Get rid of falling back on numpy in dpnp.put

[antonwolfy]

The PR proposes to get rid of call_origin (i.e. falling back on numpy) in dpnp.put function.
Additionally, the safe casting of indices to intp dtype is added to the implementation.

The tests are updated to remove temporary workaround of dpctl#1382, because it has been resolved.
Also they are separated to class TestPut and extended with more test scenarios.

• Have you provided a meaningful PR description?
• Have you added a test, reproducer or referred to issue with a reproducer?
• Have you tested your changes locally for CPU and GPU devices?
• Have you made sure that new changes do not introduce compiler warnings?
• Have you checked performance impact of proposed changes?
• If this PR is a work in progress, are you filing the PR as a draft?

[github-actions]

View rendered docs @ https://intelpython.github.io/dpnp/index.html

[vlad-perevezentsev]

@antonwolfy
1.It would be good if you could change test_put... in test_indexing.py by implementing them via class TestPut
2. Should we support the case where a and ind is a bool data type?
3.Would it make sense to add tests for test_sycl_queue and test_usm_type?

[antonwolfy]

1.It would be good if you could change test_put... in test_indexing.py by implementing them via class TestPut
2. Should we support the case where a and ind is a bool data type?
3.Would it make sense to add tests for test_sycl_queue and test_usm_type?

1. The existing tests are separated into class TestPut. Additionally the test scenarios are extended to cover more use cases.
2. The safe casting to intp dtype is added to the implementation. The tests are extended to cover that.
3. dpnp.put doesn't create any result array, but updating input array in-place. What kind of tests do you propose to add to test_sycl_queue and test_usm_type?

[vlad-perevezentsev]

LGTM!
Thank you @antonwolfy