Work around sub_group load/store issues

[oleksandr-pavlyk]

Work-around issue with sub_group::load, sub_group::store functions

These functions are from oneAPI sycl_ext_oneapi_group_load_store extension.
https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/experimental/sycl_ext_oneapi_group_load_store.asciidoc

The current implementation of these primitives seem to require certain pointer alignment.

This PR adds bool template parameter to contig functors for all elementwise functions,
and the elementwise_common_impl checks alignment all pointers. If any is not aligned to
the expected boundary, a generic implementation is used instead of the one using
sg.load/sg.store.

• 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?

[ndgrigorian]

@oleksandr-pavlyk
The contiguous copy, where, and clip kernels also use sg load/store. They will need to be changed as well.

[github-actions]

View rendered docs @ https://intelpython.github.io/dpctl/pulls/1485/index.html

[coveralls]

coverage: 85.998%. remained the same
when pulling 86153f3 on work-around-sg-loadstore-issues
into 0b63d4f on master.

[github-actions]

Array API standard conformance tests for dpctl=0.15.1dev2=py310h15de555_24 ran successfully.
Passed: 876
Failed: 56
Skipped: 59

[ndgrigorian]

Assuming CI succeeds, LGTM

[github-actions]

Array API standard conformance tests for dpctl=0.15.1dev2=py310h15de555_26 ran successfully.
Passed: 876
Failed: 56
Skipped: 59