深入理解epoll原理

这里先补充一下三种IO多路复用的应用场景

1. select 应用场景
select 的 timeout 参数精度为 1ns，而 poll 和 epoll 为 1ms，因此 select 更加适用于实时性要求比较高的场景，比如核反应堆的控制。select 可移植性更好，几乎被所有主流平台所支持。

2. poll 应用场景
poll 没有最大描述符数量的限制，如果平台支持并且对实时性要求不高，应该使用 poll 而不是 select。

3. epoll 应用场景
只需要运行在 Linux 平台上，有大量的描述符需要同时轮询，并且这些连接最好是长连接。需要同时监控小于 1000 个描述符，就没有必要使用 epoll，因为这个应用场景下并不能体现 epoll 的优势。需要监控的描述符状态变化多，而且都是非常短暂的，也没有必要使用 epoll。因为 epoll 中的所有描述符都存储在内核中，造成每次需要对描述符的状态改变都需要通过 epoll_ctl() 进行系统调用，频繁系统调用降低效率。并且 epoll 的描述符存储在内核，不容易调试。

1.epoll相比poll，select的优点

（1）select，poll实现需要自己不断轮询所有fd集合，直到设备就绪，期间可能要睡眠和唤醒多次交替。而epoll其实也需要调用epoll_wait不断轮询就绪链表，期间也可能多次睡眠和唤醒交替，但是它是设备就绪时，调用回调函数，把就绪fd放入就绪链表中，并唤醒在epoll_wait中进入睡眠的进程。虽然都要睡眠和交替，但是select和poll在“醒着”的时候要遍历整个fd集合，而epoll在“醒着”的时候只要判断一下就绪链表是否为空就行了，这节省了大量的CPU时间。这就是回调机制带来的性能提升。

（2）select，poll每次调用都要把fd集合从用户态往内核态拷贝一次，并且要把current往设备等待队列中挂一次，而epoll在注册时候已经把fd拷贝到内核空间中，epollwait只要一次拷贝，而且把current往等待队列上挂也只挂一次（在epoll_wait的开始，注意这里的等待队列并不是设备等待队列，只是一个epoll内部定义的等待队列）。这也能节省不少的开销。

epoll相对于poll不同在于，epoll在Linux内核中申请了简易文件系统，把poll调用分成三个步骤：调用epoll_create建立epoll对象，即在epoll文件系统中给这个句柄分配fd资源，调用epoll_ctl向epoll对象中添加100万个连接的套接字，调用epoll_wait收集发生事件的连接。

在epoll_ctl函数中。每次注册新的事件到epoll句柄中时（在epoll_ctl中指定EPOLL_CTL_ADD），会把所有的fd拷贝进内核，而不是在epoll_wait的时候重复拷贝。epoll保证了每个fd在整个过程中只会拷贝一次。poll的解决方案不像select或poll一样每次都把current轮流加入fd对应的设备等待队列中，而只在epoll_ctl时把current挂一遍（这一遍必不可少）并为每个fd指定一个回调函数，当设备就绪，唤醒等待队列上的等待者时，就会调用这个回调函数，而这个回调函数会把就绪的fd加入一个就绪链表）。epoll_wait的工作实际上就是在这个就绪链表中查看有没有就绪的fd（利用schedule_timeout()实现睡一会，判断一会的效果，和select实现中的第7步是类似的）。它所支持的FD上限是最大可以打开文件的数目，这个数字一般远大于2048,举个例子,在1GB内存的机器上大约是10万左右，具体数目可以cat /proc/sys/fs/file-max察看,一般来说这个数目和系统内存关系很大。

 

注意，epoll是基于slab高速缓存创建epitem，对于频繁使用或销毁的结构特别高效，epi_cache也是由内核模块初始化时调用eventpoll_init进行创建。可见不论是epoll fd本身还是epitem都是基于内核内存分配创建，没有其他博客的mmap之说。

 

2.epoll的源码解析

1.执行epoll_create时，创建了红黑树和就绪list链表。

2.执行epoll_ctl时，如果增加fd（socket），则检查在红黑树中是否存在，存在立即返回，不存在则添加到红黑树上，然后向内核注册回调函数，用于当中断事件来临时向准备就绪list链表中插入数据。

3.执行epoll_wait时立刻返回准备就绪链表里的数据即可。

在深入了解epoll的实现之前, 先来了解内核的3个方面

1. 等待队列 waitqueue

我们简单解释一下等待队列:队列头(wait_queue_head_t)往往是资源生产者, 队列成员(wait_queue_t)往往是资源消费者,当头的资源ready后, 会逐个执行每个成员指定的回调函数,来通知它们资源已经ready了, 等待队列大致就这个意思.

2. 内核的poll机制

被Poll的fd, 必须在实现上支持内核的Poll技术,比如fd是某个字符设备,或者是个socket, 它必须实现file_operations中的poll操作, 给自己分配有一个等待队列头.主动poll fd的某个进程必须分配一个等待队列成员, 添加到fd的对待队列里面去, 并指定资源ready时的回调函数.用socket做例子, 它必须有实现一个poll操作, 这个Poll是发起轮询的代码必须主动调用的, 该函数中必须调用poll_wait(),poll_wait会将发起者作为等待队列成员加入到socket的等待队列中去.这样socket发生状态变化时可以通过队列头逐个通知所有关心它的进程.这一点必须很清楚的理解, 否则会想不明白epoll是如何得知fd的状态发生变化的.

