Copy from/to multidimensional buffers

[sommerlukas]

Using typed memoryviews restricts queue memcpy to buffers with a single dimension only.

With this change, multi-dimensional buffers with contiguous layout can now be memcpy'd.

• 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?
• Have you added documentation for your changes, if necessary?
• Have you added your changes to the changelog?

[oleksandr-pavlyk]

Thank you for the fix @sommerlukas !

The change looks good to me, lets wait for the CI though

[coveralls]

coverage: 88.224% (-0.02%) from 88.24%
when pulling a8bc75e on sommerlukas:nd-memory-copy
into d84cb16 on IntelPython:master.

[ndgrigorian]

LGTM! Thank you @sommerlukas

[ndgrigorian]

Noting that the new test passes on both HIP and CUDA devices when the nd-memory-copy branch is built

[oleksandr-pavlyk]

LGTM! Thank you @sommerlukas