summaryrefslogtreecommitdiff
path: root/src/arch/x86/isa/microops/regop.isa
blob: f37b4327bdafc8d1517ca93311455088a397f61e (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
// Copyright (c) 2007 The Hewlett-Packard Development Company
// All rights reserved.
//
// Redistribution and use of this software in source and binary forms,
// with or without modification, are permitted provided that the
// following conditions are met:
//
// The software must be used only for Non-Commercial Use which means any
// use which is NOT directed to receiving any direct monetary
// compensation for, or commercial advantage from such use.  Illustrative
// examples of non-commercial use are academic research, personal study,
// teaching, education and corporate research & development.
// Illustrative examples of commercial use are distributing products for
// commercial advantage and providing services using the software for
// commercial advantage.
//
// If you wish to use this software or functionality therein that may be
// covered by patents for commercial use, please contact:
//     Director of Intellectual Property Licensing
//     Office of Strategy and Technology
//     Hewlett-Packard Company
//     1501 Page Mill Road
//     Palo Alto, California  94304
//
// Redistributions of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.  Redistributions
// in binary form must reproduce the above copyright notice, this list of
// conditions and the following disclaimer in the documentation and/or
// other materials provided with the distribution.  Neither the name of
// the COPYRIGHT HOLDER(s), HEWLETT-PACKARD COMPANY, nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.  No right of
// sublicense is granted herewith.  Derivatives of the software and
// output created using the software may be prepared, but only for
// Non-Commercial Uses.  Derivatives of the software may be shared with
// others provided: (i) the others agree to abide by the list of
// conditions herein which includes the Non-Commercial Use restrictions;
// and (ii) such Derivatives of the software include the above copyright
// notice to acknowledge the contribution from this software where
// applicable, this list of conditions and the disclaimer below.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
//
// Authors: Gabe Black

//////////////////////////////////////////////////////////////////////////
//
// RegOp Microop templates
//
//////////////////////////////////////////////////////////////////////////

def template MicroRegOpExecute {{
        Fault %(class_name)s::execute(%(CPU_exec_context)s *xc,
                Trace::InstRecord *traceData) const
        {
            Fault fault = NoFault;

            DPRINTF(X86, "The data size is %d\n", dataSize);
            %(op_decl)s;
            %(op_rd)s;

            if(%(cond_check)s)
            {
                %(code)s;
                %(flag_code)s;
            }
            else
            {
                %(else_code)s;
            }

            //Write the resulting state to the execution context
            if(fault == NoFault)
            {
                %(op_wb)s;
            }
            return fault;
        }
}};

def template MicroRegOpImmExecute {{
        Fault %(class_name)s::execute(%(CPU_exec_context)s *xc,
                Trace::InstRecord *traceData) const
        {
            Fault fault = NoFault;

            %(op_decl)s;
            %(op_rd)s;

            if(%(cond_check)s)
            {
                %(code)s;
                %(flag_code)s;
            }
            else
            {
                %(else_code)s;
            }

            //Write the resulting state to the execution context
            if(fault == NoFault)
            {
                %(op_wb)s;
            }
            return fault;
        }
}};

def template MicroRegOpDeclare {{
    class %(class_name)s : public %(base_class)s
    {
      protected:
        void buildMe();

      public:
        %(class_name)s(ExtMachInst _machInst,
                const char * instMnem,
                bool isMicro, bool isDelayed, bool isFirst, bool isLast,
                RegIndex _src1, RegIndex _src2, RegIndex _dest,
                uint8_t _dataSize, uint16_t _ext);

        %(class_name)s(ExtMachInst _machInst,
                const char * instMnem,
                RegIndex _src1, RegIndex _src2, RegIndex _dest,
                uint8_t _dataSize, uint16_t _ext);

        %(BasicExecDeclare)s
    };
}};

def template MicroRegOpImmDeclare {{

    class %(class_name)s : public %(base_class)s
    {
      protected:
        void buildMe();

      public:
        %(class_name)s(ExtMachInst _machInst,
                const char * instMnem,
                bool isMicro, bool isDelayed, bool isFirst, bool isLast,
                RegIndex _src1, uint16_t _imm8, RegIndex _dest,
                uint8_t _dataSize, uint16_t _ext);

        %(class_name)s(ExtMachInst _machInst,
                const char * instMnem,
                RegIndex _src1, uint16_t _imm8, RegIndex _dest,
                uint8_t _dataSize, uint16_t _ext);

