed · 501 skipped · 12,762 total
Contamination audit
Clean
No sign the agent saw the real fix
Patch size
98vs 168 merged
Changed lines, ours vs merged
The issue
View on GitHubInitial Checks
- I confirm that I'm using Pydantic V2
Description
This just hit us in production:
Traceback (most recent call last):
...
File "fastapi/routing.py", line 120, in app
response = await f(request)
File "fastapi/routing.py", line 674, in app
raw_response = await run_endpoint_function(
File "fastapi/routing.py", line 330, in run_endpoint_function
return await run_in_threadpool(dependant.call, **values)
File "starlette/concurrency.py", line 34, in run_in_threadpool
return await anyio.to_thread.run_sync(func)
File "anyio/to_thread.py", line 63, in run_sync
return await get_async_backend().run_sync_in_worker_thread(
File "anyio/_backends/_asyncio.py", line 2518, in run_sync_in_worker_thread
return await future
File "anyio/_backends/_asyncio.py", line 1002, in run
result = context.run(func, *args)
...
File "...", line 157, in ...
MyPydanticClass(
File "pydantic/_internal/_mock_val_ser.py", line 165, in handler
if cls.model_rebuild(raise_errors=False, _parent_namespace_depth=5) is not False:
File "pydantic/main.py", line 683, in model_rebuild
return _model_construction.complete_model_class(
File "pydantic/_internal/_model_construction.py", line 642, in complete_model_class
cls.__pydantic_fields__, cls.__pydantic_extra_info__ = rebuild_model_fields(
File "pydantic/_internal/_fields.py", line 464, in rebuild_model_fields
new_field = _recreate_field_info(
File "pydantic/_internal/_fields.py", line 502, in _recreate_field_info
ann = _typing_extra.eval_type(
File "pydantic/_internal/_typing_extra.py", line 442, in eval_type
return eval_type_backport(value, globalns, localns)
File "pydantic/_internal/_typing_extra.py", line 474, in eval_type_backport
return _eval_type_backport(value, globalns, localns, type_params)
File "pydantic/_internal/_typing_extra.py", line 511, in _eval_type_backport
return _eval_type(value, globalns, localns, type_params)
File "pydantic/_internal/_typing_extra.py", line 543, in _eval_type
evaluated = typing._eval_type( # type: ignore
File "typing.py", line 461, in _eval_type
return evaluate_forward_ref(t, globals=globalns, locals=localns,
File "typing.py", line 1000, in evaluate_forward_ref
value = forward_ref.evaluate(globals=globals, locals=locals,
File "annotationlib.py", line 207, in evaluate
return eval(code, globals=globals, locals=locals)
File "<string>", line 1, in <module>
from __future__ import annotations
File "pydantic/_internal/_namespace_utils.py", line 97, in __getitem__
return self.data[key]
File "functools.py", line 1126, in __get__
val = self.func(instance)
File "pydantic/_internal/_namespace_utils.py", line 91, in data
return {k: v for ns in self._namespaces for k, v in ns.items()}
RuntimeError: dictionary changed size during iteration
Reproducer below.
Example Code
"""
NOTE: Repro script generated with Claude Fable
Repro: concurrent lazy `model_rebuild()` raises
`RuntimeError: dictionary changed size during iteration`.
Setup: a module using `from __future__ import annotations` defines models whose
field types are only defined *later* in the same module. Building such a model's
schema fails during class creation, so pydantic leaves it incomplete (mock
validator) and defers the build to the first instantiation via
`model_rebuild(raise_errors=False)`.
If the first instantiations happen concurrently (e.g. FastAPI's threadpool
serving the first requests after process start), multiple threads enter
`model_rebuild()` for the same class simultaneously. `model_rebuild()` mutates
the class `__dict__` (delattr/setattr of `__pydantic_core_schema__`,
`__pydantic_validator__`, `__pydantic_serializer__`), while
`NsResolver.types_namespace` includes the live `vars(cls)` of that class in the
`LazyLocalNamespace` another thread is iterating during annotation evaluation
(`LazyLocalNamespace.data`). Result:
RuntimeError: dictionary changed size during iteration
The same race also intermittently produces
AttributeError: type object 'Box0' has no attribute '__pydantic_core_schema__'
from the delattr block at the top of `model_rebuild()`.
