MOVQ	s1+16(FP), RARG1
	MOVQ	s2+24(FP), RARG2
	MOVQ	result+32(FP), RARG3

	get_tls(R12)
	MOVQ	g(R12), R14
	MOVQ	g_m(R14), R13

	// Switch to g0 stack.
	MOVQ	SP, R12		// callee-saved, preserved across the CALL
	MOVQ	m_g0(R13), R10
	CMPQ	R10, R14
	JE	call	// already on g0
	MOVQ	(g_sched+gobuf_sp)(R10), SP
call:
	ANDQ	$~15, SP	// alignment for gcc ABI
	CALL	AX
	MOVQ	R12, SP
	RET

	// will not be usable.
	MOVW	g_m(g), R8
	MOVW	m_g0(R8), R3
	MOVW	R13, (g_sched+gobuf_sp)(R3)

havem:
	// Now there's a valid m, and we're running on its m->g0.
	// Save current m->g0->sched.sp on stack and then set it to SP.
	// Save current sp in m->g0->sched.sp in preparation for
	// switch back to m->curg stack.
	// NOTE: unwindm knows that the saved g->sched.sp is at 4(R13) aka savedsp-12(SP).
	MOVW	m_g0(R8), R3

	JMP	racecall<>(SB)

// Switches SP to g0 stack and calls R1. Arguments are already set.
TEXT	racecall<>(SB), NOSPLIT, $0-0
	BL	runtime·save_g(SB)		// Save g for callbacks.
	MOVD	R15, R7				// Save SP.
	MOVD	g_m(g), R8			// R8 = thread.
	MOVD	m_g0(R8), R8			// R8 = g0.
	CMPBEQ	R8, g, call			// Already on g0?
	MOVD	(g_sched+gobuf_sp)(R8), R15	// Switch SP to g0.
call:	SUB	$160, R15			// Allocate C frame.

	MOVQ	arg3+32(FP), RARG3
	JMP	racecall<>(SB)

// Switches SP to g0 stack and calls (AX). Arguments already set.
TEXT	racecall<>(SB), NOSPLIT|NOFRAME, $0-0
	MOVQ	g_m(R14), R13
	// Switch to g0 stack.
	MOVQ	SP, R12		// callee-saved, preserved across the CALL
	MOVQ	m_g0(R13), R10
	CMPQ	R10, R14
	JE	call	// already on g0
	MOVQ	(g_sched+gobuf_sp)(R10), SP
call:
	ANDQ	$~15, SP	// alignment for gcc ABI

	JMP	msancall<>(SB)

// Switches SP to g0 stack and calls (AX). Arguments already set.
TEXT	msancall<>(SB), NOSPLIT, $0-0
	get_tls(R12)
	MOVQ	g(R12), R14
	MOVQ	SP, R12		// callee-saved, preserved across the CALL
	CMPQ	R14, $0
	JE	call	// no g; still on a system stack

	MOVQ	g_m(R14), R13
	// Switch to g0 stack.
	MOVQ	m_g0(R13), R10
	CMPQ	R10, R14
	JE	call	// already on g0

	MOVQ	(g_sched+gobuf_sp)(R10), SP
call:

	// Use all 16 registers to do float64 --> float32 conversion.
	g0, g1, g2, g3, g4, g5, g6, g7, g8, g9, g10, g11, g12, g13, g14, g15 :=

{
    "formatVersion": "1.0",
    "variants": [
        {
            "name": "api",
            "dependencies": [
                {
                    "group": "g0",
                    "module": "m0"
                },
                {
                    "group": "g1",
                    "module": "m1",
                    "version": {
                        "prefers": "v1"
                    }
                },

	ADD	$const_stackGuard, R3
	MOVD	R3, g_stackguard0(g)
	MOVD	R3, g_stackguard1(g)

	// set the per-goroutine and per-mach "registers"
	MOVD	$runtime·m0(SB), R3

	// save m->g0 = g0
	MOVD	g, m_g0(R3)
	// save m0 to g0->m
	MOVD	R3, g_m(g)

	BL	runtime·check(SB)

	// args are already prepared
	BL	runtime·args(SB)
	BL	runtime·osinit(SB)
	BL	runtime·schedinit(SB)

func notetsleep(n *note, ns int64) bool {
	throw("notetsleep not supported by js")
	return false
}

// same as runtime·notetsleep, but called on user g (not g0)
func notetsleepg(n *note, ns int64) bool {
	gp := getg()
	if gp == gp.m.g0 {
		throw("notetsleepg on g0")
	}

	if ns >= 0 {
		deadline := nanotime() + ns
		delay := ns/1000000 + 1 // round up
		if delay > 1<<31-1 {

	MOVD    runtime·tls_g(SB), R10	// g offset in TLS
	MOVD    0(R10), g
	MOVD	g_m(g), R7		// m for g
	MOVD	R1, R16			// callee-saved, preserved across C call
	MOVD	m_g0(R7), R10		// g0 for m
	CMP	R10, g			// same g0?
	BEQ	call			// already on g0
	MOVD	(g_sched+gobuf_sp)(R10), R1 // switch R1
call:
	// prepare frame for C ABI
	SUB	$32, R1			// create frame for callee saving LR, CR, R2 etc.