概括
Go的垃圾回收官方形容为 非分代 非紧缩 写屏障 三色并发标记清理算法。
非分代:不像Java那样分为年轻代和年老代,自然也没有minor gc和maj o gc的区别。
非紧缩:在垃圾回收之后不会进行内存整理以清除内存碎片。
写屏障:在并发标记的过程中,如果应用程序(mutator)修改了对象图,就可能出现标记遗漏的可能,写屏障就是为了处理标记遗漏的问题。
三色:将GC中的对象按照搜索的情况分成三种:
- 黑色: 对象在这次GC中已标记,且这个对象包含的子对象也已标记
- 灰色: 对象在这次GC中已标记, 但这个对象包含的子对象未标记
- 白色: 对象在这次GC中未标记
并发:可以和应用程序(mutator)在一定程度上并发执行。
标记清理:GC算法分为两个大步骤:标记阶段找出要回收的对象,清理阶段则回收未被标记的对象(要被回收的对象)
触发时机
- gcTriggerAlways: 强制触发GC,没找到什么情况下使用这个
- gcTriggerHeap: 当前分配的内存达到一定值(动态计算)就触发GC
- gcTriggerTime: 当一定时间(2分钟)没有执行过GC就触发GC
- gcTriggerCycle: 要求启动新一轮的GC, 已启动则跳过, 手动触发GC的runtime.GC()会使用这个条件
func gcStart(mode gcMode, trigger gcTrigger) {
// Since this is called from malloc and malloc is called in
// the guts of a number of libraries that might be holding
// locks, don't attempt to start GC in non-preemptible or
// potentially unstable situations.
mp := acquirem()
if gp := getg(); gp == mp.g0 || mp.locks > 1 || mp.preemptoff != "" {
releasem(mp)
return
}
releasem(mp)
mp = nil
// 检查GC条件是否满足,和下面的test()构成双检查锁,如果满足GC条件但目前处于GC清理阶段,那就参与清理
for trigger.test() && gosweepone() != ^uintptr(0) {
sweep.nbgsweep++
}
// 加锁检查
semacquire(&work.startSema)
if !trigger.test() {
semrelease(&work.startSema)
return
}
/*************** ..... *****************/
}
在trigger.test()函数中,检查是否满足GC触发的条件
func (t gcTrigger) test() bool {
if !memstats.enablegc || panicking != 0 {
return false
}
if t.kind == gcTriggerAlways {
return true
}
if gcphase != _GCoff {
return false
}
switch t.kind {
case gcTriggerHeap:
// Non-atomic access to heap_live for performance. If
// we are going to trigger on this, this thread just
// atomically wrote heap_live anyway and we'll see our
// own write.
return memstats.heap_live >= memstats.gc_trigger
case gcTriggerTime:
if gcpercent < 0 {
return false
}
lastgc := int64(atomic.Load64(&memstats.last_gc_nanotime))
// forcegcperiod = 2分钟
return lastgc != 0 && t.now-lastgc > forcegcperiod
case gcTriggerCycle:
// t.n > work.cycles, but accounting for wraparound.
return int32(t.n-work.cycles) > 0
}
return true
}
const (
// gcTriggerAlways indicates that a cycle should be started
// unconditionally, even if GOGC is off or we're in a cycle
// right now. This cannot be consolidated with other cycles.
gcTriggerAlways gcTriggerKind = iota
// gcTriggerHeap indicates that a cycle should be started when
// the heap size reaches the trigger heap size computed by the
// controller.
gcTriggerHeap
// gcTriggerTime indicates that a cycle should be started when
// it's been more than forcegcperiod nanoseconds since the
// previous GC cycle.
gcTriggerTime
// gcTriggerCycle indicates that a cycle should be started if
// we have not yet started cycle number gcTrigger.n (relative
// to work.cycles).
