github.com/hikaru7719/go@v0.0.0-20181025140707-c8b2ac68906a/src/cmd/compile/internal/ssa/gen/AMD64Ops.go (about)

     1  // Copyright 2015 The Go Authors. All rights reserved.
     2  // Use of this source code is governed by a BSD-style
     3  // license that can be found in the LICENSE file.
     4  
     5  // +build ignore
     6  
     7  package main
     8  
     9  import "strings"
    10  
    11  // Notes:
    12  //  - Integer types live in the low portion of registers. Upper portions are junk.
    13  //  - Boolean types use the low-order byte of a register. 0=false, 1=true.
    14  //    Upper bytes are junk.
    15  //  - Floating-point types live in the low natural slot of an sse2 register.
    16  //    Unused portions are junk.
    17  //  - We do not use AH,BH,CH,DH registers.
    18  //  - When doing sub-register operations, we try to write the whole
    19  //    destination register to avoid a partial-register write.
    20  //  - Unused portions of AuxInt (or the Val portion of ValAndOff) are
    21  //    filled by sign-extending the used portion.  Users of AuxInt which interpret
    22  //    AuxInt as unsigned (e.g. shifts) must be careful.
    23  //  - All SymOff opcodes require their offset to fit in an int32.
    24  
    25  // Suffixes encode the bit width of various instructions.
    26  // Q (quad word) = 64 bit
    27  // L (long word) = 32 bit
    28  // W (word)      = 16 bit
    29  // B (byte)      = 8 bit
    30  
    31  // copied from ../../amd64/reg.go
    32  var regNamesAMD64 = []string{
    33  	"AX",
    34  	"CX",
    35  	"DX",
    36  	"BX",
    37  	"SP",
    38  	"BP",
    39  	"SI",
    40  	"DI",
    41  	"R8",
    42  	"R9",
    43  	"R10",
    44  	"R11",
    45  	"R12",
    46  	"R13",
    47  	"R14",
    48  	"R15",
    49  	"X0",
    50  	"X1",
    51  	"X2",
    52  	"X3",
    53  	"X4",
    54  	"X5",
    55  	"X6",
    56  	"X7",
    57  	"X8",
    58  	"X9",
    59  	"X10",
    60  	"X11",
    61  	"X12",
    62  	"X13",
    63  	"X14",
    64  	"X15",
    65  
    66  	// pseudo-registers
    67  	"SB",
    68  }
    69  
    70  func init() {
    71  	// Make map from reg names to reg integers.
    72  	if len(regNamesAMD64) > 64 {
    73  		panic("too many registers")
    74  	}
    75  	num := map[string]int{}
    76  	for i, name := range regNamesAMD64 {
    77  		num[name] = i
    78  	}
    79  	buildReg := func(s string) regMask {
    80  		m := regMask(0)
    81  		for _, r := range strings.Split(s, " ") {
    82  			if n, ok := num[r]; ok {
    83  				m |= regMask(1) << uint(n)
    84  				continue
    85  			}
    86  			panic("register " + r + " not found")
    87  		}
    88  		return m
    89  	}
    90  
    91  	// Common individual register masks
    92  	var (
    93  		ax         = buildReg("AX")
    94  		cx         = buildReg("CX")
    95  		dx         = buildReg("DX")
    96  		gp         = buildReg("AX CX DX BX BP SI DI R8 R9 R10 R11 R12 R13 R14 R15")
    97  		fp         = buildReg("X0 X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11 X12 X13 X14 X15")
    98  		gpsp       = gp | buildReg("SP")
    99  		gpspsb     = gpsp | buildReg("SB")
   100  		callerSave = gp | fp
   101  	)
   102  	// Common slices of register masks
   103  	var (
   104  		gponly = []regMask{gp}
   105  		fponly = []regMask{fp}
   106  	)
   107  
   108  	// Common regInfo
   109  	var (
   110  		gp01      = regInfo{inputs: nil, outputs: gponly}
   111  		gp11      = regInfo{inputs: []regMask{gp}, outputs: gponly}
   112  		gp11sp    = regInfo{inputs: []regMask{gpsp}, outputs: gponly}
   113  		gp11sb    = regInfo{inputs: []regMask{gpspsb}, outputs: gponly}
   114  		gp21      = regInfo{inputs: []regMask{gp, gp}, outputs: gponly}
   115  		gp21sp    = regInfo{inputs: []regMask{gpsp, gp}, outputs: gponly}
   116  		gp21sb    = regInfo{inputs: []regMask{gpspsb, gpsp}, outputs: gponly}
   117  		gp21shift = regInfo{inputs: []regMask{gp, cx}, outputs: []regMask{gp}}
   118  		gp11div   = regInfo{inputs: []regMask{ax, gpsp &^ dx}, outputs: []regMask{ax, dx}}
   119  		gp21hmul  = regInfo{inputs: []regMask{ax, gpsp}, outputs: []regMask{dx}, clobbers: ax}
   120  
   121  		gp2flags     = regInfo{inputs: []regMask{gpsp, gpsp}}
   122  		gp1flags     = regInfo{inputs: []regMask{gpsp}}
   123  		gp0flagsLoad = regInfo{inputs: []regMask{gpspsb, 0}}
   124  		gp1flagsLoad = regInfo{inputs: []regMask{gpspsb, gpsp, 0}}
   125  		flagsgp      = regInfo{inputs: nil, outputs: gponly}
   126  
   127  		gp11flags = regInfo{inputs: []regMask{gp}, outputs: []regMask{gp, 0}}
   128  
   129  		readflags = regInfo{inputs: nil, outputs: gponly}
   130  		flagsgpax = regInfo{inputs: nil, clobbers: ax, outputs: []regMask{gp &^ ax}}
   131  
   132  		gpload    = regInfo{inputs: []regMask{gpspsb, 0}, outputs: gponly}
   133  		gp21load  = regInfo{inputs: []regMask{gp, gpspsb, 0}, outputs: gponly}
   134  		gploadidx = regInfo{inputs: []regMask{gpspsb, gpsp, 0}, outputs: gponly}
   135  		gp21pax   = regInfo{inputs: []regMask{gp &^ ax, gp}, outputs: []regMask{gp &^ ax}, clobbers: ax}
   136  
   137  		gpstore         = regInfo{inputs: []regMask{gpspsb, gpsp, 0}}
   138  		gpstoreconst    = regInfo{inputs: []regMask{gpspsb, 0}}
   139  		gpstoreidx      = regInfo{inputs: []regMask{gpspsb, gpsp, gpsp, 0}}
   140  		gpstoreconstidx = regInfo{inputs: []regMask{gpspsb, gpsp, 0}}
   141  		gpstorexchg     = regInfo{inputs: []regMask{gp, gpspsb, 0}, outputs: []regMask{gp}}
   142  		cmpxchg         = regInfo{inputs: []regMask{gp, ax, gp, 0}, outputs: []regMask{gp, 0}, clobbers: ax}
   143  
   144  		fp01     = regInfo{inputs: nil, outputs: fponly}
   145  		fp21     = regInfo{inputs: []regMask{fp, fp}, outputs: fponly}
   146  		fp21load = regInfo{inputs: []regMask{fp, gpspsb, 0}, outputs: fponly}
   147  		fpgp     = regInfo{inputs: fponly, outputs: gponly}
   148  		gpfp     = regInfo{inputs: gponly, outputs: fponly}
   149  		fp11     = regInfo{inputs: fponly, outputs: fponly}
   150  		fp2flags = regInfo{inputs: []regMask{fp, fp}}
   151  
   152  		fpload    = regInfo{inputs: []regMask{gpspsb, 0}, outputs: fponly}
   153  		fploadidx = regInfo{inputs: []regMask{gpspsb, gpsp, 0}, outputs: fponly}
   154  
   155  		fpstore    = regInfo{inputs: []regMask{gpspsb, fp, 0}}
   156  		fpstoreidx = regInfo{inputs: []regMask{gpspsb, gpsp, fp, 0}}
   157  	)
   158  
   159  	var AMD64ops = []opData{
   160  		// fp ops
   161  		{name: "ADDSS", argLength: 2, reg: fp21, asm: "ADDSS", commutative: true, resultInArg0: true}, // fp32 add
   162  		{name: "ADDSD", argLength: 2, reg: fp21, asm: "ADDSD", commutative: true, resultInArg0: true}, // fp64 add
   163  		{name: "SUBSS", argLength: 2, reg: fp21, asm: "SUBSS", resultInArg0: true},                    // fp32 sub
   164  		{name: "SUBSD", argLength: 2, reg: fp21, asm: "SUBSD", resultInArg0: true},                    // fp64 sub
   165  		{name: "MULSS", argLength: 2, reg: fp21, asm: "MULSS", commutative: true, resultInArg0: true}, // fp32 mul
   166  		{name: "MULSD", argLength: 2, reg: fp21, asm: "MULSD", commutative: true, resultInArg0: true}, // fp64 mul
   167  		{name: "DIVSS", argLength: 2, reg: fp21, asm: "DIVSS", resultInArg0: true},                    // fp32 div
   168  		{name: "DIVSD", argLength: 2, reg: fp21, asm: "DIVSD", resultInArg0: true},                    // fp64 div
   169  
   170  		{name: "MOVSSload", argLength: 2, reg: fpload, asm: "MOVSS", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"}, // fp32 load
   171  		{name: "MOVSDload", argLength: 2, reg: fpload, asm: "MOVSD", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"}, // fp64 load
   172  		{name: "MOVSSconst", reg: fp01, asm: "MOVSS", aux: "Float32", rematerializeable: true},                               // fp32 constant
   173  		{name: "MOVSDconst", reg: fp01, asm: "MOVSD", aux: "Float64", rematerializeable: true},                               // fp64 constant
   174  		{name: "MOVSSloadidx1", argLength: 3, reg: fploadidx, asm: "MOVSS", aux: "SymOff", symEffect: "Read"},                // fp32 load indexed by i
   175  		{name: "MOVSSloadidx4", argLength: 3, reg: fploadidx, asm: "MOVSS", aux: "SymOff", symEffect: "Read"},                // fp32 load indexed by 4*i
   176  		{name: "MOVSDloadidx1", argLength: 3, reg: fploadidx, asm: "MOVSD", aux: "SymOff", symEffect: "Read"},                // fp64 load indexed by i
   177  		{name: "MOVSDloadidx8", argLength: 3, reg: fploadidx, asm: "MOVSD", aux: "SymOff", symEffect: "Read"},                // fp64 load indexed by 8*i
   178  
   179  		{name: "MOVSSstore", argLength: 3, reg: fpstore, asm: "MOVSS", aux: "SymOff", faultOnNilArg0: true, symEffect: "Write"}, // fp32 store
   180  		{name: "MOVSDstore", argLength: 3, reg: fpstore, asm: "MOVSD", aux: "SymOff", faultOnNilArg0: true, symEffect: "Write"}, // fp64 store
   181  		{name: "MOVSSstoreidx1", argLength: 4, reg: fpstoreidx, asm: "MOVSS", aux: "SymOff", symEffect: "Write"},                // fp32 indexed by i store
   182  		{name: "MOVSSstoreidx4", argLength: 4, reg: fpstoreidx, asm: "MOVSS", aux: "SymOff", symEffect: "Write"},                // fp32 indexed by 4i store
   183  		{name: "MOVSDstoreidx1", argLength: 4, reg: fpstoreidx, asm: "MOVSD", aux: "SymOff", symEffect: "Write"},                // fp64 indexed by i store
   184  		{name: "MOVSDstoreidx8", argLength: 4, reg: fpstoreidx, asm: "MOVSD", aux: "SymOff", symEffect: "Write"},                // fp64 indexed by 8i store
   185  
   186  		{name: "ADDSSload", argLength: 3, reg: fp21load, asm: "ADDSS", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true, symEffect: "Read"}, // fp32 arg0 + tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
   187  		{name: "ADDSDload", argLength: 3, reg: fp21load, asm: "ADDSD", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true, symEffect: "Read"}, // fp64 arg0 + tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
   188  		{name: "SUBSSload", argLength: 3, reg: fp21load, asm: "SUBSS", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true, symEffect: "Read"}, // fp32 arg0 - tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
   189  		{name: "SUBSDload", argLength: 3, reg: fp21load, asm: "SUBSD", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true, symEffect: "Read"}, // fp64 arg0 - tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
