并发原语Sync-Pool源码剖析

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

var pool *sync.Pool
type Person struct {
	Name string
}

func init() {
	pool = &sync.Pool{
		New: func() interface{}{
			fmt.Println("creating a new person")
			return new(Person)
		},
	}
}

func main() {
	person := pool.Get().(*Person)
	fmt.Println("Get Pool Object：", person)

	person.Name = "first"
	pool.Put(person)

	fmt.Println("Get Pool Object：",pool.Get().(*Person))
	fmt.Println("Get Pool Object：",pool.Get().(*Person))

}

1
2
3
4
5

creating a new person
Get Pool Object： &{}
Get Pool Object： &{first}
creating a new person
Get Pool Object： &{}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

// A Pool is a set of temporary objects that may be individually saved and
// retrieved.
//
// Any item stored in the Pool may be removed automatically at any time without
// notification. If the Pool holds the only reference when this happens, the
// item might be deallocated.
//
// A Pool is safe for use by multiple goroutines simultaneously.
//
// Pool's purpose is to cache allocated but unused items for later reuse,
// relieving pressure on the garbage collector. That is, it makes it easy to
// build efficient, thread-safe free lists. However, it is not suitable for all
// free lists.
//
// An appropriate use of a Pool is to manage a group of temporary items
// silently shared among and potentially reused by concurrent independent
// clients of a package. Pool provides a way to amortize allocation overhead
// across many clients.
//
// An example of good use of a Pool is in the fmt package, which maintains a
// dynamically-sized store of temporary output buffers. The store scales under
// load (when many goroutines are actively printing) and shrinks when
// quiescent.
//
// On the other hand, a free list maintained as part of a short-lived object is
// not a suitable use for a Pool, since the overhead does not amortize well in
// that scenario. It is more efficient to have such objects implement their own
// free list.
//
// A Pool must not be copied after first use.
type Pool struct {
	// 用来标记，当前的 struct 是不能够被 copy 的
	noCopy noCopy
	// local字段存储的是一个poolLocal数组的指针，poolLocal数组大小是goroutine中P的数量，访问时，P的id对应poolLocal数组下标索引，所以Pool的最大个数runtime.GOMAXPROCS()
	// 通过这样的设计，每个P都有了自己的本地空间，多个 goroutine 使用同一个 Pool 时，减少了竞争，提升了性能。
	//有当前主要的空闲可用的元素都存放在 local 字段中,请求元素时也是优先从 local 字段中查找可用的元素。
	local     unsafe.Pointer // local fixed-size per-P pool, actual type is [P]poolLocal
	// 上面数组的大小，即 P 的个数
	localSize uintptr        // size of the local array
	// 同 local 和 localSize，只是在 gc 的过程中保留一次
	victim     unsafe.Pointer // local from previous cycle
	victimSize uintptr        // size of victims array

	// New optionally specifies a function to generate
	// a value when Get would otherwise return nil.
	// It may not be changed concurrently with calls to Get.
	// New函数是在创建pool的时候设置的，当pool没有缓存对象的时候，会调用New方法生成一个新的对象
	New func() interface{}
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20

type poolLocal struct {
	poolLocalInternal

	// Prevents false sharing on widespread platforms with
	// 128 mod (cache line size) = 0 .
	// poolLocal里面有一个pad数组用来占位用，防止在 cache line 上分配多个 poolLocalInternal从而造成false sharing
	// pad是防止伪共享
	pad [128 - unsafe.Sizeof(poolLocalInternal{})%128]byte
}

// Local per-P Pool appendix.
// 当前调度器的内部资源
type poolLocalInternal struct {
	// private,代表一个缓存的元素,而且只能由相应的一个 P 存取。因为一个 P 同时只能执行一个 goroutine,所以不会有并发的问题。
	// 当前调度器的私有资源
	private interface{} // Can be used only by the respective P.
	// shared,可以由任意的 P 访问,但是只有本地的 P 才能 pushHead/popHead,其它 P 可以 popTail,相当于只有一个本地的 P 作为生产者(Producer),多个 P 作为消费者(Consumer),它是使用一个 local-free 的 queue 列表实现的。
	// 所有调度器的公有资源
	shared  poolChain   // Local P can pushHead/popHead; any P can popTail.
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18

