Function compilation not being updated for specific types on 32-bit systems

[kc611]

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'.

Note: This is only observed in 32-bit systems and not 64-bit ones.

import numba
import numpy as np

@numba.njit
def foo(x):
    print("from within: ",x)
    return x

foo(0)
print( foo( np.iinfo(np.int32).max + 1) )
# Returns a negative integer (overflow), the function signature stays as `int32` argument instead of an `int64`/`uint32`

First encountered in #8041

[sklam]

Note:

• check typeof logic in dispatching. seems like it's always picking intp.