Each round executes the module source into a fresh module (fresh, incomplete
classes -- like a freshly started process) and races threads instantiating the
models. Typically fires within a few dozen rounds (< 1 min).
Run: python pydantic_rebuild_race_repro.py
"""
import sys
import threading
import time
import traceback
import types
import pydantic
print(f"python {sys.version}")
print(f"pydantic {pydantic.VERSION}")
sys.setswitchinterval(1e-5) # more frequent thread preemption; widens the race window
N_GENERICS = 2 # generic origins shared by all models (parametrizing triggers origin.model_rebuild)
N_LEAVES = 16 # leaf models, defined at the end of the module so everything before stays incomplete
N_MODELS = 4 # top-level models, the ones instantiated by the racing threads
N_METHODS = 60 # dummy methods to enlarge each class __dict__ (the dict iterated during annotation eval)
N_THREADS = 32
ROUNDS = 200
RACE_WINDOW_S = 0.5
def make_module_source() -> str:
boxes = " | ".join(f"Box{g}[Leaf{l}]" for g in range(N_GENERICS) for l in range(N_LEAVES))
methods = [f" def method{m}(self) -> None: ..." for m in range(N_METHODS)]
lines = ["from __future__ import annotations", "from pydantic import BaseModel", ""]
# Top-level models come first: all their field types are still undefined here,
# so their schema build is deferred until first instantiation.
for i in range(N_MODELS):
lines += [f"class Model{i}(BaseModel):", f" items: list[{boxes}] = []", *methods, ""]
for g in range(N_GENERICS):
lines += [f"class Box{g}[T](BaseModel):", " value: T | None = None", " ref: Leaf0 | None = None", *methods, ""]
for l in range(N_LEAVES):
lines += [f"class Leaf{l}(BaseModel):", " x: int = 0", ""]
return "\n".join(lines)
MODULE_SOURCE = make_module_source()
errors: list[str] = []
errors_lock = threading.Lock()
def worker(barrier: threading.Barrier, model: type) -> None:
barrier.wait()
end = time.monotonic() + RACE_WINDOW_S
while time.monotonic() < end:
try:
model() # first call hits the mock validator -> lazy model_rebuild()
except Exception:
with errors_lock:
errors.append(traceback.format_exc())
for rnd in range(ROUNDS):
# Fresh module => fresh incomplete classes, like a freshly started process.
mod = types.ModuleType(f"repro_generated_{rnd}")
sys.modules[mod.__name__] = mod
exec(compile(MODULE_SOURCE, mod.__name__, "exec"), mod.__dict__)
models = [getattr(mod, f"Model{i}") for i in range(N_MODELS)]
assert not models[0].__pydantic_complete__
barrier = threading.Barrier(N_THREADS)
threads = [threading.Thread(target=worker, args=(barrier, models[i % N_MODELS])) for i in range(N_THREADS)]
for t in threads:
t.start()
for t in threads:
t.join()
hits = [tb for tb in errors if "changed size during iteration" in tb]
if hits:
print(hits[0], file=sys.stderr)
print(f"REPRODUCED in round {rnd}: RuntimeError: dictionary changed size during iteration")
sys.exit(1)
if errors:
first_line = errors[0].strip().splitlines()[-1]
print(f"round {rnd}: {len(errors)} sibling race errors (e.g. {first_line})")
errors.clear()
print("not reproduced")
Python, Pydantic & OS Version
In Prod: Ubuntu Docker, in Dev:
pydantic version: 2.13.4
pydantic-core version: 2.46.4
pydantic-core build: profile=release pgo=false
python version: 3.14.6 | packaged by conda-forge | (main, Jun 12 2026, 08:47:56) [Clang 20.1.8 ]
platform: macOS-26.5.1-arm64-arm-64bit-Mach-O
related packages: pyright-1.1.410 fastapi-0.136.3 pydantic-extra-types-2.11.1 email-validator-2.3.0 pydantic-settings-2.14.1 typing_extensions-4.15.0
commit: unknownThe two patches
