CODEWITHSUNDEEP
cinteger-overflow
Marcos Gonzalez
Marcos Gonzalez

Reputation: 1096

How can a C program show the value of the largest possible unsigned long long?

Shouldn't the following program print out the value of 2^64 -1 as an integer?

#include <stdio.h>
#include <math.h>

int main(void) 
{
    printf ("%llu\n" , ( unsigned long long int ) ( pow (2 ,  8 * sizeof (unsigned long long) ) - 1 ) ) ;

    return 0;
}

Upvotes: 1

Views: 49

Answers (1)

M Oehm
M Oehm

Reputation: 29126

The function pow returns a double, usually a 64-bit floating-point type with 53 bits of mantissa. That is not enough to resolve numbers in the vicinity of 264 with a granularity of 1.

The next representable double value smaller than 264 is 264 − 2048 = 264 − 211. 264 − 1 is still 264, which is the nearest representable value. Converting that to unsigned long long, you get 0.

If you need the value of the largest possible unsigned long long, you can use ULLONG_MAX from <limits.h>:

printf ("%llu\n", ULLONG_MAX);    // need <limits.h>

The one's complement of (unsigned long long) 0 – an unsigned long long with all bits set – should work, too:

printf ("%llu\n", ~0ull);

Upvotes: 3

Related Questions