python内存泄漏的一个简单例子及简易定位方法

以下面的 huge_mem 为例子，最终可以看到随着每次操作，该对象膨胀到十分巨大

例子及定位方法

# -*- coding: utf-8 -*-
import sys
import tracemalloc
import random
import weakref

from PyQt5.QtWidgets import QWidget, QPushButton, QApplication, QVBoxLayout, QHBoxLayout


class Person(object):

    def __init__(self, name=""):
        self._name = name
        self._best_friend = None

    @property
    def best_friend(self):
        return self._best_friend

    @best_friend.setter
    def best_friend(self, friend):
        self._best_friend = weakref.ref(friend)


class MemoryLeak(QWidget):

    def __init__(self):
        super().__init__()
        self.mem_list = []
        # 在程序启动的时候开始跟踪内存分配
        tracemalloc.start()
        self.zhang_san = Person(name='张三')
        self.li_si = Person("李四")
        self.initUI()

    def initUI(self):
        self.vLayout = QVBoxLayout()
        self.twoButtonShow()
        self.setLayout(self.vLayout)
        self.setWindowTitle('内存泄漏测试')
        self.show()

    def twoButtonShow(self):
        hLayout = QHBoxLayout()
        self.vLayout.addLayout(hLayout)
        leakBtn = QPushButton('开始泄露', self)
        getLeakPointBtn = QPushButton('获取泄漏点', self)
        hLayout.addWidget(leakBtn)
        hLayout.addWidget(getLeakPointBtn)
        leakBtn.clicked.connect(self.startLeak)
        getLeakPointBtn.clicked.connect(self.getLeakPoint)

    def startLeak(self):
        for _ in range(10000):
            huge_mem = random.random()
            self.mem_list.append(huge_mem)

    def getLeakPoint(self):
        # 获取Python分配的内存块traceback的快照
        snapshot = tracemalloc.take_snapshot()
        top_stats = snapshot.statistics('lineno')
        # 打印内存占用率最高的前十
        print("[ Top 10 ]")
        for stat in top_stats[:10]:
            print(stat)


if __name__ == '__main__':
    app = QApplication(sys.argv)
    ex = MemoryLeak()
    sys.exit(app.exec_())

在程序刚启动的时候点击一次**“获取内存”**：可以看到如下输出

[ Top 10 ]
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:38: size=15.6 KiB, count=202, average=79 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:47: size=13.5 KiB, count=184, average=75 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:45: size=1416 B, count=16, average=88 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:71: size=1112 B, count=1, average=1112 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:39: size=440 B, count=1, average=440 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:48: size=216 B, count=2, average=108 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:13: size=144 B, count=3, average=48 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:72: size=108 B, count=1, average=108 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:51: size=72 B, count=1, average=72 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:34: size=48 B, count=1, average=48 B

随后我们触发10次 “开始泄露”，重新获取内存可以得到如下：

F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:56: size=2344 KiB, count=99996, average=24 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:57: size=805 KiB, count=1, average=805 KiB
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:38: size=15.6 KiB, count=202, average=79 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:47: size=13.5 KiB, count=184, average=75 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:45: size=1416 B, count=16, average=88 B
E:\Python3.8.5\lib\tracemalloc.py:509: size=1248 B, count=3, average=416 B
F:/PythonXSLWorkSpace/PythonBaseUse/MemoryLeak/bigLeak.py:71: size=1112 B, count=1, average=1112 B
E:\Python3.8.5\lib\tracemalloc.py:209: size=896 B, count=3, average=299 B
E:\Python3.8.5\lib\tracemalloc.py:165: size=816 B, count=2, average=408 B
E:\Python3.8.5\lib\tracemalloc.py:65: size=640 B, count=10, average=64 B

触发多次泄露之后可以看到此时占据榜首的是56行也就是 huge_mem 所在行，由此我们可以定义到内存泄露的对象及所在位置

PS.我们要在内存开始泄露前追踪内存分配，也就是调用tracemalloc.start()
tracemalloc文档