// poolChain is a dynamically-sized version of poolDequeue.
//
// This is implemented as a doubly-linked list queue of poolDequeues
// where each dequeue is double the size of the previous one. Once a
// dequeue fills up, this allocates a new one and only ever pushes to
// the latest dequeue. Pops happen from the other end of the list and
// once a dequeue is exhausted, it gets removed from the list.
type poolChain struct {
	// head is the poolDequeue to push to. This is only accessed
	// by the producer, so doesn't need to be synchronized.
	// 头指针，只能单一producer操作(push, pop)
	head *poolChainElt

	// tail is the poolDequeue to popTail from. This is accessed
	// by consumers, so reads and writes must be atomic.
	// 尾指针，可以被多个consumer pop，必须是原子操作
	tail *poolChainElt
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15

type poolChainElt struct {
	poolDequeue

	// next and prev link to the adjacent poolChainElts in this
	// poolChain.
	//
	// next is written atomically by the producer and read
	// atomically by the consumer. It only transitions from nil to
	// non-nil.
	//
	// prev is written atomically by the consumer and read
	// atomically by the producer. It only transitions from
	// non-nil to nil.
	next, prev *poolChainElt
}

1
2
3
4
5
6

func storePoolChainElt(pp **poolChainElt, v *poolChainElt) {
	atomic.StorePointer((*unsafe.Pointer)(unsafe.Pointer(pp)), unsafe.Pointer(v))
}
func loadPoolChainElt(pp **poolChainElt) *poolChainElt {
	return (*poolChainElt)(atomic.LoadPointer((*unsafe.Pointer)(unsafe.Pointer(pp))))
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

func (c *poolChain) pushHead(val interface{}) {
	d := c.head
	// 如果head为nil，说明队列现在是空的，那么新建一个节点，将head和tail都指向这个节点
	// 如果链表空，则创建一个新的链表
	if d == nil {
		// Initialize the chain.
		// 固定长度为 8，必须为 2 的指数
		const initSize = 8 // Must be a power of 2
		d = new(poolChainElt)
		d.vals = make([]eface, initSize)
		c.head = d
		// 设置尾指针
		storePoolChainElt(&c.tail, d)
	}
	// 将val push到head的环形队列中，如果push成功了，可以返回了
	if d.pushHead(val) {
		return
	}

	// The current dequeue is full. Allocate a new one of twice
	// the size.
	// 如果没push成功，则说明head的环形队列满了，就再创建一个两倍head大小的节点[最大(1 << 32) / 4]，
	newSize := len(d.vals) * 2
	if newSize >= dequeueLimit {
		// Can't make it any bigger.
		newSize = dequeueLimit
	}
	// 创建一个新的poolChainElt并在链表头部插入
	d2 := &poolChainElt{prev: d}
	d2.vals = make([]eface, newSize)
	// 将新节点作为head，并且处理好新head和旧head的next，prev关系
	c.head = d2
	storePoolChainElt(&d.next, d2)
	// 将val push到head的环形队列中
	d2.pushHead(val)
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18

func (c *poolChain) popHead() (interface{}, bool) {
	// 先在head环形队列中popHead试试
	d := c.head
	for d != nil {
		// 能取到就返回，取不到说明队列空了
		// d 是一个 poolDequeue，如果 d.popHead 是并发安全的，
		// 那么这里取 val 也是并发安全的。若 d.popHead 失败，则
		// 说明需要重新尝试。这个过程会持续到整个链表为空。
		if val, ok := d.popHead(); ok {
			return val, ok
		}
		// There may still be unconsumed elements in the
		// previous dequeue, so try backing up.
		// 如果空了，当前节点就没用了，就删掉当前节点，去prev节点并且把prev节点作为新head再取一值递归下去，
		d = loadPoolChainElt(&d.prev)
	}
	return nil, false
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41

