concurrent.futures.ThreadPoolExecutor does not free memory when shutdown

Bug report

Memory allocated in threads is never freed when using a ThreadPoolExecutor. The expected behavior is that when the executor is shut down, all memory allocated in its threads should be freed. The below code demonstrates this leak in that the the memory usage before allocating memory in threads is significantly less than after.

from concurrent.futures import ThreadPoolExecutor, as_completed
import resource

def process_user(x):
    return bytearray(10000000) 

print('Before', resource.getrusage(resource.RUSAGE_SELF).ru_maxrss/1024, 'MB')

def leak_memory():
    with ThreadPoolExecutor(max_workers=20) as executor:
        futures = [executor.submit(process_user, i) for i in range(100)]
        for future in as_completed(futures):
            cur = resource.getrusage(resource.RUSAGE_SELF).ru_maxrss/1024
            print('Step', cur)

leak_memory()

print('After', resource.getrusage(resource.RUSAGE_SELF).ru_maxrss/1024, 'MB')

Your environment

Python 3.9.x/3.10.x. Mac os and debian.

Reproduisez l’exemple avec Python 3.9 ou 3.10 sous macOS ou Debian, en vous concentrant sur l’arrêt de concurrent.futures.ThreadPoolExecutor et sur la mesure de resource.RUSAGE_SELF ru_maxrss. Établissez si la valeur signalée devrait diminuer après l’arrêt, puis identifiez l’implémentation ou les tests concernés et définissez le comportement mémoire attendu après l’arrêt.