Robert Griesemer <******@****.***> 1606801129 -0800

	// s390x:`MOVHBR\t\(.*\), [R]`,`CMPW\t.*, [$]`
	return s == "xx"
}

func CompareString2(s string) bool {
	// amd64:`CMPL\t\(.*\), [$]`
	// arm64:`MOVWU\t\(.*\), [R]`,`CMPW\t.*, [R]`
	// ppc64le:`MOVWZ\t\(.*\), [R]`,`CMPW\t.*, [R]`
	// s390x:`MOVWBR\t\(.*\), [R]`,`CMPW\t.*, [$]`
	return s == "xxxx"
}

func CompareString3(s string) bool {
	// amd64:`CMPQ\t\(.*\), [A-Z]`
	// arm64:-`CMPW\t`

	RET

	PCALIGN	$16
cmp4:	// 4 - 7B
	CMP	R9,$4
	BLT	cmp2
	ANDCC	$3,R9,R9
	_LWBEX	(R0)(R5),R10
	_LWBEX	(R0)(R6),R11
	_LWBEX	(R9)(R5),R12
	_LWBEX	(R9)(R6),R14
	RLDIMI	$32,R10,$0,R12
	RLDIMI	$32,R11,$0,R14
	CMPU	R12,R14
	BR	cmp0

	PCALIGN $16
cmp2:	// 2 - 3B
	CMP	R9,$2
	BLT	cmp1
	ANDCC	$1,R9,R9
	_LHBEX	(R0)(R5),R10
	_LHBEX	(R0)(R6),R11
	_LHBEX	(R9)(R5),R12

	RET

cmp4:	// Length 4 - 7
	CMPU	R4,$4
	BLT	cmp2
	MOVD	$-4,R11
	ADD	$-4,R4,R4

	_LWBEX	(R0)(R3),R10
	_LWBEX	(R11)(R9),R11
	CMPB	R10,R5,R10
	CMPB	R11,R5,R11
	CNTLZW	R10,R10
	CNTLZW	R11,R11
	CMPU	R10,$32
	CMPU	R11,$32,CR1
	SRD	$3,R10,R3
	SRD	$3,R11,R11
	BNE	found

	ADD	R4,R11,R4
	MOVD	$-1,R3
	ISEL	CR1EQ,R3,R4,R3
	RET

cmp2:	// Length 2 - 3

			Goto("b1")),
		Bloc("b1",
			Valu("cmp1", OpLess64, c.config.Types.Bool, 0, nil, "arg1", "arg2"),
			If("cmp1", "b2", "b3")),
		Bloc("b2",
			Valu("cmp2", OpLess64, c.config.Types.Bool, 0, nil, "arg2", "arg3"),
			Goto("b3")),
		Bloc("b3",
			Valu("phi2", OpPhi, c.config.Types.Bool, 0, nil, "cmp1", "cmp2"),
			If("phi2", "b4", "b5")),
		Bloc("b4",
			Valu("cmp3", OpLess64, c.config.Types.Bool, 0, nil, "arg3", "arg1"),

loop3:
	MOVW (DI), SI
	CMPW SI,R8
	JZ   partial_success3
	ADDQ $1,DI
	CMPQ DI,DX
	JB loop3
	JMP fail
partial_success3:
	MOVW 1(DI), SI
	CMPW SI,BX
	JZ success
	ADDQ $1,DI
	CMPQ DI,DX
	JB loop3
	JMP fail
_4_or_more:
	CMPQ AX, $4
	JA   _5_or_more
	MOVL (R8), R8
	LEAQ -3(DI)(DX*1), DX
loop4:
	MOVL (DI), SI
	CMPL SI,R8
	JZ   success
	ADDQ $1,DI
	CMPQ DI,DX
	JB loop4
	JMP fail

	// Shuffle X0 around so that each byte contains
	// the character we're looking for.
	MOVD AX, X0
	PUNPCKLBW X0, X0
	PUNPCKLBW X0, X0
	PSHUFL $0, X0, X0

	CMPQ BX, $16
	JLT small

	MOVQ SI, DI

	CMPQ BX, $32
	JA avx2
sse:
	LEAQ	-16(SI)(BX*1), AX	// AX = address of last 16 bytes
	JMP	sseloopentry

	PCALIGN $16
sseloop:
	// Move the next 16-byte chunk of the data into X1.

var x32 [2]uint32
var x64 [2]uint64

func compMem1() int {
	// amd64:`CMPB\tcommand-line-arguments.x\+1\(SB\), [$]0`
	if x[1] {
		return 1
	}
	// amd64:`CMPB\tcommand-line-arguments.x8\+1\(SB\), [$]7`
	if x8[1] == 7 {
		return 1
	}
	// amd64:`CMPW\tcommand-line-arguments.x16\+2\(SB\), [$]7`
	if x16[1] == 7 {
		return 1
	}
	// amd64:`CMPL\tcommand-line-arguments.x32\+4\(SB\), [$]7`
	if x32[1] == 7 {
		return 1
	}

#include "textflag.h"

TEXT ·Count(SB),NOSPLIT,$0-40
#ifndef hasPOPCNT
	CMPB	internal∕cpu·X86+const_offsetX86HasPOPCNT(SB), $1
	JEQ	2(PC)
	JMP	·countGeneric(SB)
#endif
	MOVQ	b_base+0(FP), SI
	MOVQ	b_len+8(FP), BX
	MOVB	c+24(FP), AL
	LEAQ	ret+32(FP), R8
	JMP	countbody<>(SB)

TEXT ·CountString(SB),NOSPLIT,$0-32
#ifndef hasPOPCNT
	CMPB	internal∕cpu·X86+const_offsetX86HasPOPCNT(SB), $1
	JEQ	2(PC)

(CMP(Q|L|W|B)loadidx1 {sym} [off] ptr idx x mem) => (CMP(Q|L|W|B) (MOV(Q|L|W|B)loadidx1 {sym} [off] ptr idx mem) x)
(CMPQloadidx8 {sym} [off] ptr idx x mem) => (CMPQ (MOVQloadidx8 {sym} [off] ptr idx mem) x)
(CMPLloadidx4 {sym} [off] ptr idx x mem) => (CMPL (MOVLloadidx4 {sym} [off] ptr idx mem) x)
(CMPWloadidx2 {sym} [off] ptr idx x mem) => (CMPW (MOVWloadidx2 {sym} [off] ptr idx mem) x)