lib/MC/MCMachOStreamer.cpp

   1 //===- lib/MC/MCMachOStreamer.cpp - Mach-O Object Output ------------===//
   2 //
   3 //                     The LLVM Compiler Infrastructure
   4 //
   5 // This file is distributed under the University of Illinois Open Source
   6 // License. See LICENSE.TXT for details.
   7 //
   8 //===----------------------------------------------------------------------===//
   9
  10 #include "llvm/MC/MCStreamer.h"
  11
  12 #include "llvm/MC/MCAssembler.h"
  13 #include "llvm/MC/MCContext.h"
  14 #include "llvm/MC/MCCodeEmitter.h"
  15 #include "llvm/MC/MCExpr.h"
  16 #include "llvm/MC/MCInst.h"
  17 #include "llvm/MC/MCSection.h"
  18 #include "llvm/MC/MCSymbol.h"
  19 #include "llvm/MC/MCMachOSymbolFlags.h"
  20 #include "llvm/Support/ErrorHandling.h"
  21 #include "llvm/Support/raw_ostream.h"
  22 #include "llvm/Target/TargetAsmBackend.h"
  23
  24 using namespace llvm;
  25
  26 namespace {
  27
  28 class MCMachOStreamer : public MCStreamer {
  29
  30 private:
  31   MCAssembler Assembler;
  32   MCSectionData *CurSectionData;
  33
  34   /// Track the current atom for each section.
  35   DenseMap<const MCSectionData*, MCSymbolData*> CurrentAtomMap;
  36
  37 private:
  38   MCFragment *getCurrentFragment() const {
  39     assert(CurSectionData && "No current section!");
  40
  41     if (!CurSectionData->empty())
  42       return &CurSectionData->getFragmentList().back();
  43
  44     return 0;
  45   }
  46
  47   /// Get a data fragment to write into, creating a new one if the current
  48   /// fragment is not a data fragment.
  49   MCDataFragment *getOrCreateDataFragment() const {
  50     MCDataFragment *F = dyn_cast_or_null<MCDataFragment>(getCurrentFragment());
  51     if (!F)
  52       F = createDataFragment();
  53     return F;
  54   }
  55
  56   /// Create a new data fragment in the current section.
  57   MCDataFragment *createDataFragment() const {
  58     MCDataFragment *DF = new MCDataFragment(CurSectionData);
  59     DF->setAtom(CurrentAtomMap.lookup(CurSectionData));
  60     return DF;
  61   }
  62
  63 public:
  64   MCMachOStreamer(MCContext &Context, TargetAsmBackend &TAB,
  65                   raw_ostream &_OS, MCCodeEmitter *_Emitter)
  66     : MCStreamer(Context), Assembler(Context, TAB, *_Emitter, _OS),
  67       CurSectionData(0) {}
  68   ~MCMachOStreamer() {}
  69
  70   MCAssembler &getAssembler() { return Assembler; }
  71
  72   const MCExpr *AddValueSymbols(const MCExpr *Value) {
  73     switch (Value->getKind()) {
  74     case MCExpr::Target: assert(0 && "Can't handle target exprs yet!");
  75     case MCExpr::Constant:
  76       break;
  77
  78     case MCExpr::Binary: {
  79       const MCBinaryExpr *BE = cast<MCBinaryExpr>(Value);
  80       AddValueSymbols(BE->getLHS());
  81       AddValueSymbols(BE->getRHS());
  82       break;
  83     }
  84
  85     case MCExpr::SymbolRef:
  86       Assembler.getOrCreateSymbolData(
  87         cast<MCSymbolRefExpr>(Value)->getSymbol());
  88       break;
  89
  90     case MCExpr::Unary:
  91       AddValueSymbols(cast<MCUnaryExpr>(Value)->getSubExpr());
  92       break;
  93     }
  94
  95     return Value;
  96   }
  97
  98   /// @name MCStreamer Interface
  99   /// @{
 100
 101   virtual void SwitchSection(const MCSection *Section);
 102   virtual void EmitLabel(MCSymbol *Symbol);
 103   virtual void EmitAssemblerFlag(MCAssemblerFlag Flag);
 104   virtual void EmitAssignment(MCSymbol *Symbol, const MCExpr *Value);
 105   virtual void EmitSymbolAttribute(MCSymbol *Symbol, MCSymbolAttr Attribute);
 106   virtual void EmitSymbolDesc(MCSymbol *Symbol, unsigned DescValue);
 107   virtual void EmitCommonSymbol(MCSymbol *Symbol, uint64_t Size,
 108                                 unsigned ByteAlignment);
 109   virtual void BeginCOFFSymbolDef(const MCSymbol *Symbol) {
 110     assert(0 && "macho doesn't support this directive");
 111   }
 112   virtual void EmitCOFFSymbolStorageClass(int StorageClass) {