gcTriggerCycle
)
算法过程
- Sweep Termination: 对未清扫的span进行清扫, 只有上一轮的GC的清扫工作完成才可以开始新一轮的GC
- Mark: 扫描所有根对象, 和根对象可以到达的所有对象, 标记它们不被回收
- Mark Termination: 完成标记工作, 重新扫描部分根对象(要求STW)
-
Sweep: 按标记结果清扫span
func gcStart(mode gcMode, trigger gcTrigger) {
// 拿到锁,保证只有一个执行流进入到这个临界区
semacquire(&worldsema)
// 启动后台扫描任务(G)
if mode == gcBackgroundMode {
gcBgMarkStartWorkers()
}
gcResetMarkState()
work.stwprocs, work.maxprocs = gomaxprocs, gomaxprocs
if work.stwprocs > ncpu {
work.stwprocs = ncpu
}
work.heap0 = atomic.Load64(&memstats.heap_live)
work.pauseNS = 0
work.mode = mode
now := nanotime()
work.tSweepTerm = now
work.pauseStart = now
if trace.enabled {
traceGCSTWStart(1)
}
systemstack(stopTheWorldWithSema)
// Finish sweep before we start concurrent scan.
systemstack(func() {
finishsweep_m()
})
// clearpools before we start the GC. If we wait they memory will not be
// reclaimed until the next GC cycle.
clearpools()
work.cycles++
if mode == gcBackgroundMode { // Do as much work concurrently as possible
gcController.startCycle()
work.heapGoal = memstats.next_gc
// Enter concurrent mark phase and enable
// write barriers.
setGCPhase(_GCmark)
gcBgMarkPrepare() // Must happen before assist enable.
gcMarkRootPrepare()
// Mark all active tinyalloc blocks. Since we're
// allocating from these, they need to be black like
// other allocations. The alternative is to blacken
// the tiny block on every allocation from it, which
// would slow down the tiny allocator.
gcMarkTinyAllocs()
// At this point all Ps have enabled the write
// barrier, thus maintaining the no white to
// black invariant. Enable mutator assists to
// put back-pressure on fast allocating
// mutators.
atomic.Store(&gcBlackenEnabled, 1)
// Assists and workers can start the moment we start
// the world.
gcController.markStartTime = now
// Concurrent mark.
systemstack(func() {
now = startTheWorldWithSema(trace.enabled)
})
work.pauseNS += now - work.pauseStart
work.tMark = now
}
semrelease(&work.startSema)
}
关键函数及路径:
- gcBgMarkStartWorkers():准备后台标记工作goroutine(allp), 启动后等待该任务通知信号量bgMarkReady再继续,notewakeup(&work.bgMarkReady)
- gcResetMarkState():重置一些全局状态和所有gorontine的栈(一种根对象)扫描状态
- systemstack(stopTheWorldWithSema):启动stop the world
- systemstack(func(){finishsweep_m()}): 不断去除要清理的span进行清理,然后重置gcmark位
- clearpools(): 清扫sched.sudogcache和sched.deferpool,不知道在干嘛......
- gcController.startCycle():启动新一轮GC,设置gc controller的状态位和计算一些估计值
- setGCPhase(_GCmark):设置GC阶段,启用写屏障
- gcBgMarkPrepare():设置后台标记任务计数;work.nproc = ^uint32(0),work.nwait = ^uint32(0)
- gcMarkRootPrepare(): 计算扫描根对象的任务数量
- gcMarkTinyAllocs(): 标记所有tiny alloc等待合并的对象
- atomic.Store(&gcBlackenEnabled, 1): 启用辅助GC
- systemstack(func(){now=startTheWorldWithSema(trace.enable)}): 停止stop the world
func gcBgMarkWorker(_p_ *p) {
/********** ....... ***********/
// 通知gcBgMarkStartWorkers可以继续处理
notewakeup(&work.bgMarkReady)
for {
// 切换到g0运行
systemstack(func() {
// Mark our goroutine preemptible so its stack
// can be scanned. This lets two mark workers
// scan each other (otherwise, they would
// deadlock). We must not modify anything on
// the G stack. However, stack shrinking is
// disabled for mark workers, so it is safe to
// read from the G stack.
casgstatus(gp, _Grunning, _Gwaiting)
switch _p_.gcMarkWorkerMode {
default:
throw("gcBgMarkWorker: unexpected gcMarkWorkerMode")
case gcMarkWorkerDedicatedMode:
gcDrain(&_p_.gcw, gcDrainUntilPreempt|gcDrainFlushBgCredit)
if gp.preempt {
lock(&sched.lock)
for {
gp, _ := runqget(_p_)
if gp == nil {
break
}
globrunqput(gp)
}
unlock(&sched.lock)
}
// Go back to draining, this time
// without preemption.
