0. 简介
传统的并发编程模型是基于线程和共享内存的同步访问控制的，共享数据受锁的保护，线程将争夺这些锁以访问数据。通常而言，使用线程安全的数据结构会使得这更加容易。Go的并发原语（goroutine和channel）提供了一种优雅的方式来构建并发模型。Go鼓励在goroutine之间使用channel来传递数据，而不是显式地使用锁来限制对共享数据的访问。
Do not communicate by sharing memory; instead, share memory by communicating.
这就是Go的并发哲学，它依赖CSP(Communicating Sequential Processes)模型，它经常被认为是Go在并发编程上成功的关键因素。
如果说goroutine是Go语言程序的并发体的话，那么channel就是他们之间的通信机制，前面的系列博客对goroutine及其调度机制进行了介绍，本文将介绍一下二者之间的通信机制——channel。
1. channel数据结构
type hchan struct {
qcount uint // total data in the queue
dataqsiz uint // size of the circular queue
buf unsafe.Pointer // points to an array of dataqsiz elements
elemsize uint16
closed uint32
elemtype *_type // element type
sendx uint // send index
recvx uint // receive index
recvq waitq // list of recv waiters
sendq waitq // list of send waiters
// lock protects all fields in hchan, as well as several
// fields in sudogs blocked on this channel.
//
// Do not change another G's status while holding this lock
// (in particular, do not ready a G), as this can deadlock
// with stack shrinking.
lock mutex
}在runtime/chan.go中，channel被定义如上，其中：
-
buf：是有缓存的channel持有的，用来存储缓存数据，收个循环链表； -
dataqsiz：上述缓存数据的循环链表的最大容量，理解为cap()； -
qcount：上述缓存数据的循环链表的长度，理解为len()； -
recvx和sendx：表示上述缓存的接收或者发送位置； -
recvq和sendq：分别是接收和发送的goroutine抽象（sudog）队列，是个双向链表； -
lock：互斥锁，用来保证channel数据的线程安全。
2. channel创建
func makechan64(t *chantype, size int64) *hchan {
if int64(int(size)) != size {
panic(plainError("makechan: size out of range"))
}
return makechan(t, int(size))
}
func makechan(t *chantype, size int) *hchan {
elem := t.elem
// compiler checks this but be safe.
if elem.size >= 1<<16 {
throw("makechan: invalid channel element type")
}
if hchanSize%maxAlign != 0 || elem.align > maxAlign {
throw("makechan: bad alignment")
}
mem, overflow := math.MulUintptr(elem.size, uintptr(size))
if overflow || mem > maxAlloc-hchanSize || size < 0 {
panic(plainError("makechan: size out of range"))
}
// Hchan does not contain pointers interesting for GC when elements stored in buf do not contain pointers.
// buf points into the same allocation, elemtype is persistent.
// SudoG's are referenced from their owning thread so they can't be collected.
// TODO(dvyukov,rlh): Rethink when collector can move allocated objects.
var c *hchan
switch {
case mem == 0:
// Queue or element size is zero.
c = (*hchan)(mallocgc(hchanSize, nil, true))
// Race detector uses this location for synchronization.
c.buf = c.raceaddr()
case elem.ptrdata == 0:
// Elements do not contain pointers.
// Allocate hchan and buf in one call.
c = (*hchan)(mallocgc(hchanSize+mem, nil, true))
c.buf = add(unsafe.Pointer(c), hchanSize)
default:
// Elements contain pointers.
c = new(hchan)
c.buf = mallocgc(mem, elem, true)
}
c.elemsize = uint16(elem.size)
c.elemtype = elem
c.dataqsiz = uint(size)
lockInit(&c.lock, lockRankHchan)
if debugChan {
print("makechan: chan=", c, "; elemsize=", elem.size, "; dataqsiz=", size, "\n")
}
return c
}所有的调用最后都会走到runtime.makechan函数，函数做的事情比较简单，就是初始化一个runtime.hchan的对象，和map一样，channel对外就是一个指针对象（切片和字符串则不是指针对象，以切片为例，可以参考链接）。可以看到：
- 如果当前
channel没有缓存，那么就只会runtime.hchan分配一段空间； - 如果当前
channel中存储的类型不是指针类型，那么会为当前的runtime.hchan和底层的连续数组分配一块连续的内存空间； - 其他情况下，那么则为
runtime.hchan和其缓存各自分配一段内存；
3. 数据发送
// entry point for c <- x from compiled code
//go:nosplit
func chansend1(c *hchan, elem unsafe.Pointer) {
chansend(c, elem, true, getcallerpc())
}channel的数据发送会调用runtime.chansend1函数，而该函数则只是调用了runtime.chansend函数，该函数比较长，我们一点一点分析：
3.1 空通道的数据发送
func chansend(c *hchan, ep unsafe.Pointer, block bool, callerpc uintptr) bool {
if c == nil {
if !block {
return false
}
gopark(nil, nil, waitReasonChanSendNilChan, traceEvGoStop, 2)
throw("unreachable")
}
...