        %(BasicExecDeclare)s
    };
}};

def template MicroRegOpConstructor {{

    inline void %(class_name)s::buildMe()
    {
        %(constructor)s;
    }

    inline %(class_name)s::%(class_name)s(
            ExtMachInst machInst, const char * instMnem,
            RegIndex _src1, RegIndex _src2, RegIndex _dest,
            uint8_t _dataSize, uint16_t _ext) :
        %(base_class)s(machInst, "%(mnemonic)s", instMnem,
                false, false, false, false,
                _src1, _src2, _dest, _dataSize, _ext,
                %(op_class)s)
    {
        buildMe();
    }

    inline %(class_name)s::%(class_name)s(
            ExtMachInst machInst, const char * instMnem,
            bool isMicro, bool isDelayed, bool isFirst, bool isLast,
            RegIndex _src1, RegIndex _src2, RegIndex _dest,
            uint8_t _dataSize, uint16_t _ext) :
        %(base_class)s(machInst, "%(mnemonic)s", instMnem,
                isMicro, isDelayed, isFirst, isLast,
                _src1, _src2, _dest, _dataSize, _ext,
                %(op_class)s)
    {
        buildMe();
    }
}};

def template MicroRegOpImmConstructor {{

    inline void %(class_name)s::buildMe()
    {
        %(constructor)s;
    }

    inline %(class_name)s::%(class_name)s(
            ExtMachInst machInst, const char * instMnem,
            RegIndex _src1, uint16_t _imm8, RegIndex _dest,
            uint8_t _dataSize, uint16_t _ext) :
        %(base_class)s(machInst, "%(mnemonic)s", instMnem,
                false, false, false, false,
                _src1, _imm8, _dest, _dataSize, _ext,
                %(op_class)s)
    {
        buildMe();
    }

    inline %(class_name)s::%(class_name)s(
            ExtMachInst machInst, const char * instMnem,
            bool isMicro, bool isDelayed, bool isFirst, bool isLast,
            RegIndex _src1, uint16_t _imm8, RegIndex _dest,
            uint8_t _dataSize, uint16_t _ext) :
        %(base_class)s(machInst, "%(mnemonic)s", instMnem,
                isMicro, isDelayed, isFirst, isLast,
                _src1, _imm8, _dest, _dataSize, _ext,
                %(op_class)s)
    {
        buildMe();
    }
}};

output header {{
    void
    divide(uint64_t dividend, uint64_t divisor,
            uint64_t &quotient, uint64_t &remainder);
}};

output decoder {{
    void
    divide(uint64_t dividend, uint64_t divisor,
            uint64_t &quotient, uint64_t &remainder)
    {
        //Check for divide by zero.
        if (divisor == 0)
            panic("Divide by zero!\\n");
        //If the divisor is bigger than the dividend, don't do anything.
        if (divisor <= dividend) {
            //Shift the divisor so it's msb lines up with the dividend.
            int dividendMsb = findMsbSet(dividend);
            int divisorMsb = findMsbSet(divisor);
            int shift = dividendMsb - divisorMsb;
            divisor <<= shift;
            //Compute what we'll add to the quotient if the divisor isn't
            //now larger than the dividend.
            uint64_t quotientBit = 1;
            quotientBit <<= shift;
            //If we need to step back a bit (no pun intended) because the
            //divisor got too to large, do that here. This is the "or two"
            //part of one or two bit division.
            if (divisor > dividend) {
                quotientBit >>= 1;
                divisor >>= 1;
            }
            //Decrement the remainder and increment the quotient.
            quotient += quotientBit;
            remainder -= divisor;
        }
    }
}};

let {{
    # Make these empty strings so that concatenating onto
    # them will always work.
    header_output = ""
    decoder_output = ""
    exec_output = ""

    immTemplates = (
            MicroRegOpImmDeclare,
            MicroRegOpImmConstructor,
            MicroRegOpImmExecute)

    regTemplates = (
            MicroRegOpDeclare,
            MicroRegOpConstructor,
            MicroRegOpExecute)

    class RegOpMeta(type):
        def buildCppClasses(self, name, Name, suffix, \
                code, flag_code, cond_check, else_code):

            # Globals to stick the output in
            global header_output
            global decoder_output
            global exec_output

            # Stick all the code together so it can be searched at once
            allCode = "|".join((code, flag_code, cond_check, else_code))