gcDrain(&_p_.gcw, gcDrainNoBlock|gcDrainFlushBgCredit)
case gcMarkWorkerFractionalMode:
gcDrain(&_p_.gcw, gcDrainFractional|gcDrainUntilPreempt|gcDrainFlushBgCredit)
case gcMarkWorkerIdleMode:
gcDrain(&_p_.gcw, gcDrainIdle|gcDrainUntilPreempt|gcDrainFlushBgCredit)
}
casgstatus(gp, _Gwaiting, _Grunning)
})
/******** ...... ***********/
// 判断是否所有后台标记任务都完成, 并且没有更多的任务
if incnwait == work.nproc && !gcMarkWorkAvailable(nil) {
gcMarkDone()
}
}
}
- gcDrain()是执行标记的函数
- 当所有标记任务完成时,执行gcMarkDone()函数
func gcDrain(gcw *gcWork, flags gcDrainFlags) {
initScanWork := gcw.scanWork
// 如果根对象未扫描完,则先扫描根对象,Jobs为根对象总数,next相当于一个对象任务的取数器
if work.markrootNext < work.markrootJobs {
for !(preemptible && gp.preempt) {
job := atomic.Xadd(&work.markrootNext, +1) - 1
if job >= work.markrootJobs {
break
}
// 将会扫描根对象,并把它加入到标记队列gcWork中之中,也就是把对象变成灰色
markroot(gcw, job)
if check != nil && check() {
goto done
}
}
}
// 当根对象全部put到标记队列中, 消费标记队列,根据对象图进行消费
for !(preemptible && gp.preempt) {
if work.full == 0 {
gcw.balance()
}
var b uintptr
if blocking {
b = gcw.get()
} else {
b = gcw.tryGetFast()
if b == 0 {
b = gcw.tryGet()
}
}
if b == 0 {
// work barrier reached or tryGet failed.
break
}
scanobject(b, gcw)
// 如果已经扫描了一定数量的对象(gcCreditSlack的值是2000)
if gcw.scanWork >= gcCreditSlack {
// 把扫描的对象数量添加到全局
atomic.Xaddint64(&gcController.scanWork, gcw.scanWork)
// 减少辅助GC的工作量和唤醒等待中的G
if flushBgCredit {
gcFlushBgCredit(gcw.scanWork - initScanWork)
initScanWork = 0
}
idleCheck -= gcw.scanWork
gcw.scanWork = 0
// 如果是idle模式且达到了检查的扫描量, 则检查是否有其他任务(G), 如果有则跳出循环
if idle && idleCheck <= 0 {
idleCheck += idleCheckThreshold
if pollWork() {
break
}
}
}
}
done:
// 把扫描的对象数量添加到全局
if gcw.scanWork > 0 {
atomic.Xaddint64(&gcController.scanWork, gcw.scanWork)
// 减少辅助GC的工作量和唤醒等待中的G
if flushBgCredit {
gcFlushBgCredit(gcw.scanWork - initScanWork)
}
gcw.scanWork = 0
}
}
func gcMarkDone() {
semacquire(&work.markDoneSema)
// Re-check transition condition under transition lock.
if !(gcphase == _GCmark && work.nwait == work.nproc && !gcMarkWorkAvailable(nil)) {
semrelease(&work.markDoneSema)
return
}
// 暂时禁止启动新的后台标记任务
atomic.Xaddint64(&gcController.dedicatedMarkWorkersNeeded, -0xffffffff)
prevFractionalGoal := gcController.fractionalUtilizationGoal
gcController.fractionalUtilizationGoal = 0
// 转换到Mark Termination阶段,进入STW阶段
systemstack(stopTheWorldWithSema)
// 标记对根对象的扫描已完成
work.markrootDone = true
// 禁止辅助GC和后台任务
atomic.Store(&gcBlackenEnabled, 0)
// 唤醒所有因为辅助GC而休眠的G
gcWakeAllAssists()
semrelease(&work.markDoneSema)
// 计算下一次触发gc需要的heap大小
nextTriggerRatio := gcController.endCycle()
// 计算下一次触发gc需要的heap大小
gcMarkTermination(nextTriggerRatio)
}
func gcMarkTermination(nextTriggerRatio float64) {
// 禁止辅助GC和后台标记任务的运行
// 重新允许本地标记队列(下次GC使用)
