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
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
|
//===-- ARMAsmParser.cpp - Parse ARM assembly to MCInst instructions ------===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
#include "ARM.h"
#include "ARMAddressingModes.h"
#include "ARMSubtarget.h"
#include "llvm/MC/MCParser/MCAsmLexer.h"
#include "llvm/MC/MCParser/MCAsmParser.h"
#include "llvm/MC/MCParser/MCParsedAsmOperand.h"
#include "llvm/MC/MCContext.h"
#include "llvm/MC/MCStreamer.h"
#include "llvm/MC/MCExpr.h"
#include "llvm/MC/MCInst.h"
#include "llvm/Target/TargetRegistry.h"
#include "llvm/Target/TargetAsmParser.h"
#include "llvm/Support/SourceMgr.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/StringSwitch.h"
#include "llvm/ADT/Twine.h"
using namespace llvm;
// The shift types for register controlled shifts in arm memory addressing
enum ShiftType {
Lsl,
Lsr,
Asr,
Ror,
Rrx
};
namespace {
class ARMOperand;
class ARMAsmParser : public TargetAsmParser {
MCAsmParser &Parser;
TargetMachine &TM;
MCAsmParser &getParser() const { return Parser; }
MCAsmLexer &getLexer() const { return Parser.getLexer(); }
void Warning(SMLoc L, const Twine &Msg) { Parser.Warning(L, Msg); }
bool Error(SMLoc L, const Twine &Msg) { return Parser.Error(L, Msg); }
int TryParseRegister();
ARMOperand *TryParseRegisterWithWriteBack();
ARMOperand *ParseRegisterList();
ARMOperand *ParseMemory();
ARMOperand *ParseOperand();
bool ParseMemoryOffsetReg(bool &Negative,
bool &OffsetRegShifted,
enum ShiftType &ShiftType,
const MCExpr *&ShiftAmount,
const MCExpr *&Offset,
bool &OffsetIsReg,
int &OffsetRegNum,
SMLoc &E);
bool ParseShift(enum ShiftType &St, const MCExpr *&ShiftAmount, SMLoc &E);
bool ParseDirectiveWord(unsigned Size, SMLoc L);
bool ParseDirectiveThumb(SMLoc L);
bool ParseDirectiveThumbFunc(SMLoc L);
bool ParseDirectiveCode(SMLoc L);
bool ParseDirectiveSyntax(SMLoc L);
bool MatchAndEmitInstruction(SMLoc IDLoc,
SmallVectorImpl<MCParsedAsmOperand*> &Operands,
MCStreamer &Out);
/// @name Auto-generated Match Functions
/// {
#define GET_ASSEMBLER_HEADER
#include "ARMGenAsmMatcher.inc"
/// }
public:
ARMAsmParser(const Target &T, MCAsmParser &_Parser, TargetMachine &_TM)
: TargetAsmParser(T), Parser(_Parser), TM(_TM) {
// Initialize the set of available features.
setAvailableFeatures(ComputeAvailableFeatures(
&TM.getSubtarget<ARMSubtarget>()));
}
virtual bool ParseInstruction(StringRef Name, SMLoc NameLoc,
SmallVectorImpl<MCParsedAsmOperand*> &Operands);
virtual bool ParseDirective(AsmToken DirectiveID);
};
} // end anonymous namespace
namespace {
/// ARMOperand - Instances of this class represent a parsed ARM machine
/// instruction.
class ARMOperand : public MCParsedAsmOperand {
enum KindTy {
CondCode,
Immediate,
Memory,
Register,
RegisterList,
DPRRegisterList,
SPRRegisterList,
Token
} Kind;
SMLoc StartLoc, EndLoc;
union {
struct {
ARMCC::CondCodes Val;
} CC;
struct {
const char *Data;
unsigned Length;
} Tok;
struct {
unsigned RegNum;
bool Writeback;
} Reg;
struct {
SmallVector<unsigned, 32> *Registers;
} RegList;
struct {
const MCExpr *Val;
} Imm;
// This is for all forms of ARM address expressions
struct {
unsigned BaseRegNum;
unsigned OffsetRegNum; // used when OffsetIsReg is true
const MCExpr *Offset; // used when OffsetIsReg is false
const MCExpr *ShiftAmount; // used when OffsetRegShifted is true
enum ShiftType ShiftType; // used when OffsetRegShifted is true
unsigned OffsetRegShifted : 1; // only used when OffsetIsReg is true
unsigned Preindexed : 1;
unsigned Postindexed : 1;
unsigned OffsetIsReg : 1;
unsigned Negative : 1; // only used when OffsetIsReg is true
unsigned Writeback : 1;
} Mem;
};
ARMOperand(KindTy K) : MCParsedAsmOperand(), Kind(K) {}
public:
ARMOperand(const ARMOperand &o) : MCParsedAsmOperand() {
Kind = o.Kind;
StartLoc = o.StartLoc;
EndLoc = o.EndLoc;
switch (Kind) {
case CondCode:
CC = o.CC;
break;
case Token:
Tok = o.Tok;
break;
case Register:
Reg = o.Reg;
break;
case RegisterList:
case DPRRegisterList:
case SPRRegisterList:
RegList = o.RegList;
break;
case Immediate:
Imm = o.Imm;
break;
case Memory:
Mem = o.Mem;
break;
}
}
~ARMOperand() {
if (isRegList())
delete RegList.Registers;
}
/// getStartLoc - Get the location of the first token of this operand.