 113     assert(0 && "macho doesn't support this directive");
 114   }
 115   virtual void EmitCOFFSymbolType(int Type) {
 116     assert(0 && "macho doesn't support this directive");
 117   }
 118   virtual void EndCOFFSymbolDef() {
 119     assert(0 && "macho doesn't support this directive");
 120   }
 121   virtual void EmitELFSize(MCSymbol *Symbol, const MCExpr *Value) {
 122     assert(0 && "macho doesn't support this directive");
 123   }
 124   virtual void EmitLocalCommonSymbol(MCSymbol *Symbol, uint64_t Size) {
 125     assert(0 && "macho doesn't support this directive");
 126   }
 127   virtual void EmitZerofill(const MCSection *Section, MCSymbol *Symbol = 0,
 128                             unsigned Size = 0, unsigned ByteAlignment = 0);
 129   virtual void EmitTBSSSymbol(MCSymbol *Symbol, uint64_t Size,
 130                               unsigned ByteAlignment = 0);
 131   virtual void EmitBytes(StringRef Data, unsigned AddrSpace);
 132   virtual void EmitValue(const MCExpr *Value, unsigned Size,unsigned AddrSpace);
 133   virtual void EmitGPRel32Value(const MCExpr *Value) {
 134     assert(0 && "macho doesn't support this directive");
 135   }
 136   virtual void EmitValueToAlignment(unsigned ByteAlignment, int64_t Value = 0,
 137                                     unsigned ValueSize = 1,
 138                                     unsigned MaxBytesToEmit = 0);
 139   virtual void EmitCodeAlignment(unsigned ByteAlignment,
 140                                  unsigned MaxBytesToEmit = 0);
 141   virtual void EmitValueToOffset(const MCExpr *Offset,
 142                                  unsigned char Value = 0);
 143
 144   virtual void EmitFileDirective(StringRef Filename) {
 145     report_fatal_error("unsupported directive: '.file'");
 146   }
 147   virtual void EmitDwarfFileDirective(unsigned FileNo, StringRef Filename) {
 148     report_fatal_error("unsupported directive: '.file'");
 149   }
 150
 151   virtual void EmitInstruction(const MCInst &Inst);
 152   virtual void Finish();
 153
 154   /// @}
 155 };
 156
 157 } // end anonymous namespace.
 158
 159 void MCMachOStreamer::SwitchSection(const MCSection *Section) {
 160   assert(Section && "Cannot switch to a null section!");
 161
 162   // If already in this section, then this is a noop.
 163   if (Section == CurSection) return;
 164
 165   CurSection = Section;
 166   CurSectionData = &Assembler.getOrCreateSectionData(*Section);
 167 }
 168
 169 void MCMachOStreamer::EmitLabel(MCSymbol *Symbol) {
 170   assert(Symbol->isUndefined() && "Cannot define a symbol twice!");
 171   assert(!Symbol->isVariable() && "Cannot emit a variable symbol!");
 172   assert(CurSection && "Cannot emit before setting section!");
 173
 174   MCSymbolData &SD = Assembler.getOrCreateSymbolData(*Symbol);
 175
 176   // Update the current atom map, if necessary.
 177   bool MustCreateFragment = false;
 178   if (Assembler.isSymbolLinkerVisible(&SD)) {
 179     CurrentAtomMap[CurSectionData] = &SD;
 180
 181     // We have to create a new fragment, fragments cannot span atoms.
 182     MustCreateFragment = true;
 183   }
 184
 185   // FIXME: This is wasteful, we don't necessarily need to create a data
 186   // fragment. Instead, we should mark the symbol as pointing into the data
 187   // fragment if it exists, otherwise we should just queue the label and set its
 188   // fragment pointer when we emit the next fragment.
 189   MCDataFragment *F =
 190     MustCreateFragment ? createDataFragment() : getOrCreateDataFragment();
 191   assert(!SD.getFragment() && "Unexpected fragment on symbol data!");
 192   SD.setFragment(F);
 193   SD.setOffset(F->getContents().size());
 194
 195   // This causes the reference type flag to be cleared. Darwin 'as' was "trying"
 196   // to clear the weak reference and weak definition bits too, but the
 197   // implementation was buggy. For now we just try to match 'as', for
 198   // diffability.
