default:
				return intelReg[a]
			}
		}
	}
	return strings.ToLower(arg.String())
}

var intelOp = map[Op]string{
	JAE:       "jnb",
	JA:        "jnbe",
	JGE:       "jnl",
	JNE:       "jnz",
	JG:        "jnle",
	JE:        "jz",
	SETAE:     "setnb",
	SETA:      "setnbe",
	SETGE:     "setnl",
	SETNE:     "setnz",
	SETG:      "setnle",
	SETE:      "setz",

	instructions["JNA"] = x86.AJLS  /* alternate */
	instructions["JNAE"] = x86.AJCS /* alternate */
	instructions["JNB"] = x86.AJCC  /* alternate */
	instructions["JNBE"] = x86.AJHI /* alternate */
	instructions["JNC"] = x86.AJCC  /* alternate */
	instructions["JNE"] = x86.AJNE  /* not equal (ZF = 0) */
	instructions["JNG"] = x86.AJLE  /* alternate */
	instructions["JNGE"] = x86.AJLT /* alternate */

	CMPQ	R8, $8
	JB	small

	CMPQ	R8, $63
	JBE	loop
#ifndef hasAVX2
	CMPB	internal∕cpu·X86+const_offsetX86HasAVX2(SB), $1
	JEQ     big_loop_avx2
	JMP	big_loop
#else
	JMP	big_loop_avx2
#endif
loop:
	CMPQ	R8, $16
	JBE	_0through16
	MOVOU	(SI), X0
	MOVOU	(DI), X1
	PCMPEQB X0, X1
	PMOVMSKB X1, AX
	XORQ	$0xffff, AX	// convert EQ to NE
	JNE	diff16	// branch if at least one byte is not equal
	ADDQ	$16, SI

	TESTL	BX, BX
	JEQ	_0
	CMPL	BX, $2
	JBE	_1or2
	CMPL	BX, $4
	JB	_3
	JE	_4
	CMPL	BX, $8
	JBE	_5through8
	CMPL	BX, $16
	JBE	_9through16
#ifdef GO386_softfloat
	JMP	nosse2
#endif
	PXOR	X0, X0
	CMPL	BX, $32
	JBE	_17through32
	CMPL	BX, $64
	JBE	_33through64
	CMPL	BX, $128
	JBE	_65through128
	CMPL	BX, $256
	JBE	_129through256

loop:
	MOVOU	X0, 0(DI)
	MOVOU	X0, 16(DI)

	MOVL	ptr+0(FP), DI
	MOVL	n+4(FP), BX
	XORL	AX, AX

tail:
	TESTL	BX, BX
	JEQ	_0
	CMPL	BX, $2
	JBE	_1or2
	CMPL	BX, $4
	JB	_3
	JE	_4
	CMPL	BX, $8
	JBE	_5through8
	CMPL	BX, $16
	JBE	_9through16
	MOVL	BX, CX
	SHRL	$2, CX
	REP
	STOSL
	ANDL	$3, BX
	JNE	tail
	RET

_1or2:
	MOVB	AX, (DI)
	MOVB	AX, -1(DI)(BX*1)
	RET
_0:
	RET
_3:
	MOVW	AX, (DI)

	TESTQ	BX, BX
	JEQ	_0
	CMPQ	BX, $2
	JBE	_1or2
	CMPQ	BX, $4
	JBE	_3or4
	CMPQ	BX, $8
	JB	_5through7
	JE	_8
	CMPQ	BX, $16
	JBE	_9through16
	CMPQ	BX, $32
	JBE	_17through32
	CMPQ	BX, $64
	JBE	_33through64
	CMPQ	BX, $128
	JBE	_65through128
	CMPQ	BX, $256
	JBE	_129through256

	CMPB	internal∕cpu·X86+const_offsetX86HasERMS(SB), $1 // enhanced REP MOVSB/STOSB
	JNE	skip_erms

	TESTL	BX, BX
	JEQ	move_0
	CMPL	BX, $2
	JBE	move_1or2
	CMPL	BX, $4
	JB	move_3
	JE	move_4
	CMPL	BX, $8
	JBE	move_5through8
	CMPL	BX, $16
	JBE	move_9through16
#ifdef GO386_softfloat
	JMP	nosse2
#endif
	CMPL	BX, $32
	JBE	move_17through32
	CMPL	BX, $64
	JBE	move_33through64
	CMPL	BX, $128
	JBE	move_65through128

nosse2:
/*
 * check and set for backwards
 */

	XORL	$0xffff, BX	// convert EQ to NE
	JNE	diff16	// branch if at least one byte is not equal
	ADDL	$16, SI
	ADDL	$16, DI
	SUBL	$16, BP
	JMP	largeloop

diff16:
	BSFL	BX, BX	// index of first byte that differs
	XORL	DX, DX
	MOVB	(SI)(BX*1), CX
	CMPB	CX, (DI)(BX*1)
	SETHI	DX
	LEAL	-1(DX*2), DX	// convert 1/0 to +1/-1
	MOVL	DX, (AX)
	RET

mediumloop:
	CMPL	BP, $4
	JBE	_0through4
	MOVL	(SI), BX
	MOVL	(DI), CX

	// CX = size (want in BX)
	CMPQ	AX, BX
	JNE	neq
	MOVQ	$1, AX	// return 1
	RET
neq:
	MOVQ	AX, SI
	MOVQ	BX, DI
	MOVQ	CX, BX
	JMP	memeqbody<>(SB)

// memequal_varlen(a, b unsafe.Pointer) bool
TEXT runtime·memequal_varlen<ABIInternal>(SB),NOSPLIT,$0-17
	// AX = a       (want in SI)
	// BX = b       (want in DI)
	// 8(DX) = size (want in BX)
	CMPQ	AX, BX
	JNE	neq
	MOVQ	$1, AX	// return 1
	RET
neq:

	SUBQ    $128, inl
	LEAQ    128(inp), inp

	MOVQ $8, itr1
	MOVQ $2, itr2

	CMPQ inl, $128
	JBE  sealAVX2Tail128
	CMPQ inl, $256
	JBE  sealAVX2Tail256
	CMPQ inl, $384
	JBE  sealAVX2Tail384
	CMPQ inl, $512
	JBE  sealAVX2Tail512

	// We have 448 bytes to hash, but main loop hashes 512 bytes at a time - perform some rounds, before the main loop