diff options
author | Rafael Espindola <rafael.espindola@gmail.com> | 2013-04-26 20:07:33 +0000 |
---|---|---|
committer | Rafael Espindola <rafael.espindola@gmail.com> | 2013-04-26 20:07:33 +0000 |
commit | 2173e1839c2d00f7f980450dd537047b7b376e6b (patch) | |
tree | 88a1b23ce0c311290c8aae04f3b4cc7d21cec65f | |
parent | fbb2c7adc23b20141c5f78bbee07ae22028473a7 (diff) |
Use llvm/Object/MachO.h in macho-dumper. Drop the old macho parser.
For Mach-O there were 2 implementations for parsing object files. A
standalone llvm/Object/MachOObject.h and llvm/Object/MachO.h which
implements the generic interface in llvm/Object/ObjectFile.h.
This patch adds the missing features to MachO.h, moves macho-dump to
use MachO.h and removes ObjectFile.h.
In addition to making sure that check-all is clean, I checked that the
new version produces exactly the same output in all Mach-O files in a
llvm+clang build directory (including executables and shared
libraries).
To test the performance, I ran macho-dump over all the files in a
llvm+clang build directory again, but this time redirecting the output
to /dev/null. Both the old and new versions take about 4.6 seconds
(2.5 user) to finish.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@180624 91177308-0d34-0410-b5e6-96231b3b80d8
-rw-r--r-- | include/llvm/ADT/InMemoryStruct.h | 77 | ||||
-rw-r--r-- | include/llvm/Object/MachO.h | 22 | ||||
-rw-r--r-- | include/llvm/Object/MachOObject.h | 212 | ||||
-rw-r--r-- | lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.h | 2 | ||||
-rw-r--r-- | lib/Object/CMakeLists.txt | 1 | ||||
-rw-r--r-- | lib/Object/MachOObject.cpp | 422 | ||||
-rw-r--r-- | lib/Object/MachOObjectFile.cpp | 161 | ||||
-rw-r--r-- | tools/llvm-rtdyld/llvm-rtdyld.cpp | 2 | ||||
-rw-r--r-- | tools/macho-dump/macho-dump.cpp | 424 |
9 files changed, 347 insertions, 976 deletions
diff --git a/include/llvm/ADT/InMemoryStruct.h b/include/llvm/ADT/InMemoryStruct.h deleted file mode 100644 index a56084501a..0000000000 --- a/include/llvm/ADT/InMemoryStruct.h +++ /dev/null @@ -1,77 +0,0 @@ -//===- InMemoryStruct.h - Indirect Struct Access Smart Pointer --*- C++ -*-===// -// -// The LLVM Compiler Infrastructure -// -// This file is distributed under the University of Illinois Open Source -// License. See LICENSE.TXT for details. -// -//===----------------------------------------------------------------------===// - -#ifndef LLVM_ADT_INMEMORYSTRUCT_H -#define LLVM_ADT_INMEMORYSTRUCT_H - -#include <cassert> - -namespace llvm { - -/// \brief Helper object for abstracting access to an in-memory structure which -/// may require some kind of temporary storage. -/// -/// This class is designed to be used for accessing file data structures which -/// in the common case can be accessed from a direct pointer to a memory mapped -/// object, but which in some cases may require indirect access to a temporary -/// structure (which, for example, may have undergone endianness translation). -template<typename T> -class InMemoryStruct { - typedef T value_type; - typedef value_type &reference; - typedef value_type *pointer; - typedef const value_type &const_reference; - typedef const value_type *const_pointer; - - /// \brief The smart pointer target. - value_type *Target; - - /// \brief A temporary object which can be used as a target of the smart - /// pointer. - value_type Contents; - -private: - -public: - InMemoryStruct() : Target(0) {} - InMemoryStruct(reference Value) : Target(&Contents), Contents(Value) {} - InMemoryStruct(pointer Value) : Target(Value) {} - InMemoryStruct(const InMemoryStruct<T> &Value) { *this = Value; } - - void operator=(const InMemoryStruct<T> &Value) { - if (Value.Target != &Value.Contents) { - Target = Value.Target; - } else { - Target = &Contents; - Contents = Value.Contents; - } - } - - const_reference operator*() const { - assert(Target && "Cannot dereference null pointer"); - return *Target; - } - reference operator*() { - assert(Target && "Cannot dereference null pointer"); - return *Target; - } - - const_pointer operator->() const { - return Target; - } - pointer operator->() { - return Target; - } - - operator bool() const { return Target != 0; } -}; - -} - -#endif diff --git a/include/llvm/Object/MachO.h b/include/llvm/Object/MachO.h