[FEA]: Provide mdspan padded layouts

[fbusato]

Is this a duplicate?

• I confirmed there appear to be no duplicate issues for this request and that I agree to the Code of Conduct

Area

libcu++

Is your feature request related to a problem? Please describe.

mdspan padded layouts are essential to efficiently perform computation on matrices with padding and to select sub matrices with compile-time dimensions

Describe the solution you'd like

Provide C++26 mdspan padded layouts and backport to C++14/17:

• cuda::std::layout_left_padded,
• cuda::std::layout_right_padded

Describe alternatives you've considered

No response

Additional context

No response