Bit #11 - overflow flag

The overflow flag (OF) is sometimes perceived as two's complement form of the carry flag, which is not really the case. OF is set when the result of the operation is either too small or too big a number to fit into the destination operand. For example, consider the addition of two 8-bit positive values, 0x74 and 0x7f. The resulting value of such an addition is 0xf3, which is still 8-bit, which is fine for unsigned numbers, but since we added two values that we considered to be signed, there has to be the sign bit and there are no more bits to store the 9-bit signed result. The same would happen if we try to add two negative 8-bit values, 0x82 and 0x81. The meaning of the addition of two negative numbers is in the fact subtraction of a positive number from a negative number, which should result in an even smaller number. Thus, 0x82 + 0x81 would result in 0x103, where the ninth bit, 1, is the sign bit, but it cannot be stored in an 8-bit operand. The same applies to larger operands (16, 32, and 64-bit).