   190  		{name: "MULSSload", argLength: 3, reg: fp21load, asm: "MULSS", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true, symEffect: "Read"}, // fp32 arg0 * tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
   191  		{name: "MULSDload", argLength: 3, reg: fp21load, asm: "MULSD", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true, symEffect: "Read"}, // fp64 arg0 * tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
   192  		{name: "DIVSSload", argLength: 3, reg: fp21load, asm: "DIVSS", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true, symEffect: "Read"}, // fp32 arg0 / tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
   193  		{name: "DIVSDload", argLength: 3, reg: fp21load, asm: "DIVSD", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true, symEffect: "Read"}, // fp64 arg0 / tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
   194  
   195  		// binary ops
   196  		{name: "ADDQ", argLength: 2, reg: gp21sp, asm: "ADDQ", commutative: true, clobberFlags: true},                                                                   // arg0 + arg1
   197  		{name: "ADDL", argLength: 2, reg: gp21sp, asm: "ADDL", commutative: true, clobberFlags: true},                                                                   // arg0 + arg1
   198  		{name: "ADDQconst", argLength: 1, reg: gp11sp, asm: "ADDQ", aux: "Int32", typ: "UInt64", clobberFlags: true},                                                    // arg0 + auxint
   199  		{name: "ADDLconst", argLength: 1, reg: gp11sp, asm: "ADDL", aux: "Int32", clobberFlags: true},                                                                   // arg0 + auxint
   200  		{name: "ADDQconstmodify", argLength: 2, reg: gpstoreconst, asm: "ADDQ", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // add ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   201  		{name: "ADDLconstmodify", argLength: 2, reg: gpstoreconst, asm: "ADDL", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // add ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   202  
   203  		{name: "SUBQ", argLength: 2, reg: gp21, asm: "SUBQ", resultInArg0: true, clobberFlags: true},                    // arg0 - arg1
   204  		{name: "SUBL", argLength: 2, reg: gp21, asm: "SUBL", resultInArg0: true, clobberFlags: true},                    // arg0 - arg1
   205  		{name: "SUBQconst", argLength: 1, reg: gp11, asm: "SUBQ", aux: "Int32", resultInArg0: true, clobberFlags: true}, // arg0 - auxint
   206  		{name: "SUBLconst", argLength: 1, reg: gp11, asm: "SUBL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // arg0 - auxint
   207  
   208  		{name: "MULQ", argLength: 2, reg: gp21, asm: "IMULQ", commutative: true, resultInArg0: true, clobberFlags: true}, // arg0 * arg1
   209  		{name: "MULL", argLength: 2, reg: gp21, asm: "IMULL", commutative: true, resultInArg0: true, clobberFlags: true}, // arg0 * arg1
   210  		{name: "MULQconst", argLength: 1, reg: gp11, asm: "IMUL3Q", aux: "Int32", clobberFlags: true},                    // arg0 * auxint
   211  		{name: "MULLconst", argLength: 1, reg: gp11, asm: "IMUL3L", aux: "Int32", clobberFlags: true},                    // arg0 * auxint
   212  
   213  		{name: "MULLU", argLength: 2, reg: regInfo{inputs: []regMask{ax, gpsp}, outputs: []regMask{ax, 0}, clobbers: dx}, typ: "(UInt32,Flags)", asm: "MULL", commutative: true, clobberFlags: true}, // Let x = arg0*arg1 (full 32x32->64  unsigned multiply). Returns uint32(x), and flags set to overflow if uint32(x) != x.
   214  		{name: "MULQU", argLength: 2, reg: regInfo{inputs: []regMask{ax, gpsp}, outputs: []regMask{ax, 0}, clobbers: dx}, typ: "(UInt64,Flags)", asm: "MULQ", commutative: true, clobberFlags: true}, // Let x = arg0*arg1 (full 64x64->128 unsigned multiply). Returns uint64(x), and flags set to overflow if uint64(x) != x.
   215  
   216  		{name: "HMULQ", argLength: 2, reg: gp21hmul, commutative: true, asm: "IMULQ", clobberFlags: true}, // (arg0 * arg1) >> width
   217  		{name: "HMULL", argLength: 2, reg: gp21hmul, commutative: true, asm: "IMULL", clobberFlags: true}, // (arg0 * arg1) >> width
   218  		{name: "HMULQU", argLength: 2, reg: gp21hmul, commutative: true, asm: "MULQ", clobberFlags: true}, // (arg0 * arg1) >> width
   219  		{name: "HMULLU", argLength: 2, reg: gp21hmul, commutative: true, asm: "MULL", clobberFlags: true}, // (arg0 * arg1) >> width
   220  
   221  		{name: "AVGQU", argLength: 2, reg: gp21, commutative: true, resultInArg0: true, clobberFlags: true}, // (arg0 + arg1) / 2 as unsigned, all 64 result bits
   222  
   223  		// For DIVQ, DIVL and DIVW, AuxInt non-zero means that the divisor has been proved to be not -1.
   224  		{name: "DIVQ", argLength: 2, reg: gp11div, typ: "(Int64,Int64)", asm: "IDIVQ", aux: "Bool", clobberFlags: true}, // [arg0 / arg1, arg0 % arg1]
   225  		{name: "DIVL", argLength: 2, reg: gp11div, typ: "(Int32,Int32)", asm: "IDIVL", aux: "Bool", clobberFlags: true}, // [arg0 / arg1, arg0 % arg1]
   226  		{name: "DIVW", argLength: 2, reg: gp11div, typ: "(Int16,Int16)", asm: "IDIVW", aux: "Bool", clobberFlags: true}, // [arg0 / arg1, arg0 % arg1]
   227  
   228  		{name: "DIVQU", argLength: 2, reg: gp11div, typ: "(UInt64,UInt64)", asm: "DIVQ", clobberFlags: true}, // [arg0 / arg1, arg0 % arg1]
   229  		{name: "DIVLU", argLength: 2, reg: gp11div, typ: "(UInt32,UInt32)", asm: "DIVL", clobberFlags: true}, // [arg0 / arg1, arg0 % arg1]
   230  		{name: "DIVWU", argLength: 2, reg: gp11div, typ: "(UInt16,UInt16)", asm: "DIVW", clobberFlags: true}, // [arg0 / arg1, arg0 % arg1]
   231  
   232  		{name: "MULQU2", argLength: 2, reg: regInfo{inputs: []regMask{ax, gpsp}, outputs: []regMask{dx, ax}}, commutative: true, asm: "MULQ", clobberFlags: true}, // arg0 * arg1, returns (hi, lo)
   233  		{name: "DIVQU2", argLength: 3, reg: regInfo{inputs: []regMask{dx, ax, gpsp}, outputs: []regMask{ax, dx}}, asm: "DIVQ", clobberFlags: true},                // arg0:arg1 / arg2 (128-bit divided by 64-bit), returns (q, r)
   234  
   235  		{name: "ANDQ", argLength: 2, reg: gp21, asm: "ANDQ", commutative: true, resultInArg0: true, clobberFlags: true},                                                 // arg0 & arg1
   236  		{name: "ANDL", argLength: 2, reg: gp21, asm: "ANDL", commutative: true, resultInArg0: true, clobberFlags: true},                                                 // arg0 & arg1
   237  		{name: "ANDQconst", argLength: 1, reg: gp11, asm: "ANDQ", aux: "Int32", resultInArg0: true, clobberFlags: true},                                                 // arg0 & auxint
   238  		{name: "ANDLconst", argLength: 1, reg: gp11, asm: "ANDL", aux: "Int32", resultInArg0: true, clobberFlags: true},                                                 // arg0 & auxint
   239  		{name: "ANDQconstmodify", argLength: 2, reg: gpstoreconst, asm: "ANDQ", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // and ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   240  		{name: "ANDLconstmodify", argLength: 2, reg: gpstoreconst, asm: "ANDL", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // and ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   241  
   242  		{name: "ORQ", argLength: 2, reg: gp21, asm: "ORQ", commutative: true, resultInArg0: true, clobberFlags: true},                                                 // arg0 | arg1
   243  		{name: "ORL", argLength: 2, reg: gp21, asm: "ORL", commutative: true, resultInArg0: true, clobberFlags: true},                                                 // arg0 | arg1
   244  		{name: "ORQconst", argLength: 1, reg: gp11, asm: "ORQ", aux: "Int32", resultInArg0: true, clobberFlags: true},                                                 // arg0 | auxint
   245  		{name: "ORLconst", argLength: 1, reg: gp11, asm: "ORL", aux: "Int32", resultInArg0: true, clobberFlags: true},                                                 // arg0 | auxint
   246  		{name: "ORQconstmodify", argLength: 2, reg: gpstoreconst, asm: "ORQ", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // or ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   247  		{name: "ORLconstmodify", argLength: 2, reg: gpstoreconst, asm: "ORL", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // or ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   248  
   249  		{name: "XORQ", argLength: 2, reg: gp21, asm: "XORQ", commutative: true, resultInArg0: true, clobberFlags: true},                                                 // arg0 ^ arg1
   250  		{name: "XORL", argLength: 2, reg: gp21, asm: "XORL", commutative: true, resultInArg0: true, clobberFlags: true},                                                 // arg0 ^ arg1
   251  		{name: "XORQconst", argLength: 1, reg: gp11, asm: "XORQ", aux: "Int32", resultInArg0: true, clobberFlags: true},                                                 // arg0 ^ auxint
   252  		{name: "XORLconst", argLength: 1, reg: gp11, asm: "XORL", aux: "Int32", resultInArg0: true, clobberFlags: true},                                                 // arg0 ^ auxint
   253  		{name: "XORQconstmodify", argLength: 2, reg: gpstoreconst, asm: "XORQ", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // xor ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   254  		{name: "XORLconstmodify", argLength: 2, reg: gpstoreconst, asm: "XORL", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // xor ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   255  
   256  		{name: "CMPQ", argLength: 2, reg: gp2flags, asm: "CMPQ", typ: "Flags"},                    // arg0 compare to arg1
   257  		{name: "CMPL", argLength: 2, reg: gp2flags, asm: "CMPL", typ: "Flags"},                    // arg0 compare to arg1
   258  		{name: "CMPW", argLength: 2, reg: gp2flags, asm: "CMPW", typ: "Flags"},                    // arg0 compare to arg1
   259  		{name: "CMPB", argLength: 2, reg: gp2flags, asm: "CMPB", typ: "Flags"},                    // arg0 compare to arg1
   260  		{name: "CMPQconst", argLength: 1, reg: gp1flags, asm: "CMPQ", typ: "Flags", aux: "Int32"}, // arg0 compare to auxint
   261  		{name: "CMPLconst", argLength: 1, reg: gp1flags, asm: "CMPL", typ: "Flags", aux: "Int32"}, // arg0 compare to auxint
   262  		{name: "CMPWconst", argLength: 1, reg: gp1flags, asm: "CMPW", typ: "Flags", aux: "Int16"}, // arg0 compare to auxint
   263  		{name: "CMPBconst", argLength: 1, reg: gp1flags, asm: "CMPB", typ: "Flags", aux: "Int8"},  // arg0 compare to auxint
   264  
   265  		// compare *(arg0+auxint+aux) to arg1 (in that order). arg2=mem.
