(Neg(32|16|8) ...) => (NEG ...)
(Neg(32|64)F ...) => (NEG(F|D) ...)

(Com(32|16|8) x) => (NORconst [0] x)

(Sqrt ...) => (SQRTD ...)
(Sqrt32 ...) => (SQRTF ...)

// TODO: optimize this case?
(Ctz32NonZero ...) => (Ctz32 ...)

// count trailing zero
// 32 - CLZ(x&-x - 1)
(Ctz32 <t> x) => (SUB (MOVWconst [32]) (CLZ <t> (SUBconst <t> [1] (AND <t> x (NEG <t> x)))))

// bit length

	{name: "Ctz8NonZero", argLength: 1},  // same as above, but arg[0] known to be non-zero, returns 0-7
	{name: "Ctz16NonZero", argLength: 1}, // same as above, but arg[0] known to be non-zero, returns 0-15
	{name: "Ctz32NonZero", argLength: 1}, // same as above, but arg[0] known to be non-zero, returns 0-31
	{name: "Ctz64NonZero", argLength: 1}, // same as above, but arg[0] known to be non-zero, returns 0-63

(Ctz8  x) => (BSFL (ORLconst <typ.UInt32> [1<<8 ] x))

(Ctz64NonZero x) && buildcfg.GOAMD64 >= 3 => (TZCNTQ x)
(Ctz32NonZero x) && buildcfg.GOAMD64 >= 3 => (TZCNTL x)
(Ctz16NonZero x) && buildcfg.GOAMD64 >= 3 => (TZCNTL x)
(Ctz8NonZero  x) && buildcfg.GOAMD64 >= 3 => (TZCNTL x)
(Ctz64NonZero x) && buildcfg.GOAMD64 <  3 => (Select0 (BSFQ x))
(Ctz32NonZero x) && buildcfg.GOAMD64 <  3 => (BSFL x)
(Ctz16NonZero x) && buildcfg.GOAMD64 <  3 => (BSFL x)

(MOVDaddr {sym} [n] p:(ADD x y)) && sym == nil && n == 0 => p
(MOVDaddr {sym} [n] ptr) && sym == nil && n == 0 && (ptr.Op == OpArgIntReg || ptr.Op == OpPhi) => ptr

// TODO: optimize these cases?
(Ctz32NonZero ...) => (Ctz32 ...)
(Ctz64NonZero ...) => (Ctz64 ...)

(Ctz64 x) && buildcfg.GOPPC64<=8 => (POPCNTD (ANDN <typ.Int64> (ADDconst <typ.Int64> [-1] x) x))
(Ctz64 x) => (CNTTZD x)

(OffPtr [off] ptr) && is32Bit(off) => (ADDconst [int32(off)] ptr)
(OffPtr [off] ptr) => (ADD (MOVDconst [off]) ptr)

// TODO: optimize these cases?
(Ctz64NonZero ...) => (Ctz64 ...)
(Ctz32NonZero ...) => (Ctz32 ...)

// Ctz(x) = 64 - findLeftmostOne((x-1)&^x)
(Ctz64 <t> x) => (SUB (MOVDconst [64]) (FLOGR (AND <t> (SUBconst <t> [1] x) (NOT <t> x))))

(Neg(32|64)F ...) => (NEG(F|D) ...)

(Com(32|16|8) ...) => (MVN ...)

(Sqrt ...) => (SQRTD ...)
(Sqrt32 ...) => (SQRTF ...)
(Abs ...) => (ABSD ...)

// TODO: optimize this for ARMv5 and ARMv6
(Ctz32NonZero ...) => (Ctz32 ...)
(Ctz16NonZero ...) => (Ctz32 ...)
(Ctz8NonZero ...) => (Ctz32 ...)

// count trailing zero for ARMv5 and ARMv6
// 32 - CLZ(x&-x - 1)
(Ctz32 <t> x) && buildcfg.GOARM.Version<=6 =>

func rewriteValueAMD64_OpCtz32NonZero(v *Value) bool {
	v_0 := v.Args[0]
	// match: (Ctz32NonZero x)
	// cond: buildcfg.GOAMD64 >= 3
	// result: (TZCNTL x)
	for {
		x := v_0
		if !(buildcfg.GOAMD64 >= 3) {
			break
		}
		v.reset(OpAMD64TZCNTL)
		v.AddArg(x)
		return true
	}
	// match: (Ctz32NonZero x)
	// cond: buildcfg.GOAMD64 < 3
	// result: (BSFL x)
	for {

	},
	{
		name:    "Ctz8NonZero",
		argLen:  1,
		generic: true,
	},
	{
		name:    "Ctz16NonZero",
		argLen:  1,
		generic: true,
	},
	{
		name:    "Ctz32NonZero",
		argLen:  1,
		generic: true,
	},
	{
		name:    "Ctz64NonZero",
		argLen:  1,
		generic: true,
	},
	{
		name:    "BitLen8",
		argLen:  1,
		generic: true,