Most 64-bit Linux systems and macOS use LP64 with 64-bit long and pointers. Windows is slightly different and uses LLP64.

This means that on LP64 systems long and long long are the same size.

As usual cppreference.com has a good page on Fundamental Types.

The current definition of the byteSwap functions use unsigned short, unsigned int and unsigned long long for 16-, 32-, and 64-bit uints.

The definitions I've seen for the various builtin bswap functions are defined using fixed with types. For example gcc has the folllwing:

Built-in Function: uint16_t __builtin_bswap16 (uint16_t x)
Returns x with the order of the bytes reversed; for example, 0xaabb becomes 0xbbaa. Byte here always means exactly 8 bits.

Built-in Function: uint32_t __builtin_bswap32 (uint32_t x)
Similar to __builtin_bswap16, except the argument and return types are 32-bit.

Built-in Function: uint64_t __builtin_bswap64 (uint64_t x)
Similar to __builtin_bswap32, except the argument and return types are 64-bit.

And macOS uses something similar for the OSSwap functions.

I think it would be great if TagLib followed suit.