segfault (crash) when calling numpy np.isnan repeatedly in nopython parallel function

[limwz01]

Reporting a bug

• I have tried using the latest released version of Numba (most recent is
  visible in the change log (https://github.com/numba/numba/blob/main/CHANGE_LOG).
• I have included a self contained code sample to reproduce the problem.
  i.e. it's possible to run as 'python bug.py'.

numba crashes when calling numpy np.isnan repeatedly in nopython parallel function:

#!/usr/bin/env python3
import numba
import numpy as np


def main():
    bug()


@numba.njit(parallel=True)
def bug():
    for j in range(50000):
        np.isnan(np.full(1, 0.0))


if __name__ == "__main__":
    main()

Result with Python 3.8 on Ubuntu 20.04 (other versions may need different number of iterations?):

Segmentation fault (core dumped)

[sklam]

I can confirm the segfault on OSX with py3.9 after increasing number of iteration to 5000000.

[sklam]

This is another instance of #9029 and will be fixed by #9048.