茅厕级C++任务队列——最大的优势就是没有优势

本文主要是介绍茅厕级C++任务队列——最大的优势就是没有优势，对大家解决编程问题具有一定的参考价值，需要的程序猿们随着小编来一起学习吧！

#include <iostream>
#include <functional>
#include <queue>
#include <atomic>
#include <mutex>
#include <thread>
#include <future>
#include <memory>
#include <chrono>
#include <list>
using namespace std;

//状态类别
enum class Status
{
    //停止
    STOP,
    //开始
    START
};
/**
 * @brief 用于初步封装线程
 * 
 */
class ThreadPackage
{
private:
    //线程需要执行的目标函数封装
    shared_ptr<function<void()>> _fun = nullptr;
    //所属线程的智能指针
    shared_ptr<thread> t1 = nullptr;

public:
    //等待释放的线程
    std::mutex _wait_release_m;
    //初始线程状态
    Status _thread_status = Status::STOP;
    //默认构造函数
    ThreadPackage(){};
    //析构函数
    ~ThreadPackage();
    //设置目标功能函数
    void setWork(shared_ptr<function<void()>> fun);
    //开始执行
    void start();
};
ThreadPackage::~ThreadPackage()
{
    //查看是否执行完成，未执行完成则等待完成
    if (t1->joinable())
    {
        t1->join();
    }
};
void ThreadPackage::start()
{
    //初始化状态
    _thread_status = Status::START;
    // 封装线程
    t1 = make_shared<thread>([this]() {
        std::unique_lock<std::mutex> lk(_wait_release_m);
        this->_thread_status = Status::START;
        (*(this->_fun))();
        this->_thread_status = Status::STOP;
    });
}
void ThreadPackage::setWork(shared_ptr<function<void()>> fun)
{
    this->_fun = fun;
}

class ThreadPool
{

public:
    ThreadPool(int pool_size);
    ThreadPool(){};
    ~ThreadPool(){};
    // 添加任务
    template <typename F, typename... Args>
    decltype(auto) add_task(F &&func, Args &&...args);

private:
    //检查线程是否执行完成
    shared_ptr<thread> t_check;
    //用于释放已结束的线程
    shared_ptr<thread> t_realse;
    //添加任务到执行队列的线程
    shared_ptr<thread> t_dispatch;

    //同时执行的最大线程数
    std::atomic<int> _pool_size;
    //线程池状态
    Status _status = Status::STOP;
    //待执行任务队列
    queue<function<void()>> _wait_work_q;
    //等待添加到待执行任务队列锁
    std::mutex _wait_add_work_m;
    //等待添加到待执行任务队列的队列
    queue<function<void()>> _add_wait_work_q;
    //正在工作线程列表的锁
    std::mutex _working_m;
    //执行的线程链表
    list<shared_ptr<ThreadPackage>> _starting_l;
    //等待释放列表锁
    std::mutex _wait_release_m;
    //等待释放的线程链表
    list<shared_ptr<ThreadPackage>> _wait_release_l;
    //释放函数
    void release();
    //添加任务到执行队列的函数
    void dispatch();
    //检查线程是否执行完成的函数
    void check();
};
void ThreadPool::release()
{
    while (true)
    {
        {
            std::unique_lock<std::mutex> lk(_wait_release_m);
            for (auto i = _wait_release_l.begin(); i != _wait_release_l.end(); ++i)
            {
                if ((*i)->_thread_status == Status::STOP)
                {
                    std::unique_lock<std::mutex> lk((*i)->_wait_release_m);
                    _wait_release_l.erase(i);
                    break;
                }
            }
        }