   266  		{name: "CMPQload", argLength: 3, reg: gp1flagsLoad, asm: "CMPQ", aux: "SymOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   267  		{name: "CMPLload", argLength: 3, reg: gp1flagsLoad, asm: "CMPL", aux: "SymOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   268  		{name: "CMPWload", argLength: 3, reg: gp1flagsLoad, asm: "CMPW", aux: "SymOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   269  		{name: "CMPBload", argLength: 3, reg: gp1flagsLoad, asm: "CMPB", aux: "SymOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   270  
   271  		// compare *(arg0+ValAndOff(AuxInt).Off()+aux) to ValAndOff(AuxInt).Val() (in that order). arg1=mem.
   272  		{name: "CMPQconstload", argLength: 2, reg: gp0flagsLoad, asm: "CMPQ", aux: "SymValAndOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   273  		{name: "CMPLconstload", argLength: 2, reg: gp0flagsLoad, asm: "CMPL", aux: "SymValAndOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   274  		{name: "CMPWconstload", argLength: 2, reg: gp0flagsLoad, asm: "CMPW", aux: "SymValAndOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   275  		{name: "CMPBconstload", argLength: 2, reg: gp0flagsLoad, asm: "CMPB", aux: "SymValAndOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   276  
   277  		{name: "UCOMISS", argLength: 2, reg: fp2flags, asm: "UCOMISS", typ: "Flags"}, // arg0 compare to arg1, f32
   278  		{name: "UCOMISD", argLength: 2, reg: fp2flags, asm: "UCOMISD", typ: "Flags"}, // arg0 compare to arg1, f64
   279  
   280  		{name: "BTL", argLength: 2, reg: gp2flags, asm: "BTL", typ: "Flags"},                                           // test whether bit arg0%32 in arg1 is set
   281  		{name: "BTQ", argLength: 2, reg: gp2flags, asm: "BTQ", typ: "Flags"},                                           // test whether bit arg0%64 in arg1 is set
   282  		{name: "BTCL", argLength: 2, reg: gp21, asm: "BTCL", resultInArg0: true, clobberFlags: true},                   // complement bit arg1%32 in arg0
   283  		{name: "BTCQ", argLength: 2, reg: gp21, asm: "BTCQ", resultInArg0: true, clobberFlags: true},                   // complement bit arg1%64 in arg0
   284  		{name: "BTRL", argLength: 2, reg: gp21, asm: "BTRL", resultInArg0: true, clobberFlags: true},                   // reset bit arg1%32 in arg0
   285  		{name: "BTRQ", argLength: 2, reg: gp21, asm: "BTRQ", resultInArg0: true, clobberFlags: true},                   // reset bit arg1%64 in arg0
   286  		{name: "BTSL", argLength: 2, reg: gp21, asm: "BTSL", resultInArg0: true, clobberFlags: true},                   // set bit arg1%32 in arg0
   287  		{name: "BTSQ", argLength: 2, reg: gp21, asm: "BTSQ", resultInArg0: true, clobberFlags: true},                   // set bit arg1%64 in arg0
   288  		{name: "BTLconst", argLength: 1, reg: gp1flags, asm: "BTL", typ: "Flags", aux: "Int8"},                         // test whether bit auxint in arg0 is set, 0 <= auxint < 32
   289  		{name: "BTQconst", argLength: 1, reg: gp1flags, asm: "BTQ", typ: "Flags", aux: "Int8"},                         // test whether bit auxint in arg0 is set, 0 <= auxint < 64
   290  		{name: "BTCLconst", argLength: 1, reg: gp11, asm: "BTCL", resultInArg0: true, clobberFlags: true, aux: "Int8"}, // complement bit auxint in arg0, 0 <= auxint < 32
   291  		{name: "BTCQconst", argLength: 1, reg: gp11, asm: "BTCQ", resultInArg0: true, clobberFlags: true, aux: "Int8"}, // complement bit auxint in arg0, 0 <= auxint < 64
   292  		{name: "BTRLconst", argLength: 1, reg: gp11, asm: "BTRL", resultInArg0: true, clobberFlags: true, aux: "Int8"}, // reset bit auxint in arg0, 0 <= auxint < 32
   293  		{name: "BTRQconst", argLength: 1, reg: gp11, asm: "BTRQ", resultInArg0: true, clobberFlags: true, aux: "Int8"}, // reset bit auxint in arg0, 0 <= auxint < 64
   294  		{name: "BTSLconst", argLength: 1, reg: gp11, asm: "BTSL", resultInArg0: true, clobberFlags: true, aux: "Int8"}, // set bit auxint in arg0, 0 <= auxint < 32
   295  		{name: "BTSQconst", argLength: 1, reg: gp11, asm: "BTSQ", resultInArg0: true, clobberFlags: true, aux: "Int8"}, // set bit auxint in arg0, 0 <= auxint < 64
   296  
   297  		// direct bit operation on memory operand
   298  		{name: "BTCQmodify", argLength: 3, reg: gpstore, asm: "BTCQ", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"},     // complement bit arg1 in 64-bit arg0+auxint+aux, arg2=mem
   299  		{name: "BTCLmodify", argLength: 3, reg: gpstore, asm: "BTCL", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"},     // complement bit arg1 in 32-bit arg0+auxint+aux, arg2=mem
   300  		{name: "BTSQmodify", argLength: 3, reg: gpstore, asm: "BTSQ", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"},     // set bit arg1 in 64-bit arg0+auxint+aux, arg2=mem
   301  		{name: "BTSLmodify", argLength: 3, reg: gpstore, asm: "BTSL", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"},     // set bit arg1 in 32-bit arg0+auxint+aux, arg2=mem
   302  		{name: "BTRQmodify", argLength: 3, reg: gpstore, asm: "BTRQ", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"},     // reset bit arg1 in 64-bit arg0+auxint+aux, arg2=mem
   303  		{name: "BTRLmodify", argLength: 3, reg: gpstore, asm: "BTRL", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"},     // reset bit arg1 in 32-bit arg0+auxint+aux, arg2=mem
   304  		{name: "BTCQconstmodify", argLength: 2, reg: gpstoreconst, asm: "BTCQ", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // complement bit ValAndOff(AuxInt).Val() in 64-bit arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   305  		{name: "BTCLconstmodify", argLength: 2, reg: gpstoreconst, asm: "BTCL", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // complement bit ValAndOff(AuxInt).Val() in 32-bit arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   306  		{name: "BTSQconstmodify", argLength: 2, reg: gpstoreconst, asm: "BTSQ", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // set bit ValAndOff(AuxInt).Val() in 64-bit arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   307  		{name: "BTSLconstmodify", argLength: 2, reg: gpstoreconst, asm: "BTSL", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // set bit ValAndOff(AuxInt).Val() in 32-bit arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   308  		{name: "BTRQconstmodify", argLength: 2, reg: gpstoreconst, asm: "BTRQ", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // reset bit ValAndOff(AuxInt).Val() in 64-bit arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   309  		{name: "BTRLconstmodify", argLength: 2, reg: gpstoreconst, asm: "BTRL", aux: "SymValAndOff", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // reset bit ValAndOff(AuxInt).Val() in 32-bit arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   310  
   311  		{name: "TESTQ", argLength: 2, reg: gp2flags, commutative: true, asm: "TESTQ", typ: "Flags"}, // (arg0 & arg1) compare to 0
   312  		{name: "TESTL", argLength: 2, reg: gp2flags, commutative: true, asm: "TESTL", typ: "Flags"}, // (arg0 & arg1) compare to 0
   313  		{name: "TESTW", argLength: 2, reg: gp2flags, commutative: true, asm: "TESTW", typ: "Flags"}, // (arg0 & arg1) compare to 0
   314  		{name: "TESTB", argLength: 2, reg: gp2flags, commutative: true, asm: "TESTB", typ: "Flags"}, // (arg0 & arg1) compare to 0
   315  		{name: "TESTQconst", argLength: 1, reg: gp1flags, asm: "TESTQ", typ: "Flags", aux: "Int32"}, // (arg0 & auxint) compare to 0
   316  		{name: "TESTLconst", argLength: 1, reg: gp1flags, asm: "TESTL", typ: "Flags", aux: "Int32"}, // (arg0 & auxint) compare to 0
   317  		{name: "TESTWconst", argLength: 1, reg: gp1flags, asm: "TESTW", typ: "Flags", aux: "Int16"}, // (arg0 & auxint) compare to 0
   318  		{name: "TESTBconst", argLength: 1, reg: gp1flags, asm: "TESTB", typ: "Flags", aux: "Int8"},  // (arg0 & auxint) compare to 0
   319  
   320  		{name: "SHLQ", argLength: 2, reg: gp21shift, asm: "SHLQ", resultInArg0: true, clobberFlags: true},              // arg0 << arg1, shift amount is mod 64
   321  		{name: "SHLL", argLength: 2, reg: gp21shift, asm: "SHLL", resultInArg0: true, clobberFlags: true},              // arg0 << arg1, shift amount is mod 32
   322  		{name: "SHLQconst", argLength: 1, reg: gp11, asm: "SHLQ", aux: "Int8", resultInArg0: true, clobberFlags: true}, // arg0 << auxint, shift amount 0-63
   323  		{name: "SHLLconst", argLength: 1, reg: gp11, asm: "SHLL", aux: "Int8", resultInArg0: true, clobberFlags: true}, // arg0 << auxint, shift amount 0-31
   324  		// Note: x86 is weird, the 16 and 8 byte shifts still use all 5 bits of shift amount!
