[dev.ssa] cmd/compile: initial ARM64 SSA port

Mostly copied from ARM port, with instruction names and Prog fields
adjusted, and 64-bit int ops added. Not complete.

Fib compiles and runs correctly.

Change-Id: Id3ecb0d4b571200a035344b3e8e4408769f76221
Reviewed-on: https://go-review.googlesource.com/25130
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

7 files changed, 7551 insertions, 0 deletions

diff --git a/src/cmd/compile/internal/arm64/galign.go b/src/cmd/compile/internal/arm64/galign.go
index 7acc4e08eb..12f9b0515f 100644
--- a/src/cmd/compile/internal/arm64/galign.go
+++ b/src/cmd/compile/internal/arm64/galign.go
@@ -6,6 +6,7 @@ package arm64
 
 import (
 	"cmd/compile/internal/gc"
+	"cmd/compile/internal/ssa"
 	"cmd/internal/obj/arm64"
 )
 
@@ -61,6 +62,11 @@ func Main() {
 	gc.Thearch.Doregbits = doregbits
 	gc.Thearch.Regnames = regnames
 
+	gc.Thearch.SSARegToReg = ssaRegToReg
+	gc.Thearch.SSAMarkMoves = func(s *gc.SSAGenState, b *ssa.Block) {}
+	gc.Thearch.SSAGenValue = ssaGenValue
+	gc.Thearch.SSAGenBlock = ssaGenBlock
+
 	gc.Main()
 	gc.Exit(0)
 }
diff --git a/src/cmd/compile/internal/arm64/ssa.go b/src/cmd/compile/internal/arm64/ssa.go
new file mode 100644
index 0000000000..90dab7f4dc
--- /dev/null
+++ b/src/cmd/compile/internal/arm64/ssa.go
@@ -0,0 +1,628 @@
+// Copyright 2016 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package arm64
+
+import (
+	"math"
+
+	"cmd/compile/internal/gc"
+	"cmd/compile/internal/ssa"
+	"cmd/internal/obj"
+	"cmd/internal/obj/arm64"
+)
+
+var ssaRegToReg = []int16{
+	arm64.REG_R0,
+	arm64.REG_R1,
+	arm64.REG_R2,
+	arm64.REG_R3,
+	arm64.REG_R4,
+	arm64.REG_R5,
+	arm64.REG_R6,
+	arm64.REG_R7,
+	arm64.REG_R8,
+	arm64.REG_R9,
+	arm64.REG_R10,
+	arm64.REG_R11,
+	arm64.REG_R12,
+	arm64.REG_R13,
+	arm64.REG_R14,
+	arm64.REG_R15,
+	arm64.REG_R16,
+	arm64.REG_R17,
+	arm64.REG_R18,
+	arm64.REG_R19,
+	arm64.REG_R20,
+	arm64.REG_R21,
+	arm64.REG_R22,
+	arm64.REG_R23,
+	arm64.REG_R24,
+	arm64.REG_R25,
+	arm64.REG_R26,
+	// R27 = REGTMP not used in regalloc
+	arm64.REGG, // R28
+	arm64.REG_R29,
+	// R30 = REGLINK not used in regalloc
+	arm64.REGSP, // R31
+
+	arm64.REG_F0,
+	arm64.REG_F1,
+	arm64.REG_F2,
+	arm64.REG_F3,
+	arm64.REG_F4,
+	arm64.REG_F5,
+	arm64.REG_F6,
+	arm64.REG_F7,
+	arm64.REG_F8,
+	arm64.REG_F9,
+	arm64.REG_F10,
+	arm64.REG_F11,
+	arm64.REG_F12,
+	arm64.REG_F13,
+	arm64.REG_F14,
+	arm64.REG_F15,
+	arm64.REG_F16,
+	arm64.REG_F17,
+	arm64.REG_F18,
+	arm64.REG_F19,
+	arm64.REG_F20,
+	arm64.REG_F21,
+	arm64.REG_F22,
+	arm64.REG_F23,
+	arm64.REG_F24,
+	arm64.REG_F25,
+	arm64.REG_F26,
+	arm64.REG_F27,
+	arm64.REG_F28,
+	arm64.REG_F29,
+	arm64.REG_F30,
+	arm64.REG_F31,
+
+	arm64.REG_NZCV, // flag
+	0,              // SB isn't a real register.  We fill an Addr.Reg field with 0 in this case.
+}
+
+// Smallest possible faulting page at address zero,
+// see ../../../../runtime/mheap.go:/minPhysPageSize
+const minZeroPage = 4096
+
+// loadByType returns the load instruction of the given type.
+func loadByType(t ssa.Type) obj.As {
+	if t.IsFloat() {
+		switch t.Size() {
+		case 4:
+			return arm64.AFMOVS
+		case 8:
+			return arm64.AFMOVD
+		}
+	} else {
+		switch t.Size() {
+		case 1:
+			if t.IsSigned() {
+				return arm64.AMOVB
+			} else {
+				return arm64.AMOVBU
+			}
+		case 2:
+			if t.IsSigned() {
+				return arm64.AMOVH
+			} else {
+				return arm64.AMOVHU
+			}
+		case 4:
+			if t.IsSigned() {
+				return arm64.AMOVW
+			} else {
+				return arm64.AMOVWU
+			}
+		case 8:
+			return arm64.AMOVD
+		}
+	}
+	panic("bad load type")
+}
+
+// storeByType returns the store instruction of the given type.
+func storeByType(t ssa.Type) obj.As {
+	if t.IsFloat() {
+		switch t.Size() {
+		case 4:
+			return arm64.AFMOVS
+		case 8:
+			return arm64.AFMOVD
+		}
+	} else {
+		switch t.Size() {
+		case 1:
+			return arm64.AMOVB
+		case 2:
+			return arm64.AMOVH
+		case 4:
+			return arm64.AMOVW
+		case 8:
+			return arm64.AMOVD
+		}
+	}
+	panic("bad store type")
+}
+
+func ssaGenValue(s *gc.SSAGenState, v *ssa.Value) {
+	s.SetLineno(v.Line)
+	switch v.Op {
+	case ssa.OpInitMem:
+		// memory arg needs no code
+	case ssa.OpArg:
+		// input args need no code
+	case ssa.OpSP, ssa.OpSB, ssa.OpGetG:
+		// nothing to do
+	case ssa.OpCopy, ssa.OpARM64MOVDconvert, ssa.OpARM64MOVDreg:
+		if v.Type.IsMemory() {
+			return
+		}
+		x := gc.SSARegNum(v.Args[0])
+		y := gc.SSARegNum(v)
+		if x == y {
+			return
+		}
+		as := arm64.AMOVD
+		if v.Type.IsFloat() {
+			switch v.Type.Size() {
+			case 4:
+				as = arm64.AFMOVS
+			case 8:
+				as = arm64.AFMOVD
+			default:
+				panic("bad float size")
+			}
+		}
+		p := gc.Prog(as)
+		p.From.Type = obj.TYPE_REG
+		p.From.Reg = x
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = y
+	case ssa.OpLoadReg:
+		if v.Type.IsFlags() {
+			v.Unimplementedf("load flags not implemented: %v", v.LongString())
+			return
+		}
+		p := gc.Prog(loadByType(v.Type))
+		n, off := gc.AutoVar(v.Args[0])
+		p.From.Type = obj.TYPE_MEM
+		p.From.Node = n
+		p.From.Sym = gc.Linksym(n.Sym)
+		p.From.Offset = off
+		if n.Class == gc.PPARAM || n.Class == gc.PPARAMOUT {
+			p.From.Name = obj.NAME_PARAM
+			p.From.Offset += n.Xoffset
+		} else {
+			p.From.Name = obj.NAME_AUTO
+		}
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = gc.SSARegNum(v)
+	case ssa.OpPhi:
+		gc.CheckLoweredPhi(v)
+	case ssa.OpStoreReg:
+		if v.Type.IsFlags() {
+			v.Unimplementedf("store flags not implemented: %v", v.LongString())
+			return
+		}
+		p := gc.Prog(storeByType(v.Type))
+		p.From.Type = obj.TYPE_REG
+		p.From.Reg = gc.SSARegNum(v.Args[0])
+		n, off := gc.AutoVar(v)
+		p.To.Type = obj.TYPE_MEM
+		p.To.Node = n
+		p.To.Sym = gc.Linksym(n.Sym)
+		p.To.Offset = off
+		if n.Class == gc.PPARAM || n.Class == gc.PPARAMOUT {
+			p.To.Name = obj.NAME_PARAM
+			p.To.Offset += n.Xoffset
+		} else {
+			p.To.Name = obj.NAME_AUTO
+		}
+	case ssa.OpARM64ADD,
+		ssa.OpARM64SUB,
+		ssa.OpARM64AND,
+		ssa.OpARM64OR,
+		ssa.OpARM64XOR,
+		ssa.OpARM64BIC,
+		ssa.OpARM64MUL,
+		ssa.OpARM64DIV,
+		ssa.OpARM64UDIV,
+		ssa.OpARM64DIVW,
+		ssa.OpARM64UDIVW,
+		ssa.OpARM64MOD,
+		ssa.OpARM64UMOD,
+		ssa.OpARM64MODW,
+		ssa.OpARM64UMODW,
+		ssa.OpARM64FADDS,
+		ssa.OpARM64FADDD,
+		ssa.OpARM64FSUBS,
+		ssa.OpARM64FSUBD,
+		ssa.OpARM64FMULS,
+		ssa.OpARM64FMULD,
+		ssa.OpARM64FDIVS,
+		ssa.OpARM64FDIVD:
+		r := gc.SSARegNum(v)
+		r1 := gc.SSARegNum(v.Args[0])
+		r2 := gc.SSARegNum(v.Args[1])
+		p := gc.Prog(v.Op.Asm())
+		p.From.Type = obj.TYPE_REG
+		p.From.Reg = r2
+		p.Reg = r1
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = r
+	case ssa.OpARM64ADDconst,
+		ssa.OpARM64SUBconst,
+		ssa.OpARM64ANDconst,
+		ssa.OpARM64ORconst,
+		ssa.OpARM64XORconst,
+		ssa.OpARM64BICconst:
+		p := gc.Prog(v.Op.Asm())
+		p.From.Type = obj.TYPE_CONST
+		p.From.Offset = v.AuxInt
+		p.Reg = gc.SSARegNum(v.Args[0])
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = gc.SSARegNum(v)
+	case ssa.OpARM64MOVDconst:
+		p := gc.Prog(v.Op.Asm())
+		p.From.Type = obj.TYPE_CONST
+		p.From.Offset = v.AuxInt
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = gc.SSARegNum(v)
+	case ssa.OpARM64FMOVSconst,
+		ssa.OpARM64FMOVDconst:
+		p := gc.Prog(v.Op.Asm())
+		p.From.Type = obj.TYPE_FCONST
+		p.From.Val = math.Float64frombits(uint64(v.AuxInt))
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = gc.SSARegNum(v)
+	case ssa.OpARM64CMP,
+		ssa.OpARM64CMPW,
+		ssa.OpARM64CMN,
+		ssa.OpARM64CMNW,
+		ssa.OpARM64FCMPS,
+		ssa.OpARM64FCMPD:
+		p := gc.Prog(v.Op.Asm())
+		p.From.Type = obj.TYPE_REG
+		p.From.Reg = gc.SSARegNum(v.Args[1])
+		p.Reg = gc.SSARegNum(v.Args[0])
+	case ssa.OpARM64CMPconst,
+		ssa.OpARM64CMPWconst,
+		ssa.OpARM64CMNconst,
+		ssa.OpARM64CMNWconst:
+		p := gc.Prog(v.Op.Asm())
+		p.From.Type = obj.TYPE_CONST
+		p.From.Offset = v.AuxInt
+		p.Reg = gc.SSARegNum(v.Args[0])
+	case ssa.OpARM64MOVDaddr:
+		p := gc.Prog(arm64.AMOVD)
+		p.From.Type = obj.TYPE_ADDR
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = gc.SSARegNum(v)
+
+		var wantreg string
+		// MOVD $sym+off(base), R
+		// the assembler expands it as the following:
+		// - base is SP: add constant offset to SP (R13)
+		//               when constant is large, tmp register (R11) may be used
+		// - base is SB: load external address from constant pool (use relocation)
+		switch v.Aux.(type) {
+		default:
+			v.Fatalf("aux is of unknown type %T", v.Aux)
+		case *ssa.ExternSymbol:
+			wantreg = "SB"
+			gc.AddAux(&p.From, v)
+		case *ssa.ArgSymbol, *ssa.AutoSymbol:
+			wantreg = "SP"
+			gc.AddAux(&p.From, v)
+		case nil:
+			// No sym, just MOVD $off(SP), R
+			wantreg = "SP"
+			p.From.Reg = arm64.REGSP
+			p.From.Offset = v.AuxInt
+		}
+		if reg := gc.SSAReg(v.Args[0]); reg.Name() != wantreg {
+			v.Fatalf("bad reg %s for symbol type %T, want %s", reg.Name(), v.Aux, wantreg)
+		}
+	case ssa.OpARM64MOVBload,
+		ssa.OpARM64MOVBUload,
+		ssa.OpARM64MOVHload,
+		ssa.OpARM64MOVHUload,
+		ssa.OpARM64MOVWload,
+		ssa.OpARM64MOVWUload,
+		ssa.OpARM64MOVDload,
+		ssa.OpARM64FMOVSload,
+		ssa.OpARM64FMOVDload:
+		p := gc.Prog(v.Op.Asm())
+		p.From.Type = obj.TYPE_MEM
+		p.From.Reg = gc.SSARegNum(v.Args[0])
+		gc.AddAux(&p.From, v)
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = gc.SSARegNum(v)
+	case ssa.OpARM64MOVBstore,
+		ssa.OpARM64MOVHstore,
+		ssa.OpARM64MOVWstore,
+		ssa.OpARM64MOVDstore,
+		ssa.OpARM64FMOVSstore,
+		ssa.OpARM64FMOVDstore:
+		p := gc.Prog(v.Op.Asm())
+		p.From.Type = obj.TYPE_REG
+		p.From.Reg = gc.SSARegNum(v.Args[1])
+		p.To.Type = obj.TYPE_MEM
+		p.To.Reg = gc.SSARegNum(v.Args[0])
+		gc.AddAux(&p.To, v)
+	case ssa.OpARM64MOVBreg,
+		ssa.OpARM64MOVBUreg,
+		ssa.OpARM64MOVHreg,
+		ssa.OpARM64MOVHUreg,
+		ssa.OpARM64MOVWreg,
+		ssa.OpARM64MOVWUreg:
+		a := v.Args[0]
+		for a.Op == ssa.OpCopy || a.Op == ssa.OpARM64MOVDreg {
+			a = a.Args[0]
+		}
+		if a.Op == ssa.OpLoadReg {
+			t := a.Type
+			switch {
+			case v.Op == ssa.OpARM64MOVBreg && t.Size() == 1 && t.IsSigned(),
+				v.Op == ssa.OpARM64MOVBUreg && t.Size() == 1 && !t.IsSigned(),
+				v.Op == ssa.OpARM64MOVHreg && t.Size() == 2 && t.IsSigned(),
+				v.Op == ssa.OpARM64MOVHUreg && t.Size() == 2 && !t.IsSigned(),
+				v.Op == ssa.OpARM64MOVWreg && t.Size() == 4 && t.IsSigned(),
+				v.Op == ssa.OpARM64MOVWUreg && t.Size() == 4 && !t.IsSigned():
+				// arg is a proper-typed load, already zero/sign-extended, don't extend again
+				if gc.SSARegNum(v) == gc.SSARegNum(v.Args[0]) {
+					return
+				}
+				p := gc.Prog(arm64.AMOVD)
+				p.From.Type = obj.TYPE_REG
+				p.From.Reg = gc.SSARegNum(v.Args[0])
+				p.To.Type = obj.TYPE_REG
+				p.To.Reg = gc.SSARegNum(v)
+				return
+			default:
+			}
+		}
+		fallthrough
+	case ssa.OpARM64MVN,
+		ssa.OpARM64NEG,
+		ssa.OpARM64FNEGS,
+		ssa.OpARM64FNEGD,
+		ssa.OpARM64FSQRTD,
+		ssa.OpARM64FCVTZSSW,
+		ssa.OpARM64FCVTZSDW,
+		ssa.OpARM64FCVTZUSW,
+		ssa.OpARM64FCVTZUDW,
+		ssa.OpARM64FCVTZSS,
+		ssa.OpARM64FCVTZSD,
+		ssa.OpARM64FCVTZUS,
+		ssa.OpARM64FCVTZUD,
+		ssa.OpARM64SCVTFWS,
+		ssa.OpARM64SCVTFWD,
+		ssa.OpARM64SCVTFS,
+		ssa.OpARM64SCVTFD,
+		ssa.OpARM64UCVTFWS,
+		ssa.OpARM64UCVTFWD,
+		ssa.OpARM64UCVTFS,
+		ssa.OpARM64UCVTFD,
+		ssa.OpARM64FCVTSD,
+		ssa.OpARM64FCVTDS:
+		p := gc.Prog(v.Op.Asm())
+		p.From.Type = obj.TYPE_REG
+		p.From.Reg = gc.SSARegNum(v.Args[0])
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = gc.SSARegNum(v)
+	case ssa.OpARM64CALLstatic:
+		if v.Aux.(*gc.Sym) == gc.Deferreturn.Sym {
+			// Deferred calls will appear to be returning to
+			// the CALL deferreturn(SB) that we are about to emit.
+			// However, the stack trace code will show the line
+			// of the instruction byte before the return PC.
+			// To avoid that being an unrelated instruction,
+			// insert an actual hardware NOP that will have the right line number.
+			// This is different from obj.ANOP, which is a virtual no-op
+			// that doesn't make it into the instruction stream.
+			ginsnop()
+		}
+		p := gc.Prog(obj.ACALL)
+		p.To.Type = obj.TYPE_MEM
+		p.To.Name = obj.NAME_EXTERN
+		p.To.Sym = gc.Linksym(v.Aux.(*gc.Sym))
+		if gc.Maxarg < v.AuxInt {
+			gc.Maxarg = v.AuxInt
+		}
+	case ssa.OpARM64CALLclosure:
+		p := gc.Prog(obj.ACALL)
+		p.To.Type = obj.TYPE_MEM
+		p.To.Offset = 0
+		p.To.Reg = gc.SSARegNum(v.Args[0])
+		if gc.Maxarg < v.AuxInt {
+			gc.Maxarg = v.AuxInt
+		}
+	case ssa.OpARM64CALLdefer:
+		p := gc.Prog(obj.ACALL)
+		p.To.Type = obj.TYPE_MEM
+		p.To.Name = obj.NAME_EXTERN
+		p.To.Sym = gc.Linksym(gc.Deferproc.Sym)
+		if gc.Maxarg < v.AuxInt {
+			gc.Maxarg = v.AuxInt
+		}
+	case ssa.OpARM64CALLgo:
+		p := gc.Prog(obj.ACALL)
+		p.To.Type = obj.TYPE_MEM
+		p.To.Name = obj.NAME_EXTERN
+		p.To.Sym = gc.Linksym(gc.Newproc.Sym)
+		if gc.Maxarg < v.AuxInt {
+			gc.Maxarg = v.AuxInt
+		}
+	case ssa.OpARM64CALLinter:
+		p := gc.Prog(obj.ACALL)
+		p.To.Type = obj.TYPE_MEM
+		p.To.Offset = 0
+		p.To.Reg = gc.SSARegNum(v.Args[0])
+		if gc.Maxarg < v.AuxInt {
+			gc.Maxarg = v.AuxInt
+		}
+	case ssa.OpARM64LoweredNilCheck:
+		// Issue a load which will fault if arg is nil.
