Operator Precedence Logic Error

Description

While often just a bug, operator precedence logic errors can have serious consequences if they are used in security-critical code, such as making an authentication decision.

Demonstrations

The following examples help to illustrate the nature of this weakness and describe methods or techniques which can be used to mitigate the risk.

Note that the examples here are by no means exhaustive and any given weakness may have many subtle varieties, each of which may require different detection methods or runtime controls.

Example One

In the following example, the method validateUser makes a call to another method to authenticate a username and password for a user and returns a success or failure code.

#define FAIL 0
#define SUCCESS 1

...

int validateUser(char *username, char *password) {


  int isUser = FAIL;

  // call method to authenticate username and password

  // if authentication fails then return failure otherwise return success
  if (isUser = AuthenticateUser(username, password) == FAIL) {
    return isUser;
  }
  else {
    isUser = SUCCESS;
  }

  return isUser;

}

However, the method that authenticates the username and password is called within an if statement with incorrect operator precedence logic. Because the comparison operator "==" has a higher precedence than the assignment operator "=", the comparison operator will be evaluated first and if the method returns FAIL then the comparison will be true, the return variable will be set to true and SUCCESS will be returned. This operator precedence logic error can be easily resolved by properly using parentheses within the expression of the if statement, as shown below.

...

if ((isUser = AuthenticateUser(username, password)) == FAIL) {

...

Example Two

In this example, the method calculates the return on investment for an accounting/financial application. The return on investment is calculated by subtracting the initial investment costs from the current value and then dividing by the initial investment costs.

public double calculateReturnOnInvestment(double currentValue, double initialInvestment) {


  double returnROI = 0.0;

  // calculate return on investment
  returnROI = currentValue - initialInvestment / initialInvestment;

  return returnROI;

}

However, the return on investment calculation will not produce correct results because of the incorrect operator precedence logic in the equation. The divide operator has a higher precedence than the minus operator, therefore the equation will divide the initial investment costs by the initial investment costs which will only subtract one from the current value. Again this operator precedence logic error can be resolved by the correct use of parentheses within the equation, as shown below.

...

returnROI = (currentValue - initialInvestment) / initialInvestment;

...

Note that the initialInvestment variable in this example should be validated to ensure that it is greater than zero to avoid a potential divide by zero error (CWE-369).

See Also