bpo-1635741: port _ctypes extension module to multiphase initialization

[kumaraditya303]

https://bugs.python.org/issue1635741

[shihai1991]

I am just curious about those codes in _ctypes_mod_exec, do we need judge the attr manually?

_unpickle = PyObject_GetAttrString(mod, "_unpickle");
    if (_unpickle == NULL) {
        return -1;
    }

[vstinner]

You should not convert an extension to the multi-phase init API until all static types defined by the extension are defined as heap types.

See _ctypes_add_types(): there are many types defined as static types. Just one example: PyCArg_Type.

[bedevere-bot]

A Python core developer has requested some changes be made to your pull request before we can consider merging it. If you could please address their requests along with any other requests in other reviews from core developers that would be appreciated.

Once you have made the requested changes, please leave a comment on this pull request containing the phrase I have made the requested changes; please review again. I will then notify any core developers who have left a review that you're ready for them to take another look at this pull request.

[erlend-aasland]

You should not convert an extension to the multi-phase init API until all static types defined by the extension are defined as heap types.

True; but sadly, converting static types to heap types has been put on hold by the SC. The good news is that work is in progress to resolve the "deadlock", so we can continue applying PEP 489 and PEP 630 to the standard library 😃

See SC communications as of February 2021:

"The Steering Council discussed the ongoing work on porting types in the standard library to heap-types and the subinterpreter-related changes. It was decided that through Pablo, the Steering Council will ask the core developers driving those changes to create an informational PEP and not to make any more changes in this area after beta 1, as per our general policy."

[vstinner]

ask the core developers driving those changes to create an informational PEP

Eric is working on a PEP: python/peps#2212

Not sure if it fulfill SC requirements to convert more static types to heap types.

[kumaraditya303]

IIUC converting static types to heap types is a different issue https://bugs.python.org/issue40077 which is not I am tackling in this PR, also itertools module also creates many static types but still uses multiphase initialization

cpython/Modules/itertoolsmodule.c

Line 4907 in be578e0

return PyModuleDef_Init(&itertoolsmodule);

so how is _ctypes different ?

[erlend-aasland]

It is subinterpreter related. Also, we first convert global state to module state, then apply multi-phase init, as Victor already pointed out.

Some modules have been converted to multi-phase init prematurely. That does not alter the status quo, or set a precedent.

[corona10]

FYI, Please read https://github.com/vstinner/talks/blob/main/2021-PyconUS/subinterpreters.pdf how we've converted modules :)

[corona10]

Same as Victor

[erlend-aasland]

FYI: python/steering-council#99

[erlend-aasland]

FYI, PEP-687 was just accepted.