lock(&work.wbufSpans.lock)
	if gcphase != _GCoff || work.wbufSpans.free.isEmpty() {
		unlock(&work.wbufSpans.lock)
		return false
	}
	systemstack(func() {
		gp := getg().m.curg
		for i := 0; i < batchSize && !(preemptible && gp.preempt); i++ {
			span := work.wbufSpans.free.first
			if span == nil {
				break
			}
			work.wbufSpans.free.remove(span)

			// If this is a self-scan, put the user G in
			// _Gwaiting to prevent self-deadlock. It may
			// already be in _Gwaiting if this is a mark
			// worker or we're in mark termination.
			userG := getg().m.curg
			selfScan := gp == userG && readgstatus(userG) == _Grunning
			if selfScan {
				casGToWaitingForGC(userG, _Grunning, waitReasonGarbageCollectionScan)
			}

			// TODO: suspendG blocks (and spins) until gp

		// Examine the execution tracer's view of the CPU profile samples. Filter it
		// to only include samples from the single test goroutine. Use the goroutine
		// ID that was recorded in the events: that should reflect getg().m.curg,
		// same as the profiler's labels (even when the M is using its g0 stack).
		totalTraceSamples := 0
		traceSamples := 0
		traceStacks := make(map[string]int)
		r, err := trace.NewReader(bytes.NewReader(tb))

// See racecallback for command codes.
TEXT	runtime·racecallbackthunk(SB), NOSPLIT|NOFRAME, $0
	// Handle command raceGetProcCmd (0) here.
	// First, code below assumes that we are on curg, while raceGetProcCmd
	// can be executed on g0. Second, it is called frequently, so will
	// benefit from this fast path.
	MOVD	$0, R0		// clear R0 since we came from C code
	CMP	R3, $0
	BNE	rest

	MOVW	m_vdsoSP(R6), R8
	MOVW	R7, 4(R13)
	MOVW	R8, 8(R13)

	MOVW	$sp-4(FP), R7 // caller's SP
	MOVW	LR, m_vdsoPC(R6)
	MOVW	R7, m_vdsoSP(R6)

	MOVW	m_curg(R6), R7

	CMP	g, R7		// Only switch if on curg.
	B.NE	noswitch

	MOVW	m_g0(R6), R7
	MOVW	(g_sched+gobuf_sp)(R7), R13	 // Set SP to g0 stack

noswitch:
	BIC	$0x7, R13	// Align for C code

	// Store g on gsignal's stack, so if we receive a signal

		if gp.trace.statusWasTraced(gen) {
			return
		}
		// Scribble down information about this goroutine.
		ug := untracedG{gp: gp, mid: -1}
		systemstack(func() {
			me := getg().m.curg
			// We don't have to handle this G status transition because we
			// already eliminated ourselves from consideration above.
			casGToWaitingForGC(me, _Grunning, waitReasonTraceGoroutineStatus)

	semacquire1(addr, false, 0, 0, waitReasonSemacquire)
}

func semacquire1(addr *uint32, lifo bool, profile semaProfileFlags, skipframes int, reason waitReason) {
	gp := getg()
	if gp != gp.m.curg {
		throw("semacquire not on the G stack")
	}

	// Easy case.
	if cansemacquire(addr) {
		return
	}

	// Harder case:
	//	increment waiter count

	sigmask       sigset            // storage for saved signal mask
	tls           [tlsSlots]uintptr // thread-local storage (for x86 extern register)
	mstartfn      func()
	curg          *g       // current running goroutine
	caughtsig     guintptr // goroutine running during fatal signal
	p             puintptr // attached p for executing go code (nil if not executing go code)
	nextp         puintptr

}

// GoStart emits a GoStart event.
//
// Must be called with a valid P.
func (tl traceLocker) GoStart() {
	gp := getg().m.curg
	pp := gp.m.p
	w := tl.eventWriter(traceGoRunnable, traceProcRunning)
	w = w.write(traceEvGoStart, traceArg(gp.goid), gp.trace.nextSeq(tl.gen))
	if pp.ptr().gcMarkWorkerMode != gcMarkWorkerNotWorker {

func deductAssistCredit(size uintptr) *g {
	var assistG *g
	if gcBlackenEnabled != 0 {
		// Charge the current user G for this allocation.
		assistG = getg()
		if assistG.m.curg != nil {
			assistG = assistG.m.curg
		}
		// Charge the allocation against the G. We'll account
		// for internal fragmentation at the end of mallocgc.
		assistG.gcAssistBytes -= int64(size)

		if assistG.gcAssistBytes < 0 {