X-Git-Url: http://demsky.eecs.uci.edu/git/?a=blobdiff_plain;f=lib%2FObject%2FMachOObjectFile.cpp;h=a930117019e7edbfc87228b6f83a373cbe17a5ce;hb=bb96dfcf4f58ae7152f4be92254a0bb1c0995e6d;hp=f4df8e012cf0aa2846981ff24866ec3b2bfbaf1d;hpb=b08c6df6787971502bd51e30b0f1038c1ea0dc2c;p=oota-llvm.git diff --git a/lib/Object/MachOObjectFile.cpp b/lib/Object/MachOObjectFile.cpp index f4df8e012cf..a930117019e 100644 --- a/lib/Object/MachOObjectFile.cpp +++ b/lib/Object/MachOObjectFile.cpp @@ -14,11 +14,11 @@ #include "llvm/Object/MachO.h" #include "llvm/ADT/Triple.h" -#include "llvm/Object/MachOFormat.h" -#include "llvm/Support/Casting.h" #include "llvm/Support/DataExtractor.h" #include "llvm/Support/Format.h" +#include "llvm/Support/Host.h" #include "llvm/Support/MemoryBuffer.h" +#include "llvm/Support/raw_ostream.h" #include #include #include @@ -29,363 +29,743 @@ using namespace object; namespace llvm { namespace object { -MachOObjectFileBase::MachOObjectFileBase(MemoryBuffer *Object, bool Is64bits, - error_code &ec) - : ObjectFile(getMachOType(true, Is64bits), Object) { +struct nlist_base { + uint32_t n_strx; + uint8_t n_type; + uint8_t n_sect; + uint16_t n_desc; +}; + +struct section_base { + char sectname[16]; + char segname[16]; +}; + +template +static void SwapValue(T &Value) { + Value = sys::SwapByteOrder(Value); } -bool MachOObjectFileBase::is64Bit() const { - return isa(this); +template +static void SwapStruct(T &Value); + +template<> +void SwapStruct(MachO::any_relocation_info &H) { + SwapValue(H.r_word0); + SwapValue(H.r_word1); } -const MachOObjectFileBase::LoadCommand * -MachOObjectFileBase::getLoadCommandInfo(unsigned Index) const { - uint64_t Offset; - uint64_t NewOffset = getHeaderSize(); - const LoadCommand *Load; - unsigned I = 0; - do { - Offset = NewOffset; - StringRef Data = getData(Offset, sizeof(LoadCommand)); - Load = reinterpret_cast(Data.data()); - NewOffset = Offset + Load->Size; - ++I; - } while (I != Index + 1); +template<> +void SwapStruct(MachO::load_command &L) { + SwapValue(L.cmd); + SwapValue(L.cmdsize); +} - return Load; +template<> +void SwapStruct(nlist_base &S) { + SwapValue(S.n_strx); + SwapValue(S.n_desc); } -void MachOObjectFileBase::ReadULEB128s(uint64_t Index, - SmallVectorImpl &Out) const { - DataExtractor extractor(ObjectFile::getData(), true, 0); +template<> +void SwapStruct(MachO::section &S) { + SwapValue(S.addr); + SwapValue(S.size); + SwapValue(S.offset); + SwapValue(S.align); + SwapValue(S.reloff); + SwapValue(S.nreloc); + SwapValue(S.flags); + SwapValue(S.reserved1); + SwapValue(S.reserved2); +} - uint32_t offset = Index; - uint64_t data = 0; - while (uint64_t delta = extractor.getULEB128(&offset)) { - data += delta; - Out.push_back(data); +template<> +void SwapStruct(MachO::section_64 &S) { + SwapValue(S.addr); + SwapValue(S.size); + SwapValue(S.offset); + SwapValue(S.align); + SwapValue(S.reloff); + SwapValue(S.nreloc); + SwapValue(S.flags); + SwapValue(S.reserved1); + SwapValue(S.reserved2); + SwapValue(S.reserved3); +} + +template<> +void SwapStruct(MachO::nlist &S) { + SwapValue(S.n_strx); + SwapValue(S.n_desc); + SwapValue(S.n_value); +} + +template<> +void SwapStruct(MachO::nlist_64 &S) { + SwapValue(S.n_strx); + SwapValue(S.n_desc); + SwapValue(S.n_value); +} + +template<> +void SwapStruct(MachO::mach_header &H) { + SwapValue(H.magic); + SwapValue(H.cputype); + SwapValue(H.cpusubtype); + SwapValue(H.filetype); + SwapValue(H.ncmds); + SwapValue(H.sizeofcmds); + SwapValue(H.flags); +} + +template<> +void SwapStruct(MachO::mach_header_64 &H) { + SwapValue(H.magic); + SwapValue(H.cputype); + SwapValue(H.cpusubtype); + SwapValue(H.filetype); + SwapValue(H.ncmds); + SwapValue(H.sizeofcmds); + SwapValue(H.flags); + SwapValue(H.reserved); +} + +template<> +void SwapStruct(MachO::symtab_command &C) { + SwapValue(C.cmd); + SwapValue(C.cmdsize); + SwapValue(C.symoff); + SwapValue(C.nsyms); + SwapValue(C.stroff); + SwapValue(C.strsize); +} + +template<> +void SwapStruct(MachO::dysymtab_command &C) { + SwapValue(C.cmd); + SwapValue(C.cmdsize); + SwapValue(C.ilocalsym); + SwapValue(C.nlocalsym); + SwapValue(C.iextdefsym); + SwapValue(C.nextdefsym); + SwapValue(C.iundefsym); + SwapValue(C.nundefsym); + SwapValue(C.tocoff); + SwapValue(C.ntoc); + SwapValue(C.modtaboff); + SwapValue(C.nmodtab); + SwapValue(C.extrefsymoff); + SwapValue(C.nextrefsyms); + SwapValue(C.indirectsymoff); + SwapValue(C.nindirectsyms); + SwapValue(C.extreloff); + SwapValue(C.nextrel); + SwapValue(C.locreloff); + SwapValue(C.nlocrel); +} + +template<> +void SwapStruct(MachO::linkedit_data_command &C) { + SwapValue(C.cmd); + SwapValue(C.cmdsize); + SwapValue(C.dataoff); + SwapValue(C.datasize); +} + +template<> +void SwapStruct(MachO::segment_command &C) { + SwapValue(C.cmd); + SwapValue(C.cmdsize); + SwapValue(C.vmaddr); + SwapValue(C.vmsize); + SwapValue(C.fileoff); + SwapValue(C.filesize); + SwapValue(C.maxprot); + SwapValue(C.initprot); + SwapValue(C.nsects); + SwapValue(C.flags); +} + +template<> +void SwapStruct(MachO::segment_command_64 &C) { + SwapValue(C.cmd); + SwapValue(C.cmdsize); + SwapValue(C.vmaddr); + SwapValue(C.vmsize); + SwapValue(C.fileoff); + SwapValue(C.filesize); + SwapValue(C.maxprot); + SwapValue(C.initprot); + SwapValue(C.nsects); + SwapValue(C.flags); +} + +template<> +void SwapStruct(uint32_t &C) { + SwapValue(C); +} + +template<> +void SwapStruct(MachO::linker_options_command &C) { + SwapValue(C.cmd); + SwapValue(C.cmdsize); + SwapValue(C.count); +} + +template<> +void SwapStruct(MachO::data_in_code_entry &C) { + SwapValue(C.offset); + SwapValue(C.length); + SwapValue(C.kind); +} + +template +T getStruct(const MachOObjectFile *O, const char *P) { + T Cmd; + memcpy(&Cmd, P, sizeof(T)); + if (O->isLittleEndian() != sys::IsLittleEndianHost) + SwapStruct(Cmd); + return Cmd; +} + +static uint32_t +getSegmentLoadCommandNumSections(const MachOObjectFile *O, + const MachOObjectFile::LoadCommandInfo &L) { + if (O->is64Bit()) { + MachO::segment_command_64 S = O->getSegment64LoadCommand(L); + return S.nsects; } + MachO::segment_command S = O->getSegmentLoadCommand(L); + return S.nsects; } -const MachOObjectFileBase::Header *MachOObjectFileBase::getHeader() const { - StringRef Data = getData(0, sizeof(Header)); - return reinterpret_cast(Data.data()); +static const char * +getSectionPtr(const MachOObjectFile *O, MachOObjectFile::LoadCommandInfo L, + unsigned Sec) { + uintptr_t CommandAddr = reinterpret_cast(L.Ptr); + + bool Is64 = O->is64Bit(); + unsigned SegmentLoadSize = Is64 ? sizeof(MachO::segment_command_64) : + sizeof(MachO::segment_command); + unsigned SectionSize = Is64 ? sizeof(MachO::section_64) : + sizeof(MachO::section); + + uintptr_t SectionAddr = CommandAddr + SegmentLoadSize + Sec * SectionSize; + return reinterpret_cast(SectionAddr); } -unsigned MachOObjectFileBase::getHeaderSize() const { - return is64Bit() ? macho::Header64Size : macho::Header32Size; +static const char *getPtr(const MachOObjectFile *O, size_t Offset) { + return O->getData().substr(Offset, 1).data(); } -StringRef