What would be the result of the following code in big endian?

Question

Apparently it seems Ideone and Codepad both uses Little endian to compile and run their code.

If for example I want to test the following on Big endian CPU then I'm curious to know does casting the unsigned char[] array to unsigned short* really changes its output?

#include <iostream>

int main()
{
   unsigned char a[2] = {0, 1};
   unsigned short *s = (unsigned short*)a;
   std::cout << *s;
}

As far as I have known that if you cast an int to char* then it does depend on the endianness but what is case with arrays or any other type you try to reinterpret?

http://ideone.com/4UP0J0

Answer 1

Endian is a feature of the CPU, not the compiler...

Assuming your short type if 16 bits, and the unsigned chars are 8 bits then the memory layout of a will be

0x00 0x01

On a bit endian CPU this means that *s will be 1 (0x0001). On a little endian CPU *s will be 256 (0x0100).

As an aside, it's not so easy to find big endian CPUs any more. This page has a list of CPU architectures from the last few decades and lists the endianness of each one.