IP Address overlapping/in range of CIDR

Question

I have an IP address in unsigned long format, coding in C++ I have a CIDR notation IP address range such as "10.2.3.98/24"

How do i check if my IP address overlaps with the mentioned range?

Do you need help with parsing the strings as given, or have you already done that and need help with the bit masking? — Greg Hewgill, Nov 19 '12 at 23:39
my problem is to go from IP address which i have as unsigned long to the representation of IP range notation start and end in unsigned long so I can do my range checking. In other words, how does one go from "10.2.3.98/24" to low and high range in unsigned long — reza, Nov 20 '12 at 00:04

score 1 · Answer 1 · answered Nov 20 '12 at 03:25

To be as simple as possible, the part after the slash are the bits to keep, basically. So for example /24 means the most significant 3 bytes (24 bits) are preserved. Hence you can see if an address fits by masking it and checking for equality. The adress AND mask itself would be the min; If you are looking for max you can OR with the inverse of the mask.

score 0 · Accepted Answer · answered Nov 20 '12 at 06:32

This should work if you already know ip addresses as unsigned long and numbers:

bool cidr_overlap(uint32_t ip1, int n1,
                  uint32_t ip2, int n2)
{
    return (ip1 <= (ip2 | ((1ul << (32-n2))-1)))
        || (ip2 <= (ip1 | ((1ul << (32-n1))-1)));
}

score 0 · Answer 3 · answered Apr 05 '17 at 05:00

Let's Assume Your IP addresses and Masks as Follows and IP addresses are in integer form.

Example 3232235896/30 ==> (actual IP 192.168.1.120/30)

Lets say You need to find out overlap of (ip_one , mask_one) and (ip_two , mask_two)

uint32_t mask_one_max = ((1ul << (32 - mask_one)) - 1);
uint32_t mask_one_min = ~mask_one_max;

uint32_t mask_two_max = ((1ul << (32 - mask_two)) - 1);
uint32_t mask_two_min = ~mask_two_max;

return (((ip_one & mask_one_min) <= (ip_two | mask_two_max)) && ((ip_two & mask_two_min) <= (ip_one | mask_one_max)));

This will return true if overlapping occurs.

The Solution is Proposed based on the Generic way of finding two integer ranges overlap. As you can see in the solution I first convert the CIDR ranges to range of Integers and use them to find the overlap.

score 0 · Answer 4 · answered Nov 18 '21 at 10:00

This function checks if two networks overlap.

#include <arpa/inet.h>
#include <netinet/in.h>

static inline bool cidr_overlap(struct in_addr ip1, int n1, 
                                struct in_addr ip2, int n2)
{
    uint32_t mask1 = ~(((uint32_t)1 << (32 - n1)) - 1);
    uint32_t mask2 = ~(((uint32_t)1 << (32 - n2)) - 1);

    return (htonl(ip1.s_addr) & mask1 & mask2) == 
           (htonl(ip2.s_addr) & mask1 & mask2);
}

IP Address overlapping/in range of CIDR

4 Answers4

Linked