Improve performance of dpnp.matmul and dpnp.dot with out keyword

[vtavana]

In this PR, implementation of dpnp.matmul and dpnp.dot is updated to improve the performance when out keyword is present and its feature matches the appropriate shape, dtype, sycl_queue, and usm_type that is needed for performing the calculation in the corresponding OneMKL BLAS routines.

Sample timing including this update

>>> import dpnp
>>> a = dpnp.ones((21846, 3), device='cpu')
>>> b = dpnp.ones((21846, 3), device='cpu')
>>> c = dpnp.ones((21846, 21846), device='cpu')

>>> %timeit dpnp.matmul(a, b.T, c)
229 ms ± 22 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

Sample timing excluding this update

>>> import dpnp
>>> a = dpnp.ones((21846, 3), device='cpu')
>>> b = dpnp.ones((21846, 3), device='cpu')
>>> c = dpnp.ones((21846, 21846), device='cpu')

>>> %timeit dpnp.matmul(a, b.T, c)
798 ms ± 30.7 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

• 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/pull//index.html

[npolina4]

Performance improved.
Thanks @vtavana!

[antonwolfy]

Thank you @vtavana for resolving the issue so quickly