vload_half and vload_store implementations are overly simplified when 16bit-storage extension is not available

[dneto0]

• They assume the pointer argument points to an even-numbered index into a half array. That's why they can get away with casting out to pointer-to-int and then using the GLSL std 450 half-packing instruction.
• The vstore_half implementation assumes there is no contention for the word we're storing into. This can be fixed with a retry loop around the atomic_xor trick.

[dneto0]

We could remove support for these functions and force use of a new __clspv_vload_half(size_t, uint*ptr) that forces the pointer to be aligned.

[dneto0]

Better name is __clspv_vloada_half since the pointer is aligned more stringently than just pointer-to-half.