386Ops.go

Documentation: cmd/compile/internal/ssa/_gen

     1  // Copyright 2016 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  package main
     6  
     7  import "strings"
     8  
     9  // Notes:
    10  //  - Integer types live in the low portion of registers. Upper portions are junk.
    11  //  - Boolean types use the low-order byte of a register. 0=false, 1=true.
    12  //    Upper bytes are junk.
    13  //  - Floating-point types live in the low natural slot of an sse2 register.
    14  //    Unused portions are junk.
    15  //  - We do not use AH,BH,CH,DH registers.
    16  //  - When doing sub-register operations, we try to write the whole
    17  //    destination register to avoid a partial-register write.
    18  //  - Unused portions of AuxInt (or the Val portion of ValAndOff) are
    19  //    filled by sign-extending the used portion.  Users of AuxInt which interpret
    20  //    AuxInt as unsigned (e.g. shifts) must be careful.
    21  
    22  // Suffixes encode the bit width of various instructions.
    23  // L (long word) = 32 bit
    24  // W (word)      = 16 bit
    25  // B (byte)      = 8 bit
    26  
    27  // copied from ../../x86/reg.go
    28  var regNames386 = []string{
    29  	"AX",
    30  	"CX",
    31  	"DX",
    32  	"BX",
    33  	"SP",
    34  	"BP",
    35  	"SI",
    36  	"DI",
    37  	"X0",
    38  	"X1",
    39  	"X2",
    40  	"X3",
    41  	"X4",
    42  	"X5",
    43  	"X6",
    44  	"X7",
    45  
    46  	// If you add registers, update asyncPreempt in runtime
    47  
    48  	// pseudo-registers
    49  	"SB",
    50  }
    51  
    52  func init() {
    53  	// Make map from reg names to reg integers.
    54  	if len(regNames386) > 64 {
    55  		panic("too many registers")
    56  	}
    57  	num := map[string]int{}
    58  	for i, name := range regNames386 {
    59  		num[name] = i
    60  	}
    61  	buildReg := func(s string) regMask {
    62  		m := regMask(0)
    63  		for _, r := range strings.Split(s, " ") {
    64  			if n, ok := num[r]; ok {
    65  				m |= regMask(1) << uint(n)
    66  				continue
    67  			}
    68  			panic("register " + r + " not found")
    69  		}
    70  		return m
    71  	}
    72  
    73  	// Common individual register masks
    74  	var (
    75  		ax         = buildReg("AX")
    76  		cx         = buildReg("CX")
    77  		dx         = buildReg("DX")
    78  		bx         = buildReg("BX")
    79  		si         = buildReg("SI")
    80  		gp         = buildReg("AX CX DX BX BP SI DI")
    81  		fp         = buildReg("X0 X1 X2 X3 X4 X5 X6 X7")
    82  		gpsp       = gp | buildReg("SP")
    83  		gpspsb     = gpsp | buildReg("SB")
    84  		callerSave = gp | fp
    85  	)
    86  	// Common slices of register masks
    87  	var (
    88  		gponly = []regMask{gp}
    89  		fponly = []regMask{fp}
    90  	)
    91  
    92  	// Common regInfo
    93  	var (
    94  		gp01      = regInfo{inputs: nil, outputs: gponly}
    95  		gp11      = regInfo{inputs: []regMask{gp}, outputs: gponly}
    96  		gp11sp    = regInfo{inputs: []regMask{gpsp}, outputs: gponly}
    97  		gp11sb    = regInfo{inputs: []regMask{gpspsb}, outputs: gponly}
    98  		gp21      = regInfo{inputs: []regMask{gp, gp}, outputs: gponly}
    99  		gp11carry = regInfo{inputs: []regMask{gp}, outputs: []regMask{gp, 0}}
   100  		gp21carry = regInfo{inputs: []regMask{gp, gp}, outputs: []regMask{gp, 0}}
   101  		gp1carry1 = regInfo{inputs: []regMask{gp}, outputs: gponly}
   102  		gp2carry1 = regInfo{inputs: []regMask{gp, gp}, outputs: gponly}
   103  		gp21sp    = regInfo{inputs: []regMask{gpsp, gp}, outputs: gponly}
   104  		gp21sb    = regInfo{inputs: []regMask{gpspsb, gpsp}, outputs: gponly}
   105  		gp21shift = regInfo{inputs: []regMask{gp, cx}, outputs: []regMask{gp}}
   106  		gp11div   = regInfo{inputs: []regMask{ax, gpsp &^ dx}, outputs: []regMask{ax}, clobbers: dx}
   107  		gp21hmul  = regInfo{inputs: []regMask{ax, gpsp}, outputs: []regMask{dx}, clobbers: ax}
   108  		gp11mod   = regInfo{inputs: []regMask{ax, gpsp &^ dx}, outputs: []regMask{dx}, clobbers: ax}
   109  		gp21mul   = regInfo{inputs: []regMask{ax, gpsp}, outputs: []regMask{dx, ax}}
   110  
   111  		gp2flags     = regInfo{inputs: []regMask{gpsp, gpsp}}
   112  		gp1flags     = regInfo{inputs: []regMask{gpsp}}
   113  		gp0flagsLoad = regInfo{inputs: []regMask{gpspsb, 0}}
   114  		gp1flagsLoad = regInfo{inputs: []regMask{gpspsb, gpsp, 0}}
   115  		flagsgp      = regInfo{inputs: nil, outputs: gponly}
   116  
   117  		readflags = regInfo{inputs: nil, outputs: gponly}
   118  		flagsgpax = regInfo{inputs: nil, clobbers: ax, outputs: []regMask{gp &^ ax}}
   119  
   120  		gpload      = regInfo{inputs: []regMask{gpspsb, 0}, outputs: gponly}
   121  		gp21load    = regInfo{inputs: []regMask{gp, gpspsb, 0}, outputs: gponly}
   122  		gploadidx   = regInfo{inputs: []regMask{gpspsb, gpsp, 0}, outputs: gponly}
   123  		gp21loadidx = regInfo{inputs: []regMask{gp, gpspsb, gpsp, 0}, outputs: gponly}
   124  
   125  		gpstore         = regInfo{inputs: []regMask{gpspsb, gpsp, 0}}
   126  		gpstoreconst    = regInfo{inputs: []regMask{gpspsb, 0}}
   127  		gpstoreidx      = regInfo{inputs: []regMask{gpspsb, gpsp, gpsp, 0}}
   128  		gpstoreconstidx = regInfo{inputs: []regMask{gpspsb, gpsp, 0}}
   129  
   130  		fp01     = regInfo{inputs: nil, outputs: fponly}
   131  		fp21     = regInfo{inputs: []regMask{fp, fp}, outputs: fponly}
   132  		fp21load = regInfo{inputs: []regMask{fp, gpspsb, 0}, outputs: fponly}
   133  		fpgp     = regInfo{inputs: fponly, outputs: gponly}
   134  		gpfp     = regInfo{inputs: gponly, outputs: fponly}
   135  		fp11     = regInfo{inputs: fponly, outputs: fponly}
   136  		fp2flags = regInfo{inputs: []regMask{fp, fp}}
   137  
   138  		fpload    = regInfo{inputs: []regMask{gpspsb, 0}, outputs: fponly}
   139  		fploadidx = regInfo{inputs: []regMask{gpspsb, gpsp, 0}, outputs: fponly}
   140  
   141  		fpstore    = regInfo{inputs: []regMask{gpspsb, fp, 0}}
   142  		fpstoreidx = regInfo{inputs: []regMask{gpspsb, gpsp, fp, 0}}
