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
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
|
//===- lib/MC/MCAssembler.cpp - Assembler Backend Implementation ----------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
#include "llvm/MC/MCAssembler.h"
#include "llvm/ADT/ArrayRef.h"
#include "llvm/ADT/SmallString.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/ADT/Twine.h"
#include "llvm/MC/MCAsmBackend.h"
#include "llvm/MC/MCAsmInfo.h"
#include "llvm/MC/MCCodeEmitter.h"
#include "llvm/MC/MCCodeView.h"
#include "llvm/MC/MCContext.h"
#include "llvm/MC/MCDwarf.h"
#include "llvm/MC/MCExpr.h"
#include "llvm/MC/MCFixup.h"
#include "llvm/MC/MCInst.h"
#include "llvm/MC/MCObjectWriter.h"
#include "llvm/MC/MCSection.h"
#include "llvm/MC/MCSymbol.h"
#include "llvm/MC/MCValue.h"
#include "llvm/Support/Alignment.h"
#include "llvm/Support/Casting.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/EndianStream.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/LEB128.h"
#include "llvm/Support/raw_ostream.h"
#include <cassert>
#include <cstdint>
#include <tuple>
#include <utility>
using namespace llvm;
namespace llvm {
class MCSubtargetInfo;
}
#define DEBUG_TYPE "assembler"
namespace {
namespace stats {
STATISTIC(EmittedFragments, "Number of emitted assembler fragments - total");
STATISTIC(EmittedRelaxableFragments,
"Number of emitted assembler fragments - relaxable");
STATISTIC(EmittedDataFragments,
"Number of emitted assembler fragments - data");
STATISTIC(EmittedAlignFragments,
"Number of emitted assembler fragments - align");
STATISTIC(EmittedFillFragments,
"Number of emitted assembler fragments - fill");
STATISTIC(EmittedNopsFragments, "Number of emitted assembler fragments - nops");
STATISTIC(EmittedOrgFragments, "Number of emitted assembler fragments - org");
STATISTIC(evaluateFixup, "Number of evaluated fixups");
STATISTIC(ObjectBytes, "Number of emitted object file bytes");
STATISTIC(RelaxationSteps, "Number of assembler layout and relaxation steps");
STATISTIC(RelaxedInstructions, "Number of relaxed instructions");
} // end namespace stats
} // end anonymous namespace
// FIXME FIXME FIXME: There are number of places in this file where we convert
// what is a 64-bit assembler value used for computation into a value in the
// object file, which may truncate it. We should detect that truncation where
// invalid and report errors back.
/* *** */
MCAssembler::MCAssembler(MCContext &Context,
std::unique_ptr<MCAsmBackend> Backend,
std::unique_ptr<MCCodeEmitter> Emitter,
std::unique_ptr<MCObjectWriter> Writer)
: Context(Context), Backend(std::move(Backend)),
Emitter(std::move(Emitter)), Writer(std::move(Writer)) {
if (this->Backend)
this->Backend->setAssembler(this);
if (this->Writer)
this->Writer->setAssembler(this);
}
void MCAssembler::reset() {
HasLayout = false;
HasFinalLayout = false;
RelaxAll = false;
Sections.clear();
Symbols.clear();
ThumbFuncs.clear();
// reset objects owned by us
if (getBackendPtr())
getBackendPtr()->reset();
if (getEmitterPtr())
getEmitterPtr()->reset();
if (Writer)
Writer->reset();
}
bool MCAssembler::registerSection(MCSection &Section) {
if (Section.isRegistered())
return false;
assert(Section.curFragList()->Head && "allocInitialFragment not called");
Sections.push_back(&Section);
Section.setIsRegistered(true);
return true;
}
bool MCAssembler::isThumbFunc(const MCSymbol *Symbol) const {
if (ThumbFuncs.count(Symbol))
return true;
if (!Symbol->isVariable())
return false;
const MCExpr *Expr = Symbol->getVariableValue();
MCValue V;
if (!Expr->evaluateAsRelocatable(V, nullptr))
return false;
if (V.getSubSym() || V.getSpecifier())
return false;
auto *Sym = V.getAddSym();
if (!Sym || V.getSpecifier())
return false;
if (!isThumbFunc(Sym))
return false;
ThumbFuncs.insert(Symbol); // Cache it.
return true;
}
bool MCAssembler::evaluateFixup(const MCFragment &F, MCFixup &Fixup,
MCValue &Target, uint64_t &Value,
bool RecordReloc,
MutableArrayRef<char> Contents) const {
++stats::evaluateFixup;
// FIXME: This code has some duplication with recordRelocation. We should
// probably merge the two into a single callback that tries to evaluate a
// fixup and records a relocation if one is needed.