SMLoc getStartLoc() const { return StartLoc; }
/// getEndLoc - Get the location of the last token of this operand.
SMLoc getEndLoc() const { return EndLoc; }
ARMCC::CondCodes getCondCode() const {
assert(Kind == CondCode && "Invalid access!");
return CC.Val;
}
StringRef getToken() const {
assert(Kind == Token && "Invalid access!");
return StringRef(Tok.Data, Tok.Length);
}
unsigned getReg() const {
assert(Kind == Register && "Invalid access!");
return Reg.RegNum;
}
const SmallVectorImpl<unsigned> &getRegList() const {
assert((Kind == RegisterList || Kind == DPRRegisterList ||
Kind == SPRRegisterList) && "Invalid access!");
return *RegList.Registers;
}
const MCExpr *getImm() const {
assert(Kind == Immediate && "Invalid access!");
return Imm.Val;
}
bool isCondCode() const { return Kind == CondCode; }
bool isImm() const { return Kind == Immediate; }
bool isReg() const { return Kind == Register; }
bool isRegList() const { return Kind == RegisterList; }
bool isDPRRegList() const { return Kind == DPRRegisterList; }
bool isSPRRegList() const { return Kind == SPRRegisterList; }
bool isToken() const { return Kind == Token; }
bool isMemory() const { return Kind == Memory; }
bool isMemMode5() const {
if (!isMemory() || Mem.OffsetIsReg || Mem.OffsetRegShifted ||
Mem.Writeback || Mem.Negative)
return false;
// If there is an offset expression, make sure it's valid.
if (!Mem.Offset)
return true;
const MCConstantExpr *CE = dyn_cast<MCConstantExpr>(Mem.Offset);
if (!CE)
return false;
// The offset must be a multiple of 4 in the range 0-1020.
int64_t Value = CE->getValue();
return ((Value & 0x3) == 0 && Value <= 1020 && Value >= -1020);
}
void addExpr(MCInst &Inst, const MCExpr *Expr) const {
// Add as immediates when possible. Null MCExpr = 0.
if (Expr == 0)
Inst.addOperand(MCOperand::CreateImm(0));
else if (const MCConstantExpr *CE = dyn_cast<MCConstantExpr>(Expr))
Inst.addOperand(MCOperand::CreateImm(CE->getValue()));
else
Inst.addOperand(MCOperand::CreateExpr(Expr));
}
void addCondCodeOperands(MCInst &Inst, unsigned N) const {
assert(N == 2 && "Invalid number of operands!");
Inst.addOperand(MCOperand::CreateImm(unsigned(getCondCode())));
// FIXME: What belongs here?
Inst.addOperand(MCOperand::CreateReg(0));
}
void addRegOperands(MCInst &Inst, unsigned N) const {
assert(N == 1 && "Invalid number of operands!");
Inst.addOperand(MCOperand::CreateReg(getReg()));
}
void addRegListOperands(MCInst &Inst, unsigned N) const {
assert(N == 1 && "Invalid number of operands!");
const SmallVectorImpl<unsigned> &RegList = getRegList();
for (SmallVectorImpl<unsigned>::const_iterator
I = RegList.begin(), E = RegList.end(); I != E; ++I)
Inst.addOperand(MCOperand::CreateReg(*I));
}
void addDPRRegListOperands(MCInst &Inst, unsigned N) const {
addRegListOperands(Inst, N);
}
void addSPRRegListOperands(MCInst &Inst, unsigned N) const {
addRegListOperands(Inst, N);
}
void addImmOperands(MCInst &Inst, unsigned N) const {
assert(N == 1 && "Invalid number of operands!");
addExpr(Inst, getImm());
}
void addMemMode5Operands(MCInst &Inst, unsigned N) const {
assert(N == 2 && isMemMode5() && "Invalid number of operands!");
Inst.addOperand(MCOperand::CreateReg(Mem.BaseRegNum));
assert(!Mem.OffsetIsReg && "Invalid mode 5 operand");
// FIXME: #-0 is encoded differently than #0. Does the parser preserve
// the difference?