   143  	)
   144  
   145  	var _386ops = []opData{
   146  		// fp ops
   147  		{name: "ADDSS", argLength: 2, reg: fp21, asm: "ADDSS", commutative: true, resultInArg0: true}, // fp32 add
   148  		{name: "ADDSD", argLength: 2, reg: fp21, asm: "ADDSD", commutative: true, resultInArg0: true}, // fp64 add
   149  		{name: "SUBSS", argLength: 2, reg: fp21, asm: "SUBSS", resultInArg0: true},                    // fp32 sub
   150  		{name: "SUBSD", argLength: 2, reg: fp21, asm: "SUBSD", resultInArg0: true},                    // fp64 sub
   151  		{name: "MULSS", argLength: 2, reg: fp21, asm: "MULSS", commutative: true, resultInArg0: true}, // fp32 mul
   152  		{name: "MULSD", argLength: 2, reg: fp21, asm: "MULSD", commutative: true, resultInArg0: true}, // fp64 mul
   153  		{name: "DIVSS", argLength: 2, reg: fp21, asm: "DIVSS", resultInArg0: true},                    // fp32 div
   154  		{name: "DIVSD", argLength: 2, reg: fp21, asm: "DIVSD", resultInArg0: true},                    // fp64 div
   155  
   156  		{name: "MOVSSload", argLength: 2, reg: fpload, asm: "MOVSS", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"}, // fp32 load
   157  		{name: "MOVSDload", argLength: 2, reg: fpload, asm: "MOVSD", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"}, // fp64 load
   158  		{name: "MOVSSconst", reg: fp01, asm: "MOVSS", aux: "Float32", rematerializeable: true},                               // fp32 constant
   159  		{name: "MOVSDconst", reg: fp01, asm: "MOVSD", aux: "Float64", rematerializeable: true},                               // fp64 constant
   160  		{name: "MOVSSloadidx1", argLength: 3, reg: fploadidx, asm: "MOVSS", aux: "SymOff", symEffect: "Read"},                // fp32 load indexed by i
   161  		{name: "MOVSSloadidx4", argLength: 3, reg: fploadidx, asm: "MOVSS", aux: "SymOff", symEffect: "Read"},                // fp32 load indexed by 4*i
   162  		{name: "MOVSDloadidx1", argLength: 3, reg: fploadidx, asm: "MOVSD", aux: "SymOff", symEffect: "Read"},                // fp64 load indexed by i
   163  		{name: "MOVSDloadidx8", argLength: 3, reg: fploadidx, asm: "MOVSD", aux: "SymOff", symEffect: "Read"},                // fp64 load indexed by 8*i
   164  
   165  		{name: "MOVSSstore", argLength: 3, reg: fpstore, asm: "MOVSS", aux: "SymOff", faultOnNilArg0: true, symEffect: "Write"}, // fp32 store
   166  		{name: "MOVSDstore", argLength: 3, reg: fpstore, asm: "MOVSD", aux: "SymOff", faultOnNilArg0: true, symEffect: "Write"}, // fp64 store
   167  		{name: "MOVSSstoreidx1", argLength: 4, reg: fpstoreidx, asm: "MOVSS", aux: "SymOff", symEffect: "Write"},                // fp32 indexed by i store
   168  		{name: "MOVSSstoreidx4", argLength: 4, reg: fpstoreidx, asm: "MOVSS", aux: "SymOff", symEffect: "Write"},                // fp32 indexed by 4i store
   169  		{name: "MOVSDstoreidx1", argLength: 4, reg: fpstoreidx, asm: "MOVSD", aux: "SymOff", symEffect: "Write"},                // fp64 indexed by i store
   170  		{name: "MOVSDstoreidx8", argLength: 4, reg: fpstoreidx, asm: "MOVSD", aux: "SymOff", symEffect: "Write"},                // fp64 indexed by 8i store
   171  
   172  		{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
   173  		{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
   174  		{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
   175  		{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
   176  		{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
   177  		{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
   178  		{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
   179  		{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
   180  
   181  		// binary ops
   182  		{name: "ADDL", argLength: 2, reg: gp21sp, asm: "ADDL", commutative: true, clobberFlags: true},                // arg0 + arg1
   183  		{name: "ADDLconst", argLength: 1, reg: gp11sp, asm: "ADDL", aux: "Int32", typ: "UInt32", clobberFlags: true}, // arg0 + auxint
   184  
   185  		{name: "ADDLcarry", argLength: 2, reg: gp21carry, asm: "ADDL", commutative: true, resultInArg0: true},                // arg0 + arg1, generates <carry,result> pair
   186  		{name: "ADDLconstcarry", argLength: 1, reg: gp11carry, asm: "ADDL", aux: "Int32", resultInArg0: true},                // arg0 + auxint, generates <carry,result> pair
   187  		{name: "ADCL", argLength: 3, reg: gp2carry1, asm: "ADCL", commutative: true, resultInArg0: true, clobberFlags: true}, // arg0+arg1+carry(arg2), where arg2 is flags
   188  		{name: "ADCLconst", argLength: 2, reg: gp1carry1, asm: "ADCL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // arg0+auxint+carry(arg1), where arg1 is flags
   189  
   190  		{name: "SUBL", argLength: 2, reg: gp21, asm: "SUBL", resultInArg0: true, clobberFlags: true},                    // arg0 - arg1
   191  		{name: "SUBLconst", argLength: 1, reg: gp11, asm: "SUBL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // arg0 - auxint
   192  
   193  		{name: "SUBLcarry", argLength: 2, reg: gp21carry, asm: "SUBL", resultInArg0: true},                                   // arg0-arg1, generates <borrow,result> pair
   194  		{name: "SUBLconstcarry", argLength: 1, reg: gp11carry, asm: "SUBL", aux: "Int32", resultInArg0: true},                // arg0-auxint, generates <borrow,result> pair
   195  		{name: "SBBL", argLength: 3, reg: gp2carry1, asm: "SBBL", resultInArg0: true, clobberFlags: true},                    // arg0-arg1-borrow(arg2), where arg2 is flags
   196  		{name: "SBBLconst", argLength: 2, reg: gp1carry1, asm: "SBBL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // arg0-auxint-borrow(arg1), where arg1 is flags
   197  
   198  		{name: "MULL", argLength: 2, reg: gp21, asm: "IMULL", commutative: true, resultInArg0: true, clobberFlags: true}, // arg0 * arg1
   199  		{name: "MULLconst", argLength: 1, reg: gp11, asm: "IMUL3L", aux: "Int32", clobberFlags: true},                    // arg0 * auxint
   200  
   201  		{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.
