MOVQ	a1+8(FP), DI
	MOVQ	a2+16(FP), SI
	MOVQ	a3+24(FP), DX
	MOVQ	a4+32(FP), R10
	MOVQ	a5+40(FP), R8
	MOVQ	a6+48(FP), R9
	MOVQ	a7+56(FP), R11
	MOVQ	a8+64(FP), R12
	MOVQ	a9+72(FP), R13
	SUBQ	$32, SP
	MOVQ	R11, 8(SP)
	MOVQ	R12, 16(SP)
	MOVQ	R13, 24(SP)
	ADDQ	$0x2000000, AX
	SYSCALL
	JCC	ok9
	ADDQ	$32, SP
	MOVQ	$-1, r1+80(FP)
	MOVQ	$0, r2+88(FP)
	MOVQ	AX, err+96(FP)
	CALL	runtime·exitsyscall<ABIInternal>(SB)
	RET

	MOVV R3, 24(R29)
	MOVV R4, 32(R29)
	MOVV R5, 40(R29)
	MOVV R6, 48(R29)
	MOVV R7, 56(R29)
	MOVV R8, 64(R29)
	MOVV R9, 72(R29)
	MOVV R10, 80(R29)
	MOVV R11, 88(R29)
	MOVV R12, 96(R29)
	MOVV R13, 104(R29)
	MOVV R14, 112(R29)
	MOVV R15, 120(R29)
	MOVV R16, 128(R29)
	MOVV R17, 136(R29)
	MOVV R18, 144(R29)
	MOVV R19, 152(R29)
	MOVV R20, 160(R29)
	MOVV R21, 168(R29)
	MOVV R22, 176(R29)

func (c *sigctxt) r10() uint64 { return c.regs().r10 }
func (c *sigctxt) r11() uint64 { return c.regs().r11 }
func (c *sigctxt) r12() uint64 { return c.regs().r12 }
func (c *sigctxt) r13() uint64 { return c.regs().r13 }
func (c *sigctxt) r14() uint64 { return c.regs().r14 }
func (c *sigctxt) r15() uint64 { return c.regs().r15 }

//go:nosplit
//go:nowritebarrierrec
func (c *sigctxt) rip() uint64 { return c.regs().rip }

	MOVV R1, -480(R3)
	SUBV $480, R3
	MOVV R4, 8(R3)
	MOVV R5, 16(R3)
	MOVV R6, 24(R3)
	MOVV R7, 32(R3)
	MOVV R8, 40(R3)
	MOVV R9, 48(R3)
	MOVV R10, 56(R3)
	MOVV R11, 64(R3)
	MOVV R12, 72(R3)
	MOVV R13, 80(R3)
	MOVV R14, 88(R3)
	MOVV R15, 96(R3)
	MOVV R16, 104(R3)
	MOVV R17, 112(R3)
	MOVV R18, 120(R3)
	MOVV R19, 128(R3)
	MOVV R20, 136(R3)
	MOVV R21, 144(R3)
	MOVV R23, 152(R3)
	MOVV R24, 160(R3)

	MOVW R3, 12(R29)
	MOVW R4, 16(R29)
	MOVW R5, 20(R29)
	MOVW R6, 24(R29)
	MOVW R7, 28(R29)
	MOVW R8, 32(R29)
	MOVW R9, 36(R29)
	MOVW R10, 40(R29)
	MOVW R11, 44(R29)
	MOVW R12, 48(R29)
	MOVW R13, 52(R29)
	MOVW R14, 56(R29)
	MOVW R15, 60(R29)
	MOVW R16, 64(R29)
	MOVW R17, 68(R29)
	MOVW R18, 72(R29)
	MOVW R19, 76(R29)
	MOVW R20, 80(R29)
	MOVW R21, 84(R29)
	MOVW R22, 88(R29)
	MOVW R24, 92(R29)

	print("r8      ", hex(c.r8()), "\n")
	print("r9      ", hex(c.r9()), "\n")
	print("r10     ", hex(c.r10()), "\n")
	print("r11     ", hex(c.r11()), "\n")
	print("r12     ", hex(c.r12()), "\n")
	print("r13     ", hex(c.r13()), "\n")
	print("r14     ", hex(c.r14()), "\n")
	print("r15     ", hex(c.r15()), "\n")
	print("r16     ", hex(c.r16()), "\n")
	print("r17     ", hex(c.r17()), "\n")
	print("r18     ", hex(c.r18()), "\n")

	MOVL $495, AX		// lwp_create
	SYSCALL
	MOVL	AX, ret+8(FP)
	RET

TEXT runtime·lwp_start(SB),NOSPLIT,$0
	MOVQ	DI, R13 // m

	// set up FS to point at m->tls
	LEAQ	m_tls(R13), DI
	CALL	runtime·settls(SB)	// smashes DI

	// set up m, g
	get_tls(CX)
	MOVQ	m_g0(R13), DI
	MOVQ	R13, g_m(DI)
	MOVQ	DI, g(CX)

	CALL	runtime·stackcheck(SB)
	CALL	runtime·mstart(SB)

	MOVQ 0, AX			// crash (not reached)

	return ptracePtr(PTRACE_SETREGS, pid, 0, unsafe.Pointer(regs))
}

// PtraceRegsAmd64 is the registers used by amd64 binaries.
type PtraceRegsAmd64 struct {
	R15      uint64
	R14      uint64
	R13      uint64
	R12      uint64
	Rbp      uint64
	Rbx      uint64
	R11      uint64
	R10      uint64
	R9       uint64
	R8       uint64
	Rax      uint64
	Rcx      uint64
	Rdx      uint64
	Rsi      uint64

	print("r8   ", hex(c.r8()), "\t")
	print("r9   ", hex(c.r9()), "\n")
	print("r10  ", hex(c.r10()), "\t")
	print("r11  ", hex(c.r11()), "\n")
	print("r12  ", hex(c.r12()), "\t")
	print("r13  ", hex(c.r13()), "\n")
	print("r14  ", hex(c.r14()), "\t")
	print("r15  ", hex(c.r15()), "\n")
	print("r16  ", hex(c.r16()), "\t")
	print("r17  ", hex(c.r17()), "\n")
	print("r18  ", hex(c.r18()), "\t")

	MOVD	dig+0(FP), R0
	MOVD	p+8(FP), R1
	MOVD	p_len+16(FP), R2
	AND	$~63, R2
	CBZ	R2, zero

	ADD	R1, R2, R21
	LDPW	(0*8)(R0), (R4, R5)
	LDPW	(1*8)(R0), (R6, R7)

loop:
	MOVW	R4, R12
	MOVW	R5, R13
	MOVW	R6, R14
	MOVW	R7, R15

	MOVW	(0*4)(R1), R8
	MOVW	R7, R9

#define ROUND1(a, b, c, d, index, const, shift) \
	ADDW	$const, a; \
	ADDW	R8, a; \
	MOVW	(index*4)(R1), R8; \
	EORW	c, R9; \