// 设置当前GC阶段到完成标记阶段, 并启用写屏障
atomic.Store(&gcBlackenEnabled, 0)
gcBlackenPromptly = false
setGCPhase(_GCmarktermination)
systemstack(func() {gcMark(startTime)})
systemstack(func() {
// 设置当前GC阶段到关闭, 并禁用写屏障
setGCPhase(_GCoff)
// 唤醒后台清扫任务, 将在STW结束后开始运行
gcSweep(work.mode)
})
// 更新下一次触发gc需要的heap大小(gc_trigger)
gcSetTriggerRatio(nextTriggerRatio)
// 重置清扫状态
sweep.nbgsweep = 0
sweep.npausesweep = 0
// 统计执行GC的次数然后唤醒等待清扫的G
lock(&work.sweepWaiters.lock)
memstats.numgc++
injectglist(work.sweepWaiters.head.ptr())
work.sweepWaiters.head = 0
unlock(&work.sweepWaiters.lock)
// 重新启动世界
systemstack(func() { startTheWorldWithSema(true) })
// 移动标记队列使用的缓冲区到自由列表, 使得它们可以被回收
prepareFreeWorkbufs()
// 释放未使用的栈
systemstack(freeStackSpans)
semrelease(&worldsema)
// 重新允许当前的G被抢占
releasem(mp)
mp = nil
当标记的扫描工作完成之后,会进入到GC Mark Termination阶段,也就是gcMarkDone()函数,关键路径:
- systemstack(stopTheWorldWithSema):启动STW
- gcWakeAllAssists():唤醒所有因辅助gc而休眠的G
- nextTriggerRatio:=gcController.endCycle():计算下一次触发gc需要的heap大小
- setGCPhase(_GCmarktermination):启用写屏障
- systemstack(func() {gcMark(startTime)}): 再次执行标记
- systemstack(func(){setGCPhase(_GCoff);gcSweep(work.mode)}):关闭写屏障,唤醒后台清扫任务,将在STW结束后开始运行
- gcSetTriggerRatio(nextTriggerRatio):更新下次触发gc时的heap大小
- systemstack(func() { startTheWorldWithSema(true) }): 停止STW
STW分析:web程序中,我们关注最大停顿时间
STW出现在两个位置,分别是在初始标记阶段Mark和并发标记完成后重标记Mark Termination:
初始标记阶段:
- systemstack(stopTheWorldWithSema):启动stop the world
- systemstack(func(){finishsweep_m()}): 不断去除要清理的span进行清理,然后重置gcmark位
- clearpools(): 清扫sched.sudogcache和sched.deferpool,不知道在干嘛......
- gcController.startCycle():启动新一轮GC,设置gc controller的状态位和计算一些估计值
- gcMarkRootPrepare(): 计算扫描根对象的任务数量
- gcMarkTinyAllocs(): 涂灰所有tiny alloc等待合并的对象
- systemstack(func(){now=startTheWorldWithSema(trace.enable)}): 停止stop the world
找出其中比较耗时的阶段:
- finishsweep_m():如果上一次GC清扫阶段没有完成,那么在新的一轮GC阶段中就会在阻塞在这里,使得原本可以和应用程序并行的清扫阶段被放进STW。所以,如果频繁的执行GC,就可能会使得GC的最大停顿时间变长。
- clearpools():时间复杂度大概为:O(5*L),L为_defer中链表的长度。
- gcController.startCycle():O(P),P为go的P的数量,和cpu数有关,时间复杂度可以忽略
- gcMarkRootPrepare(): O(全局变量区),包括bss段和data段
- gcMarkTinyAllocs(): O(P)
个人觉得,对STW影响最大的是finishsweep_m()阶段,所有我们应该尽量避免让go在清扫期执行新一轮的GC。
重新标记阶段
- systemstack(stopTheWorldWithSema):启动STW
- gcWakeAllAssists():唤醒所有因辅助gc而休眠的G
- nextTriggerRatio:=gcController.endCycle():计算下一次触发gc需要的heap大小
- setGCPhase(_GCmarktermination):启用写屏障
- systemstack(func() {gcMark(startTime)}): 再次执行标记
- systemstack(func(){setGCPhase(_GCoff);gcSweep(work.mode)}):关闭写屏障,唤醒后台清扫任务,将在STW结束后开始运行
- gcSetTriggerRatio(nextTriggerRatio):更新下次触发gc时的heap大小
- systemstack(func() { startTheWorldWithSema(true) }): 停止STW
找出其中比较耗时的阶段:
- gcWakeAllAssists():O(G),将所有可运行的G插入到调度链表
- systemstack(func() {gcMark(startTime)}):
有疑问加站长微信联系(非本文作者)