   202  
   203  		{name: "HMULL", argLength: 2, reg: gp21hmul, commutative: true, asm: "IMULL", clobberFlags: true}, // (arg0 * arg1) >> width
   204  		{name: "HMULLU", argLength: 2, reg: gp21hmul, commutative: true, asm: "MULL", clobberFlags: true}, // (arg0 * arg1) >> width
   205  
   206  		{name: "MULLQU", argLength: 2, reg: gp21mul, commutative: true, asm: "MULL", clobberFlags: true}, // arg0 * arg1, high 32 in result[0], low 32 in result[1]
   207  
   208  		{name: "AVGLU", argLength: 2, reg: gp21, commutative: true, resultInArg0: true, clobberFlags: true}, // (arg0 + arg1) / 2 as unsigned, all 32 result bits
   209  
   210  		// For DIVL, DIVW, MODL and MODW, AuxInt non-zero means that the divisor has been proved to be not -1.
   211  		{name: "DIVL", argLength: 2, reg: gp11div, asm: "IDIVL", aux: "Bool", clobberFlags: true}, // arg0 / arg1
   212  		{name: "DIVW", argLength: 2, reg: gp11div, asm: "IDIVW", aux: "Bool", clobberFlags: true}, // arg0 / arg1
   213  		{name: "DIVLU", argLength: 2, reg: gp11div, asm: "DIVL", clobberFlags: true},              // arg0 / arg1
   214  		{name: "DIVWU", argLength: 2, reg: gp11div, asm: "DIVW", clobberFlags: true},              // arg0 / arg1
   215  
   216  		{name: "MODL", argLength: 2, reg: gp11mod, asm: "IDIVL", aux: "Bool", clobberFlags: true}, // arg0 % arg1
   217  		{name: "MODW", argLength: 2, reg: gp11mod, asm: "IDIVW", aux: "Bool", clobberFlags: true}, // arg0 % arg1
   218  		{name: "MODLU", argLength: 2, reg: gp11mod, asm: "DIVL", clobberFlags: true},              // arg0 % arg1
   219  		{name: "MODWU", argLength: 2, reg: gp11mod, asm: "DIVW", clobberFlags: true},              // arg0 % arg1
   220  
   221  		{name: "ANDL", argLength: 2, reg: gp21, asm: "ANDL", commutative: true, resultInArg0: true, clobberFlags: true}, // arg0 & arg1
   222  		{name: "ANDLconst", argLength: 1, reg: gp11, asm: "ANDL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // arg0 & auxint
   223  
   224  		{name: "ORL", argLength: 2, reg: gp21, asm: "ORL", commutative: true, resultInArg0: true, clobberFlags: true}, // arg0 | arg1
   225  		{name: "ORLconst", argLength: 1, reg: gp11, asm: "ORL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // arg0 | auxint
   226  
   227  		{name: "XORL", argLength: 2, reg: gp21, asm: "XORL", commutative: true, resultInArg0: true, clobberFlags: true}, // arg0 ^ arg1
   228  		{name: "XORLconst", argLength: 1, reg: gp11, asm: "XORL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // arg0 ^ auxint
   229  
   230  		{name: "CMPL", argLength: 2, reg: gp2flags, asm: "CMPL", typ: "Flags"},                    // arg0 compare to arg1
   231  		{name: "CMPW", argLength: 2, reg: gp2flags, asm: "CMPW", typ: "Flags"},                    // arg0 compare to arg1
   232  		{name: "CMPB", argLength: 2, reg: gp2flags, asm: "CMPB", typ: "Flags"},                    // arg0 compare to arg1
   233  		{name: "CMPLconst", argLength: 1, reg: gp1flags, asm: "CMPL", typ: "Flags", aux: "Int32"}, // arg0 compare to auxint
   234  		{name: "CMPWconst", argLength: 1, reg: gp1flags, asm: "CMPW", typ: "Flags", aux: "Int16"}, // arg0 compare to auxint
   235  		{name: "CMPBconst", argLength: 1, reg: gp1flags, asm: "CMPB", typ: "Flags", aux: "Int8"},  // arg0 compare to auxint
   236  
   237  		// compare *(arg0+auxint+aux) to arg1 (in that order). arg2=mem.
   238  		{name: "CMPLload", argLength: 3, reg: gp1flagsLoad, asm: "CMPL", aux: "SymOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   239  		{name: "CMPWload", argLength: 3, reg: gp1flagsLoad, asm: "CMPW", aux: "SymOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   240  		{name: "CMPBload", argLength: 3, reg: gp1flagsLoad, asm: "CMPB", aux: "SymOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   241  
   242  		// compare *(arg0+ValAndOff(AuxInt).Off()+aux) to ValAndOff(AuxInt).Val() (in that order). arg1=mem.
   243  		{name: "CMPLconstload", argLength: 2, reg: gp0flagsLoad, asm: "CMPL", aux: "SymValAndOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   244  		{name: "CMPWconstload", argLength: 2, reg: gp0flagsLoad, asm: "CMPW", aux: "SymValAndOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   245  		{name: "CMPBconstload", argLength: 2, reg: gp0flagsLoad, asm: "CMPB", aux: "SymValAndOff", typ: "Flags", symEffect: "Read", faultOnNilArg0: true},
   246  
   247  		{name: "UCOMISS", argLength: 2, reg: fp2flags, asm: "UCOMISS", typ: "Flags"}, // arg0 compare to arg1, f32
   248  		{name: "UCOMISD", argLength: 2, reg: fp2flags, asm: "UCOMISD", typ: "Flags"}, // arg0 compare to arg1, f64
   249  
   250  		{name: "TESTL", argLength: 2, reg: gp2flags, commutative: true, asm: "TESTL", typ: "Flags"}, // (arg0 & arg1) compare to 0
   251  		{name: "TESTW", argLength: 2, reg: gp2flags, commutative: true, asm: "TESTW", typ: "Flags"}, // (arg0 & arg1) compare to 0
   252  		{name: "TESTB", argLength: 2, reg: gp2flags, commutative: true, asm: "TESTB", typ: "Flags"}, // (arg0 & arg1) compare to 0
   253  		{name: "TESTLconst", argLength: 1, reg: gp1flags, asm: "TESTL", typ: "Flags", aux: "Int32"}, // (arg0 & auxint) compare to 0
   254  		{name: "TESTWconst", argLength: 1, reg: gp1flags, asm: "TESTW", typ: "Flags", aux: "Int16"}, // (arg0 & auxint) compare to 0
   255  		{name: "TESTBconst", argLength: 1, reg: gp1flags, asm: "TESTB", typ: "Flags", aux: "Int8"},  // (arg0 & auxint) compare to 0
   256  
   257  		{name: "SHLL", argLength: 2, reg: gp21shift, asm: "SHLL", resultInArg0: true, clobberFlags: true},               // arg0 << arg1, shift amount is mod 32
   258  		{name: "SHLLconst", argLength: 1, reg: gp11, asm: "SHLL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // arg0 << auxint, shift amount 0-31
   259  		// Note: x86 is weird, the 16 and 8 byte shifts still use all 5 bits of shift amount!