            # If op2 is used anywhere, make register and immediate versions
            # of this code.
            matcher = re.compile("(?<!\\w)(?P<prefix>s?)op2(?P<typeQual>\\.\\w+)?")
            match = matcher.search(allCode)
            if match:
                typeQual = ""
                if match.group("typeQual"):
                    typeQual = match.group("typeQual")
                src2_name = "%spsrc2%s" % (match.group("prefix"), typeQual)
                self.buildCppClasses(name, Name, suffix,
                        matcher.sub(src2_name, code),
                        matcher.sub(src2_name, flag_code),
                        matcher.sub(src2_name, cond_check),
                        matcher.sub(src2_name, else_code))
                self.buildCppClasses(name + "i", Name, suffix + "Imm",
                        matcher.sub("imm8", code),
                        matcher.sub("imm8", flag_code),
                        matcher.sub("imm8", cond_check),
                        matcher.sub("imm8", else_code))
                return

            # If there's something optional to do with flags, generate
            # a version without it and fix up this version to use it.
            if flag_code != "" or cond_check != "true":
                self.buildCppClasses(name, Name, suffix,
                        code, "", "true", else_code)
                suffix = "Flags" + suffix

            # If psrc1 or psrc2 is used, we need to actually insert code to
            # compute it.
            matcher = re.compile("(?<!\w)psrc1(?!\w)")
            if matcher.search(allCode):
                code = "uint64_t psrc1 = pick(SrcReg1, 0, dataSize);" + code
            matcher = re.compile("(?<!\w)psrc2(?!\w)")
            if matcher.search(allCode):
                code = "uint64_t psrc2 = pick(SrcReg2, 1, dataSize);" + code
            # Also make available versions which do sign extension
            matcher = re.compile("(?<!\w)spsrc1(?!\w)")
            if matcher.search(allCode):
                code = "int64_t spsrc1 = signedPick(SrcReg1, 0, dataSize);" + code
            matcher = re.compile("(?<!\w)spsrc2(?!\w)")
            if matcher.search(allCode):
                code = "int64_t spsrc2 = signedPick(SrcReg2, 1, dataSize);" + code

            base = "X86ISA::RegOp"

            # If imm8 shows up in the code, use the immediate templates, if
            # not, hopefully the register ones will be correct.
            templates = regTemplates
            matcher = re.compile("(?<!\w)imm8(?!\w)")
            if matcher.search(allCode):
                base += "Imm"
                templates = immTemplates

            # Get everything ready for the substitution
            iop = InstObjParams(name, Name + suffix, base,
                    {"code" : code,
                     "flag_code" : flag_code,
                     "cond_check" : cond_check,
                     "else_code" : else_code})

            # Generate the actual code (finally!)
            header_output += templates[0].subst(iop)
            decoder_output += templates[1].subst(iop)
            exec_output += templates[2].subst(iop)


        def __new__(mcls, Name, bases, dict):
            abstract = False
            name = Name.lower()
            if "abstract" in dict:
                abstract = dict['abstract']
                del dict['abstract']

            cls = super(RegOpMeta, mcls).__new__(mcls, Name, bases, dict)
            if not abstract:
                cls.className = Name
                cls.base_mnemonic = name
                code = cls.code
                flag_code = cls.flag_code
                cond_check = cls.cond_check
                else_code = cls.else_code

                # Set up the C++ classes
                mcls.buildCppClasses(cls, name, Name, "",
                        code, flag_code, cond_check, else_code)

                # Hook into the microassembler dict
                global microopClasses
                microopClasses[name] = cls

                allCode = "|".join((code, flag_code, cond_check, else_code))

                # If op2 is used anywhere, make register and immediate versions
                # of this code.
                matcher = re.compile("op2(?P<typeQual>\\.\\w+)?")
                if matcher.search(allCode):
                    microopClasses[name + 'i'] = cls
            return cls


    class RegOp(X86Microop):
        __metaclass__ = RegOpMeta
        # This class itself doesn't act as a microop
        abstract = True

