holding 2^63 -1 in long long

Question

I ran the following two pieces of code on Windows XP (Code:Block, MinGW), and Ubuntu (11.04, G++)

I have trouble running the following code

#include <iostream>

using namespace std;

int main(){

    long long a = 9223372036854775807;
    cout <<  a;
    return 0;

}

That number is 2^63 -1. But I will get an error stating:

C:\Documents and Settings\JohnWong\My Documents\codeblock\343_hw_1\main.cpp|9|error: integer constant is too large for "long" type|

On ubuntu - it compiled, but the answer retunred is 9223372036854775808, notice the 8 at the end....

Now if I run this code, using the power function, I am okay.

#include <iostream>
#include <iomanip>
#include <math.h>
using namespace std;

int main(){

    long long a = pow(2,64);
    cout << "a:   " << setprecision(20) << a << endl;
    cout << "a-1:   " << setprecision(20) << a-1 << endl;
    cout << "a-2:   " <<  setprecision(20) << a-2 << endl;
    cout << "a+0:   " << setprecision(20) << a+0 << endl;
    cout << "a+1:   " << setprecision(20) << a+1 << endl;
    cout << "a+2:   " << setprecision(20) << a+2 << endl;
    cout << "a+3:   " << setprecision(20) << a+3 << endl;

    return 0;
}

I will get the values I want (anything from +1 will cause an overflow, that's okay).

On Ubuntu the outputs looks the same. Good.

So what's going on here? Why constant is not good??? I even tried intmax_t and int64_t as datatype running the first code.

Can someone explain this behavior? Thanks!

Answer 1

long long a = 9223372036854775807LL;

The LL makes the literal a long long literal. Otherwise the literal defaults to being a long literal and then is casted over to a long long before being stored in a.