func (c *poolChain) popTail() (interface{}, bool) {
	d := loadPoolChainElt(&c.tail)
	// 如果tail为nil，说明队列是空的，直接返回
	if d == nil {
		return nil, false
	}
	// 如果tail非nil，就取取试试，有东西就返回
	for {
		// It's important that we load the next pointer
		// *before* popping the tail. In general, d may be
		// transiently empty, but if next is non-nil before
		// the pop and the pop fails, then d is permanently
		// empty, which is the only condition under which it's
		// safe to drop d from the chain.
		d2 := loadPoolChainElt(&d.next)
		// 尝试从当前poolChainElt的队列尾部取，成功则直接返回
		if val, ok := d.popTail(); ok {
			return val, ok
		}
		// 如果没取出来东西，那么说明tail节点没存东西了，递归去prev节点环形队列中popTail，并且把prev节点作为tail，能取到就返回，取不到就是空了
		if d2 == nil {
			// This is the only dequeue. It's empty right
			// now, but could be pushed to in the future.
			// d2为空表明链表只有一个节点，而从该节点取对象已失败，则返回
			return nil, false
		}

		// The tail of the chain has been drained, so move on
		// to the next dequeue. Try to drop it from the chain
		// so the next pop doesn't have to look at the empty
		// dequeue again.
		if atomic.CompareAndSwapPointer((*unsafe.Pointer)(unsafe.Pointer(&c.tail)), unsafe.Pointer(d), unsafe.Pointer(d2)) {
			// We won the race. Clear the prev pointer so
			// the garbage collector can collect the empty
			// dequeue and so popHead doesn't back up
			// further than necessary.
			storePoolChainElt(&d2.prev, nil)
		}
		d = d2
	}
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58

// poolDequeue is a lock-free fixed-size single-producer,
// multi-consumer queue. The single producer can both push and pop
// from the head, and consumers can pop from the tail.
//
// It has the added feature that it nils out unused slots to avoid
// unnecessary retention of objects. This is important for sync.Pool,
// but not typically a property considered in the literature.
type poolDequeue struct {
	// headTail packs together a 32-bit head index and a 32-bit
	// tail index. Both are indexes into vals modulo len(vals)-1.
	//
	// tail = index of oldest data in queue
	// head = index of next slot to fill
	//
	// Slots in the range [tail, head) are owned by consumers.
	// A consumer continues to own a slot outside this range until
	// it nils the slot, at which point ownership passes to the
	// producer.
	//
	// The head index is stored in the most-significant bits so
	// that we can atomically add to it and the overflow is
	// harmless.
	headTail uint64

	// vals is a ring buffer of interface{} values stored in this
	// dequeue. The size of this must be a power of 2.
	//
	// vals[i].typ is nil if the slot is empty and non-nil
	// otherwise. A slot is still in use until *both* the tail
	// index has moved beyond it and typ has been set to nil. This
	// is set to nil atomically by the consumer and read
	// atomically by the producer.
	vals []eface
}
// 存储元素的结构体，类型指针和值指针
// Pool 底层用 eface 来存储单个 Object, 包括 typ 指针: Object 的类型，val 指针: Object 的值
type eface struct {
	typ, val unsafe.Pointer
}



// dequeueLimit is the maximum size of a poolDequeue.
//
// This must be at most (1<<dequeueBits)/2 because detecting fullness
// depends on wrapping around the ring buffer without wrapping around
// the index. We divide by 4 so this fits in an int on 32-bit.

// 为什么dequeueLimit是(1 << 32) / 4 = 1 << 30 ?
// dequeueLimit 必须是2的幂(上边解释过)
// head和tail都是32位，最大是1 << 31，如果都用的话，head和tail就是无符号整型，无符号整型使用的时候会有很多上溢的错误，这类错误是不容易检测的，所以相比之下还不如用31位有符号整型，有错就报出来
const dequeueLimit = (1 << dequeueBits) / 4

// dequeueNil is used in poolDeqeue to represent interface{}(nil).
// Since we use nil to represent empty slots, we need a sentinel value
// to represent nil.
type dequeueNil *struct{}

1
2
3
4
5
6

[hhhhhhhh hhhhhhhh hhhhhhhh hhhhhhhh tttttttt tttttttt tttttttt tttttttt]
1. headTail表示下标，高32位表示头下标，低32位表示尾下标，poolDequeue定义了，head tail的pack和unpack函数方便转化，
实际用的时候都会mod ( len(vals) - 1 ) 来防止溢出
2. head和tail永远只用32位表示，溢出后会从0开始，这也满足循环队列的设计
3. 队列为空的条件  tail == head
4. 队列满的条件    (tail+uint32(len(d.vals)))&(1<<dequeueBits-1) == head tail加上队列长度和head相等(实际上就是队列已有的空间都有值了,满了)