if (Mem.Offset) {
const MCConstantExpr *CE = dyn_cast<MCConstantExpr>(Mem.Offset);
assert(CE && "Non-constant mode 5 offset operand!");
// The MCInst offset operand doesn't include the low two bits (like
// the instruction encoding).
int64_t Offset = CE->getValue() / 4;
if (Offset >= 0)
Inst.addOperand(MCOperand::CreateImm(ARM_AM::getAM5Opc(ARM_AM::add,
Offset)));
else
Inst.addOperand(MCOperand::CreateImm(ARM_AM::getAM5Opc(ARM_AM::sub,
-Offset)));
} else {
Inst.addOperand(MCOperand::CreateImm(0));
}
}
virtual void dump(raw_ostream &OS) const;
static ARMOperand *CreateCondCode(ARMCC::CondCodes CC, SMLoc S) {
ARMOperand *Op = new ARMOperand(CondCode);
Op->CC.Val = CC;
Op->StartLoc = S;
Op->EndLoc = S;
return Op;
}
static ARMOperand *CreateToken(StringRef Str, SMLoc S) {
ARMOperand *Op = new ARMOperand(Token);
Op->Tok.Data = Str.data();
Op->Tok.Length = Str.size();
Op->StartLoc = S;
Op->EndLoc = S;
return Op;
}
static ARMOperand *CreateReg(unsigned RegNum, bool Writeback, SMLoc S,
SMLoc E) {
ARMOperand *Op = new ARMOperand(Register);
Op->Reg.RegNum = RegNum;
Op->Reg.Writeback = Writeback;
Op->StartLoc = S;
Op->EndLoc = E;
return Op;
}
static ARMOperand *
CreateRegList(const SmallVectorImpl<std::pair<unsigned, SMLoc> > &Regs,
SMLoc StartLoc, SMLoc EndLoc) {
KindTy Kind = RegisterList;
if (ARM::DPRRegClass.contains(Regs.front().first))
Kind = DPRRegisterList;
else if (ARM::SPRRegClass.contains(Regs.front().first))
Kind = SPRRegisterList;
ARMOperand *Op = new ARMOperand(Kind);
Op->RegList.Registers = new SmallVector<unsigned, 32>();
for (SmallVectorImpl<std::pair<unsigned, SMLoc> >::const_iterator
I = Regs.begin(), E = Regs.end(); I != E; ++I)
Op->RegList.Registers->push_back(I->first);
std::sort(Op->RegList.Registers->begin(), Op->RegList.Registers->end());
Op->StartLoc = StartLoc;
Op->EndLoc = EndLoc;
return Op;
}
static ARMOperand *CreateImm(const MCExpr *Val, SMLoc S, SMLoc E) {
ARMOperand *Op = new ARMOperand(Immediate);
Op->Imm.Val = Val;
Op->StartLoc = S;
Op->EndLoc = E;
return Op;
}
static ARMOperand *CreateMem(unsigned BaseRegNum, bool OffsetIsReg,
const MCExpr *Offset, unsigned OffsetRegNum,
bool OffsetRegShifted, enum ShiftType ShiftType,
const MCExpr *ShiftAmount, bool Preindexed,
bool Postindexed, bool Negative, bool Writeback,
SMLoc S, SMLoc E) {
ARMOperand *Op = new ARMOperand(Memory);
Op->Mem.BaseRegNum = BaseRegNum;
Op->Mem.OffsetIsReg = OffsetIsReg;
Op->Mem.Offset = Offset;
Op->Mem.OffsetRegNum = OffsetRegNum;
Op->Mem.OffsetRegShifted = OffsetRegShifted;
Op->Mem.ShiftType = ShiftType;
Op->Mem.ShiftAmount = ShiftAmount;
Op->Mem.Preindexed = Preindexed;
Op->Mem.Postindexed = Postindexed;
Op->Mem.Negative = Negative;
Op->Mem.Writeback = Writeback;
Op->StartLoc = S;
Op->EndLoc = E;
return Op;
}
};
} // end anonymous namespace.