pydantic/_internal/_namespace_utils.pyindex d308f7e4..fa003ac9 100644@@ -88,7 +88,14 @@ class LazyLocalNamespace(Mapping[str, Any]): @cached_property def data(self) -> dict[str, Any]:- return {k: v for ns in self._namespaces for k, v in ns.items()}+ # Using `dict.update()` instead of iterating over the namespaces in Python code, as the+ # namespaces (e.g. a class `__dict__`) can be mutated concurrently by another thread+ # (e.g. during concurrent `model_rebuild()` calls). `dict.update()` runs in C, so no+ # thread switch can happen during the iteration (at least on GIL-enabled builds):+ data: dict[str, Any] = {}+ for ns in self._namespaces:+ data.update(ns)+ return data def __len__(self) -> int: return len(self.data)pydantic/main.pyindex 758a839c..50fcee5e 100644@@ -693,7 +693,11 @@ class BaseModel(metaclass=_model_construction.ModelMetaclass): # pydantic-core. We do so only if they aren't mock instances, otherwise — as `model_rebuild()` # isn't thread-safe — concurrent model instantiations can lead to the parent validator being used. # Same applies for the core schema that can be reused in schema generation.- delattr(cls, attr)+ try:+ delattr(cls, attr)+ except AttributeError:+ # Can happen if the attribute was deleted concurrently by another thread:+ pass if _types_namespace is not None: rebuild_ns = _types_namespacetests/test_forward_ref.pyindex 9b0325d2..11a5ef78 100644@@ -1144,6 +1144,48 @@ def test_lazy_local_namespace_len() -> None: assert len(namespace) == 2 +def test_lazy_local_namespace_concurrent_mutation() -> None:+ """The namespace can wrap a live class `__dict__`, which can be mutated by another thread+ while `LazyLocalNamespace.data` is being computed. Simulate this deterministically+ by mutating the class dict from a trace function during the computation.+ """++ class Cls:+ a = int++ namespace = LazyLocalNamespace(vars(Cls))+ data_code = LazyLocalNamespace.data.func.__code__+ counter = 0++ def local_trace(frame, event, arg):+ nonlocal counter+ if event in {'line', 'opcode'} and counter < 100:+ # Simulate another thread adding an attribute to the class:+ setattr(Cls, f'new_attr_{counter}', counter)+ counter += 1+ return local_trace++ def global_trace(frame, event, arg):+ # On Python < 3.12, the dict comprehension runs in a nested `<dictcomp>` frame:+ if frame.f_code is data_code or (+ frame.f_code.co_name == '<dictcomp>' and frame.f_back is not None and frame.f_back.f_code is data_code+ ):+ frame.f_trace_opcodes = True+ return local_trace+ return None++ old_trace = sys.gettrace()+ sys.settrace(global_trace)+ try:+ data = namespace.data+ finally:+ sys.settrace(old_trace)++ assert counter > 0+ assert data['a'] is int+ assert namespace['a'] is int++ @pytest.mark.xfail( reason='While `get_cls_type_hints` uses the correct module ns for each base, `collect_model_fields` ' 'will still use the `FieldInfo` instances from each base (see the `parent_fields_lookup` logic). 'tests/test_main.pyindex c87dae79..c8959341 100644@@ -1,4 +1,5 @@ import json+import linecache import platform import re import sys@@ -2557,6 +2558,46 @@ def test_model_rebuild_zero_depth(): assert m.model_dump() == {'x': 42} +def test_model_rebuild_concurrent_attribute_deletion() -> None:+ """`model_rebuild()` deletes the core schema/validator/serializer attributes if present.++ If another thread deletes one of them between the presence check and the deletion,+ no `AttributeError` should be raised. Simulate this deterministically with a trace function.+ """++ class Model(BaseModel):+ x: int++ assert '__pydantic_core_schema__' in Model.__dict__++ rebuild_code = BaseModel.model_rebuild.__func__.__code__+ deleted: list[str] = []++ def local_trace(frame, event, arg):+ if event == 'line' and linecache.getline(frame.f_code.co_filename, frame.f_lineno).strip() == (+ 'delattr(cls, attr)'+ ):+ attr = frame.f_locals['attr']+ if attr == '__pydantic_core_schema__' and attr in Model.__dict__:+ # Simulate another thread deleting the attribute right before `delattr()`:+ delattr(Model, attr)+ deleted.append(attr)+ return local_trace++ def global_trace(frame, event, arg):+ return local_trace if frame.f_code is rebuild_code else None++ old_trace = sys.gettrace()+ sys.settrace(global_trace)+ try:+ assert Model.model_rebuild(force=True) is True+ finally:+ sys.settrace(old_trace)++ assert deleted == ['__pydantic_core_schema__']+ assert Model(x=1).x == 1++ @pytest.fixture(scope='session', name='InnerEqualityModel') def inner_equality_fixture(): class InnerEqualityModel(BaseModel):