// On error claim to have completely evaluated the fixup, to prevent any
// further processing from being done.
const MCExpr *Expr = Fixup.getValue();
Value = 0;
if (!Expr->evaluateAsRelocatable(Target, this)) {
reportError(Fixup.getLoc(), "expected relocatable expression");
return true;
}
bool IsResolved = false;
if (auto State = getBackend().evaluateFixup(F, Fixup, Target, Value)) {
IsResolved = *State;
} else {
const MCSymbol *Add = Target.getAddSym();
const MCSymbol *Sub = Target.getSubSym();
Value += Target.getConstant();
if (Add && Add->isDefined())
Value += getSymbolOffset(*Add);
if (Sub && Sub->isDefined())
Value -= getSymbolOffset(*Sub);
if (Fixup.isPCRel()) {
Value -= getFragmentOffset(F) + Fixup.getOffset();
if (Add && !Sub && !Add->isUndefined() && !Add->isAbsolute()) {
IsResolved = getWriter().isSymbolRefDifferenceFullyResolvedImpl(
*Add, F, false, true);
}
} else {
IsResolved = Target.isAbsolute();
}
}
if (!RecordReloc)
return IsResolved;
if (IsResolved && mc::isRelocRelocation(Fixup.getKind()))
IsResolved = false;
getBackend().applyFixup(F, Fixup, Target, Contents, Value, IsResolved);
return true;
}
uint64_t MCAssembler::computeFragmentSize(const MCFragment &F) const {
assert(getBackendPtr() && "Requires assembler backend");
switch (F.getKind()) {
case MCFragment::FT_Data:
case MCFragment::FT_Relaxable:
case MCFragment::FT_LEB:
case MCFragment::FT_Dwarf:
case MCFragment::FT_DwarfFrame:
case MCFragment::FT_CVInlineLines:
case MCFragment::FT_CVDefRange:
case MCFragment::FT_PseudoProbe:
return F.getSize();
case MCFragment::FT_Fill: {
auto &FF = cast<MCFillFragment>(F);
int64_t NumValues = 0;
if (!FF.getNumValues().evaluateKnownAbsolute(NumValues, *this)) {
recordError(FF.getLoc(), "expected assembly-time absolute expression");
return 0;
}
int64_t Size = NumValues * FF.getValueSize();
if (Size < 0) {
recordError(FF.getLoc(), "invalid number of bytes");
return 0;
}
return Size;
}
case MCFragment::FT_Nops:
return cast<MCNopsFragment>(F).getNumBytes();
case MCFragment::FT_BoundaryAlign:
return cast<MCBoundaryAlignFragment>(F).getSize();
case MCFragment::FT_SymbolId:
return 4;
case MCFragment::FT_Align: {
const MCAlignFragment &AF = cast<MCAlignFragment>(F);
unsigned Offset = getFragmentOffset(AF);
unsigned Size = offsetToAlignment(Offset, AF.getAlignment());
// Insert extra Nops for code alignment if the target define
// shouldInsertExtraNopBytesForCodeAlign target hook.
if (AF.getParent()->useCodeAlign() && AF.hasEmitNops() &&
getBackend().shouldInsertExtraNopBytesForCodeAlign(AF, Size))
return Size;
// If we are padding with nops, force the padding to be larger than the
// minimum nop size.
if (Size > 0 && AF.hasEmitNops()) {
while (Size % getBackend().getMinimumNopSize())
Size += AF.getAlignment().value();
}
if (Size > AF.getMaxBytesToEmit())
return 0;
return Size;
}
case MCFragment::FT_Org: {
const MCOrgFragment &OF = cast<MCOrgFragment>(F);
MCValue Value;
if (!OF.getOffset().evaluateAsValue(Value, *this)) {
recordError(OF.getLoc(), "expected assembly-time absolute expression");
return 0;
}
uint64_t FragmentOffset = getFragmentOffset(OF);
int64_t TargetLocation = Value.getConstant();
if (const auto *SA = Value.getAddSym()) {
uint64_t Val;
if (!getSymbolOffset(*SA, Val)) {
recordError(OF.getLoc(), "expected absolute expression");
return 0;
}
TargetLocation += Val;
}
int64_t Size = TargetLocation - FragmentOffset;
if (Size < 0 || Size >= 0x40000000) {
recordError(OF.getLoc(), "invalid .org offset '" + Twine(TargetLocation) +
"' (at offset '" + Twine(FragmentOffset) +
"')");
return 0;
}
return Size;
}
}
llvm_unreachable("invalid fragment kind");
}
// Simple getSymbolOffset helper for the non-variable case.
static bool getLabelOffset(const MCAssembler &Asm, const MCSymbol &S,
bool ReportError, uint64_t &Val) {
if (!S.getFragment()) {
if (ReportError)
reportFatalUsageError("cannot evaluate undefined symbol '" + S.getName() +
"'");
return false;
}
Val = Asm.getFragmentOffset(*S.getFragment()) + S.getOffset();
return true;
}
static bool getSymbolOffsetImpl(const MCAssembler &Asm, const MCSymbol &S,
bool ReportError, uint64_t &Val) {
if (!S.isVariable())
return getLabelOffset(Asm, S, ReportError, Val);
// If SD is a variable, evaluate it.