重要数据结构： 每创建一个epollfd, 内核就会分配一个eventpoll与之对应, 可以说是内核态的epollfd.

struct eventpoll {
    /* Protect the this structure access */
    spinlock_t lock;
    /*
     * This mutex is used to ensure that files are not removed
     * while epoll is using them. This is held during the event
     * collection loop, the file cleanup path, the epoll file exit
     * code and the ctl operations.
     */
    /* 添加, 修改或者删除监听fd的时候, 以及epoll_wait返回, 向用户空间
     * 传递数据时都会持有这个互斥锁, 所以在用户空间可以放心的在多个线程
     * 中同时执行epoll相关的操作, 内核级已经做了保护. */
    struct mutex mtx;
    /* Wait queue used by sys_epoll_wait() */
    /* 调用epoll_wait()时, 我们就是"睡"在了这个等待队列上... */
    wait_queue_head_t wq;
    /* Wait queue used by file->poll() */
    /* 这个用于epollfd本事被poll的时候... */
    wait_queue_head_t poll_wait;
    /* List of ready file descriptors */
    /* 所有已经ready的epitem都在这个链表里面 */
    struct list_head rdllist;
    /* RB tree root used to store monitored fd structs */
    /* 所有要监听的epitem都在这里 */
    struct rb_root rbr;
    /*
        这是一个单链表链接着所有的struct epitem当event转移到用户空间时
     */
     * This is a single linked list that chains all the "struct epitem" that
     * happened while transfering ready events to userspace w/out
     * holding ->lock.
     */
    struct epitem *ovflist;
    /* The user that created the eventpoll descriptor */
    /* 这里保存了一些用户变量, 比如fd监听数量的最大值等等 */
    struct user_struct *user;
};

/* epitem 表示一个被监听的fd */
struct epitem {
    /* RB tree node used to link this structure to the eventpoll RB tree */
    /* rb_node, 当使用epoll_ctl()将一批fds加入到某个epollfd时, 内核会分配
     * 一批的epitem与fds们对应, 而且它们以rb_tree的形式组织起来, tree的root
     * 保存在epollfd, 也就是struct eventpoll中.
     * 在这里使用rb_tree的原因我认为是提高查找,插入以及删除的速度.
     * rb_tree对以上3个操作都具有O(lgN)的时间复杂度 */
    struct rb_node rbn;
    /* List header used to link this structure to the eventpoll ready list */
    /* 链表节点, 所有已经ready的epitem都会被链到eventpoll的rdllist中 */
    struct list_head rdllink;
    /*
     * Works together "struct eventpoll"->ovflist in keeping the
     * single linked chain of items.
     */
    /* 这个在代码中再解释... */
    struct epitem *next;
    /* The file descriptor information this item refers to */
    /* epitem对应的fd和struct file */
    struct epoll_filefd ffd;
    /* Number of active wait queue attached to poll operations */
    int nwait;
    /* List containing poll wait queues */
    struct list_head pwqlist;
    /* The "container" of this item */
    /* 当前epitem属于哪个eventpoll */
    struct eventpoll *ep;
    /* List header used to link this item to the "struct file" items list */
    struct list_head fllink;
    /* The structure that describe the interested events and the source fd */
    /* 当前的epitem关系哪些events, 这个数据是调用epoll_ctl时从用户态传递过来 */
    struct epoll_event event;
};
struct epoll_filefd {
    struct file *file;
    int fd;
};
/* poll所用到的钩子Wait structure used by the poll hooks */
struct eppoll_entry {
    /* List header used to link this structure to the "struct epitem" */
    struct list_head llink;
    /* The "base" pointer is set to the container "struct epitem" */
    struct epitem *base;
    /*
     * Wait queue item that will be linked to the target file wait
     * queue head.
     */
    wait_queue_t wait;
    /* The wait queue head that linked the "wait" wait queue item */
    wait_queue_head_t *whead;
};
/* Wrapper struct used by poll queueing */
struct ep_pqueue {
    poll_table pt;
    struct epitem *epi;
};
/* Used by the ep_send_events() function as callback private data */
struct ep_send_events_data {
    int maxevents;
    struct epoll_event __user *events;
};

epoll_create内核源码：

SYSCALL_DEFINE1(epoll_create1, int, flags)
{
    int error;
    struct eventpoll *ep = NULL;//主描述符
    /* Check the EPOLL_* constant for consistency.  */
    /* 这句没啥用处... */
    BUILD_BUG_ON(EPOLL_CLOEXEC != O_CLOEXEC);
    /* 对于epoll来讲, 目前唯一有效的FLAG就是CLOEXEC */
    if (flags & ~EPOLL_CLOEXEC)
        return -EINVAL;
    /*
     * Create the internal data structure ("struct eventpoll").
     */
    /* 分配一个struct eventpoll, 分配和初始化细节我们随后深聊~ */
    error = ep_alloc(&ep);
    if (error < 0)
        return error;
    /*
     * Creates all the items needed to setup an eventpoll file. That is,
     * a file structure and a free file descriptor.
     */
    /* 这里是创建一个匿名fd, epollfd本身并不存在一个真正的文件与之对应, 所以内核需要创建一个
     * "虚拟"的文件, 并为之分配真正的struct file结构, 而且有真正的fd.
     * 这里2个参数比较关键:
     * eventpoll_fops, fops就是file operations, 就是当你对这个文件(这里是虚拟的)进行操作(比如读)时,
     * fops里面的函数指针指向真正的操作实现, 类似C++里面虚函数和子类的概念.
     * epoll只实现了poll和release(就是close)操作, 其它文件系统操作都有VFS全权处理了.
     * ep, ep就是struct epollevent, 它会作为一个私有数据保存在struct file的private指针里面.
     * 其实说白了, 就是为了能通过fd找到struct file, 通过struct file能找到eventpoll结构.
     */
    error = anon_inode_getfd("[eventpoll]", &eventpoll_fops, ep,
                 O_RDWR | (flags & O_CLOEXEC));
    if (error < 0)
        ep_free(ep);
    return error;
}