        # Default template parameter values
        flag_code = ""
        cond_check = "true"
        else_code = ";"

        def __init__(self, dest, src1, op2, flags = None, dataSize = "env.dataSize"):
            self.dest = dest
            self.src1 = src1
            self.op2 = op2
            self.flags = flags
            self.dataSize = dataSize
            if flags is None:
                self.ext = 0
            else:
                if not isinstance(flags, (list, tuple)):
                    raise Exception, "flags must be a list or tuple of flags"
                self.ext = " | ".join(flags)
                self.className += "Flags"

        def getAllocator(self, *microFlags):
            className = self.className
            if self.mnemonic == self.base_mnemonic + 'i':
                className += "Imm"
            allocator = '''new %(class_name)s(machInst, mnemonic
                    %(flags)s, %(src1)s, %(op2)s, %(dest)s,
                    %(dataSize)s, %(ext)s)''' % {
                "class_name" : className,
                "flags" : self.microFlagsText(microFlags),
                "src1" : self.src1, "op2" : self.op2,
                "dest" : self.dest,
                "dataSize" : self.dataSize,
                "ext" : self.ext}
            return allocator

    class LogicRegOp(RegOp):
        abstract = True
        flag_code = '''
            //Don't have genFlags handle the OF or CF bits
            uint64_t mask = CFBit | ECFBit | OFBit;
            ccFlagBits = genFlags(ccFlagBits, ext & ~mask, DestReg, psrc1, op2);
            //If a logic microop wants to set these, it wants to set them to 0.
            ccFlagBits &= ~(CFBit & ext);
            ccFlagBits &= ~(ECFBit & ext);
            ccFlagBits &= ~(OFBit & ext);
        '''

    class FlagRegOp(RegOp):
        abstract = True
        flag_code = \
            "ccFlagBits = genFlags(ccFlagBits, ext, DestReg, psrc1, op2);"

    class SubRegOp(RegOp):
        abstract = True
        flag_code = \
            "ccFlagBits = genFlags(ccFlagBits, ext, DestReg, psrc1, ~op2, true);"

    class CondRegOp(RegOp):
        abstract = True
        cond_check = "checkCondition(ccFlagBits, ext)"

    class RdRegOp(RegOp):
        abstract = True
        def __init__(self, dest, src1=None, dataSize="env.dataSize"):
            if not src1:
                src1 = dest
            super(RdRegOp, self).__init__(dest, src1, "NUM_INTREGS", None, dataSize)

    class WrRegOp(RegOp):
        abstract = True
        def __init__(self, src1, src2, flags=None, dataSize="env.dataSize"):
            super(WrRegOp, self).__init__("NUM_INTREGS", src1, src2, flags, dataSize)

    class Add(FlagRegOp):
        code = 'DestReg = merge(DestReg, psrc1 + op2, dataSize);'

    class Or(LogicRegOp):
        code = 'DestReg = merge(DestReg, psrc1 | op2, dataSize);'

    class Adc(FlagRegOp):
        code = '''
            CCFlagBits flags = ccFlagBits;
            DestReg = merge(DestReg, psrc1 + op2 + flags.cf, dataSize);
            '''

    class Sbb(SubRegOp):
        code = '''
            CCFlagBits flags = ccFlagBits;
            DestReg = merge(DestReg, psrc1 - op2 - flags.cf, dataSize);
            '''

    class And(LogicRegOp):
        code = 'DestReg = merge(DestReg, psrc1 & op2, dataSize)'

    class Sub(SubRegOp):
        code = 'DestReg = merge(DestReg, psrc1 - op2, dataSize)'

    class Xor(LogicRegOp):
        code = 'DestReg = merge(DestReg, psrc1 ^ op2, dataSize)'

    # Neither of these is quite correct because it assumes that right shifting
    # a signed or unsigned value does sign or zero extension respectively.
    # The C standard says that what happens on a right shift with a 1 in the
    # MSB position is undefined. On x86 and under likely most compilers the
    # "right thing" happens, but this isn't a guarantee.
    class Mul1s(WrRegOp):
        code = '''
            ProdLow = psrc1 * op2;
            int halfSize = (dataSize * 8) / 2;
            int64_t spsrc1_h = spsrc1 >> halfSize;
            int64_t spsrc1_l = spsrc1 & mask(halfSize);
            int64_t spsrc2_h = sop2 >> halfSize;
            int64_t spsrc2_l = sop2 & mask(halfSize);
            ProdHi = ((spsrc1_l * spsrc2_h + spsrc1_h * spsrc2_l +
                      ((spsrc1_l * spsrc2_l) >> halfSize)) >> halfSize) +
                     spsrc1_h * spsrc2_h;
            '''

    class Mul1u(WrRegOp):
        code = '''
            ProdLow = psrc1 * op2;
            int halfSize = (dataSize * 8) / 2;
            uint64_t psrc1_h = psrc1 >> halfSize;
            uint64_t psrc1_l = psrc1 & mask(halfSize);
            uint64_t psrc2_h = op2 >> halfSize;
            uint64_t psrc2_l = op2 & mask(halfSize);
            ProdHi = ((psrc1_l * psrc2_h + psrc1_h * psrc2_l +
                      ((psrc1_l * psrc2_l) >> halfSize)) >> halfSize) +
                     psrc1_h * psrc2_h;
            '''