 199   //
 200   // FIXME: Cleanup this code, these bits should be emitted based on semantic
 201   // properties, not on the order of definition, etc.
 202   SD.setFlags(SD.getFlags() & ~SF_ReferenceTypeMask);
 203
 204   Symbol->setSection(*CurSection);
 205 }
 206
 207 void MCMachOStreamer::EmitAssemblerFlag(MCAssemblerFlag Flag) {
 208   switch (Flag) {
 209   case MCAF_SubsectionsViaSymbols:
 210     Assembler.setSubsectionsViaSymbols(true);
 211     return;
 212   }
 213
 214   assert(0 && "invalid assembler flag!");
 215 }
 216
 217 void MCMachOStreamer::EmitAssignment(MCSymbol *Symbol, const MCExpr *Value) {
 218   // FIXME: Lift context changes into super class.
 219   Assembler.getOrCreateSymbolData(*Symbol);
 220   Symbol->setVariableValue(AddValueSymbols(Value));
 221 }
 222
 223 void MCMachOStreamer::EmitSymbolAttribute(MCSymbol *Symbol,
 224                                           MCSymbolAttr Attribute) {
 225   // Indirect symbols are handled differently, to match how 'as' handles
 226   // them. This makes writing matching .o files easier.
 227   if (Attribute == MCSA_IndirectSymbol) {
 228     // Note that we intentionally cannot use the symbol data here; this is
 229     // important for matching the string table that 'as' generates.
 230     IndirectSymbolData ISD;
 231     ISD.Symbol = Symbol;
 232     ISD.SectionData = CurSectionData;
 233     Assembler.getIndirectSymbols().push_back(ISD);
 234     return;
 235   }
 236
 237   // Adding a symbol attribute always introduces the symbol, note that an
 238   // important side effect of calling getOrCreateSymbolData here is to register
 239   // the symbol with the assembler.
 240   MCSymbolData &SD = Assembler.getOrCreateSymbolData(*Symbol);
 241
 242   // The implementation of symbol attributes is designed to match 'as', but it
 243   // leaves much to desired. It doesn't really make sense to arbitrarily add and
 244   // remove flags, but 'as' allows this (in particular, see .desc).
 245   //
 246   // In the future it might be worth trying to make these operations more well
 247   // defined.
 248   switch (Attribute) {
 249   case MCSA_Invalid:
 250   case MCSA_ELF_TypeFunction:
 251   case MCSA_ELF_TypeIndFunction:
 252   case MCSA_ELF_TypeObject:
 253   case MCSA_ELF_TypeTLS:
 254   case MCSA_ELF_TypeCommon:
 255   case MCSA_ELF_TypeNoType:
 256   case MCSA_IndirectSymbol:
 257   case MCSA_Hidden:
 258   case MCSA_Internal:
 259   case MCSA_Protected:
 260   case MCSA_Weak:
 261   case MCSA_Local:
 262     assert(0 && "Invalid symbol attribute for Mach-O!");
 263     break;
 264
 265   case MCSA_Global:
 266     SD.setExternal(true);
 267     // This effectively clears the undefined lazy bit, in Darwin 'as', although
 268     // it isn't very consistent because it implements this as part of symbol
 269     // lookup.
 270     //
 271     // FIXME: Cleanup this code, these bits should be emitted based on semantic
 272     // properties, not on the order of definition, etc.
 273     SD.setFlags(SD.getFlags() & ~SF_ReferenceTypeUndefinedLazy);
 274     break;
 275
 276   case MCSA_LazyReference:
 277     // FIXME: This requires -dynamic.
 278     SD.setFlags(SD.getFlags() | SF_NoDeadStrip);
 279     if (Symbol->isUndefined())
 280       SD.setFlags(SD.getFlags() | SF_ReferenceTypeUndefinedLazy);
 281     break;
 282
 283     // Since .reference sets the no dead strip bit, it is equivalent to
 284     // .no_dead_strip in practice.
 285   case MCSA_Reference:
 286   case MCSA_NoDeadStrip:
 287     SD.setFlags(SD.getFlags() | SF_NoDeadStrip);
 288     break;
 289
 290   case MCSA_PrivateExtern:
 291     SD.setExternal(true);
 292     SD.setPrivateExtern(true);
 293     break;
 294
 295   case MCSA_WeakReference:
 296     // FIXME: This requires -dynamic.
 297     if (Symbol->isUndefined())
 298       SD.setFlags(SD.getFlags() | SF_WeakReference);
 299     break;
 300
 301   case MCSA_WeakDefinition:
 302     // FIXME: 'as' enforces that this is defined and global. The manual claims
 303     // it has to be in a coalesced section, but this isn't enforced.
