case APOPW, APOPL, APOPQ,
		ALEAL, ALEAQ,
		AIMUL3W, AIMUL3L, AIMUL3Q,
		APEXTRB, APEXTRW, APEXTRD, APEXTRQ, AVPEXTRB, AVPEXTRW, AVPEXTRD, AVPEXTRQ, AEXTRACTPS,
		ABSFW, ABSFL, ABSFQ, ABSRW, ABSRL, ABSRQ, APOPCNTW, APOPCNTL, APOPCNTQ, ALZCNTW, ALZCNTL, ALZCNTQ,
		ASHLXL, ASHLXQ, ASHRXL, ASHRXQ, ASARXL, ASARXQ:
		// These instructions are pure writes to To. They don't use its old value.
		m &^= readTo
	case AXORL, AXORQ:

(BitLen64 x) => (SUB (MOVDconst [64]) (FLOGR x))

// POPCNT treats the input register as a vector of 8 bytes, producing
// a population count for each individual byte. For inputs larger than
// a single byte we therefore need to sum the individual bytes produced
// by the POPCNT instruction. For example, the following instruction
// sequence could be used to calculate the population count of a 4-byte

	AMOVDBR

	// conditional moves
	AMOVDEQ
	AMOVDGE
	AMOVDGT
	AMOVDLE
	AMOVDLT
	AMOVDNE
	ALOCR
	ALOCGR

	// find leftmost one
	AFLOGR

	// population count
	APOPCNT

	// integer bitwise
	AAND
	AANDW
	AOR
	AORW
	AXOR
	AXORW
	ASLW
	ASLD
	ASRW
	ASRAW
	ASRD
	ASRAD
	ARLL
	ARLLG
	ARNSBG
	ARXSBG
	AROSBG
	ARNSBGT

		{name: "POPCNTD", argLength: 1, reg: gp11, asm: "POPCNTD"}, // number of set bits in arg0
		{name: "POPCNTW", argLength: 1, reg: gp11, asm: "POPCNTW"}, // number of set bits in each word of arg0 placed in corresponding word
		{name: "POPCNTB", argLength: 1, reg: gp11, asm: "POPCNTB"}, // number of set bits in each byte of arg0 placed in corresponding byte

		{name: "I64Rotl", asm: "I64Rotl", argLength: 2, reg: gp21, typ: "Int64"},     // rotl(arg0, arg1)
		{name: "I64Popcnt", asm: "I64Popcnt", argLength: 1, reg: gp11, typ: "Int64"}, // popcnt(arg0)
	}

	archs = append(archs, arch{
		name:            "Wasm",
		pkg:             "cmd/internal/obj/wasm",
		genfile:         "../../wasm/ssa.go",
		ops:             WasmOps,
		blocks:          nil,

(Ctz64NonZero ...) => (Ctz64 ...)

(Ctz64 x) && buildcfg.GOPPC64<=8 => (POPCNTD (ANDN <typ.Int64> (ADDconst <typ.Int64> [-1] x) x))
(Ctz64 x) => (CNTTZD x)
(Ctz32 x) && buildcfg.GOPPC64<=8 => (POPCNTW (MOVWZreg (ANDN <typ.Int> (ADDconst <typ.Int> [-1] x) x)))
(Ctz32 x) => (CNTTZW (MOVWZreg x))
(Ctz16 x) => (POPCNTW (MOVHZreg (ANDN <typ.Int16> (ADDconst <typ.Int16> [-1] x) x)))
(Ctz8 x)  => (POPCNTB (MOVBZreg (ANDN <typ.UInt8> (ADDconst <typ.UInt8> [-1] x) x)))

(Bswap(64|32) ...) => (BSWAP(Q|L) ...)
(Bswap16 x) => (ROLWconst [8] x)

(PopCount(64|32) ...) => (POPCNT(Q|L) ...)
(PopCount16 x) => (POPCNTL (MOVWQZX <typ.UInt32> x))
(PopCount8 x) => (POPCNTL (MOVBQZX <typ.UInt32> x))

(Sqrt ...) => (SQRTSD ...)
(Sqrt32 ...) => (SQRTSS ...)

(RoundToEven x) => (ROUNDSD [0] x)
(Floor x)       => (ROUNDSD [1] x)

	{APMULLW, ymm, Py1, opBytes{0xd5, Pe, 0xd5}},
	{APMULULQ, ymm, Py1, opBytes{0xf4, Pe, 0xf4}},
	{APOPAL, ynone, P32, opBytes{0x61}},
	{APOPAW, ynone, Pe, opBytes{0x61}},
	{APOPCNTW, yml_rl, Pef3, opBytes{0xb8}},
	{APOPCNTL, yml_rl, Pf3, opBytes{0xb8}},
	{APOPCNTQ, yml_rl, Pfw, opBytes{0xb8}},
	{APOPFL, ynone, P32, opBytes{0x9d}},
	{APOPFQ, ynone, Py, opBytes{0x9d}},
	{APOPFW, ynone, Pe, opBytes{0x9d}},

		v.AddArg2(y, x)
		return true
	}
}
func rewriteValuePPC64_OpCtz16(v *Value) bool {
	v_0 := v.Args[0]
	b := v.Block
	typ := &b.Func.Config.Types
	// match: (Ctz16 x)
	// result: (POPCNTW (MOVHZreg (ANDN <typ.Int16> (ADDconst <typ.Int16> [-1] x) x)))
	for {
		x := v_0
		v.reset(OpPPC64POPCNTW)
		v0 := b.NewValue0(v.Pos, OpPPC64MOVHZreg, typ.Int64)
		v1 := b.NewValue0(v.Pos, OpPPC64ANDN, typ.Int16)

				{0, 1073733624}, // R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29
			},
		},
	},
	{
		name:   "POPCNTW",
		argLen: 1,
		asm:    ppc64.APOPCNTW,
		reg: regInfo{
			inputs: []inputInfo{
				{0, 1073733630}, // SP SB R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29
			},