`prefetch` on sycl with cuda backend on Windows

[mmoadeli]

Windows doesn't support cudaMemPrefetchAsync() which is used in the call to prefetch. Would be a need to investigate alternatives for Windows.

intel/llvm#13957 may be used to further information.

[JackAKirk]

Is this not already dealt with by

unified-runtime/source/adapters/cuda/enqueue.cpp

Line 1644 in f06bc02

// for managed memory. Therefore, ignore prefetch hint if concurrent managed

Or maybe we need a similar fix in other parts of the code?