MCValue Target;
if (!S.getVariableValue()->evaluateAsValue(Target, Asm))
reportFatalUsageError("cannot evaluate equated symbol '" + S.getName() +
"'");
uint64_t Offset = Target.getConstant();
const MCSymbol *A = Target.getAddSym();
if (A) {
uint64_t ValA;
// FIXME: On most platforms, `Target`'s component symbols are labels from
// having been simplified during evaluation, but on Mach-O they can be
// variables due to PR19203. This, and the line below for `B` can be
// restored to call `getLabelOffset` when PR19203 is fixed.
if (!getSymbolOffsetImpl(Asm, *A, ReportError, ValA))
return false;
Offset += ValA;
}
const MCSymbol *B = Target.getSubSym();
if (B) {
uint64_t ValB;
if (!getSymbolOffsetImpl(Asm, *B, ReportError, ValB))
return false;
Offset -= ValB;
}
Val = Offset;
return true;
}
bool MCAssembler::getSymbolOffset(const MCSymbol &S, uint64_t &Val) const {
return getSymbolOffsetImpl(*this, S, false, Val);
}
uint64_t MCAssembler::getSymbolOffset(const MCSymbol &S) const {
uint64_t Val;
getSymbolOffsetImpl(*this, S, true, Val);
return Val;
}
const MCSymbol *MCAssembler::getBaseSymbol(const MCSymbol &Symbol) const {
assert(HasLayout);
if (!Symbol.isVariable())
return &Symbol;
const MCExpr *Expr = Symbol.getVariableValue();
MCValue Value;
if (!Expr->evaluateAsValue(Value, *this)) {
reportError(Expr->getLoc(), "expression could not be evaluated");
return nullptr;
}
const MCSymbol *SymB = Value.getSubSym();
if (SymB) {
reportError(Expr->getLoc(),
Twine("symbol '") + SymB->getName() +
"' could not be evaluated in a subtraction expression");
return nullptr;
}
const MCSymbol *A = Value.getAddSym();
if (!A)
return nullptr;
const MCSymbol &ASym = *A;
if (ASym.isCommon()) {
reportError(Expr->getLoc(), "Common symbol '" + ASym.getName() +
"' cannot be used in assignment expr");
return nullptr;
}
return &ASym;
}
uint64_t MCAssembler::getSectionAddressSize(const MCSection &Sec) const {
assert(HasLayout);
// The size is the last fragment's end offset.
const MCFragment &F = *Sec.curFragList()->Tail;
return getFragmentOffset(F) + computeFragmentSize(F);
}
uint64_t MCAssembler::getSectionFileSize(const MCSection &Sec) const {
// Virtual sections have no file size.
if (Sec.isVirtualSection())
return 0;
return getSectionAddressSize(Sec);
}
bool MCAssembler::registerSymbol(const MCSymbol &Symbol) {
bool Changed = !Symbol.isRegistered();
if (Changed) {
Symbol.setIsRegistered(true);
Symbols.push_back(&Symbol);
}
return Changed;
}
/// Write the fragment \p F to the output file.
static void writeFragment(raw_ostream &OS, const MCAssembler &Asm,
const MCFragment &F) {
// FIXME: Embed in fragments instead?
uint64_t FragmentSize = Asm.computeFragmentSize(F);
llvm::endianness Endian = Asm.getBackend().Endian;
// This variable (and its dummy usage) is to participate in the assert at
// the end of the function.
