`GPUArray.__div__` does not comply with numpy

[kaushikcfd]

Here's an MWE:

>>> import pycuda.autoinit
>>> import pycuda.gpuarray as cu_np
>>> a = cu_np.zeros(10, dtype="int32") + 1
>>> b = cu_np.zeros(10, dtype="int32") + 2
>>> a / b
array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=int32)

while, Numpy returns an array of 0.5's

>>> import numpy as np
>>> a = np.zeros(10, "int32") + 1
>>> b = np.zeros(10, "int32") + 2
>>> a / b
array([0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5])

[inducer]

That's clearly a bug. Believe it or not, PyCUDA predates __truediv__, I think. :astonished:

[kaushikcfd]

There's a proposed fix for this at https://gitlab.tiker.net/inducer/pycuda/-/merge_requests/66.