pydantic/dataclasses.pyindex 62d380f54..474289ec3 100644@@ -330,9 +330,13 @@ def rebuild_dataclass( return None for attr in ('__pydantic_core_schema__', '__pydantic_validator__', '__pydantic_serializer__'):- if attr in cls.__dict__ and not isinstance(getattr(cls, attr), _mock_val_ser.MockValSer):+ if attr in cls.__dict__ and not isinstance(+ getattr(cls, attr), (_mock_val_ser.MockCoreSchema, _mock_val_ser.MockValSer)+ ): # Deleting the validator/serializer is necessary as otherwise they can get reused in # pydantic-core. Same applies for the core schema that can be reused in schema generation.+ # We do so only if they aren't mock instances, otherwise concurrent reads of these attributes+ # (e.g. when instantiating the dataclass in another thread) can resolve them from the parent class. delattr(cls, attr) cls.__pydantic_complete__ = Falsepydantic/main.pyindex 758a839c6..6bb1554e5 100644@@ -8,6 +8,7 @@ from __future__ import annotations as _annotations import operator import sys+import threading import types import warnings from collections.abc import Generator, Mapping@@ -78,6 +79,18 @@ IncEx: TypeAlias = set[int] | set[str] | Mapping[int, Union['IncEx', bool]] | Ma _object_setattr = _model_construction.object_setattr +_rebuild_lock = threading.RLock()+"""+A lock used to make model rebuilds thread-safe when first instantiating an incomplete model.++Rebuilding a model isn't thread-safe (the class attributes are mutated during the rebuild,+while other threads may be reading them to perform validation/serialization), so `model_rebuild()`+calls are serialized using this lock. The lock is reentrant, as rebuilding a model can trigger the+rebuild of another one (e.g. when a generic origin is rebuilt during parametrization in+`__class_getitem__()`). For the same reason, the lock is global and not per-class: two threads+holding their own class's lock could otherwise request the other's and deadlock.+"""+ def _check_frozen(model_cls: type[BaseModel], name: str, value: Any) -> None: if model_cls.model_config.get('frozen'):@@ -681,42 +694,54 @@ class BaseModel(metaclass=_model_construction.ModelMetaclass): Returns `None` if the schema is already "complete" and rebuilding was not required. If rebuilding _was_ required, returns `True` if rebuilding was successful, otherwise `False`. """- already_complete = cls.__pydantic_complete__- if already_complete and not force:+ # As the rebuild lock is global (not per model class), avoid needlessly holding it if the model is+ # already complete:+ if cls.__pydantic_complete__ and not force: return None - cls.__pydantic_complete__ = False+ with _rebuild_lock:+ # Re-check inside the lock, as another thread may have rebuilt the model while we were waiting:+ already_complete = cls.__pydantic_complete__+ if already_complete and not force:+ return None - for attr in ('__pydantic_core_schema__', '__pydantic_validator__', '__pydantic_serializer__'):- if attr in cls.__dict__ and not isinstance(getattr(cls, attr), _mock_val_ser.MockValSer):- # Deleting the validator/serializer is necessary as otherwise they can get reused in- # pydantic-core. We do so only if they aren't mock instances, otherwise — as `model_rebuild()`- # isn't thread-safe — concurrent model instantiations can lead to the parent validator being used.- # Same applies for the core schema that can be reused in schema generation.- delattr(cls, attr)+ cls.__pydantic_complete__ = False - if _types_namespace is not None:- rebuild_ns = _types_namespace- elif _parent_namespace_depth > 0:- rebuild_ns = _typing_extra.parent_frame_namespace(parent_depth=_parent_namespace_depth, force=True) or {}- else:- rebuild_ns = {}+ for attr in ('__pydantic_core_schema__', '__pydantic_validator__', '__pydantic_serializer__'):+ if attr in cls.__dict__ and not isinstance(+ getattr(cls, attr), (_mock_val_ser.MockCoreSchema, _mock_val_ser.MockValSer)+ ):+ # Deleting the validator/serializer is necessary as otherwise they can get reused in+ # pydantic-core. Same applies for the core schema that can be reused in schema generation.+ # We do so only if they aren't mock instances, otherwise concurrent reads of these attributes+ # — performed without holding the rebuild lock (e.g. when instantiating the model) — can+ # resolve them from the parent class.