   260  
   261  		{name: "SHRL", argLength: 2, reg: gp21shift, asm: "SHRL", resultInArg0: true, clobberFlags: true},               // unsigned arg0 >> arg1, shift amount is mod 32
   262  		{name: "SHRW", argLength: 2, reg: gp21shift, asm: "SHRW", resultInArg0: true, clobberFlags: true},               // unsigned arg0 >> arg1, shift amount is mod 32
   263  		{name: "SHRB", argLength: 2, reg: gp21shift, asm: "SHRB", resultInArg0: true, clobberFlags: true},               // unsigned arg0 >> arg1, shift amount is mod 32
   264  		{name: "SHRLconst", argLength: 1, reg: gp11, asm: "SHRL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // unsigned arg0 >> auxint, shift amount 0-31
   265  		{name: "SHRWconst", argLength: 1, reg: gp11, asm: "SHRW", aux: "Int16", resultInArg0: true, clobberFlags: true}, // unsigned arg0 >> auxint, shift amount 0-15
   266  		{name: "SHRBconst", argLength: 1, reg: gp11, asm: "SHRB", aux: "Int8", resultInArg0: true, clobberFlags: true},  // unsigned arg0 >> auxint, shift amount 0-7
   267  
   268  		{name: "SARL", argLength: 2, reg: gp21shift, asm: "SARL", resultInArg0: true, clobberFlags: true},               // signed arg0 >> arg1, shift amount is mod 32
   269  		{name: "SARW", argLength: 2, reg: gp21shift, asm: "SARW", resultInArg0: true, clobberFlags: true},               // signed arg0 >> arg1, shift amount is mod 32
   270  		{name: "SARB", argLength: 2, reg: gp21shift, asm: "SARB", resultInArg0: true, clobberFlags: true},               // signed arg0 >> arg1, shift amount is mod 32
   271  		{name: "SARLconst", argLength: 1, reg: gp11, asm: "SARL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // signed arg0 >> auxint, shift amount 0-31
   272  		{name: "SARWconst", argLength: 1, reg: gp11, asm: "SARW", aux: "Int16", resultInArg0: true, clobberFlags: true}, // signed arg0 >> auxint, shift amount 0-15
   273  		{name: "SARBconst", argLength: 1, reg: gp11, asm: "SARB", aux: "Int8", resultInArg0: true, clobberFlags: true},  // signed arg0 >> auxint, shift amount 0-7
   274  
   275  		{name: "ROLL", argLength: 2, reg: gp21shift, asm: "ROLL", resultInArg0: true, clobberFlags: true},               //     32 bits of arg0 rotate left by arg1
   276  		{name: "ROLW", argLength: 2, reg: gp21shift, asm: "ROLW", resultInArg0: true, clobberFlags: true},               // low 16 bits of arg0 rotate left by arg1
   277  		{name: "ROLB", argLength: 2, reg: gp21shift, asm: "ROLB", resultInArg0: true, clobberFlags: true},               // low  8 bits of arg0 rotate left by arg1
   278  		{name: "ROLLconst", argLength: 1, reg: gp11, asm: "ROLL", aux: "Int32", resultInArg0: true, clobberFlags: true}, // arg0 rotate left auxint, rotate amount 0-31
   279  		{name: "ROLWconst", argLength: 1, reg: gp11, asm: "ROLW", aux: "Int16", resultInArg0: true, clobberFlags: true}, // arg0 rotate left auxint, rotate amount 0-15
   280  		{name: "ROLBconst", argLength: 1, reg: gp11, asm: "ROLB", aux: "Int8", resultInArg0: true, clobberFlags: true},  // arg0 rotate left auxint, rotate amount 0-7
   281  
   282  		// binary-op with a memory source operand
   283  		{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
   284  		{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
   285  		{name: "MULLload", argLength: 3, reg: gp21load, asm: "IMULL", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true, symEffect: "Read"}, // arg0 * tmp, tmp loaded from  arg1+auxint+aux, arg2 = mem
   286  		{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
   287  		{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
   288  		{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
   289  
   290  		// binary-op with an indexed memory source operand
   291  		{name: "ADDLloadidx4", argLength: 4, reg: gp21loadidx, asm: "ADDL", aux: "SymOff", resultInArg0: true, clobberFlags: true, symEffect: "Read"},  // arg0 + tmp, tmp loaded from  arg1+arg2*4+auxint+aux, arg3 = mem
   292  		{name: "SUBLloadidx4", argLength: 4, reg: gp21loadidx, asm: "SUBL", aux: "SymOff", resultInArg0: true, clobberFlags: true, symEffect: "Read"},  // arg0 - tmp, tmp loaded from  arg1+arg2*4+auxint+aux, arg3 = mem
   293  		{name: "MULLloadidx4", argLength: 4, reg: gp21loadidx, asm: "IMULL", aux: "SymOff", resultInArg0: true, clobberFlags: true, symEffect: "Read"}, // arg0 * tmp, tmp loaded from  arg1+arg2*4+auxint+aux, arg3 = mem
   294  		{name: "ANDLloadidx4", argLength: 4, reg: gp21loadidx, asm: "ANDL", aux: "SymOff", resultInArg0: true, clobberFlags: true, symEffect: "Read"},  // arg0 & tmp, tmp loaded from  arg1+arg2*4+auxint+aux, arg3 = mem
   295  		{name: "ORLloadidx4", argLength: 4, reg: gp21loadidx, asm: "ORL", aux: "SymOff", resultInArg0: true, clobberFlags: true, symEffect: "Read"},    // arg0 | tmp, tmp loaded from  arg1+arg2*4+auxint+aux, arg3 = mem
   296  		{name: "XORLloadidx4", argLength: 4, reg: gp21loadidx, asm: "XORL", aux: "SymOff", resultInArg0: true, clobberFlags: true, symEffect: "Read"},  // arg0 ^ tmp, tmp loaded from  arg1+arg2*4+auxint+aux, arg3 = mem
   297  
   298  		// unary ops
   299  		{name: "NEGL", argLength: 1, reg: gp11, asm: "NEGL", resultInArg0: true, clobberFlags: true}, // -arg0
   300  
   301  		{name: "NOTL", argLength: 1, reg: gp11, asm: "NOTL", resultInArg0: true}, // ^arg0
   302  
   303  		{name: "BSFL", argLength: 1, reg: gp11, asm: "BSFL", clobberFlags: true}, // arg0 # of low-order zeroes ; undef if zero
   304  		{name: "BSFW", argLength: 1, reg: gp11, asm: "BSFW", clobberFlags: true}, // arg0 # of low-order zeroes ; undef if zero
   305  		{name: "LoweredCtz32", argLength: 1, reg: gp11, clobberFlags: true},      // arg0 # of low-order zeroes
   306  
   307  		{name: "BSRL", argLength: 1, reg: gp11, asm: "BSRL", clobberFlags: true}, // arg0 # of high-order zeroes ; undef if zero
   308  		{name: "BSRW", argLength: 1, reg: gp11, asm: "BSRW", clobberFlags: true}, // arg0 # of high-order zeroes ; undef if zero
   309  
   310  		{name: "BSWAPL", argLength: 1, reg: gp11, asm: "BSWAPL", resultInArg0: true}, // arg0 swap bytes
   311  
   312  		{name: "SQRTSD", argLength: 1, reg: fp11, asm: "SQRTSD"}, // sqrt(arg0)
   313  		{name: "SQRTSS", argLength: 1, reg: fp11, asm: "SQRTSS"}, // sqrt(arg0), float32
   314  
   315  		{name: "SBBLcarrymask", argLength: 1, reg: flagsgp, asm: "SBBL"}, // (int32)(-1) if carry is set, 0 if carry is clear.