index d8598ebdcf..98d3788e44 100644 --- a/include/llvm/Object/MachO.h +++ b/include/llvm/Object/MachO.h @@ -106,6 +106,9 @@ public: virtual StringRef getLoadName() const; + relocation_iterator getSectionRelBegin(unsigned Index) const; + relocation_iterator getSectionRelEnd(unsigned Index) const; + // In a MachO file, sections have a segment name. This is used in the .o // files. They have a single segment, but this field specifies which segment // a section should be put in in the final object. @@ -134,14 +137,32 @@ public: // MachO specific structures. macho::Section getSection(DataRefImpl DRI) const; macho::Section64 getSection64(DataRefImpl DRI) const; + macho::Section getSection(const LoadCommandInfo &L, unsigned Index) const; + macho::Section64 getSection64(const LoadCommandInfo &L, unsigned Index) const; macho::SymbolTableEntry getSymbolTableEntry(DataRefImpl DRI) const; macho::Symbol64TableEntry getSymbol64TableEntry(DataRefImpl DRI) const; + macho::LinkeditDataLoadCommand getLinkeditDataLoadCommand(const LoadCommandInfo &L) const; + macho::SegmentLoadCommand + getSegmentLoadCommand(const LoadCommandInfo &L) const; + macho::Segment64LoadCommand + getSegment64LoadCommand(const LoadCommandInfo &L) const; + macho::LinkerOptionsLoadCommand + getLinkerOptionsLoadCommand(const LoadCommandInfo &L) const; + macho::RelocationEntry getRelocation(DataRefImpl Rel) const; macho::Header getHeader() const; + macho::Header64Ext getHeader64Ext() const; + macho::IndirectSymbolTableEntry + getIndirectSymbolTableEntry(const macho::DysymtabLoadCommand &DLC, + unsigned Index) const; + macho::DataInCodeTableEntry getDataInCodeTableEntry(uint32_t DataOffset, + unsigned Index) const; macho::SymtabLoadCommand getSymtabLoadCommand() const; + macho::DysymtabLoadCommand getDysymtabLoadCommand() const; + StringRef getStringTableData() const; bool is64Bit() const; void ReadULEB128s(uint64_t Index, SmallVectorImpl<uint64_t> &Out) const; @@ -153,6 +174,7 @@ private: typedef SmallVector<const char*, 1> SectionList; SectionList Sections; const char *SymtabLoadCmd; + const char *DysymtabLoadCmd; }; } diff --git a/include/llvm/Object/MachOObject.h b/include/llvm/Object/MachOObject.h deleted file mode 100644 index 50ee25b463..0000000000 --- a/include/llvm/Object/MachOObject.h +++ /dev/null @@ -1,212 +0,0 @@ -//===- MachOObject.h - Mach-O Object File Wrapper ---------------*- C++ -*-===// -// -// The LLVM Compiler Infrastructure -// -// This file is distributed under the University of Illinois Open Source -// License. See LICENSE.TXT for details. -// -//===----------------------------------------------------------------------===// - -#ifndef LLVM_OBJECT_MACHOOBJECT_H -#define LLVM_OBJECT_MACHOOBJECT_H - -#include "llvm/ADT/InMemoryStruct.h" -#include "llvm/ADT/OwningPtr.h" -#include "llvm/ADT/StringRef.h" -#include "llvm/Object/MachOFormat.h" -#include <string> - -namespace llvm { - -class MemoryBuffer; -class raw_ostream; - -namespace object { - -/// \brief Wrapper object for manipulating Mach-O object files. -/// -/// This class is designed to implement a full-featured, efficient, portable, -/// and robust Mach-O interface to Mach-O object files. It does not attempt to -/// smooth over rough edges in the Mach-O format or generalize access to object -/// independent features. -/// -/// The class is designed around accessing the Mach-O object which is expected -/// to be fully loaded into memory. -/// -/// This class is *not* suitable for concurrent use. For efficient operation, -/// the class uses APIs which rely on the ability to cache the results of -/// certain calls in internal objects which are not safe for concurrent -/// access. This allows the API to be zero-copy on the common paths. -// -// FIXME: It would be cool if we supported a "paged" MemoryBuffer -// implementation. This would allow us to implement a more sensible version of -// MemoryObject which can work like a MemoryBuffer, but be more efficient for -// objects which are in the current address space. -class MachOObject { -public: - struct LoadCommandInfo { - /// The load command information. - macho::LoadCommand Command; - - /// The offset to the start of the load command in memory. - uint64_t Offset; - }; - -private: - OwningPtr<MemoryBuffer> Buffer; - - /// Whether