CON35-C. Avoid deadlock by locking in a predefined order

Mutexes are used to prevent multiple threads from accessing shared resources at the same time. Sometimes, when locking mutexes, multiple threads hold each other's lock, and the program consequently deadlocks. Four conditions are required for deadlock to occur:

Mutual exclusion
Hold and wait
No preemption
Circular wait

Deadlock needs all four conditions, so preventing deadlock requires preventing any one of the four conditions. One simple solution is to lock the mutexes in a predefined order, which prevents circular wait.

Noncompliant Code Example

This noncompliant code example has behavior that depends on the runtime environment and the platform's scheduler. However, with proper timing, the main() function will deadlock when running thr1 and thr2. Thread thr1 tries to lock ba2's mutex, while thr2 tries to lock ba1's mutex in the deposit() function, and the program will hang.

#include <stdlib.h>
#include <threads.h>
 
typedef struct {
  int balance;
  mtx_t balance_mutex;
} bank_account;

typedef struct {
  bank_account *from;
  bank_account *to;
  int amount;
} transaction;

void create_bank_account(bank_account **ba, int initial_amount) {
  bank_account *nba = (bank_account *)
                        malloc(sizeof(bank_account));
  if (nba == NULL) {
    /* Handle error */
  }

  nba->balance = initial_amount;
  if (thrd_success != mtx_init(&nba->balance_mutex, mtx_plain)) {
    /* Handle error */
  }

  *ba = nba;
}

int deposit(void *ptr) {
  transaction *args = (transaction *)ptr;

  if (thrd_success != mtx_lock(&(args->from->balance_mutex))) {
    /* Handle error */
  }

  /* Not enough balance to transfer */
  if (args->from->balance < args->amount) {
    if (thrd_success != mtx_unlock(&(args->from->balance_mutex))) {
      /* Handle error */
    }
    return -1;  /* Indicate error */
  }

  if (thrd_success != mtx_lock(&(args->to->balance_mutex))) {
    /* Handle error */
  }

  args->from->balance -= args->amount;
  args->to->balance += args->amount;

  if (thrd_success != mtx_unlock(&(args->from->balance_mutex))) {
    /* Handle error */
  }
  if (thrd_success != mtx_unlock(&(args->to->balance_mutex))) {
    /* Handle error */
  }

  free(ptr);
  
  return 0;
}

int main(void) {
  thrd_t thr1, thr2;
  transaction *arg1;
  transaction *arg2;
  bank_account *ba1;
  bank_account *ba2;

  create_bank_account(&ba1, 1000);
  create_bank_account(&ba2, 1000);

  arg1 = (transaction *)malloc(sizeof(transaction));
  if (arg1 == NULL) {
    /* Handle error */
  }
  arg2 = (transaction *)malloc(sizeof(transaction));
  if (arg2 == NULL) {
    /* Handle error */
  }

  arg1->from = ba1;
  arg1->to = ba2;
  arg1->amount = 100;

  arg2->from = ba2;
  arg2->to = ba1;
  arg2->amount = 100;

  /* Perform the deposits */
  if (thrd_success != thrd_create(&thr1, deposit, (void *)arg1)) {
    /* Handle error */
  }
  if (thrd_success != thrd_create(&thr2, deposit, (void *)arg2)) {
    /* Handle error */
  }
  return 0;
}

Compliant Solution

This compliant solution eliminates the circular wait condition by establishing a predefined order for locking in the deposit() function. Each thread will lock on the basis of the bank_account ID, which is set when the bank_account struct is initialized.

#include <stdlib.h>
#include <threads.h>
 
typedef struct {
  int balance;
  mtx_t balance_mutex;
 
  /* Should never be changed after initialized */ 
  unsigned int id;
} bank_account;

unsigned int global_id = 1;

void create_bank_account(bank_account **ba, int initial_amount) {
  bank_account *nba = (bank_account *)
                        malloc(sizeof(bank_account));
  if (nba == NULL) {
    /* Handle error */
  }

  nba->balance = initial_amount;
  if (thrd_success != mtx_init(&nba->balance_mutex, mtx_plain)) {
    /* Handle error */
  }

  nba->id = global_id++;
  *ba = nba;
}

int deposit(void *ptr) {
  transaction *args = (transaction *)ptr;
  int result = -1;
  mtx_t *first;
  mtx_t *second;

  if (args->from->id == args->to->id)
    return -1;  /* Indicate error */

  /* Ensure proper ordering for locking */
  if (args->from->id < args->to->id) {
    first = &args->from->balance_mutex;
    second = &args->to->balance_mutex;
  } else {
    first = &args->to->balance_mutex;
    second = &args->from->balance_mutex;
  }
  if (thrd_success != mtx_lock(first)) {
    /* Handle error */
  }
  if (thrd_success != mtx_lock(second)) {
    /* Handle error */
  }

  /* Not enough balance to transfer */
  if (args->from->balance >= args->amount) {
    args->from->balance -= args->amount;
    args->to->balance += args->amount;
    result = 0;
  }

  if (thrd_success != mtx_unlock(second)) {
    /* Handle error */
  }
  if (thrd_success != mtx_unlock(first)) {
    /* Handle error */
  }
  free(ptr);
  return result;
}

Risk Assessment

Deadlock prevents multiple threads from progressing, halting program execution. A denial-of-service attack is possible because the attacker can force deadlock situations. Deadlock is likely to occur in multithreaded programs that manage multiple shared resources.

Rule	Severity	Likelihood	Remediation Cost	Priority	Level
CON35-C	Low	Probable	Medium	P4	L3

Related Vulnerabilities

Search for vulnerabilities resulting from the violation of this rule on the CERT website.

Automated Detection

Tool	Version	Checker	Description
Coverity	6.5	DEADLOCK	Fully implemented

Related Guidelines

CERT Oracle Secure Coding Standard for Java	LCK07-J. Avoid deadlock by requesting and releasing locks in the same order
MITRE CWE	CWE-764, Multiple locks of critical resources

Space shortcuts

Page tree