+		p := gc.Prog(arm64.AMOVB)
+		p.From.Type = obj.TYPE_MEM
+		p.From.Reg = gc.SSARegNum(v.Args[0])
+		gc.AddAux(&p.From, v)
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = arm64.REGTMP
+		if gc.Debug_checknil != 0 && v.Line > 1 { // v.Line==1 in generated wrappers
+			gc.Warnl(v.Line, "generated nil check")
+		}
+	case ssa.OpVarDef:
+		gc.Gvardef(v.Aux.(*gc.Node))
+	case ssa.OpVarKill:
+		gc.Gvarkill(v.Aux.(*gc.Node))
+	case ssa.OpVarLive:
+		gc.Gvarlive(v.Aux.(*gc.Node))
+	case ssa.OpKeepAlive:
+		if !v.Args[0].Type.IsPtrShaped() {
+			v.Fatalf("keeping non-pointer alive %v", v.Args[0])
+		}
+		n, off := gc.AutoVar(v.Args[0])
+		if n == nil {
+			v.Fatalf("KeepLive with non-spilled value %s %s", v, v.Args[0])
+		}
+		if off != 0 {
+			v.Fatalf("KeepLive with non-zero offset spill location %s:%d", n, off)
+		}
+		gc.Gvarlive(n)
+	case ssa.OpARM64Equal,
+		ssa.OpARM64NotEqual,
+		ssa.OpARM64LessThan,
+		ssa.OpARM64LessEqual,
+		ssa.OpARM64GreaterThan,
+		ssa.OpARM64GreaterEqual,
+		ssa.OpARM64LessThanU,
+		ssa.OpARM64LessEqualU,
+		ssa.OpARM64GreaterThanU,
+		ssa.OpARM64GreaterEqualU:
+		// generate boolean values using CSET
+		p := gc.Prog(arm64.ACSET)
+		p.From.Type = obj.TYPE_REG
+		p.From.Reg = condBits[v.Op]
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = gc.SSARegNum(v)
+	case ssa.OpSelect0, ssa.OpSelect1:
+		// nothing to do
+	case ssa.OpARM64LoweredGetClosurePtr:
+		// Closure pointer is R26 (arm64.REGCTXT).
+		gc.CheckLoweredGetClosurePtr(v)
+	case ssa.OpARM64FlagEQ,
+		ssa.OpARM64FlagLT_ULT,
+		ssa.OpARM64FlagLT_UGT,
+		ssa.OpARM64FlagGT_ULT,
+		ssa.OpARM64FlagGT_UGT:
+		v.Fatalf("Flag* ops should never make it to codegen %v", v.LongString())
+	case ssa.OpARM64InvertFlags:
+		v.Fatalf("InvertFlags should never make it to codegen %v", v.LongString())
+	default:
+		v.Unimplementedf("genValue not implemented: %s", v.LongString())
+	}
+}
+
+var condBits = map[ssa.Op]int16{
+	ssa.OpARM64Equal:         arm64.COND_EQ,
+	ssa.OpARM64NotEqual:      arm64.COND_NE,
+	ssa.OpARM64LessThan:      arm64.COND_LT,
+	ssa.OpARM64LessThanU:     arm64.COND_LO,
+	ssa.OpARM64LessEqual:     arm64.COND_LE,
+	ssa.OpARM64LessEqualU:    arm64.COND_LS,
+	ssa.OpARM64GreaterThan:   arm64.COND_GT,
+	ssa.OpARM64GreaterThanU:  arm64.COND_HI,
+	ssa.OpARM64GreaterEqual:  arm64.COND_GE,
+	ssa.OpARM64GreaterEqualU: arm64.COND_HS,
+}
+
+var blockJump = map[ssa.BlockKind]struct {
+	asm, invasm obj.As
+}{
+	ssa.BlockARM64EQ:  {arm64.ABEQ, arm64.ABNE},
+	ssa.BlockARM64NE:  {arm64.ABNE, arm64.ABEQ},
+	ssa.BlockARM64LT:  {arm64.ABLT, arm64.ABGE},
+	ssa.BlockARM64GE:  {arm64.ABGE, arm64.ABLT},
+	ssa.BlockARM64LE:  {arm64.ABLE, arm64.ABGT},
+	ssa.BlockARM64GT:  {arm64.ABGT, arm64.ABLE},
+	ssa.BlockARM64ULT: {arm64.ABLO, arm64.ABHS},
+	ssa.BlockARM64UGE: {arm64.ABHS, arm64.ABLO},
+	ssa.BlockARM64UGT: {arm64.ABHI, arm64.ABLS},
+	ssa.BlockARM64ULE: {arm64.ABLS, arm64.ABHI},
+}
+
+func ssaGenBlock(s *gc.SSAGenState, b, next *ssa.Block) {
+	s.SetLineno(b.Line)
+
+	switch b.Kind {
+	case ssa.BlockPlain, ssa.BlockCall, ssa.BlockCheck:
+		if b.Succs[0].Block() != next {
+			p := gc.Prog(obj.AJMP)
+			p.To.Type = obj.TYPE_BRANCH
+			s.Branches = append(s.Branches, gc.Branch{P: p, B: b.Succs[0].Block()})
+		}
+
+	case ssa.BlockDefer:
+		// defer returns in R0:
+		// 0 if we should continue executing
+		// 1 if we should jump to deferreturn call
+		p := gc.Prog(arm64.ACMP)
+		p.From.Type = obj.TYPE_CONST
+		p.From.Offset = 0
+		p.Reg = arm64.REG_R0
+		p = gc.Prog(arm64.ABNE)
+		p.To.Type = obj.TYPE_BRANCH
+		s.Branches = append(s.Branches, gc.Branch{P: p, B: b.Succs[1].Block()})
+		if b.Succs[0].Block() != next {
+			p := gc.Prog(obj.AJMP)
+			p.To.Type = obj.TYPE_BRANCH
+			s.Branches = append(s.Branches, gc.Branch{P: p, B: b.Succs[0].Block()})
+		}
+
+	case ssa.BlockExit:
+		gc.Prog(obj.AUNDEF) // tell plive.go that we never reach here
+
+	case ssa.BlockRet:
+		gc.Prog(obj.ARET)
+
+	case ssa.BlockRetJmp:
+		p := gc.Prog(obj.ARET)
+		p.To.Type = obj.TYPE_MEM
+		p.To.Name = obj.NAME_EXTERN
+		p.To.Sym = gc.Linksym(b.Aux.(*gc.Sym))
+
+	case ssa.BlockARM64EQ, ssa.BlockARM64NE,
+		ssa.BlockARM64LT, ssa.BlockARM64GE,
+		ssa.BlockARM64LE, ssa.BlockARM64GT,
+		ssa.BlockARM64ULT, ssa.BlockARM64UGT,
+		ssa.BlockARM64ULE, ssa.BlockARM64UGE:
+		jmp := blockJump[b.Kind]
+		var p *obj.Prog
+		switch next {
+		case b.Succs[0].Block():
+			p = gc.Prog(jmp.invasm)
+			p.To.Type = obj.TYPE_BRANCH
+			s.Branches = append(s.Branches, gc.Branch{P: p, B: b.Succs[1].Block()})
+		case b.Succs[1].Block():
+			p = gc.Prog(jmp.asm)
+			p.To.Type = obj.TYPE_BRANCH
+			s.Branches = append(s.Branches, gc.Branch{P: p, B: b.Succs[0].Block()})
+		default:
+			p = gc.Prog(jmp.asm)
+			p.To.Type = obj.TYPE_BRANCH
+			s.Branches = append(s.Branches, gc.Branch{P: p, B: b.Succs[0].Block()})
+			q := gc.Prog(obj.AJMP)
+			q.To.Type = obj.TYPE_BRANCH
+			s.Branches = append(s.Branches, gc.Branch{P: q, B: b.Succs[1].Block()})
+		}
+
+	default:
+		b.Unimplementedf("branch not implemented: %s. Control: %s", b.LongString(), b.Control.LongString())
+	}
+}
diff --git a/src/cmd/compile/internal/ssa/config.go b/src/cmd/compile/internal/ssa/config.go
index 9c1daa9c7b..4a9817423f 100644
--- a/src/cmd/compile/internal/ssa/config.go
+++ b/src/cmd/compile/internal/ssa/config.go
@@ -162,6 +162,17 @@ func NewConfig(arch string, fe Frontend, ctxt *obj.Link, optimize bool) *Config
 		c.flagRegMask = flagRegMaskARM
 		c.FPReg = framepointerRegARM
 		c.hasGReg = true
+	case "arm64":
+		c.IntSize = 8
+		c.PtrSize = 8
+		c.lowerBlock = rewriteBlockARM64
+		c.lowerValue = rewriteValueARM64
+		c.registers = registersARM64[:]
+		c.gpRegMask = gpRegMaskARM64
+		c.fpRegMask = fpRegMaskARM64
+		c.flagRegMask = flagRegMaskARM64
+		c.FPReg = framepointerRegARM64
+		c.hasGReg = true
 	case "ppc64le":
 		c.IntSize = 8
 		c.PtrSize = 8
diff --git a/src/cmd/compile/internal/ssa/gen/ARM64.rules b/src/cmd/compile/internal/ssa/gen/ARM64.rules
new file mode 100644
index 0000000000..9290b3c186
--- /dev/null
+++ b/src/cmd/compile/internal/ssa/gen/ARM64.rules
@@ -0,0 +1,323 @@
+// Copyright 2016 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+(AddPtr x y) -> (ADD x y)
+(Add64 x y) -> (ADD x y)
+(Add32 x y) -> (ADD x y)
+(Add16 x y) -> (ADD x y)
+(Add8 x y) -> (ADD x y)
+(Add32F x y) -> (FADDS x y)
+(Add64F x y) -> (FADDD x y)
+
+(SubPtr x y) -> (SUB x y)
+(Sub64 x y) -> (SUB x y)
+(Sub32 x y) -> (SUB x y)
+(Sub16 x y) -> (SUB x y)
+(Sub8 x y) -> (SUB x y)
+(Sub32F x y) -> (FSUBS x y)
+(Sub64F x y) -> (FSUBD x y)
+
+(Mul64 x y) -> (MUL x y)
+(Mul32 x y) -> (MUL x y)
+(Mul16 x y) -> (MUL x y)
+(Mul8 x y) -> (MUL x y)
+(Mul32F x y) -> (FMULS x y)
+(Mul64F x y) -> (FMULD x y)
+
+(Div64 x y) -> (DIV x y)
+(Div64u x y) -> (UDIV x y)
+(Div32 x y) -> (DIVW x y)
+(Div32u x y) -> (UDIVW x y)
+(Div16 x y) -> (DIVW (SignExt16to32 x) (SignExt16to32 y))
+(Div16u x y) -> (UDIVW (ZeroExt16to32 x) (ZeroExt16to32 y))
+(Div8 x y) -> (DIVW (SignExt8to32 x) (SignExt8to32 y))
+(Div8u x y) -> (UDIVW (ZeroExt8to32 x) (ZeroExt8to32 y))
+(Div32F x y) -> (FDIVS x y)
+(Div64F x y) -> (FDIVD x y)
+
+(Mod64 x y) -> (MOD x y)
+(Mod64u x y) -> (UMOD x y)
+(Mod32 x y) -> (MODW x y)
+(Mod32u x y) -> (UMODW x y)
+(Mod16 x y) -> (MODW (SignExt16to32 x) (SignExt16to32 y))
+(Mod16u x y) -> (UMODW (ZeroExt16to32 x) (ZeroExt16to32 y))
+(Mod8 x y) -> (MODW (SignExt8to32 x) (SignExt8to32 y))
+(Mod8u x y) -> (UMODW (ZeroExt8to32 x) (ZeroExt8to32 y))
+
+(And64 x y) -> (AND x y)
+(And32 x y) -> (AND x y)
+(And16 x y) -> (AND x y)
+(And8 x y) -> (AND x y)
+
+(Or64 x y) -> (OR x y)
+(Or32 x y) -> (OR x y)
+(Or16 x y) -> (OR x y)
+(Or8 x y) -> (OR x y)
+
+(Xor64 x y) -> (XOR x y)
+(Xor32 x y) -> (XOR x y)
+(Xor16 x y) -> (XOR x y)
+(Xor8 x y) -> (XOR x y)
+
+// unary ops
+(Neg64 x) -> (NEG x)
+(Neg32 x) -> (NEG x)
+(Neg16 x) -> (NEG x)
+(Neg8 x) -> (NEG x)
+(Neg32F x) -> (FNEGS x)
+(Neg64F x) -> (FNEGD x)
+
+(Com64 x) -> (MVN x)
+(Com32 x) -> (MVN x)
+(Com16 x) -> (MVN x)
+(Com8 x) -> (MVN x)
+
+// boolean ops -- booleans are represented with 0=false, 1=true
+(AndB x y) -> (AND x y)
+(OrB x y) -> (OR x y)
+(EqB x y) -> (XORconst [1] (XOR <config.fe.TypeBool()> x y))
+(NeqB x y) -> (XOR x y)
+(Not x) -> (XORconst [1] x)
+
+// constants
+(Const64 [val]) -> (MOVDconst [val])
+(Const32 [val]) -> (MOVDconst [val])
+(Const16 [val]) -> (MOVDconst [val])
+(Const8 [val]) -> (MOVDconst [val])
+(Const32F [val]) -> (FMOVSconst [val])
+(Const64F [val]) -> (FMOVDconst [val])
+(ConstNil) -> (MOVDconst [0])
+(ConstBool [b]) -> (MOVDconst [b])
+
+// truncations
+// Because we ignore high parts of registers, truncates are just copies.
+(Trunc16to8 x) -> x
+(Trunc32to8 x) -> x
+(Trunc32to16 x) -> x
+(Trunc64to8 x) -> x
+(Trunc64to16 x) -> x
+(Trunc64to32 x) -> x
+
+// Zero-/Sign-extensions
+(ZeroExt8to16 x) -> (MOVBUreg x)
+(ZeroExt8to32 x) -> (MOVBUreg x)
+(ZeroExt16to32 x) -> (MOVHUreg x)
+(ZeroExt8to64 x) -> (MOVBUreg x)
+(ZeroExt16to64 x) -> (MOVHUreg x)
+(ZeroExt32to64 x) -> (MOVWUreg x)
+
+(SignExt8to16 x) -> (MOVBreg x)
+(SignExt8to32 x) -> (MOVBreg x)
+(SignExt16to32 x) -> (MOVHreg x)
+(SignExt8to64 x) -> (MOVBreg x)
+(SignExt16to64 x) -> (MOVHreg x)
+(SignExt32to64 x) -> (MOVWreg x)
+
+// float <-> int conversion
+(Cvt32to32F x) -> (SCVTFWS x)
+(Cvt32to64F x) -> (SCVTFWD x)
+(Cvt64to32F x) -> (SCVTFS x)
+(Cvt64to64F x) -> (SCVTFD x)
+(Cvt32Uto32F x) -> (UCVTFWS x)
+(Cvt32Uto64F x) -> (UCVTFWD x)
+//(Cvt64Uto32F x) -> (UCVTFS x)
+//(Cvt64Uto64F x) -> (UCVTFD x)
+(Cvt32Fto32 x) -> (FCVTZSSW x)
+(Cvt64Fto32 x) -> (FCVTZSDW x)
+(Cvt32Fto64 x) -> (FCVTZSS x)
+(Cvt64Fto64 x) -> (FCVTZSD x)
+(Cvt32Fto32U x) -> (FCVTZUSW x)
+(Cvt64Fto32U x) -> (FCVTZUDW x)
+//(Cvt32Fto64U x) -> (FCVTZUS x)
+//(Cvt64Fto64U x) -> (FCVTZUD x)
+(Cvt32Fto64F x) -> (FCVTSD x)
+(Cvt64Fto32F x) -> (FCVTDS x)
+
+// comparisons
+(Eq8 x y)  -> (Equal (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+(Eq16 x y) -> (Equal (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+(Eq32 x y) -> (Equal (CMPW x y))
+(Eq64 x y) -> (Equal (CMP x y))
+(EqPtr x y) -> (Equal (CMP x y))
+(Eq32F x y) -> (Equal (FCMPS x y))
+(Eq64F x y) -> (Equal (FCMPD x y))
+
+(Neq8 x y)  -> (NotEqual (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+(Neq16 x y) -> (NotEqual (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+(Neq32 x y) -> (NotEqual (CMPW x y))
+(Neq64 x y) -> (NotEqual (CMP x y))
+(NeqPtr x y) -> (NotEqual (CMP x y))
+(Neq32F x y) -> (NotEqual (FCMPS x y))
+(Neq64F x y) -> (NotEqual (FCMPD x y))
+
+(Less8 x y)  -> (LessThan (CMPW (SignExt8to32 x) (SignExt8to32 y)))
+(Less16 x y) -> (LessThan (CMPW (SignExt16to32 x) (SignExt16to32 y)))
+(Less32 x y) -> (LessThan (CMPW x y))
+(Less64 x y) -> (LessThan (CMP x y))
+(Less32F x y) -> (GreaterThan (FCMPS y x)) // reverse operands to work around NaN
+(Less64F x y) -> (GreaterThan (FCMPD y x)) // reverse operands to work around NaN
+
+(Less8U x y)  -> (LessThanU (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+(Less16U x y) -> (LessThanU (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+(Less32U x y) -> (LessThanU (CMPW x y))
+(Less64U x y) -> (LessThanU (CMP x y))
+
+(Leq8 x y)  -> (LessEqual (CMPW (SignExt8to32 x) (SignExt8to32 y)))
+(Leq16 x y) -> (LessEqual (CMPW (SignExt16to32 x) (SignExt16to32 y)))
+(Leq32 x y) -> (LessEqual (CMPW x y))
+(Leq64 x y) -> (LessEqual (CMP x y))
+(Leq32F x y) -> (GreaterEqual (FCMPS y x)) // reverse operands to work around NaN
+(Leq64F x y) -> (GreaterEqual (FCMPD y x)) // reverse operands to work around NaN
+
+(Leq8U x y)  -> (LessEqualU (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+(Leq16U x y) -> (LessEqualU (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+(Leq32U x y) -> (LessEqualU (CMPW x y))
+(Leq64U x y) -> (LessEqualU (CMP x y))
+
+(Greater8 x y)  -> (GreaterThan (CMPW (SignExt8to32 x) (SignExt8to32 y)))
+(Greater16 x y) -> (GreaterThan (CMPW (SignExt16to32 x) (SignExt16to32 y)))
+(Greater32 x y) -> (GreaterThan (CMPW x y))
+(Greater64 x y) -> (GreaterThan (CMP x y))
+(Greater32F x y) -> (GreaterThan (FCMPS x y))
+(Greater64F x y) -> (GreaterThan (FCMPD x y))
+
+(Greater8U x y)  -> (GreaterThanU (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+(Greater16U x y) -> (GreaterThanU (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+(Greater32U x y) -> (GreaterThanU (CMPW x y))
+(Greater64U x y) -> (GreaterThanU (CMP x y))
+
+(Geq8 x y)  -> (GreaterEqual (CMPW (SignExt8to32 x) (SignExt8to32 y)))
+(Geq16 x y) -> (GreaterEqual (CMPW (SignExt16to32 x) (SignExt16to32 y)))
+(Geq32 x y) -> (GreaterEqual (CMPW x y))
+(Geq64 x y) -> (GreaterEqual (CMP x y))
+(Geq32F x y) -> (GreaterEqual (FCMPS x y))
+(Geq64F x y) -> (GreaterEqual (FCMPD x y))
+
+(Geq8U x y)  -> (GreaterEqualU (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+(Geq16U x y) -> (GreaterEqualU (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+(Geq32U x y) -> (GreaterEqualU (CMPW x y))
+(Geq64U x y) -> (GreaterEqualU (CMP x y))
+
+(OffPtr [off] ptr:(SP)) -> (MOVDaddr [off] ptr)
+(OffPtr [off] ptr) -> (ADDconst [off] ptr)
+
+(Addr {sym} base) -> (MOVDaddr {sym} base)
+
+// loads
+(Load <t> ptr mem) && t.IsBoolean() -> (MOVBUload ptr mem)
+(Load <t> ptr mem) && (is8BitInt(t) && isSigned(t)) -> (MOVBload ptr mem)
+(Load <t> ptr mem) && (is8BitInt(t) && !isSigned(t)) -> (MOVBUload ptr mem)
+(Load <t> ptr mem) && (is16BitInt(t) && isSigned(t)) -> (MOVHload ptr mem)
+(Load <t> ptr mem) && (is16BitInt(t) && !isSigned(t)) -> (MOVHUload ptr mem)
+(Load <t> ptr mem) && (is32BitInt(t) && isSigned(t)) -> (MOVWload ptr mem)
+(Load <t> ptr mem) && (is32BitInt(t) && !isSigned(t)) -> (MOVWUload ptr mem)
+(Load <t> ptr mem) && (is64BitInt(t) || isPtr(t)) -> (MOVDload ptr mem)
+(Load <t> ptr mem) && is32BitFloat(t) -> (FMOVSload ptr mem)
+(Load <t> ptr mem) && is64BitFloat(t) -> (FMOVDload ptr mem)
+
+// stores
+(Store [1] ptr val mem) -> (MOVBstore ptr val mem)
+(Store [2] ptr val mem) -> (MOVHstore ptr val mem)
+(Store [4] ptr val mem) && !is32BitFloat(val.Type) -> (MOVWstore ptr val mem)
+(Store [8] ptr val mem) && !is64BitFloat(val.Type) -> (MOVDstore ptr val mem)
+(Store [4] ptr val mem) && is32BitFloat(val.Type) -> (FMOVSstore ptr val mem)
+(Store [8] ptr val mem) && is64BitFloat(val.Type) -> (FMOVDstore ptr val mem)
+
+// calls
+(StaticCall [argwid] {target} mem) -> (CALLstatic [argwid] {target} mem)
+(ClosureCall [argwid] entry closure mem) -> (CALLclosure [argwid] entry closure mem)
+(DeferCall [argwid] mem) -> (CALLdefer [argwid] mem)
+(GoCall [argwid] mem) -> (CALLgo [argwid] mem)
+(InterCall [argwid] entry mem) -> (CALLinter [argwid] entry mem)
+
+// checks
+(NilCheck ptr mem) -> (LoweredNilCheck ptr mem)
+(IsNonNil ptr) -> (NotEqual (CMPconst [0] ptr))
+(IsInBounds idx len) -> (LessThanU (CMP idx len))
+(IsSliceInBounds idx len) -> (LessEqualU (CMP idx len))
+
+// pseudo-ops
+(GetClosurePtr) -> (LoweredGetClosurePtr)
+(Convert x mem) -> (MOVDconvert x mem)
+
+// Absorb pseudo-ops into blocks.
