In the NEWTABLE opcode, the field B and C,which is the array and hash size of the table,respectively,use the so-called "floating point byte" with unsigned int.
I guess the "floating point byte" can support more data range than the raw unsigned int, is it right?
