标签:结果 lifo llb class app 生产者消费者模型 func gil for
#线程中是不是会产生数据不安全
# 共享内存
a = 0
def add_f():
global a
for i in range(200000):
a += 1
?
def sub_f():
global a
for i in range(200000):
a -= 1
?
from threading import Thread
?
t1 = Thread(target=add_f)
t1.start()
t2 = Thread(target=sub_f)
t2.start()
t1.join()
t2.join()
print(a)
a = 0
def func():
global a
a -= 1
import dis
dis.dis(func)
?
#即便是线程 即便有GIL 也会出现数据不安全的问题
# 1.操作的是全局变量
# 2.做一下操作
# += -= *= /+ 先计算再赋值才容易出现数据不安全的问题
# 包括 lst[0] += 1 dic[‘key‘]-=1
?
a = 0
def func():
global a
a += 1
?
import dis
dis.dis(func)
?
?
a = 0
def add_f(lock):
global a
for i in range(200000):
with lock:
a += 1
?
def sub_f(lock):
global a
for i in range(200000):
with lock:
a -= 1
?
from threading import Thread,Lock
lock = Lock()
t1 = Thread(target=add_f,args=(lock,))
t1.start()
t2 = Thread(target=sub_f,args=(lock,))
t2.start()
t1.join()
t2.join()
print(a)
#加锁会影响程序的执行效率,但是保证了数据的安全
?
#互斥锁是锁中的一种:在同一个线程中,不能连续acquire多次
from threading import Lock
lock = Lock()
lock.acquire()
print(‘*‘*20)
lock.release()
lock.acquire()
print(‘-‘*20)
lock.