+(If (Equal cc) yes no) -> (EQ cc yes no)
+(If (NotEqual cc) yes no) -> (NE cc yes no)
+(If (LessThan cc) yes no) -> (LT cc yes no)
+(If (LessThanU cc) yes no) -> (ULT cc yes no)
+(If (LessEqual cc) yes no) -> (LE cc yes no)
+(If (LessEqualU cc) yes no) -> (ULE cc yes no)
+(If (GreaterThan cc) yes no) -> (GT cc yes no)
+(If (GreaterThanU cc) yes no) -> (UGT cc yes no)
+(If (GreaterEqual cc) yes no) -> (GE cc yes no)
+(If (GreaterEqualU cc) yes no) -> (UGE cc yes no)
+
+(If cond yes no) -> (NE (CMPconst [0] cond) yes no)
+
+// Absorb boolean tests into block
+(NE (CMPconst [0] (Equal cc)) yes no) -> (EQ cc yes no)
+(NE (CMPconst [0] (NotEqual cc)) yes no) -> (NE cc yes no)
+(NE (CMPconst [0] (LessThan cc)) yes no) -> (LT cc yes no)
+(NE (CMPconst [0] (LessThanU cc)) yes no) -> (ULT cc yes no)
+(NE (CMPconst [0] (LessEqual cc)) yes no) -> (LE cc yes no)
+(NE (CMPconst [0] (LessEqualU cc)) yes no) -> (ULE cc yes no)
+(NE (CMPconst [0] (GreaterThan cc)) yes no) -> (GT cc yes no)
+(NE (CMPconst [0] (GreaterThanU cc)) yes no) -> (UGT cc yes no)
+(NE (CMPconst [0] (GreaterEqual cc)) yes no) -> (GE cc yes no)
+(NE (CMPconst [0] (GreaterEqualU cc)) yes no) -> (UGE cc yes no)
+
+// Optimizations
+
+// fold offset into address
+(ADDconst [off1] (MOVDaddr [off2] {sym} ptr)) -> (MOVDaddr [off1+off2] {sym} ptr)
+
+// fold address into load/store
+(MOVBload [off1] {sym} (ADDconst [off2] ptr) mem) -> (MOVBload [off1+off2] {sym} ptr mem)
+(MOVBUload [off1] {sym} (ADDconst [off2] ptr) mem) -> (MOVBUload [off1+off2] {sym} ptr mem)
+(MOVHload [off1] {sym} (ADDconst [off2] ptr) mem) -> (MOVHload [off1+off2] {sym} ptr mem)
+(MOVHUload [off1] {sym} (ADDconst [off2] ptr) mem) -> (MOVHUload [off1+off2] {sym} ptr mem)
+(MOVWload [off1] {sym} (ADDconst [off2] ptr) mem) -> (MOVWload [off1+off2] {sym} ptr mem)
+(MOVWUload [off1] {sym} (ADDconst [off2] ptr) mem) -> (MOVWUload [off1+off2] {sym} ptr mem)
+(MOVDload [off1] {sym} (ADDconst [off2] ptr) mem) -> (MOVDload [off1+off2] {sym} ptr mem)
+(FMOVSload [off1] {sym} (ADDconst [off2] ptr) mem) -> (FMOVSload [off1+off2] {sym} ptr mem)
+(FMOVDload [off1] {sym} (ADDconst [off2] ptr) mem) -> (FMOVDload [off1+off2] {sym} ptr mem)
+
+(MOVBstore [off1] {sym} (ADDconst [off2] ptr) val mem) -> (MOVBstore [off1+off2] {sym} ptr val mem)
+(MOVHstore [off1] {sym} (ADDconst [off2] ptr) val mem) -> (MOVHstore [off1+off2] {sym} ptr val mem)
+(MOVWstore [off1] {sym} (ADDconst [off2] ptr) val mem) -> (MOVWstore [off1+off2] {sym} ptr val mem)
+(MOVDstore [off1] {sym} (ADDconst [off2] ptr) val mem) -> (MOVDstore [off1+off2] {sym} ptr val mem)
+(FMOVSstore [off1] {sym} (ADDconst [off2] ptr) val mem) -> (FMOVSstore [off1+off2] {sym} ptr val mem)
+(FMOVDstore [off1] {sym} (ADDconst [off2] ptr) val mem) -> (FMOVDstore [off1+off2] {sym} ptr val mem)
+
+(MOVBload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem) && canMergeSym(sym1,sym2) ->
+	(MOVBload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+(MOVBUload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem) && canMergeSym(sym1,sym2) ->
+	(MOVBUload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+(MOVHload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem) && canMergeSym(sym1,sym2) ->
+	(MOVHload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+(MOVHUload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem) && canMergeSym(sym1,sym2) ->
+	(MOVHUload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+(MOVWload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem) && canMergeSym(sym1,sym2) ->
+	(MOVWload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+(MOVWUload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem) && canMergeSym(sym1,sym2) ->
+	(MOVWUload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+(MOVDload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem) && canMergeSym(sym1,sym2) ->
+	(MOVDload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+(FMOVSload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem) && canMergeSym(sym1,sym2) ->
+	(FMOVSload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+(FMOVDload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem) && canMergeSym(sym1,sym2) ->
+	(FMOVDload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+
+(MOVBstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem) && canMergeSym(sym1,sym2) ->
+	(MOVBstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+(MOVHstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem) && canMergeSym(sym1,sym2) ->
+	(MOVHstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+(MOVWstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem) && canMergeSym(sym1,sym2) ->
+	(MOVWstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+(MOVDstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem) && canMergeSym(sym1,sym2) ->
+	(MOVDstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+(FMOVSstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem) && canMergeSym(sym1,sym2) ->
+	(FMOVSstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+(FMOVDstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem) && canMergeSym(sym1,sym2) ->
+	(FMOVDstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
diff --git a/src/cmd/compile/internal/ssa/gen/ARM64Ops.go b/src/cmd/compile/internal/ssa/gen/ARM64Ops.go
new file mode 100644
index 0000000000..9925a65b00
--- /dev/null
+++ b/src/cmd/compile/internal/ssa/gen/ARM64Ops.go
@@ -0,0 +1,347 @@
+// Copyright 2016 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// +build ignore
+
+package main
+
+import "strings"
+
+// Notes:
+//  - Integer types live in the low portion of registers. Upper portions are junk.
+//  - Boolean types use the low-order byte of a register. 0=false, 1=true.
+//    Upper bytes are junk.
+//  - *const instructions may use a constant larger than the instuction can encode.
+//    In this case the assembler expands to multiple instructions and uses tmp
+//    register (R27).
+
+// Suffixes encode the bit width of various instructions.
+// D (double word) = 64 bit
+// W (word)        = 32 bit
+// H (half word)   = 16 bit
+// HU              = 16 bit unsigned
+// B (byte)        = 8 bit
+// BU              = 8 bit unsigned
+// S (single)      = 32 bit float
+// D (double)      = 64 bit float
+
+// Note: registers not used in regalloc are not included in this list,
+// so that regmask stays within int64
+// Be careful when hand coding regmasks.
+var regNamesARM64 = []string{
+	"R0",
+	"R1",
+	"R2",
+	"R3",
+	"R4",
+	"R5",
+	"R6",
+	"R7",
+	"R8",
+	"R9",
+	"R10",
+	"R11",
+	"R12",
+	"R13",
+	"R14",
+	"R15",
+	"R16",
+	"R17",
+	"R18", // platform register?
+	"R19",
+	"R20",
+	"R21",
+	"R22",
+	"R23",
+	"R24",
+	"R25",
+	"R26",
+	// R27 = REGTMP not used in regalloc
+	"g",   // aka R28
+	"R29", // frame pointer?
+	// R30 = REGLINK not used in regalloc
+	"SP", // aka R31
+
+	"F0",
+	"F1",
+	"F2",
+	"F3",
+	"F4",
+	"F5",
+	"F6",
+	"F7",
+	"F8",
+	"F9",
+	"F10",
+	"F11",
+	"F12",
+	"F13",
+	"F14",
+	"F15",
+	"F16",
+	"F17",
+	"F18",
+	"F19",
+	"F20",
+	"F21",
+	"F22",
+	"F23",
+	"F24",
+	"F25",
+	"F26",
+	"F27",
+	"F28", // 0.0
+	"F29", // 0.5
+	"F30", // 1.0
+	"F31", // 2.0
+
+	// pseudo-registers
+	"FLAGS",
+	"SB",
+}
+
+func init() {
+	// Make map from reg names to reg integers.
+	if len(regNamesARM64) > 64 {
+		panic("too many registers")
+	}
+	num := map[string]int{}
+	for i, name := range regNamesARM64 {
+		num[name] = i
+	}
+	buildReg := func(s string) regMask {
+		m := regMask(0)
+		for _, r := range strings.Split(s, " ") {
+			if n, ok := num[r]; ok {
+				m |= regMask(1) << uint(n)
+				continue
+			}
+			panic("register " + r + " not found")
+		}
+		return m
+	}
+
+	// Common individual register masks
+	var (
+		gp         = buildReg("R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29")
+		gpg        = gp | buildReg("g")
+		gpsp       = gp | buildReg("SP")
+		gpspg      = gpg | buildReg("SP")
+		gpspsbg    = gpspg | buildReg("SB")
+		flags      = buildReg("FLAGS")
+		fp         = buildReg("F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27")
+		callerSave = gp | fp | flags | buildReg("g") // runtime.setg (and anything calling it) may clobber g
+	)
+	// Common regInfo
+	var (
+		gp01     = regInfo{inputs: []regMask{}, outputs: []regMask{gp}}
+		gp11     = regInfo{inputs: []regMask{gpg}, outputs: []regMask{gp}}
+		gp11sp   = regInfo{inputs: []regMask{gpspg}, outputs: []regMask{gp}}
+		gp1flags = regInfo{inputs: []regMask{gpg}, outputs: []regMask{flags}}
+		//gp1flags1 = regInfo{inputs: []regMask{gp, flags}, outputs: []regMask{gp}}
+		gp21 = regInfo{inputs: []regMask{gpg, gpg}, outputs: []regMask{gp}}
+		//gp21cf    = regInfo{inputs: []regMask{gpg, gpg}, outputs: []regMask{gp}, clobbers: flags} // cf: clobbers flags
+		gp2flags = regInfo{inputs: []regMask{gpg, gpg}, outputs: []regMask{flags}}
+		//gp2flags1 = regInfo{inputs: []regMask{gp, gp, flags}, outputs: []regMask{gp}}
+		//gp22      = regInfo{inputs: []regMask{gpg, gpg}, outputs: []regMask{gp, gp}}
+		//gp31      = regInfo{inputs: []regMask{gp, gp, gp}, outputs: []regMask{gp}}
+		//gp3flags  = regInfo{inputs: []regMask{gp, gp, gp}, outputs: []regMask{flags}}
+		//gp3flags1 = regInfo{inputs: []regMask{gp, gp, gp, flags}, outputs: []regMask{gp}}
+		gpload  = regInfo{inputs: []regMask{gpspsbg}, outputs: []regMask{gp}}
+		gpstore = regInfo{inputs: []regMask{gpspsbg, gpg}, outputs: []regMask{}}
+		//gp2load   = regInfo{inputs: []regMask{gpspsbg, gpg}, outputs: []regMask{gp}}
+		//gp2store  = regInfo{inputs: []regMask{gpspsbg, gpg, gpg}, outputs: []regMask{}}
+		fp01 = regInfo{inputs: []regMask{}, outputs: []regMask{fp}}
+		fp11 = regInfo{inputs: []regMask{fp}, outputs: []regMask{fp}}
+		//fp1flags  = regInfo{inputs: []regMask{fp}, outputs: []regMask{flags}}
+		fpgp      = regInfo{inputs: []regMask{fp}, outputs: []regMask{gp}}
+		gpfp      = regInfo{inputs: []regMask{gp}, outputs: []regMask{fp}}
+		fp21      = regInfo{inputs: []regMask{fp, fp}, outputs: []regMask{fp}}
+		fp2flags  = regInfo{inputs: []regMask{fp, fp}, outputs: []regMask{flags}}
+		fpload    = regInfo{inputs: []regMask{gpspsbg}, outputs: []regMask{fp}}
+		fpstore   = regInfo{inputs: []regMask{gpspsbg, fp}, outputs: []regMask{}}
+		readflags = regInfo{inputs: []regMask{flags}, outputs: []regMask{gp}}
+	)
+	ops := []opData{
+		// binary ops
+		{name: "ADD", argLength: 2, reg: gp21, asm: "ADD", commutative: true},   // arg0 + arg1
+		{name: "ADDconst", argLength: 1, reg: gp11sp, asm: "ADD", aux: "Int32"}, // arg0 + auxInt
+		{name: "SUB", argLength: 2, reg: gp21, asm: "SUB"},                      // arg0 - arg1
+		{name: "SUBconst", argLength: 1, reg: gp11, asm: "SUB", aux: "Int32"},   // arg0 - auxInt
+		{name: "MUL", argLength: 2, reg: gp21, asm: "MUL", commutative: true},   // arg0 * arg1
+		{name: "DIV", argLength: 2, reg: gp21, asm: "SDIV"},                     // arg0 / arg1, signed
+		{name: "UDIV", argLength: 2, reg: gp21, asm: "UDIV"},                    // arg0 / arg1, unsighed
+		{name: "DIVW", argLength: 2, reg: gp21, asm: "SDIVW"},                   // arg0 / arg1, signed, 32 bit
+		{name: "UDIVW", argLength: 2, reg: gp21, asm: "UDIVW"},                  // arg0 / arg1, unsighed, 32 bit
+		{name: "MOD", argLength: 2, reg: gp21, asm: "REM"},                      // arg0 % arg1, signed
+		{name: "UMOD", argLength: 2, reg: gp21, asm: "UREM"},                    // arg0 % arg1, unsigned
+		{name: "MODW", argLength: 2, reg: gp21, asm: "REMW"},                    // arg0 % arg1, signed, 32 bit
+		{name: "UMODW", argLength: 2, reg: gp21, asm: "UREMW"},                  // arg0 % arg1, unsigned, 32 bit
+
+		{name: "FADDS", argLength: 2, reg: fp21, asm: "FADDS", commutative: true}, // arg0 + arg1
+		{name: "FADDD", argLength: 2, reg: fp21, asm: "FADDD", commutative: true}, // arg0 + arg1
+		{name: "FSUBS", argLength: 2, reg: fp21, asm: "FSUBS"},                    // arg0 - arg1
+		{name: "FSUBD", argLength: 2, reg: fp21, asm: "FSUBD"},                    // arg0 - arg1
+		{name: "FMULS", argLength: 2, reg: fp21, asm: "FMULS", commutative: true}, // arg0 * arg1
+		{name: "FMULD", argLength: 2, reg: fp21, asm: "FMULD", commutative: true}, // arg0 * arg1
+		{name: "FDIVS", argLength: 2, reg: fp21, asm: "FDIVS"},                    // arg0 / arg1
+		{name: "FDIVD", argLength: 2, reg: fp21, asm: "FDIVD"},                    // arg0 / arg1
+
+		{name: "AND", argLength: 2, reg: gp21, asm: "AND", commutative: true}, // arg0 & arg1
+		{name: "ANDconst", argLength: 1, reg: gp11, asm: "AND", aux: "Int32"}, // arg0 & auxInt
+		{name: "OR", argLength: 2, reg: gp21, asm: "ORR", commutative: true},  // arg0 | arg1
+		{name: "ORconst", argLength: 1, reg: gp11, asm: "ORR", aux: "Int32"},  // arg0 | auxInt
+		{name: "XOR", argLength: 2, reg: gp21, asm: "EOR", commutative: true}, // arg0 ^ arg1
+		{name: "XORconst", argLength: 1, reg: gp11, asm: "EOR", aux: "Int32"}, // arg0 ^ auxInt
+		{name: "BIC", argLength: 2, reg: gp21, asm: "BIC"},                    // arg0 &^ arg1
+		{name: "BICconst", argLength: 1, reg: gp11, asm: "BIC", aux: "Int32"}, // arg0 &^ auxInt
+
+		// unary ops
+		{name: "MVN", argLength: 1, reg: gp11, asm: "MVN"},       // ^arg0
+		{name: "NEG", argLength: 1, reg: gp11, asm: "NEG"},       // -arg0
+		{name: "FNEGS", argLength: 1, reg: fp11, asm: "FNEGS"},   // -arg0, float32
+		{name: "FNEGD", argLength: 1, reg: fp11, asm: "FNEGD"},   // -arg0, float64
+		{name: "FSQRTD", argLength: 1, reg: fp11, asm: "FSQRTD"}, // sqrt(arg0), float64
+
+		// comparisons
+		{name: "CMP", argLength: 2, reg: gp2flags, asm: "CMP", typ: "Flags"},                      // arg0 compare to arg1
+		{name: "CMPconst", argLength: 1, reg: gp1flags, asm: "CMP", aux: "Int32", typ: "Flags"},   // arg0 compare to auxInt
+		{name: "CMPW", argLength: 2, reg: gp2flags, asm: "CMPW", typ: "Flags"},                    // arg0 compare to arg1, 32 bit
+		{name: "CMPWconst", argLength: 1, reg: gp1flags, asm: "CMPW", aux: "Int32", typ: "Flags"}, // arg0 compare to auxInt, 32 bit
+		{name: "CMN", argLength: 2, reg: gp2flags, asm: "CMN", typ: "Flags"},                      // arg0 compare to -arg1
+		{name: "CMNconst", argLength: 1, reg: gp1flags, asm: "CMN", aux: "Int32", typ: "Flags"},   // arg0 compare to -auxInt
+		{name: "CMNW", argLength: 2, reg: gp2flags, asm: "CMNW", typ: "Flags"},                    // arg0 compare to -arg1, 32 bit
+		{name: "CMNWconst", argLength: 1, reg: gp1flags, asm: "CMNW", aux: "Int32", typ: "Flags"}, // arg0 compare to -auxInt, 32 bit
+		{name: "FCMPS", argLength: 2, reg: fp2flags, asm: "FCMPS", typ: "Flags"},                  // arg0 compare to arg1, float32
+		{name: "FCMPD", argLength: 2, reg: fp2flags, asm: "FCMPD", typ: "Flags"},                  // arg0 compare to arg1, float64
+
+		// moves
+		{name: "MOVDconst", argLength: 0, reg: gp01, aux: "Int32", asm: "MOVD", typ: "UInt32", rematerializeable: true},      // 32 low bits of auxint
+		{name: "FMOVSconst", argLength: 0, reg: fp01, aux: "Float64", asm: "FMOVS", typ: "Float32", rematerializeable: true}, // auxint as 64-bit float, convert to 32-bit float
+		{name: "FMOVDconst", argLength: 0, reg: fp01, aux: "Float64", asm: "FMOVD", typ: "Float64", rematerializeable: true}, // auxint as 64-bit float
+
+		{name: "MOVDaddr", argLength: 1, reg: regInfo{inputs: []regMask{buildReg("SP") | buildReg("SB")}, outputs: []regMask{gp}}, aux: "SymOff", asm: "MOVD", rematerializeable: true}, // arg0 + auxInt + aux.(*gc.Sym), arg0=SP/SB
+
+		{name: "MOVBload", argLength: 2, reg: gpload, aux: "SymOff", asm: "MOVB", typ: "Int8"},      // load from arg0 + auxInt + aux.  arg1=mem.
