MOVV	R4, name(SB)		// 1e00001ac403c029
	MOVB	R4, name(SB)		// 1e00001ac4030029
	MOVBU	R4, name(SB)		// 1e00001ac4030029
	MOVF	F4, name(SB)		// 1e00001ac403402b
	MOVD	F4, name(SB)		// 1e00001ac403c02b
	MOVW	name(SB), R4		// 1e00001ac4038028
	MOVWU	name(SB), R4		// 1e00001ac403802a
	MOVV	name(SB), R4		// 1e00001ac403c028
	MOVB	name(SB), R4		// 1e00001ac4030028
	MOVBU	name(SB), R4		// 1e00001ac403002a

	MOVWU	R4, result+16(FP)	// 64608029
	MOVV	R4, result+16(FP)	// 6460c029
	MOVB	R4, result+16(FP)	// 64600029
	MOVBU	R4, result+16(FP)	// 64600029
	MOVWL	R4, result+16(FP)	// 6460002f
	MOVVL	R4, result+16(FP)	// 6460802f
	MOVW	R4, 1(R5)		// a4048029
	MOVWU	R4, 1(R5)		// a4048029
	MOVV	R4, 1(R5)		// a404c029
	MOVB	R4, 1(R5)		// a4040029
	MOVBU	R4, 1(R5)		// a4040029
	MOVWL	R4, 1(R5)		// a404002f
	MOVVL	R4, 1(R5)		// a404802f

	RET

TEXT ·Store8(SB), NOSPLIT, $0-9
	MOVV	ptr+0(FP), R1
	MOVB	val+8(FP), R2
	SYNC
	MOVB	R2, 0(R1)
	SYNC
	RET

TEXT ·Store64(SB), NOSPLIT, $0-16
	MOVV	ptr+0(FP), R1
	MOVV	val+8(FP), R2
	SYNC
	MOVV	R2, 0(R1)
	SYNC
	RET

// void	Or8(byte volatile*, byte);
TEXT ·Or8(SB), NOSPLIT, $0-9
	MOVV	ptr+0(FP), R1
	MOVBU	val+8(FP), R2

	"CMOVZ",
	"CMPEQD",
	"CMPEQF",
	"CMPGED",
	"CMPGEF",
	"CMPGTD",
	"CMPGTF",
	"DIV",
	"DIVD",
	"DIVF",
	"DIVU",
	"DIVW",
	"GOK",
	"LL",
	"LLV",
	"LUI",
	"MADD",
	"MOVB",
	"MOVBU",
	"MOVD",
	"MOVDF",
	"MOVDW",
	"MOVF",
	"MOVFD",
	"MOVFW",
	"MOVH",
	"MOVHU",
	"MOVW",
	"MOVWD",
	"MOVWF",
	"MOVWL",
	"MOVWR",
	"MSUB",
	"MUL",
	"MULD",
	"MULF",

// func And8(ptr *uint8, val uint8)
TEXT ·And8(SB), NOSPLIT, $0-9
	MOV	ptr+0(FP), A0
	MOVBU	val+8(FP), A1
	AND	$3, A0, A2
	AND	$-4, A0
	SLL	$3, A2
	XOR	$255, A1
	SLL	A2, A1
	XOR	$-1, A1
	AMOANDW A1, (A0), ZERO
	RET

// func Or8(ptr *uint8, val uint8)
TEXT ·Or8(SB), NOSPLIT, $0-9
	MOV	ptr+0(FP), A0
	MOVBU	val+8(FP), A1
	AND	$3, A0, A2
	AND	$-4, A0
	SLL	$3, A2
	SLL	A2, A1

	MOV	$1234, 8(SP)			// ERROR "constant load must target register"
	MOVB	$1, X5				// ERROR "unsupported constant load"
	MOVH	$1, X5				// ERROR "unsupported constant load"
	MOVW	$1, X5				// ERROR "unsupported constant load"
	MOVF	$1, X5				// ERROR "unsupported constant load"
	MOVBU	X5, (X6)			// ERROR "unsupported unsigned store"
	MOVHU	X5, (X6)			// ERROR "unsupported unsigned store"

	"LU32ID",
	"LU52ID",
	"PCALAU12I",
	"PCADDU12I",
	"JIRL",
	"BGE",
	"BLT",
	"BLTU",
	"BGEU",
	"DIV",
	"DIVD",
	"DIVF",
	"DIVU",
	"DIVW",
	"LL",
	"LLV",
	"LUI",
	"MOVB",
	"MOVBU",
	"MOVD",
	"MOVDF",
	"MOVDW",
	"MOVF",
	"MOVFD",
	"MOVFW",
	"MOVH",
	"MOVHU",
	"MOVW",
	"MOVWD",
	"MOVWF",
	"MOVWL",
	"MOVWR",
	"MUL",
	"MULD",
	"MULF",
	"MULU",

	MOVD.P -8(R10), R8         <=>      ldr x8, [x10],#-8
	MOVB.W 16(R16), R10        <=>      ldrsb x10, [x16,#16]!
	MOVBU.W 16(R16), R10       <=>      ldrb x10, [x16,#16]!

3. Go uses a series of MOV instructions as load and store.

64-bit variant ldr, str, stur => MOVD;
32-bit variant str, stur, ldrsw => MOVW;
32-bit variant ldr => MOVWU;
ldrb => MOVBU; ldrh => MOVHU;
ldrsb, sturb, strb => MOVB;
ldrsh, sturh, strh =>  MOVH.

TEXT ·Cas(SB), NOSPLIT, $0-17
	MOVD	ptr+0(FP), R0
	MOVW	old+8(FP), R1
	MOVW	new+12(FP), R2
#ifndef GOARM64_LSE
	MOVBU	internal∕cpu·ARM64+const_offsetARM64HasATOMICS(SB), R4
	CBZ 	R4, load_store_loop
#endif
	MOVD	R1, R3
	CASALW	R3, (R0), R2
	CMP 	R1, R3
	CSET	EQ, R0
	MOVB	R0, ret+16(FP)
	RET
#ifndef GOARM64_LSE
load_store_loop:
	LDAXRW	(R0), R3
	CMPW	R1, R3
	BNE	ok
	STLXRW	R2, (R0), R3

	BLT	X5, X6, check4_unaligned
compare8_unaligned:
	MOVBU	0(X10), X8
	MOVBU	1(X10), X15
	MOVBU	2(X10), X17
	MOVBU	3(X10), X19
	MOVBU	4(X10), X21
	MOVBU	5(X10), X23
	MOVBU	6(X10), X25
	MOVBU	7(X10), X29
	MOVBU	0(X12), X9
	MOVBU	1(X12), X16
	MOVBU	2(X12), X18
	MOVBU	3(X12), X20
	MOVBU	4(X12), X22
	MOVBU	5(X12), X24
	MOVBU	6(X12), X28
	MOVBU	7(X12), X30
	BNE	X8, X9, cmp1a
	BNE	X15, X16, cmp1b