uint64_t Start = OS.tell();
(void) Start;
++stats::EmittedFragments;
switch (F.getKind()) {
case MCFragment::FT_Data:
case MCFragment::FT_Relaxable:
case MCFragment::FT_LEB:
case MCFragment::FT_Dwarf:
case MCFragment::FT_DwarfFrame:
case MCFragment::FT_CVInlineLines:
case MCFragment::FT_CVDefRange:
case MCFragment::FT_PseudoProbe: {
if (F.getKind() == MCFragment::FT_Data)
++stats::EmittedDataFragments;
else if (F.getKind() == MCFragment::FT_Relaxable)
++stats::EmittedRelaxableFragments;
const auto &EF = cast<MCFragment>(F);
OS << StringRef(EF.getContents().data(), EF.getContents().size());
OS << StringRef(EF.getVarContents().data(), EF.getVarContents().size());
break;
}
case MCFragment::FT_Align: {
++stats::EmittedAlignFragments;
const MCAlignFragment &AF = cast<MCAlignFragment>(F);
assert(AF.getFillLen() && "Invalid virtual align in concrete fragment!");
uint64_t Count = FragmentSize / AF.getFillLen();
assert(FragmentSize % AF.getFillLen() == 0 &&
"computeFragmentSize computed size is incorrect");
// See if we are aligning with nops, and if so do that first to try to fill
// the Count bytes. Then if that did not fill any bytes or there are any
// bytes left to fill use the Value and ValueSize to fill the rest.
// If we are aligning with nops, ask that target to emit the right data.
if (AF.hasEmitNops()) {
if (!Asm.getBackend().writeNopData(OS, Count, AF.getSubtargetInfo()))
report_fatal_error("unable to write nop sequence of " +
Twine(Count) + " bytes");
break;
}
// Otherwise, write out in multiples of the value size.
for (uint64_t i = 0; i != Count; ++i) {
switch (AF.getFillLen()) {
default: llvm_unreachable("Invalid size!");
case 1:
OS << char(AF.getFill());
break;
case 2:
support::endian::write<uint16_t>(OS, AF.getFill(), Endian);
break;
case 4:
support::endian::write<uint32_t>(OS, AF.getFill(), Endian);
break;
case 8:
support::endian::write<uint64_t>(OS, AF.getFill(), Endian);
break;
}
}
break;
}
case MCFragment::FT_Fill: {
++stats::EmittedFillFragments;
const MCFillFragment &FF = cast<MCFillFragment>(F);
uint64_t V = FF.getValue();
unsigned VSize = FF.getValueSize();
const unsigned MaxChunkSize = 16;
char Data[MaxChunkSize];
assert(0 < VSize && VSize <= MaxChunkSize && "Illegal fragment fill size");
// Duplicate V into Data as byte vector to reduce number of
// writes done. As such, do endian conversion here.
for (unsigned I = 0; I != VSize; ++I) {
unsigned index = Endian == llvm::endianness::little ? I : (VSize - I - 1);
Data[I] = uint8_t(V >> (index * 8));
}
for (unsigned I = VSize; I < MaxChunkSize; ++I)
Data[I] = Data[I - VSize];
// Set to largest multiple of VSize in Data.
const unsigned NumPerChunk = MaxChunkSize / VSize;
// Set ChunkSize to largest multiple of VSize in Data
const unsigned ChunkSize = VSize * NumPerChunk;
// Do copies by chunk.
StringRef Ref(Data, ChunkSize);
for (uint64_t I = 0, E = FragmentSize / ChunkSize; I != E; ++I)
OS << Ref;
// do remainder if needed.
unsigned TrailingCount = FragmentSize % ChunkSize;
if (TrailingCount)
OS.write(Data, TrailingCount);
break;
}
case MCFragment::FT_Nops: {
++stats::EmittedNopsFragments;
const MCNopsFragment &NF = cast<MCNopsFragment>(F);
int64_t NumBytes = NF.getNumBytes();
int64_t ControlledNopLength = NF.getControlledNopLength();
int64_t MaximumNopLength =
Asm.getBackend().getMaximumNopSize(*NF.getSubtargetInfo());
assert(NumBytes > 0 && "Expected positive NOPs fragment size");
assert(ControlledNopLength >= 0 && "Expected non-negative NOP size");
if (ControlledNopLength > MaximumNopLength) {
Asm.reportError(NF.getLoc(), "illegal NOP size " +
std::to_string(ControlledNopLength) +
". (expected within [0, " +
std::to_string(MaximumNopLength) + "])");
// Clamp the NOP length as reportError does not stop the execution
// immediately.