   316  		// Note: SBBW and SBBB are subsumed by SBBL
   317  
   318  		{name: "SETEQ", argLength: 1, reg: readflags, asm: "SETEQ"}, // extract == condition from arg0
   319  		{name: "SETNE", argLength: 1, reg: readflags, asm: "SETNE"}, // extract != condition from arg0
   320  		{name: "SETL", argLength: 1, reg: readflags, asm: "SETLT"},  // extract signed < condition from arg0
   321  		{name: "SETLE", argLength: 1, reg: readflags, asm: "SETLE"}, // extract signed <= condition from arg0
   322  		{name: "SETG", argLength: 1, reg: readflags, asm: "SETGT"},  // extract signed > condition from arg0
   323  		{name: "SETGE", argLength: 1, reg: readflags, asm: "SETGE"}, // extract signed >= condition from arg0
   324  		{name: "SETB", argLength: 1, reg: readflags, asm: "SETCS"},  // extract unsigned < condition from arg0
   325  		{name: "SETBE", argLength: 1, reg: readflags, asm: "SETLS"}, // extract unsigned <= condition from arg0
   326  		{name: "SETA", argLength: 1, reg: readflags, asm: "SETHI"},  // extract unsigned > condition from arg0
   327  		{name: "SETAE", argLength: 1, reg: readflags, asm: "SETCC"}, // extract unsigned >= condition from arg0
   328  		{name: "SETO", argLength: 1, reg: readflags, asm: "SETOS"},  // extract if overflow flag is set from arg0
   329  		// Need different opcodes for floating point conditions because
   330  		// any comparison involving a NaN is always FALSE and thus
   331  		// the patterns for inverting conditions cannot be used.
   332  		{name: "SETEQF", argLength: 1, reg: flagsgpax, asm: "SETEQ", clobberFlags: true}, // extract == condition from arg0
   333  		{name: "SETNEF", argLength: 1, reg: flagsgpax, asm: "SETNE", clobberFlags: true}, // extract != condition from arg0
   334  		{name: "SETORD", argLength: 1, reg: flagsgp, asm: "SETPC"},                       // extract "ordered" (No Nan present) condition from arg0
   335  		{name: "SETNAN", argLength: 1, reg: flagsgp, asm: "SETPS"},                       // extract "unordered" (Nan present) condition from arg0
   336  
   337  		{name: "SETGF", argLength: 1, reg: flagsgp, asm: "SETHI"},  // extract floating > condition from arg0
   338  		{name: "SETGEF", argLength: 1, reg: flagsgp, asm: "SETCC"}, // extract floating >= condition from arg0
   339  
   340  		{name: "MOVBLSX", argLength: 1, reg: gp11, asm: "MOVBLSX"}, // sign extend arg0 from int8 to int32
   341  		{name: "MOVBLZX", argLength: 1, reg: gp11, asm: "MOVBLZX"}, // zero extend arg0 from int8 to int32
   342  		{name: "MOVWLSX", argLength: 1, reg: gp11, asm: "MOVWLSX"}, // sign extend arg0 from int16 to int32
   343  		{name: "MOVWLZX", argLength: 1, reg: gp11, asm: "MOVWLZX"}, // zero extend arg0 from int16 to int32
   344  
   345  		{name: "MOVLconst", reg: gp01, asm: "MOVL", typ: "UInt32", aux: "Int32", rematerializeable: true}, // 32 low bits of auxint
   346  
   347  		{name: "CVTTSD2SL", argLength: 1, reg: fpgp, asm: "CVTTSD2SL"}, // convert float64 to int32
   348  		{name: "CVTTSS2SL", argLength: 1, reg: fpgp, asm: "CVTTSS2SL"}, // convert float32 to int32
   349  		{name: "CVTSL2SS", argLength: 1, reg: gpfp, asm: "CVTSL2SS"},   // convert int32 to float32
   350  		{name: "CVTSL2SD", argLength: 1, reg: gpfp, asm: "CVTSL2SD"},   // convert int32 to float64
   351  		{name: "CVTSD2SS", argLength: 1, reg: fp11, asm: "CVTSD2SS"},   // convert float64 to float32
   352  		{name: "CVTSS2SD", argLength: 1, reg: fp11, asm: "CVTSS2SD"},   // convert float32 to float64
   353  
   354  		{name: "PXOR", argLength: 2, reg: fp21, asm: "PXOR", commutative: true, resultInArg0: true}, // exclusive or, applied to X regs for float negation.
   355  
   356  		{name: "LEAL", argLength: 1, reg: gp11sb, aux: "SymOff", rematerializeable: true, symEffect: "Addr"}, // arg0 + auxint + offset encoded in aux
   357  		{name: "LEAL1", argLength: 2, reg: gp21sb, commutative: true, aux: "SymOff", symEffect: "Addr"},      // arg0 + arg1 + auxint + aux
   358  		{name: "LEAL2", argLength: 2, reg: gp21sb, aux: "SymOff", symEffect: "Addr"},                         // arg0 + 2*arg1 + auxint + aux
   359  		{name: "LEAL4", argLength: 2, reg: gp21sb, aux: "SymOff", symEffect: "Addr"},                         // arg0 + 4*arg1 + auxint + aux
   360  		{name: "LEAL8", argLength: 2, reg: gp21sb, aux: "SymOff", symEffect: "Addr"},                         // arg0 + 8*arg1 + auxint + aux
   361  		// Note: LEAL{1,2,4,8} must not have OpSB as either argument.