+		{name: "MOVBUload", argLength: 2, reg: gpload, aux: "SymOff", asm: "MOVBU", typ: "UInt8"},   // load from arg0 + auxInt + aux.  arg1=mem.
+		{name: "MOVHload", argLength: 2, reg: gpload, aux: "SymOff", asm: "MOVH", typ: "Int16"},     // load from arg0 + auxInt + aux.  arg1=mem.
+		{name: "MOVHUload", argLength: 2, reg: gpload, aux: "SymOff", asm: "MOVHU", typ: "UInt16"},  // load from arg0 + auxInt + aux.  arg1=mem.
+		{name: "MOVWload", argLength: 2, reg: gpload, aux: "SymOff", asm: "MOVW", typ: "Int32"},     // load from arg0 + auxInt + aux.  arg1=mem.
+		{name: "MOVWUload", argLength: 2, reg: gpload, aux: "SymOff", asm: "MOVWU", typ: "UInt32"},  // load from arg0 + auxInt + aux.  arg1=mem.
+		{name: "MOVDload", argLength: 2, reg: gpload, aux: "SymOff", asm: "MOVD", typ: "UInt64"},    // load from arg0 + auxInt + aux.  arg1=mem.
+		{name: "FMOVSload", argLength: 2, reg: fpload, aux: "SymOff", asm: "FMOVS", typ: "Float32"}, // load from arg0 + auxInt + aux.  arg1=mem.
+		{name: "FMOVDload", argLength: 2, reg: fpload, aux: "SymOff", asm: "FMOVD", typ: "Float64"}, // load from arg0 + auxInt + aux.  arg1=mem.
+
+		{name: "MOVBstore", argLength: 3, reg: gpstore, aux: "SymOff", asm: "MOVB", typ: "Mem"},   // store 1 byte of arg1 to arg0 + auxInt + aux.  arg2=mem.
+		{name: "MOVHstore", argLength: 3, reg: gpstore, aux: "SymOff", asm: "MOVH", typ: "Mem"},   // store 2 bytes of arg1 to arg0 + auxInt + aux.  arg2=mem.
+		{name: "MOVWstore", argLength: 3, reg: gpstore, aux: "SymOff", asm: "MOVW", typ: "Mem"},   // store 4 bytes of arg1 to arg0 + auxInt + aux.  arg2=mem.
+		{name: "MOVDstore", argLength: 3, reg: gpstore, aux: "SymOff", asm: "MOVD", typ: "Mem"},   // store 8 bytes of arg1 to arg0 + auxInt + aux.  arg2=mem.
+		{name: "FMOVSstore", argLength: 3, reg: fpstore, aux: "SymOff", asm: "FMOVS", typ: "Mem"}, // store 4 bytes of arg1 to arg0 + auxInt + aux.  arg2=mem.
+		{name: "FMOVDstore", argLength: 3, reg: fpstore, aux: "SymOff", asm: "FMOVD", typ: "Mem"}, // store 8 bytes of arg1 to arg0 + auxInt + aux.  arg2=mem.
+
+		// conversions
+		{name: "MOVBreg", argLength: 1, reg: gp11, asm: "MOVB"},   // move from arg0, sign-extended from byte
+		{name: "MOVBUreg", argLength: 1, reg: gp11, asm: "MOVBU"}, // move from arg0, unsign-extended from byte
+		{name: "MOVHreg", argLength: 1, reg: gp11, asm: "MOVH"},   // move from arg0, sign-extended from half
+		{name: "MOVHUreg", argLength: 1, reg: gp11, asm: "MOVHU"}, // move from arg0, unsign-extended from half
+		{name: "MOVWreg", argLength: 1, reg: gp11, asm: "MOVW"},   // move from arg0, sign-extended from word
+		{name: "MOVWUreg", argLength: 1, reg: gp11, asm: "MOVWU"}, // move from arg0, unsign-extended from word
+		{name: "MOVDreg", argLength: 1, reg: gp11, asm: "MOVD"},   // move from arg0
+
+		{name: "SCVTFWS", argLength: 1, reg: gpfp, asm: "SCVTFWS"},   // int32 -> float32
+		{name: "SCVTFWD", argLength: 1, reg: gpfp, asm: "SCVTFWD"},   // int32 -> float64
+		{name: "UCVTFWS", argLength: 1, reg: gpfp, asm: "UCVTFWS"},   // uint32 -> float32
+		{name: "UCVTFWD", argLength: 1, reg: gpfp, asm: "UCVTFWD"},   // uint32 -> float64
+		{name: "SCVTFS", argLength: 1, reg: gpfp, asm: "SCVTFS"},     // int64 -> float32
+		{name: "SCVTFD", argLength: 1, reg: gpfp, asm: "SCVTFD"},     // int64 -> float64
+		{name: "UCVTFS", argLength: 1, reg: gpfp, asm: "UCVTFS"},     // uint64 -> float32
+		{name: "UCVTFD", argLength: 1, reg: gpfp, asm: "UCVTFD"},     // uint64 -> float64
+		{name: "FCVTZSSW", argLength: 1, reg: fpgp, asm: "FCVTZSSW"}, // float32 -> int32
+		{name: "FCVTZSDW", argLength: 1, reg: fpgp, asm: "FCVTZSDW"}, // float64 -> int32
+		{name: "FCVTZUSW", argLength: 1, reg: fpgp, asm: "FCVTZUSW"}, // float32 -> uint32
+		{name: "FCVTZUDW", argLength: 1, reg: fpgp, asm: "FCVTZUDW"}, // float64 -> uint32
+		{name: "FCVTZSS", argLength: 1, reg: fpgp, asm: "FCVTZSS"},   // float32 -> int64
+		{name: "FCVTZSD", argLength: 1, reg: fpgp, asm: "FCVTZSD"},   // float64 -> int64
+		{name: "FCVTZUS", argLength: 1, reg: fpgp, asm: "FCVTZUS"},   // float32 -> uint64
+		{name: "FCVTZUD", argLength: 1, reg: fpgp, asm: "FCVTZUD"},   // float64 -> uint64
+		{name: "FCVTSD", argLength: 1, reg: fp11, asm: "FCVTSD"},     // float32 -> float64
+		{name: "FCVTDS", argLength: 1, reg: fp11, asm: "FCVTDS"},     // float64 -> float32
+
+		// function calls
+		{name: "CALLstatic", argLength: 1, reg: regInfo{clobbers: callerSave}, aux: "SymOff"},                                              // call static function aux.(*gc.Sym).  arg0=mem, auxint=argsize, returns mem
+		{name: "CALLclosure", argLength: 3, reg: regInfo{inputs: []regMask{gpsp, buildReg("R26"), 0}, clobbers: callerSave}, aux: "Int64"}, // call function via closure.  arg0=codeptr, arg1=closure, arg2=mem, auxint=argsize, returns mem
+		{name: "CALLdefer", argLength: 1, reg: regInfo{clobbers: callerSave}, aux: "Int64"},                                                // call deferproc.  arg0=mem, auxint=argsize, returns mem
+		{name: "CALLgo", argLength: 1, reg: regInfo{clobbers: callerSave}, aux: "Int64"},                                                   // call newproc.  arg0=mem, auxint=argsize, returns mem
+		{name: "CALLinter", argLength: 2, reg: regInfo{inputs: []regMask{gp}, clobbers: callerSave}, aux: "Int64"},                         // call fn by pointer.  arg0=codeptr, arg1=mem, auxint=argsize, returns mem
+
+		// pseudo-ops
+		{name: "LoweredNilCheck", argLength: 2, reg: regInfo{inputs: []regMask{gpg}}}, // panic if arg0 is nil.  arg1=mem.
+
+		{name: "Equal", argLength: 1, reg: readflags},         // bool, true flags encode x==y false otherwise.
+		{name: "NotEqual", argLength: 1, reg: readflags},      // bool, true flags encode x!=y false otherwise.
+		{name: "LessThan", argLength: 1, reg: readflags},      // bool, true flags encode signed x<y false otherwise.
+		{name: "LessEqual", argLength: 1, reg: readflags},     // bool, true flags encode signed x<=y false otherwise.
+		{name: "GreaterThan", argLength: 1, reg: readflags},   // bool, true flags encode signed x>y false otherwise.
+		{name: "GreaterEqual", argLength: 1, reg: readflags},  // bool, true flags encode signed x>=y false otherwise.
+		{name: "LessThanU", argLength: 1, reg: readflags},     // bool, true flags encode unsigned x<y false otherwise.
+		{name: "LessEqualU", argLength: 1, reg: readflags},    // bool, true flags encode unsigned x<=y false otherwise.
+		{name: "GreaterThanU", argLength: 1, reg: readflags},  // bool, true flags encode unsigned x>y false otherwise.
+		{name: "GreaterEqualU", argLength: 1, reg: readflags}, // bool, true flags encode unsigned x>=y false otherwise.
+
+		// Scheduler ensures LoweredGetClosurePtr occurs only in entry block,
+		// and sorts it to the very beginning of the block to prevent other
+		// use of R26 (arm64.REGCTXT, the closure pointer)
+		{name: "LoweredGetClosurePtr", reg: regInfo{outputs: []regMask{buildReg("R26")}}},
+
+		// MOVDconvert converts between pointers and integers.
+		// We have a special op for this so as to not confuse GC
+		// (particularly stack maps).  It takes a memory arg so it
+		// gets correctly ordered with respect to GC safepoints.
+		// arg0=ptr/int arg1=mem, output=int/ptr
+		{name: "MOVDconvert", argLength: 2, reg: gp11, asm: "MOVD"},
+
+		// Constant flag values. For any comparison, there are 5 possible
+		// outcomes: the three from the signed total order (<,==,>) and the
+		// three from the unsigned total order. The == cases overlap.
+		// Note: there's a sixth "unordered" outcome for floating-point
+		// comparisons, but we don't use such a beast yet.
+		// These ops are for temporary use by rewrite rules. They
+		// cannot appear in the generated assembly.
+		{name: "FlagEQ"},     // equal
+		{name: "FlagLT_ULT"}, // signed < and unsigned <
+		{name: "FlagLT_UGT"}, // signed < and unsigned >
+		{name: "FlagGT_UGT"}, // signed > and unsigned <
+		{name: "FlagGT_ULT"}, // signed > and unsigned >
+
+		// (InvertFlags (CMP a b)) == (CMP b a)
+		// InvertFlags is a pseudo-op which can't appear in assembly output.