void ARMOperand::dump(raw_ostream &OS) const {
switch (Kind) {
case CondCode:
OS << ARMCondCodeToString(getCondCode());
break;
case Immediate:
getImm()->print(OS);
break;
case Memory:
OS << "<memory" << (!Mem.Writeback ? ">" : "!>");
break;
case Register:
OS << "<register " << getReg() << (!Reg.Writeback ? ">" : "!>");
break;
case RegisterList:
case DPRRegisterList:
case SPRRegisterList: {
OS << "<register_list ";
const SmallVectorImpl<unsigned> &RegList = getRegList();
for (SmallVectorImpl<unsigned>::const_iterator
I = RegList.begin(), E = RegList.end(); I != E; ) {
OS << *I;
if (++I < E) OS << ", ";
}
OS << ">";
break;
}
case Token:
OS << "'" << getToken() << "'";
break;
}
}
/// @name Auto-generated Match Functions
/// {
static unsigned MatchRegisterName(StringRef Name);
/// }
/// Try to parse a register name. The token must be an Identifier when called,
/// and if it is a register name the token is eaten and the register number is
/// returned. Otherwise return -1.
///
int ARMAsmParser::TryParseRegister() {
const AsmToken &Tok = Parser.getTok();
assert(Tok.is(AsmToken::Identifier) && "Token is not an Identifier");
// FIXME: Validate register for the current architecture; we have to do
// validation later, so maybe there is no need for this here.
unsigned RegNum = MatchRegisterName(Tok.getString());
if (RegNum == 0)
return -1;
Parser.Lex(); // Eat identifier token.
return RegNum;
}
/// Try to parse a register name. The token must be an Identifier when called,
/// and if it is a register name the token is eaten and the register number is
/// returned. Otherwise return -1.
///
/// TODO this is likely to change to allow different register types and or to
/// parse for a specific register type.
ARMOperand *ARMAsmParser::TryParseRegisterWithWriteBack() {
SMLoc S = Parser.getTok().getLoc();
int RegNo = TryParseRegister();
if (RegNo == -1)
return 0;
SMLoc E = Parser.getTok().getLoc();
bool Writeback = false;
const AsmToken &ExclaimTok = Parser.getTok();
if (ExclaimTok.is(AsmToken::Exclaim)) {
E = ExclaimTok.getLoc();
Writeback = true;
Parser.Lex(); // Eat exclaim token
}
return ARMOperand::CreateReg(RegNo, Writeback, S, E);
}
/// Parse a register list, return it if successful else return null. The first
/// token must be a '{' when called.
ARMOperand *ARMAsmParser::ParseRegisterList() {
assert(Parser.getTok().is(AsmToken::LCurly) &&
"Token is not a Left Curly Brace");
SMLoc S = Parser.getTok().getLoc();
// Read the rest of the registers in the list.
unsigned PrevRegNum = 0;
SmallVector<std::pair<unsigned, SMLoc>, 32> Registers;
do {
bool IsRange = Parser.getTok().is(AsmToken::Minus);
Parser.Lex(); // Eat non-identifier token.
const AsmToken &RegTok = Parser.getTok();
SMLoc RegLoc = RegTok.getLoc();
if (RegTok.isNot(AsmToken::Identifier)) {
Error(RegLoc, "register expected");
return 0;
}
int RegNum = TryParseRegister();
if (RegNum == -1) {
Error(RegLoc, "register expected");
return 0;
}
if (IsRange) {
int Reg = PrevRegNum;
do {
++Reg;
Registers.push_back(std::make_pair(Reg, RegLoc));
} while (Reg != RegNum);
} else {
Registers.push_back(std::make_pair(RegNum, RegLoc));
}
PrevRegNum = RegNum;
} while (Parser.getTok().is(AsmToken::Comma) ||
Parser.getTok().is(AsmToken::Minus));
// Process the right curly brace of the list.
const AsmToken &RCurlyTok = Parser.getTok();
if (RCurlyTok.isNot(AsmToken::RCurly)) {
Error(RCurlyTok.getLoc(), "'}' expected");
return 0;
}
SMLoc E = RCurlyTok.getLoc();
Parser.Lex(); // Eat right curly brace token.
// Verify the register list.
SmallVectorImpl<std::pair<unsigned, SMLoc> >::const_iterator
RI = Registers.begin(), RE = Registers.end();
DenseMap<unsigned, bool> RegMap;
RegMap[RI->first] = true;
unsigned HighRegNum = RI->first;
bool EmittedWarning = false;
for (++RI; RI != RE; ++RI) {
const std::pair<unsigned, SMLoc> &RegInfo = *RI;
unsigned Reg = RegInfo.first;
if (RegMap[Reg]) {
Error(RegInfo.second, "register duplicated in register list");
return 0;
}
if (!EmittedWarning && Reg < HighRegNum)
Warning(RegInfo.second,
"register not in ascending order in register list");
RegMap[Reg] = true;
HighRegNum = std::max(Reg, HighRegNum);
}
return ARMOperand::CreateRegList(Registers, S, E);
}
/// Parse an ARM memory expression, return false if successful else return true
/// or an error. The first token must be a '[' when called.
