Incorrect Short Circuit Evaluation

The software contains a conditional statement with multiple logical expressions in which one of the non-leading expressions may produce side effects. This may lead to an unexpected state in the program after the execution of the conditional, because short-circuiting logic may prevent the side effects from occurring.


Description

Usage of short circuit evaluation, though well-defined in the C standard, may alter control flow in a way that introduces logic errors that are difficult to detect, possibly causing errors later during the software's execution. If an attacker can discover such an inconsistency, it may be exploitable to gain arbitrary control over a system.

If the first condition of an "or" statement is assumed to be true under normal circumstances, or if the first condition of an "and" statement is assumed to be false, then any subsequent conditional may contain its own logic errors that are not detected during code review or testing.

Finally, the usage of short circuit evaluation may decrease the maintainability of the code.

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

The following function attempts to take a size value from a user and allocate an array of that size (we ignore bounds checking for simplicity). The function tries to initialize each spot with the value of its index, that is, A[len-1] = len - 1; A[len-2] = len - 2; ... A[1] = 1; A[0] = 0; However, since the programmer uses the prefix decrement operator, when the conditional is evaluated with i == 1, the decrement will result in a 0 value for the first part of the predicate, causing the second portion to be bypassed via short-circuit evaluation. This means we cannot be sure of what value will be in A[0] when we return the array to the user.

#define PRIV_ADMIN 0
#define PRIV_REGULAR 1
typedef struct{
  int privileges;
  int id;
} user_t;
user_t *Add_Regular_Users(int num_users){
  user_t* users = (user_t*)calloc(num_users, sizeof(user_t));
  int i = num_users;
  while( --i && (users[i].privileges = PRIV_REGULAR) ){
    users[i].id = i;
  }
  return users;
}
int main(){
  user_t* test;
  int i;
  test = Add_Regular_Users(25);
  for(i = 0; i < 25; i++) printf("user %d has privilege level %d\n", test[i].id, test[i].privileges);
}

When compiled and run, the above code will output a privilege level of 1, or PRIV_REGULAR for every user but the user with id 0 since the prefix increment operator used in the if statement will reach zero and short circuit before setting the 0th user's privilege level. Since we used calloc, this privilege will be set to 0, or PRIV_ADMIN.

See Also

SFP Secondary Cluster: Glitch in Computation

This category identifies Software Fault Patterns (SFPs) within the Glitch in Computation cluster (SFP1).

CERT C++ Secure Coding Section 03 - Expressions (EXP)

Weaknesses in this category are related to rules in the Expressions (EXP) section of the CERT C++ Secure Coding Standard. Since not all rules map to specific weaknesse...

Comprehensive CWE Dictionary

This view (slice) covers all the elements in CWE.

Weaknesses Introduced During Implementation

This view (slice) lists weaknesses that can be introduced during implementation.


Common Weakness Enumeration content on this website is copyright of The MITRE Corporation unless otherwise specified. Use of the Common Weakness Enumeration and the associated references on this website are subject to the Terms of Use as specified by The MITRE Corporation.