   325  
   326  		{name: "SHRQ", argLength: 2, reg: gp21shift, asm: "SHRQ", resultInArg0: true, clobberFlags: true},              // unsigned arg0 >> arg1, shift amount is mod 64
   327  		{name: "SHRL", argLength: 2, reg: gp21shift, asm: "SHRL", resultInArg0: true, clobberFlags: true},              // unsigned uint32(arg0) >> arg1, shift amount is mod 32
   328  		{name: "SHRW", argLength: 2, reg: gp21shift, asm: "SHRW", resultInArg0: true, clobberFlags: true},              // unsigned uint16(arg0) >> arg1, shift amount is mod 32
   329  		{name: "SHRB", argLength: 2, reg: gp21shift, asm: "SHRB", resultInArg0: true, clobberFlags: true},              // unsigned uint8(arg0) >> arg1, shift amount is mod 32
   330  		{name: "SHRQconst", argLength: 1, reg: gp11, asm: "SHRQ", aux: "Int8", resultInArg0: true, clobberFlags: true}, // unsigned arg0 >> auxint, shift amount 0-63
   331  		{name: "SHRLconst", argLength: 1, reg: gp11, asm: "SHRL", aux: "Int8", resultInArg0: true, clobberFlags: true}, // unsigned uint32(arg0) >> auxint, shift amount 0-31
   332  		{name: "SHRWconst", argLength: 1, reg: gp11, asm: "SHRW", aux: "Int8", resultInArg0: true, clobberFlags: true}, // unsigned uint16(arg0) >> auxint, shift amount 0-15
   333  		{name: "SHRBconst", argLength: 1, reg: gp11, asm: "SHRB", aux: "Int8", resultInArg0: true, clobberFlags: true}, // unsigned uint8(arg0) >> auxint, shift amount 0-7
   334  
   335  		{name: "SARQ", argLength: 2, reg: gp21shift, asm: "SARQ", resultInArg0: true, clobberFlags: true},              // signed arg0 >> arg1, shift amount is mod 64
   336  		{name: "SARL", argLength: 2, reg: gp21shift, asm: "SARL", resultInArg0: true, clobberFlags: true},              // signed int32(arg0) >> arg1, shift amount is mod 32
   337  		{name: "SARW", argLength: 2, reg: gp21shift, asm: "SARW", resultInArg0: true, clobberFlags: true},              // signed int16(arg0) >> arg1, shift amount is mod 32
   338  		{name: "SARB", argLength: 2, reg: gp21shift, asm: "SARB", resultInArg0: true, clobberFlags: true},              // signed int8(arg0) >> arg1, shift amount is mod 32
   339  		{name: "SARQconst", argLength: 1, reg: gp11, asm: "SARQ", aux: "Int8", resultInArg0: true, clobberFlags: true}, // signed arg0 >> auxint, shift amount 0-63
   340  		{name: "SARLconst", argLength: 1, reg: gp11, asm: "SARL", aux: "Int8", resultInArg0: true, clobberFlags: true}, // signed int32(arg0) >> auxint, shift amount 0-31
   341  		{name: "SARWconst", argLength: 1, reg: gp11, asm: "SARW", aux: "Int8", resultInArg0: true, clobberFlags: true}, // signed int16(arg0) >> auxint, shift amount 0-15
   342  		{name: "SARBconst", argLength: 1, reg: gp11, asm: "SARB", aux: "Int8", resultInArg0: true, clobberFlags: true}, // signed int8(arg0) >> auxint, shift amount 0-7
   343  
   344  		{name: "ROLQ", argLength: 2, reg: gp21shift, asm: "ROLQ", resultInArg0: true, clobberFlags: true},              // arg0 rotate left arg1 bits.
   345  		{name: "ROLL", argLength: 2, reg: gp21shift, asm: "ROLL", resultInArg0: true, clobberFlags: true},              // arg0 rotate left arg1 bits.
   346  		{name: "ROLW", argLength: 2, reg: gp21shift, asm: "ROLW", resultInArg0: true, clobberFlags: true},              // arg0 rotate left arg1 bits.
   347  		{name: "ROLB", argLength: 2, reg: gp21shift, asm: "ROLB", resultInArg0: true, clobberFlags: true},              // arg0 rotate left arg1 bits.
   348  		{name: "RORQ", argLength: 2, reg: gp21shift, asm: "RORQ", resultInArg0: true, clobberFlags: true},              // arg0 rotate right arg1 bits.
   349  		{name: "RORL", argLength: 2, reg: gp21shift, asm: "RORL", resultInArg0: true, clobberFlags: true},              // arg0 rotate right arg1 bits.
   350  		{name: "RORW", argLength: 2, reg: gp21shift, asm: "RORW", resultInArg0: true, clobberFlags: true},              // arg0 rotate right arg1 bits.
   351  		{name: "RORB", argLength: 2, reg: gp21shift, asm: "RORB", resultInArg0: true, clobberFlags: true},              // arg0 rotate right arg1 bits.
   352  		{name: "ROLQconst", argLength: 1, reg: gp11, asm: "ROLQ", aux: "Int8", resultInArg0: true, clobberFlags: true}, // arg0 rotate left auxint, rotate amount 0-63
   353  		{name: "ROLLconst", argLength: 1, reg: gp11, asm: "ROLL", aux: "Int8", resultInArg0: true, clobberFlags: true}, // arg0 rotate left auxint, rotate amount 0-31
   354  		{name: "ROLWconst", argLength: 1, reg: gp11, asm: "ROLW", aux: "Int8", resultInArg0: true, clobberFlags: true}, // arg0 rotate left auxint, rotate amount 0-15
   355  		{name: "ROLBconst", argLength: 1, reg: gp11, asm: "ROLB", aux: "Int8", resultInArg0: true, clobberFlags: true}, // arg0 rotate left auxint, rotate amount 0-7
   356  
   357  		{name: "ADDLload", argLength: 3, reg: gp21load, asm: "ADDL", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"}, // arg0 + tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   358  		{name: "ADDQload", argLength: 3, reg: gp21load, asm: "ADDQ", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"}, // arg0 + tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   359  		{name: "SUBQload", argLength: 3, reg: gp21load, asm: "SUBQ", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"}, // arg0 - tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   360  		{name: "SUBLload", argLength: 3, reg: gp21load, asm: "SUBL", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"}, // arg0 - tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   361  		{name: "ANDLload", argLength: 3, reg: gp21load, asm: "ANDL", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"}, // arg0 & tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   362  		{name: "ANDQload", argLength: 3, reg: gp21load, asm: "ANDQ", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"}, // arg0 & tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   363  		{name: "ORQload", argLength: 3, reg: gp21load, asm: "ORQ", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"},   // arg0 | tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   364  		{name: "ORLload", argLength: 3, reg: gp21load, asm: "ORL", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"},   // arg0 | tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   365  		{name: "XORQload", argLength: 3, reg: gp21load, asm: "XORQ", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"}, // arg0 ^ tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   366  		{name: "XORLload", argLength: 3, reg: gp21load, asm: "XORL", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"}, // arg0 ^ tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   367  
   368  		// direct binary-op on memory (read-modify-write)
   369  		{name: "ADDQmodify", argLength: 3, reg: gpstore, asm: "ADDQ", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) += arg1, arg2=mem
   370  		{name: "SUBQmodify", argLength: 3, reg: gpstore, asm: "SUBQ", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) -= arg1, arg2=mem
   371  		{name: "ANDQmodify", argLength: 3, reg: gpstore, asm: "ANDQ", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) &= arg1, arg2=mem
   372  		{name: "ORQmodify", argLength: 3, reg: gpstore, asm: "ORQ", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"},   // *(arg0+auxint+aux) |= arg1, arg2=mem
   373  		{name: "XORQmodify", argLength: 3, reg: gpstore, asm: "XORQ", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) ^= arg1, arg2=mem
   374  		{name: "ADDLmodify", argLength: 3, reg: gpstore, asm: "ADDL", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) += arg1, arg2=mem
   375  		{name: "SUBLmodify", argLength: 3, reg: gpstore, asm: "SUBL", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) -= arg1, arg2=mem
   376  		{name: "ANDLmodify", argLength: 3, reg: gpstore, asm: "ANDL", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) &= arg1, arg2=mem
   377  		{name: "ORLmodify", argLength: 3, reg: gpstore, asm: "ORL", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"},   // *(arg0+auxint+aux) |= arg1, arg2=mem
   378  		{name: "XORLmodify", argLength: 3, reg: gpstore, asm: "XORL", aux: "SymOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) ^= arg1, arg2=mem
   379  
   380  		// unary ops
   381  		{name: "NEGQ", argLength: 1, reg: gp11, asm: "NEGQ", resultInArg0: true, clobberFlags: true}, // -arg0
   382  		{name: "NEGL", argLength: 1, reg: gp11, asm: "NEGL", resultInArg0: true, clobberFlags: true}, // -arg0
   383  
   384  		{name: "NOTQ", argLength: 1, reg: gp11, asm: "NOTQ", resultInArg0: true, clobberFlags: true}, // ^arg0
   385  		{name: "NOTL", argLength: 1, reg: gp11, asm: "NOTL", resultInArg0: true, clobberFlags: true}, // ^arg0
   386  
   387  		// BS{F,R}Q returns a tuple [result, flags]
   388  		// result is undefined if the input is zero.