ControlledNopLength = MaximumNopLength;
}
// Use maximum value if the size of each NOP is not specified
if (!ControlledNopLength)
ControlledNopLength = MaximumNopLength;
while (NumBytes) {
uint64_t NumBytesToEmit =
(uint64_t)std::min(NumBytes, ControlledNopLength);
assert(NumBytesToEmit && "try to emit empty NOP instruction");
if (!Asm.getBackend().writeNopData(OS, NumBytesToEmit,
NF.getSubtargetInfo())) {
report_fatal_error("unable to write nop sequence of the remaining " +
Twine(NumBytesToEmit) + " bytes");
break;
}
NumBytes -= NumBytesToEmit;
}
break;
}
case MCFragment::FT_BoundaryAlign: {
const MCBoundaryAlignFragment &BF = cast<MCBoundaryAlignFragment>(F);
if (!Asm.getBackend().writeNopData(OS, FragmentSize, BF.getSubtargetInfo()))
report_fatal_error("unable to write nop sequence of " +
Twine(FragmentSize) + " bytes");
break;
}
case MCFragment::FT_SymbolId: {
const MCSymbolIdFragment &SF = cast<MCSymbolIdFragment>(F);
support::endian::write<uint32_t>(OS, SF.getSymbol()->getIndex(), Endian);
break;
}
case MCFragment::FT_Org: {
++stats::EmittedOrgFragments;
const MCOrgFragment &OF = cast<MCOrgFragment>(F);
for (uint64_t i = 0, e = FragmentSize; i != e; ++i)
OS << char(OF.getValue());
break;
}
}
assert(OS.tell() - Start == FragmentSize &&
"The stream should advance by fragment size");
}
void MCAssembler::writeSectionData(raw_ostream &OS,
const MCSection *Sec) const {
assert(getBackendPtr() && "Expected assembler backend");
// Ignore virtual sections.
if (Sec->isVirtualSection()) {
assert(getSectionFileSize(*Sec) == 0 && "Invalid size for section!");
// Check that contents are only things legal inside a virtual section.
for (const MCFragment &F : *Sec) {
switch (F.getKind()) {
default: llvm_unreachable("Invalid fragment in virtual section!");
case MCFragment::FT_Data: {
// Check that we aren't trying to write a non-zero contents (or fixups)
// into a virtual section. This is to support clients which use standard
// directives to fill the contents of virtual sections.
if (F.getFixups().size() || F.getVarFixups().size())
reportError(SMLoc(), Sec->getVirtualSectionKind() + " section '" +
Sec->getName() + "' cannot have fixups");
for (char C : F.getContents())
if (C) {
reportError(SMLoc(), Sec->getVirtualSectionKind() + " section '" +
Sec->getName() +
"' cannot have non-zero initializers");
break;
}
break;
}
case MCFragment::FT_Align:
// Check that we aren't trying to write a non-zero value into a virtual
// section.
assert((cast<MCAlignFragment>(F).getFillLen() == 0 ||
cast<MCAlignFragment>(F).getFill() == 0) &&
"Invalid align in virtual section!");
break;
case MCFragment::FT_Fill:
assert((cast<MCFillFragment>(F).getValue() == 0) &&
"Invalid fill in virtual section!");
break;
case MCFragment::FT_Org:
break;
}
}
return;
}
uint64_t Start = OS.tell();
(void)Start;
for (const MCFragment &F : *Sec)
writeFragment(OS, *this, F);
flushPendingErrors();
assert(getContext().hadError() ||
OS.tell() - Start == getSectionAddressSize(*Sec));
}
void MCAssembler::layout() {
assert(getBackendPtr() && "Expected assembler backend");
DEBUG_WITH_TYPE("mc-dump-pre", {
errs() << "assembler backend - pre-layout\n--\n";
dump();
});
// Assign section ordinals.
unsigned SectionIndex = 0;
for (MCSection &Sec : *this) {
Sec.setOrdinal(SectionIndex++);
// Chain together fragments from all subsections.
if (Sec.Subsections.size() > 1) {
MCFragment Dummy;
MCFragment *Tail = &Dummy;
for (auto &[_, List] : Sec.Subsections) {
assert(List.Head);
Tail->Next = List.Head;
Tail = List.Tail;
}
Sec.Subsections.clear();
Sec.Subsections.push_back({0u, {Dummy.getNext(), Tail}});
Sec.CurFragList = &Sec.Subsections[0].second;
unsigned FragmentIndex = 0;
for (MCFragment &Frag : Sec)
Frag.setLayoutOrder(FragmentIndex++);
}
}
// Layout until everything fits.
this->HasLayout = true;
for (MCSection &Sec : *this)
layoutSection(Sec);
unsigned FirstStable = Sections.size();
while ((FirstStable = relaxOnce(FirstStable)) > 0)
if (getContext().hadError())
return;
// Some targets might want to adjust fragment offsets. If so, perform another
// layout iteration.
if (getBackend().finishLayout(*this))
for (MCSection &Sec : *this)
layoutSection(Sec);
flushPendingErrors();
DEBUG_WITH_TYPE("mc-dump", {
errs() << "assembler backend - final-layout\n--\n";
dump(); });
// Allow the object writer a chance to perform post-layout binding (for
// example, to set the index fields in the symbol data).
getWriter().executePostLayoutBinding();
// Fragment sizes are finalized. For RISC-V linker relaxation, this flag
// helps check whether a PC-relative fixup is fully resolved.
this->HasFinalLayout = true;
// Evaluate and apply the fixups, generating relocation entries as necessary.
for (MCSection &Sec : *this) {
for (MCFragment &F : Sec) {
// Process fragments with fixups here.
if (F.isEncoded()) {
auto Contents = F.getContents();
for (MCFixup &Fixup : F.getFixups()) {
uint64_t FixedValue;
MCValue Target;
evaluateFixup(F, Fixup, Target, FixedValue,
/*RecordReloc=*/true, Contents);
}
// In the variable part, fixup offsets are relative to the fixed part's
// start. Extend the variable contents to the left to account for the
// fixed part size.