epoll_ctl内核源码：

/*
* 来看epoll_ctl
* epfd 就是epollfd
* op ADD,MOD,DEL
* fd 需要监听的描述符
* event 我们关心的events
*/
SYSCALL_DEFINE4(epoll_ctl, int, epfd, int, op, int, fd,
        struct epoll_event __user *, event)
{
    int error;
    struct file *file, *tfile;
    struct eventpoll *ep;
    struct epitem *epi;
    struct epoll_event epds;
    error = -EFAULT;
    /*
     * 错误处理以及从用户空间将epoll_event结构copy到内核空间.
     */
    if (ep_op_has_event(op) &&
        copy_from_user(&epds, event, sizeof(struct epoll_event)))
        goto error_return;
    /* Get the "struct file *" for the eventpoll file */
    /* 取得struct file结构, epfd既然是真正的fd, 那么内核空间
     * 就会有与之对于的一个struct file结构
     * 这个结构在epoll_create1()中, 由函数anon_inode_getfd()分配 */
    error = -EBADF;
    file = fget(epfd);
    if (!file)
        goto error_return;
    /* Get the "struct file *" for the target file */
    /* 我们需要监听的fd, 它当然也有个struct file结构, 上下2个不要搞混了哦 */
    tfile = fget(fd);
    if (!tfile)
        goto error_fput;
    /* The target file descriptor must support poll */
    error = -EPERM;
    /* 如果监听的文件不支持poll, 那就没辙了.
     * 你知道什么情况下, 文件会不支持poll吗?
     */
    if (!tfile->f_op || !tfile->f_op->poll)
        goto error_tgt_fput;
    /*
     * We have to check that the file structure underneath the file descriptor
     * the user passed to us _is_ an eventpoll file. And also we do not permit
     * adding an epoll file descriptor inside itself.
     */
    error = -EINVAL;
    /* epoll不能自己监听自己... */
    if (file == tfile || !is_file_epoll(file))
        goto error_tgt_fput;
    /*
     * At this point it is safe to assume that the "private_data" contains
     * our own data structure.
     */
    /* 取到我们的eventpoll结构, 来自与epoll_create1()中的分配 */
    ep = file->private_data;
    /* 接下来的操作有可能修改数据结构内容, 锁之~ */
    mutex_lock(&ep->mtx);
    /*
     * Try to lookup the file inside our RB tree, Since we grabbed "mtx"
     * above, we can be sure to be able to use the item looked up by
     * ep_find() till we release the mutex.
     */
    /* 对于每一个监听的fd, 内核都有分配一个epitem结构,
     * 而且我们也知道, epoll是不允许重复添加fd的,
     * 所以我们首先查找该fd是不是已经存在了.
     * ep_find()其实就是RBTREE查找, 跟C++STL的map差不多一回事, O(lgn)的时间复杂度.
     */
    epi = ep_find(ep, tfile, fd);
    error = -EINVAL;
    switch (op) {
        /* 首先我们关心添加 */
    case EPOLL_CTL_ADD:
        if (!epi) {
            /* 之前的find没有找到有效的epitem, 证明是第一次插入, 接受!
             * 这里我们可以知道, POLLERR和POLLHUP事件内核总是会关心的
             * */
            epds.events |= POLLERR | POLLHUP;
            /* rbtree插入, 详情见ep_insert()的分析
             * 其实我觉得这里有insert的话, 之前的find应该
             * 是可以省掉的... */
            error = ep_insert(ep, &epds, tfile, fd);
        } else
            /* 找到了!? 重复添加! */
            error = -EEXIST;
        break;
        /* 删除和修改操作都比较简单 */
    case EPOLL_CTL_DEL:
        if (epi)
            error = ep_remove(ep, epi);
        else
            error = -ENOENT;
        break;
    case EPOLL_CTL_MOD:
        if (epi) {
            epds.events |= POLLERR | POLLHUP;
            error = ep_modify(ep, epi, &epds);
        } else
            error = -ENOENT;
        break;
    }
    mutex_unlock(&ep->mtx);
error_tgt_fput:
    fput(tfile);
error_fput:
    fput(file);
error_return:
    return error;
}
/* 分配一个eventpoll结构 */
static int ep_alloc(struct eventpoll **pep)
{
    int error;
    struct user_struct *user;
    struct eventpoll *ep;
    /* 获取当前用户的一些信息, 比如是不是root啦, 最大监听fd数目啦 */
    user = get_current_user();
    error = -ENOMEM;
    ep = kzalloc(sizeof(*ep), GFP_KERNEL);
    if (unlikely(!ep))
        goto free_uid;
    /* 这些都是初始化啦 */
    spin_lock_init(&ep->lock);
    mutex_init(&ep->mtx);
    init_waitqueue_head(&ep->wq);//初始化自己睡在的等待队列
    init_waitqueue_head(&ep->poll_wait);//初始化
    INIT_LIST_HEAD(&ep->rdllist);//初始化就绪链表
    ep->rbr = RB_ROOT;
    ep->ovflist = EP_UNACTIVE_PTR;
    ep->user = user;
    *pep = ep;
    return 0;
free_uid:
    free_uid(user);
    return error;
}
/*
 * Must be called with "mtx" held.
 */
/*