/// TODO Only preindexing and postindexing addressing are started, unindexed
/// with option, etc are still to do.
ARMOperand *ARMAsmParser::ParseMemory() {
SMLoc S, E;
assert(Parser.getTok().is(AsmToken::LBrac) &&
"Token is not a Left Bracket");
S = Parser.getTok().getLoc();
Parser.Lex(); // Eat left bracket token.
const AsmToken &BaseRegTok = Parser.getTok();
if (BaseRegTok.isNot(AsmToken::Identifier)) {
Error(BaseRegTok.getLoc(), "register expected");
return 0;
}
int BaseRegNum = TryParseRegister();
if (BaseRegNum == -1) {
Error(BaseRegTok.getLoc(), "register expected");
return 0;
}
bool Preindexed = false;
bool Postindexed = false;
bool OffsetIsReg = false;
bool Negative = false;
bool Writeback = false;
// First look for preindexed address forms, that is after the "[Rn" we now
// have to see if the next token is a comma.
const AsmToken &Tok = Parser.getTok();
if (Tok.is(AsmToken::Comma)) {
Preindexed = true;
Parser.Lex(); // Eat comma token.
int OffsetRegNum;
bool OffsetRegShifted;
enum ShiftType ShiftType;
const MCExpr *ShiftAmount;
const MCExpr *Offset;
if (ParseMemoryOffsetReg(Negative, OffsetRegShifted, ShiftType, ShiftAmount,
Offset, OffsetIsReg, OffsetRegNum, E))
return 0;
const AsmToken &RBracTok = Parser.getTok();
if (RBracTok.isNot(AsmToken::RBrac)) {
Error(RBracTok.getLoc(), "']' expected");
return 0;
}
E = RBracTok.getLoc();
Parser.Lex(); // Eat right bracket token.
const AsmToken &ExclaimTok = Parser.getTok();
if (ExclaimTok.is(AsmToken::Exclaim)) {
E = ExclaimTok.getLoc();
Writeback = true;
Parser.Lex(); // Eat exclaim token
}
return ARMOperand::CreateMem(BaseRegNum, OffsetIsReg, Offset, OffsetRegNum,
OffsetRegShifted, ShiftType, ShiftAmount,
Preindexed, Postindexed, Negative, Writeback,
S, E);
}
// The "[Rn" we have so far was not followed by a comma.
else if (Tok.is(AsmToken::RBrac)) {
// If there's anything other than the right brace, this is a post indexing
// addressing form.
E = Tok.getLoc();
Parser.Lex(); // Eat right bracket token.
int OffsetRegNum = 0;
bool OffsetRegShifted = false;
enum ShiftType ShiftType;
const MCExpr *ShiftAmount;
const MCExpr *Offset = 0;
const AsmToken &NextTok = Parser.getTok();
if (NextTok.isNot(AsmToken::EndOfStatement)) {
Postindexed = true;
Writeback = true;
if (NextTok.isNot(AsmToken::Comma)) {
Error(NextTok.getLoc(), "',' expected");
return 0;
}
Parser.Lex(); // Eat comma token.
if (ParseMemoryOffsetReg(Negative, OffsetRegShifted, ShiftType,
ShiftAmount, Offset, OffsetIsReg, OffsetRegNum,
E))
return 0;
}
return ARMOperand::CreateMem(BaseRegNum, OffsetIsReg, Offset, OffsetRegNum,
OffsetRegShifted, ShiftType, ShiftAmount,
Preindexed, Postindexed, Negative, Writeback,
S, E);
}
return 0;
}
/// Parse the offset of a memory operand after we have seen "[Rn," or "[Rn],"
/// we will parse the following (were +/- means that a plus or minus is
/// optional):
/// +/-Rm
/// +/-Rm, shift
/// #offset
/// we return false on success or an error otherwise.
bool ARMAsmParser::ParseMemoryOffsetReg(bool &Negative,
bool &OffsetRegShifted,
enum ShiftType &ShiftType,
const MCExpr *&ShiftAmount,
const MCExpr *&Offset,
bool &OffsetIsReg,
int &OffsetRegNum,
SMLoc &E) {
Negative = false;
OffsetRegShifted = false;
OffsetIsReg = false;
OffsetRegNum = -1;
const AsmToken &NextTok = Parser.getTok();
E = NextTok.getLoc();
if (NextTok.is(AsmToken::Plus))
Parser.Lex(); // Eat plus token.