+ delattr(cls, attr)++ if _types_namespace is not None:+ rebuild_ns = _types_namespace+ elif _parent_namespace_depth > 0:+ rebuild_ns = (+ _typing_extra.parent_frame_namespace(parent_depth=_parent_namespace_depth, force=True) or {}+ )+ else:+ rebuild_ns = {} - parent_ns = _model_construction.unpack_lenient_weakvaluedict(cls.__pydantic_parent_namespace__) or {}+ parent_ns = _model_construction.unpack_lenient_weakvaluedict(cls.__pydantic_parent_namespace__) or {} - ns_resolver = _namespace_utils.NsResolver(- parent_namespace={**rebuild_ns, **parent_ns},- )+ ns_resolver = _namespace_utils.NsResolver(+ parent_namespace={**rebuild_ns, **parent_ns},+ ) - return _model_construction.complete_model_class(- cls,- _config.ConfigWrapper(cls.model_config, check=False),- ns_resolver,- raise_errors=raise_errors,- # If the model was already complete, we don't need to call the hook again.- call_on_complete_hook=not already_complete,- is_force_rebuild=force,- )+ return _model_construction.complete_model_class(+ cls,+ _config.ConfigWrapper(cls.model_config, check=False),+ ns_resolver,+ raise_errors=raise_errors,+ # If the model was already complete, we don't need to call the hook again.+ call_on_complete_hook=not already_complete,+ is_force_rebuild=force,+ ) @classmethod def model_validate(tests/test_main.pyindex c87dae793..f924cec99 100644@@ -2,6 +2,7 @@ import json import platform import re import sys+import threading import typing import warnings from collections import defaultdict@@ -2541,6 +2542,84 @@ def test_model_rebuild_localns(): C.model_rebuild(_types_namespace={'A': A}) +@pytest.mark.timeout(5)+def test_model_rebuild_nested_during_parametrization():+ T = TypeVar('T')++ class Model(BaseModel):+ b: Optional['Box[Leaf]'] = None++ class Box(BaseModel, Generic[T]):+ leaf: Optional['Leaf'] = None++ class Leaf(BaseModel):+ x: int = 0++ assert not Model.__pydantic_complete__+ assert not Box.__pydantic_complete__++ # Rebuilding `Model` evaluates `Box[Leaf]`, and `__class_getitem__` then calls+ # `Box.model_rebuild()` on the same thread. Any locking added to the rebuild logic+ # must support this reentrancy:+ m = Model(b={'leaf': {'x': 1}})+ assert m.b.leaf.x == 1+ assert Model.__pydantic_complete__+++@pytest.mark.skipif(sys.platform == 'emscripten', reason='no threading on emscripten')+def test_model_rebuild_already_rebuilt_by_other_thread():+ schema_gen_entered = threading.Event()+ resume_schema_gen = threading.Event()+ hook_calls: list[type] = []++ class Model(BaseModel):+ x: 'Gate'++ @classmethod+ def __pydantic_on_complete__(cls) -> None:+ hook_calls.append(cls)++ class Gate:+ @classmethod+ def __get_pydantic_core_schema__(cls, source_type, handler):+ schema_gen_entered.set()+ resume_schema_gen.wait(timeout=10)+ return core_schema.int_schema()++ assert not Model.__pydantic_complete__++ results: dict[str, bool | None] = {}+ t2_started = threading.Event()++ def rebuild(key: str) -> None:+ if key == 't2':+ t2_started.set()+ results[key] = Model.model_rebuild(_types_namespace={'Gate': Gate})++ # The first thread acquires the rebuild lock and waits in schema generation:+ t1 = threading.Thread(target=rebuild, args=('t1',), daemon=True)+ t1.start()+ assert schema_gen_entered.wait(timeout=10)++ # While the first thread is waiting on `resume_schema_gen`, `Model` can't become complete,+ # so the second thread is guaranteed to pass the unlocked completeness check and block on the lock:+ t2 = threading.Thread(target=rebuild, args=('t2',), daemon=True)+ t2.start()+ assert t2_started.wait(timeout=10)+ assert t2.is_alive()+ assert 't2' not in results++ resume_schema_gen.set()+ t1.join(timeout=10)+ t2.join(timeout=10)+ assert not t1.is_alive()+ assert not t2.is_alive()++ assert results == {'t1': True, 't2': None}+ assert hook_calls == [Model]+ assert Model(x=1).x == 1++ def test_model_rebuild_zero_depth(): class Model(BaseModel): x: 'X_Type'
What the agent reported
I fixed both errors from the report. The changes are left as uncommitted edits.