the object is little endian. - bool IsLittleEndian; - /// Whether the object is 64-bit. - bool Is64Bit; - /// Whether the object is swapped endianness from the host. - bool IsSwappedEndian; - /// Whether the string table has been registered. - bool HasStringTable; - - /// The cached information on the load commands. - LoadCommandInfo *LoadCommands; - mutable unsigned NumLoadedCommands; - - /// The cached copy of the header. - macho::Header Header; - macho::Header64Ext Header64Ext; - - /// Cache string table information. - StringRef StringTable; - -private: - MachOObject(MemoryBuffer *Buffer, bool IsLittleEndian, bool Is64Bit); - -public: - ~MachOObject(); - - /// \brief Load a Mach-O object from a MemoryBuffer object. - /// - /// \param Buffer - The buffer to load the object from. This routine takes - /// exclusive ownership of the buffer (which is passed to the returned object - /// on success). - /// \param ErrorStr [out] - If given, will be set to a user readable error - /// message on failure. - /// \returns The loaded object, or null on error. - static MachOObject *LoadFromBuffer(MemoryBuffer *Buffer, - std::string *ErrorStr = 0); - - /// @name File Information - /// @{ - - bool isLittleEndian() const { return IsLittleEndian; } - bool isSwappedEndian() const { return IsSwappedEndian; } - bool is64Bit() const { return Is64Bit; } - - unsigned getHeaderSize() const { - return Is64Bit ? macho::Header64Size : macho::Header32Size; - } - - StringRef getData(size_t Offset, size_t Size) const; - - /// @} - /// @name String Table Data - /// @{ - - StringRef getStringTableData() const { - assert(HasStringTable && "String table has not been registered!"); - return StringTable; - } - - StringRef getStringAtIndex(unsigned Index) const { - size_t End = getStringTableData().find('\0', Index); - return getStringTableData().slice(Index, End); - } - - void RegisterStringTable(macho::SymtabLoadCommand &SLC); - - /// @} - /// @name Object Header Access - /// @{ - - const macho::Header &getHeader() const { return Header; } - const macho::Header64Ext &getHeader64Ext() const { - assert(is64Bit() && "Invalid access!"); - return Header64Ext; - } - - /// @} - /// @name Object Structure Access - /// @{ - - // TODO: Would be useful to have an iterator based version - // of this. - /// \brief Retrieve the information for the given load command. - const LoadCommandInfo &getLoadCommandInfo(unsigned Index) const; - - void ReadSegmentLoadCommand( - const LoadCommandInfo &LCI, - InMemoryStruct<macho::SegmentLoadCommand> &Res) const; - void ReadSegment64LoadCommand( - const LoadCommandInfo &LCI, - InMemoryStruct<macho::Segment64LoadCommand> &Res) const; - void ReadSymtabLoadCommand( - const LoadCommandInfo &LCI, - InMemoryStruct<macho::SymtabLoadCommand> &Res) const; - void ReadDysymtabLoadCommand( - const LoadCommandInfo &LCI, - InMemoryStruct<macho::DysymtabLoadCommand> &Res) const; - void ReadLinkeditDataLoadCommand( - const LoadCommandInfo &LCI, - InMemoryStruct<macho::LinkeditDataLoadCommand> &Res) const; - void ReadLinkerOptionsLoadCommand( - const LoadCommandInfo &LCI, - InMemoryStruct<macho::LinkerOptionsLoadCommand> &Res) const; - void ReadIndirectSymbolTableEntry( - const macho::DysymtabLoadCommand &DLC, - unsigned Index, - InMemoryStruct<macho::IndirectSymbolTableEntry> &Res) const; - void ReadSection( - const LoadCommandInfo &LCI, - unsigned Index, - InMemoryStruct<macho::Section> &Res) const; - void ReadSection64( - const LoadCommandInfo &LCI, - unsigned Index, - InMemoryStruct<macho::Section64> &Res) const; - void ReadRelocationEntry( - uint64_t RelocationTableOffset, unsigned Index, - InMemoryStruct<macho::RelocationEntry> &Res) const; - void ReadSymbolTableEntry( - uint64_t SymbolTableOffset, unsigned Index, - InMemoryStruct<macho::SymbolTableEntry> &Res) const; - void ReadSymbol64TableEntry( - uint64_t SymbolTableOffset, unsigned Index, - InMemoryStruct<macho::Symbol64TableEntry> &Res) const; - void ReadDataInCodeTableEntry( - uint64_t TableOffset, unsigned Index, - InMemoryStruct<macho::DataInCodeTableEntry> &Res) const; - void ReadULEB128s(uint64_t Index, SmallVectorImpl<uint64_t> &Out) const; - - /// @} - - /// @name Object Dump Facilities - /// @{ - /// dump - Support for