In general, killing threads abruptly is considered a bad programming practice. Killing a thread abruptly might leave a critical resource that must be closed properly, open. But you might want to kill a thread once some specific time period has passed or some interrupt has been generated. There are the various methods by which you can kill a thread in python.
- Raising exceptions in a python thread
- Set/Reset stop flag
- Using traces to kill threads
- Using the multiprocessing module to kill threads
- Killing Python thread by setting it as daemon
- Using a hidden function _stop()
This method uses the function PyThreadState_SetAsyncExc() to raise an exception in the a thread. For Example,
# Python program raising
# exceptions in a python
# thread
import threading
import ctypes
import time
class thread_with_exception(threading.Thread):
def __init__(self, name):
threading.Thread.__init__(self)
self.name = name
def run(self):
# target function of the thread class
try:
while True:
print('running ' + self.name)
finally:
print('ended')
def get_id(self):
# returns id of the respective thread
if hasattr(self, '_thread_id'):
return self._thread_id
for id, thread in threading._active.items():
if thread is self:
return id
def raise_exception(self):
thread_id = self.get_id()
res = ctypes.pythonapi.PyThreadState_SetAsyncExc(thread_id,
ctypes.py_object(SystemExit))
if res > 1:
ctypes.pythonapi.PyThreadState_SetAsyncExc(thread_id, 0)
print('Exception raise failure')
t1 = thread_with_exception('Thread 1')
t1.start()
time.sleep(2)
t1.raise_exception()
t1.join() When we run the code above in a machine and you will notice, as soon as the function raise_exception() is called, the target function run() ends. This is because as soon as an exception is raised, program control jumps out of the try block and run() function is terminated. After that join() function can be called to kill the thread. In the absence of the function run_exception(), the target function run() keeps running forever and join() function is never called to kill the thread.
In order to kill a threads, we can declare a stop flag and this flag will be check occasionally by the thread. For Example
# Python program showing
# how to kill threads
# using set/reset stop
# flag
import threading
import time
def run():
while True:
print('thread running')
global stop_threads
if stop_threads:
break
stop_threads = False
t1 = threading.Thread(target = run)
t1.start()
time.sleep(1)
stop_threads = True
t1.join()
print('thread killed')In the above code, as soon as the global variable stop_threads is set, the target function run() ends and the thread t1 can be killed by using t1.join(). But one may refrain from using global variable due to certain reasons. For those situations, function objects can be passed to provide a similar functionality as shown below.
# Python program killing
# threads using stop
# flag
import threading
import time
def run(stop):
while True:
print('thread running')
if stop():
break
def main():
stop_threads = False
t1 = threading.Thread(target = run, args =(lambda : stop_threads, ))
t1.start()
time.sleep(1)
stop_threads = True
t1.join()
print('thread killed')
main()The function object passed in the above code always returns the value of the local variable stop_threads. This value is checked in the function run(), and as soon as stop_threads is reset, the run() function ends and the thread can be killed.
This methods works by installing traces in each thread. Each trace terminates itself on the detection of some stimulus or flag, thus instantly killing the associated thread. For Example
# Python program using
# traces to kill threads
import sys
import trace
import threading
import time
class thread_with_trace(threading.Thread):
def __init__(self, *args, **keywords):
threading.Thread.__init__(self, *args, **keywords)
self.killed = False
def start(self):
self.__run_backup = self.run
self.run = self.__run
threading.Thread.start(self)
def __run(self):
sys.settrace(self.globaltrace)
self.__run_backup()
self.run = self.__run_backup
def globaltrace(self, frame, event, arg):
if event == 'call':
return self.localtrace
else:
return None
def localtrace(self, frame, event, arg):
if self.killed:
if event == 'line':
raise SystemExit()
return self.localtrace
def kill(self):
self.killed = True
def func():
while True:
print('thread running')
t1 = thread_with_trace(target = func)
t1.start()
time.sleep(2)
t1.kill()
t1.join()
if not t1.isAlive():
print('thread killed')In this code, start() is slightly modified to set the system trace function using settrace(). The local trace function is defined such that, whenever the kill flag (killed) of the respective thread is set, a SystemExit exception is raised upon the excution of the next line of code, which end the execution of the target function func. Now the thread can be killed with join().
