"cysignals.signals.SignalError: Segmentation fault" when using multiprocessing

[maxale]

Steps To Reproduce

The following code results in cysignals.signals.SignalError: Segmentation fault:

import multiprocessing
with multiprocessing.Pool() as pool:
    for t in pool.imap_unordered( print, (d for d in range(1,100) if is_squarefree(d)) ):
        pass

Expected Behavior

Segmentation fault should not happen.

Actual Behavior

The complete error message:

Traceback (most recent call last):
  File "/usr/lib/python3.10/threading.py", line 1016, in _bootstrap_inner
    self.run()
  File "/usr/lib/python3.10/threading.py", line 953, in run
    self._target(*self._args, **self._kwargs)
  File "/usr/lib/python3.10/multiprocessing/pool.py", line 535, in _handle_tasks
    for task in taskseq:
  File "/usr/lib/python3.10/multiprocessing/pool.py", line 391, in _guarded_task_generation
    for i, x in enumerate(iterable):
  File "/tmp/ipykernel_2667401/2212641732.py", line 4, in <genexpr>
  File "/home/sc_serv/sage/src/sage/arith/misc.py", line 2993, in is_squarefree
    return e.is_squarefree()
  File "sage/rings/integer.pyx", line 6090, in sage.rings.integer.Integer.is_squarefree
  File "sage/rings/integer.pyx", line 6182, in sage.rings.integer.Integer.__pari__
  File "sage/libs/pari/convert_sage.pyx", line 377, in sage.libs.pari.convert_sage.new_gen_from_integer
  File "sage/libs/pari/convert_sage.pyx", line 386, in sage.libs.pari.convert_sage.new_gen_from_integer
  File "sage/libs/pari/convert_gmp.pyx", line 52, in sage.libs.pari.convert_gmp.new_gen_from_mpz_t
cysignals.signals.SignalError: Segmentation fault

Additional Information

Somehow the issue is related to the use of is_squarefree(). Replacing it with, say, is_prime() makes the code works fine.

Environment

- **OS**: Ubuntu 22.04.3 LTS
- **Sage Version**: 10.2.beta1

Checklist

• I have searched the existing issues for a bug report that matches the one I want to file, without success.
• I have read the documentation and troubleshoot guide

[maxale]

The issue seems to be more widespread and not tied to is_squarefree(). I also see the following segmentation fault in different set-up when is_squarefree() is not used:

Exception in thread Thread-3 (_handle_results):
Traceback (most recent call last):
  File "/usr/lib/python3.10/threading.py", line 1016, in _bootstrap_inner
    self.run()
  File "/usr/lib/python3.10/threading.py", line 953, in run
    self._target(*self._args, **self._kwargs)
  File "/usr/lib/python3.10/multiprocessing/pool.py", line 579, in _handle_results
    task = get()
  File "/usr/lib/python3.10/multiprocessing/connection.py", line 251, in recv
    return _ForkingPickler.loads(buf.getbuffer())
  File "cypari2/gen.pyx", line 4706, in cypari2.gen.objtogen
  File "cypari2/gen.pyx", line 4813, in cypari2.gen.objtogen
  File "cypari2/convert.pyx", line 556, in cypari2.convert.PyObject_AsGEN
cysignals.signals.SignalError: Segmentation fault

[yyyyx4]

This seems to be related to PARI conversions: For example,

import multiprocessing
with multiprocessing.Pool() as pool:
    for t in pool.imap_unordered( print, (d for d in ZZ.range(1,100) if is_prime(d)) ):
        pass

works, but

import multiprocessing
with multiprocessing.Pool() as pool:
    for t in pool.imap_unordered( print, (d for d in ZZ.range(2^99+1,2^99+100) if is_prime(d)) ):
        pass

fails. The difference is that a fast path is used for small integers whereas PARI is invoked for larger sizes.

Smaller failing example:

sage: import multiprocessing
sage: list(multiprocessing.Pool().map(print, [pari(42)]))

[yyyyx4]

Is this a duplicate of #28800?