+		{name: "InvertFlags", argLength: 1}, // reverse direction of arg0
+	}
+
+	blocks := []blockData{
+		{name: "EQ"},
+		{name: "NE"},
+		{name: "LT"},
+		{name: "LE"},
+		{name: "GT"},
+		{name: "GE"},
+		{name: "ULT"},
+		{name: "ULE"},
+		{name: "UGT"},
+		{name: "UGE"},
+	}
+
+	archs = append(archs, arch{
+		name:            "ARM64",
+		pkg:             "cmd/internal/obj/arm64",
+		genfile:         "../../arm64/ssa.go",
+		ops:             ops,
+		blocks:          blocks,
+		regnames:        regNamesARM64,
+		gpregmask:       gp,
+		fpregmask:       fp,
+		flagmask:        flags,
+		framepointerreg: -1, // not used
+	})
+}
diff --git a/src/cmd/compile/internal/ssa/opGen.go b/src/cmd/compile/internal/ssa/opGen.go
index 60350ebaee..f898931479 100644
--- a/src/cmd/compile/internal/ssa/opGen.go
+++ b/src/cmd/compile/internal/ssa/opGen.go
@@ -6,6 +6,7 @@ package ssa
 import (
 	"cmd/internal/obj"
 	"cmd/internal/obj/arm"
+	"cmd/internal/obj/arm64"
 	"cmd/internal/obj/ppc64"
 	"cmd/internal/obj/x86"
 )
@@ -54,6 +55,17 @@ const (
 	BlockARMUGT
 	BlockARMUGE
 
+	BlockARM64EQ
+	BlockARM64NE
+	BlockARM64LT
+	BlockARM64LE
+	BlockARM64GT
+	BlockARM64GE
+	BlockARM64ULT
+	BlockARM64ULE
+	BlockARM64UGT
+	BlockARM64UGE
+
 	BlockPPC64EQ
 	BlockPPC64NE
 	BlockPPC64LT
@@ -120,6 +132,17 @@ var blockString = [...]string{
 	BlockARMUGT: "UGT",
 	BlockARMUGE: "UGE",
 
+	BlockARM64EQ:  "EQ",
+	BlockARM64NE:  "NE",
+	BlockARM64LT:  "LT",
+	BlockARM64LE:  "LE",
+	BlockARM64GT:  "GT",
+	BlockARM64GE:  "GE",
+	BlockARM64ULT: "ULT",
+	BlockARM64ULE: "ULE",
+	BlockARM64UGT: "UGT",
+	BlockARM64UGE: "UGE",
+
 	BlockPPC64EQ:  "EQ",
 	BlockPPC64NE:  "NE",
 	BlockPPC64LT:  "LT",
@@ -774,6 +797,119 @@ const (
 	OpARMFlagGT_ULT
 	OpARMInvertFlags
 
+	OpARM64ADD
+	OpARM64ADDconst
+	OpARM64SUB
+	OpARM64SUBconst
+	OpARM64MUL
+	OpARM64DIV
+	OpARM64UDIV
+	OpARM64DIVW
+	OpARM64UDIVW
+	OpARM64MOD
+	OpARM64UMOD
+	OpARM64MODW
+	OpARM64UMODW
+	OpARM64FADDS
+	OpARM64FADDD
+	OpARM64FSUBS
+	OpARM64FSUBD
+	OpARM64FMULS
+	OpARM64FMULD
+	OpARM64FDIVS
+	OpARM64FDIVD
+	OpARM64AND
+	OpARM64ANDconst
+	OpARM64OR
+	OpARM64ORconst
+	OpARM64XOR
+	OpARM64XORconst
+	OpARM64BIC
+	OpARM64BICconst
+	OpARM64MVN
+	OpARM64NEG
+	OpARM64FNEGS
+	OpARM64FNEGD
+	OpARM64FSQRTD
+	OpARM64CMP
+	OpARM64CMPconst
+	OpARM64CMPW
+	OpARM64CMPWconst
+	OpARM64CMN
+	OpARM64CMNconst
+	OpARM64CMNW
+	OpARM64CMNWconst
+	OpARM64FCMPS
+	OpARM64FCMPD
+	OpARM64MOVDconst
+	OpARM64FMOVSconst
+	OpARM64FMOVDconst
+	OpARM64MOVDaddr
+	OpARM64MOVBload
+	OpARM64MOVBUload
+	OpARM64MOVHload
+	OpARM64MOVHUload
+	OpARM64MOVWload
+	OpARM64MOVWUload
+	OpARM64MOVDload
+	OpARM64FMOVSload
+	OpARM64FMOVDload
+	OpARM64MOVBstore
+	OpARM64MOVHstore
+	OpARM64MOVWstore
+	OpARM64MOVDstore
+	OpARM64FMOVSstore
+	OpARM64FMOVDstore
+	OpARM64MOVBreg
+	OpARM64MOVBUreg
+	OpARM64MOVHreg
+	OpARM64MOVHUreg
+	OpARM64MOVWreg
+	OpARM64MOVWUreg
+	OpARM64MOVDreg
+	OpARM64SCVTFWS
+	OpARM64SCVTFWD
+	OpARM64UCVTFWS
+	OpARM64UCVTFWD
+	OpARM64SCVTFS
+	OpARM64SCVTFD
+	OpARM64UCVTFS
+	OpARM64UCVTFD
+	OpARM64FCVTZSSW
+	OpARM64FCVTZSDW
+	OpARM64FCVTZUSW
+	OpARM64FCVTZUDW
+	OpARM64FCVTZSS
+	OpARM64FCVTZSD
+	OpARM64FCVTZUS
+	OpARM64FCVTZUD
+	OpARM64FCVTSD
+	OpARM64FCVTDS
+	OpARM64CALLstatic
+	OpARM64CALLclosure
+	OpARM64CALLdefer
+	OpARM64CALLgo
+	OpARM64CALLinter
+	OpARM64LoweredNilCheck
+	OpARM64Equal
+	OpARM64NotEqual
+	OpARM64LessThan
+	OpARM64LessEqual
+	OpARM64GreaterThan
+	OpARM64GreaterEqual
+	OpARM64LessThanU
+	OpARM64LessEqualU
+	OpARM64GreaterThanU
+	OpARM64GreaterEqualU
+	OpARM64LoweredGetClosurePtr
+	OpARM64MOVDconvert
+	OpARM64FlagEQ
+	OpARM64FlagLT_ULT
+	OpARM64FlagLT_UGT
+	OpARM64FlagGT_UGT
+	OpARM64FlagGT_ULT
+	OpARM64InvertFlags
+
 	OpPPC64ADD
 	OpPPC64ADDconst
 	OpPPC64FADD
@@ -9825,6 +9961,1429 @@ var opcodeTable = [...]opInfo{
 		name:        "ADD",
 		argLen:      2,
 		commutative: true,
+		asm:         arm64.AADD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "ADDconst",
+		auxType: auxInt32,
+		argLen:  1,
+		asm:     arm64.AADD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 1073741823}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "SUB",
+		argLen: 2,
+		asm:    arm64.ASUB,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "SUBconst",
+		auxType: auxInt32,
+		argLen:  1,
+		asm:     arm64.ASUB,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:        "MUL",
+		argLen:      2,
+		commutative: true,
+		asm:         arm64.AMUL,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "DIV",
+		argLen: 2,
+		asm:    arm64.ASDIV,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "UDIV",
+		argLen: 2,
+		asm:    arm64.AUDIV,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "DIVW",
+		argLen: 2,
+		asm:    arm64.ASDIVW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "UDIVW",
+		argLen: 2,
+		asm:    arm64.AUDIVW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "MOD",
+		argLen: 2,
+		asm:    arm64.AREM,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "UMOD",
+		argLen: 2,
+		asm:    arm64.AUREM,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "MODW",
+		argLen: 2,
+		asm:    arm64.AREMW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "UMODW",
+		argLen: 2,
+		asm:    arm64.AUREMW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:        "FADDS",
+		argLen:      2,
+		commutative: true,
+		asm:         arm64.AFADDS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{1, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:        "FADDD",
+		argLen:      2,
+		commutative: true,
+		asm:         arm64.AFADDD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{1, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "FSUBS",
+		argLen: 2,
+		asm:    arm64.AFSUBS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{1, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "FSUBD",
+		argLen: 2,
+		asm:    arm64.AFSUBD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{1, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:        "FMULS",
+		argLen:      2,
+		commutative: true,
+		asm:         arm64.AFMULS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{1, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:        "FMULD",
+		argLen:      2,
+		commutative: true,
+		asm:         arm64.AFMULD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{1, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "FDIVS",
+		argLen: 2,
+		asm:    arm64.AFDIVS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{1, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "FDIVD",
+		argLen: 2,
+		asm:    arm64.AFDIVD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{1, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:        "AND",
+		argLen:      2,
+		commutative: true,
+		asm:         arm64.AAND,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "ANDconst",
+		auxType: auxInt32,
+		argLen:  1,
+		asm:     arm64.AAND,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:        "OR",
+		argLen:      2,
+		commutative: true,
+		asm:         arm64.AORR,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "ORconst",
+		auxType: auxInt32,
+		argLen:  1,
+		asm:     arm64.AORR,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:        "XOR",
+		argLen:      2,
+		commutative: true,
+		asm:         arm64.AEOR,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "XORconst",
+		auxType: auxInt32,
+		argLen:  1,
+		asm:     arm64.AEOR,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "BIC",
+		argLen: 2,
+		asm:    arm64.ABIC,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "BICconst",
+		auxType: auxInt32,
+		argLen:  1,
+		asm:     arm64.ABIC,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "MVN",
+		argLen: 1,
+		asm:    arm64.AMVN,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "NEG",
+		argLen: 1,
+		asm:    arm64.ANEG,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "FNEGS",
+		argLen: 1,
+		asm:    arm64.AFNEGS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "FNEGD",
+		argLen: 1,
+		asm:    arm64.AFNEGD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "FSQRTD",
+		argLen: 1,
+		asm:    arm64.AFSQRTD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "CMP",
+		argLen: 2,
+		asm:    arm64.ACMP,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+		},
+	},
+	{
+		name:    "CMPconst",
+		auxType: auxInt32,
+		argLen:  1,
+		asm:     arm64.ACMP,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+		},
+	},
+	{
+		name:   "CMPW",
+		argLen: 2,
+		asm:    arm64.ACMPW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+		},
+	},
+	{
+		name:    "CMPWconst",
+		auxType: auxInt32,
+		argLen:  1,
+		asm:     arm64.ACMPW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+		},
+	},
+	{
+		name:   "CMN",
+		argLen: 2,
+		asm:    arm64.ACMN,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+		},
+	},
+	{
+		name:    "CMNconst",
+		auxType: auxInt32,
+		argLen:  1,
+		asm:     arm64.ACMN,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+		},
+	},
+	{
+		name:   "CMNW",
+		argLen: 2,
+		asm:    arm64.ACMNW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{1, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+		},
+	},
+	{
+		name:    "CMNWconst",
+		auxType: auxInt32,
+		argLen:  1,
+		asm:     arm64.ACMNW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+		},
+	},
+	{
+		name:   "FCMPS",
+		argLen: 2,
+		asm:    arm64.AFCMPS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{1, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+		},
+	},
+	{
+		name:   "FCMPD",
+		argLen: 2,
+		asm:    arm64.AFCMPD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{1, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+		},
+	},
+	{
+		name:              "MOVDconst",
+		auxType:           auxInt32,
+		argLen:            0,
+		rematerializeable: true,
+		asm:               arm64.AMOVD,
+		reg: regInfo{
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:              "FMOVSconst",
+		auxType:           auxFloat64,
+		argLen:            0,
+		rematerializeable: true,
+		asm:               arm64.AFMOVS,
+		reg: regInfo{
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:              "FMOVDconst",
+		auxType:           auxFloat64,
+		argLen:            0,
+		rematerializeable: true,
+		asm:               arm64.AFMOVD,
+		reg: regInfo{
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:              "MOVDaddr",
+		auxType:           auxSymOff,
+		argLen:            1,
+		rematerializeable: true,
+		asm:               arm64.AMOVD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 9223372037391646720}, // SP SB
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "MOVBload",
+		auxType: auxSymOff,
+		argLen:  2,
+		asm:     arm64.AMOVB,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "MOVBUload",
+		auxType: auxSymOff,
+		argLen:  2,
+		asm:     arm64.AMOVBU,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "MOVHload",
+		auxType: auxSymOff,
+		argLen:  2,
+		asm:     arm64.AMOVH,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "MOVHUload",
+		auxType: auxSymOff,
+		argLen:  2,
+		asm:     arm64.AMOVHU,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "MOVWload",
+		auxType: auxSymOff,
+		argLen:  2,
+		asm:     arm64.AMOVW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "MOVWUload",
+		auxType: auxSymOff,
+		argLen:  2,
+		asm:     arm64.AMOVWU,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "MOVDload",
+		auxType: auxSymOff,
+		argLen:  2,
+		asm:     arm64.AMOVD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:    "FMOVSload",
+		auxType: auxSymOff,
+		argLen:  2,
+		asm:     arm64.AFMOVS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:    "FMOVDload",
+		auxType: auxSymOff,
+		argLen:  2,
+		asm:     arm64.AFMOVD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:    "MOVBstore",
+		auxType: auxSymOff,
+		argLen:  3,
+		asm:     arm64.AMOVB,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{1, 536870911},           // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+		},
+	},
+	{
+		name:    "MOVHstore",
+		auxType: auxSymOff,
+		argLen:  3,
+		asm:     arm64.AMOVH,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{1, 536870911},           // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+		},
+	},
+	{
+		name:    "MOVWstore",
+		auxType: auxSymOff,
+		argLen:  3,
+		asm:     arm64.AMOVW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{1, 536870911},           // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+		},
+	},
+	{
+		name:    "MOVDstore",
+		auxType: auxSymOff,
+		argLen:  3,
+		asm:     arm64.AMOVD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{1, 536870911},           // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+		},
+	},
+	{
+		name:    "FMOVSstore",
+		auxType: auxSymOff,
+		argLen:  3,
+		asm:     arm64.AFMOVS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{1, 288230375077969920},  // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+		},
+	},
+	{
+		name:    "FMOVDstore",
+		auxType: auxSymOff,
+		argLen:  3,
+		asm:     arm64.AFMOVD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{1, 288230375077969920},  // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+				{0, 9223372037928517631}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 SP SB
+			},
+		},
+	},
+	{
+		name:   "MOVBreg",
+		argLen: 1,
+		asm:    arm64.AMOVB,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "MOVBUreg",
+		argLen: 1,
+		asm:    arm64.AMOVBU,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "MOVHreg",
+		argLen: 1,
+		asm:    arm64.AMOVH,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "MOVHUreg",
+		argLen: 1,
+		asm:    arm64.AMOVHU,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "MOVWreg",
+		argLen: 1,
+		asm:    arm64.AMOVW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "MOVWUreg",
+		argLen: 1,
+		asm:    arm64.AMOVWU,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "MOVDreg",
+		argLen: 1,
+		asm:    arm64.AMOVD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "SCVTFWS",
+		argLen: 1,
+		asm:    arm64.ASCVTFWS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "SCVTFWD",
+		argLen: 1,
+		asm:    arm64.ASCVTFWD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "UCVTFWS",
+		argLen: 1,
+		asm:    arm64.AUCVTFWS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "UCVTFWD",
+		argLen: 1,
+		asm:    arm64.AUCVTFWD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "SCVTFS",
+		argLen: 1,
+		asm:    arm64.ASCVTFS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "SCVTFD",
+		argLen: 1,
+		asm:    arm64.ASCVTFD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "UCVTFS",
+		argLen: 1,
+		asm:    arm64.AUCVTFS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "UCVTFD",
+		argLen: 1,
+		asm:    arm64.AUCVTFD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "FCVTZSSW",
+		argLen: 1,
+		asm:    arm64.AFCVTZSSW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "FCVTZSDW",
+		argLen: 1,
+		asm:    arm64.AFCVTZSDW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "FCVTZUSW",
+		argLen: 1,
+		asm:    arm64.AFCVTZUSW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "FCVTZUDW",
+		argLen: 1,
+		asm:    arm64.AFCVTZUDW,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "FCVTZSS",
+		argLen: 1,
+		asm:    arm64.AFCVTZSS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "FCVTZSD",
+		argLen: 1,
+		asm:    arm64.AFCVTZSD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "FCVTZUS",
+		argLen: 1,
+		asm:    arm64.AFCVTZUS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "FCVTZUD",
+		argLen: 1,
+		asm:    arm64.AFCVTZUD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "FCVTSD",
+		argLen: 1,
+		asm:    arm64.AFCVTSD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:   "FCVTDS",
+		argLen: 1,
+		asm:    arm64.AFCVTDS,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+			outputs: []outputInfo{
+				{0, 288230375077969920}, // F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27
+			},
+		},
+	},
+	{
+		name:    "CALLstatic",
+		auxType: auxSymOff,
+		argLen:  1,
+		reg: regInfo{
+			clobbers: 4899916394042228735, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27 FLAGS
+		},
+	},
+	{
+		name:    "CALLclosure",
+		auxType: auxInt64,
+		argLen:  3,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{1, 67108864},  // R26
+				{0, 939524095}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29 SP
+			},
+			clobbers: 4899916394042228735, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27 FLAGS
+		},
+	},
+	{
+		name:    "CALLdefer",
+		auxType: auxInt64,
+		argLen:  1,
+		reg: regInfo{
+			clobbers: 4899916394042228735, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27 FLAGS
+		},
+	},
+	{
+		name:    "CALLgo",
+		auxType: auxInt64,
+		argLen:  1,
+		reg: regInfo{
+			clobbers: 4899916394042228735, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27 FLAGS
+		},
+	},
+	{
+		name:    "CALLinter",
+		auxType: auxInt64,
+		argLen:  2,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+			clobbers: 4899916394042228735, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29 F0 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 F21 F22 F23 F24 F25 F26 F27 FLAGS
+		},
+	},
+	{
+		name:   "LoweredNilCheck",
+		argLen: 2,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+		},
+	},
+	{
+		name:   "Equal",
+		argLen: 1,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "NotEqual",
+		argLen: 1,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "LessThan",
+		argLen: 1,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "LessEqual",
+		argLen: 1,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "GreaterThan",
+		argLen: 1,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "GreaterEqual",
+		argLen: 1,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "LessThanU",
+		argLen: 1,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "LessEqualU",
+		argLen: 1,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "GreaterThanU",
+		argLen: 1,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "GreaterEqualU",
+		argLen: 1,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 4611686018427387904}, // FLAGS
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "LoweredGetClosurePtr",
+		argLen: 0,
+		reg: regInfo{
+			outputs: []outputInfo{
+				{0, 67108864}, // R26
+			},
+		},
+	},
+	{
+		name:   "MOVDconvert",
+		argLen: 2,
+		asm:    arm64.AMOVD,
+		reg: regInfo{
+			inputs: []inputInfo{
+				{0, 536870911}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 g R29
+			},
+			outputs: []outputInfo{
+				{0, 402653183}, // R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R29
+			},
+		},
+	},
+	{
+		name:   "FlagEQ",
+		argLen: 0,
+		reg:    regInfo{},
+	},
+	{
+		name:   "FlagLT_ULT",
+		argLen: 0,
+		reg:    regInfo{},
+	},
+	{
+		name:   "FlagLT_UGT",
+		argLen: 0,
+		reg:    regInfo{},
+	},
+	{
+		name:   "FlagGT_UGT",
+		argLen: 0,
+		reg:    regInfo{},
+	},
+	{
+		name:   "FlagGT_ULT",
+		argLen: 0,
+		reg:    regInfo{},
+	},
+	{
+		name:   "InvertFlags",
+		argLen: 1,
+		reg:    regInfo{},
+	},
+
+	{
+		name:        "ADD",
+		argLen:      2,
+		commutative: true,
 		asm:         ppc64.AADD,
 		reg: regInfo{
 			inputs: []inputInfo{
@@ -12354,6 +13913,76 @@ var gpRegMaskARM = regMask(5119)
 var fpRegMaskARM = regMask(4294901760)
 var flagRegMaskARM = regMask(4294967296)
 var framepointerRegARM = int8(-1)
+var registersARM64 = [...]Register{
+	{0, "R0"},
+	{1, "R1"},
+	{2, "R2"},
+	{3, "R3"},
+	{4, "R4"},
+	{5, "R5"},
+	{6, "R6"},
+	{7, "R7"},
+	{8, "R8"},
+	{9, "R9"},
+	{10, "R10"},
+	{11, "R11"},
+	{12, "R12"},
+	{13, "R13"},
+	{14, "R14"},
+	{15, "R15"},
+	{16, "R16"},
+	{17, "R17"},
+	{18, "R18"},
+	{19, "R19"},
+	{20, "R20"},
+	{21, "R21"},
+	{22, "R22"},
+	{23, "R23"},
+	{24, "R24"},
+	{25, "R25"},
+	{26, "R26"},
+	{27, "g"},
+	{28, "R29"},
+	{29, "SP"},
+	{30, "F0"},
+	{31, "F1"},
+	{32, "F2"},
+	{33, "F3"},
+	{34, "F4"},
+	{35, "F5"},
+	{36, "F6"},
+	{37, "F7"},
+	{38, "F8"},
+	{39, "F9"},
+	{40, "F10"},
+	{41, "F11"},
+	{42, "F12"},
+	{43, "F13"},
+	{44, "F14"},
+	{45, "F15"},
+	{46, "F16"},
+	{47, "F17"},
+	{48, "F18"},
+	{49, "F19"},
+	{50, "F20"},
+	{51, "F21"},
+	{52, "F22"},
+	{53, "F23"},
+	{54, "F24"},
+	{55, "F25"},
+	{56, "F26"},
+	{57, "F27"},
+	{58, "F28"},
+	{59, "F29"},
+	{60, "F30"},
+	{61, "F31"},
+	{62, "FLAGS"},
+	{63, "SB"},
+}
+var gpRegMaskARM64 = regMask(402653183)
+var fpRegMaskARM64 = regMask(288230375077969920)
+var flagRegMaskARM64 = regMask(4611686018427387904)
+var framepointerRegARM64 = int8(-1)
 var registersPPC64 = [...]Register{
 	{0, "R0"},
 	{1, "SP"},
diff --git a/src/cmd/compile/internal/ssa/rewriteARM64.go b/src/cmd/compile/internal/ssa/rewriteARM64.go
new file mode 100644
index 0000000000..72717f66be
--- /dev/null
+++ b/src/cmd/compile/internal/ssa/rewriteARM64.go
@@ -0,0 +1,4607 @@
+// autogenerated from gen/ARM64.rules: do not edit!