    class Mulel(RdRegOp):
        code = 'DestReg = merge(SrcReg1, ProdLow, dataSize);'

    class Muleh(RdRegOp):
        def __init__(self, dest, src1=None, flags=None, dataSize="env.dataSize"):
            if not src1:
                src1 = dest
            super(RdRegOp, self).__init__(dest, src1, "NUM_INTREGS", flags, dataSize)
        code = 'DestReg = merge(SrcReg1, ProdHi, dataSize);'
        flag_code = '''
            if (ProdHi)
                ccFlagBits = ccFlagBits | (ext & (CFBit | OFBit | ECFBit));
            else
                ccFlagBits = ccFlagBits & ~(ext & (CFBit | OFBit | ECFBit));
        '''

    # One or two bit divide
    class Div1(WrRegOp):
        code = '''
            //These are temporaries so that modifying them later won't make
            //the ISA parser think they're also sources.
            uint64_t quotient = 0;
            uint64_t remainder = psrc1;
            //Similarly, this is a temporary so changing it doesn't make it
            //a source.
            uint64_t divisor = op2;
            //This is a temporary just for consistency and clarity.
            uint64_t dividend = remainder;
            //Do the division.
            divide(dividend, divisor, quotient, remainder);
            //Record the final results.
            Remainder = remainder;
            Quotient = quotient;
            Divisor = divisor;
            '''

    # Step divide
    class Div2(RegOp):
        code = '''
            uint64_t dividend = Remainder;
            uint64_t divisor = Divisor;
            uint64_t quotient = Quotient;
            uint64_t remainder = dividend;
            int remaining = op2;
            //If we overshot, do nothing. This lets us unrool division loops a
            //little.
            if (remaining) {
                //Shift in bits from the low order portion of the dividend
                while(dividend < divisor && remaining) {
                    dividend = (dividend << 1) | bits(SrcReg1, remaining - 1);
                    quotient <<= 1;
                    remaining--;
                }
                remainder = dividend;
                //Do the division.
                divide(dividend, divisor, quotient, remainder);
            }
            //Keep track of how many bits there are still to pull in.
            DestReg = merge(DestReg, remaining, dataSize);
            //Record the final results
            Remainder = remainder;
            Quotient = quotient;
        '''
        flag_code = '''
            if (DestReg == 0)
                ccFlagBits = ccFlagBits | (ext & EZFBit);
            else
                ccFlagBits = ccFlagBits & ~(ext & EZFBit);
        '''

    class Divq(RdRegOp):
        code = 'DestReg = merge(SrcReg1, Quotient, dataSize);'

    class Divr(RdRegOp):
        code = 'DestReg = merge(SrcReg1, Remainder, dataSize);'

    class Mov(CondRegOp):
        code = 'DestReg = merge(SrcReg1, op2, dataSize)'
        else_code = 'DestReg=DestReg;'

    # Shift instructions

    class Sll(RegOp):
        code = '''
            uint8_t shiftAmt = (op2 & ((dataSize == 8) ? mask(6) : mask(5)));
            DestReg = merge(DestReg, psrc1 << shiftAmt, dataSize);
            '''
        flag_code = '''
            // If the shift amount is zero, no flags should be modified.
            if (shiftAmt) {
                //Zero out any flags we might modify. This way we only have to
                //worry about setting them.
                ccFlagBits = ccFlagBits & ~(ext & (CFBit | ECFBit | OFBit));
                int CFBits = 0;
                //Figure out if we -would- set the CF bits if requested.
                if (bits(SrcReg1, dataSize * 8 - shiftAmt))
                    CFBits = 1;
                //If some combination of the CF bits need to be set, set them.
                if ((ext & (CFBit | ECFBit)) && CFBits)
                    ccFlagBits = ccFlagBits | (ext & (CFBit | ECFBit));
                //Figure out what the OF bit should be.
                if ((ext & OFBit) && (CFBits ^ bits(DestReg, dataSize * 8 - 1)))
                    ccFlagBits = ccFlagBits | OFBit;
                //Use the regular mechanisms to calculate the other flags.
                ccFlagBits = genFlags(ccFlagBits, ext & ~(CFBit | ECFBit | OFBit),
                        DestReg, psrc1, op2);
            }
        '''