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16

const dequeueBits = 32

// 将headTail分解为head和tail
func (d *poolDequeue) unpack(ptrs uint64) (head, tail uint32) {
	const mask = 1<<dequeueBits - 1
	head = uint32((ptrs >> dequeueBits) & mask)
	tail = uint32(ptrs & mask)
	return
}

// 将head和tail组合成headTail
func (d *poolDequeue) pack(head, tail uint32) uint64 {
	const mask = 1<<dequeueBits - 1
	return (uint64(head) << dequeueBits) |
		uint64(tail&mask)
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38

// popHead removes and returns the element at the head of the queue.
// It returns false if the queue is empty. It must only be called by a
// single producer.
// 获取并删除队首元素
func (d *poolDequeue) popHead() (interface{}, bool) {
	var slot *eface
	for {
		ptrs := atomic.LoadUint64(&d.headTail)
		head, tail := d.unpack(ptrs)
		if tail == head {
			// Queue is empty.
			// 队列为空，获取对象失败
			return nil, false
		}

		// Confirm tail and decrement head. We do this before
		// reading the value to take back ownership of this
		// slot.
		// 因为head是下一个对象存储的位置，因此从队列头部获取对象需先将head-1
		head--
		ptrs2 := d.pack(head, tail)
		if atomic.CompareAndSwapUint64(&d.headTail, ptrs, ptrs2) {
			// We successfully took back slot.
			slot = &d.vals[head&uint32(len(d.vals)-1)]
			break
		}
	}

	val := *(*interface{})(unsafe.Pointer(slot))
	if val == dequeueNil(nil) {
		val = nil
	}
	// Zero the slot. Unlike popTail, this isn't racing with
	// pushHead, so we don't need to be careful here.
	// 重置slot
	*slot = eface{}
	return val, true
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

// pushHead adds val at the head of the queue. It returns false if the
// queue is full. It must only be called by a single producer.
// 添加元素到队首
func (d *poolDequeue) pushHead(val interface{}) bool {
	ptrs := atomic.LoadUint64(&d.headTail)
	// 解析出head、tail的索引
	head, tail := d.unpack(ptrs)
	if (tail+uint32(len(d.vals)))&(1<<dequeueBits-1) == head {
		// 队列已满，插入失败
		// Queue is full.
		return false
	}
	// 找到head的槽位
	slot := &d.vals[head&uint32(len(d.vals)-1)]

	// Check if the head slot has been released by popTail.
	// 此处可能与 popTail 发生竞争，参见 popTail
	// 检测这个槽位有没被popTail释放
	typ := atomic.LoadPointer(&slot.typ)
	if typ != nil {
		// Another goroutine is still cleaning up the tail, so
		// the queue is actually still full.
		return false
	}

	// The head slot is free, so we own it.
	if val == nil {
		val = dequeueNil(nil)
	}
	*(*interface{})(unsafe.Pointer(slot)) = val

	// Increment head. This passes ownership of slot to popTail
	// and acts as a store barrier for writing the slot.
	atomic.AddUint64(&d.headTail, 1<<dequeueBits)
	return true
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

// popTail removes and returns the element at the tail of the queue.
// It returns false if the queue is empty. It may be called by any
// number of consumers.
// 获取并删除队尾元素
func (d *poolDequeue) popTail() (interface{}, bool) {
	var slot *eface
	for {
		ptrs := atomic.LoadUint64(&d.headTail)
		head, tail := d.unpack(ptrs)
		if tail == head {
			// Queue is empty.
			// 队列为空，直接返回
			return nil, false
		}

		// Confirm head and tail (for our speculative check
		// above) and increment tail. If this succeeds, then
		// we own the slot at tail.
		// tail+1，表明尾部的对象已被获取
		ptrs2 := d.pack(head, tail+1)
		if atomic.CompareAndSwapUint64(&d.headTail, ptrs, ptrs2) {
			// Success.
			slot = &d.vals[tail&uint32(len(d.vals)-1)]
			break
		}
	}