 * ep_insert()在epoll_ctl()中被调用, 完成往epollfd里面添加一个监听fd的工作
 * tfile是fd在内核态的struct file结构
 */
static int ep_insert(struct eventpoll *ep, struct epoll_event *event,
             struct file *tfile, int fd)
{
    int error, revents, pwake = 0;
    unsigned long flags;
    struct epitem *epi;
    struct ep_pqueue epq;
    /* 查看是否达到当前用户的最大监听数 */
    if (unlikely(atomic_read(&ep->user->epoll_watches) >=
             max_user_watches))
        return -ENOSPC;
    /* 从著名的slab中分配一个epitem */
    if (!(epi = kmem_cache_alloc(epi_cache, GFP_KERNEL)))
        return -ENOMEM;
    /* Item initialization follow here ... */
    /* 这些都是相关成员的初始化... */
    INIT_LIST_HEAD(&epi->rdllink);
    INIT_LIST_HEAD(&epi->fllink);
    INIT_LIST_HEAD(&epi->pwqlist);
    epi->ep = ep;
    /* 这里保存了我们需要监听的文件fd和它的file结构 */
    ep_set_ffd(&epi->ffd, tfile, fd);
    epi->event = *event;
    epi->nwait = 0;
    /* 这个指针的初值不是NULL哦... */
    epi->next = EP_UNACTIVE_PTR;
    /* Initialize the poll table using the queue callback */
    /* 好, 我们终于要进入到poll的正题了 */
    epq.epi = epi;
    /* 初始化一个poll_table
     * 其实就是指定调用poll_wait(注意不是epoll_wait!!!)时的回调函数,和我们关心哪些events,
     * ep_ptable_queue_proc()就是我们的回调啦, 初值是所有event都关心 */
    init_poll_funcptr(&epq.pt, ep_ptable_queue_proc);
    /*
     * Attach the item to the poll hooks and get current event bits.
     * We can safely use the file* here because its usage count has
     * been increased by the caller of this function. Note that after
     * this operation completes, the poll callback can start hitting
     * the new item.
     */
    /* 这一部很关键, 也比较难懂, 完全是内核的poll机制导致的...
     * 首先, f_op->poll()一般来说只是个wrapper, 它会调用真正的poll实现,
     * 拿UDP的socket来举例, 这里就是这样的调用流程: f_op->poll(), sock_poll(),
     * udp_poll(), datagram_poll(), sock_poll_wait(), 最后调用到我们上面指定的
     * ep_ptable_queue_proc()这个回调函数...(好深的调用路径...).
     * 完成这一步, 我们的epitem就跟这个socket关联起来了, 当它有状态变化时,
     * 会通过ep_poll_callback()来通知.
     * 最后, 这个函数还会查询当前的fd是不是已经有啥event已经ready了, 有的话
     * 会将event返回. */
    revents = tfile->f_op->poll(tfile, &epq.pt);
    /*
     * We have to check if something went wrong during the poll wait queue
     * install process. Namely an allocation for a wait queue failed due
     * high memory pressure.
     */
    error = -ENOMEM;
    if (epi->nwait < 0)
        goto error_unregister;
    /* Add the current item to the list of active epoll hook for this file */
    /* 这个就是每个文件会将所有监听自己的epitem链起来 */
    spin_lock(&tfile->f_lock);
    list_add_tail(&epi->fllink, &tfile->f_ep_links);
    spin_unlock(&tfile->f_lock);
    /*
     * Add the current item to the RB tree. All RB tree operations are
     * protected by "mtx", and ep_insert() is called with "mtx" held.
     */
    /* 都搞定后, 将epitem插入到对应的eventpoll中去 */
    ep_rbtree_insert(ep, epi);
    /* We have to drop the new item inside our item list to keep track of it */
    spin_lock_irqsave(&ep->lock, flags);
    /* If the file is already "ready" we drop it inside the ready list */
    /* 到达这里后, 如果我们监听的fd已经有事件发生, 那就要处理一下 */
    if ((revents & event->events) && !ep_is_linked(&epi->rdllink)) {
        /* 将当前的epitem加入到ready list中去 */
        list_add_tail(&epi->rdllink, &ep->rdllist);
        /* Notify waiting tasks that events are available */
        /* 谁在epoll_wait, 就唤醒它... */
        if (waitqueue_active(&ep->wq))
            wake_up_locked(&ep->wq);
        /* 谁在epoll当前的epollfd, 也唤醒它... */
        if (waitqueue_active(&ep->poll_wait))
            pwake++;
    }
    spin_unlock_irqrestore(&ep->lock, flags);
    atomic_inc(&ep->user->epoll_watches);
    /* We have to call this outside the lock */
    if (pwake)
        ep_poll_safewake(&ep->poll_wait);
    return 0;
error_unregister:
    ep_unregister_pollwait(ep, epi);
    /*
     * We need to do this because an event could have been arrived on some
     * allocated wait queue. Note that we don't care about the ep->ovflist
     * list, since that is used/cleaned only inside a section bound by "mtx".
     * And ep_insert() is called with "mtx" held.
     */