   389  		// flags are set to "equal" if the input is zero, "not equal" otherwise.
   390  		// BS{F,R}L returns only the result.
   391  		{name: "BSFQ", argLength: 1, reg: gp11flags, asm: "BSFQ", typ: "(UInt64,Flags)"},        // # of low-order zeroes in 64-bit arg
   392  		{name: "BSFL", argLength: 1, reg: gp11, asm: "BSFL", typ: "UInt32", clobberFlags: true}, // # of low-order zeroes in 32-bit arg
   393  		{name: "BSRQ", argLength: 1, reg: gp11flags, asm: "BSRQ", typ: "(UInt64,Flags)"},        // # of high-order zeroes in 64-bit arg
   394  		{name: "BSRL", argLength: 1, reg: gp11, asm: "BSRL", typ: "UInt32", clobberFlags: true}, // # of high-order zeroes in 32-bit arg
   395  
   396  		// CMOV instructions: 64, 32 and 16-bit sizes.
   397  		// if arg2 encodes a true result, return arg1, else arg0
   398  		{name: "CMOVQEQ", argLength: 3, reg: gp21, asm: "CMOVQEQ", resultInArg0: true},
   399  		{name: "CMOVQNE", argLength: 3, reg: gp21, asm: "CMOVQNE", resultInArg0: true},
   400  		{name: "CMOVQLT", argLength: 3, reg: gp21, asm: "CMOVQLT", resultInArg0: true},
   401  		{name: "CMOVQGT", argLength: 3, reg: gp21, asm: "CMOVQGT", resultInArg0: true},
   402  		{name: "CMOVQLE", argLength: 3, reg: gp21, asm: "CMOVQLE", resultInArg0: true},
   403  		{name: "CMOVQGE", argLength: 3, reg: gp21, asm: "CMOVQGE", resultInArg0: true},
   404  		{name: "CMOVQLS", argLength: 3, reg: gp21, asm: "CMOVQLS", resultInArg0: true},
   405  		{name: "CMOVQHI", argLength: 3, reg: gp21, asm: "CMOVQHI", resultInArg0: true},
   406  		{name: "CMOVQCC", argLength: 3, reg: gp21, asm: "CMOVQCC", resultInArg0: true},
   407  		{name: "CMOVQCS", argLength: 3, reg: gp21, asm: "CMOVQCS", resultInArg0: true},
   408  
   409  		{name: "CMOVLEQ", argLength: 3, reg: gp21, asm: "CMOVLEQ", resultInArg0: true},
   410  		{name: "CMOVLNE", argLength: 3, reg: gp21, asm: "CMOVLNE", resultInArg0: true},
   411  		{name: "CMOVLLT", argLength: 3, reg: gp21, asm: "CMOVLLT", resultInArg0: true},
   412  		{name: "CMOVLGT", argLength: 3, reg: gp21, asm: "CMOVLGT", resultInArg0: true},
   413  		{name: "CMOVLLE", argLength: 3, reg: gp21, asm: "CMOVLLE", resultInArg0: true},
   414  		{name: "CMOVLGE", argLength: 3, reg: gp21, asm: "CMOVLGE", resultInArg0: true},
   415  		{name: "CMOVLLS", argLength: 3, reg: gp21, asm: "CMOVLLS", resultInArg0: true},
   416  		{name: "CMOVLHI", argLength: 3, reg: gp21, asm: "CMOVLHI", resultInArg0: true},
   417  		{name: "CMOVLCC", argLength: 3, reg: gp21, asm: "CMOVLCC", resultInArg0: true},
   418  		{name: "CMOVLCS", argLength: 3, reg: gp21, asm: "CMOVLCS", resultInArg0: true},
   419  
   420  		{name: "CMOVWEQ", argLength: 3, reg: gp21, asm: "CMOVWEQ", resultInArg0: true},
   421  		{name: "CMOVWNE", argLength: 3, reg: gp21, asm: "CMOVWNE", resultInArg0: true},
   422  		{name: "CMOVWLT", argLength: 3, reg: gp21, asm: "CMOVWLT", resultInArg0: true},
   423  		{name: "CMOVWGT", argLength: 3, reg: gp21, asm: "CMOVWGT", resultInArg0: true},
   424  		{name: "CMOVWLE", argLength: 3, reg: gp21, asm: "CMOVWLE", resultInArg0: true},
   425  		{name: "CMOVWGE", argLength: 3, reg: gp21, asm: "CMOVWGE", resultInArg0: true},
   426  		{name: "CMOVWLS", argLength: 3, reg: gp21, asm: "CMOVWLS", resultInArg0: true},
   427  		{name: "CMOVWHI", argLength: 3, reg: gp21, asm: "CMOVWHI", resultInArg0: true},
   428  		{name: "CMOVWCC", argLength: 3, reg: gp21, asm: "CMOVWCC", resultInArg0: true},
   429  		{name: "CMOVWCS", argLength: 3, reg: gp21, asm: "CMOVWCS", resultInArg0: true},
   430  
   431  		// CMOV with floating point instructions. We need separate pseudo-op to handle
   432  		// InvertFlags correctly, and to generate special code that handles NaN (unordered flag).
   433  		// NOTE: the fact that CMOV*EQF here is marked to generate CMOV*NE is not a bug. See
   434  		// code generation in amd64/ssa.go.
   435  		{name: "CMOVQEQF", argLength: 3, reg: gp21pax, asm: "CMOVQNE", resultInArg0: true},
   436  		{name: "CMOVQNEF", argLength: 3, reg: gp21, asm: "CMOVQNE", resultInArg0: true},
   437  		{name: "CMOVQGTF", argLength: 3, reg: gp21, asm: "CMOVQHI", resultInArg0: true},
   438  		{name: "CMOVQGEF", argLength: 3, reg: gp21, asm: "CMOVQCC", resultInArg0: true},
   439  		{name: "CMOVLEQF", argLength: 3, reg: gp21pax, asm: "CMOVLNE", resultInArg0: true},
   440  		{name: "CMOVLNEF", argLength: 3, reg: gp21, asm: "CMOVLNE", resultInArg0: true},
   441  		{name: "CMOVLGTF", argLength: 3, reg: gp21, asm: "CMOVLHI", resultInArg0: true},
   442  		{name: "CMOVLGEF", argLength: 3, reg: gp21, asm: "CMOVLCC", resultInArg0: true},
   443  		{name: "CMOVWEQF", argLength: 3, reg: gp21pax, asm: "CMOVWNE", resultInArg0: true},
   444  		{name: "CMOVWNEF", argLength: 3, reg: gp21, asm: "CMOVWNE", resultInArg0: true},
   445  		{name: "CMOVWGTF", argLength: 3, reg: gp21, asm: "CMOVWHI", resultInArg0: true},
   446  		{name: "CMOVWGEF", argLength: 3, reg: gp21, asm: "CMOVWCC", resultInArg0: true},
   447  
   448  		{name: "BSWAPQ", argLength: 1, reg: gp11, asm: "BSWAPQ", resultInArg0: true, clobberFlags: true}, // arg0 swap bytes
   449  		{name: "BSWAPL", argLength: 1, reg: gp11, asm: "BSWAPL", resultInArg0: true, clobberFlags: true}, // arg0 swap bytes
   450  
   451  		// POPCNT instructions aren't guaranteed to be on the target platform (they are SSE4).
   452  		// Any use must be preceded by a successful check of runtime.support_popcnt.
   453  		{name: "POPCNTQ", argLength: 1, reg: gp11, asm: "POPCNTQ", clobberFlags: true}, // count number of set bits in arg0
   454  		{name: "POPCNTL", argLength: 1, reg: gp11, asm: "POPCNTL", clobberFlags: true}, // count number of set bits in arg0
   455  
   456  		{name: "SQRTSD", argLength: 1, reg: fp11, asm: "SQRTSD"}, // sqrt(arg0)
   457  
   458  		// ROUNDSD instruction isn't guaranteed to be on the target platform (it is SSE4.1)
   459  		// Any use must be preceded by a successful check of runtime.support_sse41.
   460  		{name: "ROUNDSD", argLength: 1, reg: fp11, aux: "Int8", asm: "ROUNDSD"}, // rounds arg0 depending on auxint, 1 means math.Floor, 2 Ceil, 3 Trunc
   461  
   462  		{name: "SBBQcarrymask", argLength: 1, reg: flagsgp, asm: "SBBQ"}, // (int64)(-1) if carry is set, 0 if carry is clear.
   463  		{name: "SBBLcarrymask", argLength: 1, reg: flagsgp, asm: "SBBL"}, // (int32)(-1) if carry is set, 0 if carry is clear.