	// We now own slot.
	val := *(*interface{})(unsafe.Pointer(slot))
	if val == dequeueNil(nil) {
		val = nil
	}

	// Tell pushHead that we're done with this slot. Zeroing the
	// slot is also important so we don't leave behind references
	// that could keep this object live longer than necessary.
	//
	// We write to val first and then publish that we're done with
	// this slot by atomically writing to typ.
	// 注意：此处可能与 pushHead 发生竞争，解决方案是：
	// 1. 让 pushHead 先读取 typ 的值，如果 typ 值不为 nil，则说明 popTail 尚未清理完 slot
	// 2. 让 popTail 先清理掉 val 中的内容，在清理掉 typ，从而确保不会与 pushHead 对 slot 的写行为发生竞争
	// 将slot置空
	slot.val = nil
	atomic.StorePointer(&slot.typ, nil)
	// At this point pushHead owns the slot.

	return val, true
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

// Get selects an arbitrary item from the Pool, removes it from the
// Pool, and returns it to the caller.
// Get may choose to ignore the pool and treat it as empty.
// Callers should not assume any relation between values passed to Put and
// the values returned by Get.
//
// If Get would otherwise return nil and p.New is non-nil, Get returns
// the result of calling p.New.
// Get 从 Pool 中选择一个任意的对象，将其移出 Pool, 并返回给调用方。
// Get 可能会返回一个非零值对象（被其他人使用过），因此调用方不应假设
// 返回的对象具有任何形式的状态。
func (p *Pool) Get() interface{} {
	if race.Enabled {
		race.Disable()
	}
	// 把当前goroutine固定在当前的P上
	// 获取一个 poolLocal
	l, pid := p.pin()
	// 优先从local的private字段取,快速
	// 先从 private 获取对象
	x := l.private
	l.private = nil
	if x == nil {
		// Try to pop the head of the local shard. We prefer
		// the head over the tail for temporal locality of
		// reuse.
		// 从当前的local.shared弹出一个,注意是从head读取并移除
		// 尝试从 localPool 的 shared 队列队头读取，
		// 因为队头的内存局部性比队尾更好。
		x, _ = l.shared.popHead()
		if x == nil {
			// 当前P的private和share都取不到，则去其他P的share读取
			x = p.getSlow(pid)
		}
	}
	runtime_procUnpin()
	if race.Enabled {
		race.Enable()
		if x != nil {
			race.Acquire(poolRaceAddr(x))
		}
	}
	// 如果没有获取到,尝试使用New函数生成一个新的
	// 如果 getSlow 还是获取不到，则 New 一个
	if x == nil && p.New != nil {
		x = p.New()
	}
	return x
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

// pin pins the current goroutine to P, disables preemption and
// returns poolLocal pool for the P and the P's id.
// Caller must call runtime_procUnpin() when done with the pool.
// 将当前goroutine和P绑定，禁止抢占，并返回对应的poolLocal和P的id
// 调用方在调用完成后必须调用runtime_procUnpin方法取消抢占
func (p *Pool) pin() (*poolLocal, int) {

	// 返回当前 P.id
	pid := runtime_procPin()
	// In pinSlow we store to local and then to localSize, here we load in opposite order.
	// Since we've disabled preemption, GC cannot happen in between.
	// Thus here we must observe local at least as large localSize.
	// We can observe a newer/larger local, it is fine (we must observe its zero-initialized-ness).
	// 在 pinSlow 中会存储 localSize 后再存储 local，因此这里反过来读取
	// 因为我们已经禁用了抢占，这时不会发生 GC
	// 因此，我们必须观察 local 和 localSize 是否对应
	// 观察到一个全新或很大的的 local 是正常行为
	s := atomic.LoadUintptr(&p.localSize) // load-acquire
	l := p.local                          // load-consume
	// 因为可能存在动态的 P（运行时调整 P 的个数）procresize/GOMAXPROCS
	// 如果 P.id 没有越界，则直接返回
	if uintptr(pid) < s {
		// 对应的poolLocal已创建，调用indexLocal取出对应的poolLocal返回
		return indexLocal(l, pid), pid
	}
	// 没有结果时，涉及全局加锁
	// 例如重新分配数组内存，添加到全局列表
	return p.pinSlow()
}

