CERT C: Rec. CON05-C

Do not perform operations that can block while holding a lock

Description

This checker is deactivated in a default Polyspace^® as You Code analysis. See Checkers Deactivated in Polyspace as You Code Analysis (Polyspace Access).

Rule Definition

Do not perform operations that can block while holding a lock.¹

Polyspace Implementation

The rule checker checks for Blocking operation while holding lock.

Examples

expand all

Blocking operation while holding lock

Issue

Blocking operation while holding lock occurs when a task (thread) performs a potentially lengthy operation while holding a lock.

The checker considers calls to these functions as potentially lengthy:

Functions that access a network such as recv
System call functions such as fork, pipe and system
Functions for I/O operations such as getchar and scanf
File handling functions such as fopen, remove and lstat
Directory manipulation functions such as mkdir and rmdir

The checker automatically detects certain primitives that hold and release a lock, for instance, pthread_mutex_lock and pthread_mutex_unlock. For the full list of primitives that are automatically detected, see Auto-Detection of Thread Creation and Critical Section in Polyspace.

Risk

If a thread performs a lengthy operation when holding a lock, other threads that use the lock have to wait for the lock to be available. As a result, system performance can slow down or deadlocks can occur.

Fix

Perform the blocking operation before holding the lock or after releasing the lock.

Some functions detected by this checker can be called in a way that does not make them potentially lengthy. For instance, the function recv can be called with the parameter O_NONBLOCK which causes the call to fail if no message is available. When called with this parameter, recv does not wait for a message to become available.

Example - Network I/O Operations with recv While Holding Lock

#include <pthread.h>
#include <sys/socket.h>

pthread_mutexattr_t attr;
pthread_mutex_t mutex;
 
void thread_foo(void *ptr) {
  unsigned int num;
  int result;
  int sock;
 
  /* sock is a connected TCP socket */
 
  if ((result = pthread_mutex_lock(&mutex)) != 0) {
    /* Handle Error */
  }
 
  if ((result = recv(sock, (void *)&num, sizeof(unsigned int), 0)) < 0) { //Noncompliant
    /* Handle Error */
  }
 
  /* ... */
 
  if ((result = pthread_mutex_unlock(&mutex)) != 0) {
    /* Handle Error */
  }
}
 
int main() {
  pthread_t thread;
  int result;
 
  if ((result = pthread_mutexattr_settype(
      &attr, PTHREAD_MUTEX_ERRORCHECK)) != 0) {
    /* Handle Error */
  }
 
  if ((result = pthread_mutex_init(&mutex, &attr)) != 0) {
    /* Handle Error */
  }
 
  if (pthread_create(&thread, NULL,(void*(*)(void*))& thread_foo, NULL) != 0) {
    /* Handle Error */
  }
 
  /* ... */
 
  pthread_join(thread, NULL);
 
  if ((result = pthread_mutex_destroy(&mutex)) != 0) {
    /* Handle Error */
  }
 
  return 0;
}

In this example, in each thread created with pthread_create, the function thread_foo performs a network I/O operation with recv after acquiring a lock with pthread_mutex_lock. Other threads using the same lock variable mutex have to wait for the operation to complete and the lock to become available.

Correction — Perform Blocking Operation Before Acquiring Lock

One possible correction is to call recv before acquiring the lock.

#include <pthread.h>
#include <sys/socket.h>

pthread_mutexattr_t attr;
pthread_mutex_t mutex;
 
void thread_foo(void *ptr) {
  unsigned int num;
  int result;
  int sock;
 
  /* sock is a connected TCP socket */
  if ((result = recv(sock, (void *)&num, sizeof(unsigned int), 0)) < 0) {
    /* Handle Error */
  }
  
  if ((result = pthread_mutex_lock(&mutex)) != 0) {
    /* Handle Error */
  }
 
    /* ... */
 
  if ((result = pthread_mutex_unlock(&mutex)) != 0) {
    /* Handle Error */
  }
}
 
int main() {
  pthread_t thread;
  int result;
 
  if ((result = pthread_mutexattr_settype(
      &attr, PTHREAD_MUTEX_ERRORCHECK)) != 0) {
    /* Handle Error */
  }
 
  if ((result = pthread_mutex_init(&mutex, &attr)) != 0) {
    /* Handle Error */
  }
 
  if (pthread_create(&thread, NULL,(void*(*)(void*))& thread_foo, NULL) != 0) {
    /* Handle Error */
  }
 
  /* ... */
 
  pthread_join(thread, NULL);
 
  if ((result = pthread_mutex_destroy(&mutex)) != 0) {
    /* Handle Error */
  }
 
  return 0;
}

Check Information

Group: Rec. 14. Concurrency (CON)

Version History

Introduced in R2019a

¹ This software has been created by MathWorks incorporating portions of: the “SEI CERT-C Website,” © 2017 Carnegie Mellon University, the SEI CERT-C++ Web site © 2017 Carnegie Mellon University, ”SEI CERT C Coding Standard – Rules for Developing safe, Reliable and Secure systems – 2016 Edition,” © 2016 Carnegie Mellon University, and “SEI CERT C++ Coding Standard – Rules for Developing safe, Reliable and Secure systems in C++ – 2016 Edition” © 2016 Carnegie Mellon University, with special permission from its Software Engineering Institute.

ANY MATERIAL OF CARNEGIE MELLON UNIVERSITY AND/OR ITS SOFTWARE ENGINEERING INSTITUTE CONTAINED HEREIN IS FURNISHED ON AN "AS-IS" BASIS. CARNEGIE MELLON UNIVERSITY MAKES NO WARRANTIES OF ANY KIND, EITHER EXPRESSED OR IMPLIED, AS TO ANY MATTER INCLUDING, BUT NOT LIMITED TO, WARRANTY OF FITNESS FOR PURPOSE OR MERCHANTABILITY, EXCLUSIVITY, OR RESULTS OBTAINED FROM USE OF THE MATERIAL. CARNEGIE MELLON UNIVERSITY DOES NOT MAKE ANY WARRANTY OF ANY KIND WITH RESPECT TO FREEDOM FROM PATENT, TRADEMARK, OR COPYRIGHT INFRINGEMENT.

This software and associated documentation has not been reviewed nor is it endorsed by Carnegie Mellon University or its Software Engineering Institute.