   464  		// Note: SBBW and SBBB are subsumed by SBBL
   465  
   466  		{name: "SETEQ", argLength: 1, reg: readflags, asm: "SETEQ"}, // extract == condition from arg0
   467  		{name: "SETNE", argLength: 1, reg: readflags, asm: "SETNE"}, // extract != condition from arg0
   468  		{name: "SETL", argLength: 1, reg: readflags, asm: "SETLT"},  // extract signed < condition from arg0
   469  		{name: "SETLE", argLength: 1, reg: readflags, asm: "SETLE"}, // extract signed <= condition from arg0
   470  		{name: "SETG", argLength: 1, reg: readflags, asm: "SETGT"},  // extract signed > condition from arg0
   471  		{name: "SETGE", argLength: 1, reg: readflags, asm: "SETGE"}, // extract signed >= condition from arg0
   472  		{name: "SETB", argLength: 1, reg: readflags, asm: "SETCS"},  // extract unsigned < condition from arg0
   473  		{name: "SETBE", argLength: 1, reg: readflags, asm: "SETLS"}, // extract unsigned <= condition from arg0
   474  		{name: "SETA", argLength: 1, reg: readflags, asm: "SETHI"},  // extract unsigned > condition from arg0
   475  		{name: "SETAE", argLength: 1, reg: readflags, asm: "SETCC"}, // extract unsigned >= condition from arg0
   476  		{name: "SETO", argLength: 1, reg: readflags, asm: "SETOS"},  // extract if overflow flag is set from arg0
   477  		// Variants that store result to memory
   478  		{name: "SETEQstore", argLength: 3, reg: gpstoreconst, asm: "SETEQ", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // extract == condition from arg1 to arg0+auxint+aux, arg2=mem
   479  		{name: "SETNEstore", argLength: 3, reg: gpstoreconst, asm: "SETNE", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // extract != condition from arg1 to arg0+auxint+aux, arg2=mem
   480  		{name: "SETLstore", argLength: 3, reg: gpstoreconst, asm: "SETLT", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"},  // extract signed < condition from arg1 to arg0+auxint+aux, arg2=mem
   481  		{name: "SETLEstore", argLength: 3, reg: gpstoreconst, asm: "SETLE", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // extract signed <= condition from arg1 to arg0+auxint+aux, arg2=mem
   482  		{name: "SETGstore", argLength: 3, reg: gpstoreconst, asm: "SETGT", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"},  // extract signed > condition from arg1 to arg0+auxint+aux, arg2=mem
   483  		{name: "SETGEstore", argLength: 3, reg: gpstoreconst, asm: "SETGE", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // extract signed >= condition from arg1 to arg0+auxint+aux, arg2=mem
   484  		{name: "SETBstore", argLength: 3, reg: gpstoreconst, asm: "SETCS", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"},  // extract unsigned < condition from arg1 to arg0+auxint+aux, arg2=mem
   485  		{name: "SETBEstore", argLength: 3, reg: gpstoreconst, asm: "SETLS", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // extract unsigned <= condition from arg1 to arg0+auxint+aux, arg2=mem
   486  		{name: "SETAstore", argLength: 3, reg: gpstoreconst, asm: "SETHI", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"},  // extract unsigned > condition from arg1 to arg0+auxint+aux, arg2=mem
   487  		{name: "SETAEstore", argLength: 3, reg: gpstoreconst, asm: "SETCC", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // extract unsigned >= condition from arg1 to arg0+auxint+aux, arg2=mem
   488  		// Need different opcodes for floating point conditions because
   489  		// any comparison involving a NaN is always FALSE and thus
   490  		// the patterns for inverting conditions cannot be used.
   491  		{name: "SETEQF", argLength: 1, reg: flagsgpax, asm: "SETEQ", clobberFlags: true}, // extract == condition from arg0
   492  		{name: "SETNEF", argLength: 1, reg: flagsgpax, asm: "SETNE", clobberFlags: true}, // extract != condition from arg0
   493  		{name: "SETORD", argLength: 1, reg: flagsgp, asm: "SETPC"},                       // extract "ordered" (No Nan present) condition from arg0
   494  		{name: "SETNAN", argLength: 1, reg: flagsgp, asm: "SETPS"},                       // extract "unordered" (Nan present) condition from arg0
   495  
   496  		{name: "SETGF", argLength: 1, reg: flagsgp, asm: "SETHI"},  // extract floating > condition from arg0
   497  		{name: "SETGEF", argLength: 1, reg: flagsgp, asm: "SETCC"}, // extract floating >= condition from arg0
   498  
   499  		{name: "MOVBQSX", argLength: 1, reg: gp11, asm: "MOVBQSX"}, // sign extend arg0 from int8 to int64
   500  		{name: "MOVBQZX", argLength: 1, reg: gp11, asm: "MOVBLZX"}, // zero extend arg0 from int8 to int64
   501  		{name: "MOVWQSX", argLength: 1, reg: gp11, asm: "MOVWQSX"}, // sign extend arg0 from int16 to int64
   502  		{name: "MOVWQZX", argLength: 1, reg: gp11, asm: "MOVWLZX"}, // zero extend arg0 from int16 to int64
   503  		{name: "MOVLQSX", argLength: 1, reg: gp11, asm: "MOVLQSX"}, // sign extend arg0 from int32 to int64
   504  		{name: "MOVLQZX", argLength: 1, reg: gp11, asm: "MOVL"},    // zero extend arg0 from int32 to int64
   505  
   506  		{name: "MOVLconst", reg: gp01, asm: "MOVL", typ: "UInt32", aux: "Int32", rematerializeable: true}, // 32 low bits of auxint
   507  		{name: "MOVQconst", reg: gp01, asm: "MOVQ", typ: "UInt64", aux: "Int64", rematerializeable: true}, // auxint
   508  
   509  		{name: "CVTTSD2SL", argLength: 1, reg: fpgp, asm: "CVTTSD2SL"}, // convert float64 to int32
   510  		{name: "CVTTSD2SQ", argLength: 1, reg: fpgp, asm: "CVTTSD2SQ"}, // convert float64 to int64
   511  		{name: "CVTTSS2SL", argLength: 1, reg: fpgp, asm: "CVTTSS2SL"}, // convert float32 to int32
   512  		{name: "CVTTSS2SQ", argLength: 1, reg: fpgp, asm: "CVTTSS2SQ"}, // convert float32 to int64
   513  		{name: "CVTSL2SS", argLength: 1, reg: gpfp, asm: "CVTSL2SS"},   // convert int32 to float32
   514  		{name: "CVTSL2SD", argLength: 1, reg: gpfp, asm: "CVTSL2SD"},   // convert int32 to float64
   515  		{name: "CVTSQ2SS", argLength: 1, reg: gpfp, asm: "CVTSQ2SS"},   // convert int64 to float32
   516  		{name: "CVTSQ2SD", argLength: 1, reg: gpfp, asm: "CVTSQ2SD"},   // convert int64 to float64
   517  		{name: "CVTSD2SS", argLength: 1, reg: fp11, asm: "CVTSD2SS"},   // convert float64 to float32
   518  		{name: "CVTSS2SD", argLength: 1, reg: fp11, asm: "CVTSS2SD"},   // convert float32 to float64
   519  
   520  		// Move values between int and float registers, with no conversion.
   521  		// TODO: should we have generic versions of these?
   522  		{name: "MOVQi2f", argLength: 1, reg: gpfp, typ: "Float64"}, // move 64 bits from int to float reg
   523  		{name: "MOVQf2i", argLength: 1, reg: fpgp, typ: "UInt64"},  // move 64 bits from float to int reg
   524  		{name: "MOVLi2f", argLength: 1, reg: gpfp, typ: "Float32"}, // move 32 bits from int to float reg
   525  		{name: "MOVLf2i", argLength: 1, reg: fpgp, typ: "UInt32"},  // move 32 bits from float to int reg, zero extend
   526  
   527  		{name: "PXOR", argLength: 2, reg: fp21, asm: "PXOR", commutative: true, resultInArg0: true}, // exclusive or, applied to X regs for float negation.
   528  
   529  		{name: "LEAQ", argLength: 1, reg: gp11sb, asm: "LEAQ", aux: "SymOff", rematerializeable: true, symEffect: "Addr"}, // arg0 + auxint + offset encoded in aux
   530  		{name: "LEAL", argLength: 1, reg: gp11sb, asm: "LEAL", aux: "SymOff", rematerializeable: true, symEffect: "Addr"}, // arg0 + auxint + offset encoded in aux
   531  		{name: "LEAW", argLength: 1, reg: gp11sb, asm: "LEAW", aux: "SymOff", rematerializeable: true, symEffect: "Addr"}, // arg0 + auxint + offset encoded in aux
   532  		{name: "LEAQ1", argLength: 2, reg: gp21sb, asm: "LEAQ", commutative: true, aux: "SymOff", symEffect: "Addr"},      // arg0 + arg1 + auxint + aux
   533  		{name: "LEAL1", argLength: 2, reg: gp21sb, asm: "LEAL", commutative: true, aux: "SymOff", symEffect: "Addr"},      // arg0 + arg1 + auxint + aux
   534  		{name: "LEAW1", argLength: 2, reg: gp21sb, asm: "LEAW", commutative: true, aux: "SymOff", symEffect: "Addr"},      // arg0 + arg1 + auxint + aux
   535  		{name: "LEAQ2", argLength: 2, reg: gp21sb, asm: "LEAQ", aux: "SymOff", symEffect: "Addr"},                         // arg0 + 2*arg1 + auxint + aux
   536  		{name: "LEAL2", argLength: 2, reg: gp21sb, asm: "LEAL", aux: "SymOff", symEffect: "Addr"},                         // arg0 + 2*arg1 + auxint + aux
   537  		{name: "LEAW2", argLength: 2, reg: gp21sb, asm: "LEAW", aux: "SymOff", symEffect: "Addr"},                         // arg0 + 2*arg1 + auxint + aux
   538  		{name: "LEAQ4", argLength: 2, reg: gp21sb, asm: "LEAQ", aux: "SymOff", symEffect: "Addr"},                         // arg0 + 4*arg1 + auxint + aux
   539  		{name: "LEAL4", argLength: 2, reg: gp21sb, asm: "LEAL", aux: "SymOff", symEffect: "Addr"},                         // arg0 + 4*arg1 + auxint + aux
   540  		{name: "LEAW4", argLength: 2, reg: gp21sb, asm: "LEAW", aux: "SymOff", symEffect: "Addr"},                         // arg0 + 4*arg1 + auxint + aux
   541  		{name: "LEAQ8", argLength: 2, reg: gp21sb, asm: "LEAQ", aux: "SymOff", symEffect: "Addr"},                         // arg0 + 8*arg1 + auxint + aux
   542  		{name: "LEAL8", argLength: 2, reg: gp21sb, asm: "LEAL", aux: "SymOff", symEffect: "Addr"},                         // arg0 + 8*arg1 + auxint + aux
   543  		{name: "LEAW8", argLength: 2, reg: gp21sb, asm: "LEAW", aux: "SymOff", symEffect: "Addr"},                         // arg0 + 8*arg1 + auxint + aux
   544  		// Note: LEAx{1,2,4,8} must not have OpSB as either argument.
   545  
   546  		// auxint+aux == add auxint and the offset of the symbol in aux (if any) to the effective address
   547  		{name: "MOVBload", argLength: 2, reg: gpload, asm: "MOVBLZX", aux: "SymOff", typ: "UInt8", faultOnNilArg0: true, symEffect: "Read"},  // load byte from arg0+auxint+aux. arg1=mem.  Zero extend.