MachOObjectFileBase::getData(size_t Offset, size_t Size) const { - return ObjectFile::getData().substr(Offset, Size); +static nlist_base +getSymbolTableEntryBase(const MachOObjectFile *O, DataRefImpl DRI) { + const char *P = reinterpret_cast(DRI.p); + return getStruct(O, P); } -ObjectFile *ObjectFile::createMachOObjectFile(MemoryBuffer *Buffer) { - StringRef Magic = Buffer->getBuffer().slice(0, 4); +static StringRef parseSegmentOrSectionName(const char *P) { + if (P[15] == 0) + // Null terminated. + return P; + // Not null terminated, so this is a 16 char string. + return StringRef(P, 16); +} + +// Helper to advance a section or symbol iterator multiple increments at a time. +template +static error_code advance(T &it, size_t Val) { error_code ec; - bool Is64Bits = Magic == "\xFE\xED\xFA\xCF" || Magic == "\xCF\xFA\xED\xFE"; - ObjectFile *Ret; - if (Is64Bits) - Ret = new MachOObjectFile64Le(Buffer, ec); - else - Ret = new MachOObjectFile32Le(Buffer, ec); - if (ec) - return NULL; - return Ret; -} - -/*===-- Symbols -----------------------------------------------------------===*/ - -void MachOObjectFileBase::moveToNextSymbol(DataRefImpl &DRI) const { - uint32_t LoadCommandCount = getHeader()->NumLoadCommands; - while (DRI.d.a < LoadCommandCount) { - const LoadCommand *Command = getLoadCommandInfo(DRI.d.a); - if (Command->Type == macho::LCT_Symtab) { - const SymtabLoadCommand *SymtabLoadCmd = - reinterpret_cast(Command); - if (DRI.d.b < SymtabLoadCmd->NumSymbolTableEntries) - return; + while (Val--) { + it.increment(ec); + } + return ec; +} + +template +static void advanceTo(T &it, size_t Val) { + if (error_code ec = advance(it, Val)) + report_fatal_error(ec.message()); +} + +static unsigned getCPUType(const MachOObjectFile *O) { + return O->getHeader().cputype; +} + +static void printRelocationTargetName(const MachOObjectFile *O, + const MachO::any_relocation_info &RE, + raw_string_ostream &fmt) { + bool IsScattered = O->isRelocationScattered(RE); + + // Target of a scattered relocation is an address. In the interest of + // generating pretty output, scan through the symbol table looking for a + // symbol that aligns with that address. If we find one, print it. + // Otherwise, we just print the hex address of the target. + if (IsScattered) { + uint32_t Val = O->getPlainRelocationSymbolNum(RE); + + error_code ec; + for (symbol_iterator SI = O->begin_symbols(), SE = O->end_symbols(); + SI != SE; SI.increment(ec)) { + if (ec) report_fatal_error(ec.message()); + + uint64_t Addr; + StringRef Name; + + if ((ec = SI->getAddress(Addr))) + report_fatal_error(ec.message()); + if (Addr != Val) continue; + if ((ec = SI->getName(Name))) + report_fatal_error(ec.message()); + fmt << Name; + return; } - DRI.d.a++; - DRI.d.b = 0; + // If we couldn't find a symbol that this relocation refers to, try + // to find a section beginning instead. + for (section_iterator SI = O->begin_sections(), SE = O->end_sections(); + SI != SE; SI.increment(ec)) { + if (ec) report_fatal_error(ec.message()); + + uint64_t Addr; + StringRef Name; + + if ((ec = SI->getAddress(Addr))) + report_fatal_error(ec.message()); + if (Addr != Val) continue; + if ((ec = SI->getName(Name))) + report_fatal_error(ec.message()); + fmt << Name; + return; + } + + fmt << format("0x%x", Val); + return; } -} -const MachOObjectFileBase::SymbolTableEntryBase * -MachOObjectFileBase::getSymbolTableEntryBase(DataRefImpl DRI) const { - const LoadCommand *Command = getLoadCommandInfo(DRI.d.a); - const SymtabLoadCommand *SymtabLoadCmd = - reinterpret_cast(Command); - return getSymbolTableEntryBase(DRI, SymtabLoadCmd); + StringRef S; + bool isExtern = O->getPlainRelocationExternal(RE); + uint64_t Val = O->getPlainRelocationSymbolNum(RE); + + if (isExtern) { + symbol_iterator SI = O->begin_symbols(); + advanceTo(SI, Val); + SI->getName(S); + } else { + section_iterator SI = O->begin_sections(); + // Adjust for the fact that sections are 1-indexed. + advanceTo(SI, Val - 1); + SI->getName(S); + } + + fmt << S; } -const MachOObjectFileBase::SymbolTableEntryBase * -MachOObjectFileBase::getSymbolTableEntryBase(DataRefImpl DRI, - const SymtabLoadCommand *SymtabLoadCmd) const { - uint64_t SymbolTableOffset = SymtabLoadCmd->SymbolTableOffset; - unsigned Index = DRI.d.b; +static uint32_t +getPlainRelocationAddress(const MachO::any_relocation_info &RE) { + return RE.r_word0; +} - unsigned SymbolTableEntrySize = is64Bit() ? - sizeof(MachOObjectFile64Le::SymbolTableEntry) : - sizeof(MachOObjectFile32Le::SymbolTableEntry); +static unsigned +getScatteredRelocationAddress(const MachO::any_relocation_info &RE) { + return RE.r_word0 & 0xffffff; +} - uint64_t Offset = SymbolTableOffset + Index * SymbolTableEntrySize; - StringRef Data = getData(Offset, SymbolTableEntrySize); - return reinterpret_cast(Data.data()); +static bool getPlainRelocationPCRel(const MachOObjectFile *O, + const MachO::any_relocation_info &RE) { + if (O->isLittleEndian()) + return (RE.r_word1 >> 24) & 1; + return (RE.r_word1 >> 7) & 1; } -error_code MachOObjectFileBase::getSymbolNext(DataRefImpl DRI, - SymbolRef &Result) const { - DRI.d.b++; - moveToNextSymbol(DRI); - Result = SymbolRef(DRI, this); - return object_error::success; +static bool +getScatteredRelocationPCRel(const MachOObjectFile *O, + const MachO::any_relocation_info &RE) { + return (RE.r_word0 >> 30) & 1; } -error_code MachOObjectFileBase::getSymbolName(DataRefImpl DRI, - StringRef &Result) const { - const LoadCommand *Command = getLoadCommandInfo(DRI.d.a); - const SymtabLoadCommand *SymtabLoadCmd = - reinterpret_cast(Command); +static unsigned getPlainRelocationLength(const MachOObjectFile *O, + const MachO::any_relocation_info &RE) { + if (O->isLittleEndian()) + return (RE.r_word1 >> 25) & 3; + return (RE.r_word1 >> 5) & 3; +} - StringRef StringTable = getData(SymtabLoadCmd->StringTableOffset, - SymtabLoadCmd->StringTableSize); +static unsigned +getScatteredRelocationLength(const MachO::any_relocation_info &RE) { + return (RE.r_word0 >> 28) & 3; +} - const SymbolTableEntryBase *Entry = - getSymbolTableEntryBase(DRI, SymtabLoadCmd); - uint32_t StringIndex = Entry->StringIndex; +static unsigned getPlainRelocationType(const MachOObjectFile *O, + const MachO::any_relocation_info &RE) { + if (O->isLittleEndian()) + return RE.r_word1 >> 28; + return RE.r_word1 & 0xf; +} - const char *Start = &StringTable.data()[StringIndex]; - Result = StringRef(Start); +static unsigned +getScatteredRelocationType(const MachO::any_relocation_info &RE) { + return (RE.r_word0 >> 24) & 0xf; +} - return object_error::success; +static uint32_t getSectionFlags(const MachOObjectFile *O, + DataRefImpl Sec) { + if (O->is64Bit()) { + MachO::section_64 Sect = O->getSection64(Sec); + return Sect.flags; + } + MachO::section Sect = O->getSection(Sec); + return Sect.flags; } -error_code MachOObjectFileBase::getSymbolNMTypeChar(DataRefImpl DRI, - char &Result) const { - const SymbolTableEntryBase *Entry = getSymbolTableEntryBase(DRI); - uint8_t Type = Entry->Type; - uint16_t Flags = Entry->Flags; +MachOObjectFile::MachOObjectFile(MemoryBuffer *Object, + bool IsLittleEndian, bool Is64bits, + error_code &ec) + : ObjectFile(getMachOType(IsLittleEndian, Is64bits), Object), + SymtabLoadCmd(NULL), DysymtabLoadCmd(NULL), DataInCodeLoadCmd(NULL) { + uint32_t LoadCommandCount = this->getHeader().ncmds; + MachO::LoadCommandType SegmentLoadType = is64Bit() ? + MachO::LC_SEGMENT_64 : MachO::LC_SEGMENT; + + MachOObjectFile::LoadCommandInfo Load = getFirstLoadCommandInfo(); + for (unsigned I = 0; ; ++I) { + if (Load.C.cmd == MachO::LC_SYMTAB) { + assert(!SymtabLoadCmd && "Multiple symbol tables"); + SymtabLoadCmd = Load.Ptr; + } else if (Load.C.cmd == MachO::LC_DYSYMTAB) { + assert(!DysymtabLoadCmd && "Multiple dynamic symbol tables"); + DysymtabLoadCmd = Load.Ptr; + } else if (Load.C.cmd == MachO::LC_DATA_IN_CODE) { + assert(!DataInCodeLoadCmd && "Multiple data in code tables"); + DataInCodeLoadCmd = Load.Ptr; + } else if (Load.C.cmd == SegmentLoadType) { + uint32_t NumSections = getSegmentLoadCommandNumSections(this, Load); + for (unsigned J = 0; J < NumSections; ++J) { + const char *Sec = getSectionPtr(this, Load, J); + Sections.push_back(Sec); + } + } - char Char; - switch (Type & macho::STF_TypeMask) { - case macho::STT_Undefined: - Char = 'u'; - break; - case macho::STT_Absolute: - case macho::STT_Section: - Char = 's'; - break; - default: - Char = '?'; + if (I == LoadCommandCount - 1) break; + else + Load = getNextLoadCommandInfo(Load); } +} - if (Flags & (macho::STF_External | macho::STF_PrivateExtern)) - Char = toupper(static_cast(Char)); - Result = Char; +error_code MachOObjectFile::getSymbolNext(DataRefImpl Symb, + SymbolRef &Res) const { + unsigned SymbolTableEntrySize = is64Bit() ? + sizeof(MachO::nlist_64) : + sizeof(MachO::nlist); + Symb.p += SymbolTableEntrySize; + Res = SymbolRef(Symb, this); return object_error::success; } -error_code MachOObjectFileBase::getSymbolFlags(DataRefImpl DRI, - uint32_t &Result) const { - const SymbolTableEntryBase *Entry = getSymbolTableEntryBase(DRI); - uint8_t MachOType = Entry->Type; - uint16_t MachOFlags = Entry->Flags; - - // TODO: Correctly set SF_ThreadLocal - Result = SymbolRef::SF_None; - - if ((MachOType & MachO::NlistMaskType) == MachO::NListTypeUndefined) - Result |= SymbolRef::SF_Undefined; - - if (MachOFlags & macho::STF_StabsEntryMask) - Result |= SymbolRef::SF_FormatSpecific; +error_code MachOObjectFile::getSymbolName(DataRefImpl Symb, + StringRef &Res) const { + StringRef StringTable = getStringTableData(); + nlist_base Entry = getSymbolTableEntryBase(this, Symb); + const char *Start = &StringTable.data()[Entry.n_strx]; + Res = StringRef(Start); + return object_error::success; +} - if (MachOType & MachO::NlistMaskExternal) { - Result |= SymbolRef::SF_Global; - if ((MachOType & MachO::NlistMaskType) == MachO::NListTypeUndefined) - Result |= SymbolRef::SF_Common; +error_code MachOObjectFile::getSymbolAddress(DataRefImpl Symb, + uint64_t &Res) const { + if (is64Bit()) { + MachO::nlist_64 Entry = getSymbol64TableEntry(Symb); + Res = Entry.n_value; + } else { + MachO::nlist Entry = getSymbolTableEntry(Symb); + Res = Entry.n_value; } + return object_error::success; +} - if (MachOFlags & (MachO::NListDescWeakRef | MachO::NListDescWeakDef)) - Result |= SymbolRef::SF_Weak; +error_code +MachOObjectFile::getSymbolFileOffset(DataRefImpl Symb, + uint64_t &Res) const { + nlist_base Entry = getSymbolTableEntryBase(this, Symb); + getSymbolAddress(Symb, Res); + if (Entry.n_sect) { + uint64_t Delta; + DataRefImpl SecRel; + SecRel.d.a = Entry.n_sect-1; + if (is64Bit()) { + MachO::section_64 Sec = getSection64(SecRel); + Delta = Sec.offset - Sec.addr; + } else { + MachO::section Sec = getSection(SecRel); + Delta = Sec.offset - Sec.addr; + } - if ((MachOType & MachO::NlistMaskType) == MachO::NListTypeAbsolute) - Result |= SymbolRef::SF_Absolute; + Res += Delta; + } return object_error::success; } -error_code MachOObjectFileBase::getSymbolSection(DataRefImpl Symb, - section_iterator &Res) const { - const SymbolTableEntryBase *Entry = getSymbolTableEntryBase(Symb); - uint8_t index = Entry->SectionIndex; - - if (index == 0) - Res = end_sections(); - else - Res = section_iterator(SectionRef(Sections[index-1], this)); +error_code MachOObjectFile::getSymbolAlignment(DataRefImpl DRI, + uint32_t &Result) const { + uint32_t flags; + this->getSymbolFlags(DRI, flags); + if (flags & SymbolRef::SF_Common) { + nlist_base Entry = getSymbolTableEntryBase(this, DRI); + Result = 1 << MachO::GET_COMM_ALIGN(Entry.n_desc); + } else { + Result = 0; + } + return object_error::success; +} +error_code MachOObjectFile::getSymbolSize(DataRefImpl DRI, + uint64_t &Result) const { + uint64_t BeginOffset; + uint64_t EndOffset = 0; + uint8_t SectionIndex; + + nlist_base Entry = getSymbolTableEntryBase(this, DRI); + uint64_t Value; + getSymbolAddress(DRI, Value); + + BeginOffset = Value; + + SectionIndex = Entry.n_sect; + if (!SectionIndex) { + uint32_t flags = SymbolRef::SF_None; + this->getSymbolFlags(DRI, flags); + if (flags & SymbolRef::SF_Common) + Result = Value; + else + Result = UnknownAddressOrSize; + return object_error::success; + } + // Unfortunately symbols are unsorted so we need to touch all + // symbols from load command + error_code ec; + for (symbol_iterator I = begin_symbols(), E = end_symbols(); I != E; + I.increment(ec)) { + DataRefImpl DRI = I->getRawDataRefImpl(); + Entry = getSymbolTableEntryBase(this, DRI); + getSymbolAddress(DRI, Value); + if (Entry.n_sect == SectionIndex && Value > BeginOffset) + if (!EndOffset || Value < EndOffset) + EndOffset = Value; + } + if (!EndOffset) { + uint64_t Size; + DataRefImpl Sec; + Sec.d.a = SectionIndex-1; + getSectionSize(Sec, Size); + getSectionAddress(Sec, EndOffset); + EndOffset += Size; + } + Result = EndOffset - BeginOffset; return object_error::success; } -error_code MachOObjectFileBase::getSymbolType(DataRefImpl Symb, - SymbolRef::Type &Res) const { - const SymbolTableEntryBase *Entry = getSymbolTableEntryBase(Symb); - uint8_t n_type = Entry->Type; +error_code MachOObjectFile::getSymbolType(DataRefImpl Symb, + SymbolRef::Type &Res) const { + nlist_base Entry = getSymbolTableEntryBase(this, Symb); + uint8_t n_type = Entry.n_type; Res = SymbolRef::ST_Other; // If this is a STAB debugging symbol, we can do nothing more. - if (n_type & MachO::NlistMaskStab) { + if (n_type & MachO::N_STAB) { Res = SymbolRef::ST_Debug; return object_error::success; } - switch (n_type & MachO::NlistMaskType) { - case MachO::NListTypeUndefined : + switch (n_type & MachO::N_TYPE) { + case MachO::N_UNDF : Res = SymbolRef::ST_Unknown; break; - case MachO::NListTypeSection : + case MachO::N_SECT : Res = SymbolRef::ST_Function; break; } return object_error::success; } -error_code MachOObjectFileBase::getSymbolValue(DataRefImpl Symb, - uint64_t &Val) const { - report_fatal_error("getSymbolValue unimplemented in MachOObjectFileBase"); -} +error_code MachOObjectFile::getSymbolFlags(DataRefImpl DRI, + uint32_t &Result) const { + nlist_base Entry = getSymbolTableEntryBase(this, DRI); -symbol_iterator MachOObjectFileBase::begin_symbols() const { - // DRI.d.a = segment number; DRI.d.b = symbol index. - DataRefImpl DRI; - moveToNextSymbol(DRI); - return symbol_iterator(SymbolRef(DRI, this)); -} + uint8_t MachOType = Entry.n_type; + uint16_t MachOFlags = Entry.n_desc; -symbol_iterator MachOObjectFileBase::end_symbols() const { - DataRefImpl DRI; - DRI.d.a = getHeader()->NumLoadCommands; - return symbol_iterator(SymbolRef(DRI, this)); -} + // TODO: Correctly set SF_ThreadLocal + Result = SymbolRef::SF_None; -symbol_iterator MachOObjectFileBase::begin_dynamic_symbols() const { - // TODO: implement - report_fatal_error("Dynamic symbols unimplemented in MachOObjectFileBase"); -} + if ((MachOType & MachO::N_TYPE) == MachO::N_UNDF) + Result |= SymbolRef::SF_Undefined; -symbol_iterator MachOObjectFileBase::end_dynamic_symbols() const { - // TODO: implement - report_fatal_error("Dynamic symbols unimplemented in MachOObjectFileBase"); -} + if (MachOType & MachO::N_STAB) + Result |= SymbolRef::SF_FormatSpecific; -library_iterator MachOObjectFileBase::begin_libraries_needed() const { - // TODO: implement - report_fatal_error("Needed libraries unimplemented in MachOObjectFileBase"); -} + if (MachOType & MachO::N_EXT) { + Result |= SymbolRef::SF_Global; + if ((MachOType & MachO::N_TYPE) == MachO::N_UNDF) { + uint64_t Value; + getSymbolAddress(DRI, Value); + if (Value) + Result |= SymbolRef::SF_Common; + } + } -library_iterator MachOObjectFileBase::end_libraries_needed() const { - // TODO: implement - report_fatal_error("Needed libraries unimplemented in MachOObjectFileBase"); -} + if (MachOFlags & (MachO::N_WEAK_REF | MachO::N_WEAK_DEF)) + Result |= SymbolRef::SF_Weak; -StringRef MachOObjectFileBase::getLoadName() const { - // TODO: Implement - report_fatal_error("get_load_name() unimplemented in MachOObjectFileBase"); + if ((MachOType & MachO::N_TYPE) == MachO::N_ABS) + Result |= SymbolRef::SF_Absolute; + + return object_error::success; } -/*===-- Sections ----------------------------------------------------------===*/ +error_code +MachOObjectFile::getSymbolSection(DataRefImpl Symb, + section_iterator &Res) const { + nlist_base Entry = getSymbolTableEntryBase(this, Symb); + uint8_t index = Entry.n_sect; -std::size_t MachOObjectFileBase::getSectionIndex(DataRefImpl Sec) const { - SectionList::const_iterator loc = - std::find(Sections.begin(), Sections.end(), Sec); - assert(loc != Sections.end() && "Sec is not a valid section!"); - return std::distance(Sections.begin(), loc); + if (index == 0) { + Res = end_sections(); + } else { + DataRefImpl DRI; + DRI.d.a = index - 1; + Res = section_iterator(SectionRef(DRI, this)); + } + + return object_error::success; } -const MachOObjectFileBase::SectionBase* -MachOObjectFileBase::getSectionBase(DataRefImpl DRI) const { - const LoadCommand *Command = getLoadCommandInfo(DRI.d.a); - uintptr_t CommandAddr = reinterpret_cast(Command); +error_code MachOObjectFile::getSymbolValue(DataRefImpl Symb, + uint64_t &Val) const { + report_fatal_error("getSymbolValue unimplemented in MachOObjectFile"); +} - bool Is64 = is64Bit(); - unsigned SegmentLoadSize = - Is64 ? sizeof(MachOObjectFile64Le::SegmentLoadCommand) : - sizeof(MachOObjectFile32Le::SegmentLoadCommand); - unsigned SectionSize = Is64 ? sizeof(MachOObjectFile64Le::Section) : - sizeof(MachOObjectFile32Le::Section); +error_code MachOObjectFile::getSectionNext(DataRefImpl Sec, + SectionRef &Res) const { + Sec.d.a++; + Res = SectionRef(Sec, this); + return object_error::success; +} - uintptr_t SectionAddr = CommandAddr + SegmentLoadSize + DRI.d.b * SectionSize; - return reinterpret_cast(SectionAddr); +error_code +MachOObjectFile::getSectionName(DataRefImpl Sec, StringRef &Result) const { + ArrayRef Raw = getSectionRawName(Sec); + Result = parseSegmentOrSectionName(Raw.data()); + return object_error::success; } -static StringRef parseSegmentOrSectionName(const char *P) { - if (P[15] == 0) - // Null terminated. - return P; - // Not null terminated, so this is a 16 char string. - return StringRef(P, 16); +error_code +MachOObjectFile::getSectionAddress(DataRefImpl Sec, uint64_t &Res) const { + if (is64Bit()) { + MachO::section_64 Sect = getSection64(Sec); + Res = Sect.addr; + } else { + MachO::section Sect = getSection(Sec); + Res = Sect.addr; + } + return object_error::success; } -ArrayRef MachOObjectFileBase::getSectionRawName(DataRefImpl DRI) const { - const SectionBase *Base = getSectionBase(DRI); - return ArrayRef(Base->Name); +error_code +MachOObjectFile::getSectionSize(DataRefImpl Sec, uint64_t &Res) const { + if (is64Bit()) { + MachO::section_64 Sect = getSection64(Sec); + Res = Sect.size; + } else { + MachO::section Sect = getSection(Sec); + Res = Sect.size; + } + + return object_error::success; } -error_code MachOObjectFileBase::getSectionName(DataRefImpl DRI, - StringRef &Result) const { - ArrayRef Raw = getSectionRawName(DRI); - Result = parseSegmentOrSectionName(Raw.data()); +error_code +MachOObjectFile::getSectionContents(DataRefImpl Sec, StringRef &Res) const { + uint32_t Offset; + uint64_t Size; + + if (is64Bit()) { + MachO::section_64 Sect = getSection64(Sec); + Offset = Sect.offset; + Size = Sect.size; + } else { + MachO::section Sect = getSection(Sec); + Offset = Sect.offset; + Size = Sect.size; + } + + Res = this->getData().substr(Offset, Size); return object_error::success; } -ArrayRef -MachOObjectFileBase::getSectionRawFinalSegmentName(DataRefImpl Sec) const { - const SectionBase *Base = getSectionBase(Sec); - return ArrayRef(Base->SegmentName); +error_code +MachOObjectFile::getSectionAlignment(DataRefImpl Sec, uint64_t &Res) const { + uint32_t Align; + if (is64Bit()) { + MachO::section_64 Sect = getSection64(Sec); + Align = Sect.align; + } else { + MachO::section Sect = getSection(Sec); + Align = Sect.align; + } + + Res = uint64_t(1) << Align; + return object_error::success; } -StringRef -MachOObjectFileBase::getSectionFinalSegmentName(DataRefImpl DRI) const { - ArrayRef Raw = getSectionRawFinalSegmentName(DRI); - return parseSegmentOrSectionName(Raw.data()); +error_code +MachOObjectFile::isSectionText(DataRefImpl Sec, bool &Res) const { + uint32_t Flags = getSectionFlags(this, Sec); + Res = Flags & MachO::S_ATTR_PURE_INSTRUCTIONS; + return object_error::success; } -error_code MachOObjectFileBase::isSectionData(DataRefImpl DRI, - bool &Result) const { +error_code MachOObjectFile::isSectionData(DataRefImpl DRI, bool &Result) const { // FIXME: Unimplemented. Result = false; return object_error::success; } -error_code MachOObjectFileBase::isSectionBSS(DataRefImpl DRI, - bool &Result) const { +error_code MachOObjectFile::isSectionBSS(DataRefImpl DRI, bool &Result) const { // FIXME: Unimplemented. Result = false; return object_error::success; } error_code -MachOObjectFileBase::isSectionRequiredForExecution(DataRefImpl Sec, - bool &Result) const { +MachOObjectFile::isSectionRequiredForExecution(DataRefImpl Sec, + bool &Result) const { // FIXME: Unimplemented. Result = true; return object_error::success; } -error_code MachOObjectFileBase::isSectionVirtual(DataRefImpl Sec, - bool &Result) const { +error_code MachOObjectFile::isSectionVirtual(DataRefImpl Sec, + bool &Result) const { // FIXME: Unimplemented. Result = false; return object_error::success; } -error_code MachOObjectFileBase::isSectionReadOnlyData(DataRefImpl Sec, - bool &Result) const { +error_code +MachOObjectFile::isSectionZeroInit(DataRefImpl Sec, bool &Res) const { + uint32_t Flags = getSectionFlags(this, Sec); + unsigned SectionType = Flags & MachO::SECTION_TYPE; + Res = SectionType == MachO::S_ZEROFILL || + SectionType == MachO::S_GB_ZEROFILL; + return object_error::success; +} + +error_code MachOObjectFile::isSectionReadOnlyData(DataRefImpl Sec, + bool &Result) const { // Consider using the code from isSectionText to look for __const sections. // Alternately, emit S_ATTR_PURE_INSTRUCTIONS and/or S_ATTR_SOME_INSTRUCTIONS // to use section attributes to distinguish code from data. @@ -395,177 +775,834 @@ error_code MachOObjectFileBase::isSectionReadOnlyData(DataRefImpl Sec, return object_error::success; } -relocation_iterator MachOObjectFileBase::getSectionRelBegin(DataRefImpl Sec) const { - DataRefImpl ret; - ret.d.b = getSectionIndex(Sec); - return relocation_iterator(RelocationRef(ret, this)); +error_code +MachOObjectFile::sectionContainsSymbol(DataRefImpl Sec, DataRefImpl Symb, + bool &Result) const { + SymbolRef::Type ST; + this->getSymbolType(Symb, ST); + if (ST == SymbolRef::ST_Unknown) { + Result = false; + return object_error::success; + } + + uint64_t SectBegin, SectEnd; + getSectionAddress(Sec, SectBegin); + getSectionSize(Sec, SectEnd); + SectEnd += SectBegin; + + uint64_t SymAddr; + getSymbolAddress(Symb, SymAddr); + Result = (SymAddr >= SectBegin) && (SymAddr < SectEnd); + + return object_error::success; } -section_iterator MachOObjectFileBase::end_sections() const { - DataRefImpl DRI; - DRI.d.a = getHeader()->NumLoadCommands; - return section_iterator(SectionRef(DRI, this)); +relocation_iterator MachOObjectFile::section_rel_begin(DataRefImpl Sec) const { + uint32_t Offset; + if (is64Bit()) { + MachO::section_64 Sect = getSection64(Sec); + Offset = Sect.reloff; + } else { + MachO::section Sect = getSection(Sec); + Offset = Sect.reloff; + } + + DataRefImpl Ret; + Ret.p = reinterpret_cast(getPtr(this, Offset)); + return relocation_iterator(RelocationRef(Ret, this)); } -/*===-- Relocations -------------------------------------------------------===*/ +relocation_iterator +MachOObjectFile::section_rel_end(DataRefImpl Sec) const { + uint32_t Offset; + uint32_t Num; + if (is64Bit()) { + MachO::section_64 Sect = getSection64(Sec); + Offset = Sect.reloff; + Num = Sect.nreloc; + } else { + MachO::section Sect = getSection(Sec); + Offset = Sect.reloff; + Num = Sect.nreloc; + } + + const MachO::any_relocation_info *P = + reinterpret_cast(getPtr(this, Offset)); + + DataRefImpl Ret; + Ret.p = reinterpret_cast(P + Num); + return relocation_iterator(RelocationRef(Ret, this)); +} -error_code MachOObjectFileBase::getRelocationNext(DataRefImpl Rel, - RelocationRef &Res) const { - ++Rel.d.a; +error_code MachOObjectFile::getRelocationNext(DataRefImpl Rel, + RelocationRef &Res) const { + const MachO::any_relocation_info *P = + reinterpret_cast(Rel.p); + Rel.p = reinterpret_cast(P + 1); Res = RelocationRef(Rel, this); return object_error::success; } -// Helper to advance a section or symbol iterator multiple increments at a time. -template -error_code advance(T &it, size_t Val) { - error_code ec; - while (Val--) { - it.increment(ec); - } - return ec; +error_code +MachOObjectFile::getRelocationAddress(DataRefImpl Rel, uint64_t &Res) const { + report_fatal_error("getRelocationAddress not implemented in MachOObjectFile"); } -template -void advanceTo(T &it, size_t Val) { - if (error_code ec = advance(it, Val)) - report_fatal_error(ec.message()); +error_code MachOObjectFile::getRelocationOffset(DataRefImpl Rel, + uint64_t &Res) const { + MachO::any_relocation_info RE = getRelocation(Rel); + Res = getAnyRelocationAddress(RE); + return object_error::success; } -void -MachOObjectFileBase::printRelocationTargetName(const RelocationEntry *RE, - raw_string_ostream &fmt) const { - unsigned Arch = getArch(); - bool isScattered = (Arch != Triple::x86_64) && - (RE->Word0 & macho::RF_Scattered); +symbol_iterator +MachOObjectFile::getRelocationSymbol(DataRefImpl Rel) const { + MachO::any_relocation_info RE = getRelocation(Rel); + uint32_t SymbolIdx = getPlainRelocationSymbolNum(RE); + bool isExtern = getPlainRelocationExternal(RE); + if (!isExtern) + return end_symbols(); - // Target of a scattered relocation is an address. In the interest of - // generating pretty output, scan through the symbol table looking for a - // symbol that aligns with that address. If we find one, print it. - // Otherwise, we just print the hex address of the target. - if (isScattered) { - uint32_t Val = RE->Word1; + MachO::symtab_command S = getSymtabLoadCommand(); + unsigned SymbolTableEntrySize = is64Bit() ? + sizeof(MachO::nlist_64) : + sizeof(MachO::nlist); + uint64_t Offset = S.symoff + SymbolIdx * SymbolTableEntrySize; + DataRefImpl Sym; + Sym.p = reinterpret_cast(getPtr(this, Offset)); + return symbol_iterator(SymbolRef(Sym, this)); +} - error_code ec; - for (symbol_iterator SI = begin_symbols(), SE = end_symbols(); SI != SE; - SI.increment(ec)) { - if (ec) report_fatal_error(ec.message()); +error_code MachOObjectFile::getRelocationType(DataRefImpl Rel, + uint64_t &Res) const { + MachO::any_relocation_info RE = getRelocation(Rel); + Res = getAnyRelocationType(RE); + return object_error::success; +} - uint64_t Addr; - StringRef Name; +error_code +MachOObjectFile::getRelocationTypeName(DataRefImpl Rel, + SmallVectorImpl &Result) const { + StringRef res; + uint64_t RType; + getRelocationType(Rel, RType); + + unsigned Arch = this->getArch(); + + switch (Arch) { + case Triple::x86: { + static const char *const Table[] = { + "GENERIC_RELOC_VANILLA", + "GENERIC_RELOC_PAIR", + "GENERIC_RELOC_SECTDIFF", + "GENERIC_RELOC_PB_LA_PTR", + "GENERIC_RELOC_LOCAL_SECTDIFF", + "GENERIC_RELOC_TLV" }; + + if (RType > 5) + res = "Unknown"; + else + res = Table[RType]; + break; + } + case Triple::x86_64: { + static const char *const Table[] = { + "X86_64_RELOC_UNSIGNED", + "X86_64_RELOC_SIGNED", + "X86_64_RELOC_BRANCH", + "X86_64_RELOC_GOT_LOAD", + "X86_64_RELOC_GOT", + "X86_64_RELOC_SUBTRACTOR", + "X86_64_RELOC_SIGNED_1", + "X86_64_RELOC_SIGNED_2", + "X86_64_RELOC_SIGNED_4", + "X86_64_RELOC_TLV" }; + + if (RType > 9) + res = "Unknown"; + else + res = Table[RType]; + break; + } + case Triple::arm: { + static const char *const Table[] = { + "ARM_RELOC_VANILLA", + "ARM_RELOC_PAIR", + "ARM_RELOC_SECTDIFF", + "ARM_RELOC_LOCAL_SECTDIFF", + "ARM_RELOC_PB_LA_PTR", + "ARM_RELOC_BR24", + "ARM_THUMB_RELOC_BR22", + "ARM_THUMB_32BIT_BRANCH", + "ARM_RELOC_HALF", + "ARM_RELOC_HALF_SECTDIFF" }; + + if (RType > 9) + res = "Unknown"; + else + res = Table[RType]; + break; + } + case Triple::ppc: { + static const char *const