   362  
   363  		// auxint+aux == add auxint and the offset of the symbol in aux (if any) to the effective address
   364  		{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.
   365  		{name: "MOVBLSXload", argLength: 2, reg: gpload, asm: "MOVBLSX", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"},             // ditto, sign extend to int32
   366  		{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.
   367  		{name: "MOVWLSXload", argLength: 2, reg: gpload, asm: "MOVWLSX", aux: "SymOff", faultOnNilArg0: true, symEffect: "Read"},             // ditto, sign extend to int32
   368  		{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.
   369  		{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
   370  		{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
   371  		{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
   372  
   373  		// direct binary-op on memory (read-modify-write)
   374  		{name: "ADDLmodify", argLength: 3, reg: gpstore, asm: "ADDL", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, clobberFlags: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) += arg1, arg2=mem
   375  		{name: "SUBLmodify", argLength: 3, reg: gpstore, asm: "SUBL", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, clobberFlags: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) -= arg1, arg2=mem
   376  		{name: "ANDLmodify", argLength: 3, reg: gpstore, asm: "ANDL", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, clobberFlags: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) &= arg1, arg2=mem
   377  		{name: "ORLmodify", argLength: 3, reg: gpstore, asm: "ORL", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, clobberFlags: true, symEffect: "Read,Write"},   // *(arg0+auxint+aux) |= arg1, arg2=mem
   378  		{name: "XORLmodify", argLength: 3, reg: gpstore, asm: "XORL", aux: "SymOff", typ: "Mem", faultOnNilArg0: true, clobberFlags: true, symEffect: "Read,Write"}, // *(arg0+auxint+aux) ^= arg1, arg2=mem
   379  
   380  		// direct binary-op on indexed memory (read-modify-write)
   381  		{name: "ADDLmodifyidx4", argLength: 4, reg: gpstoreidx, asm: "ADDL", aux: "SymOff", typ: "Mem", clobberFlags: true, symEffect: "Read,Write"}, // *(arg0+arg1*4+auxint+aux) += arg2, arg3=mem
   382  		{name: "SUBLmodifyidx4", argLength: 4, reg: gpstoreidx, asm: "SUBL", aux: "SymOff", typ: "Mem", clobberFlags: true, symEffect: "Read,Write"}, // *(arg0+arg1*4+auxint+aux) -= arg2, arg3=mem
   383  		{name: "ANDLmodifyidx4", argLength: 4, reg: gpstoreidx, asm: "ANDL", aux: "SymOff", typ: "Mem", clobberFlags: true, symEffect: "Read,Write"}, // *(arg0+arg1*4+auxint+aux) &= arg2, arg3=mem
   384  		{name: "ORLmodifyidx4", argLength: 4, reg: gpstoreidx, asm: "ORL", aux: "SymOff", typ: "Mem", clobberFlags: true, symEffect: "Read,Write"},   // *(arg0+arg1*4+auxint+aux) |= arg2, arg3=mem
   385  		{name: "XORLmodifyidx4", argLength: 4, reg: gpstoreidx, asm: "XORL", aux: "SymOff", typ: "Mem", clobberFlags: true, symEffect: "Read,Write"}, // *(arg0+arg1*4+auxint+aux) ^= arg2, arg3=mem
   386  
   387  		// direct binary-op on memory with a constant (read-modify-write)
   388  		{name: "ADDLconstmodify", argLength: 2, reg: gpstoreconst, asm: "ADDL", aux: "SymValAndOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // add ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   389  		{name: "ANDLconstmodify", argLength: 2, reg: gpstoreconst, asm: "ANDL", aux: "SymValAndOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // and ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   390  		{name: "ORLconstmodify", argLength: 2, reg: gpstoreconst, asm: "ORL", aux: "SymValAndOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"},   // or  ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   391  		{name: "XORLconstmodify", argLength: 2, reg: gpstoreconst, asm: "XORL", aux: "SymValAndOff", typ: "Mem", clobberFlags: true, faultOnNilArg0: true, symEffect: "Read,Write"}, // xor ValAndOff(AuxInt).Val() to arg0+ValAndOff(AuxInt).Off()+aux, arg1=mem
   392  
   393  		// direct binary-op on indexed memory with a constant (read-modify-write)
   394  		{name: "ADDLconstmodifyidx4", argLength: 3, reg: gpstoreconstidx, asm: "ADDL", aux: "SymValAndOff", typ: "Mem", clobberFlags: true, symEffect: "Read,Write"}, // add ValAndOff(AuxInt).Val() to arg0+arg1*4+ValAndOff(AuxInt).Off()+aux, arg2=mem
   395  		{name: "ANDLconstmodifyidx4", argLength: 3, reg: gpstoreconstidx, asm: "ANDL", aux: "SymValAndOff", typ: "Mem", clobberFlags: true, symEffect: "Read,Write"}, // and ValAndOff(AuxInt).Val() to arg0+arg1*4+ValAndOff(AuxInt).Off()+aux, arg2=mem
   396  		{name: "ORLconstmodifyidx4", argLength: 3, reg: gpstoreconstidx, asm: "ORL", aux: "SymValAndOff", typ: "Mem", clobberFlags: true, symEffect: "Read,Write"},   // or  ValAndOff(AuxInt).Val() to arg0+arg1*4+ValAndOff(AuxInt).Off()+aux, arg2=mem
   397  		{name: "XORLconstmodifyidx4", argLength: 3, reg: gpstoreconstidx, asm: "XORL", aux: "SymValAndOff", typ: "Mem", clobberFlags: true, symEffect: "Read,Write"}, // xor ValAndOff(AuxInt).Val() to arg0+arg1*4+ValAndOff(AuxInt).Off()+aux, arg2=mem
   398  
   399  		// indexed loads/stores
   400  		{name: "MOVBloadidx1", argLength: 3, reg: gploadidx, commutative: true, asm: "MOVBLZX", aux: "SymOff", symEffect: "Read"}, // load a byte from arg0+arg1+auxint+aux. arg2=mem
   401  		{name: "MOVWloadidx1", argLength: 3, reg: gploadidx, commutative: true, asm: "MOVWLZX", aux: "SymOff", symEffect: "Read"}, // load 2 bytes from arg0+arg1+auxint+aux. arg2=mem
   402  		{name: "MOVWloadidx2", argLength: 3, reg: gploadidx, asm: "MOVWLZX", aux: "SymOff", symEffect: "Read"},                    // load 2 bytes from arg0+2*arg1+auxint+aux. arg2=mem
   403  		{name: "MOVLloadidx1", argLength: 3, reg: gploadidx, commutative: true, asm: "MOVL", aux: "SymOff", symEffect: "Read"},    // load 4 bytes from arg0+arg1+auxint+aux. arg2=mem
   404  		{name: "MOVLloadidx4", argLength: 3, reg: gploadidx, asm: "MOVL", aux: "SymOff", symEffect: "Read"},                       // load 4 bytes from arg0+4*arg1+auxint+aux. arg2=mem
   405  		// TODO: sign-extending indexed loads
   406  		{name: "MOVBstoreidx1", argLength: 4, reg: gpstoreidx, commutative: true, asm: "MOVB", aux: "SymOff", symEffect: "Write"}, // store byte in arg2 to arg0+arg1+auxint+aux. arg3=mem
   407  		{name: "MOVWstoreidx1", argLength: 4, reg: gpstoreidx, commutative: true, asm: "MOVW", aux: "SymOff", symEffect: "Write"}, // store 2 bytes in arg2 to arg0+arg1+auxint+aux. arg3=mem
   408  		{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
   409  		{name: "MOVLstoreidx1", argLength: 4, reg: gpstoreidx, commutative: true, asm: "MOVL", aux: "SymOff", symEffect: "Write"}, // store 4 bytes in arg2 to arg0+arg1+auxint+aux. arg3=mem
   410  		{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
   411  		// TODO: add size-mismatched indexed loads, like MOVBstoreidx4.