Contents = MutableArrayRef(F.getParent()->ContentStorage)
.slice(F.VarContentStart - Contents.size(), F.getSize());
for (MCFixup &Fixup : F.getVarFixups()) {
uint64_t FixedValue;
MCValue Target;
evaluateFixup(F, Fixup, Target, FixedValue,
/*RecordReloc=*/true, Contents);
}
} else if (auto *AF = dyn_cast<MCAlignFragment>(&F)) {
// For RISC-V linker relaxation, an alignment relocation might be
// needed.
if (AF->hasEmitNops())
getBackend().shouldInsertFixupForCodeAlign(*this, *AF);
}
}
}
}
void MCAssembler::Finish() {
layout();
// Write the object file.
stats::ObjectBytes += getWriter().writeObject();
HasLayout = false;
assert(PendingErrors.empty());
}
bool MCAssembler::fixupNeedsRelaxation(const MCFragment &F,
const MCFixup &Fixup) const {
assert(getBackendPtr() && "Expected assembler backend");
MCValue Target;
uint64_t Value;
bool Resolved = evaluateFixup(F, const_cast<MCFixup &>(Fixup), Target, Value,
/*RecordReloc=*/false, {});
return getBackend().fixupNeedsRelaxationAdvanced(F, Fixup, Target, Value,
Resolved);
}
bool MCAssembler::relaxInstruction(MCFragment &F) {
assert(getEmitterPtr() &&
"Expected CodeEmitter defined for relaxInstruction");
// If this inst doesn't ever need relaxation, ignore it. This occurs when we
// are intentionally pushing out inst fragments, or because we relaxed a
// previous instruction to one that doesn't need relaxation.
if (!getBackend().mayNeedRelaxation(F.getOpcode(), F.getOperands(),
*F.getSubtargetInfo()))
return false;
bool DoRelax = false;
for (const MCFixup &Fixup : F.getVarFixups())
if ((DoRelax = fixupNeedsRelaxation(F, Fixup)))
break;
if (!DoRelax)
return false;
++stats::RelaxedInstructions;
// TODO Refactor relaxInstruction to accept MCFragment and remove
// `setInst`.
MCInst Relaxed = F.getInst();
getBackend().relaxInstruction(Relaxed, *F.getSubtargetInfo());
// Encode the new instruction.
F.setInst(Relaxed);
SmallVector<char, 16> Data;
SmallVector<MCFixup, 1> Fixups;
getEmitter().encodeInstruction(Relaxed, Data, Fixups, *F.getSubtargetInfo());
F.setVarContents(Data);
F.setVarFixups(Fixups);
return true;
}
bool MCAssembler::relaxLEB(MCFragment &F) {
const unsigned OldSize = F.getVarSize();
unsigned PadTo = OldSize;
int64_t Value;
F.clearVarFixups();
// Use evaluateKnownAbsolute for Mach-O as a hack: .subsections_via_symbols
// requires that .uleb128 A-B is foldable where A and B reside in different
// fragments. This is used by __gcc_except_table.
bool Abs = getWriter().getSubsectionsViaSymbols()
? F.getLEBValue().evaluateKnownAbsolute(Value, *this)
: F.getLEBValue().evaluateAsAbsolute(Value, *this);
if (!Abs) {
bool Relaxed, UseZeroPad;
std::tie(Relaxed, UseZeroPad) = getBackend().relaxLEB128(F, Value);
if (!Relaxed) {
reportError(F.getLEBValue().getLoc(),
Twine(F.isLEBSigned() ? ".s" : ".u") +
"leb128 expression is not absolute");
F.setLEBValue(MCConstantExpr::create(0, Context));
}
uint8_t Tmp[10]; // maximum size: ceil(64/7)
PadTo = std::max(PadTo, encodeULEB128(uint64_t(Value), Tmp));
if (UseZeroPad)
Value = 0;
}
uint8_t Data[16];
size_t Size = 0;
// The compiler can generate EH table assembly that is impossible to assemble
// without either adding padding to an LEB fragment or adding extra padding
// to a later alignment fragment. To accommodate such tables, relaxation can
// only increase an LEB fragment size here, not decrease it. See PR35809.
if (F.isLEBSigned())
Size = encodeSLEB128(Value, Data, PadTo);
else
Size = encodeULEB128(Value, Data, PadTo);
F.setVarContents({reinterpret_cast<char *>(Data), Size});
return OldSize != Size;
}
/// Check if the branch crosses the boundary.