}可以看到，如果通道是nil，那么往这个通道中写数据时：
- 非阻塞写会直接返回（在单
channel发送+default分支的select操作时会调用runtime.selectnbsend函数，从而会非阻塞写）； - 阻塞写（正常的
ch <- v）时则会通过gopark函数让出CPU调度权，阻塞此goroutine；
3.2 直接发送
func chansend(c *hchan, ep unsafe.Pointer, block bool, callerpc uintptr) bool {
...
if c.closed != 0 {
unlock(&c.lock)
panic(plainError("send on closed channel"))
}
if sg := c.recvq.dequeue(); sg != nil {
// Found a waiting receiver. We pass the value we want to send
// directly to the receiver, bypassing the channel buffer (if any).
send(c, sg, ep, func() { unlock(&c.lock) }, 3)
return true
}
...
}可以发现，当channel被关闭后再发送数据，那么会导致panic。
如果目标channel没有关闭，且有已经处于读等待的goroutine，那么会直接从recvq中取出最先陷入等待的goroutine，并通过runtime.send函数向其发送数据：
func send(c *hchan, sg *sudog, ep unsafe.Pointer, unlockf func(), skip int) {
if raceenabled {
if c.dataqsiz == 0 {
racesync(c, sg)
} else {
// Pretend we go through the buffer, even though
// we copy directly. Note that we need to increment
// the head/tail locations only when raceenabled.
racenotify(c, c.recvx, nil)
racenotify(c, c.recvx, sg)
c.recvx++
if c.recvx == c.dataqsiz {
c.recvx = 0
}
c.sendx = c.recvx // c.sendx = (c.sendx+1) % c.dataqsiz
}
}
if sg.elem != nil {
sendDirect(c.elemtype, sg, ep)
sg.elem = nil
}
gp := sg.g
unlockf()
gp.param = unsafe.Pointer(sg)
sg.success = true
if sg.releasetime != 0 {
sg.releasetime = cputicks()
}
goready(gp, skip+1)
}可以看到，以上函数做了两件事：
- 调用
sendDirect函数将发送的数据拷贝到接收协程的变量所在的地址上； - 通过
goready函数唤醒协程，将其状态置为_Grunnable后放置到处理器的队列的下一个待处理goroutine；
3.3 缓存区
如果没有已经处于读等待的goroutine，且创建的channel包含缓存，并且缓存还没有满，那么会执行以下代码：
func chansend(c *hchan, ep unsafe.Pointer, block bool, callerpc uintptr) bool {
...
if c.qcount < c.dataqsiz {
// Space is available in the channel buffer. Enqueue the element to send.
qp := chanbuf(c, c.sendx)
if raceenabled {
racenotify(c, c.sendx, nil)
}
typedmemmove(c.elemtype, qp, ep)
c.sendx++
if c.sendx == c.dataqsiz {
c.sendx = 0
}
c.qcount++
unlock(&c.lock)
return true
}
...
}在这里会首先通过runtime.chanbuf函数计算出下一个可以存储的位置，然后通过runtime.typedmemmove将发送的数据拷贝到缓冲区中并增加sendx索引和qcount计数器。等待有接收数据的goroutine时可以直接从缓存中读取。
3.4 阻塞发送
如果既没有等待读的goroutine，又没有缓存区或着缓存区满了，那么就会阻塞发送数据：
func chansend(c *hchan, ep unsafe.Pointer, block bool, callerpc uintptr) bool {
...
if !block {
unlock(&c.lock)
return false
}
// Block on the channel. Some receiver will complete our operation for us.
gp := getg()
mysg := acquireSudog()
mysg.releasetime = 0
if t0 != 0 {
mysg.releasetime = -1
}
// No stack splits between assigning elem and enqueuing mysg
// on gp.waiting where copystack can find it.
mysg.elem = ep
mysg.waitlink = nil
mysg.g = gp
mysg.isSelect = false
mysg.c = c
gp.waiting = mysg
gp.param = nil
c.sendq.enqueue(mysg)
// Signal to anyone trying to shrink our stack that we're about
// to park on a channel. The window between when this G's status
// changes and when we set gp.activeStackChans is not safe for
// stack shrinking.
atomic.Store8(&gp.parkingOnChan, 1)
gopark(chanparkcommit, unsafe.Pointer(&c.lock), waitReasonChanSend, traceEvGoBlockSend, 2)
// Ensure the value being sent is kept alive until the
// receiver copies it out. The sudog has a pointer to the
// stack object, but sudogs aren't considered as roots of the
// stack tracer.