else if (NextTok.is(AsmToken::Minus)) {
Negative = true;
Parser.Lex(); // Eat minus token
}
// See if there is a register following the "[Rn," or "[Rn]," we have so far.
const AsmToken &OffsetRegTok = Parser.getTok();
if (OffsetRegTok.is(AsmToken::Identifier)) {
SMLoc CurLoc = OffsetRegTok.getLoc();
OffsetRegNum = TryParseRegister();
if (OffsetRegNum != -1) {
OffsetIsReg = true;
E = CurLoc;
}
}
// If we parsed a register as the offset then there can be a shift after that.
if (OffsetRegNum != -1) {
// Look for a comma then a shift
const AsmToken &Tok = Parser.getTok();
if (Tok.is(AsmToken::Comma)) {
Parser.Lex(); // Eat comma token.
const AsmToken &Tok = Parser.getTok();
if (ParseShift(ShiftType, ShiftAmount, E))
return Error(Tok.getLoc(), "shift expected");
OffsetRegShifted = true;
}
}
else { // the "[Rn," or "[Rn,]" we have so far was not followed by "Rm"
// Look for #offset following the "[Rn," or "[Rn],"
const AsmToken &HashTok = Parser.getTok();
if (HashTok.isNot(AsmToken::Hash))
return Error(HashTok.getLoc(), "'#' expected");
Parser.Lex(); // Eat hash token.
if (getParser().ParseExpression(Offset))
return true;
E = SMLoc::getFromPointer(Parser.getTok().getLoc().getPointer() - 1);
}
return false;
}
/// ParseShift as one of these two:
/// ( lsl | lsr | asr | ror ) , # shift_amount
/// rrx
/// and returns true if it parses a shift otherwise it returns false.
bool ARMAsmParser::ParseShift(ShiftType &St, const MCExpr *&ShiftAmount,
SMLoc &E) {
const AsmToken &Tok = Parser.getTok();
if (Tok.isNot(AsmToken::Identifier))
return true;
StringRef ShiftName = Tok.getString();
if (ShiftName == "lsl" || ShiftName == "LSL")
St = Lsl;
else if (ShiftName == "lsr" || ShiftName == "LSR")
St = Lsr;
else if (ShiftName == "asr" || ShiftName == "ASR")
St = Asr;
else if (ShiftName == "ror" || ShiftName == "ROR")
St = Ror;
else if (ShiftName == "rrx" || ShiftName == "RRX")
St = Rrx;
else
return true;
Parser.Lex(); // Eat shift type token.
// Rrx stands alone.
if (St == Rrx)
return false;
// Otherwise, there must be a '#' and a shift amount.
const AsmToken &HashTok = Parser.getTok();
if (HashTok.isNot(AsmToken::Hash))
return Error(HashTok.getLoc(), "'#' expected");
Parser.Lex(); // Eat hash token.
if (getParser().ParseExpression(ShiftAmount))
return true;
return false;
}
/// Parse a arm instruction operand. For now this parses the operand regardless
/// of the mnemonic.
ARMOperand *ARMAsmParser::ParseOperand() {
SMLoc S, E;
switch (getLexer().getKind()) {
default:
Error(Parser.getTok().getLoc(), "unexpected token in operand");
return 0;
case AsmToken::Identifier:
if (ARMOperand *Op = TryParseRegisterWithWriteBack())
return Op;
// This was not a register so parse other operands that start with an
// identifier (like labels) as expressions and create them as immediates.
const MCExpr *IdVal;
S = Parser.getTok().getLoc();
if (getParser().ParseExpression(IdVal))
return 0;
E = SMLoc::getFromPointer(Parser.getTok().getLoc().getPointer() - 1);
return ARMOperand::CreateImm(IdVal, S, E);
case AsmToken::LBrac:
return ParseMemory();
case AsmToken::LCurly:
return ParseRegisterList();
case AsmToken::Hash:
// #42 -> immediate.