    spin_lock_irqsave(&ep->lock, flags);
    if (ep_is_linked(&epi->rdllink))
        list_del_init(&epi->rdllink);
    spin_unlock_irqrestore(&ep->lock, flags);
    kmem_cache_free(epi_cache, epi);
    return error;
}
/*
 * This is the callback that is used to add our wait queue to the
 * target file wakeup lists.
 */
/*
 * 该函数在调用f_op->poll()时会被调用.
 * 也就是epoll主动poll某个fd时, 用来将epitem与指定的fd关联起来的.
 * 关联的办法就是使用等待队列(waitqueue)
 */
static void ep_ptable_queue_proc(struct file *file, wait_queue_head_t *whead,
                 poll_table *pt)
{
    struct epitem *epi = ep_item_from_epqueue(pt);
    struct eppoll_entry *pwq;
    if (epi->nwait >= 0 && (pwq = kmem_cache_alloc(pwq_cache, GFP_KERNEL))) {
        /* 初始化等待队列, 指定ep_poll_callback为唤醒时的回调函数,
         * 当我们监听的fd发生状态改变时, 也就是队列头被唤醒时,
         * 指定的回调函数将会被调用. */
        init_waitqueue_func_entry(&pwq->wait, ep_poll_callback);
        pwq->whead = whead;
        pwq->base = epi;
        /* 将刚分配的等待队列成员加入到头中, 头是由fd持有的 */
        add_wait_queue(whead, &pwq->wait);
        list_add_tail(&pwq->llink, &epi->pwqlist);
        /* nwait记录了当前epitem加入到了多少个等待队列中,
         * 我认为这个值最大也只会是1... */
        epi->nwait++;
    } else {
        /* We have to signal that an error occurred */
        epi->nwait = -1;
    }
}
/*
 * This is the callback that is passed to the wait queue wakeup
 * machanism. It is called by the stored file descriptors when they
 * have events to report.
 */
/*
 * 这个是关键性的回调函数, 当我们监听的fd发生状态改变时, 它会被调用.
 * 参数key被当作一个unsigned long整数使用, 携带的是events.
 */
static int ep_poll_callback(wait_queue_t *wait, unsigned mode, int sync, void *key)
{
    int pwake = 0;
    unsigned long flags;
    struct epitem *epi = ep_item_from_wait(wait);//从等待队列获取epitem.需要知道哪个进程挂载到这个设备
    struct eventpoll *ep = epi->ep;//获取
    spin_lock_irqsave(&ep->lock, flags);
    /*
     * If the event mask does not contain any poll(2) event, we consider the
     * descriptor to be disabled. This condition is likely the effect of the
     * EPOLLONESHOT bit that disables the descriptor when an event is received,
     * until the next EPOLL_CTL_MOD will be issued.
     */
    if (!(epi->event.events & ~EP_PRIVATE_BITS))
        goto out_unlock;
    /*
     * Check the events coming with the callback. At this stage, not
     * every device reports the events in the "key" parameter of the
     * callback. We need to be able to handle both cases here, hence the
     * test for "key" != NULL before the event match test.
     */
    /* 没有我们关心的event... */
    if (key && !((unsigned long) key & epi->event.events))
        goto out_unlock;
    /*
     * If we are trasfering events to userspace, we can hold no locks
     * (because we're accessing user memory, and because of linux f_op->poll()
     * semantics). All the events that happens during that period of time are
     * chained in ep->ovflist and requeued later on.
     */
    /*
     * 这里看起来可能有点费解, 其实干的事情比较简单:
     * 如果该callback被调用的同时, epoll_wait()已经返回了,
     * 也就是说, 此刻应用程序有可能已经在循环获取events,
     * 这种情况下, 内核将此刻发生event的epitem用一个单独的链表
     * 链起来, 不发给应用程序, 也不丢弃, 而是在下一次epoll_wait
     * 时返回给用户.
     */
    if (unlikely(ep->ovflist != EP_UNACTIVE_PTR)) {
        if (epi->next == EP_UNACTIVE_PTR) {
            epi->next = ep->ovflist;
            ep->ovflist = epi;
        }
        goto out_unlock;
    }
    /* If this file is already in the ready list we exit soon */
    /* 将当前的epitem放入ready list */
    if (!ep_is_linked(&epi->rdllink))
        list_add_tail(&epi->rdllink, &ep->rdllist);
    /*
     * Wake up ( if active ) both the eventpoll wait list and the ->poll()
     * wait list.
     */
    /* 唤醒epoll_wait... */
    if (waitqueue_active(&ep->wq))
        wake_up_locked(&ep->wq);
    /* 如果epollfd也在被poll, 那就唤醒队列里面的所有成员. */
    if (waitqueue_active(&ep->poll_wait))
        pwake++;
out_unlock:
    spin_unlock_irqrestore(&ep->lock, flags);
    /* We have to call this outside the lock */
    if (pwake)
        ep_poll_safewake(&ep->poll_wait);
    return 1;
}