+// generated with: cd gen; go run *.go
+
+package ssa
+
+import "math"
+
+var _ = math.MinInt8 // in case not otherwise used
+func rewriteValueARM64(v *Value, config *Config) bool {
+	switch v.Op {
+	case OpARM64ADDconst:
+		return rewriteValueARM64_OpARM64ADDconst(v, config)
+	case OpAdd16:
+		return rewriteValueARM64_OpAdd16(v, config)
+	case OpAdd32:
+		return rewriteValueARM64_OpAdd32(v, config)
+	case OpAdd32F:
+		return rewriteValueARM64_OpAdd32F(v, config)
+	case OpAdd64:
+		return rewriteValueARM64_OpAdd64(v, config)
+	case OpAdd64F:
+		return rewriteValueARM64_OpAdd64F(v, config)
+	case OpAdd8:
+		return rewriteValueARM64_OpAdd8(v, config)
+	case OpAddPtr:
+		return rewriteValueARM64_OpAddPtr(v, config)
+	case OpAddr:
+		return rewriteValueARM64_OpAddr(v, config)
+	case OpAnd16:
+		return rewriteValueARM64_OpAnd16(v, config)
+	case OpAnd32:
+		return rewriteValueARM64_OpAnd32(v, config)
+	case OpAnd64:
+		return rewriteValueARM64_OpAnd64(v, config)
+	case OpAnd8:
+		return rewriteValueARM64_OpAnd8(v, config)
+	case OpAndB:
+		return rewriteValueARM64_OpAndB(v, config)
+	case OpClosureCall:
+		return rewriteValueARM64_OpClosureCall(v, config)
+	case OpCom16:
+		return rewriteValueARM64_OpCom16(v, config)
+	case OpCom32:
+		return rewriteValueARM64_OpCom32(v, config)
+	case OpCom64:
+		return rewriteValueARM64_OpCom64(v, config)
+	case OpCom8:
+		return rewriteValueARM64_OpCom8(v, config)
+	case OpConst16:
+		return rewriteValueARM64_OpConst16(v, config)
+	case OpConst32:
+		return rewriteValueARM64_OpConst32(v, config)
+	case OpConst32F:
+		return rewriteValueARM64_OpConst32F(v, config)
+	case OpConst64:
+		return rewriteValueARM64_OpConst64(v, config)
+	case OpConst64F:
+		return rewriteValueARM64_OpConst64F(v, config)
+	case OpConst8:
+		return rewriteValueARM64_OpConst8(v, config)
+	case OpConstBool:
+		return rewriteValueARM64_OpConstBool(v, config)
+	case OpConstNil:
+		return rewriteValueARM64_OpConstNil(v, config)
+	case OpConvert:
+		return rewriteValueARM64_OpConvert(v, config)
+	case OpCvt32Fto32:
+		return rewriteValueARM64_OpCvt32Fto32(v, config)
+	case OpCvt32Fto32U:
+		return rewriteValueARM64_OpCvt32Fto32U(v, config)
+	case OpCvt32Fto64:
+		return rewriteValueARM64_OpCvt32Fto64(v, config)
+	case OpCvt32Fto64F:
+		return rewriteValueARM64_OpCvt32Fto64F(v, config)
+	case OpCvt32Uto32F:
+		return rewriteValueARM64_OpCvt32Uto32F(v, config)
+	case OpCvt32Uto64F:
+		return rewriteValueARM64_OpCvt32Uto64F(v, config)
+	case OpCvt32to32F:
+		return rewriteValueARM64_OpCvt32to32F(v, config)
+	case OpCvt32to64F:
+		return rewriteValueARM64_OpCvt32to64F(v, config)
+	case OpCvt64Fto32:
+		return rewriteValueARM64_OpCvt64Fto32(v, config)
+	case OpCvt64Fto32F:
+		return rewriteValueARM64_OpCvt64Fto32F(v, config)
+	case OpCvt64Fto32U:
+		return rewriteValueARM64_OpCvt64Fto32U(v, config)
+	case OpCvt64Fto64:
+		return rewriteValueARM64_OpCvt64Fto64(v, config)
+	case OpCvt64to32F:
+		return rewriteValueARM64_OpCvt64to32F(v, config)
+	case OpCvt64to64F:
+		return rewriteValueARM64_OpCvt64to64F(v, config)
+	case OpDeferCall:
+		return rewriteValueARM64_OpDeferCall(v, config)
+	case OpDiv16:
+		return rewriteValueARM64_OpDiv16(v, config)
+	case OpDiv16u:
+		return rewriteValueARM64_OpDiv16u(v, config)
+	case OpDiv32:
+		return rewriteValueARM64_OpDiv32(v, config)
+	case OpDiv32F:
+		return rewriteValueARM64_OpDiv32F(v, config)
+	case OpDiv32u:
+		return rewriteValueARM64_OpDiv32u(v, config)
+	case OpDiv64:
+		return rewriteValueARM64_OpDiv64(v, config)
+	case OpDiv64F:
+		return rewriteValueARM64_OpDiv64F(v, config)
+	case OpDiv64u:
+		return rewriteValueARM64_OpDiv64u(v, config)
+	case OpDiv8:
+		return rewriteValueARM64_OpDiv8(v, config)
+	case OpDiv8u:
+		return rewriteValueARM64_OpDiv8u(v, config)
+	case OpEq16:
+		return rewriteValueARM64_OpEq16(v, config)
+	case OpEq32:
+		return rewriteValueARM64_OpEq32(v, config)
+	case OpEq32F:
+		return rewriteValueARM64_OpEq32F(v, config)
+	case OpEq64:
+		return rewriteValueARM64_OpEq64(v, config)
+	case OpEq64F:
+		return rewriteValueARM64_OpEq64F(v, config)
+	case OpEq8:
+		return rewriteValueARM64_OpEq8(v, config)
+	case OpEqB:
+		return rewriteValueARM64_OpEqB(v, config)
+	case OpEqPtr:
+		return rewriteValueARM64_OpEqPtr(v, config)
+	case OpARM64FMOVDload:
+		return rewriteValueARM64_OpARM64FMOVDload(v, config)
+	case OpARM64FMOVDstore:
+		return rewriteValueARM64_OpARM64FMOVDstore(v, config)
+	case OpARM64FMOVSload:
+		return rewriteValueARM64_OpARM64FMOVSload(v, config)
+	case OpARM64FMOVSstore:
+		return rewriteValueARM64_OpARM64FMOVSstore(v, config)
+	case OpGeq16:
+		return rewriteValueARM64_OpGeq16(v, config)
+	case OpGeq16U:
+		return rewriteValueARM64_OpGeq16U(v, config)
+	case OpGeq32:
+		return rewriteValueARM64_OpGeq32(v, config)
+	case OpGeq32F:
+		return rewriteValueARM64_OpGeq32F(v, config)
+	case OpGeq32U:
+		return rewriteValueARM64_OpGeq32U(v, config)
+	case OpGeq64:
+		return rewriteValueARM64_OpGeq64(v, config)
+	case OpGeq64F:
+		return rewriteValueARM64_OpGeq64F(v, config)
+	case OpGeq64U:
+		return rewriteValueARM64_OpGeq64U(v, config)
+	case OpGeq8:
+		return rewriteValueARM64_OpGeq8(v, config)
+	case OpGeq8U:
+		return rewriteValueARM64_OpGeq8U(v, config)
+	case OpGetClosurePtr:
+		return rewriteValueARM64_OpGetClosurePtr(v, config)
+	case OpGoCall:
+		return rewriteValueARM64_OpGoCall(v, config)
+	case OpGreater16:
+		return rewriteValueARM64_OpGreater16(v, config)
+	case OpGreater16U:
+		return rewriteValueARM64_OpGreater16U(v, config)
+	case OpGreater32:
+		return rewriteValueARM64_OpGreater32(v, config)
+	case OpGreater32F:
+		return rewriteValueARM64_OpGreater32F(v, config)
+	case OpGreater32U:
+		return rewriteValueARM64_OpGreater32U(v, config)
+	case OpGreater64:
+		return rewriteValueARM64_OpGreater64(v, config)
+	case OpGreater64F:
+		return rewriteValueARM64_OpGreater64F(v, config)
+	case OpGreater64U:
+		return rewriteValueARM64_OpGreater64U(v, config)
+	case OpGreater8:
+		return rewriteValueARM64_OpGreater8(v, config)
+	case OpGreater8U:
+		return rewriteValueARM64_OpGreater8U(v, config)
+	case OpInterCall:
+		return rewriteValueARM64_OpInterCall(v, config)
+	case OpIsInBounds:
+		return rewriteValueARM64_OpIsInBounds(v, config)
+	case OpIsNonNil:
+		return rewriteValueARM64_OpIsNonNil(v, config)
+	case OpIsSliceInBounds:
+		return rewriteValueARM64_OpIsSliceInBounds(v, config)
+	case OpLeq16:
+		return rewriteValueARM64_OpLeq16(v, config)
+	case OpLeq16U:
+		return rewriteValueARM64_OpLeq16U(v, config)
+	case OpLeq32:
+		return rewriteValueARM64_OpLeq32(v, config)
+	case OpLeq32F:
+		return rewriteValueARM64_OpLeq32F(v, config)
+	case OpLeq32U:
+		return rewriteValueARM64_OpLeq32U(v, config)
+	case OpLeq64:
+		return rewriteValueARM64_OpLeq64(v, config)
+	case OpLeq64F:
+		return rewriteValueARM64_OpLeq64F(v, config)
+	case OpLeq64U:
+		return rewriteValueARM64_OpLeq64U(v, config)
+	case OpLeq8:
+		return rewriteValueARM64_OpLeq8(v, config)
+	case OpLeq8U:
+		return rewriteValueARM64_OpLeq8U(v, config)
+	case OpLess16:
+		return rewriteValueARM64_OpLess16(v, config)
+	case OpLess16U:
+		return rewriteValueARM64_OpLess16U(v, config)
+	case OpLess32:
+		return rewriteValueARM64_OpLess32(v, config)
+	case OpLess32F:
+		return rewriteValueARM64_OpLess32F(v, config)
+	case OpLess32U:
+		return rewriteValueARM64_OpLess32U(v, config)
+	case OpLess64:
+		return rewriteValueARM64_OpLess64(v, config)
+	case OpLess64F:
+		return rewriteValueARM64_OpLess64F(v, config)
+	case OpLess64U:
+		return rewriteValueARM64_OpLess64U(v, config)
+	case OpLess8:
+		return rewriteValueARM64_OpLess8(v, config)
+	case OpLess8U:
+		return rewriteValueARM64_OpLess8U(v, config)
+	case OpLoad:
+		return rewriteValueARM64_OpLoad(v, config)
+	case OpARM64MOVBUload:
+		return rewriteValueARM64_OpARM64MOVBUload(v, config)
+	case OpARM64MOVBload:
+		return rewriteValueARM64_OpARM64MOVBload(v, config)
+	case OpARM64MOVBstore:
+		return rewriteValueARM64_OpARM64MOVBstore(v, config)
+	case OpARM64MOVDload:
+		return rewriteValueARM64_OpARM64MOVDload(v, config)
+	case OpARM64MOVDstore:
+		return rewriteValueARM64_OpARM64MOVDstore(v, config)
+	case OpARM64MOVHUload:
+		return rewriteValueARM64_OpARM64MOVHUload(v, config)
+	case OpARM64MOVHload:
+		return rewriteValueARM64_OpARM64MOVHload(v, config)
+	case OpARM64MOVHstore:
+		return rewriteValueARM64_OpARM64MOVHstore(v, config)
+	case OpARM64MOVWUload:
+		return rewriteValueARM64_OpARM64MOVWUload(v, config)
+	case OpARM64MOVWload:
+		return rewriteValueARM64_OpARM64MOVWload(v, config)
+	case OpARM64MOVWstore:
+		return rewriteValueARM64_OpARM64MOVWstore(v, config)
+	case OpMod16:
+		return rewriteValueARM64_OpMod16(v, config)
+	case OpMod16u:
+		return rewriteValueARM64_OpMod16u(v, config)
+	case OpMod32:
+		return rewriteValueARM64_OpMod32(v, config)
+	case OpMod32u:
+		return rewriteValueARM64_OpMod32u(v, config)
+	case OpMod64:
+		return rewriteValueARM64_OpMod64(v, config)
+	case OpMod64u:
+		return rewriteValueARM64_OpMod64u(v, config)
+	case OpMod8:
+		return rewriteValueARM64_OpMod8(v, config)
+	case OpMod8u:
+		return rewriteValueARM64_OpMod8u(v, config)
+	case OpMul16:
+		return rewriteValueARM64_OpMul16(v, config)
+	case OpMul32:
+		return rewriteValueARM64_OpMul32(v, config)
+	case OpMul32F:
+		return rewriteValueARM64_OpMul32F(v, config)
+	case OpMul64:
+		return rewriteValueARM64_OpMul64(v, config)
+	case OpMul64F:
+		return rewriteValueARM64_OpMul64F(v, config)
+	case OpMul8:
+		return rewriteValueARM64_OpMul8(v, config)
+	case OpNeg16:
+		return rewriteValueARM64_OpNeg16(v, config)
+	case OpNeg32:
+		return rewriteValueARM64_OpNeg32(v, config)
+	case OpNeg32F:
+		return rewriteValueARM64_OpNeg32F(v, config)
+	case OpNeg64:
+		return rewriteValueARM64_OpNeg64(v, config)
+	case OpNeg64F:
+		return rewriteValueARM64_OpNeg64F(v, config)
+	case OpNeg8:
+		return rewriteValueARM64_OpNeg8(v, config)
+	case OpNeq16:
+		return rewriteValueARM64_OpNeq16(v, config)
+	case OpNeq32:
+		return rewriteValueARM64_OpNeq32(v, config)
+	case OpNeq32F:
+		return rewriteValueARM64_OpNeq32F(v, config)
+	case OpNeq64:
+		return rewriteValueARM64_OpNeq64(v, config)
+	case OpNeq64F:
+		return rewriteValueARM64_OpNeq64F(v, config)
+	case OpNeq8:
+		return rewriteValueARM64_OpNeq8(v, config)
+	case OpNeqB:
+		return rewriteValueARM64_OpNeqB(v, config)
+	case OpNeqPtr:
+		return rewriteValueARM64_OpNeqPtr(v, config)
+	case OpNilCheck:
+		return rewriteValueARM64_OpNilCheck(v, config)
+	case OpNot:
+		return rewriteValueARM64_OpNot(v, config)
+	case OpOffPtr:
+		return rewriteValueARM64_OpOffPtr(v, config)
+	case OpOr16:
+		return rewriteValueARM64_OpOr16(v, config)
+	case OpOr32:
+		return rewriteValueARM64_OpOr32(v, config)
+	case OpOr64:
+		return rewriteValueARM64_OpOr64(v, config)
+	case OpOr8:
+		return rewriteValueARM64_OpOr8(v, config)
+	case OpOrB:
+		return rewriteValueARM64_OpOrB(v, config)
+	case OpSignExt16to32:
+		return rewriteValueARM64_OpSignExt16to32(v, config)
+	case OpSignExt16to64:
+		return rewriteValueARM64_OpSignExt16to64(v, config)
+	case OpSignExt32to64:
+		return rewriteValueARM64_OpSignExt32to64(v, config)
+	case OpSignExt8to16:
+		return rewriteValueARM64_OpSignExt8to16(v, config)
+	case OpSignExt8to32:
+		return rewriteValueARM64_OpSignExt8to32(v, config)
+	case OpSignExt8to64:
+		return rewriteValueARM64_OpSignExt8to64(v, config)
+	case OpStaticCall:
+		return rewriteValueARM64_OpStaticCall(v, config)
+	case OpStore:
+		return rewriteValueARM64_OpStore(v, config)
+	case OpSub16:
+		return rewriteValueARM64_OpSub16(v, config)
+	case OpSub32:
+		return rewriteValueARM64_OpSub32(v, config)
+	case OpSub32F:
+		return rewriteValueARM64_OpSub32F(v, config)
+	case OpSub64:
+		return rewriteValueARM64_OpSub64(v, config)
+	case OpSub64F:
+		return rewriteValueARM64_OpSub64F(v, config)
+	case OpSub8:
+		return rewriteValueARM64_OpSub8(v, config)
+	case OpSubPtr:
+		return rewriteValueARM64_OpSubPtr(v, config)
+	case OpTrunc16to8:
+		return rewriteValueARM64_OpTrunc16to8(v, config)
+	case OpTrunc32to16:
+		return rewriteValueARM64_OpTrunc32to16(v, config)
+	case OpTrunc32to8:
+		return rewriteValueARM64_OpTrunc32to8(v, config)
+	case OpTrunc64to16:
+		return rewriteValueARM64_OpTrunc64to16(v, config)
+	case OpTrunc64to32:
+		return rewriteValueARM64_OpTrunc64to32(v, config)
+	case OpTrunc64to8:
+		return rewriteValueARM64_OpTrunc64to8(v, config)
+	case OpXor16:
+		return rewriteValueARM64_OpXor16(v, config)
+	case OpXor32:
+		return rewriteValueARM64_OpXor32(v, config)
+	case OpXor64:
+		return rewriteValueARM64_OpXor64(v, config)
+	case OpXor8:
+		return rewriteValueARM64_OpXor8(v, config)
+	case OpZeroExt16to32:
+		return rewriteValueARM64_OpZeroExt16to32(v, config)
+	case OpZeroExt16to64:
+		return rewriteValueARM64_OpZeroExt16to64(v, config)
+	case OpZeroExt32to64:
+		return rewriteValueARM64_OpZeroExt32to64(v, config)
+	case OpZeroExt8to16:
+		return rewriteValueARM64_OpZeroExt8to16(v, config)
+	case OpZeroExt8to32:
+		return rewriteValueARM64_OpZeroExt8to32(v, config)
+	case OpZeroExt8to64:
+		return rewriteValueARM64_OpZeroExt8to64(v, config)
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64ADDconst(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (ADDconst [off1] (MOVDaddr [off2] {sym} ptr))
+	// cond:
+	// result: (MOVDaddr [off1+off2] {sym} ptr)
+	for {
+		off1 := v.AuxInt
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym := v_0.Aux
+		ptr := v_0.Args[0]
+		v.reset(OpARM64MOVDaddr)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpAdd16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Add16 x y)
+	// cond:
+	// result: (ADD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64ADD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAdd32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Add32 x y)
+	// cond:
+	// result: (ADD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64ADD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAdd32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Add32F x y)
+	// cond:
+	// result: (FADDS x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64FADDS)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAdd64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Add64 x y)
+	// cond:
+	// result: (ADD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64ADD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAdd64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Add64F x y)
+	// cond:
+	// result: (FADDD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64FADDD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAdd8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Add8 x y)
+	// cond:
+	// result: (ADD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64ADD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAddPtr(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (AddPtr x y)
+	// cond:
+	// result: (ADD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64ADD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAddr(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Addr {sym} base)
+	// cond:
+	// result: (MOVDaddr {sym} base)
+	for {
+		sym := v.Aux
+		base := v.Args[0]
+		v.reset(OpARM64MOVDaddr)
+		v.Aux = sym
+		v.AddArg(base)
+		return true
+	}
+}
+func rewriteValueARM64_OpAnd16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (And16 x y)
+	// cond:
+	// result: (AND x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64AND)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAnd32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (And32 x y)
+	// cond:
+	// result: (AND x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64AND)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAnd64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (And64 x y)
+	// cond:
+	// result: (AND x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64AND)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAnd8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (And8 x y)
+	// cond:
+	// result: (AND x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64AND)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpAndB(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (AndB x y)
+	// cond:
+	// result: (AND x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64AND)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpClosureCall(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (ClosureCall [argwid] entry closure mem)
+	// cond:
+	// result: (CALLclosure [argwid] entry closure mem)
+	for {
+		argwid := v.AuxInt
+		entry := v.Args[0]
+		closure := v.Args[1]
+		mem := v.Args[2]
+		v.reset(OpARM64CALLclosure)
+		v.AuxInt = argwid
+		v.AddArg(entry)
+		v.AddArg(closure)
+		v.AddArg(mem)
+		return true
+	}
+}
+func rewriteValueARM64_OpCom16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Com16 x)
+	// cond:
+	// result: (MVN x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MVN)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCom32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Com32 x)
+	// cond:
+	// result: (MVN x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MVN)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCom64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Com64 x)
+	// cond:
+	// result: (MVN x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MVN)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCom8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Com8 x)
+	// cond:
+	// result: (MVN x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MVN)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpConst16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Const16 [val])
+	// cond:
+	// result: (MOVDconst [val])
+	for {
+		val := v.AuxInt
+		v.reset(OpARM64MOVDconst)
+		v.AuxInt = val
+		return true
+	}
+}
+func rewriteValueARM64_OpConst32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Const32 [val])
+	// cond:
+	// result: (MOVDconst [val])
+	for {
+		val := v.AuxInt
+		v.reset(OpARM64MOVDconst)
+		v.AuxInt = val
+		return true
+	}
+}
+func rewriteValueARM64_OpConst32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Const32F [val])
+	// cond:
+	// result: (FMOVSconst [val])
+	for {
+		val := v.AuxInt
+		v.reset(OpARM64FMOVSconst)
+		v.AuxInt = val
+		return true
+	}
+}
+func rewriteValueARM64_OpConst64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Const64 [val])
+	// cond:
+	// result: (MOVDconst [val])
+	for {
+		val := v.AuxInt
+		v.reset(OpARM64MOVDconst)
+		v.AuxInt = val
+		return true
+	}
+}
+func rewriteValueARM64_OpConst64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Const64F [val])
+	// cond:
+	// result: (FMOVDconst [val])
+	for {
+		val := v.AuxInt
+		v.reset(OpARM64FMOVDconst)
+		v.AuxInt = val
+		return true
+	}
+}
+func rewriteValueARM64_OpConst8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Const8 [val])
+	// cond:
+	// result: (MOVDconst [val])
+	for {
+		val := v.AuxInt
+		v.reset(OpARM64MOVDconst)
+		v.AuxInt = val
+		return true
+	}
+}
+func rewriteValueARM64_OpConstBool(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (ConstBool [b])
+	// cond:
+	// result: (MOVDconst [b])
+	for {
+		b := v.AuxInt
+		v.reset(OpARM64MOVDconst)
+		v.AuxInt = b
+		return true
+	}
+}
+func rewriteValueARM64_OpConstNil(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (ConstNil)
+	// cond:
+	// result: (MOVDconst [0])
+	for {
+		v.reset(OpARM64MOVDconst)
+		v.AuxInt = 0
+		return true
+	}
+}
+func rewriteValueARM64_OpConvert(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Convert x mem)
+	// cond:
+	// result: (MOVDconvert x mem)
+	for {
+		x := v.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64MOVDconvert)
+		v.AddArg(x)
+		v.AddArg(mem)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt32Fto32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt32Fto32 x)
+	// cond:
+	// result: (FCVTZSSW x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64FCVTZSSW)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt32Fto32U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt32Fto32U x)
+	// cond:
+	// result: (FCVTZUSW x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64FCVTZUSW)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt32Fto64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt32Fto64 x)
+	// cond:
+	// result: (FCVTZSS x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64FCVTZSS)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt32Fto64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt32Fto64F x)
+	// cond:
+	// result: (FCVTSD x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64FCVTSD)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt32Uto32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt32Uto32F x)
+	// cond:
+	// result: (UCVTFWS x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64UCVTFWS)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt32Uto64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt32Uto64F x)
+	// cond:
+	// result: (UCVTFWD x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64UCVTFWD)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt32to32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt32to32F x)
+	// cond:
+	// result: (SCVTFWS x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64SCVTFWS)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt32to64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt32to64F x)
+	// cond:
+	// result: (SCVTFWD x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64SCVTFWD)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt64Fto32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt64Fto32 x)
+	// cond:
+	// result: (FCVTZSDW x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64FCVTZSDW)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt64Fto32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt64Fto32F x)
+	// cond:
+	// result: (FCVTDS x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64FCVTDS)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt64Fto32U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt64Fto32U x)
+	// cond:
+	// result: (FCVTZUDW x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64FCVTZUDW)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt64Fto64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt64Fto64 x)
+	// cond:
+	// result: (FCVTZSD x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64FCVTZSD)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt64to32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt64to32F x)
+	// cond:
+	// result: (SCVTFS x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64SCVTFS)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpCvt64to64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Cvt64to64F x)
+	// cond:
+	// result: (SCVTFD x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64SCVTFD)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpDeferCall(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (DeferCall [argwid] mem)
+	// cond:
+	// result: (CALLdefer [argwid] mem)
+	for {
+		argwid := v.AuxInt
+		mem := v.Args[0]
+		v.reset(OpARM64CALLdefer)
+		v.AuxInt = argwid
+		v.AddArg(mem)
+		return true
+	}
+}
+func rewriteValueARM64_OpDiv16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Div16 x y)
+	// cond:
+	// result: (DIVW (SignExt16to32 x) (SignExt16to32 y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64DIVW)
+		v0 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v0.AddArg(x)
+		v.AddArg(v0)
+		v1 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v1.AddArg(y)
+		v.AddArg(v1)
+		return true
+	}
+}
+func rewriteValueARM64_OpDiv16u(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Div16u x y)
+	// cond:
+	// result: (UDIVW (ZeroExt16to32 x) (ZeroExt16to32 y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64UDIVW)
+		v0 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v0.AddArg(x)
+		v.AddArg(v0)
+		v1 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v1.AddArg(y)
+		v.AddArg(v1)
+		return true
+	}
+}
+func rewriteValueARM64_OpDiv32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Div32 x y)
+	// cond:
+	// result: (DIVW x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64DIVW)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpDiv32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Div32F x y)
+	// cond:
+	// result: (FDIVS x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64FDIVS)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpDiv32u(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Div32u x y)
+	// cond:
+	// result: (UDIVW x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64UDIVW)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpDiv64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Div64 x y)
+	// cond:
+	// result: (DIV x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64DIV)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpDiv64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Div64F x y)
+	// cond:
+	// result: (FDIVD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64FDIVD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpDiv64u(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Div64u x y)
+	// cond:
+	// result: (UDIV x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64UDIV)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpDiv8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Div8 x y)
+	// cond:
+	// result: (DIVW (SignExt8to32 x) (SignExt8to32 y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64DIVW)
+		v0 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v0.AddArg(x)
+		v.AddArg(v0)
+		v1 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v1.AddArg(y)
+		v.AddArg(v1)
+		return true
+	}
+}
+func rewriteValueARM64_OpDiv8u(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Div8u x y)
+	// cond:
+	// result: (UDIVW (ZeroExt8to32 x) (ZeroExt8to32 y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64UDIVW)
+		v0 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v0.AddArg(x)
+		v.AddArg(v0)
+		v1 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v1.AddArg(y)
+		v.AddArg(v1)
+		return true
+	}
+}
+func rewriteValueARM64_OpEq16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Eq16 x y)
+	// cond:
+	// result: (Equal (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64Equal)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpEq32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Eq32 x y)
+	// cond:
+	// result: (Equal (CMPW x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64Equal)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpEq32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Eq32F x y)
+	// cond:
+	// result: (Equal (FCMPS x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64Equal)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPS, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpEq64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Eq64 x y)
+	// cond:
+	// result: (Equal (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64Equal)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpEq64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Eq64F x y)
+	// cond:
+	// result: (Equal (FCMPD x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64Equal)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPD, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpEq8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Eq8 x y)
+	// cond:
+	// result: (Equal (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64Equal)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpEqB(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (EqB x y)
+	// cond:
+	// result: (XORconst [1] (XOR <config.fe.TypeBool()> x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64XORconst)
+		v.AuxInt = 1
+		v0 := b.NewValue0(v.Line, OpARM64XOR, config.fe.TypeBool())
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpEqPtr(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (EqPtr x y)
+	// cond:
+	// result: (Equal (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64Equal)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpARM64FMOVDload(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (FMOVDload [off1] {sym} (ADDconst [off2] ptr) mem)
+	// cond:
+	// result: (FMOVDload [off1+off2] {sym} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64FMOVDload)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (FMOVDload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (FMOVDload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64FMOVDload)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64FMOVDstore(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (FMOVDstore [off1] {sym} (ADDconst [off2] ptr) val mem)
+	// cond:
+	// result: (FMOVDstore [off1+off2] {sym} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		v.reset(OpARM64FMOVDstore)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (FMOVDstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (FMOVDstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64FMOVDstore)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64FMOVSload(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (FMOVSload [off1] {sym} (ADDconst [off2] ptr) mem)
+	// cond:
+	// result: (FMOVSload [off1+off2] {sym} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64FMOVSload)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (FMOVSload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (FMOVSload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64FMOVSload)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64FMOVSstore(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (FMOVSstore [off1] {sym} (ADDconst [off2] ptr) val mem)
+	// cond:
+	// result: (FMOVSstore [off1+off2] {sym} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		v.reset(OpARM64FMOVSstore)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (FMOVSstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (FMOVSstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64FMOVSstore)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpGeq16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Geq16 x y)
+	// cond:
+	// result: (GreaterEqual (CMPW (SignExt16to32 x) (SignExt16to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGeq16U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Geq16U x y)
+	// cond:
+	// result: (GreaterEqualU (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqualU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGeq32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Geq32 x y)
+	// cond:
+	// result: (GreaterEqual (CMPW x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGeq32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Geq32F x y)
+	// cond:
+	// result: (GreaterEqual (FCMPS x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqual)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPS, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGeq32U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Geq32U x y)
+	// cond:
+	// result: (GreaterEqualU (CMPW x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqualU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGeq64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Geq64 x y)
+	// cond:
+	// result: (GreaterEqual (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGeq64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Geq64F x y)
+	// cond:
+	// result: (GreaterEqual (FCMPD x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqual)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPD, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGeq64U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Geq64U x y)
+	// cond:
+	// result: (GreaterEqualU (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqualU)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGeq8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Geq8 x y)
+	// cond:
+	// result: (GreaterEqual (CMPW (SignExt8to32 x) (SignExt8to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGeq8U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Geq8U x y)
+	// cond:
+	// result: (GreaterEqualU (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqualU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGetClosurePtr(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (GetClosurePtr)
+	// cond:
+	// result: (LoweredGetClosurePtr)
+	for {
+		v.reset(OpARM64LoweredGetClosurePtr)
+		return true
+	}
+}
+func rewriteValueARM64_OpGoCall(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (GoCall [argwid] mem)
+	// cond:
+	// result: (CALLgo [argwid] mem)
+	for {
+		argwid := v.AuxInt
+		mem := v.Args[0]
+		v.reset(OpARM64CALLgo)
+		v.AuxInt = argwid
+		v.AddArg(mem)
+		return true
+	}
+}
+func rewriteValueARM64_OpGreater16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Greater16 x y)
+	// cond:
+	// result: (GreaterThan (CMPW (SignExt16to32 x) (SignExt16to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThan)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGreater16U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Greater16U x y)
+	// cond:
+	// result: (GreaterThanU (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThanU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGreater32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Greater32 x y)
+	// cond:
+	// result: (GreaterThan (CMPW x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThan)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGreater32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Greater32F x y)
+	// cond:
+	// result: (GreaterThan (FCMPS x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThan)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPS, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGreater32U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Greater32U x y)
+	// cond:
+	// result: (GreaterThanU (CMPW x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThanU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGreater64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Greater64 x y)
+	// cond:
+	// result: (GreaterThan (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThan)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGreater64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Greater64F x y)
+	// cond:
+	// result: (GreaterThan (FCMPD x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThan)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPD, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGreater64U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Greater64U x y)
+	// cond:
+	// result: (GreaterThanU (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThanU)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGreater8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Greater8 x y)
+	// cond:
+	// result: (GreaterThan (CMPW (SignExt8to32 x) (SignExt8to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThan)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpGreater8U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Greater8U x y)
+	// cond:
+	// result: (GreaterThanU (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThanU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpInterCall(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (InterCall [argwid] entry mem)
+	// cond:
+	// result: (CALLinter [argwid] entry mem)
+	for {
+		argwid := v.AuxInt
+		entry := v.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64CALLinter)
+		v.AuxInt = argwid
+		v.AddArg(entry)
+		v.AddArg(mem)
+		return true
+	}
+}
+func rewriteValueARM64_OpIsInBounds(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (IsInBounds idx len)
+	// cond:
+	// result: (LessThanU (CMP idx len))
+	for {
+		idx := v.Args[0]
+		len := v.Args[1]
+		v.reset(OpARM64LessThanU)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(idx)
+		v0.AddArg(len)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpIsNonNil(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (IsNonNil ptr)
+	// cond:
+	// result: (NotEqual (CMPconst [0] ptr))
+	for {
+		ptr := v.Args[0]
+		v.reset(OpARM64NotEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMPconst, TypeFlags)
+		v0.AuxInt = 0
+		v0.AddArg(ptr)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpIsSliceInBounds(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (IsSliceInBounds idx len)
+	// cond:
+	// result: (LessEqualU (CMP idx len))
+	for {
+		idx := v.Args[0]
+		len := v.Args[1]
+		v.reset(OpARM64LessEqualU)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(idx)
+		v0.AddArg(len)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLeq16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Leq16 x y)
+	// cond:
+	// result: (LessEqual (CMPW (SignExt16to32 x) (SignExt16to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLeq16U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Leq16U x y)
+	// cond:
+	// result: (LessEqualU (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessEqualU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLeq32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Leq32 x y)
+	// cond:
+	// result: (LessEqual (CMPW x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLeq32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Leq32F x y)
+	// cond:
+	// result: (GreaterEqual (FCMPS y x))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqual)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPS, TypeFlags)
+		v0.AddArg(y)
+		v0.AddArg(x)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLeq32U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Leq32U x y)
+	// cond:
+	// result: (LessEqualU (CMPW x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessEqualU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLeq64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Leq64 x y)
+	// cond:
+	// result: (LessEqual (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLeq64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Leq64F x y)
+	// cond:
+	// result: (GreaterEqual (FCMPD y x))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterEqual)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPD, TypeFlags)
+		v0.AddArg(y)
+		v0.AddArg(x)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLeq64U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Leq64U x y)
+	// cond:
+	// result: (LessEqualU (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessEqualU)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLeq8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Leq8 x y)
+	// cond:
+	// result: (LessEqual (CMPW (SignExt8to32 x) (SignExt8to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLeq8U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Leq8U x y)
+	// cond:
+	// result: (LessEqualU (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessEqualU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLess16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Less16 x y)
+	// cond:
+	// result: (LessThan (CMPW (SignExt16to32 x) (SignExt16to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessThan)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLess16U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Less16U x y)
+	// cond:
+	// result: (LessThanU (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessThanU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLess32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Less32 x y)
+	// cond:
+	// result: (LessThan (CMPW x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessThan)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLess32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Less32F x y)
+	// cond:
+	// result: (GreaterThan (FCMPS y x))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThan)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPS, TypeFlags)
+		v0.AddArg(y)
+		v0.AddArg(x)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLess32U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Less32U x y)
+	// cond:
+	// result: (LessThanU (CMPW x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessThanU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLess64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Less64 x y)
+	// cond:
+	// result: (LessThan (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessThan)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLess64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Less64F x y)
+	// cond:
+	// result: (GreaterThan (FCMPD y x))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64GreaterThan)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPD, TypeFlags)
+		v0.AddArg(y)
+		v0.AddArg(x)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLess64U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Less64U x y)
+	// cond:
+	// result: (LessThanU (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessThanU)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLess8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Less8 x y)
+	// cond:
+	// result: (LessThan (CMPW (SignExt8to32 x) (SignExt8to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessThan)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLess8U(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Less8U x y)
+	// cond:
+	// result: (LessThanU (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64LessThanU)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpLoad(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Load <t> ptr mem)
+	// cond: t.IsBoolean()
+	// result: (MOVBUload ptr mem)
+	for {
+		t := v.Type
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		if !(t.IsBoolean()) {
+			break
+		}
+		v.reset(OpARM64MOVBUload)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Load <t> ptr mem)
+	// cond: (is8BitInt(t) && isSigned(t))
+	// result: (MOVBload ptr mem)
+	for {
+		t := v.Type
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		if !(is8BitInt(t) && isSigned(t)) {
+			break
+		}
+		v.reset(OpARM64MOVBload)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Load <t> ptr mem)
+	// cond: (is8BitInt(t) && !isSigned(t))
+	// result: (MOVBUload ptr mem)
+	for {
+		t := v.Type
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		if !(is8BitInt(t) && !isSigned(t)) {
+			break
+		}
+		v.reset(OpARM64MOVBUload)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Load <t> ptr mem)
+	// cond: (is16BitInt(t) && isSigned(t))
+	// result: (MOVHload ptr mem)
+	for {
+		t := v.Type
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		if !(is16BitInt(t) && isSigned(t)) {
+			break
+		}
+		v.reset(OpARM64MOVHload)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Load <t> ptr mem)
+	// cond: (is16BitInt(t) && !isSigned(t))
+	// result: (MOVHUload ptr mem)
+	for {
+		t := v.Type
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		if !(is16BitInt(t) && !isSigned(t)) {
+			break
+		}
+		v.reset(OpARM64MOVHUload)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Load <t> ptr mem)
+	// cond: (is32BitInt(t) && isSigned(t))
+	// result: (MOVWload ptr mem)
+	for {
+		t := v.Type
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		if !(is32BitInt(t) && isSigned(t)) {
+			break
+		}
+		v.reset(OpARM64MOVWload)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Load <t> ptr mem)
+	// cond: (is32BitInt(t) && !isSigned(t))
+	// result: (MOVWUload ptr mem)
+	for {
+		t := v.Type
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		if !(is32BitInt(t) && !isSigned(t)) {
+			break
+		}
+		v.reset(OpARM64MOVWUload)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Load <t> ptr mem)
+	// cond: (is64BitInt(t) || isPtr(t))
+	// result: (MOVDload ptr mem)
+	for {
+		t := v.Type
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		if !(is64BitInt(t) || isPtr(t)) {
+			break
+		}
+		v.reset(OpARM64MOVDload)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Load <t> ptr mem)
+	// cond: is32BitFloat(t)
+	// result: (FMOVSload ptr mem)
+	for {
+		t := v.Type
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		if !(is32BitFloat(t)) {
+			break
+		}
+		v.reset(OpARM64FMOVSload)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Load <t> ptr mem)
+	// cond: is64BitFloat(t)
+	// result: (FMOVDload ptr mem)
+	for {
+		t := v.Type
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		if !(is64BitFloat(t)) {
+			break
+		}
+		v.reset(OpARM64FMOVDload)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVBUload(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVBUload [off1] {sym} (ADDconst [off2] ptr) mem)
+	// cond:
+	// result: (MOVBUload [off1+off2] {sym} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64MOVBUload)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVBUload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVBUload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVBUload)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVBload(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVBload [off1] {sym} (ADDconst [off2] ptr) mem)
+	// cond:
+	// result: (MOVBload [off1+off2] {sym} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64MOVBload)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVBload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVBload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVBload)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVBstore(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVBstore [off1] {sym} (ADDconst [off2] ptr) val mem)
+	// cond:
+	// result: (MOVBstore [off1+off2] {sym} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		v.reset(OpARM64MOVBstore)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVBstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVBstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVBstore)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVDload(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVDload [off1] {sym} (ADDconst [off2] ptr) mem)
+	// cond:
+	// result: (MOVDload [off1+off2] {sym} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64MOVDload)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVDload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVDload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVDload)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVDstore(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVDstore [off1] {sym} (ADDconst [off2] ptr) val mem)
+	// cond:
+	// result: (MOVDstore [off1+off2] {sym} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		v.reset(OpARM64MOVDstore)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVDstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVDstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVDstore)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVHUload(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVHUload [off1] {sym} (ADDconst [off2] ptr) mem)
+	// cond:
+	// result: (MOVHUload [off1+off2] {sym} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64MOVHUload)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVHUload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVHUload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVHUload)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVHload(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVHload [off1] {sym} (ADDconst [off2] ptr) mem)
+	// cond:
+	// result: (MOVHload [off1+off2] {sym} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64MOVHload)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVHload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVHload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVHload)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVHstore(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVHstore [off1] {sym} (ADDconst [off2] ptr) val mem)
+	// cond:
+	// result: (MOVHstore [off1+off2] {sym} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		v.reset(OpARM64MOVHstore)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVHstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVHstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVHstore)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVWUload(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVWUload [off1] {sym} (ADDconst [off2] ptr) mem)
+	// cond:
+	// result: (MOVWUload [off1+off2] {sym} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64MOVWUload)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVWUload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVWUload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVWUload)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVWload(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVWload [off1] {sym} (ADDconst [off2] ptr) mem)
+	// cond:
+	// result: (MOVWload [off1+off2] {sym} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64MOVWload)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVWload [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVWload [off1+off2] {mergeSym(sym1,sym2)} ptr mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		mem := v.Args[1]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVWload)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpARM64MOVWstore(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (MOVWstore [off1] {sym} (ADDconst [off2] ptr) val mem)
+	// cond:
+	// result: (MOVWstore [off1+off2] {sym} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64ADDconst {
+			break
+		}
+		off2 := v_0.AuxInt
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		v.reset(OpARM64MOVWstore)
+		v.AuxInt = off1 + off2
+		v.Aux = sym
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (MOVWstore [off1] {sym1} (MOVDaddr [off2] {sym2} ptr) val mem)
+	// cond: canMergeSym(sym1,sym2)
+	// result: (MOVWstore [off1+off2] {mergeSym(sym1,sym2)} ptr val mem)
+	for {
+		off1 := v.AuxInt
+		sym1 := v.Aux
+		v_0 := v.Args[0]
+		if v_0.Op != OpARM64MOVDaddr {
+			break
+		}
+		off2 := v_0.AuxInt
+		sym2 := v_0.Aux
+		ptr := v_0.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		if !(canMergeSym(sym1, sym2)) {
+			break
+		}
+		v.reset(OpARM64MOVWstore)
+		v.AuxInt = off1 + off2
+		v.Aux = mergeSym(sym1, sym2)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpMod16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mod16 x y)
+	// cond:
+	// result: (MODW (SignExt16to32 x) (SignExt16to32 y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64MODW)
+		v0 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v0.AddArg(x)
+		v.AddArg(v0)