 304     SD.setFlags(SD.getFlags() | SF_WeakDefinition);
 305     break;
 306   }
 307 }
 308
 309 void MCMachOStreamer::EmitSymbolDesc(MCSymbol *Symbol, unsigned DescValue) {
 310   // Encode the 'desc' value into the lowest implementation defined bits.
 311   assert(DescValue == (DescValue & SF_DescFlagsMask) &&
 312          "Invalid .desc value!");
 313   Assembler.getOrCreateSymbolData(*Symbol).setFlags(DescValue&SF_DescFlagsMask);
 314 }
 315
 316 void MCMachOStreamer::EmitCommonSymbol(MCSymbol *Symbol, uint64_t Size,
 317                                        unsigned ByteAlignment) {
 318   // FIXME: Darwin 'as' does appear to allow redef of a .comm by itself.
 319   assert(Symbol->isUndefined() && "Cannot define a symbol twice!");
 320
 321   MCSymbolData &SD = Assembler.getOrCreateSymbolData(*Symbol);
 322   SD.setExternal(true);
 323   SD.setCommon(Size, ByteAlignment);
 324 }
 325
 326 void MCMachOStreamer::EmitZerofill(const MCSection *Section, MCSymbol *Symbol,
 327                                    unsigned Size, unsigned ByteAlignment) {
 328   MCSectionData &SectData = Assembler.getOrCreateSectionData(*Section);
 329
 330   // The symbol may not be present, which only creates the section.
 331   if (!Symbol)
 332     return;
 333
 334   // FIXME: Assert that this section has the zerofill type.
 335
 336   assert(Symbol->isUndefined() && "Cannot define a symbol twice!");
 337
 338   MCSymbolData &SD = Assembler.getOrCreateSymbolData(*Symbol);
 339
 340   // Emit an align fragment if necessary.
 341   if (ByteAlignment != 1)
 342     new MCAlignFragment(ByteAlignment, 0, 0, ByteAlignment, &SectData);
 343
 344   MCFragment *F = new MCFillFragment(0, 0, Size, &SectData);
 345   SD.setFragment(F);
 346   if (Assembler.isSymbolLinkerVisible(&SD))
 347     F->setAtom(&SD);
 348
 349   Symbol->setSection(*Section);
 350
 351   // Update the maximum alignment on the zero fill section if necessary.
 352   if (ByteAlignment > SectData.getAlignment())
 353     SectData.setAlignment(ByteAlignment);
 354 }
 355
 356 void MCMachOStreamer::EmitTBSSSymbol(MCSymbol *Symbol, uint64_t Size,
 357                                      unsigned ByteAlignment) {
 358   assert(false && "Implement me!");
 359 }
 360
 361 void MCMachOStreamer::EmitBytes(StringRef Data, unsigned AddrSpace) {
 362   getOrCreateDataFragment()->getContents().append(Data.begin(), Data.end());
 363 }
 364
 365 void MCMachOStreamer::EmitValue(const MCExpr *Value, unsigned Size,
 366                                 unsigned AddrSpace) {
 367   MCDataFragment *DF = getOrCreateDataFragment();
 368
 369   // Avoid fixups when possible.
 370   int64_t AbsValue;
 371   if (AddValueSymbols(Value)->EvaluateAsAbsolute(AbsValue)) {
 372     // FIXME: Endianness assumption.
 373     for (unsigned i = 0; i != Size; ++i)
 374       DF->getContents().push_back(uint8_t(AbsValue >> (i * 8)));
 375   } else {
 376     DF->addFixup(MCAsmFixup(DF->getContents().size(), *AddValueSymbols(Value),
 377                             MCFixup::getKindForSize(Size)));
 378     DF->getContents().resize(DF->getContents().size() + Size, 0);
 379   }
 380 }
 381
 382 void MCMachOStreamer::EmitValueToAlignment(unsigned ByteAlignment,
 383                                            int64_t Value, unsigned ValueSize,
 384                                            unsigned MaxBytesToEmit) {
 385   if (MaxBytesToEmit == 0)
 386     MaxBytesToEmit = ByteAlignment;
 387   MCFragment *F = new MCAlignFragment(ByteAlignment, Value, ValueSize,
 388                                       MaxBytesToEmit, CurSectionData);
 389   F->setAtom(CurrentAtomMap.lookup(CurSectionData));
 390
 391   // Update the maximum alignment on the current section if necessary.