epoll_wait内核源码：

/*
 * Implement the event wait interface for the eventpoll file. It is the kernel
 * part of the user space epoll_wait(2).
 */
SYSCALL_DEFINE4(epoll_wait, int, epfd, struct epoll_event __user *, events,
        int, maxevents, int, timeout)
{
    int error;
    struct file *file;
    struct eventpoll *ep;
    /* The maximum number of event must be greater than zero */
    if (maxevents <= 0 || maxevents > EP_MAX_EVENTS)
        return -EINVAL;
    /* Verify that the area passed by the user is writeable */
    /* 这个地方有必要说明一下:
     * 内核对应用程序采取的策略是"绝对不信任",
     * 所以内核跟应用程序之间的数据交互大都是copy, 不允许(也时候也是不能...)指针引用.
     * epoll_wait()需要内核返回数据给用户空间, 内存由用户程序提供,
     * 所以内核会用一些手段来验证这一段内存空间是不是有效的.
     */
    if (!access_ok(VERIFY_WRITE, events, maxevents * sizeof(struct epoll_event))) {
        error = -EFAULT;
        goto error_return;
    }
    /* Get the "struct file *" for the eventpoll file */
    error = -EBADF;
    /* 获取epollfd的struct file, epollfd也是文件嘛 */
    file = fget(epfd);
    if (!file)
        goto error_return;
    /*
     * We have to check that the file structure underneath the fd
     * the user passed to us _is_ an eventpoll file.
     */
    error = -EINVAL;
    /* 检查一下它是不是一个真正的epollfd... */
    if (!is_file_epoll(file))
        goto error_fput;
    /*
     * At this point it is safe to assume that the "private_data" contains
     * our own data structure.
     */
    /* 获取eventpoll结构 */
    ep = file->private_data;
    /* Time to fish for events ... */
    /* OK, 睡觉, 等待事件到来~~ */
    error = ep_poll(ep, events, maxevents, timeout);
error_fput:
    fput(file);
error_return:
    return error;
}
/* 这个函数真正将执行epoll_wait的进程带入睡眠状态... */
static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events,
           int maxevents, long timeout)
{
    int res, eavail;
    unsigned long flags;
    long jtimeout;
    wait_queue_t wait;//等待队列
    /*
     * Calculate the timeout by checking for the "infinite" value (-1)
     * and the overflow condition. The passed timeout is in milliseconds,
     * that why (t * HZ) / 1000.
     */
    /* 计算睡觉时间, 毫秒要转换为HZ */
    jtimeout = (timeout < 0 || timeout >= EP_MAX_MSTIMEO) ?
        MAX_SCHEDULE_TIMEOUT : (timeout * HZ + 999) / 1000;
retry:
    spin_lock_irqsave(&ep->lock, flags);
    res = 0;
    /* 如果ready list不为空, 就不睡了, 直接干活... */
    if (list_empty(&ep->rdllist)) {
        /*
         * We don't have any available event to return to the caller.
         * We need to sleep here, and we will be wake up by
         * ep_poll_callback() when events will become available.
         */
        /* OK, 初始化一个等待队列, 准备直接把自己挂起,
         * 注意current是一个宏, 代表当前进程 */
        init_waitqueue_entry(&wait, current);//初始化等待队列,wait表示当前进程
        __add_wait_queue_exclusive(&ep->wq, &wait);//挂载到ep结构的等待队列
        for (;;) {
            /*
             * We don't want to sleep if the ep_poll_callback() sends us
             * a wakeup in between. That's why we set the task state
             * to TASK_INTERRUPTIBLE before doing the checks.
             */
            /* 将当前进程设置位睡眠, 但是可以被信号唤醒的状态,
             * 注意这个设置是"将来时", 我们此刻还没睡! */
            set_current_state(TASK_INTERRUPTIBLE);
            /* 如果这个时候, ready list里面有成员了,
             * 或者睡眠时间已经过了, 就直接不睡了... */
            if (!list_empty(&ep->rdllist) || !jtimeout)
                break;
            /* 如果有信号产生, 也起床... */
            if (signal_pending(current)) {
                res = -EINTR;
                break;
            }
            /* 啥事都没有,解锁, 睡觉... */
            spin_unlock_irqrestore(&ep->lock, flags);
            /* jtimeout这个时间后, 会被唤醒,
             * ep_poll_callback()如果此时被调用,
             * 那么我们就会直接被唤醒, 不用等时间了...
             * 再次强调一下ep_poll_callback()的调用时机是由被监听的fd
             * 的具体实现, 比如socket或者某个设备驱动来决定的,
             * 因为等待队列头是他们持有的, epoll和当前进程
             * 只是单纯的等待...
             **/
            jtimeout = schedule_timeout(jtimeout);//睡觉
            spin_lock_irqsave(&ep->lock, flags);
        }
        __remove_wait_queue(&ep->wq, &wait);
        /* OK 我们醒来了... */
        set_current_state(TASK_RUNNING);
    }
    /* Is it worth to try to dig for events ? */
    eavail = !list_empty(&ep->rdllist) || ep->ovflist != EP_UNACTIVE_PTR;
    spin_unlock_irqrestore(&ep->lock, flags);
    /*
     * Try to transfer events to user space. In case we get 0 events and
     * there's still timeout left over, we go trying again in search of
     * more luck.
     */
    /* 如果一切正常, 有event发生, 就开始准备数据copy给用户空间了... */
    if (!res && eavail &&
        !(res = ep_send_events(ep, events, maxevents)) && jtimeout)
        goto retry;
    return res;
}
/* 这个简单, 我们直奔下一个... */
static int ep_send_events(struct eventpoll *ep,
              struct epoll_event __user *events, int maxevents)
{
    struct ep_send_events_data esed;
    esed.maxevents = maxevents;
    esed.events = events;
    return ep_scan_ready_list(ep, ep_send_events_proc, &esed);
}
/**
 * ep_scan_ready_list - Scans the ready list in a way that makes possible for
 *                      the scan code, to call f_op->poll(). Also allows for
 *                      O(NumReady) performance.
 *
 * @ep: Pointer to the epoll private data structure.
 * @sproc: Pointer to the scan callback.
 * @priv: Private opaque data passed to the @sproc callback.
 *
 * Returns: The same integer error code returned by the @sproc callback.
 */
static int ep_scan_ready_list(struct eventpoll *ep,
                  int (*sproc)(struct eventpoll *,
                       struct list_head *, void *),
                  void *priv)
{
    int error, pwake = 0;
    unsigned long flags;
    struct epitem *epi, *nepi;
    LIST_HEAD(txlist);
    /*
     * We need to lock this because we could be hit by
     * eventpoll_release_file() and epoll_ctl().
     */
    mutex_lock(&ep->mtx);
    /*
     * Steal the ready list, and re-init the original one to the
     * empty list. Also, set ep->ovflist to NULL so that events
     * happening while looping w/out locks, are not lost. We cannot
     * have the poll callback to queue directly on ep->rdllist,
     * because we want the "sproc" callback to be able to do it
     * in a lockless way.
     */
    spin_lock_irqsave(&ep->lock, flags);
    /* 这一步要注意, 首先, 所有监听到events的epitem都链到rdllist上了,
     * 但是这一步之后, 所有的epitem都转移到了txlist上, 而rdllist被清空了,
