MOVQ msg_len+16(FP), R15

	MOVQ 0(DI), R8   // h0
	MOVQ 8(DI), R9   // h1
	MOVQ 16(DI), R10 // h2
	MOVQ 24(DI), R11 // r0
	MOVQ 32(DI), R12 // r1

	CMPQ R15, $16
	JB   bytes_between_0_and_15

loop:
	POLY1305_ADD(SI, R8, R9, R10)

multiply:
	POLY1305_MUL(R8, R9, R10, R11, R12, BX, CX, R13, R14)
	SUBQ $16, R15
	CMPQ R15, $16
	JAE  loop

bytes_between_0_and_15:
	TESTQ R15, R15
	JZ    done
	MOVQ  $1, BX

	print("rsp    ", hex(c.rsp()), "\n")
	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("rip    ", hex(c.rip()), "\n")
	print("rflags ", hex(c.rflags()), "\n")

	print("r7   ", hex(c.r7()), "\n")
	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("pc   ", hex(c.pc()), "\t")
	print("link ", hex(c.link()), "\n")
}

//go:nosplit

	print("r7   ", hex(c.r7()), "\n")
	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")

	ADD	$16, Rkeyp \
	BEQ	CR1, L_start10 \
	BEQ	CR2, L_start12 \
	LXVD2X	(R0+Rkeyp), V7 \
	LXVD2X	(R12+Rkeyp), V8 \
	ADD	$32, Rkeyp \
	L_start12: \
	LXVD2X	(R0+Rkeyp), V9 \
	LXVD2X	(R12+Rkeyp), V10 \
	ADD	$32, Rkeyp \
	L_start10: \
	LXVD2X	(R0+Rkeyp), V11 \
	LXVD2X	(R12+Rkeyp), V12 \
	LXVD2X	(R14+Rkeyp), V13 \
	LXVD2X	(R15+Rkeyp), V14 \
	LXVD2X	(R16+Rkeyp), V15 \
	LXVD2X	(R17+Rkeyp), V16 \

// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
// THE SOFTWARE.

#include "textflag.h"

#define TO	R8
#define TOE	R11
#define N	R12
#define TMP	R12				/* N and TMP don't overlap */

// See memclrNoHeapPointers Go doc for important implementation constraints.

// func memclrNoHeapPointers(ptr unsafe.Pointer, n uintptr)

func (c *sigctxt) r9() uint64  { return c.regs().r9 }
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

	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)

	MOVV R2, 16(R29)
	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)

chunk_loop:
	// chunk size is 16
	SGTU	$16, R3, R8
	BNE	R0, R8, chunk_tail_8
chunk_loop_1:
	MOVV	(R1), R6
	MOVV	(R2), R7
	BNE	R6, R7, not_eq
	MOVV	8(R1), R12
	MOVV	8(R2), R13
	ADDV	$16, R1
	ADDV	$16, R2
	SUBV	$16, R3
	BEQ	R12, R13, chunk_loop
	JMP	not_eq

chunk_tail_8:
	AND	$8, R3, R14
	BEQ	R0, R14, chunk_tail_4
	MOVV	(R1), R6
	MOVV	(R2), R7
	BNE	R6, R7, not_eq
	ADDV	$8, R1
	ADDV	$8, R2