    class Srl(RegOp):
        code = '''
            uint8_t shiftAmt = (op2 & ((dataSize == 8) ? mask(6) : mask(5)));
            // Because what happens to the bits shift -in- on a right shift
            // is not defined in the C/C++ standard, we have to mask them out
            // to be sure they're zero.
            uint64_t logicalMask = mask(dataSize * 8 - shiftAmt);
            DestReg = merge(DestReg, (psrc1 >> shiftAmt) & logicalMask, dataSize);
            '''
        flag_code = '''
            // If the shift amount is zero, no flags should be modified.
            if (shiftAmt) {
                //Zero out any flags we might modify. This way we only have to
                //worry about setting them.
                ccFlagBits = ccFlagBits & ~(ext & (CFBit | ECFBit | OFBit));
                //If some combination of the CF bits need to be set, set them.
                if ((ext & (CFBit | ECFBit)) && bits(SrcReg1, shiftAmt - 1))
                    ccFlagBits = ccFlagBits | (ext & (CFBit | ECFBit));
                //Figure out what the OF bit should be.
                if ((ext & OFBit) && bits(SrcReg1, dataSize * 8 - 1))
                    ccFlagBits = ccFlagBits | OFBit;
                //Use the regular mechanisms to calculate the other flags.
                ccFlagBits = genFlags(ccFlagBits, ext & ~(CFBit | ECFBit | OFBit),
                        DestReg, psrc1, op2);
            }
        '''

    class Sra(RegOp):
        code = '''
            uint8_t shiftAmt = (op2 & ((dataSize == 8) ? mask(6) : mask(5)));
            // Because what happens to the bits shift -in- on a right shift
            // is not defined in the C/C++ standard, we have to sign extend
            // them manually to be sure.
            uint64_t arithMask =
                -bits(psrc1, dataSize * 8 - 1) << (dataSize * 8 - shiftAmt);
            DestReg = merge(DestReg, (psrc1 >> shiftAmt) | arithMask, dataSize);
            '''
        flag_code = '''
            // If the shift amount is zero, no flags should be modified.
            if (shiftAmt) {
                //Zero out any flags we might modify. This way we only have to
                //worry about setting them.
                ccFlagBits = ccFlagBits & ~(ext & (CFBit | ECFBit | OFBit));
                //If some combination of the CF bits need to be set, set them.
                if ((ext & (CFBit | ECFBit)) && bits(SrcReg1, shiftAmt - 1))
                    ccFlagBits = ccFlagBits | (ext & (CFBit | ECFBit));
                //Use the regular mechanisms to calculate the other flags.
                ccFlagBits = genFlags(ccFlagBits, ext & ~(CFBit | ECFBit | OFBit),
                        DestReg, psrc1, op2);
            }
        '''

    class Ror(RegOp):
        code = '''
            uint8_t shiftAmt =
                (op2 & ((dataSize == 8) ? mask(6) : mask(5)));
            if(shiftAmt)
            {
                uint64_t top = psrc1 << (dataSize * 8 - shiftAmt);
                uint64_t bottom = bits(psrc1, dataSize * 8, shiftAmt);
                DestReg = merge(DestReg, top | bottom, dataSize);
            }
            else
                DestReg = DestReg;
            '''
        flag_code = '''
            // If the shift amount is zero, no flags should be modified.
            if (shiftAmt) {
                //Zero out any flags we might modify. This way we only have to
                //worry about setting them.
                ccFlagBits = ccFlagBits & ~(ext & (CFBit | ECFBit | OFBit));
                //Find the most and second most significant bits of the result.
                int msb = bits(DestReg, dataSize * 8 - 1);
                int smsb = bits(DestReg, dataSize * 8 - 2);
                //If some combination of the CF bits need to be set, set them.
                if ((ext & (CFBit | ECFBit)) && msb)
                    ccFlagBits = ccFlagBits | (ext & (CFBit | ECFBit));
                //Figure out what the OF bit should be.
                if ((ext & OFBit) && (msb ^ smsb))
                    ccFlagBits = ccFlagBits | OFBit;
                //Use the regular mechanisms to calculate the other flags.
                ccFlagBits = genFlags(ccFlagBits, ext & ~(CFBit | ECFBit | OFBit),
                        DestReg, psrc1, op2);
            }
        '''