What was wrong
When several threads create the same not-yet-built model at the same moment (like FastAPI's first requests after startup), each thread calls model_rebuild() on that class.
RuntimeError: dictionary changed size during iteration: while evaluating annotations,LazyLocalNamespace.datainpydantic/_internal/_namespace_utils.pycopied the class's live__dict__(vars(cls)) using a Python dict comprehension. Python can switch threads between steps of that loop. Meanwhile another thread'smodel_rebuild()was deleting and setting__pydantic_core_schema__,__pydantic_validator__and__pydantic_serializer__on the same class.- The occasional
AttributeError: ... has no attribute '__pydantic_core_schema__':model_rebuild()inpydantic/main.pyfirst checks that the attribute is incls.__dict__and then deletes it. Another thread could delete it between those two steps.
What I changed
LazyLocalNamespace.datanow builds its dict withdata.update(ns)for each namespace instead of the comprehension.dict.updateruns entirely in C, so on normal CPython no other thread can run while the copy is made. Later namespaces still override earlier ones, as before.- On free-threaded CPython builds (3.13t/3.14t) the copy is not fully protected. A badly timed deletion there could still raise a
KeyError. The code comment says so. model_rebuild(): thedelattris wrapped intry/except AttributeError, so an attribute another thread already deleted is simply skipped.
Test evidence
I added two tests. Each one makes the "other thread" act at the exact point of the race, using a trace function (sys.settrace) rather than real threads. Neither mocks any pydantic code.
tests/test_forward_ref.py::test_lazy_local_namespace_concurrent_mutationadds attributes to the class whileLazyLocalNamespace.datais being built.- Before the fix:
RuntimeError: dictionary changed size during iteration. - After the fix: passes.
tests/test_main.py::test_model_rebuild_concurrent_attribute_deletiondeletes__pydantic_core_schema__just beforemodel_rebuild's owndelattrruns.- Before the fix:
AttributeError: type object 'Model' has no attribute '__pydantic_core_schema__'. - After the fix: passes. It also checks that the model still validates afterwards.
- It finds that moment by matching the source line
delattr(cls, attr), so reformatting that line will make the test fail until it is updated. ./bt-test tests/test_forward_ref.py tests/test_main.py tests/test_generics.py tests/test_internal.py tests/test_type_adapter.py: 678 passed, 27 skipped, 12 xfailed (xfail = expected to fail). The final run on the first two files after a comment tweak: 321 passed, 25 skipped, 5 xfailed.- I could not run the reporter's threaded script.
pydantic_coreisn't installed in the local Python, and the environment./bt-testuses isn't reachable any other way. So the evidence is the two tests above, not a real thread-race run.
Requirements
| Item | Outcome |
|---|---|
RuntimeError: dictionary changed size during iteration in LazyLocalNamespace.data during concurrent lazy model_rebuild() | Fixed; regression test fails before and passes after |
The related AttributeError ... '__pydantic_core_schema__' from the delete step in model_rebuild() | Fixed; regression test fails before and passes after |
| The reporter's threaded reproducer (FastAPI threadpool scenario) | Covered by the two tests; the script itself couldn't be run here |
| Existing behaviour unchanged | Adjacent test files all pass; no existing tests were modified |
A final review of the diff found no blocking issues. model_rebuild() is still not fully thread-safe as a whole; this change only removes the two crashes in the report.