当我们在组件中对响应的数据做了修改，就会触发 setter 的逻辑，最后调用 dep.notify() 方法代码如下：

class Dep {
  // ...
  notify () {
  // stabilize the subscriber list first
    const subs = this.subs.slice()
    for (let i = 0, l = subs.length; i < l; i++) {
      subs[i].update()
    }
  }
}

这里的逻辑非常简单，遍历所有的 subs，也就是 Watcher 的实例数组，然后调用每一个 watcher 的 update 方法，它的定义在 src/core/observer/watcher.js中：

class Watcher{
//....
update () {
    /* istanbul ignore else */
    if (this.computed) {
      // A computed property watcher has two modes: lazy and activated.
      // It initializes as lazy by default, and only becomes activated when
      // it is depended on by at least one subscriber, which is typically
      // another computed property or a component's render function.
      if (this.dep.subs.length === 0) {
        // In lazy mode, we don't want to perform computations until necessary,
        // so we simply mark the watcher as dirty. The actual computation is
        // performed just-in-time in this.evaluate() when the computed property
        // is accessed.
        this.dirty = true
      } else {
        // In activated mode, we want to proactively perform the computation
        // but only notify our subscribers when the value has indeed changed.
        this.getAndInvoke(() => {
          this.dep.notify()
        })
      }
    } else if (this.sync) {
      this.run()
    } else {
      queueWatcher(this)
    }
  }
}

在一般组件数据更新的场景，会走到最后一个 queueWatcher(this) 的逻辑，queueWatcher 的定义在 src/core/observer/scheduler.js 中：

const queue: Array<Watcher> = []
let has: { [key: number]: ?true } = {}
let waiting = false
let flushing = false
/**
 * Push a watcher into the watcher queue.
 * Jobs with duplicate IDs will be skipped unless it's
 * pushed when the queue is being flushed.
 */
export function queueWatcher (watcher: Watcher) {
  const id = watcher.id
  if (has[id] == null) {
    has[id] = true
    if (!flushing) {
      queue.push(watcher)
    } else {
      // if already flushing, splice the watcher based on its id
      // if already past its id, it will be run next immediately.
      let i = queue.length - 1
      while (i > index && queue[i].id > watcher.id) {
        i--
      }
      queue.splice(i + 1, 0, watcher)
    }
    // queue the flush
    if (!waiting) {
      waiting = true
      nextTick(flushSchedulerQueue)
    }
  }
}

这里引入了一个队列的概念，这也是 Vue 在做派发更新的时候的一个优化的点，它并不会每次数据改变都触发 watcher 的回调，而是把这些 watcher 先添加到一个队列里，然后在 nextTick 后执行 flushSchedulerQueue。
这里有几个细节要注意一下，首先用 has 对象保证同一个 Watcher 只添加一次；接着对 flushing 的判断，else 部分的逻辑稍后我会讲；最后通过 wating 保证对 nextTick(flushSchedulerQueue) 的调用逻辑只有一次，另外 nextTick 的实现我之后会抽一小节专门去讲，目前就可以理解它是在下一个 tick，也就是异步的去执行 flushSchedulerQueue。
接下来我们来看 flushSchedulerQueue 的实现，它的定义在 src/core/observer/scheduler.js 中。

let flushing = false
let index = 0
/**
 * Flush both queues and run the watchers.
 */
function flushSchedulerQueue () {
  flushing = true
  let watcher, id

  // Sort queue before flush.
  // This ensures that:
  // 1. Components are updated from parent to child. (because parent is always
  //    created before the child)
  // 2. A component's user watchers are run before its render watcher (because
  //    user watchers are created before the render watcher)
  // 3. If a component is destroyed during a parent component's watcher run,
  //    its watchers can be skipped.
  queue.sort((a, b) => a.id - b.id)

  // do not cache length because more watchers might be pushed
  // as we run existing watchers
  for (index = 0; index < queue.length; index++) {
    watcher = queue[index]
    if (watcher.before) {
      watcher.before()
    }
    id = watcher.id
    has[id] = null
    watcher.run()
    // in dev build, check and stop circular updates.
    if (process.env.NODE_ENV !== 'production' && has[id] != null) {
      circular[id] = (circular[id] || 0) + 1
      if (circular[id] > MAX_UPDATE_COUNT) {
        warn(
          'You may have an infinite update loop ' + (
            watcher.user
              ? `in watcher with expression "${watcher.expression}"`
              : `in a component render function.`
          ),
          watcher.vm
        )
        break
      }
    }
  }