   548  		{name: "MOVBQSXload", argLength: 2, reg: gpload, asm: "MOVBQSX", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"},             // ditto, sign extend to int64
   549  		{name: "MOVWload", argLength: 2, reg: gpload, asm: "MOVWLZX", aux: "SymOff", typ: "UInt16", faultOnNilArg0: true, symEffect: "Read"}, // load 2 bytes from arg0+auxint+aux. arg1=mem.  Zero extend.
   550  		{name: "MOVWQSXload", argLength: 2, reg: gpload, asm: "MOVWQSX", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"},             // ditto, sign extend to int64
   551  		{name: "MOVLload", argLength: 2, reg: gpload, asm: "MOVL", aux: "SymOff", typ: "UInt32", faultOnNilArg0: true, symEffect: "Read"},    // load 4 bytes from arg0+auxint+aux. arg1=mem.  Zero extend.
   552  		{name: "MOVLQSXload", argLength: 2, reg: gpload, asm: "MOVLQSX", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"},             // ditto, sign extend to int64
   553  		{name: "MOVQload", argLength: 2, reg: gpload, asm: "MOVQ", aux: "SymOff", typ: "UInt64", faultOnNilArg0: true, symEffect: "Read"},    // load 8 bytes from arg0+auxint+aux. arg1=mem
   554  		{name: "MOVBstore", argLength: 3, reg: gpstore, asm: "MOVB", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"},    // store byte in arg1 to arg0+auxint+aux. arg2=mem
   555  		{name: "MOVWstore", argLength: 3, reg: gpstore, asm: "MOVW", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"},    // store 2 bytes in arg1 to arg0+auxint+aux. arg2=mem
   556  		{name: "MOVLstore", argLength: 3, reg: gpstore, asm: "MOVL", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"},    // store 4 bytes in arg1 to arg0+auxint+aux. arg2=mem
   557  		{name: "MOVQstore", argLength: 3, reg: gpstore, asm: "MOVQ", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"},    // store 8 bytes in arg1 to arg0+auxint+aux. arg2=mem
   558  		{name: "MOVOload", argLength: 2, reg: fpload, asm: "MOVUPS", aux: "SymOff", typ: "Int128", faultOnNilArg0: true, symEffect: "Read"},  // load 16 bytes from arg0+auxint+aux. arg1=mem
   559  		{name: "MOVOstore", argLength: 3, reg: fpstore, asm: "MOVUPS", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"},  // store 16 bytes in arg1 to arg0+auxint+aux. arg2=mem
   560  
   561  		// indexed loads/stores
   562  		{name: "MOVBloadidx1", argLength: 3, reg: gploadidx, commutative: true, asm: "MOVBLZX", aux: "SymOff", typ: "UInt8", symEffect: "Read"},  // load a byte from arg0+arg1+auxint+aux. arg2=mem
   563  		{name: "MOVWloadidx1", argLength: 3, reg: gploadidx, commutative: true, asm: "MOVWLZX", aux: "SymOff", typ: "UInt16", symEffect: "Read"}, // load 2 bytes from arg0+arg1+auxint+aux. arg2=mem
   564  		{name: "MOVWloadidx2", argLength: 3, reg: gploadidx, asm: "MOVWLZX", aux: "SymOff", typ: "UInt16", symEffect: "Read"},                    // load 2 bytes from arg0+2*arg1+auxint+aux. arg2=mem
   565  		{name: "MOVLloadidx1", argLength: 3, reg: gploadidx, commutative: true, asm: "MOVL", aux: "SymOff", typ: "UInt32", symEffect: "Read"},    // load 4 bytes from arg0+arg1+auxint+aux. arg2=mem
   566  		{name: "MOVLloadidx4", argLength: 3, reg: gploadidx, asm: "MOVL", aux: "SymOff", typ: "UInt32", symEffect: "Read"},                       // load 4 bytes from arg0+4*arg1+auxint+aux. arg2=mem
   567  		{name: "MOVLloadidx8", argLength: 3, reg: gploadidx, asm: "MOVL", aux: "SymOff", typ: "UInt32", symEffect: "Read"},                       // load 4 bytes from arg0+8*arg1+auxint+aux. arg2=mem
   568  		{name: "MOVQloadidx1", argLength: 3, reg: gploadidx, commutative: true, asm: "MOVQ", aux: "SymOff", typ: "UInt64", symEffect: "Read"},    // load 8 bytes from arg0+arg1+auxint+aux. arg2=mem
   569  		{name: "MOVQloadidx8", argLength: 3, reg: gploadidx, asm: "MOVQ", aux: "SymOff", typ: "UInt64", symEffect: "Read"},                       // load 8 bytes from arg0+8*arg1+auxint+aux. arg2=mem
   570  		// TODO: sign-extending indexed loads
   571  		// TODO: mark the MOVXstoreidx1 ops as commutative.  Generates too many rewrite rules at the moment.
   572  		{name: "MOVBstoreidx1", argLength: 4, reg: gpstoreidx, asm: "MOVB", aux: "SymOff", symEffect: "Write"}, // store byte in arg2 to arg0+arg1+auxint+aux. arg3=mem
   573  		{name: "MOVWstoreidx1", argLength: 4, reg: gpstoreidx, asm: "MOVW", aux: "SymOff", symEffect: "Write"}, // store 2 bytes in arg2 to arg0+arg1+auxint+aux. arg3=mem
   574  		{name: "MOVWstoreidx2", argLength: 4, reg: gpstoreidx, asm: "MOVW", aux: "SymOff", symEffect: "Write"}, // store 2 bytes in arg2 to arg0+2*arg1+auxint+aux. arg3=mem
   575  		{name: "MOVLstoreidx1", argLength: 4, reg: gpstoreidx, asm: "MOVL", aux: "SymOff", symEffect: "Write"}, // store 4 bytes in arg2 to arg0+arg1+auxint+aux. arg3=mem
   576  		{name: "MOVLstoreidx4", argLength: 4, reg: gpstoreidx, asm: "MOVL", aux: "SymOff", symEffect: "Write"}, // store 4 bytes in arg2 to arg0+4*arg1+auxint+aux. arg3=mem
   577  		{name: "MOVLstoreidx8", argLength: 4, reg: gpstoreidx, asm: "MOVL", aux: "SymOff", symEffect: "Write"}, // store 4 bytes in arg2 to arg0+8*arg1+auxint+aux. arg3=mem
   578  		{name: "MOVQstoreidx1", argLength: 4, reg: gpstoreidx, asm: "MOVQ", aux: "SymOff", symEffect: "Write"}, // store 8 bytes in arg2 to arg0+arg1+auxint+aux. arg3=mem
   579  		{name: "MOVQstoreidx8", argLength: 4, reg: gpstoreidx, asm: "MOVQ", aux: "SymOff", symEffect: "Write"}, // store 8 bytes in arg2 to arg0+8*arg1+auxint+aux. arg3=mem
   580  		// TODO: add size-mismatched indexed loads, like MOVBstoreidx4.
   581  
   582  		// For storeconst ops, the AuxInt field encodes both
   583  		// the value to store and an address offset of the store.
   584  		// Cast AuxInt to a ValAndOff to extract Val and Off fields.
   585  		{name: "MOVBstoreconst", argLength: 2, reg: gpstoreconst, asm: "MOVB", aux: "SymValAndOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // store low byte of ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux.  arg1=mem
   586  		{name: "MOVWstoreconst", argLength: 2, reg: gpstoreconst, asm: "MOVW", aux: "SymValAndOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // store low 2 bytes of ...
   587  		{name: "MOVLstoreconst", argLength: 2, reg: gpstoreconst, asm: "MOVL", aux: "SymValAndOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // store low 4 bytes of ...
   588  		{name: "MOVQstoreconst", argLength: 2, reg: gpstoreconst, asm: "MOVQ", aux: "SymValAndOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // store 8 bytes of ...
   589  
   590  		{name: "MOVBstoreconstidx1", argLength: 3, reg: gpstoreconstidx, asm: "MOVB", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store low byte of ValAndOff(AuxInt).Val() to arg0+1*arg1+ValAndOff(AuxInt).Off()+aux.  arg2=mem
   591  		{name: "MOVWstoreconstidx1", argLength: 3, reg: gpstoreconstidx, asm: "MOVW", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store low 2 bytes of ... arg1 ...
   592  		{name: "MOVWstoreconstidx2", argLength: 3, reg: gpstoreconstidx, asm: "MOVW", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store low 2 bytes of ... 2*arg1 ...
   593  		{name: "MOVLstoreconstidx1", argLength: 3, reg: gpstoreconstidx, asm: "MOVL", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store low 4 bytes of ... arg1 ...
   594  		{name: "MOVLstoreconstidx4", argLength: 3, reg: gpstoreconstidx, asm: "MOVL", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store low 4 bytes of ... 4*arg1 ...
   595  		{name: "MOVQstoreconstidx1", argLength: 3, reg: gpstoreconstidx, asm: "MOVQ", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store 8 bytes of ... arg1 ...
   596  		{name: "MOVQstoreconstidx8", argLength: 3, reg: gpstoreconstidx, asm: "MOVQ", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store 8 bytes of ... 8*arg1 ...