// TODO: ":lower16:" and ":upper16:" modifiers after # before immediate
S = Parser.getTok().getLoc();
Parser.Lex();
const MCExpr *ImmVal;
if (getParser().ParseExpression(ImmVal))
return 0;
E = SMLoc::getFromPointer(Parser.getTok().getLoc().getPointer() - 1);
return ARMOperand::CreateImm(ImmVal, S, E);
}
}
/// Parse an arm instruction mnemonic followed by its operands.
bool ARMAsmParser::ParseInstruction(StringRef Name, SMLoc NameLoc,
SmallVectorImpl<MCParsedAsmOperand*> &Operands) {
// Create the leading tokens for the mnemonic, split by '.' characters.
size_t Start = 0, Next = Name.find('.');
StringRef Head = Name.slice(Start, Next);
// Determine the predicate, if any.
//
// FIXME: We need a way to check whether a prefix supports predication,
// otherwise we will end up with an ambiguity for instructions that happen to
// end with a predicate name.
// FIXME: Likewise, some arithmetic instructions have an 's' prefix which
// indicates to update the condition codes. Those instructions have an
// additional immediate operand which encodes the prefix as reg0 or CPSR.
// Just checking for a suffix of 's' definitely creates ambiguities; e.g,
// the SMMLS instruction.
unsigned CC = StringSwitch<unsigned>(Head.substr(Head.size()-2))
.Case("eq", ARMCC::EQ)
.Case("ne", ARMCC::NE)
.Case("hs", ARMCC::HS)
.Case("lo", ARMCC::LO)
.Case("mi", ARMCC::MI)
.Case("pl", ARMCC::PL)
.Case("vs", ARMCC::VS)
.Case("vc", ARMCC::VC)
.Case("hi", ARMCC::HI)
.Case("ls", ARMCC::LS)
.Case("ge", ARMCC::GE)
.Case("lt", ARMCC::LT)
.Case("gt", ARMCC::GT)
.Case("le", ARMCC::LE)
.Case("al", ARMCC::AL)
.Default(~0U);
if (CC == ~0U ||
(CC == ARMCC::LS && (Head == "vmls" || Head == "vnmls"))) {
CC = ARMCC::AL;
} else {
Head = Head.slice(0, Head.size() - 2);
}
Operands.push_back(ARMOperand::CreateToken(Head, NameLoc));
// FIXME: Should only add this operand for predicated instructions
Operands.push_back(ARMOperand::CreateCondCode(ARMCC::CondCodes(CC), NameLoc));
// Add the remaining tokens in the mnemonic.
while (Next != StringRef::npos) {
Start = Next;
Next = Name.find('.', Start + 1);
Head = Name.slice(Start, Next);
Operands.push_back(ARMOperand::CreateToken(Head, NameLoc));
}
// Read the remaining operands.
if (getLexer().isNot(AsmToken::EndOfStatement)) {
// Read the first operand.
if (ARMOperand *Op = ParseOperand())
Operands.push_back(Op);
else {
Parser.EatToEndOfStatement();
return true;
}
while (getLexer().is(AsmToken::Comma)) {
Parser.Lex(); // Eat the comma.
// Parse and remember the operand.
if (ARMOperand *Op = ParseOperand())
Operands.push_back(Op);
else {
Parser.EatToEndOfStatement();
return true;
}
}
}
if (getLexer().isNot(AsmToken::EndOfStatement)) {
Parser.EatToEndOfStatement();
return TokError("unexpected token in argument list");
}
Parser.Lex(); // Consume the EndOfStatement
return false;
}
bool ARMAsmParser::
MatchAndEmitInstruction(SMLoc IDLoc,
SmallVectorImpl<MCParsedAsmOperand*> &Operands,
MCStreamer &Out) {
MCInst Inst;
unsigned ErrorInfo;
switch (MatchInstructionImpl(Operands, Inst, ErrorInfo)) {
case Match_Success:
Out.EmitInstruction(Inst);
return false;
case Match_MissingFeature:
Error(IDLoc, "instruction requires a CPU feature not currently enabled");
return true;
case Match_InvalidOperand: {
SMLoc ErrorLoc = IDLoc;
if (ErrorInfo != ~0U) {
if (ErrorInfo >= Operands.size())
return Error(IDLoc, "too few operands for instruction");
ErrorLoc = ((ARMOperand*)Operands[ErrorInfo])->getStartLoc();
if (ErrorLoc == SMLoc()) ErrorLoc = IDLoc;
}
return Error(ErrorLoc, "invalid operand for instruction");
}
case Match_MnemonicFail:
return Error(IDLoc, "unrecognized instruction mnemonic");
}
llvm_unreachable("Implement any new match types added!");
return true;
}
/// ParseDirective parses the arm specific directives
bool ARMAsmParser::ParseDirective(AsmToken DirectiveID) {
StringRef IDVal = DirectiveID.getIdentifier();
if (IDVal == ".word")
return ParseDirectiveWord(4, DirectiveID.getLoc());
else if (IDVal == ".thumb")
return ParseDirectiveThumb(DirectiveID.getLoc());
else if (IDVal == ".thumb_func")
return ParseDirectiveThumbFunc(DirectiveID.getLoc());
else if (IDVal == ".code")
return ParseDirectiveCode(DirectiveID.getLoc());
else if (IDVal == ".syntax")
return ParseDirectiveSyntax(DirectiveID.getLoc());
return true;
}
/// ParseDirectiveWord
/// ::= .word [ expression (, expression)* ]
bool ARMAsmParser::ParseDirectiveWord(unsigned Size, SMLoc L) {
if (getLexer().isNot(AsmToken::EndOfStatement)) {
for (;;) {
const MCExpr *Value;
if (getParser().ParseExpression(Value))
return true;
getParser().getStreamer().EmitValue(Value, Size, 0/*addrspace*/);
if (getLexer().is(AsmToken::EndOfStatement))
break;
// FIXME: Improve diagnostic.