1
2
3
4
5

func indexLocal(l unsafe.Pointer, i int) *poolLocal {
	// 简单的通过 p.local 的头指针与索引来第 i 个 pooLocal
	lp := unsafe.Pointer(uintptr(l) + uintptr(i)*unsafe.Sizeof(poolLocal{}))
	return (*poolLocal)(lp)
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43

var (
	allPoolsMu Mutex
	// allPools 是一组 pool 的集合，具有非空主缓存。
	// 有两种形式来保护它的读写：1. allPoolsMu 锁; 2. STW.
	allPools   []*Pool
)
// 对应的poolLocal不存在，则创建
func (p *Pool) pinSlow() (*poolLocal, int) {
	// Retry under the mutex.
	// Can not lock the mutex while pinned.
	// 这时取消 P 的禁止抢占，因为使用 mutex 时候 P 必须可抢占
	// 加锁期间须先取消抢占
	runtime_procUnpin()
	// 加锁
	allPoolsMu.Lock()
	defer allPoolsMu.Unlock()
	// 当锁住后，再次固定 P 取其 id
	pid := runtime_procPin()
	// poolCleanup won't be called while we are pinned.
	// 并再次检查是否符合条件，因为可能中途已被其他线程调用
	// 当再次固定 P 时 poolCleanup 不会被调用
	s := p.localSize
	l := p.local
	// 第二次检测，因为在加锁过程中可能别的goroutine调用创建了poolLocal
	if uintptr(pid) < s {
		return indexLocal(l, pid), pid
	}
	// 如果数组为空，新建
	// 将其添加到 allPools，垃圾回收器从这里获取所有 Pool 实例
	if p.local == nil {
		allPools = append(allPools, p)
	}
	// 根据 P 数量创建 slice，如果 GOMAXPROCS 在 GC 间发生变化
	// 我们重新分配此数组并丢弃旧的
	// If GOMAXPROCS changes between GCs, we re-allocate the array and lose the old one.
	size := runtime.GOMAXPROCS(0)
	local := make([]poolLocal, size)
	// 将底层数组起始指针保存到 p.local，并设置 p.localSize
	atomic.StorePointer(&p.local, unsafe.Pointer(&local[0])) // store-release
	atomic.StoreUintptr(&p.localSize, uintptr(size))         // store-release
	// 返回所需的 pollLocal
	return &local[pid], pid
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54

func (p *Pool) getSlow(pid int) interface{} {
	// See the comment in pin regarding ordering of the loads.
	size := atomic.LoadUintptr(&p.localSize) // load-acquire
	locals := p.local                        // load-consume
	// Try to steal one element from other procs.
	// 尝试从其他P获取对象，成功则直接返回
	for i := 0; i < int(size); i++ {
		// 获取目标 poolLocal, 引入 pid 保证不是自身
		l := indexLocal(locals, (pid+i+1)%int(size))
		// 从其他的 P 中固定的 localPool 的 share 队列的队尾偷一个缓存对象
		if x, _ := l.shared.popTail(); x != nil {
			return x
		}
	}

	// Try the victim cache. We do this after attempting to steal
	// from all primary caches because we want objects in the
	// victim cache to age out if at all possible.
	// 如果其它proc也没有可用元素,那么尝试从vintim中获取
	// 当 local 失败后，尝试再尝试从上一个垃圾回收周期遗留下来的 victim。
	// 如果 pid 比 victim 遗留的 localPool 还大，则说明从根据此 pid 从
	// victim 获取 localPool 会发生越界（同时也表明此时 P 的数量已经发生变化）
	// 这时无法继续读取，直接返回 nil
	// 尝试从victim cache中获取对象
	size = atomic.LoadUintptr(&p.victimSize)
	if uintptr(pid) >= size {
		return nil
	}
	// 获取 localPool，并优先读取 private
	locals = p.victim
	l := indexLocal(locals, pid)
	// 同样的逻辑,先从vintim中的local private获取
	if x := l.private; x != nil {
		l.private = nil
		return x
	}
	for i := 0; i < int(size); i++ {
		// 从vintim其它proc尝试偷取
		l := indexLocal(locals, (pid+i)%int(size))
		// 从其他的 P 中固定的 localPool 的 share 队列的队尾偷一个缓存对象
		if x, _ := l.shared.popTail(); x != nil {
			return x
		}
	}

