What methods exist for preventing and resolving deadlocks in a system?

A deadlock occurs when two or more threads are permanently blocked, waiting for resources to be released. The following methods are used to prevent and eliminate deadlocks:

1. Avoiding deadlocks:

   • Using a fixed order of lock acquisition (lock ordering).
   • Applying timeouts when attempting to acquire locks.
   • Using non-blocking algorithms and data structures.

2. Detecting deadlocks:

   • Monitoring the state of threads and resources.
   • Analyzing the wait-for graph to identify cycles.

3. Resolving deadlocks:

   • Forcibly interrupting or rolling back one of the threads.
   • Releasing resources and retrying lock acquisition.

In Python, to prevent deadlocks, you can use the threading module with timeouts:

import threading

lock1 = threading.Lock()
lock2 = threading.Lock()

def thread1():
    with lock1:
        if lock2.acquire(timeout=1):
            try:
                # critical section
                pass
            finally:
                lock2.release()

def thread2():
    with lock2:
        if lock1.acquire(timeout=1):
            try:
                # critical section
                pass
            finally:
                lock1.release()

Thus, using timeouts and proper lock acquisition order helps avoid deadlocks.