several calls to %mload_packing can be directly replaced to MLOAD_32BYTES, no need for hook + jump
we don't really care about Kernel code size now that memory is preinitialized, hence we can add dedicated macros for mstore_unpacking_foo, which skips going through the jumptable
Removes some CPU cycles:
%mload_packing
can be directly replaced toMLOAD_32BYTES
, no need for hook + jumpmstore_unpacking_foo
, which skips going through the jumptable