    class Rcr(RegOp):
        code = '''
            uint8_t shiftAmt =
                (op2 & ((dataSize == 8) ? mask(6) : mask(5)));
            if(shiftAmt)
            {
                CCFlagBits flags = ccFlagBits;
                uint64_t top = flags.cf << (dataSize * 8 - shiftAmt);
                if(shiftAmt > 1)
                    top |= psrc1 << (dataSize * 8 - shiftAmt - 1);
                uint64_t bottom = bits(psrc1, dataSize * 8, shiftAmt);
                DestReg = merge(DestReg, top | bottom, dataSize);
            }
            else
                DestReg = DestReg;
            '''
        flag_code = '''
            // If the shift amount is zero, no flags should be modified.
            if (shiftAmt) {
                //Zero out any flags we might modify. This way we only have to
                //worry about setting them.
                ccFlagBits = ccFlagBits & ~(ext & (CFBit | ECFBit | OFBit));
                //Figure out what the OF bit should be.
                if ((ext & OFBit) && ((ccFlagBits & CFBit) ^
                                      bits(SrcReg1, dataSize * 8 - 1)))
                    ccFlagBits = ccFlagBits | OFBit;
                //If some combination of the CF bits need to be set, set them.
                if ((ext & (CFBit | ECFBit)) && bits(SrcReg1, shiftAmt - 1))
                    ccFlagBits = ccFlagBits | (ext & (CFBit | ECFBit));
                //Use the regular mechanisms to calculate the other flags.
                ccFlagBits = genFlags(ccFlagBits, ext & ~(CFBit | ECFBit | OFBit),
                        DestReg, psrc1, op2);
            }
        '''

    class Rol(RegOp):
        code = '''
            uint8_t shiftAmt =
                (op2 & ((dataSize == 8) ? mask(6) : mask(5)));
            if(shiftAmt)
            {
                uint64_t top = psrc1 << shiftAmt;
                uint64_t bottom =
                    bits(psrc1, dataSize * 8 - 1, dataSize * 8 - shiftAmt);
                DestReg = merge(DestReg, top | bottom, dataSize);
            }
            else
                DestReg = DestReg;
            '''
        flag_code = '''
            // If the shift amount is zero, no flags should be modified.
            if (shiftAmt) {
                //Zero out any flags we might modify. This way we only have to
                //worry about setting them.
                ccFlagBits = ccFlagBits & ~(ext & (CFBit | ECFBit | OFBit));
                //The CF bits, if set, would be set to the lsb of the result.
                int lsb = DestReg & 0x1;
                int msb = bits(DestReg, dataSize * 8 - 1);
                //If some combination of the CF bits need to be set, set them.
                if ((ext & (CFBit | ECFBit)) && lsb)
                    ccFlagBits = ccFlagBits | (ext & (CFBit | ECFBit));
                //Figure out what the OF bit should be.
                if ((ext & OFBit) && (msb ^ lsb))
                    ccFlagBits = ccFlagBits | OFBit;
                //Use the regular mechanisms to calculate the other flags.
                ccFlagBits = genFlags(ccFlagBits, ext & ~(CFBit | ECFBit | OFBit),
                        DestReg, psrc1, op2);
            }
        '''

    class Rcl(RegOp):
        code = '''
            uint8_t shiftAmt =
                (op2 & ((dataSize == 8) ? mask(6) : mask(5)));
            if(shiftAmt)
            {
                CCFlagBits flags = ccFlagBits;
                uint64_t top = psrc1 << shiftAmt;
                uint64_t bottom = flags.cf << (shiftAmt - 1);
                if(shiftAmt > 1)
                    bottom |=
                        bits(psrc1, dataSize * 8 - 1,
                                   dataSize * 8 - shiftAmt + 1);
                DestReg = merge(DestReg, top | bottom, dataSize);
            }
            else
                DestReg = DestReg;
            '''
        flag_code = '''
            // If the shift amount is zero, no flags should be modified.
            if (shiftAmt) {
                //Zero out any flags we might modify. This way we only have to
                //worry about setting them.
                ccFlagBits = ccFlagBits & ~(ext & (CFBit | ECFBit | OFBit));
                int msb = bits(DestReg, dataSize * 8 - 1);
                int CFBits = bits(SrcReg1, dataSize * 8 - shiftAmt);
                //If some combination of the CF bits need to be set, set them.
                if ((ext & (CFBit | ECFBit)) && CFBits)
                    ccFlagBits = ccFlagBits | (ext & (CFBit | ECFBit));
                //Figure out what the OF bit should be.
                if ((ext & OFBit) && (msb ^ CFBits))
                    ccFlagBits = ccFlagBits | OFBit;
                //Use the regular mechanisms to calculate the other flags.
                ccFlagBits = genFlags(ccFlagBits, ext & ~(CFBit | ECFBit | OFBit),
                        DestReg, psrc1, op2);
            }
        '''