	// Mark the victim cache as empty for future gets don't bother
	// with it.
	// 如果victim中都没有,则把这个victim标记为空,以后的查找可以快速跳过了
	// 将 victim 缓存置空，从而确保之后的 get 操作不再读取此处的值
	// 清空 victim cache。下次就不用再从这里找了
	atomic.StoreUintptr(&p.victimSize, 0)

	return nil
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

// Put adds x to the pool.
func (p *Pool) Put(x interface{}) {
	// 如果存入对象为nil直接返回
	if x == nil {
		return
	}
	if race.Enabled {
		if fastrand()%4 == 0 {
			// Randomly drop x on floor.
			return
		}
		race.ReleaseMerge(poolRaceAddr(x))
		race.Disable()
	}
	// 获得一个 localPool
	l, _ := p.pin()
	// 优先放入 private
	if l.private == nil {
		l.private = x
		x = nil
	}
	// 如果不能放入 private 则放入 shared
	if x != nil {
		l.shared.pushHead(x)
	}
	runtime_procUnpin()
	if race.Enabled {
		race.Enable()
	}
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26

//go:linkname sync_runtime_procPin sync.runtime_procPin
//go:nosplit
func sync_runtime_procPin() int {
	return procPin()
}

//go:linkname sync_runtime_procUnpin sync.runtime_procUnpin
//go:nosplit
func sync_runtime_procUnpin() {
	procUnpin()
}

//go:nosplit
func procPin() int {
	_g_ := getg()
	mp := _g_.m

	mp.locks++
	return int(mp.p.ptr().id)
}

//go:nosplit
func procUnpin() {
	_g_ := getg()
	_g_.m.locks--
}

1
2
3
4
5
6
7

// 将缓存清理函数注册到运行时 GC 时间段
func init() {
	runtime_registerPoolCleanup(poolCleanup)
}

// 由运行时实现
func runtime_registerPoolCleanup(cleanup func())

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

// 开始 GC
func gcStart(trigger gcTrigger) {
	...
	clearpools()
	...
}

// 实现缓存清理
func clearpools() {
	// clear sync.Pools
	if poolcleanup != nil {
		poolcleanup()
	}
	...
}

var poolcleanup func()

// 利用编译器标志将 sync 包中的清理注册到运行时
//go:linkname sync_runtime_registerPoolCleanup sync.runtime_registerPoolCleanup
func sync_runtime_registerPoolCleanup(f func()) {
	poolcleanup = f
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37

var (
	allPoolsMu Mutex

	// allPools is the set of pools that have non-empty primary
	// caches. Protected by either 1) allPoolsMu and pinning or 2)
	// STW.
	allPools []*Pool

	// oldPools is the set of pools that may have non-empty victim
	// caches. Protected by STW.
	// oldPools 是一组 pool 的集合，具有非空 victim 缓存。由 STW 保护
	oldPools []*Pool
)
func poolCleanup() {
	// 该函数会注册到运行时 GC 阶段(前)，此时为 STW 状态，不需要加锁
	// 它必须不处理分配且不调用任何运行时函数。

	// 由于此时是 STW，不存在用户态代码能尝试读取 localPool，进而所有的 P 都已固定（与 goroutine 绑定）

	// 从所有的 oldPools 中删除 victim
	for _, p := range oldPools {
		p.victim = nil
		p.victimSize = 0
	}

	// 将主缓存移动到 victim 缓存
	for _, p := range allPools {
		p.victim = p.local
		p.victimSize = p.localSize

		p.local = nil
		p.localSize = 0
	}

	// 具有非空主缓存的池现在具有非空的 victim 缓存，并且没有任何 pool 具有主缓存。
	oldPools, allPools = allPools, nil
}

1
2
3
4
5
6
7
8
9

        goroutine      goroutine       goroutine
           |               |               |
           P               P               P
           |               |               |
         private        private          private
           |               |               |
    [   poolLocal      poolLocal        poolLocal  ]    sync.Pool
           |               |               |
         shared          shared          shared