if (getLexer().isNot(AsmToken::Comma))
return Error(L, "unexpected token in directive");
Parser.Lex();
}
}
Parser.Lex();
return false;
}
/// ParseDirectiveThumb
/// ::= .thumb
bool ARMAsmParser::ParseDirectiveThumb(SMLoc L) {
if (getLexer().isNot(AsmToken::EndOfStatement))
return Error(L, "unexpected token in directive");
Parser.Lex();
// TODO: set thumb mode
// TODO: tell the MC streamer the mode
// getParser().getStreamer().Emit???();
return false;
}
/// ParseDirectiveThumbFunc
/// ::= .thumbfunc symbol_name
bool ARMAsmParser::ParseDirectiveThumbFunc(SMLoc L) {
const AsmToken &Tok = Parser.getTok();
if (Tok.isNot(AsmToken::Identifier) && Tok.isNot(AsmToken::String))
return Error(L, "unexpected token in .thumb_func directive");
StringRef Name = Tok.getString();
Parser.Lex(); // Consume the identifier token.
if (getLexer().isNot(AsmToken::EndOfStatement))
return Error(L, "unexpected token in directive");
Parser.Lex();
// Mark symbol as a thumb symbol.
MCSymbol *Func = getParser().getContext().GetOrCreateSymbol(Name);
getParser().getStreamer().EmitThumbFunc(Func);
return false;
}
/// ParseDirectiveSyntax
/// ::= .syntax unified | divided
bool ARMAsmParser::ParseDirectiveSyntax(SMLoc L) {
const AsmToken &Tok = Parser.getTok();
if (Tok.isNot(AsmToken::Identifier))
return Error(L, "unexpected token in .syntax directive");
StringRef Mode = Tok.getString();
if (Mode == "unified" || Mode == "UNIFIED")
Parser.Lex();
else if (Mode == "divided" || Mode == "DIVIDED")
Parser.Lex();
else
return Error(L, "unrecognized syntax mode in .syntax directive");
if (getLexer().isNot(AsmToken::EndOfStatement))
return Error(Parser.getTok().getLoc(), "unexpected token in directive");
Parser.Lex();
// TODO tell the MC streamer the mode
// getParser().getStreamer().Emit???();
return false;
}
/// ParseDirectiveCode
/// ::= .code 16 | 32
bool ARMAsmParser::ParseDirectiveCode(SMLoc L) {
const AsmToken &Tok = Parser.getTok();
if (Tok.isNot(AsmToken::Integer))
return Error(L, "unexpected token in .code directive");
int64_t Val = Parser.getTok().getIntVal();
if (Val == 16)
Parser.Lex();
else if (Val == 32)
Parser.Lex();
else
return Error(L, "invalid operand to .code directive");
if (getLexer().isNot(AsmToken::EndOfStatement))
return Error(Parser.getTok().getLoc(), "unexpected token in directive");
Parser.Lex();
if (Val == 16)
getParser().getStreamer().EmitAssemblerFlag(MCAF_Code16);
else
getParser().getStreamer().EmitAssemblerFlag(MCAF_Code32);
return false;
}
extern "C" void LLVMInitializeARMAsmLexer();
/// Force static initialization.
extern "C" void LLVMInitializeARMAsmParser() {
RegisterAsmParser<ARMAsmParser> X(TheARMTarget);
RegisterAsmParser<ARMAsmParser> Y(TheThumbTarget);
LLVMInitializeARMAsmLexer();
}
#define GET_REGISTER_MATCHER
#define GET_MATCHER_IMPLEMENTATION
#include "ARMGenAsmMatcher.inc"
|