order of output array for binary elementwise funcs, clip, and matmul defaults to C-contiguous

[ndgrigorian]

This PR proposes a change to clip, matmul, and the BinaryElementwiseFunc class which makes C the default contiguity of outputs when order="K" and the input arrays must be cast.

This fixes a minor discrepancy where when multiple input arrays are both C- and F-contiguous and need to be cast, the output defaulted to F-contiguous instead of C-contiguous, creating unusual edge cases where type promotion (and therefore the selected device) became relevant to the output array's order, which should not be the case.

For example

In [1]: import dpctl.tensor as dpt, dpctl, numpy as np

In [2]: x1 = dpt.ones((10, 1), dtype="i4", device="cpu")

In [3]: x2 = dpt.ones((1, 5), dtype="f4", device="cpu")

In [4]: r = dpt.multiply(x1, x2)

In [5]: r.flags
Out[5]:
  C_CONTIGUOUS : False
  F_CONTIGUOUS : True
  WRITABLE : True

The selected device's type promotion graph contains float64, leading to F-contiguous outputs.

after this change:

In [1]: import dpctl.tensor as dpt, dpctl, numpy as np

In [2]: x1 = dpt.ones((10, 1), dtype="i4", device="cpu")
x
In [3]: x2 = dpt.ones((1, 5), dtype="f4", device="cpu")

In [4]: r = dpt.multiply(x1, x2)

In [5]: r.flags
Out[5]:
  C_CONTIGUOUS : True
  F_CONTIGUOUS : False
  WRITABLE : True

• Have you provided a meaningful PR description?
• Have you added a test, reproducer or referred to an 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 opening the PR as a draft?

[github-actions]

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

[oleksandr-pavlyk]

This looks uncontroversial enough for me.

[coveralls]

coverage: 87.948%. remained the same
when pulling eacef28 on fix-elementwise-order-K-broadcasting
into 7bc3124 on master.

[oleksandr-pavlyk]

LGMT!

[github-actions]

Array API standard conformance tests for dpctl=0.17.0dev0=py310h15de555_352 ran successfully.
Passed: 888
Failed: 17
Skipped: 91

[ndgrigorian]

@vtavana confirmed that dpctl test suite was unaffected by this change, I will go ahead and merge it

Internal CI failures also unrelated