Fixed data race in all_type_info in free-threading mode (#5419)

* Fix data race all_type_info_populate in free-threading mode
Description:
- fixed data race all_type_info_populate in free-threading mode
- added test

For example, we have 2 threads entering `all_type_info`.
Both enter `all_type_info_get_cache`` function and
there is a first one which inserts a tuple (type, empty_vector) to the map
and second is waiting. Inserting thread gets the (iter_to_key, True) and non-inserting thread
after waiting gets (iter_to_key, False).
Inserting thread than will add a weakref and will then call into `all_type_info_populate`.
However, non-inserting thread is not entering `if (ins.second) {` clause and
returns `ins.first->second;`` which is just empty_vector.
Finally, non-inserting thread is failing the check in `allocate_layout`:
```c++
if (n_types == 0) {
    pybind11_fail(
        "instance allocation failed: new instance has no pybind11-registered base types");
}
```

* style: pre-commit fixes

* Addressed PR comments

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
This commit is contained in:
vfdev 2024-11-07 18:32:09 +01:00 committed by GitHub
parent f46f5be4fa
commit ce2f005594
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
5 changed files with 67 additions and 12 deletions

View File

@ -117,7 +117,6 @@ PYBIND11_NOINLINE void all_type_info_populate(PyTypeObject *t, std::vector<type_
for (handle parent : reinterpret_borrow<tuple>(t->tp_bases)) { for (handle parent : reinterpret_borrow<tuple>(t->tp_bases)) {
check.push_back((PyTypeObject *) parent.ptr()); check.push_back((PyTypeObject *) parent.ptr());
} }
auto const &type_dict = get_internals().registered_types_py; auto const &type_dict = get_internals().registered_types_py;
for (size_t i = 0; i < check.size(); i++) { for (size_t i = 0; i < check.size(); i++) {
auto *type = check[i]; auto *type = check[i];
@ -176,13 +175,7 @@ PYBIND11_NOINLINE void all_type_info_populate(PyTypeObject *t, std::vector<type_
* The value is cached for the lifetime of the Python type. * The value is cached for the lifetime of the Python type.
*/ */
inline const std::vector<detail::type_info *> &all_type_info(PyTypeObject *type) { inline const std::vector<detail::type_info *> &all_type_info(PyTypeObject *type) {
auto ins = all_type_info_get_cache(type); return all_type_info_get_cache(type).first->second;
if (ins.second) {
// New cache entry: populate it
all_type_info_populate(type, ins.first->second);
}
return ins.first->second;
} }
/** /**

View File

@ -2326,13 +2326,20 @@ keep_alive_impl(size_t Nurse, size_t Patient, function_call &call, handle ret) {
inline std::pair<decltype(internals::registered_types_py)::iterator, bool> inline std::pair<decltype(internals::registered_types_py)::iterator, bool>
all_type_info_get_cache(PyTypeObject *type) { all_type_info_get_cache(PyTypeObject *type) {
auto res = with_internals([type](internals &internals) { auto res = with_internals([type](internals &internals) {
return internals auto ins = internals
.registered_types_py .registered_types_py
#ifdef __cpp_lib_unordered_map_try_emplace #ifdef __cpp_lib_unordered_map_try_emplace
.try_emplace(type); .try_emplace(type);
#else #else
.emplace(type, std::vector<detail::type_info *>()); .emplace(type, std::vector<detail::type_info *>());
#endif #endif
if (ins.second) {
// For free-threading mode, this call must be under
// the with_internals() mutex lock, to avoid that other threads
// continue running with the empty ins.first->second.
all_type_info_populate(type, ins.first->second);
}
return ins;
}); });
if (res.second) { if (res.second) {
// New cache entry created; set up a weak reference to automatically remove it if the type // New cache entry created; set up a weak reference to automatically remove it if the type

View File

@ -128,4 +128,9 @@ PYBIND11_MODULE(pybind11_tests, m, py::mod_gil_not_used()) {
for (const auto &initializer : initializers()) { for (const auto &initializer : initializers()) {
initializer(m); initializer(m);
} }
py::class_<TestContext>(m, "TestContext")
.def(py::init<>(&TestContext::createNewContextForInit))
.def("__enter__", &TestContext::contextEnter)
.def("__exit__", &TestContext::contextExit);
} }

View File

@ -96,3 +96,24 @@ void ignoreOldStyleInitWarnings(F &&body) {
)", )",
py::dict(py::arg("body") = py::cpp_function(body))); py::dict(py::arg("body") = py::cpp_function(body)));
} }
// See PR #5419 for background.
class TestContext {
public:
TestContext() = delete;
TestContext(const TestContext &) = delete;
TestContext(TestContext &&) = delete;
static TestContext *createNewContextForInit() { return new TestContext("new-context"); }
pybind11::object contextEnter() {
py::object contextObj = py::cast(*this);
return contextObj;
}
void contextExit(const pybind11::object & /*excType*/,
const pybind11::object & /*excVal*/,
const pybind11::object & /*excTb*/) {}
private:
explicit TestContext(const std::string &context) : context(context) {}
std::string context;
};

View File

@ -1,5 +1,6 @@
from __future__ import annotations from __future__ import annotations
import sys
from unittest import mock from unittest import mock
import pytest import pytest
@ -508,3 +509,31 @@ def test_pr4220_tripped_over_this():
m.Empty0().get_msg() m.Empty0().get_msg()
== "This is really only meant to exercise successful compilation." == "This is really only meant to exercise successful compilation."
) )
@pytest.mark.skipif(sys.platform.startswith("emscripten"), reason="Requires threads")
def test_all_type_info_multithreaded():
# See PR #5419 for background.
import threading
from pybind11_tests import TestContext
class Context(TestContext):
pass
num_runs = 10
num_threads = 4
barrier = threading.Barrier(num_threads)
def func():
barrier.wait()
with Context():
pass
for _ in range(num_runs):
threads = [threading.Thread(target=func) for _ in range(num_threads)]
for thread in threads:
thread.start()
for thread in threads:
thread.join()