Utility validate_type_for_device must flag complex128 as it flags double

[oleksandr-pavlyk]

Utility validate_type_for_device<T>(sycl::device &) modified to flag std::complex<double> for devices with no aspect::fp64.

• Have you provided a meaningful PR description?
• Have you tested your changes locally for CPU and GPU devices?

To test, I used:

In [1]: import dpctl.tensor as dpt, dpctl.tensor._tensor_impl as ti

In [2]: dpt.ones(10, dtype='c16')
---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
Input In [2], in <cell line: 1>()
----> 1 dpt.ones(10, dtype='c16')

File ~/repos/dpctl/dpctl/tensor/_ctors.py:708, in ones(sh, dtype, order, device, usm_type, sycl_queue)
    700 dtype = _get_dtype(dtype, sycl_queue)
    701 res = dpt.usm_ndarray(
    702     sh,
    703     dtype=dtype,
   (...)
    706     buffer_ctor_kwargs={"queue": sycl_queue},
    707 )
--> 708 hev, _ = ti._full_usm_ndarray(1, res, sycl_queue)
    709 hev.wait()
    710 return res

RuntimeError: Device Intel(R) Graphics [0x9a49] does not support type 'complex128'

In [3]: quit

[github-actions]

View rendered docs @ https://intelpython.github.io/dpctl/pulls/1040/index.html

[coveralls]

Coverage: 82.23%. Remained the same when pulling f69658f on expand-validate-type-for-device into 83ad22d on master.

[github-actions]

Array API standard conformance tests for dpctl=0.14.1dev1=py310h76be34b_22 ran successfully.
Passed: 33
Failed: 801
Skipped: 280

[github-actions]

Deleted rendered PR docs from intelpython.github.com/dpctl, latest should be updated shortly. 🤞

[github-actions]

Array API standard conformance tests for dpctl=0.14.1dev1=py310h76be34b_22 ran successfully.
Passed: 32
Failed: 802
Skipped: 280