The multiprocessing module of Python allows you to spawn processes in the similar way you spawn threads using the threading module. The interface of the multithreading module is similar to that of the threading module. For Example, in a given code we created three threads(processes) which count from 1 to 9.
# Python program creating
# three threads
import threading
import time
# counts from 1 to 9
def func(number):
for i in range(1, 10):
time.sleep(0.01)
print('Thread ' + str(number) + ': prints ' + str(number*i))
# creates 3 threads
for i in range(0, 3):
thread = threading.Thread(target=func, args=(i,))
thread.start() The functionality of the above code can also be implemented by using the multiprocessing module in a similar manner, with very few changes. See the code given below.
# Python program creating
# thread using multiprocessing
# module
import multiprocessing
import time
def func(number):
for i in range(1, 10):
time.sleep(0.01)
print('Processing ' + str(number) + ': prints ' + str(number*i))
for i in range(0, 3):
process = multiprocessing.Process(target=func, args=(i,))
process.start()Though the interface of the two modules is similar, the two modules have very different implementations. All the threads share global variables, whereas processes are completely separate from each other. Hence, killing processes is much safer as compared to killing threads. The Process class is provided a method, terminate(), to kill a process. Now, getting back to the initial problem. Suppose in the above code, we want to kill all the processes after 0.03s have passed. This functionality is achieved using the multiprocessing module in the following code.
# Python program killing
# a thread using multiprocessing
# module
import multiprocessing
import time
def func(number):
for i in range(1, 10):
time.sleep(0.01)
print('Processing ' + str(number) + ': prints ' + str(number*i))
# list of all processes, so that they can be killed afterwards
all_processes = []
for i in range(0, 3):
process = multiprocessing.Process(target=func, args=(i,))
process.start()
all_processes.append(process)
# kill all processes after 0.03s
time.sleep(0.03)
for process in all_processes:
process.terminate()Though the two modules have different implementations. This functionality provided by the multiprocessing module in the above code is similar to killing threads. Hence, the multiprocessing module can be used as a simple alternative whenever we are required to implement the killing of threads in Python.
Daemon threads are those threads which are killed when the main program exits. For Example
import threading
import time
import sys
def func():
while True:
time.sleep(0.5)
print('Thread alive, and it won't die on program termination')
t1 = threading.Thread(target=func)
t1.start()
time.sleep(2)
sys.exit()Notice that, thread t1 stays alive and prevents the main program to exit via sys.exit(). In Python, any alive non-daemon thread blocks the main program to exit. Whereas, daemon threads themselves are killed as soon as the main program exits. In other words, as soon as the main program exits, all the daemon threads are killed. To declare a thread as daemon, we set the keyword argument, daemon as True. For Example in the given code it demonstrates the property of daemon threads.
# Python program killing
# thread using daemon
import threading
import time
import sys
def func():
while True:
time.sleep(0.5)
print('Thread alive, but it will die on program termination')
t1 = threading.Thread(target=func)
t1.daemon = True
t1.start()
time.sleep(2)
sys.exit()Notice that, as soon as the main program exits, the thread t1 is killed. This method proves to be extremely useful in cases where program termination can be used to trigger the killing of threads. Note that in Python, the main program terminates as soon as all the non-daemon threads are dead, irrespective of the number of daemon threads alive. Hence, the resources held by these daemon threads, such as open files, database transactions, etc. may not be released properly. The initial thread of control in a python program is not a daemon thread. Killing a thread forcibly is not recommended unless it is known for sure, that doing so will not cause any leaks or deadlocks.
Using a hidden function _stop() :
In order to kill a thread, we use hidden function _stop() this function is not documented but might disappear in the next version of python.
# Python program killing
# a thread using ._stop()
# function
import time
import threading
class MyThread(threading.Thread):
# Thread class with a _stop() method.
# The thread itself has to check
# regularly for the stopped() condition.
def __init__(self, *args, **kwargs):
super(MyThread, self).__init__(*args, **kwargs)
self._stop = threading.Event()
# function using _stop function
def stop(self):
self._stop.set()
def stopped(self):
return self._stop.isSet()
def run(self):
while True:
if self.stopped():
return
print("Hello, world!")
time.sleep(1)
t1 = MyThread()
t1.start()
time.sleep(5)
t1.stop()
t1.join()Note: Above methods might not work in some situation or another, because python does not provide any direct method to kill threads.