[BUG]: gil_safe_call_once_and_store is not NoGil thread safe.

[albanD]

Required prerequisites

• Make sure you've read the documentation. Your issue may be addressed there.
• Search the issue tracker and Discussions to verify that this hasn't already been reported. +1 or comment there if it has.
• Consider asking first in the Gitter chat room or in a Discussion.

What version (or hash if on master) of pybind11 are you using?

0e2c3e5

Problem description

In the function,

pybind11/include/pybind11/gil_safe_call_once.h

Lines 51 to 59 in ccefee4

	if (!is_initialized_) { // This read is guarded by the GIL.
	// Multiple threads may enter here, because the GIL is released in the next line and
	// CPython API calls in the `fn()` call below may release and reacquire the GIL.
	gil_scoped_release gil_rel; // Needed to establish lock ordering.
	std::call_once(once_flag_, [&] {
	// Only one thread will ever enter here.
	gil_scoped_acquire gil_acq;
	::new (storage_) T(fn()); // fn may release, but will reacquire, the GIL.
	is_initialized_ = true; // This write is guarded by the GIL.

is_initialized_ is read outside of the critical section without any synchronization. Which can lead to a race in a NoGil build.

Reproducible example code

N/A

Is this a regression? Put the last known working version here if it is.

Not a regression

[albanD]

Thanks for the quick fix !