   412  
   413  		// For storeconst ops, the AuxInt field encodes both
   414  		// the value to store and an address offset of the store.
   415  		// Cast AuxInt to a ValAndOff to extract Val and Off fields.
   416  		{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
   417  		{name: "MOVWstoreconst", argLength: 2, reg: gpstoreconst, asm: "MOVW", aux: "SymValAndOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // store low 2 bytes of ...
   418  		{name: "MOVLstoreconst", argLength: 2, reg: gpstoreconst, asm: "MOVL", aux: "SymValAndOff", typ: "Mem", faultOnNilArg0: true, symEffect: "Write"}, // store low 4 bytes of ...
   419  
   420  		{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
   421  		{name: "MOVWstoreconstidx1", argLength: 3, reg: gpstoreconstidx, asm: "MOVW", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store low 2 bytes of ... arg1 ...
   422  		{name: "MOVWstoreconstidx2", argLength: 3, reg: gpstoreconstidx, asm: "MOVW", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store low 2 bytes of ... 2*arg1 ...
   423  		{name: "MOVLstoreconstidx1", argLength: 3, reg: gpstoreconstidx, asm: "MOVL", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store low 4 bytes of ... arg1 ...
   424  		{name: "MOVLstoreconstidx4", argLength: 3, reg: gpstoreconstidx, asm: "MOVL", aux: "SymValAndOff", typ: "Mem", symEffect: "Write"}, // store low 4 bytes of ... 4*arg1 ...
   425  
   426  		// arg0 = pointer to start of memory to zero
   427  		// arg1 = value to store (will always be zero)
   428  		// arg2 = mem
   429  		// auxint = offset into duffzero code to start executing
   430  		// returns mem
   431  		{
   432  			name:      "DUFFZERO",
   433  			aux:       "Int64",
   434  			argLength: 3,
   435  			reg: regInfo{
   436  				inputs:   []regMask{buildReg("DI"), buildReg("AX")},
   437  				clobbers: buildReg("DI CX"),
   438  				// Note: CX is only clobbered when dynamic linking.
   439  			},
   440  			faultOnNilArg0: true,
   441  		},
   442  
   443  		// arg0 = address of memory to zero
   444  		// arg1 = # of 4-byte words to zero
   445  		// arg2 = value to store (will always be zero)
   446  		// arg3 = mem
   447  		// returns mem
   448  		{
   449  			name:      "REPSTOSL",
   450  			argLength: 4,
   451  			reg: regInfo{
   452  				inputs:   []regMask{buildReg("DI"), buildReg("CX"), buildReg("AX")},
   453  				clobbers: buildReg("DI CX"),
   454  			},
   455  			faultOnNilArg0: true,
   456  		},
   457  
   458  		{name: "CALLstatic", argLength: 1, reg: regInfo{clobbers: callerSave}, aux: "CallOff", clobberFlags: true, call: true},                                              // call static function aux.(*obj.LSym).  arg0=mem, auxint=argsize, returns mem
   459  		{name: "CALLtail", argLength: 1, reg: regInfo{clobbers: callerSave}, aux: "CallOff", clobberFlags: true, call: true, tailCall: true},                                // tail call static function aux.(*obj.LSym).  arg0=mem, auxint=argsize, returns mem
   460  		{name: "CALLclosure", argLength: 3, reg: regInfo{inputs: []regMask{gpsp, buildReg("DX"), 0}, clobbers: callerSave}, aux: "CallOff", clobberFlags: true, call: true}, // call function via closure.  arg0=codeptr, arg1=closure, arg2=mem, auxint=argsize, returns mem
   461  		{name: "CALLinter", argLength: 2, reg: regInfo{inputs: []regMask{gp}, clobbers: callerSave}, aux: "CallOff", clobberFlags: true, call: true},                        // call fn by pointer.  arg0=codeptr, arg1=mem, auxint=argsize, returns mem
   462  
   463  		// arg0 = destination pointer
   464  		// arg1 = source pointer
   465  		// arg2 = mem
   466  		// auxint = offset from duffcopy symbol to call
   467  		// returns memory
   468  		{
   469  			name:      "DUFFCOPY",
   470  			aux:       "Int64",
   471  			argLength: 3,
   472  			reg: regInfo{
   473  				inputs:   []regMask{buildReg("DI"), buildReg("SI")},
   474  				clobbers: buildReg("DI SI CX"), // uses CX as a temporary
   475  			},
   476  			clobberFlags:   true,
   477  			faultOnNilArg0: true,
   478  			faultOnNilArg1: true,
   479  		},
   480  
   481  		// arg0 = destination pointer
   482  		// arg1 = source pointer
   483  		// arg2 = # of 8-byte words to copy
   484  		// arg3 = mem
   485  		// returns memory
   486  		{
   487  			name:      "REPMOVSL",
   488  			argLength: 4,
   489  			reg: regInfo{
   490  				inputs:   []regMask{buildReg("DI"), buildReg("SI"), buildReg("CX")},
   491  				clobbers: buildReg("DI SI CX"),
   492  			},
   493  			faultOnNilArg0: true,
   494  			faultOnNilArg1: true,
   495  		},
   496  
   497  		// (InvertFlags (CMPL a b)) == (CMPL b a)
   498  		// So if we want (SETL (CMPL a b)) but we can't do that because a is a constant,
   499  		// then we do (SETL (InvertFlags (CMPL b a))) instead.
   500  		// Rewrites will convert this to (SETG (CMPL b a)).
   501  		// InvertFlags is a pseudo-op which can't appear in assembly output.