Table[] = { + "PPC_RELOC_VANILLA", + "PPC_RELOC_PAIR", + "PPC_RELOC_BR14", + "PPC_RELOC_BR24", + "PPC_RELOC_HI16", + "PPC_RELOC_LO16", + "PPC_RELOC_HA16", + "PPC_RELOC_LO14", + "PPC_RELOC_SECTDIFF", + "PPC_RELOC_PB_LA_PTR", + "PPC_RELOC_HI16_SECTDIFF", + "PPC_RELOC_LO16_SECTDIFF", + "PPC_RELOC_HA16_SECTDIFF", + "PPC_RELOC_JBSR", + "PPC_RELOC_LO14_SECTDIFF", + "PPC_RELOC_LOCAL_SECTDIFF" }; + + if (RType > 15) + res = "Unknown"; + else + res = Table[RType]; + break; + } + case Triple::UnknownArch: + res = "Unknown"; + break; + } + Result.append(res.begin(), res.end()); + return object_error::success; +} - if ((ec = SI->getAddress(Addr))) - report_fatal_error(ec.message()); - if (Addr != Val) continue; - if ((ec = SI->getName(Name))) - report_fatal_error(ec.message()); - fmt << Name; - return; +error_code +MachOObjectFile::getRelocationValueString(DataRefImpl Rel, + SmallVectorImpl &Result) const { + MachO::any_relocation_info RE = getRelocation(Rel); + + unsigned Arch = this->getArch(); + + std::string fmtbuf; + raw_string_ostream fmt(fmtbuf); + unsigned Type = this->getAnyRelocationType(RE); + bool IsPCRel = this->getAnyRelocationPCRel(RE); + + // Determine any addends that should be displayed with the relocation. + // These require decoding the relocation type, which is triple-specific. + + // X86_64 has entirely custom relocation types. + if (Arch == Triple::x86_64) { + bool isPCRel = getAnyRelocationPCRel(RE); + + switch (Type) { + case MachO::X86_64_RELOC_GOT_LOAD: + case MachO::X86_64_RELOC_GOT: { + printRelocationTargetName(this, RE, fmt); + fmt << "@GOT"; + if (isPCRel) fmt << "PCREL"; + break; + } + case MachO::X86_64_RELOC_SUBTRACTOR: { + DataRefImpl RelNext = Rel; + RelNext.d.a++; + MachO::any_relocation_info RENext = getRelocation(RelNext); + + // X86_64_RELOC_SUBTRACTOR must be followed by a relocation of type + // X86_64_RELOC_UNSIGNED. + // NOTE: Scattered relocations don't exist on x86_64. + unsigned RType = getAnyRelocationType(RENext); + if (RType != MachO::X86_64_RELOC_UNSIGNED) + report_fatal_error("Expected X86_64_RELOC_UNSIGNED after " + "X86_64_RELOC_SUBTRACTOR."); + + // The X86_64_RELOC_UNSIGNED contains the minuend symbol; + // X86_64_RELOC_SUBTRACTOR contains the subtrahend. + printRelocationTargetName(this, RENext, fmt); + fmt << "-"; + printRelocationTargetName(this, RE, fmt); + break; + } + case MachO::X86_64_RELOC_TLV: + printRelocationTargetName(this, RE, fmt); + fmt << "@TLV"; + if (isPCRel) fmt << "P"; + break; + case MachO::X86_64_RELOC_SIGNED_1: + printRelocationTargetName(this, RE, fmt); + fmt << "-1"; + break; + case MachO::X86_64_RELOC_SIGNED_2: + printRelocationTargetName(this, RE, fmt); + fmt << "-2"; + break; + case MachO::X86_64_RELOC_SIGNED_4: + printRelocationTargetName(this, RE, fmt); + fmt << "-4"; + break; + default: + printRelocationTargetName(this, RE, fmt); + break; + } + // X86 and ARM share some relocation types in common. + } else if (Arch == Triple::x86 || Arch == Triple::arm || + Arch == Triple::ppc) { + // Generic relocation types... + switch (Type) { + case MachO::GENERIC_RELOC_PAIR: // prints no info + return object_error::success; + case MachO::GENERIC_RELOC_SECTDIFF: { + DataRefImpl RelNext = Rel; + RelNext.d.a++; + MachO::any_relocation_info RENext = getRelocation(RelNext); + + // X86 sect diff's must be followed by a relocation of type + // GENERIC_RELOC_PAIR. + unsigned RType = getAnyRelocationType(RENext); + + if (RType != MachO::GENERIC_RELOC_PAIR) + report_fatal_error("Expected GENERIC_RELOC_PAIR after " + "GENERIC_RELOC_SECTDIFF."); + + printRelocationTargetName(this, RE, fmt); + fmt << "-"; + printRelocationTargetName(this, RENext, fmt); + break; + } } - // If we couldn't find a symbol that this relocation refers to, try - // to find a section beginning instead. - for (section_iterator SI = begin_sections(), SE = end_sections(); SI != SE; - SI.increment(ec)) { - if (ec) report_fatal_error(ec.message()); + if (Arch == Triple::x86 || Arch == Triple::ppc) { + switch (Type) { + case MachO::GENERIC_RELOC_LOCAL_SECTDIFF: { + DataRefImpl RelNext = Rel; + RelNext.d.a++; + MachO::any_relocation_info RENext = getRelocation(RelNext); + + // X86 sect diff's must be followed by a relocation of type + // GENERIC_RELOC_PAIR. + unsigned RType = getAnyRelocationType(RENext); + if (RType != MachO::GENERIC_RELOC_PAIR) + report_fatal_error("Expected GENERIC_RELOC_PAIR after " + "GENERIC_RELOC_LOCAL_SECTDIFF."); + + printRelocationTargetName(this, RE, fmt); + fmt << "-"; + printRelocationTargetName(this, RENext, fmt); + break; + } + case MachO::GENERIC_RELOC_TLV: { + printRelocationTargetName(this, RE, fmt); + fmt << "@TLV"; + if (IsPCRel) fmt << "P"; + break; + } + default: + printRelocationTargetName(this, RE, fmt); + } + } else { // ARM-specific relocations + switch (Type) { + case MachO::ARM_RELOC_HALF: + case MachO::ARM_RELOC_HALF_SECTDIFF: { + // Half relocations steal a bit from the length field to encode + // whether this is an upper16 or a lower16 relocation. + bool isUpper = getAnyRelocationLength(RE) >> 1; + + if (isUpper) + fmt << ":upper16:("; + else + fmt << ":lower16:("; + printRelocationTargetName(this, RE, fmt); + + DataRefImpl RelNext = Rel; + RelNext.d.a++; + MachO::any_relocation_info RENext = getRelocation(RelNext); + + // ARM half relocs must be followed by a relocation of type + // ARM_RELOC_PAIR. + unsigned RType = getAnyRelocationType(RENext); + if (RType != MachO::ARM_RELOC_PAIR) + report_fatal_error("Expected ARM_RELOC_PAIR after " + "ARM_RELOC_HALF"); + + // NOTE: The half of the target virtual address is stashed in the + // address field of the secondary relocation, but we can't reverse + // engineer the constant offset from it without decoding the movw/movt + // instruction to find the other half in its immediate field. + + // ARM_RELOC_HALF_SECTDIFF encodes the second section in the + // symbol/section pointer of the follow-on relocation. + if (Type == MachO::ARM_RELOC_HALF_SECTDIFF) { + fmt << "-"; + printRelocationTargetName(this, RENext, fmt); + } + + fmt << ")"; + break; + } + default: { + printRelocationTargetName(this, RE, fmt); + } + } + } + } else + printRelocationTargetName(this, RE, fmt); - uint64_t Addr; - StringRef Name; + fmt.flush(); + Result.append(fmtbuf.begin(), fmtbuf.end()); + return object_error::success; +} - if ((ec = SI->getAddress(Addr))) - report_fatal_error(ec.message()); - if (Addr != Val) continue; - if ((ec = SI->getName(Name))) - report_fatal_error(ec.message()); - fmt << Name; - return; - } +error_code +MachOObjectFile::getRelocationHidden(DataRefImpl Rel, bool &Result) const { + unsigned Arch = getArch(); + uint64_t Type; + getRelocationType(Rel, Type); - fmt << format("0x%x", Val); - return; + Result = false; + + // On arches that use the generic relocations, GENERIC_RELOC_PAIR + // is always hidden. + if (Arch == Triple::x86 || Arch == Triple::arm || Arch == Triple::ppc) { + if (Type == MachO::GENERIC_RELOC_PAIR) Result = true; + } else if (Arch == Triple::x86_64) { + // On x86_64, X86_64_RELOC_UNSIGNED is hidden only when it follows + // an X86_64_RELOC_SUBTRACTOR. + if (Type == MachO::X86_64_RELOC_UNSIGNED && Rel.d.a > 