///
/// \param StartAddr start address of the fused/unfused branch.
/// \param Size size of the fused/unfused branch.
/// \param BoundaryAlignment alignment requirement of the branch.
/// \returns true if the branch cross the boundary.
static bool mayCrossBoundary(uint64_t StartAddr, uint64_t Size,
Align BoundaryAlignment) {
uint64_t EndAddr = StartAddr + Size;
return (StartAddr >> Log2(BoundaryAlignment)) !=
((EndAddr - 1) >> Log2(BoundaryAlignment));
}
/// Check if the branch is against the boundary.
///
/// \param StartAddr start address of the fused/unfused branch.
/// \param Size size of the fused/unfused branch.
/// \param BoundaryAlignment alignment requirement of the branch.
/// \returns true if the branch is against the boundary.
static bool isAgainstBoundary(uint64_t StartAddr, uint64_t Size,
Align BoundaryAlignment) {
uint64_t EndAddr = StartAddr + Size;
return (EndAddr & (BoundaryAlignment.value() - 1)) == 0;
}
/// Check if the branch needs padding.
///
/// \param StartAddr start address of the fused/unfused branch.
/// \param Size size of the fused/unfused branch.
/// \param BoundaryAlignment alignment requirement of the branch.
/// \returns true if the branch needs padding.
static bool needPadding(uint64_t StartAddr, uint64_t Size,
Align BoundaryAlignment) {
return mayCrossBoundary(StartAddr, Size, BoundaryAlignment) ||
isAgainstBoundary(StartAddr, Size, BoundaryAlignment);
}
bool MCAssembler::relaxBoundaryAlign(MCBoundaryAlignFragment &BF) {
// BoundaryAlignFragment that doesn't need to align any fragment should not be
// relaxed.
if (!BF.getLastFragment())
return false;
uint64_t AlignedOffset = getFragmentOffset(BF);
uint64_t AlignedSize = 0;
for (const MCFragment *F = BF.getNext();; F = F->getNext()) {
AlignedSize += computeFragmentSize(*F);
if (F == BF.getLastFragment())
break;
}
Align BoundaryAlignment = BF.getAlignment();
uint64_t NewSize = needPadding(AlignedOffset, AlignedSize, BoundaryAlignment)
? offsetToAlignment(AlignedOffset, BoundaryAlignment)
: 0U;
if (NewSize == BF.getSize())
return false;
BF.setSize(NewSize);
return true;
}
bool MCAssembler::relaxDwarfLineAddr(MCFragment &F) {
bool WasRelaxed;
if (getBackend().relaxDwarfLineAddr(F, WasRelaxed))
return WasRelaxed;
MCContext &Context = getContext();
auto OldSize = F.getVarSize();
int64_t AddrDelta;
bool Abs = F.getDwarfAddrDelta().evaluateKnownAbsolute(AddrDelta, *this);
assert(Abs && "We created a line delta with an invalid expression");
(void)Abs;
SmallVector<char, 8> Data;
MCDwarfLineAddr::encode(Context, getDWARFLinetableParams(),
F.getDwarfLineDelta(), AddrDelta, Data);
F.setVarContents(Data);
F.clearVarFixups();
return OldSize != Data.size();
}
bool MCAssembler::relaxDwarfCallFrameFragment(MCFragment &F) {
bool WasRelaxed;
if (getBackend().relaxDwarfCFA(F, WasRelaxed))
return WasRelaxed;
MCContext &Context = getContext();
int64_t Value;
bool Abs = F.getDwarfAddrDelta().evaluateAsAbsolute(Value, *this);
if (!Abs) {
reportError(F.getDwarfAddrDelta().getLoc(),
"invalid CFI advance_loc expression");
F.setDwarfAddrDelta(MCConstantExpr::create(0, Context));
return false;
}
auto OldSize = F.getVarContents().size();
SmallVector<char, 8> Data;
MCDwarfFrameEmitter::encodeAdvanceLoc(Context, Value, Data);
F.setVarContents(Data);
F.clearVarFixups();
return OldSize != Data.size();
}
bool MCAssembler::relaxCVInlineLineTable(MCCVInlineLineTableFragment &F) {
unsigned OldSize = F.getContents().size();
getContext().getCVContext().encodeInlineLineTable(*this, F);
return OldSize != F.getContents().size();
}
bool MCAssembler::relaxCVDefRange(MCCVDefRangeFragment &F) {
unsigned OldSize = F.getContents().size();
getContext().getCVContext().encodeDefRange(*this, F);
