PPractick

The problem

An embedded controller with only a few registers needs to swap two integer values in place without a spare variable. The XOR swap does it with three XORs, using the facts that x ^ x = 0 and x ^ 0 = x. It's a classic interview question because of the reasoning it requires and one sharp edge: if both references point to the same variable, the first XOR zeroes it. The arithmetic version (a = a + b; b = a − b; a = a − b) has an overflow problem instead.

Example: a = 5, b = 7 → a = 7, b = 5.

Function signature

void xorSwap(int& a, int& b);

Test cases

Input (a, b)	Output (a, b)	Explanation
(5, 7)	7 5
(-3, 12)	12 -3	Works on the bit pattern, signs included
(0, 9)	9 0
(4, 4)	4 4	Equal values in different variables are fine
xorSwap(x, x) with x = 8	8	The same variable must not be zeroed

(Confirm: no temporary variable; must work for negative numbers; swapping a variable with itself must leave it unchanged.)

How interviewers phrase it

• "Swap two numbers without using a third variable" (GeeksforGeeks), and "now do it with bit manipulation".

Clarifying questions to ask

1. Integers only? (XOR doesn't apply to floating point.)
2. Can the two references refer to the same variable?
3. Is overflow a concern? (It rules out the arithmetic version.)

Follow-ups to expect

1. "Prove the XOR swap."
2. "What breaks with aliasing?"
3. "Why not a = a + b; b = a − b; a = a − b?"
4. "Is it faster than a temporary?"
5. "Complexity?"

Related questions

• Find missing number Easy
• Introduction to Bits and Tricks Easy
• Check if the i-th bit is Set or Not Easy
• Check if a Number is Odd or Not Easy
• Check if a Number is Power of 2 or Not Easy
• Count the Number of Set Bits Easy

All 474 dsa interview questions

DSA mock interviews · System design interview questions · Free mock interview practice · Browse interviewers · Sign in