   502  		{name: "InvertFlags", argLength: 1}, // reverse direction of arg0
   503  
   504  		// Pseudo-ops
   505  		{name: "LoweredGetG", argLength: 1, reg: gp01}, // arg0=mem
   506  		// Scheduler ensures LoweredGetClosurePtr occurs only in entry block,
   507  		// and sorts it to the very beginning of the block to prevent other
   508  		// use of DX (the closure pointer)
   509  		{name: "LoweredGetClosurePtr", reg: regInfo{outputs: []regMask{buildReg("DX")}}, zeroWidth: true},
   510  		// LoweredGetCallerPC evaluates to the PC to which its "caller" will return.
   511  		// I.e., if f calls g "calls" getcallerpc,
   512  		// the result should be the PC within f that g will return to.
   513  		// See runtime/stubs.go for a more detailed discussion.
   514  		{name: "LoweredGetCallerPC", reg: gp01, rematerializeable: true},
   515  		// LoweredGetCallerSP returns the SP of the caller of the current function. arg0=mem
   516  		{name: "LoweredGetCallerSP", argLength: 1, reg: gp01, rematerializeable: true},
   517  		//arg0=ptr,arg1=mem, returns void.  Faults if ptr is nil.
   518  		{name: "LoweredNilCheck", argLength: 2, reg: regInfo{inputs: []regMask{gpsp}}, clobberFlags: true, nilCheck: true, faultOnNilArg0: true},
   519  
   520  		// LoweredWB invokes runtime.gcWriteBarrier. arg0=mem, auxint=# of write barrier slots
   521  		// It saves all GP registers if necessary, but may clobber others.
   522  		// Returns a pointer to a write barrier buffer in DI.
   523  		{name: "LoweredWB", argLength: 1, reg: regInfo{clobbers: callerSave &^ gp, outputs: []regMask{buildReg("DI")}}, clobberFlags: true, aux: "Int64"},
   524  
   525  		// There are three of these functions so that they can have three different register inputs.
   526  		// When we check 0 <= c <= cap (A), then 0 <= b <= c (B), then 0 <= a <= b (C), we want the
   527  		// default registers to match so we don't need to copy registers around unnecessarily.
   528  		{name: "LoweredPanicBoundsA", argLength: 3, aux: "Int64", reg: regInfo{inputs: []regMask{dx, bx}}, typ: "Mem", call: true}, // arg0=idx, arg1=len, arg2=mem, returns memory. AuxInt contains report code (see PanicBounds in genericOps.go).
   529  		{name: "LoweredPanicBoundsB", argLength: 3, aux: "Int64", reg: regInfo{inputs: []regMask{cx, dx}}, typ: "Mem", call: true}, // arg0=idx, arg1=len, arg2=mem, returns memory. AuxInt contains report code (see PanicBounds in genericOps.go).
   530  		{name: "LoweredPanicBoundsC", argLength: 3, aux: "Int64", reg: regInfo{inputs: []regMask{ax, cx}}, typ: "Mem", call: true}, // arg0=idx, arg1=len, arg2=mem, returns memory. AuxInt contains report code (see PanicBounds in genericOps.go).
   531  		// Extend ops are the same as Bounds ops except the indexes are 64-bit.
   532  		{name: "LoweredPanicExtendA", argLength: 4, aux: "Int64", reg: regInfo{inputs: []regMask{si, dx, bx}}, typ: "Mem", call: true}, // arg0=idxHi, arg1=idxLo, arg2=len, arg3=mem, returns memory. AuxInt contains report code (see PanicExtend in genericOps.go).
   533  		{name: "LoweredPanicExtendB", argLength: 4, aux: "Int64", reg: regInfo{inputs: []regMask{si, cx, dx}}, typ: "Mem", call: true}, // arg0=idxHi, arg1=idxLo, arg2=len, arg3=mem, returns memory. AuxInt contains report code (see PanicExtend in genericOps.go).
   534  		{name: "LoweredPanicExtendC", argLength: 4, aux: "Int64", reg: regInfo{inputs: []regMask{si, ax, cx}}, typ: "Mem", call: true}, // arg0=idxHi, arg1=idxLo, arg2=len, arg3=mem, returns memory. AuxInt contains report code (see PanicExtend in genericOps.go).
   535  
   536  		// Constant flag values. For any comparison, there are 5 possible
   537  		// outcomes: the three from the signed total order (<,==,>) and the
   538  		// three from the unsigned total order. The == cases overlap.
   539  		// Note: there's a sixth "unordered" outcome for floating-point
   540  		// comparisons, but we don't use such a beast yet.
   541  		// These ops are for temporary use by rewrite rules. They
   542  		// cannot appear in the generated assembly.
   543  		{name: "FlagEQ"},     // equal
   544  		{name: "FlagLT_ULT"}, // signed < and unsigned <
   545  		{name: "FlagLT_UGT"}, // signed < and unsigned >
   546  		{name: "FlagGT_UGT"}, // signed > and unsigned <
   547  		{name: "FlagGT_ULT"}, // signed > and unsigned >
   548  
   549  		// Special ops for PIC floating-point constants.
   550  		// MOVSXconst1 loads the address of the constant-pool entry into a register.
   551  		// MOVSXconst2 loads the constant from that address.
   552  		// MOVSXconst1 returns a pointer, but we type it as uint32 because it can never point to the Go heap.
   553  		{name: "MOVSSconst1", reg: gp01, typ: "UInt32", aux: "Float32"},
   554  		{name: "MOVSDconst1", reg: gp01, typ: "UInt32", aux: "Float64"},
   555  		{name: "MOVSSconst2", argLength: 1, reg: gpfp, asm: "MOVSS"},
   556  		{name: "MOVSDconst2", argLength: 1, reg: gpfp, asm: "MOVSD"},
   557  	}
   558  
   559  	var _386blocks = []blockData{
   560  		{name: "EQ", controls: 1},
   561  		{name: "NE", controls: 1},
   562  		{name: "LT", controls: 1},
   563  		{name: "LE", controls: 1},
   564  		{name: "GT", controls: 1},
   565  		{name: "GE", controls: 1},
   566  		{name: "OS", controls: 1},
   567  		{name: "OC", controls: 1},
   568  		{name: "ULT", controls: 1},
   569  		{name: "ULE", controls: 1},
   570  		{name: "UGT", controls: 1},
   571  		{name: "UGE", controls: 1},
   572  		{name: "EQF", controls: 1},
   573  		{name: "NEF", controls: 1},
   574  		{name: "ORD", controls: 1}, // FP, ordered comparison (parity zero)
   575  		{name: "NAN", controls: 1}, // FP, unordered comparison (parity one)
   576  	}
   577  
   578  	archs = append(archs, arch{
   579  		name:            "386",
   580  		pkg:             "cmd/internal/obj/x86",
   581  		genfile:         "../../x86/ssa.go",
   582  		ops:             _386ops,
   583  		blocks:          _386blocks,
   584  		regnames:        regNames386,
   585  		gpregmask:       gp,
   586  		fpregmask:       fp,
   587  		framepointerreg: int8(num["BP"]),
   588  		linkreg:         -1, // not used
   589  	})
   590  }
   591
View as plain text