 392   if (ByteAlignment > CurSectionData->getAlignment())
 393     CurSectionData->setAlignment(ByteAlignment);
 394 }
 395
 396 void MCMachOStreamer::EmitCodeAlignment(unsigned ByteAlignment,
 397                                         unsigned MaxBytesToEmit) {
 398   if (MaxBytesToEmit == 0)
 399     MaxBytesToEmit = ByteAlignment;
 400   MCAlignFragment *F = new MCAlignFragment(ByteAlignment, 0, 1, MaxBytesToEmit,
 401                                            CurSectionData);
 402   F->setEmitNops(true);
 403   F->setAtom(CurrentAtomMap.lookup(CurSectionData));
 404
 405   // Update the maximum alignment on the current section if necessary.
 406   if (ByteAlignment > CurSectionData->getAlignment())
 407     CurSectionData->setAlignment(ByteAlignment);
 408 }
 409
 410 void MCMachOStreamer::EmitValueToOffset(const MCExpr *Offset,
 411                                         unsigned char Value) {
 412   MCFragment *F = new MCOrgFragment(*Offset, Value, CurSectionData);
 413   F->setAtom(CurrentAtomMap.lookup(CurSectionData));
 414 }
 415
 416 void MCMachOStreamer::EmitInstruction(const MCInst &Inst) {
 417   // Scan for values.
 418   for (unsigned i = Inst.getNumOperands(); i--; )
 419     if (Inst.getOperand(i).isExpr())
 420       AddValueSymbols(Inst.getOperand(i).getExpr());
 421
 422   CurSectionData->setHasInstructions(true);
 423
 424   // FIXME-PERF: Common case is that we don't need to relax, encode directly
 425   // onto the data fragments buffers.
 426
 427   SmallVector<MCFixup, 4> Fixups;
 428   SmallString<256> Code;
 429   raw_svector_ostream VecOS(Code);
 430   Assembler.getEmitter().EncodeInstruction(Inst, VecOS, Fixups);
 431   VecOS.flush();
 432
 433   // FIXME: Eliminate this copy.
 434   SmallVector<MCAsmFixup, 4> AsmFixups;
 435   for (unsigned i = 0, e = Fixups.size(); i != e; ++i) {
 436     MCFixup &F = Fixups[i];
 437     AsmFixups.push_back(MCAsmFixup(F.getOffset(), *F.getValue(),
 438                                    F.getKind()));
 439   }
 440
 441   // See if we might need to relax this instruction, if so it needs its own
 442   // fragment.
 443   //
 444   // FIXME-PERF: Support target hook to do a fast path that avoids the encoder,
 445   // when we can immediately tell that we will get something which might need
 446   // relaxation (and compute its size).
 447   //
 448   // FIXME-PERF: We should also be smart about immediately relaxing instructions
 449   // which we can already show will never possibly fit (we can also do a very
 450   // good job of this before we do the first relaxation pass, because we have
 451   // total knowledge about undefined symbols at that point). Even now, though,
 452   // we can do a decent job, especially on Darwin where scattering means that we
 453   // are going to often know that we can never fully resolve a fixup.
 454   if (Assembler.getBackend().MayNeedRelaxation(Inst, AsmFixups)) {
 455     MCInstFragment *IF = new MCInstFragment(Inst, CurSectionData);
 456     IF->setAtom(CurrentAtomMap.lookup(CurSectionData));
 457
 458     // Add the fixups and data.
 459     //
 460     // FIXME: Revisit this design decision when relaxation is done, we may be
 461     // able to get away with not storing any extra data in the MCInst.
 462     IF->getCode() = Code;
 463     IF->getFixups() = AsmFixups;
 464
 465     return;
 466   }
 467
 468   // Add the fixups and data.
 469   MCDataFragment *DF = getOrCreateDataFragment();
 470   for (unsigned i = 0, e = AsmFixups.size(); i != e; ++i) {
 471     AsmFixups[i].Offset += DF->getContents().size();
 472     DF->addFixup(AsmFixups[i]);
 473   }
 474   DF->getContents().append(Code.begin(), Code.end());
 475 }
 476
 477 void MCMachOStreamer::Finish() {
 478   Assembler.Finish();
 479 }
 480
 481 MCStreamer *llvm::createMachOStreamer(MCContext &Context, TargetAsmBackend &TAB,
 482                                       raw_ostream &OS, MCCodeEmitter *CE,
 483                                       bool RelaxAll) {
 484   MCMachOStreamer *S = new MCMachOStreamer(Context, TAB, OS, CE);
 485   if (RelaxAll)
 486     S->getAssembler().setRelaxAll(true);
 487   return S;
 488 }