Round signed integers toward negative infinity in dpctl.tensor.floor_divide

[ndgrigorian]

Resolves #1247

This PR changes the behavior of dpctl.tensor.floor_divide to round the result of division between signed integers toward negative infinity rather than toward 0 (as is the norm in C++).

• 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]

View rendered docs @ https://intelpython.github.io/dpctl/pulls/1260/index.html

[github-actions]

Array API standard conformance tests for dpctl=0.14.4=py310h7bf5fec_8 ran successfully.
Passed: 387
Failed: 613
Skipped: 119

[coveralls]

coverage: 83.252% (+0.008%) from 83.244% when pulling 3f436b2 on floor-divide-negative-fix into d109770 on master.

[github-actions]

Array API standard conformance tests for dpctl=0.14.4=py310h7bf5fec_10 ran successfully.
Passed: 388
Failed: 612
Skipped: 119

[npolina4]

in: dpt.floor_divide(dpt.asarray(1), 0)
out: usm_ndarray(4294967295)
expected: 0

in: dpt.floor_divide(dpt.asarray(0), 0)
out: usm_ndarray(4294967295)
expected: 0

[ndgrigorian]

in: dpt.floor_divide(dpt.asarray(1), 0)
out: usm_ndarray(4294967295)
expected: 0

in: dpt.floor_divide(dpt.asarray(0), 0)
out: usm_ndarray(4294967295)
expected: 0

In [2]: x = dpt.asarray(1, dtype="u8")

In [3]: dpt.floor_divide(x, dpt.asarray(0, dtype="u8"))
Out[3]: usm_ndarray(4294967295, dtype=uint64)

In [4]: x = dpt.asarray(4294967295, dtype="u8")

In [5]: dpt.floor_divide(x, dpt.asarray(0, dtype="u8"))
Out[5]: usm_ndarray(4294967295, dtype=uint64)

In [6]: x = dpt.asarray(4294967296, dtype="u8")

In [7]: dpt.floor_divide(x, dpt.asarray(0, dtype="u8"))
Out[7]: usm_ndarray(18446744073709551615, dtype=uint64)

Another ugly case. I was trying to avoid a branch for the denominator = 0 case, but I will add a branch to resolve this after all.

[github-actions]

Array API standard conformance tests for dpctl=0.14.4=py310h7bf5fec_11 ran successfully.
Passed: 387
Failed: 613
Skipped: 119

[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=0.14.4=py310h7bf5fec_12 ran successfully.
Passed: 388
Failed: 612
Skipped: 119