return OldSize != F.getContents().size();
}
bool MCAssembler::relaxFill(MCFillFragment &F) {
uint64_t Size = computeFragmentSize(F);
if (F.getSize() == Size)
return false;
F.setSize(Size);
return true;
}
bool MCAssembler::relaxPseudoProbeAddr(MCPseudoProbeAddrFragment &PF) {
uint64_t OldSize = PF.getContents().size();
int64_t AddrDelta;
bool Abs = PF.getAddrDelta().evaluateKnownAbsolute(AddrDelta, *this);
assert(Abs && "We created a pseudo probe with an invalid expression");
(void)Abs;
SmallVector<char, 8> Data;
raw_svector_ostream OSE(Data);
// AddrDelta is a signed integer
encodeSLEB128(AddrDelta, OSE, OldSize);
PF.setContents(Data);
PF.clearFixups();
return OldSize != Data.size();
}
bool MCAssembler::relaxFragment(MCFragment &F) {
switch(F.getKind()) {
default:
return false;
case MCFragment::FT_Relaxable:
assert(!getRelaxAll() && "Did not expect a FT_Relaxable in RelaxAll mode");
return relaxInstruction(F);
case MCFragment::FT_LEB:
return relaxLEB(F);
case MCFragment::FT_Dwarf:
return relaxDwarfLineAddr(F);
case MCFragment::FT_DwarfFrame:
return relaxDwarfCallFrameFragment(F);
case MCFragment::FT_BoundaryAlign:
return relaxBoundaryAlign(cast<MCBoundaryAlignFragment>(F));
case MCFragment::FT_CVInlineLines:
return relaxCVInlineLineTable(cast<MCCVInlineLineTableFragment>(F));
case MCFragment::FT_CVDefRange:
return relaxCVDefRange(cast<MCCVDefRangeFragment>(F));
case MCFragment::FT_Fill:
return relaxFill(cast<MCFillFragment>(F));
case MCFragment::FT_PseudoProbe:
return relaxPseudoProbeAddr(cast<MCPseudoProbeAddrFragment>(F));
}
}
void MCAssembler::layoutSection(MCSection &Sec) {
uint64_t Offset = 0;
for (MCFragment &F : Sec) {
F.Offset = Offset;
Offset += computeFragmentSize(F);
}
}
unsigned MCAssembler::relaxOnce(unsigned FirstStable) {
++stats::RelaxationSteps;
PendingErrors.clear();
unsigned Res = 0;
for (unsigned I = 0; I != FirstStable; ++I) {
// Assume each iteration finalizes at least one extra fragment. If the
// layout does not converge after N+1 iterations, bail out.
auto &Sec = *Sections[I];
auto MaxIter = Sec.curFragList()->Tail->getLayoutOrder() + 1;
for (;;) {
bool Changed = false;
for (MCFragment &F : Sec)
if (relaxFragment(F))
Changed = true;
if (!Changed)
break;
// If any fragment changed size, it might impact the layout of subsequent
// sections. Therefore, we must re-evaluate all sections.
FirstStable = Sections.size();
Res = I;
if (--MaxIter == 0)
break;
layoutSection(Sec);
}
}
// The subsequent relaxOnce call only needs to visit Sections [0,Res) if no
// change occurred.
return Res;
}
void MCAssembler::reportError(SMLoc L, const Twine &Msg) const {
getContext().reportError(L, Msg);
}
void MCAssembler::recordError(SMLoc Loc, const Twine &Msg) const {
PendingErrors.emplace_back(Loc, Msg.str());
}
void MCAssembler::flushPendingErrors() const {
for (auto &Err : PendingErrors)
reportError(Err.first, Err.second);
PendingErrors.clear();
}
#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
LLVM_DUMP_METHOD void MCAssembler::dump() const{
raw_ostream &OS = errs();
DenseMap<const MCFragment *, SmallVector<const MCSymbol *, 0>> FragToSyms;
// Scan symbols and build a map of fragments to their corresponding symbols.
// For variable symbols, we don't want to call their getFragment, which might
// modify `Fragment`.
for (const MCSymbol &Sym : symbols())
if (!Sym.isVariable())
if (auto *F = Sym.getFragment())
FragToSyms.try_emplace(F).first->second.push_back(&Sym);
OS << "Sections:[";
for (const MCSection &Sec : *this) {
OS << '\n';
Sec.dump(&FragToSyms);
}
OS << "\n]\n";
}
#endif
SMLoc MCFixup::getLoc() const {
if (auto *E = getValue())
return E->getLoc();
return {};
}
|