KeepAlive(ep)
// someone woke us up.
if mysg != gp.waiting {
throw("G waiting list is corrupted")
}
gp.waiting = nil
gp.activeStackChans = false
closed := !mysg.success
gp.param = nil
if mysg.releasetime > 0 {
blockevent(mysg.releasetime-t0, 2)
}
mysg.c = nil
releaseSudog(mysg)
if closed {
if c.closed == 0 {
throw("chansend: spurious wakeup")
}
panic(plainError("send on closed channel"))
}
return true
}- 调用
runtime.getg获取此时发送数据的goroutine； - 调用
runtime.acquireSudog获取sudog结构并设置相关信息； - 将上一步获取的
sudog放到发送等待队列，并且调用gopark挂起当前协程； - 等待有接收数据的
goroutine到来后，即唤醒此goroutine，然后继续往下走；或者close了此channel，导致后续的panic。
4. 接收数据
4.1 空通道的数据接收
func chanrecv(c *hchan, ep unsafe.Pointer, block bool) (selected, received bool) {
// raceenabled: don't need to check ep, as it is always on the stack
// or is new memory allocated by reflect.
if debugChan {
print("chanrecv: chan=", c, "\n")
}
if c == nil {
if !block {
return
}
gopark(nil, nil, waitReasonChanReceiveNilChan, traceEvGoStop, 2)
throw("unreachable")
}
...
lock(&c.lock)
if c.closed != 0 && c.qcount == 0 {
if raceenabled {
raceacquire(c.raceaddr())
}
unlock(&c.lock)
if ep != nil {
typedmemclr(c.elemtype, ep)
}
return true, false
}
...
}以上是通道接收时的一部分代码，可以看到：
- 和发送数据一样，如果通道是
nil，且非阻塞读，则会返回，阻塞读后则会挂起； - 和发送数据时不一样的是，如果是一个已经关闭的通道，其实是可读的，但是读回的数据都是
零值+false。
4.2 直接接收
func chanrecv(c *hchan, ep unsafe.Pointer, block bool) (selected, received bool) {
...
if sg := c.sendq.dequeue(); sg != nil {
// Found a waiting sender. If buffer is size 0, receive value
// directly from sender. Otherwise, receive from head of queue
// and add sender's value to the tail of the queue (both map to
// the same buffer slot because the queue is full).
recv(c, sg, ep, func() { unlock(&c.lock) }, 3)
return true, true
}
...
}当channel的sendq队列中包含处于等待状态的goroutine时，会取出等待的最早的写数据goroutine，然后调用runtime.recv进行发送：
func recv(c *hchan, sg *sudog, ep unsafe.Pointer, unlockf func(), skip int) {
if c.dataqsiz == 0 {
if raceenabled {
racesync(c, sg)
}
if ep != nil {
// copy data from sender
recvDirect(c.elemtype, sg, ep)
}
} else {
// Queue is full. Take the item at the
// head of the queue. Make the sender enqueue
// its item at the tail of the queue. Since the
// queue is full, those are both the same slot.
qp := chanbuf(c, c.recvx)
if raceenabled {
racenotify(c, c.recvx, nil)
racenotify(c, c.recvx, sg)
}
// copy data from queue to receiver
if ep != nil {
typedmemmove(c.elemtype, ep, qp)
}
// copy data from sender to queue
typedmemmove(c.elemtype, qp, sg.elem)
c.recvx++
if c.recvx == c.dataqsiz {
c.recvx = 0
}
c.sendx = c.recvx // c.sendx = (c.sendx+1) % c.dataqsiz
}
sg.elem = nil
gp := sg.g
unlockf()
gp.param = unsafe.Pointer(sg)
sg.success = true
if sg.releasetime != 0 {
sg.releasetime = cputicks()
}
goready(gp, skip+1)
}该函数会根据是否存在缓存区分别处理：
- 如果不存在缓存区，则调用
runtime.recvDirect函数直接将发送goroutine存储的数据拷贝到目标内存地址中，相当于直接从这个goroutine中取数据； - 如果存在缓存区，那么先将缓存区中的数据拷贝到目标内存地址中，然后将
gp的数据拷贝到缓存区最后，相当于先从缓存队列头部取出数据给接收goroutine，在从等待发送goroutine中取出数据到缓存队列尾部，可以看出，此时队列一定是满的。
最后无论哪种情况，都需要调用goready唤醒gp。
4.3 从缓存区拿
其实这里的章节名描述并不准确，在4.2中也存在从缓存区拿数据的情况，差别在于：
- 4.2中缓存队列是满的，且还有发送阻塞等到的
goroutine； - 4.3中不存在发送阻塞等到的
goroutine。
func chanrecv(c *hchan, ep unsafe.Pointer, block bool) (selected, received bool) {
...
if c.qcount > 0 {
// Receive directly from queue
qp := chanbuf(c, c.recvx)
if raceenabled {
racenotify(c, c.recvx, nil)
}
if ep != nil {
typedmemmove(c.elemtype, ep, qp)
}
typedmemclr(c.elemtype, qp)
c.recvx++
if c.recvx == c.dataqsiz {
c.recvx = 0
}
c.qcount--
unlock(&c.lock)
return true, true
}
...
