If you could change meaning of int to be 64-bit type, then your code ABI (https://en.wikipedia.org/wiki/Application_binary_interface https://en.wikipedia.org/wi

If you could change meaning of int to be 64-bit type, then your code ABI (https://en.wikipedia.org/wiki/Application_binary_interface https://en.wikipedia.org/wiki/Application_binary_interface) would be essentially incompatible, and you wouldn't be able to use code (such as standard library, or other headers provided by operating system libraries) that was compiled on a compiler where int was 32-bit, as for instance wrong structure fields would be read. If you would try to "solve" that by assuming headers use 32-bit int definition, then you would break calling your own code, as you would use 32-bit int definition provided in headers, while your code was called with 64-bit int convention. It's probably just easier to define your own int meaning type, say, a shorter alias of int_fast32_t.