     * 要注意哦, rdllist已经被清空了! */
    list_splice_init(&ep->rdllist, &txlist);
    /* ovflist, 在ep_poll_callback()里面我解释过, 此时此刻我们不希望
     * 有新的event加入到ready list中了, 保存后下次再处理... */
    ep->ovflist = NULL;
    spin_unlock_irqrestore(&ep->lock, flags);
    /*
     * Now call the callback function.
     */
    /* 在这个回调函数里面处理每个epitem
     * sproc 就是 ep_send_events_proc, 下面会注释到. */
    error = (*sproc)(ep, &txlist, priv);
    spin_lock_irqsave(&ep->lock, flags);
    /*
     * During the time we spent inside the "sproc" callback, some
     * other events might have been queued by the poll callback.
     * We re-insert them inside the main ready-list here.
     */
    /* 现在我们来处理ovflist, 这些epitem都是我们在传递数据给用户空间时
     * 监听到了事件. */
    for (nepi = ep->ovflist; (epi = nepi) != NULL;
         nepi = epi->next, epi->next = EP_UNACTIVE_PTR) {
        /*
         * We need to check if the item is already in the list.
         * During the "sproc" callback execution time, items are
         * queued into ->ovflist but the "txlist" might already
         * contain them, and the list_splice() below takes care of them.
         */
        /* 将这些直接放入readylist */
        if (!ep_is_linked(&epi->rdllink))
            list_add_tail(&epi->rdllink, &ep->rdllist);
    }
    /*
     * We need to set back ep->ovflist to EP_UNACTIVE_PTR, so that after
     * releasing the lock, events will be queued in the normal way inside
     * ep->rdllist.
     */
    ep->ovflist = EP_UNACTIVE_PTR;
    /*
     * Quickly re-inject items left on "txlist".
     */
    /* 上一次没有处理完的epitem, 重新插入到ready list */
    list_splice(&txlist, &ep->rdllist);
    /* ready list不为空, 直接唤醒... */
    if (!list_empty(&ep->rdllist)) {
        /*
         * Wake up (if active) both the eventpoll wait list and
         * the ->poll() wait list (delayed after we release the lock).
         */
        if (waitqueue_active(&ep->wq))
            wake_up_locked(&ep->wq);
        if (waitqueue_active(&ep->poll_wait))
            pwake++;
    }
    spin_unlock_irqrestore(&ep->lock, flags);
    mutex_unlock(&ep->mtx);
    /* We have to call this outside the lock */
    if (pwake)
        ep_poll_safewake(&ep->poll_wait);
    return error;
}
/* 该函数作为callbakc在ep_scan_ready_list()中被调用
 * head是一个链表, 包含了已经ready的epitem,
 * 这个不是eventpoll里面的ready list, 而是上面函数中的txlist.
 */
static int ep_send_events_proc(struct eventpoll *ep, struct list_head *head,
                   void *priv)
{
    struct ep_send_events_data *esed = priv;
    int eventcnt;
    unsigned int revents;
    struct epitem *epi;
    struct epoll_event __user *uevent;
    /*
     * We can loop without lock because we are passed a task private list.
     * Items cannot vanish during the loop because ep_scan_ready_list() is
     * holding "mtx" during this call.
     */
    /* 扫描整个链表... */
    for (eventcnt = 0, uevent = esed->events;
         !list_empty(head) && eventcnt < esed->maxevents;) {
        /* 取出第一个成员 */
        epi = list_first_entry(head, struct epitem, rdllink);
        /* 然后从链表里面移除 */
        list_del_init(&epi->rdllink);
        /* 读取events,
         * 注意events我们ep_poll_callback()里面已经取过一次了, 为啥还要再取?
         * 1. 我们当然希望能拿到此刻的最新数据, events是会变的~
         * 2. 不是所有的poll实现, 都通过等待队列传递了events, 有可能某些驱动压根没传
         * 必须主动去读取. */
        revents = epi->ffd.file->f_op->poll(epi->ffd.file, NULL) &
            epi->event.events;
        if (revents) {
            /* 将当前的事件和用户传入的数据都copy给用户空间,
             * 就是epoll_wait()后应用程序能读到的那一堆数据. */
            if (__put_user(revents, &uevent->events) ||
                __put_user(epi->event.data, &uevent->data)) {
                list_add(&epi->rdllink, head);
                return eventcnt ? eventcnt : -EFAULT;
            }
            eventcnt++;
            uevent++;
            if (epi->event.events & EPOLLONESHOT)
                epi->event.events &= EP_PRIVATE_BITS;
            else if (!(epi->event.events & EPOLLET)) {
                /*
                 * 如果是ET, epitem是不会再进入到readly list,
                 * 除非fd再次发生了状态改变, ep_poll_callback被调用.
                 * 如果是非ET, 不管你还有没有有效的事件或者数据,
                 * 都会被重新插入到ready list, 再下一次epoll_wait
                 * 时, 会立即返回, 并通知给用户空间. 当然如果这个
                 * 被监听的fds确实没事件也没数据了, epoll_wait会返回一个0,
                 * 空转一次.
                 */
                list_add_tail(&epi->rdllink, &ep->rdllist);
            }
        }
    }
    return eventcnt;
}
/* ep_free在epollfd被close时调用,
 * 释放一些资源而已, 比较简单 */
static void ep_free(struct eventpoll *ep)
{
    struct rb_node *rbp;
    struct epitem *epi;
    /* We need to release all tasks waiting for these file */
    if (waitqueue_active(&ep->poll_wait))
        ep_poll_safewake(&ep->poll_wait);
    /*
     * We need to lock this because we could be hit by
     * eventpoll_release_file() while we're freeing the "struct eventpoll".
     * We do not need to hold "ep->mtx" here because the epoll file
     * is on the way to be removed and no one has references to it
     * anymore. The only hit might come from eventpoll_release_file() but
     * holding "epmutex" is sufficent here.
     */
    mutex_lock(&epmutex);
    /*
     * Walks through the whole tree by unregistering poll callbacks.
     */
    for (rbp = rb_first(&ep->rbr); rbp; rbp = rb_next(rbp)) {
        epi = rb_entry(rbp, struct epitem, rbn);
        ep_unregister_pollwait(ep, epi);
    }
    /*
     * Walks through the whole tree by freeing each "struct epitem". At this
     * point we are sure no poll callbacks will be lingering around, and also by
     * holding "epmutex" we can be sure that no file cleanup code will hit
     * us during this operation. So we can avoid the lock on "ep->lock".
     */
    /* 之所以在关闭epollfd之前不需要调用epoll_ctl移除已经添加的fd,
     * 是因为这里已经做了... */
    while ((rbp = rb_first(&ep->rbr)) != NULL) {
        epi = rb_entry(rbp, struct epitem, rbn);
        ep_remove(ep, epi);
    }
    mutex_unlock(&epmutex);
    mutex_destroy(&ep->mtx);
    free_uid(ep->user);
    kfree(ep);
}
/* File callbacks that implement the eventpoll file behaviour */
static const struct file_operations eventpoll_fops = {
    .release    = ep_eventpoll_release,
    .poll       = ep_eventpoll_poll
};
/* Fast test to see if the file is an evenpoll file */
static inline int is_file_epoll(struct file *f)
{
    return f->f_op == &eventpoll_fops;
}