    class Wrip(WrRegOp, CondRegOp):
        code = 'RIP = psrc1 + sop2 + CSBase'
        else_code="RIP = RIP;"

    class Br(WrRegOp, CondRegOp):
        code = 'nuIP = psrc1 + op2;'
        else_code='nuIP = nuIP;'

    class Wruflags(WrRegOp):
        code = 'ccFlagBits = psrc1 ^ op2'

    class Rdip(RdRegOp):
        code = 'DestReg = RIP - CSBase'

    class Ruflags(RdRegOp):
        code = 'DestReg = ccFlagBits'

    class Ruflag(RegOp):
        code = '''
            int flag = bits(ccFlagBits, imm8);
            DestReg = merge(DestReg, flag, dataSize);
            ccFlagBits = (flag == 0) ? (ccFlagBits | EZFBit) :
                                       (ccFlagBits & ~EZFBit);
            '''
        def __init__(self, dest, imm, flags=None, \
                dataSize="env.dataSize"):
            super(Ruflag, self).__init__(dest, \
                    "NUM_INTREGS", imm, flags, dataSize)

    class Sext(RegOp):
        code = '''
            IntReg val = psrc1;
            // Mask the bit position so that it wraps.
            int bitPos = op2 & (dataSize * 8 - 1);
            int sign_bit = bits(val, bitPos, bitPos);
            uint64_t maskVal = mask(bitPos+1);
            val = sign_bit ? (val | ~maskVal) : (val & maskVal);
            DestReg = merge(DestReg, val, dataSize);
            '''
        flag_code = '''
            if (!sign_bit)
                ccFlagBits = ccFlagBits &
                    ~(ext & (CFBit | ECFBit | ZFBit | EZFBit));
            else
                ccFlagBits = ccFlagBits |
                    (ext & (CFBit | ECFBit | ZFBit | EZFBit));
            '''

    class Zext(RegOp):
        code = 'DestReg = bits(psrc1, op2, 0);'

    class Wrcr(RegOp):
        def __init__(self, dest, src1, flags=None, dataSize="env.dataSize"):
            super(Wrcr, self).__init__(dest, \
                    src1, "NUM_INTREGS", flags, dataSize)
        code = '''
            if (dest == 1 || (dest > 4 && dest < 8) || (dest > 8)) {
                fault = new InvalidOpcode();
            } else {
                // There are *s in the line below so it doesn't confuse the
                // parser. They may be unnecessary.
                //Mis*cReg old*Val = pick(Cont*rolDest, 0, dat*aSize);
                MiscReg newVal = psrc1;

                // Check for any modifications that would cause a fault.
                switch(dest) {
                  case 0:
                    {
                        Efer efer = EferOp;
                        CR0 cr0 = newVal;
                        CR4 oldCr4 = CR4Op;
                        if (bits(newVal, 63, 32) ||
                                (!cr0.pe && cr0.pg) ||
                                (!cr0.cd && cr0.nw) ||
                                (cr0.pg && efer.lme && !oldCr4.pae))
                            fault = new GeneralProtection(0);
                    }
                    break;
                  case 2:
                    break;
                  case 3:
                    break;
                  case 4:
                    {
                        CR4 cr4 = newVal;
                        // PAE can't be disabled in long mode.
                        if (bits(newVal, 63, 11) ||
                                (machInst.mode.mode == LongMode && !cr4.pae))
                            fault = new GeneralProtection(0);
                    }
                    break;
                  case 8:
                    {
                        if (bits(newVal, 63, 4))
                            fault = new GeneralProtection(0);
                    }
                  default:
                    panic("Unrecognized control register %d.\\n", dest);
                }
                ControlDest = newVal;
            }
            '''

    class Wrbase(RegOp):
        def __init__(self, dest, src1, flags=None, dataSize="env.dataSize"):
            super(Wrbase, self).__init__(dest, \
                    src1, "NUM_INTREGS", flags, dataSize)
        code = '''
            SysSegBaseDest = psrc1;
        '''

    class Wrlimit(RegOp):
        def __init__(self, dest, src1, flags=None, dataSize="env.dataSize"):
            super(Wrlimit, self).__init__(dest, \
                    src1, "NUM_INTREGS", flags, dataSize)
        code = '''
            SysSegLimitDest = psrc1;
        '''
}};