Radix sort: handle more data types

Signed integers are currently mishandled:

>>> arr = cl.array.to_device(queue, np.array([-1, -2, 0], dtype=np.int32))
>>> RadixSort(ctx, ["int *arr"], "arr[i]", ["arr"], key_dtype=np.int32)(arr)
([array([ 0, -2, -1], dtype=int32)], <pyopencl._cl.Event object at 0x12e2da1a8>)

IEEE floating point could also be handled.

Edited May 30, 2019 by Andreas Klöckner

To upload designs, you'll need to enable LFS and have an admin enable hashed storage. More information