EOR.S	R0<<28, R1, R2       // 002e31e0
	EOR.S	R0->28, R1, R2       // 402e31e0
	EOR.S	R0@>28, R1, R2       // 602e31e0
	EOR	R0<<28, R1           // 001e21e0
	EOR	R0>>28, R1           // 201e21e0
	EOR	R0->28, R1           // 401e21e0
	EOR	R0@>28, R1           // 601e21e0
	EOR.S	R0<<28, R1           // 001e31e0
	EOR.S	R0>>28, R1           // 201e31e0
	EOR.S	R0->28, R1           // 401e31e0
	EOR.S	R0@>28, R1           // 601e31e0

L8:	MOVL AX, DX		// w = w1
	MOVL -4(SI)(BX*4), AX	// w1 = x[i-1]
	SHLL CX, AX, DX		// w<<s | w1>>ŝ
	MOVL DX, (DI)(BX*4)	// z[i] = w<<s | w1>>ŝ
	SUBL $1, BX		// i--
	JG L8			// i > 0

	// i <= 0
X8a:	SHLL CX, AX		// w1<<s
	MOVL AX, (DI)		// z[0] = w1<<s
	RET

X8b:	MOVL $0, c+28(FP)
	RET


// func shrVU(z, x []Word, s uint) (c Word)
TEXT ·shrVU(SB),NOSPLIT,$0
	MOVL z_len+4(FP), BP
	SUBL $1, BP		// n--

	TEQ	R1, R4
	BNE L4a
	MOVW	R3, c+28(FP)
	RET
L4a:
	MOVW.P	4(R2), R5
	SUB.S	R3, R5
	MOVW.P	R5, 4(R1)
	B	E4
L4:
	MOVW.P	4(R2), R5
	SBC.S	$0, R5
	MOVW.P	R5, 4(R1)
E4:
	TEQ	R1, R4
	BNE	L4

	MOVW	$0, R0
	MOVW.CC	$1, R0
	MOVW	R0, c+28(FP)
	RET


// func shlVU(z, x []Word, s uint) (c Word)
TEXT ·shlVU(SB),NOSPLIT,$0
	MOVW	z_len+4(FP), R5
	TEQ	$0, R5
	BEQ	X7

L8:	MOVQ AX, DX		// w = w1
	MOVQ -8(R8)(BX*8), AX	// w1 = x[i-1]
	SHLQ CX, AX, DX		// w<<s | w1>>ŝ
	MOVQ DX, (R10)(BX*8)	// z[i] = w<<s | w1>>ŝ
	SUBQ $1, BX		// i--
	JG L8			// i > 0

	// i <= 0
X8a:	SHLQ CX, AX		// w1<<s
	MOVQ AX, (R10)		// z[0] = w1<<s
	RET

X8b:	MOVQ $0, c+56(FP)
	RET


// func shrVU(z, x []Word, s uint) (c Word)
TEXT ·shrVU(SB),NOSPLIT,$0
	MOVQ z_len+8(FP), R11

		STPW.P	(R6, R7), 8(R10)
		VMOV	R7, V2.S[0]
		WORD	$0x4E030042 //TBL	V3.B16, [V2.B16], V2.B16
		AESE	V0.B16, V2.B16
		EORW	R13, R0
		LSLW	$1, R13
		SUBS	$1, R8
		VMOV	V2.S[0], R9
		EORW	R9, R0
		EORW	R0, R1
		EORW	R1, R2
		EORW	R2, R3
		VMOV	R3, V2.S[0]
		WORD	$0x4E040042 //TBL	V3.B16, [V2.B16], V2.B16
		AESE	V0.B16, V2.B16
		VMOV	V2.S[0], R9
		EORW	R9, R4
		EORW	R4, R5
		EORW	R5, R6

	MOVD	x+24(FP), R2
	MOVD	s+48(FP), R3
	ADD	R1<<3, R0	// R0 = &z[n]
	ADD	R1<<3, R2	// R2 = &x[n]
	CBZ	R1, len0
	CBZ	R3, copy	// if the number of shift is 0, just copy x to z
	MOVD	$64, R4
	SUB	R3, R4
	// handling the most significant element x[n-1]
	MOVD.W	-8(R2), R6
	LSR	R4, R6, R5	// return value
	LSL	R3, R6, R8	// x[i] << s
	SUB	$1, R1
one:	TBZ	$0, R1, two
	MOVD.W	-8(R2), R6

// license that can be found in the LICENSE file.

//go:build race

// This file is here only to allow external functions.

	ROUND2(Rd, Re, Ra, Rb, Rc)
	ROUND2(Rc, Rd, Re, Ra, Rb)
	ROUND2(Rb, Rc, Rd, Re, Ra)
	SUB.S	$1, Rctr
	BNE	loop2

	MOVW	$0x8F1BBCDC, Rconst
	MOVW	$4, Rctr
loop3:	ROUND3(Ra, Rb, Rc, Rd, Re)
	ROUND3(Re, Ra, Rb, Rc, Rd)
	ROUND3(Rd, Re, Ra, Rb, Rc)
	ROUND3(Rc, Rd, Re, Ra, Rb)
	ROUND3(Rb, Rc, Rd, Re, Ra)
	SUB.S	$1, Rctr
	BNE	loop3

	MOVW	$0xCA62C1D6, Rconst
	MOVW	$4, Rctr

// license that can be found in the LICENSE file.

//go:build asan

#include "go_asm.h"
#include "go_tls.h"
#include "funcdata.h"
#include "textflag.h"

// This is like race_amd64.s, but for the asan calls.
// See race_amd64.s for detailed comments.

#ifdef GOOS_windows
#define RARG0 CX
#define RARG1 DX
#define RARG2 R8
#define RARG3 R9
#else
#define RARG0 DI
#define RARG1 SI
#define RARG2 DX

// Inferno's libkern/vlop-386.s
// https://bitbucket.org/inferno-os/inferno-os/src/master/libkern/vlop-386.s
//
//         Copyright © 1994-1999 Lucent Technologies Inc. All rights reserved.
//         Revisions Copyright © 2000-2007 Vita Nuova Holdings Limited (www.vitanuova.com).  All rights reserved.
//         Portions Copyright 2009 The Go Authors. All rights reserved.
//
// Permission is hereby granted, free of charge, to any person obtaining a copy