}和发送时一样，如果缓存区有数据，那么从缓存区拷贝数据。
4.4 阻塞接收
func chanrecv(c *hchan, ep unsafe.Pointer, block bool) (selected, received bool) {
...
if !block {
unlock(&c.lock)
return false, false
}
// no sender available: block on this channel.
gp := getg()
mysg := acquireSudog()
mysg.releasetime = 0
if t0 != 0 {
mysg.releasetime = -1
}
// No stack splits between assigning elem and enqueuing mysg
// on gp.waiting where copystack can find it.
mysg.elem = ep
mysg.waitlink = nil
gp.waiting = mysg
mysg.g = gp
mysg.isSelect = false
mysg.c = c
gp.param = nil
c.recvq.enqueue(mysg)
// Signal to anyone trying to shrink our stack that we're about
// to park on a channel. The window between when this G's status
// changes and when we set gp.activeStackChans is not safe for
// stack shrinking.
atomic.Store8(&gp.parkingOnChan, 1)
gopark(chanparkcommit, unsafe.Pointer(&c.lock), waitReasonChanReceive, traceEvGoBlockRecv, 2)
// someone woke us up
if mysg != gp.waiting {
throw("G waiting list is corrupted")
}
gp.waiting = nil
gp.activeStackChans = false
if mysg.releasetime > 0 {
blockevent(mysg.releasetime-t0, 2)
}
success := mysg.success
gp.param = nil
mysg.c = nil
releaseSudog(mysg)
return true, success
}和阻塞发送类似，如果没有等待发送的goroutine，且没有缓存区或者缓存区没有数据，那这个时候就需要将此接收goroutine压到recvq中，并且gopark挂起，等待唤醒。
5. 关闭
func closechan(c *hchan) {
if c == nil {
panic(plainError("close of nil channel"))
}
lock(&c.lock)
if c.closed != 0 {
unlock(&c.lock)
panic(plainError("close of closed channel"))
}
if raceenabled {
callerpc := getcallerpc()
racewritepc(c.raceaddr(), callerpc, abi.FuncPCABIInternal(closechan))
racerelease(c.raceaddr())
}
c.closed = 1
var glist gList
// release all readers
for {
sg := c.recvq.dequeue()
if sg == nil {
break
}
if sg.elem != nil {
typedmemclr(c.elemtype, sg.elem)
sg.elem = nil
}
if sg.releasetime != 0 {
sg.releasetime = cputicks()
}
gp := sg.g
gp.param = unsafe.Pointer(sg)
sg.success = false
if raceenabled {
raceacquireg(gp, c.raceaddr())
}
glist.push(gp)
}
// release all writers (they will panic)
for {
sg := c.sendq.dequeue()
if sg == nil {
break
}
sg.elem = nil
if sg.releasetime != 0 {
sg.releasetime = cputicks()
}
gp := sg.g
gp.param = unsafe.Pointer(sg)
sg.success = false
if raceenabled {
raceacquireg(gp, c.raceaddr())
}
glist.push(gp)
}
unlock(&c.lock)
// Ready all Gs now that we've dropped the channel lock.
for !glist.empty() {
gp := glist.pop()
gp.schedlink = 0
goready(gp, 3)
}
}关闭通道的代码看上去很长，实际上在处理完一些特殊情况后，就是对发送和接收队列的数据通通使用goready唤醒。
6. 总结
在Go中，虽然极力推崇CSP哲学，推荐大家使用channel实现共享内存的保护，但是：
在幕后，通道使用锁来序列化访问并提供线程安全性。 因此，通过使用通道同步对内存的访问，你实际上就是在使用锁。 被包装在线程安全队列中的锁。 那么，与仅仅使用标准库
sync包中的互斥量相比，Go 的花式锁又如何呢？ 以下数字是通过使用 Go 的内置基准测试功能，对它们的单个集合连续调用 Put 得出的。
`> BenchmarkSimpleSet-8 3000000 391 ns/op`
`> BenchmarkSimpleChannelSet-8 1000000 1699 ns/o`
就我个人的理解而言：
- 在进行数据的传输时使用
channel； - 在进行内存数据的保护时使用
sync.Mutex； - 利用
channel和select的特性，实现类似于Linux epoll的功能。