Bitwise Operations

Bitwise operations act on the individual bits of their operands. For example, a bitwise AND (&) will perform a logical AND (&&) on each pair of bits. 14 & 9 evaluates to 8:

   1110
&  1001
--------
   1000

Logical operators treat each operand as having only one value.

Operators

C has a number of operators that work directly on bits.

Operator Name
& Bitwise AND
Bitwise OR
^ Bitwise XOR
~ Bitwise NOT
<< Bit Shift Left
>> Bit Shift Right

The first three operators are explained with truth table below.

    & ^
0 0 0 0 0
0 1 0 1 1
1 0 0 1 1
1 1 1 1 0

The right shift (>>) can be logical or arithmetic. Logical right shift pads with zeros, but arithmetic right shift pads with the most significant bit.

Examples

Code Description
x&(1<<i) Retrieve the ith bit
x│=(1<<i) Set the ith bit
x&=~(1<<i) Clear the ith bit
x^=(1<<i) Flip the ith bit

Bit-level operations often use masking. A mask is a bit pattern that indicates a selected set of bits within a word. For example, the bit-level operation x & 0xFF yields a value consisting of the least significant byte of x, but with all other bytes set to 0. The expression ~0 will yield a mask of all ones, regardless of the word size of the machine.

Sum Integers

int add(int x, int y)
{
  if (y == 0) {
    return x;
  } else {
    int carry = (x & y) << 1;
    return add(x ^ y, carry);
  }
}

Count 1 Bits

int count(int bits)
{
  int total = 0;
  while (bits) {
    total += bits & 1;
    bits >>= 1;
  }
  return total;
}

Erase Lowest Bit

x&(x - 1) equals x with its lowest set bit erased. For example, if x = 00101100, then x - 1 = 00101011, so x&(x - 1) = 00101100 & 00101011 = 00101000.

This fact can be used to reduce the time complexity of counting 1 bits.

int count(int bits)
{
  int total = 0;
  while (bits) {
    total += 1;
    bits &= bits - 1;
  }
  return total;
}

Parity

The parity of a binary word is 1 if the number of 1s in the word is odd; otherwise, it is 0. For example, the parity of 1011 is 1, and the parity of 10001000 is 0.

Parity could be computed with brute-force by counting 1 bits (see above) and modding by 2. XOR provides a more efficient solution. Note that the XOR of a group of bits is its parity. For example 1^0^1^1 is 1. Since XOR is associative and commutative, we can process multiple bits at a time.

int parity(int bits)
{
  bits ^= bits >> 32;
  bits ^= bits >> 16;
  bits ^= bits >> 8;
  bits ^= bits >> 4;
  bits ^= bits >> 2;
  bits ^= bits >> 1;
  return bits & 1;  /* extract last bit */
}