   597  
   598  		// arg0 = pointer to start of memory to zero
   599  		// arg1 = value to store (will always be zero)
   600  		// arg2 = mem
   601  		// auxint = # of bytes to zero
   602  		// returns mem
   603  		{
   604  			name:      "DUFFZERO",
   605  			aux:       "Int64",
   606  			argLength: 3,
   607  			reg: regInfo{
   608  				inputs:   []regMask{buildReg("DI"), buildReg("X0")},
   609  				clobbers: buildReg("DI"),
   610  			},
   611  			faultOnNilArg0: true,
   612  		},
   613  		{name: "MOVOconst", reg: regInfo{nil, 0, []regMask{fp}}, typ: "Int128", aux: "Int128", rematerializeable: true},
   614  
   615  		// arg0 = address of memory to zero
   616  		// arg1 = # of 8-byte words to zero
   617  		// arg2 = value to store (will always be zero)
   618  		// arg3 = mem
   619  		// returns mem
   620  		{
   621  			name:      "REPSTOSQ",
   622  			argLength: 4,
   623  			reg: regInfo{
   624  				inputs:   []regMask{buildReg("DI"), buildReg("CX"), buildReg("AX")},
   625  				clobbers: buildReg("DI CX"),
   626  			},
   627  			faultOnNilArg0: true,
   628  		},
   629  
   630  		{name: "CALLstatic", argLength: 1, reg: regInfo{clobbers: callerSave}, aux: "SymOff", clobberFlags: true, call: true, symEffect: "None"},                          // call static function aux.(*obj.LSym).  arg0=mem, auxint=argsize, returns mem
   631  		{name: "CALLclosure", argLength: 3, reg: regInfo{inputs: []regMask{gpsp, buildReg("DX"), 0}, clobbers: callerSave}, aux: "Int64", clobberFlags: true, call: true}, // call function via closure.  arg0=codeptr, arg1=closure, arg2=mem, auxint=argsize, returns mem
   632  		{name: "CALLinter", argLength: 2, reg: regInfo{inputs: []regMask{gp}, clobbers: callerSave}, aux: "Int64", clobberFlags: true, call: true},                        // call fn by pointer.  arg0=codeptr, arg1=mem, auxint=argsize, returns mem
   633  
   634  		// arg0 = destination pointer
   635  		// arg1 = source pointer
   636  		// arg2 = mem
   637  		// auxint = offset from duffcopy symbol to call
   638  		// returns memory
   639  		{
   640  			name:      "DUFFCOPY",
   641  			aux:       "Int64",
   642  			argLength: 3,
   643  			reg: regInfo{
   644  				inputs:   []regMask{buildReg("DI"), buildReg("SI")},
   645  				clobbers: buildReg("DI SI X0"), // uses X0 as a temporary
   646  			},
   647  			clobberFlags:   true,
   648  			faultOnNilArg0: true,
   649  			faultOnNilArg1: true,
   650  		},
   651  
   652  		// arg0 = destination pointer
   653  		// arg1 = source pointer
   654  		// arg2 = # of 8-byte words to copy
   655  		// arg3 = mem
   656  		// returns memory
   657  		{
   658  			name:      "REPMOVSQ",
   659  			argLength: 4,
   660  			reg: regInfo{
   661  				inputs:   []regMask{buildReg("DI"), buildReg("SI"), buildReg("CX")},
   662  				clobbers: buildReg("DI SI CX"),
   663  			},
   664  			faultOnNilArg0: true,
   665  			faultOnNilArg1: true,
   666  		},
   667  
   668  		// (InvertFlags (CMPQ a b)) == (CMPQ b a)
   669  		// So if we want (SETL (CMPQ a b)) but we can't do that because a is a constant,
   670  		// then we do (SETL (InvertFlags (CMPQ b a))) instead.
   671  		// Rewrites will convert this to (SETG (CMPQ b a)).
   672  		// InvertFlags is a pseudo-op which can't appear in assembly output.
   673  		{name: "InvertFlags", argLength: 1}, // reverse direction of arg0
   674  
   675  		// Pseudo-ops
   676  		{name: "LoweredGetG", argLength: 1, reg: gp01}, // arg0=mem
   677  		// Scheduler ensures LoweredGetClosurePtr occurs only in entry block,
   678  		// and sorts it to the very beginning of the block to prevent other
   679  		// use of DX (the closure pointer)
   680  		{name: "LoweredGetClosurePtr", reg: regInfo{outputs: []regMask{buildReg("DX")}}, zeroWidth: true},
   681  		// LoweredGetCallerPC evaluates to the PC to which its "caller" will return.
   682  		// I.e., if f calls g "calls" getcallerpc,
   683  		// the result should be the PC within f that g will return to.
   684  		// See runtime/stubs.go for a more detailed discussion.
   685  		{name: "LoweredGetCallerPC", reg: gp01, rematerializeable: true},
   686  		// LoweredGetCallerSP returns the SP of the caller of the current function.
   687  		{name: "LoweredGetCallerSP", reg: gp01, rematerializeable: true},
   688  		//arg0=ptr,arg1=mem, returns void.  Faults if ptr is nil.
   689  		{name: "LoweredNilCheck", argLength: 2, reg: regInfo{inputs: []regMask{gpsp}}, clobberFlags: true, nilCheck: true, faultOnNilArg0: true},
   690  		// LoweredWB invokes runtime.gcWriteBarrier. arg0=destptr, arg1=srcptr, arg2=mem, aux=runtime.gcWriteBarrier
   691  		// It saves all GP registers if necessary, but may clobber others.
   692  		{name: "LoweredWB", argLength: 3, reg: regInfo{inputs: []regMask{buildReg("DI"), ax}, clobbers: callerSave &^ gp}, clobberFlags: true, aux: "Sym", symEffect: "None"},
   693  
   694  		// Constant flag values. For any comparison, there are 5 possible
   695  		// outcomes: the three from the signed total order (<,==,>) and the
   696  		// three from the unsigned total order. The == cases overlap.
   697  		// Note: there's a sixth "unordered" outcome for floating-point
   698  		// comparisons, but we don't use such a beast yet.
   699  		// These ops are for temporary use by rewrite rules. They
   700  		// cannot appear in the generated assembly.
   701  		{name: "FlagEQ"},     // equal
   702  		{name: "FlagLT_ULT"}, // signed < and unsigned <
   703  		{name: "FlagLT_UGT"}, // signed < and unsigned >
   704  		{name: "FlagGT_UGT"}, // signed > and unsigned >
   705  		{name: "FlagGT_ULT"}, // signed > and unsigned <
   706  
   707  		// Atomic loads.  These are just normal loads but return <value,memory> tuples
   708  		// so they can be properly ordered with other loads.
   709  		// load from arg0+auxint+aux.  arg1=mem.
   710  		{name: "MOVLatomicload", argLength: 2, reg: gpload, asm: "MOVL", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"},
   711  		{name: "MOVQatomicload", argLength: 2, reg: gpload, asm: "MOVQ", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"},
   712  
   713  		// Atomic stores and exchanges.  Stores use XCHG to get the right memory ordering semantics.
   714  		// store arg0 to arg1+auxint+aux, arg2=mem.
   715  		// These ops return a tuple of <old contents of *(arg1+auxint+aux), memory>.
   716  		// Note: arg0 and arg1 are backwards compared to MOVLstore (to facilitate resultInArg0)!
   717  		{name: "XCHGL", argLength: 3, reg: gpstorexchg, asm: "XCHGL", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true, hasSideEffects: true, symEffect: "RdWr"},
   718  		{name: "XCHGQ", argLength: 3, reg: gpstorexchg, asm: "XCHGQ", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true, hasSideEffects: true, symEffect: "RdWr"},
   719  
   720  		// Atomic adds.
   721  		// *(arg1+auxint+aux) += arg0.  arg2=mem.
   722  		// Returns a tuple of <old contents of *(arg1+auxint+aux), memory>.
   723  		// Note: arg0 and arg1 are backwards compared to MOVLstore (to facilitate resultInArg0)!
   724  		{name: "XADDLlock", argLength: 3, reg: gpstorexchg, asm: "XADDL", typ: "(UInt32,Mem)", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, hasSideEffects: true, symEffect: "RdWr"},
   725  		{name: "XADDQlock", argLength: 3, reg: gpstorexchg, asm: "XADDQ", typ: "(UInt64,Mem)", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, hasSideEffects: true, symEffect: "RdWr"},
   726  		{name: "AddTupleFirst32", argLength: 2}, // arg1=tuple <x,y>.  Returns <x+arg0,y>.
   727  		{name: "AddTupleFirst64", argLength: 2}, // arg1=tuple <x,y>.  Returns <x+arg0,y>.
   728  
   729  		// Compare and swap.
   730  		// arg0 = pointer, arg1 = old value, arg2 = new value, arg3 = memory.
   731  		// if *(arg0+auxint+aux) == arg1 {
   732  		//   *(arg0+auxint+aux) = arg2
   733  		//   return (true, memory)
   734  		// } else {
   735  		//   return (false, memory)
   736  		// }
   737  		// Note that these instructions also return the old value in AX, but we ignore it.
   738  		// TODO: have these return flags instead of bool.  The current system generates:
   739  		//    CMPXCHGQ ...
   740  		//    SETEQ AX
   741  		//    CMPB  AX, $0
   742  		//    JNE ...
   743  		// instead of just
   744  		//    CMPXCHGQ ...
   745  		//    JEQ ...
   746  		// but we can't do that because memory-using ops can't generate flags yet
   747  		// (flagalloc wants to move flag-generating instructions around).
   748  		{name: "CMPXCHGLlock", argLength: 4, reg: cmpxchg, asm: "CMPXCHGL", aux: "SymOff", clobberFlags: true, faultOnNilArg0: true, hasSideEffects: true, symEffect: "RdWr"},
   749  		{name: "CMPXCHGQlock", argLength: 4, reg: cmpxchg, asm: "CMPXCHGQ", aux: "SymOff", clobberFlags: true, faultOnNilArg0: true, hasSideEffects: true, symEffect: "RdWr"},
   750  
   751  		// Atomic memory updates.
   752  		{name: "ANDBlock", argLength: 3, reg: gpstore, asm: "ANDB", aux: "SymOff", clobberFlags: true, faultOnNilArg0: true, hasSideEffects: true, symEffect: "RdWr"}, // *(arg0+auxint+aux) &= arg1
   753  		{name: "ORBlock", argLength: 3, reg: gpstore, asm: "ORB", aux: "SymOff", clobberFlags: true, faultOnNilArg0: true, hasSideEffects: true, symEffect: "RdWr"},   // *(arg0+auxint+aux) |= arg1
   754  	}
   755  
   756  	var AMD64blocks = []blockData{
   757  		{name: "EQ"},
   758  		{name: "NE"},
   759  		{name: "LT"},
   760  		{name: "LE"},
   761  		{name: "GT"},
   762  		{name: "GE"},
   763  		{name: "OS"},
   764  		{name: "OC"},
   765  		{name: "ULT"},
   766  		{name: "ULE"},
   767  		{name: "UGT"},
   768  		{name: "UGE"},
   769  		{name: "EQF"},
   770  		{name: "NEF"},
   771  		{name: "ORD"}, // FP, ordered comparison (parity zero)
   772  		{name: "NAN"}, // FP, unordered comparison (parity one)
   773  	}
   774  
   775  	archs = append(archs, arch{
   776  		name:            "AMD64",
   777  		pkg:             "cmd/internal/obj/x86",
   778  		genfile:         "../../amd64/ssa.go",
   779  		ops:             AMD64ops,
   780  		blocks:          AMD64blocks,
   781  		regnames:        regNamesAMD64,
   782  		gpregmask:       gp,
   783  		fpregmask:       fp,
   784  		framepointerreg: int8(num["BP"]),
   785  		linkreg:         -1, // not used
   786  	})
   787  }