  // keep copies of post queues before resetting state
  const activatedQueue = activatedChildren.slice()
  const updatedQueue = queue.slice()

  resetSchedulerState()

  // call component updated and activated hooks
  callActivatedHooks(activatedQueue)
  callUpdatedHooks(updatedQueue)

  // devtool hook
  /* istanbul ignore if */
  if (devtools && config.devtools) {
    devtools.emit('flush')
  }
}

这里有几个重要的逻辑要梳理一下，
队列排序
queue.sort((a, b) => a.id - b.id) 对队列做了从小到大的排序，这么做主要有以下要确保以下几点：
1.组件的更新由父到子；因为父组件的创建过程是先于子的，所以 watcher 的创建也是先父后子，执行顺序也应该保持先父后子。
2.用户的自定义 watcher 要优先于渲染 watcher 执行；因为用户自定义 watcher 是在渲染 watcher 之前创建的。
3.如果一个组件在父组件的 watcher 执行期间被销毁，那么它对应的 watcher 执行都可以被跳过，所以父组件的 watcher 应该先执行。
队列遍历
在对 queue 排序后，接着就是要对它做遍历，拿到对应的 watcher，执行 watcher.run()。这里需要注意一个细节，在遍历的时候每次都会对 queue.length 求值，因为在 watcher.run() 的时候，很可能用户会再次添加新的 watcher，这样会再次执行到 queueWatcher，如下：

export function queueWatcher (watcher: Watcher) {
  const id = watcher.id
  if (has[id] == null) {
    has[id] = true
    if (!flushing) {
      queue.push(watcher)
    } else {
      // if already flushing, splice the watcher based on its id
      // if already past its id, it will be run next immediately.
      let i = queue.length - 1
      while (i > index && queue[i].id > watcher.id) {
        i--
      }
      queue.splice(i + 1, 0, watcher)
    }
    // ...
  }
}

可以看到，这时候 flushing 为 true，就会执行到 else 的逻辑，然后就会从后往前找，找到第一个待插入 watcher 的 id 比当前队列中 watcher 的 id 大的位置。把 watcher 按照 id的插入到队列中，因此 queue 的长度发送了变化。
状态恢复
这个过程就是执行 resetSchedulerState 函数，它的定义在 src/core/observer/scheduler.js 中。

const queue: Array<Watcher> = []
let has: { [key: number]: ?true } = {}
let circular: { [key: number]: number } = {}
let waiting = false
let flushing = false
let index = 0
/**
 * Reset the scheduler's state.
 */
function resetSchedulerState () {
  index = queue.length = activatedChildren.length = 0
  has = {}
  if (process.env.NODE_ENV !== 'production') {
    circular = {}
  }
  waiting = flushing = false
}

逻辑非常简单，就是把这些控制流程状态的一些变量恢复到初始值，把 watcher 队列清空。
接下来我们继续分析 watcher.run() 的逻辑，它的定义在 src/core/observer/watcher.js 中。

class Watcher {
  /**
   * Scheduler job interface.
   * Will be called by the scheduler.
   */
  run () {
    if (this.active) {
      this.getAndInvoke(this.cb)
    }
  }

  getAndInvoke (cb: Function) {
    const value = this.get()
    if (
      value !== this.value ||
      // Deep watchers and watchers on Object/Arrays should fire even
      // when the value is the same, because the value may
      // have mutated.
      isObject(value) ||
      this.deep
    ) {
      // set new value
      const oldValue = this.value
      this.value = value
      this.dirty = false
      if (this.user) {
        try {
          cb.call(this.vm, value, oldValue)
        } catch (e) {
          handleError(e, this.vm, `callback for watcher "${this.expression}"`)
        }
      } else {
        cb.call(this.vm, value, oldValue)
      }
    }
  }
}

run 函数实际上就是执行 this.getAndInvoke 方法，并传入 watcher 的回调函数。getAndInvoke 函数逻辑也很简单，先通过 this.get() 得到它当前的值，然后做判断，如果满足新旧值不等、新值是对象类型、deep 模式任何一个条件，则执行 watcher 的回调，注意回调函数执行的时候会把第一个和第二个参数传入新值 value 和旧值 oldValue，这就是当我们添加自定义 watcher 的时候能在回调函数的参数中拿到新旧值的原因。
那么对于渲染 watcher 而言，它在执行 this.get() 方法求值的时候，会执行 getter 方法：

updateComponent = () => {
  vm._update(vm._render(), hydrating)
}

所以这就是当我们去修改组件相关的响应式数据的时候，会触发组件重新渲染的原因，接着就会重新执行 patch 的过程.