总结如下：

epoll_create
从slab缓存中创建一个eventpoll对象,并且创建一个匿名的fd跟fd对应的file对象,而eventpoll对象保存在struct file结构的private指针中,并且返回,该fd对应的file operations只是实现了poll跟release操作，创建eventpoll对象的初始化操作，获取当前用户信息,是不是root,最大监听fd数目等并且保存到eventpoll对象中，初始化等待队列,初始化就绪链表,初始化红黑树的头结点；

epoll_ctl
将epoll_event结构拷贝到内核空间中，并且判断加入的fd是否支持poll结构(epoll,poll,selectI/O多路复用必须支持poll操作)。并且从epfd->file->privatedata获取event_poll对象,根据op区分是添加删除还是修改,首先在eventpoll结构中的红黑树查找是否已经存在了相对应的fd,没找到就支持插入操作,否则报重复的错误。插入操作时,会创建一个与fd对应的epitem结构,并且初始化相关成员,比如保存监听的fd跟file结构之类的。重要的是指定了调用poll_wait时的回调函数用于数据就绪时唤醒进程,(其内部,初始化设备的等待队列,将该进程注册到等待队列)完成这一步, 我们的epitem就跟这个socket关联起来了, 当它有状态变化时,会通过ep_poll_callback()来通知。最后调用加入的fd的file operation->poll函数(最后会调用poll_wait操作)用于完成注册操作。最后将epitem结构添加到红黑树中；

epoll_wait
计算睡眠时间(如果有),判断eventpoll对象的链表是否为空,不为空那就干活不睡明.并且初始化一个等待队列,把自己挂上去,设置自己的进程状态，为可睡眠状态.判断是否有信号到来(有的话直接被中断醒来,),如果啥事都没有那就调用schedule_timeout进行睡眠,如果超时或者被唤醒,首先从自己初始化的等待队列删除,然后开始拷贝资源给用户空间了拷贝资源则是先把就绪事件链表转移到中间链表,然后挨个遍历拷贝到用户空间,并且挨个判断其是否为水平触发,是的话再次插入到就绪链表