Should OpenCL C have a memcpy?

At the moment the [OpenCL C spec](https://registry.khronos.org/OpenCL/specs/unified/html/OpenCL_C.html#reinterpreting-data-as-another-type) states:

> Of these, only memcpy is strictly correct in C99. Since OpenCL does not provide memcpy, other methods are needed.

Should we provide a type generic (for all OpenCL C scalars and vector types?) `memcpy` in OpenCL C? In SPIR-V we have [OpCopyMemory](https://registry.khronos.org/SPIR-V/specs/unified1/SPIRV.html#OpCopyMemory) and [OpCopyMemorySized](https://registry.khronos.org/SPIR-V/specs/unified1/SPIRV.html#OpCopyMemorySized)?

I think I'd prefer to provide a higher level function to developers rather than them trying to implement `memcpy` with loops or something that are fundamentally more difficult to optimize and more difficult to reason about pointer alignments.

Should we have a `memcpy` extension for this?

Should we have a `memcpy_aligned` version where developers can specify explicit pointer alignment?

Start with the linked OpenCL C specification section on reinterpreting data and compare it with the linked SPIR-V OpCopyMemory and OpCopyMemorySized definitions. Resolve whether memcpy and memcpy_aligned should be specified, including supported scalar or vector types and alignment semantics. Done means the proposal has a settled specification decision and documented scope.

Written by the indexing model from the issue text.