+		v1 := b.NewValue0(v.Line, OpSignExt16to32, config.fe.TypeInt32())
+		v1.AddArg(y)
+		v.AddArg(v1)
+		return true
+	}
+}
+func rewriteValueARM64_OpMod16u(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mod16u x y)
+	// cond:
+	// result: (UMODW (ZeroExt16to32 x) (ZeroExt16to32 y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64UMODW)
+		v0 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v0.AddArg(x)
+		v.AddArg(v0)
+		v1 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v1.AddArg(y)
+		v.AddArg(v1)
+		return true
+	}
+}
+func rewriteValueARM64_OpMod32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mod32 x y)
+	// cond:
+	// result: (MODW x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64MODW)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpMod32u(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mod32u x y)
+	// cond:
+	// result: (UMODW x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64UMODW)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpMod64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mod64 x y)
+	// cond:
+	// result: (MOD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64MOD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpMod64u(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mod64u x y)
+	// cond:
+	// result: (UMOD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64UMOD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpMod8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mod8 x y)
+	// cond:
+	// result: (MODW (SignExt8to32 x) (SignExt8to32 y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64MODW)
+		v0 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v0.AddArg(x)
+		v.AddArg(v0)
+		v1 := b.NewValue0(v.Line, OpSignExt8to32, config.fe.TypeInt32())
+		v1.AddArg(y)
+		v.AddArg(v1)
+		return true
+	}
+}
+func rewriteValueARM64_OpMod8u(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mod8u x y)
+	// cond:
+	// result: (UMODW (ZeroExt8to32 x) (ZeroExt8to32 y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64UMODW)
+		v0 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v0.AddArg(x)
+		v.AddArg(v0)
+		v1 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v1.AddArg(y)
+		v.AddArg(v1)
+		return true
+	}
+}
+func rewriteValueARM64_OpMul16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mul16 x y)
+	// cond:
+	// result: (MUL x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64MUL)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpMul32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mul32 x y)
+	// cond:
+	// result: (MUL x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64MUL)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpMul32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mul32F x y)
+	// cond:
+	// result: (FMULS x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64FMULS)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpMul64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mul64 x y)
+	// cond:
+	// result: (MUL x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64MUL)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpMul64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mul64F x y)
+	// cond:
+	// result: (FMULD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64FMULD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpMul8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Mul8 x y)
+	// cond:
+	// result: (MUL x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64MUL)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeg16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neg16 x)
+	// cond:
+	// result: (NEG x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64NEG)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeg32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neg32 x)
+	// cond:
+	// result: (NEG x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64NEG)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeg32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neg32F x)
+	// cond:
+	// result: (FNEGS x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64FNEGS)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeg64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neg64 x)
+	// cond:
+	// result: (NEG x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64NEG)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeg64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neg64F x)
+	// cond:
+	// result: (FNEGD x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64FNEGD)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeg8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neg8 x)
+	// cond:
+	// result: (NEG x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64NEG)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeq16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neq16 x y)
+	// cond:
+	// result: (NotEqual (CMPW (ZeroExt16to32 x) (ZeroExt16to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64NotEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt16to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeq32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neq32 x y)
+	// cond:
+	// result: (NotEqual (CMPW x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64NotEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeq32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neq32F x y)
+	// cond:
+	// result: (NotEqual (FCMPS x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64NotEqual)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPS, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeq64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neq64 x y)
+	// cond:
+	// result: (NotEqual (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64NotEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeq64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neq64F x y)
+	// cond:
+	// result: (NotEqual (FCMPD x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64NotEqual)
+		v0 := b.NewValue0(v.Line, OpARM64FCMPD, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeq8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Neq8 x y)
+	// cond:
+	// result: (NotEqual (CMPW (ZeroExt8to32 x) (ZeroExt8to32 y)))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64NotEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMPW, TypeFlags)
+		v1 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v1.AddArg(x)
+		v0.AddArg(v1)
+		v2 := b.NewValue0(v.Line, OpZeroExt8to32, config.fe.TypeUInt32())
+		v2.AddArg(y)
+		v0.AddArg(v2)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeqB(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (NeqB x y)
+	// cond:
+	// result: (XOR x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64XOR)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpNeqPtr(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (NeqPtr x y)
+	// cond:
+	// result: (NotEqual (CMP x y))
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64NotEqual)
+		v0 := b.NewValue0(v.Line, OpARM64CMP, TypeFlags)
+		v0.AddArg(x)
+		v0.AddArg(y)
+		v.AddArg(v0)
+		return true
+	}
+}
+func rewriteValueARM64_OpNilCheck(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (NilCheck ptr mem)
+	// cond:
+	// result: (LoweredNilCheck ptr mem)
+	for {
+		ptr := v.Args[0]
+		mem := v.Args[1]
+		v.reset(OpARM64LoweredNilCheck)
+		v.AddArg(ptr)
+		v.AddArg(mem)
+		return true
+	}
+}
+func rewriteValueARM64_OpNot(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Not x)
+	// cond:
+	// result: (XORconst [1] x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64XORconst)
+		v.AuxInt = 1
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpOffPtr(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (OffPtr [off] ptr:(SP))
+	// cond:
+	// result: (MOVDaddr [off] ptr)
+	for {
+		off := v.AuxInt
+		ptr := v.Args[0]
+		if ptr.Op != OpSP {
+			break
+		}
+		v.reset(OpARM64MOVDaddr)
+		v.AuxInt = off
+		v.AddArg(ptr)
+		return true
+	}
+	// match: (OffPtr [off] ptr)
+	// cond:
+	// result: (ADDconst [off] ptr)
+	for {
+		off := v.AuxInt
+		ptr := v.Args[0]
+		v.reset(OpARM64ADDconst)
+		v.AuxInt = off
+		v.AddArg(ptr)
+		return true
+	}
+}
+func rewriteValueARM64_OpOr16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Or16 x y)
+	// cond:
+	// result: (OR x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64OR)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpOr32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Or32 x y)
+	// cond:
+	// result: (OR x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64OR)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpOr64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Or64 x y)
+	// cond:
+	// result: (OR x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64OR)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpOr8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Or8 x y)
+	// cond:
+	// result: (OR x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64OR)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpOrB(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (OrB x y)
+	// cond:
+	// result: (OR x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64OR)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpSignExt16to32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (SignExt16to32 x)
+	// cond:
+	// result: (MOVHreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVHreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpSignExt16to64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (SignExt16to64 x)
+	// cond:
+	// result: (MOVHreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVHreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpSignExt32to64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (SignExt32to64 x)
+	// cond:
+	// result: (MOVWreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVWreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpSignExt8to16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (SignExt8to16 x)
+	// cond:
+	// result: (MOVBreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVBreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpSignExt8to32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (SignExt8to32 x)
+	// cond:
+	// result: (MOVBreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVBreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpSignExt8to64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (SignExt8to64 x)
+	// cond:
+	// result: (MOVBreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVBreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpStaticCall(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (StaticCall [argwid] {target} mem)
+	// cond:
+	// result: (CALLstatic [argwid] {target} mem)
+	for {
+		argwid := v.AuxInt
+		target := v.Aux
+		mem := v.Args[0]
+		v.reset(OpARM64CALLstatic)
+		v.AuxInt = argwid
+		v.Aux = target
+		v.AddArg(mem)
+		return true
+	}
+}
+func rewriteValueARM64_OpStore(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Store [1] ptr val mem)
+	// cond:
+	// result: (MOVBstore ptr val mem)
+	for {
+		if v.AuxInt != 1 {
+			break
+		}
+		ptr := v.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		v.reset(OpARM64MOVBstore)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Store [2] ptr val mem)
+	// cond:
+	// result: (MOVHstore ptr val mem)
+	for {
+		if v.AuxInt != 2 {
+			break
+		}
+		ptr := v.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		v.reset(OpARM64MOVHstore)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Store [4] ptr val mem)
+	// cond: !is32BitFloat(val.Type)
+	// result: (MOVWstore ptr val mem)
+	for {
+		if v.AuxInt != 4 {
+			break
+		}
+		ptr := v.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		if !(!is32BitFloat(val.Type)) {
+			break
+		}
+		v.reset(OpARM64MOVWstore)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Store [8] ptr val mem)
+	// cond: !is64BitFloat(val.Type)
+	// result: (MOVDstore ptr val mem)
+	for {
+		if v.AuxInt != 8 {
+			break
+		}
+		ptr := v.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		if !(!is64BitFloat(val.Type)) {
+			break
+		}
+		v.reset(OpARM64MOVDstore)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Store [4] ptr val mem)
+	// cond: is32BitFloat(val.Type)
+	// result: (FMOVSstore ptr val mem)
+	for {
+		if v.AuxInt != 4 {
+			break
+		}
+		ptr := v.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		if !(is32BitFloat(val.Type)) {
+			break
+		}
+		v.reset(OpARM64FMOVSstore)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	// match: (Store [8] ptr val mem)
+	// cond: is64BitFloat(val.Type)
+	// result: (FMOVDstore ptr val mem)
+	for {
+		if v.AuxInt != 8 {
+			break
+		}
+		ptr := v.Args[0]
+		val := v.Args[1]
+		mem := v.Args[2]
+		if !(is64BitFloat(val.Type)) {
+			break
+		}
+		v.reset(OpARM64FMOVDstore)
+		v.AddArg(ptr)
+		v.AddArg(val)
+		v.AddArg(mem)
+		return true
+	}
+	return false
+}
+func rewriteValueARM64_OpSub16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Sub16 x y)
+	// cond:
+	// result: (SUB x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64SUB)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpSub32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Sub32 x y)
+	// cond:
+	// result: (SUB x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64SUB)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpSub32F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Sub32F x y)
+	// cond:
+	// result: (FSUBS x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64FSUBS)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpSub64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Sub64 x y)
+	// cond:
+	// result: (SUB x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64SUB)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpSub64F(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Sub64F x y)
+	// cond:
+	// result: (FSUBD x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64FSUBD)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpSub8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Sub8 x y)
+	// cond:
+	// result: (SUB x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64SUB)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpSubPtr(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (SubPtr x y)
+	// cond:
+	// result: (SUB x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64SUB)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpTrunc16to8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Trunc16to8 x)
+	// cond:
+	// result: x
+	for {
+		x := v.Args[0]
+		v.reset(OpCopy)
+		v.Type = x.Type
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpTrunc32to16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Trunc32to16 x)
+	// cond:
+	// result: x
+	for {
+		x := v.Args[0]
+		v.reset(OpCopy)
+		v.Type = x.Type
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpTrunc32to8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Trunc32to8 x)
+	// cond:
+	// result: x
+	for {
+		x := v.Args[0]
+		v.reset(OpCopy)
+		v.Type = x.Type
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpTrunc64to16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Trunc64to16 x)
+	// cond:
+	// result: x
+	for {
+		x := v.Args[0]
+		v.reset(OpCopy)
+		v.Type = x.Type
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpTrunc64to32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Trunc64to32 x)
+	// cond:
+	// result: x
+	for {
+		x := v.Args[0]
+		v.reset(OpCopy)
+		v.Type = x.Type
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpTrunc64to8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Trunc64to8 x)
+	// cond:
+	// result: x
+	for {
+		x := v.Args[0]
+		v.reset(OpCopy)
+		v.Type = x.Type
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpXor16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Xor16 x y)
+	// cond:
+	// result: (XOR x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64XOR)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpXor32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Xor32 x y)
+	// cond:
+	// result: (XOR x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64XOR)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpXor64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Xor64 x y)
+	// cond:
+	// result: (XOR x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64XOR)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpXor8(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (Xor8 x y)
+	// cond:
+	// result: (XOR x y)
+	for {
+		x := v.Args[0]
+		y := v.Args[1]
+		v.reset(OpARM64XOR)
+		v.AddArg(x)
+		v.AddArg(y)
+		return true
+	}
+}
+func rewriteValueARM64_OpZeroExt16to32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (ZeroExt16to32 x)
+	// cond:
+	// result: (MOVHUreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVHUreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpZeroExt16to64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (ZeroExt16to64 x)
+	// cond:
+	// result: (MOVHUreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVHUreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpZeroExt32to64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (ZeroExt32to64 x)
+	// cond:
+	// result: (MOVWUreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVWUreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpZeroExt8to16(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (ZeroExt8to16 x)
+	// cond:
+	// result: (MOVBUreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVBUreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpZeroExt8to32(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (ZeroExt8to32 x)
+	// cond:
+	// result: (MOVBUreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVBUreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteValueARM64_OpZeroExt8to64(v *Value, config *Config) bool {
+	b := v.Block
+	_ = b
+	// match: (ZeroExt8to64 x)
+	// cond:
+	// result: (MOVBUreg x)
+	for {
+		x := v.Args[0]
+		v.reset(OpARM64MOVBUreg)
+		v.AddArg(x)
+		return true
+	}
+}
+func rewriteBlockARM64(b *Block) bool {
+	switch b.Kind {
+	case BlockIf:
+		// match: (If (Equal cc) yes no)
+		// cond:
+		// result: (EQ cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64Equal {
+				break
+			}
+			cc := v.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64EQ
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (If (NotEqual cc) yes no)
+		// cond:
+		// result: (NE cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64NotEqual {
+				break
+			}
+			cc := v.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64NE
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (If (LessThan cc) yes no)
+		// cond:
+		// result: (LT cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64LessThan {
+				break
+			}
+			cc := v.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64LT
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (If (LessThanU cc) yes no)
+		// cond:
+		// result: (ULT cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64LessThanU {
+				break
+			}
+			cc := v.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64ULT
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (If (LessEqual cc) yes no)
+		// cond:
+		// result: (LE cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64LessEqual {
+				break
+			}
+			cc := v.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64LE
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (If (LessEqualU cc) yes no)
+		// cond:
+		// result: (ULE cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64LessEqualU {
+				break
+			}
+			cc := v.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64ULE
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (If (GreaterThan cc) yes no)
+		// cond:
+		// result: (GT cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64GreaterThan {
+				break
+			}
+			cc := v.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64GT
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (If (GreaterThanU cc) yes no)
+		// cond:
+		// result: (UGT cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64GreaterThanU {
+				break
+			}
+			cc := v.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64UGT
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (If (GreaterEqual cc) yes no)
+		// cond:
+		// result: (GE cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64GreaterEqual {
+				break
+			}
+			cc := v.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64GE
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (If (GreaterEqualU cc) yes no)
+		// cond:
+		// result: (UGE cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64GreaterEqualU {
+				break
+			}
+			cc := v.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64UGE
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (If cond yes no)
+		// cond:
+		// result: (NE (CMPconst [0] cond) yes no)
+		for {
+			v := b.Control
+			cond := b.Control
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64NE
+			v0 := b.NewValue0(v.Line, OpARM64CMPconst, TypeFlags)
+			v0.AuxInt = 0
+			v0.AddArg(cond)
+			b.SetControl(v0)
+			_ = yes
+			_ = no
+			return true
+		}
+	case BlockARM64NE:
+		// match: (NE (CMPconst [0] (Equal cc)) yes no)
+		// cond:
+		// result: (EQ cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64CMPconst {
+				break
+			}
+			if v.AuxInt != 0 {
+				break
+			}
+			v_0 := v.Args[0]
+			if v_0.Op != OpARM64Equal {
+				break
+			}
+			cc := v_0.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64EQ
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (NE (CMPconst [0] (NotEqual cc)) yes no)
+		// cond:
+		// result: (NE cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64CMPconst {
+				break
+			}
+			if v.AuxInt != 0 {
+				break
+			}
+			v_0 := v.Args[0]
+			if v_0.Op != OpARM64NotEqual {
+				break
+			}
+			cc := v_0.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64NE
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (NE (CMPconst [0] (LessThan cc)) yes no)
+		// cond:
+		// result: (LT cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64CMPconst {
+				break
+			}
+			if v.AuxInt != 0 {
+				break
+			}
+			v_0 := v.Args[0]
+			if v_0.Op != OpARM64LessThan {
+				break
+			}
+			cc := v_0.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64LT
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (NE (CMPconst [0] (LessThanU cc)) yes no)
+		// cond:
+		// result: (ULT cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64CMPconst {
+				break
+			}
+			if v.AuxInt != 0 {
+				break
+			}
+			v_0 := v.Args[0]
+			if v_0.Op != OpARM64LessThanU {
+				break
+			}
+			cc := v_0.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64ULT
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (NE (CMPconst [0] (LessEqual cc)) yes no)
+		// cond:
+		// result: (LE cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64CMPconst {
+				break
+			}
+			if v.AuxInt != 0 {
+				break
+			}
+			v_0 := v.Args[0]
+			if v_0.Op != OpARM64LessEqual {
+				break
+			}
+			cc := v_0.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64LE
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (NE (CMPconst [0] (LessEqualU cc)) yes no)
+		// cond:
+		// result: (ULE cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64CMPconst {
+				break
+			}
+			if v.AuxInt != 0 {
+				break
+			}
+			v_0 := v.Args[0]
+			if v_0.Op != OpARM64LessEqualU {
+				break
+			}
+			cc := v_0.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64ULE
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (NE (CMPconst [0] (GreaterThan cc)) yes no)
+		// cond:
+		// result: (GT cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64CMPconst {
+				break
+			}
+			if v.AuxInt != 0 {
+				break
+			}
+			v_0 := v.Args[0]
+			if v_0.Op != OpARM64GreaterThan {
+				break
+			}
+			cc := v_0.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64GT
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (NE (CMPconst [0] (GreaterThanU cc)) yes no)
+		// cond:
+		// result: (UGT cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64CMPconst {
+				break
+			}
+			if v.AuxInt != 0 {
+				break
+			}
+			v_0 := v.Args[0]
+			if v_0.Op != OpARM64GreaterThanU {
+				break
+			}
+			cc := v_0.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64UGT
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (NE (CMPconst [0] (GreaterEqual cc)) yes no)
+		// cond:
+		// result: (GE cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64CMPconst {
+				break
+			}
+			if v.AuxInt != 0 {
+				break
+			}
+			v_0 := v.Args[0]
+			if v_0.Op != OpARM64GreaterEqual {
+				break
+			}
+			cc := v_0.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64GE
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+		// match: (NE (CMPconst [0] (GreaterEqualU cc)) yes no)
+		// cond:
+		// result: (UGE cc yes no)
+		for {
+			v := b.Control
+			if v.Op != OpARM64CMPconst {
+				break
+			}
+			if v.AuxInt != 0 {
+				break
+			}
+			v_0 := v.Args[0]
+			if v_0.Op != OpARM64GreaterEqualU {
+				break
+			}
+			cc := v_0.Args[0]
+			yes := b.Succs[0]
+			no := b.Succs[1]
+			b.Kind = BlockARM64UGE
+			b.SetControl(cc)
+			_ = yes
+			_ = no
+			return true
+		}
+	}
+	return false
+}