0) { + DataRefImpl RelPrev = Rel; + RelPrev.d.a--; + uint64_t PrevType; + getRelocationType(RelPrev, PrevType); + if (PrevType == MachO::X86_64_RELOC_SUBTRACTOR) + Result = true; + } } - StringRef S; - bool isExtern = (RE->Word1 >> 27) & 1; - uint32_t Val = RE->Word1 & 0xFFFFFF; + return object_error::success; +} - if (isExtern) { - symbol_iterator SI = begin_symbols(); - advanceTo(SI, Val); - SI->getName(S); - } else { - section_iterator SI = begin_sections(); - advanceTo(SI, Val); - SI->getName(S); - } +error_code MachOObjectFile::getLibraryNext(DataRefImpl LibData, + LibraryRef &Res) const { + report_fatal_error("Needed libraries unimplemented in MachOObjectFile"); +} - fmt << S; +error_code MachOObjectFile::getLibraryPath(DataRefImpl LibData, + StringRef &Res) const { + report_fatal_error("Needed libraries unimplemented in MachOObjectFile"); } -error_code MachOObjectFileBase::getLibraryNext(DataRefImpl LibData, - LibraryRef &Res) const { - report_fatal_error("Needed libraries unimplemented in MachOObjectFileBase"); +symbol_iterator MachOObjectFile::begin_symbols() const { + DataRefImpl DRI; + if (!SymtabLoadCmd) + return symbol_iterator(SymbolRef(DRI, this)); + + MachO::symtab_command Symtab = getSymtabLoadCommand(); + DRI.p = reinterpret_cast(getPtr(this, Symtab.symoff)); + return symbol_iterator(SymbolRef(DRI, this)); } -error_code MachOObjectFileBase::getLibraryPath(DataRefImpl LibData, - StringRef &Res) const { - report_fatal_error("Needed libraries unimplemented in MachOObjectFileBase"); +symbol_iterator MachOObjectFile::end_symbols() const { + DataRefImpl DRI; + if (!SymtabLoadCmd) + return symbol_iterator(SymbolRef(DRI, this)); + + MachO::symtab_command Symtab = getSymtabLoadCommand(); + unsigned SymbolTableEntrySize = is64Bit() ? + sizeof(MachO::nlist_64) : + sizeof(MachO::nlist); + unsigned Offset = Symtab.symoff + + Symtab.nsyms * SymbolTableEntrySize; + DRI.p = reinterpret_cast(getPtr(this, Offset)); + return symbol_iterator(SymbolRef(DRI, this)); } +symbol_iterator MachOObjectFile::begin_dynamic_symbols() const { + // TODO: implement + report_fatal_error("Dynamic symbols unimplemented in MachOObjectFile"); +} + +symbol_iterator MachOObjectFile::end_dynamic_symbols() const { + // TODO: implement + report_fatal_error("Dynamic symbols unimplemented in MachOObjectFile"); +} + +section_iterator MachOObjectFile::begin_sections() const { + DataRefImpl DRI; + return section_iterator(SectionRef(DRI, this)); +} -/*===-- Miscellaneous -----------------------------------------------------===*/ +section_iterator MachOObjectFile::end_sections() const { + DataRefImpl DRI; + DRI.d.a = Sections.size(); + return section_iterator(SectionRef(DRI, this)); +} -uint8_t MachOObjectFileBase::getBytesInAddress() const { +library_iterator MachOObjectFile::begin_libraries_needed() const { + // TODO: implement + report_fatal_error("Needed libraries unimplemented in MachOObjectFile"); +} + +library_iterator MachOObjectFile::end_libraries_needed() const { + // TODO: implement + report_fatal_error("Needed libraries unimplemented in MachOObjectFile"); +} + +uint8_t MachOObjectFile::getBytesInAddress() const { return is64Bit() ? 8 : 4; } -StringRef MachOObjectFileBase::getFileFormatName() const { +StringRef MachOObjectFile::getFileFormatName() const { + unsigned CPUType = getCPUType(this); if (!is64Bit()) { - switch (getHeader()->CPUType) { - case llvm::MachO::CPUTypeI386: + switch (CPUType) { + case llvm::MachO::CPU_TYPE_I386: return "Mach-O 32-bit i386"; - case llvm::MachO::CPUTypeARM: + case llvm::MachO::CPU_TYPE_ARM: return "Mach-O arm"; - case llvm::MachO::CPUTypePowerPC: + case llvm::MachO::CPU_TYPE_POWERPC: return "Mach-O 32-bit ppc"; default: - assert((getHeader()->CPUType & llvm::MachO::CPUArchABI64) == 0 && + assert((CPUType & llvm::MachO::CPU_ARCH_ABI64) == 0 && "64-bit object file when we're not 64-bit?"); return "Mach-O 32-bit unknown"; } } // Make sure the cpu type has the correct mask. - assert((getHeader()->CPUType & llvm::MachO::CPUArchABI64) - == llvm::MachO::CPUArchABI64 && - "32-bit object file when we're 64-bit?"); + assert((CPUType & llvm::MachO::CPU_ARCH_ABI64) + == llvm::MachO::CPU_ARCH_ABI64 && + "32-bit object file when we're 64-bit?"); - switch (getHeader()->CPUType) { - case llvm::MachO::CPUTypeX86_64: + switch (CPUType) { + case llvm::MachO::CPU_TYPE_X86_64: return "Mach-O 64-bit x86-64"; - case llvm::MachO::CPUTypePowerPC64: + case llvm::MachO::CPU_TYPE_POWERPC64: return "Mach-O 64-bit ppc64"; default: return "Mach-O 64-bit unknown"; } } -unsigned MachOObjectFileBase::getArch() const { - switch (getHeader()->CPUType) { - case llvm::MachO::CPUTypeI386: +Triple::ArchType MachOObjectFile::getArch(uint32_t CPUType) { + switch (CPUType) { + case llvm::MachO::CPU_TYPE_I386: return Triple::x86; - case llvm::MachO::CPUTypeX86_64: + case llvm::MachO::CPU_TYPE_X86_64: return Triple::x86_64; - case llvm::MachO::CPUTypeARM: + case llvm::MachO::CPU_TYPE_ARM: return Triple::arm; - case llvm::MachO::CPUTypePowerPC: + case llvm::MachO::CPU_TYPE_POWERPC: return Triple::ppc; - case llvm::MachO::CPUTypePowerPC64: + case llvm::MachO::CPU_TYPE_POWERPC64: return Triple::ppc64; default: return Triple::UnknownArch; } } +unsigned MachOObjectFile::getArch() const { + return getArch(getCPUType(this)); +} + +StringRef MachOObjectFile::getLoadName() const { + // TODO: Implement + report_fatal_error("get_load_name() unimplemented in MachOObjectFile"); +} + +relocation_iterator MachOObjectFile::section_rel_begin(unsigned Index) const { + DataRefImpl DRI; + DRI.d.a = Index; + return section_rel_begin(DRI); +} + +relocation_iterator MachOObjectFile::section_rel_end(unsigned Index) const { + DataRefImpl DRI; + DRI.d.a = Index; + return section_rel_end(DRI); +} + +dice_iterator MachOObjectFile::begin_dices() const { + DataRefImpl DRI; + if (!DataInCodeLoadCmd) + return dice_iterator(DiceRef(DRI, this)); + + MachO::linkedit_data_command DicLC = getDataInCodeLoadCommand(); + DRI.p = reinterpret_cast(getPtr(this, DicLC.dataoff)); + return dice_iterator(DiceRef(DRI, this)); +} + +dice_iterator MachOObjectFile::end_dices() const { + DataRefImpl DRI; + if (!DataInCodeLoadCmd) + return dice_iterator(DiceRef(DRI, this)); + + MachO::linkedit_data_command DicLC = getDataInCodeLoadCommand(); + unsigned Offset = DicLC.dataoff + DicLC.datasize; + DRI.p = reinterpret_cast(getPtr(this, Offset)); + return dice_iterator(DiceRef(DRI, this)); +} + +StringRef +MachOObjectFile::getSectionFinalSegmentName(DataRefImpl Sec) const { + ArrayRef Raw = getSectionRawFinalSegmentName(Sec); + return parseSegmentOrSectionName(Raw.data()); +} + +ArrayRef +MachOObjectFile::getSectionRawName(DataRefImpl Sec) const { + const section_base *Base = + reinterpret_cast(Sections[Sec.d.a]); + return ArrayRef(Base->sectname); +} + +ArrayRef +MachOObjectFile::getSectionRawFinalSegmentName(DataRefImpl Sec) const { + const section_base *Base = + reinterpret_cast(Sections[Sec.d.a]); + return ArrayRef(Base->segname); +} + +bool +MachOObjectFile::isRelocationScattered(const MachO::any_relocation_info &RE) + const { + if (getCPUType(this) == MachO::CPU_TYPE_X86_64) + return false; + return getPlainRelocationAddress(RE) & MachO::R_SCATTERED; +} + +unsigned MachOObjectFile::getPlainRelocationSymbolNum( + const MachO::any_relocation_info &RE) const { + if (isLittleEndian()) + return RE.r_word1 & 0xffffff; + return RE.r_word1 >> 8; +} + +bool MachOObjectFile::getPlainRelocationExternal( + const MachO::any_relocation_info &RE) const { + if (isLittleEndian()) + return (RE.r_word1 >> 27) & 1; + return (RE.r_word1 >> 4) & 1; +} + +bool MachOObjectFile::getScatteredRelocationScattered( + const MachO::any_relocation_info &RE) const { + return RE.r_word0 >> 31; +} + +uint32_t MachOObjectFile::getScatteredRelocationValue( + const MachO::any_relocation_info &RE) const { + return RE.r_word1; +} + +unsigned MachOObjectFile::getAnyRelocationAddress( + const MachO::any_relocation_info &RE) const { + if (isRelocationScattered(RE)) + return getScatteredRelocationAddress(RE); + return getPlainRelocationAddress(RE); +} + +unsigned MachOObjectFile::getAnyRelocationPCRel( + const MachO::any_relocation_info &RE) const { + if (isRelocationScattered(RE)) + return getScatteredRelocationPCRel(this, RE); + return getPlainRelocationPCRel(this, RE); +} + +unsigned MachOObjectFile::getAnyRelocationLength( + const MachO::any_relocation_info &RE) const { + if (isRelocationScattered(RE)) + return getScatteredRelocationLength(RE); + return getPlainRelocationLength(this, RE); +} + +unsigned +MachOObjectFile::getAnyRelocationType( + const MachO::any_relocation_info &RE) const { + if (isRelocationScattered(RE)) + return getScatteredRelocationType(RE); + return getPlainRelocationType(this, RE); +} + +SectionRef +MachOObjectFile::getRelocationSection( + const MachO::any_relocation_info &RE) const { + if (isRelocationScattered(RE) || getPlainRelocationExternal(RE)) + return *end_sections(); + unsigned SecNum = getPlainRelocationSymbolNum(RE) - 1; + DataRefImpl DRI; + DRI.d.a = SecNum; + return SectionRef(DRI, this); +} + +MachOObjectFile::LoadCommandInfo +MachOObjectFile::getFirstLoadCommandInfo() const { + MachOObjectFile::LoadCommandInfo Load; + + unsigned HeaderSize = is64Bit() ? sizeof(MachO::mach_header_64) : + sizeof(MachO::mach_header); + Load.Ptr = getPtr(this, HeaderSize); + Load.C = getStruct(this, Load.Ptr); + return Load; +} + +MachOObjectFile::LoadCommandInfo +MachOObjectFile::getNextLoadCommandInfo(const LoadCommandInfo &L) const { + MachOObjectFile::LoadCommandInfo Next; + Next.Ptr = L.Ptr + L.C.cmdsize; + Next.C = getStruct(this, Next.Ptr); + return Next; +} + +MachO::section MachOObjectFile::getSection(DataRefImpl DRI) const { + return getStruct(this, Sections[DRI.d.a]); +} + +MachO::section_64 MachOObjectFile::getSection64(DataRefImpl DRI) const { + return getStruct(this, Sections[DRI.d.a]); +} + +MachO::section MachOObjectFile::getSection(const LoadCommandInfo &L, + unsigned Index) const { + const char *Sec = getSectionPtr(this, L, Index); + return getStruct(this, Sec); +} + +MachO::section_64 MachOObjectFile::getSection64(const LoadCommandInfo &L, + unsigned Index) const { + const char *Sec = getSectionPtr(this, L, Index); + return getStruct(this, Sec); +} + +MachO::nlist +MachOObjectFile::getSymbolTableEntry(DataRefImpl DRI) const { + const char *P = reinterpret_cast(DRI.p); + return getStruct(this, P); +} + +MachO::nlist_64 +MachOObjectFile::getSymbol64TableEntry(DataRefImpl DRI) const { + const char *P = reinterpret_cast(DRI.p); + return getStruct(this, P); +} + +MachO::linkedit_data_command +MachOObjectFile::getLinkeditDataLoadCommand(const LoadCommandInfo &L) const { + return getStruct(this, L.Ptr); +} + +MachO::segment_command +MachOObjectFile::getSegmentLoadCommand(const LoadCommandInfo &L) const { + return getStruct(this, L.Ptr); +} + +MachO::segment_command_64 +MachOObjectFile::getSegment64LoadCommand(const LoadCommandInfo &L) const { + return getStruct(this, L.Ptr); +} + +MachO::linker_options_command +MachOObjectFile::getLinkerOptionsLoadCommand(const LoadCommandInfo &L) const { + return getStruct(this, L.Ptr); +} + +MachO::any_relocation_info +MachOObjectFile::getRelocation(DataRefImpl Rel) const { + const char *P = reinterpret_cast(Rel.p); + return getStruct(this, P); +} + +MachO::data_in_code_entry +MachOObjectFile::getDice(DataRefImpl Rel) const { + const char *P = reinterpret_cast(Rel.p); + return getStruct(this, P); +} + +MachO::mach_header MachOObjectFile::getHeader() const { + return getStruct(this, getPtr(this, 0)); +} + +MachO::mach_header_64 MachOObjectFile::getHeader64() const { + return getStruct(this, getPtr(this, 0)); +} + +uint32_t MachOObjectFile::getIndirectSymbolTableEntry( + const MachO::dysymtab_command &DLC, + unsigned Index) const { + uint64_t Offset = DLC.indirectsymoff + Index * sizeof(uint32_t); + return getStruct(this, getPtr(this, Offset)); +} + +MachO::data_in_code_entry +MachOObjectFile::getDataInCodeTableEntry(uint32_t DataOffset, + unsigned Index) const { + uint64_t Offset = DataOffset + Index * sizeof(MachO::data_in_code_entry); + return getStruct(this, getPtr(this, Offset)); +} + +MachO::symtab_command MachOObjectFile::getSymtabLoadCommand() const { + return getStruct(this, SymtabLoadCmd); +} + +MachO::dysymtab_command MachOObjectFile::getDysymtabLoadCommand() const { + return getStruct(this, DysymtabLoadCmd); +} + +MachO::linkedit_data_command +MachOObjectFile::getDataInCodeLoadCommand() const { + if (DataInCodeLoadCmd) + return getStruct(this, DataInCodeLoadCmd); + + // If there is no DataInCodeLoadCmd return a load command with zero'ed fields. + MachO::linkedit_data_command Cmd; + Cmd.cmd = MachO::LC_DATA_IN_CODE; + Cmd.cmdsize = sizeof(MachO::linkedit_data_command); + Cmd.dataoff = 0; + Cmd.datasize = 0; + return Cmd; +} + +StringRef MachOObjectFile::getStringTableData() const { + MachO::symtab_command S = getSymtabLoadCommand(); + return getData().substr(S.stroff, S.strsize); +} + +bool MachOObjectFile::is64Bit() const { + return getType() == getMachOType(false, true) || + getType() == getMachOType(true, true); +} + +void MachOObjectFile::ReadULEB128s(uint64_t Index, + SmallVectorImpl &Out) const { + DataExtractor extractor(ObjectFile::getData(), true, 0); + + uint32_t offset = Index; + uint64_t data = 0; + while (uint64_t delta = extractor.getULEB128(&offset)) { + data += delta; + Out.push_back(data); + } +} + +ErrorOr ObjectFile::createMachOObjectFile(MemoryBuffer *Buffer) { + StringRef Magic = Buffer->getBuffer().slice(0, 4); + error_code EC; + OwningPtr Ret; + if (Magic == "\xFE\xED\xFA\xCE") + Ret.reset(new MachOObjectFile(Buffer, false, false, EC)); + else if (Magic == "\xCE\xFA\xED\xFE") + Ret.reset(new MachOObjectFile(Buffer, true, false, EC)); + else if (Magic == "\xFE\xED\xFA\xCF") + Ret.reset(new MachOObjectFile(Buffer, false, true, EC)); + else if (Magic == "\xCF\xFA\xED\xFE") + Ret.reset(new MachOObjectFile(Buffer, true, true, EC)); + else { + delete Buffer; + return object_error::parse_failed; + } + + if (EC) + return EC; + return Ret.take(); +} + } // end namespace object } // end namespace llvm