        std::this_thread::yield();
    }
}

void ThreadPool::dispatch()
{
    while (true)
    {
        if (_wait_work_q.size() > 0 && _starting_l.size() < _pool_size)
        {
            auto _temp_wait_task = _wait_work_q.front();
            _wait_work_q.pop();
            auto _temp_thread_package = std::make_shared<ThreadPackage>();
            _temp_thread_package->setWork(make_shared<function<void()>>(_temp_wait_task));
            _temp_thread_package->start();
            std::unique_lock<std::mutex> lk(_working_m);
            _starting_l.emplace_back(_temp_thread_package);
        }
        if (_add_wait_work_q.size() > 0)
        {
            {
                std::unique_lock<std::mutex> lk(_wait_add_work_m);
                auto _temp_add_wait_task = _add_wait_work_q.front();
                _add_wait_work_q.pop();
                _wait_work_q.push(_temp_add_wait_task);
            }
        }
        std::this_thread::yield();
    }
}

void ThreadPool::check()
{
    while (true)
    {
        {
            std::unique_lock<std::mutex> lk(_working_m);
            for (auto i = _starting_l.begin(); i != _starting_l.end(); ++i)
            {

                if ((*i)->_thread_status == Status::STOP)
                {
                    std::unique_lock<std::mutex> lk(_wait_release_m);
                    _wait_release_l.emplace_back(*i);
                    i = _starting_l.erase(i);
                    break;
                }
            }
        }
        std::this_thread::yield();
    }
}
ThreadPool::ThreadPool(int pool_size)
{
    this->_pool_size = pool_size;
    t_check = make_shared<thread>(bind(&ThreadPool::check, this));
    t_realse = make_shared<thread>(bind(&ThreadPool::release, this));
    t_dispatch = make_shared<thread>(bind(&ThreadPool::dispatch, this));
}

/**
 * @brief 主要核心思想就这一个模板函数，用于将原始函数封装成
 * 
 * @tparam F 目标函数类型
 * @tparam Args 函数的参数类型
 * @param func 目标函数
 * @param args 目标参数
 * @return decltype(auto) 自动推导返回值
 */
template <typename F, typename... Args>
decltype(auto) ThreadPool::add_task(F &&func, Args &&...args)
{
    //推导最后的返回值类型
    using res_type = typename std::result_of<F(Args...)>::type;
    //封装为packaged_task的智能指针类型
    auto task = std::make_shared<std::packaged_task<res_type()>>(
        std::bind(std::forward<F>(func), std::forward<Args>(args)...));
    //获取用于异步胡获取返回值的future
    std::future<res_type> res = task->get_future();
    {
        //尝试获取待添加队列锁
        std::unique_lock<std::mutex> lk(_wait_add_work_m);
        //添加到待添加任务队列
        _add_wait_work_q.emplace([task]() { (*task)(); });
    }
    return res;
}
int wc(int n)
{
    int all = 1;
    for (int i = 1; i < 100000000; ++i)
    {
        all *= i;
        all %= 100000007;
    }
    cout << n << " ";
    return all;
}
int main()
{
    vector<future<int>> temp;
    auto s1 = clock();
    ThreadPool pt(10);

    for (int i = 0; i < 8; ++i)
    {
        temp.emplace_back(pt.add_task(wc, 1));
    }
    auto s2 = clock();
    for (int i = 0; i < temp.size(); ++i)
    {
        temp[i].get();
    }
    auto s3 = clock();
    cout << endl;
    for (int i = 0; i < 8; ++i)
    {
        wc(2);
    }
    auto s4 = clock();
    cout << endl;
    thread t1(wc, 3), t2(wc, 3), t3(wc, 3), t4(wc, 3), t5(wc, 3), t6(wc, 3), t7(wc, 3), t8(wc, 3);
    t1.join();
    t2.join();
    t3.join();
    t4.join();
    t5.join();
    t6.join();
    t7.join();
    t8.join();
    auto s5 = clock();
    cout << "end\n";
    std::cout << (s2 - s1) << "\t" << (s3 - s1) << "\t" << (s4 - s3) << "\t" << (s5 - s4) << std::endl;

    while (true)
    {
        std::this_thread::yield();
    }

    getchar();
    return 0;
}

这篇关于茅厕级C++任务队列——最大的优势就是没有优势的文章就介绍到这儿，希望我们推荐的文章对大家有所帮助，也希望大家多多支持为之网！