Network Programming Lab #3 — IPv4 Subnet Calculator

In the previous labs you used tools such as ipcalc, ping, and traceroute. You also learned how an IPv4 address and a netmask are related, and how the network address is calculated with a bitwise AND operation.

For this lab you will turn that calculation into a small program.

Before starting, read the notes from class:

https://github.com/networks26/BBS/blob/main/2/ip2.txt

The important ideas are:

network   = IP AND mask
broadcast = IP OR (NOT mask)

For this lab we only use prefixes /1 through /30. /31 and /32 are useful in real networks, but they have special rules and are outside this exercise.

---

Goal

Write a program that accepts an IPv4 address and a prefix length:

./ipcalc2 192.168.1.42 28

and prints:

IP: 192.168.1.42
Netmask: 255.255.255.240
Network: 192.168.1.32
Broadcast: 192.168.1.47
First usable: 192.168.1.33
Last usable: 192.168.1.46
Usable hosts: 14

You do not need to write the command-line parser. Starter programs are provided. They already:

• read the IP address from the command line;
• read the prefix length;
• convert the IPv4 address into a 32-bit value;
• check basic input errors;
• contain code for printing an IPv4 address.

Your job is to fill in the subnet calculations marked TODO.

---

Task 1 — Calculate One Network by Hand

Before programming, calculate this network yourself:

IP:     192.168.10.77
Prefix: /27

Write the IP address in binary:

Oh, how to convert?

Look:

echo "obase=2; ibase=10; 192" | bc

To convert binary to decimal:

echo "ibase=2; 101010" | bc

IP:
____________________________________________________

Write the /27 mask in binary:

Mask:
____________________________________________________

Apply bitwise AND:

IP:       __________________________________________
Mask:     __________________________________________
          -------------------------------------------
Network:  __________________________________________

Now write the answers in decimal:

Netmask:        ______________________________
Network:        ______________________________
Broadcast:      ______________________________
First usable:   ______________________________
Last usable:    ______________________________
Usable hosts:   ______________________________

Check your calculation with the system tool:

ipcalc 192.168.10.77/27

Task 2 — Choose a Programming Language

Choose one directory:

c/
python/
java/
go/
pascal/
asm/

The networking algorithm is the same in every language.

The assembly version is x86-64 GNU assembly for Linux. It uses the C library for input/output: sscanf() parses the four IPv4 octets, strtol() parses the prefix, and printf() prints the result. The subnet arithmetic itself is done in assembly.

Enter your chosen directory, for example:

cd c

Build the starter:

make

Run the example:

make run

The starter should run, but most answers will initially be 0.0.0.0 because the calculations are unfinished.

---

Task 3 — Complete the Subnet Calculator

Find the variables marked TODO in your source file.

The inputs have already been parsed for you:

ip      = IPv4 address represented as 32 bits
prefix  = number of network bits

You need to calculate:

mask
network
broadcast
first usable address
last usable address
number of usable hosts

Think about these questions before writing code:

1. If the prefix is /28, how many host bits are there?
2. How can you create a 32-bit value with 28 leading 1 bits?
3. Which bitwise operation gives the network address?
4. Which bits must be set to 1 to obtain the broadcast address?
5. Why is the first usable address network + 1?
6. Why is the last usable address broadcast - 1?
7. Why is the normal usable-host calculation 2^(host bits) - 2?

Do not call ipcalc from your program. Your program must perform the calculation itself.

After editing, rebuild and run:

make
make run

---

Task 4 — Test Your Program

A file called tests.txt is provided in the parent directory:

cat ../tests.txt

You should see several IP/prefix pairs.

Run the automatic tests:

make test

The starter version should fail. After your calculations are correct, every test should say PASS.

Example:

PASS: 192.168.1.42/24
PASS: 192.168.1.42/28
...

Do not modify test.sh to make your program pass.

---

Task 5 — Verify with ipcalc

Choose at least two entries from tests.txt and compare your program with the real ipcalc utility.

For example:

./ipcalc2 192.168.1.42 28
ipcalc 192.168.1.42/28

For Python use:

python3 ipcalc2.py 192.168.1.42 28

For Java use:

java IpCalc 192.168.1.42 28

Task 6 — Run All Test Networks with a Shell Loop

You already used while read in the previous lab. Now use it to run your own program repeatedly.

First, just read the input file:

while read ip prefix; do
    echo "$ip/$prefix"
done < ../tests.txt

Now run your program for every line.

For C, Go, Pascal, or assembly:

while read ip prefix; do
    echo "===== $ip/$prefix ====="
    ./ipcalc2 "$ip" "$prefix"
done < ../tests.txt

For Python replace ./ipcalc2 with:

python3 ipcalc2.py

For Java replace it with:

java IpCalc

When it works, run it again and save the complete output with tee:

while read ip prefix; do
    echo "===== $ip/$prefix ====="
    ./ipcalc2 "$ip" "$prefix"
done < ../tests.txt | tee results.txt

Use the correct command for your chosen language.

Read the result:

cat results.txt

Count its lines:

wc -l results.txt

Show only calculated network addresses:

grep '^Network:' results.txt

Write the number of lines in results.txt:

____________________

---

Challenge — Same Network or Router?

Suppose your machine is:

192.168.10.17/27

Consider these destinations:

192.168.10.29
192.168.10.35
8.8.8.8

For each destination, calculate its network using the same mask and decide whether the destination is on the same IP network.

192.168.10.29: same network / router needed: __________________
192.168.10.35: same network / router needed: __________________
8.8.8.8:       same network / router needed: __________________

The useful test is:

(my_ip AND mask) == (destination_ip AND mask)

If the two network addresses are equal, the destination is on the same subnet. Otherwise the packet normally has to be sent toward a router.

Optional programming challenge: extend your program, or write a second small program, that performs this test automatically.

---

Git Submission

Your final commit must contain:

• your completed source file;
• results.txt;
• this README.md with your answers.

Check your changes:

git status

Add the files explicitly. For example, if you chose C:

git add c/ipcalc2.c c/results.txt README.md

If you are already inside the c/ directory:

git add ipcalc2.c results.txt ../README.md

Commit:

git commit -m "Complete IPv4 subnet calculator"

Check the commit:

git show --stat

Push:

git push

Before finishing, make sure this succeeds in your chosen language directory:

make test