debugging, callable in GDB: V->dump() - // - void dump() const; - void dumpHeader() const; - - /// print - Implement operator<< on Value. - /// - void print(raw_ostream &O) const; - void printHeader(raw_ostream &O) const; - - /// @} -}; - -inline raw_ostream &operator<<(raw_ostream &OS, const MachOObject &V) { - V.print(OS); - return OS; -} - -} // end namespace object -} // end namespace llvm - -#endif diff --git a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.h b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.h index 62d8487078..f581b3f824 100644 --- a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.h +++ b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.h @@ -16,7 +16,7 @@ #include "RuntimeDyldImpl.h" #include "llvm/ADT/IndexedMap.h" -#include "llvm/Object/MachOObject.h" +#include "llvm/Object/MachO.h" #include "llvm/Support/Format.h" using namespace llvm; diff --git a/lib/Object/CMakeLists.txt b/lib/Object/CMakeLists.txt index c20fc0cc39..4ed129f467 100644 --- a/lib/Object/CMakeLists.txt +++ b/lib/Object/CMakeLists.txt @@ -4,7 +4,6 @@ add_llvm_library(LLVMObject COFFObjectFile.cpp ELFObjectFile.cpp Error.cpp - MachOObject.cpp MachOObjectFile.cpp Object.cpp ObjectFile.cpp diff --git a/lib/Object/MachOObject.cpp b/lib/Object/MachOObject.cpp deleted file mode 100644 index d88f96f298..0000000000 --- a/lib/Object/MachOObject.cpp +++ /dev/null @@ -1,422 +0,0 @@ -//===- MachOObject.cpp - Mach-O Object File Wrapper -----------------------===// -// -// The LLVM Compiler Infrastructure -// -// This file is distributed under the University of Illinois Open Source -// License. See LICENSE.TXT for details. -// -//===----------------------------------------------------------------------===// - -#include "llvm/Object/MachOObject.h" -#include "llvm/ADT/SmallVector.h" -#include "llvm/ADT/StringRef.h" -#include "llvm/Support/DataExtractor.h" -#include "llvm/Support/Debug.h" -#include "llvm/Support/Host.h" -#include "llvm/Support/MemoryBuffer.h" -#include "llvm/Support/SwapByteOrder.h" -#include "llvm/Support/raw_ostream.h" - -using namespace llvm; -using namespace llvm::object; - -/* Translation Utilities */ - -template<typename T> -static void SwapValue(T &Value) { - Value = sys::SwapByteOrder(Value); -} - -template<typename T> -static void SwapStruct(T &Value); - -template<typename T> -static void ReadInMemoryStruct(const MachOObject &MOO, - StringRef Buffer, uint64_t Base, - InMemoryStruct<T> &Res) { - typedef T struct_type; - uint64_t Size = sizeof(struct_type); - - // Check that the buffer contains the expected data. - if (Base + Size > Buffer.size()) { - Res = 0; - return; - } - - // Check whether we can return a direct pointer. - struct_type *Ptr = reinterpret_cast<struct_type *>( - const_cast<char *>(Buffer.data() + Base)); - if (!MOO.isSwappedEndian()) { - Res = Ptr; - return; - } - - // Otherwise, copy the struct and translate the values. - Res = *Ptr; - SwapStruct(*Res); -} - -/* *** */ - -MachOObject::MachOObject(MemoryBuffer *Buffer_, bool IsLittleEndian_, - bool Is64Bit_) - : Buffer(Buffer_), IsLittleEndian(IsLittleEndian_), Is64Bit(Is64Bit_), - IsSwappedEndian(IsLittleEndian != sys::IsLittleEndianHost), - HasStringTable(false), LoadCommands(0), NumLoadedCommands(0) { - // Load the common header. - memcpy(&Header, Buffer->getBuffer().data(), sizeof(Header)); - if (IsSwappedEndian) { - SwapValue(Header.Magic); - SwapValue(Header.CPUType); - SwapValue(Header.CPUSubtype); - SwapValue(Header.FileType); - SwapValue(Header.NumLoadCommands); - SwapValue(Header.SizeOfLoadCommands); - SwapValue(Header.Flags); - } - - if (is64Bit()) { - memcpy(&Header64Ext, Buffer->getBuffer().data() + sizeof(Header), - sizeof(Header64Ext)); - if (IsSwappedEndian) { - SwapValue(Header64Ext.Reserved); - } - } - - // Create the load command array if sane. - if (getHeader().NumLoadCommands < (1 << 20)) - LoadCommands = new LoadCommandInfo[getHeader().NumLoadCommands]; -} - -MachOObject::~MachOObject() { - delete [] LoadCommands; -} - -MachOObject *MachOObject::LoadFromBuffer(MemoryBuffer *Buffer, - std::string *ErrorStr) { - // First, check the magic value and initialize the basic object info. - bool IsLittleEndian = false, Is64Bit = false; - StringRef Magic = Buffer->getBuffer().slice(0, 4); - if (Magic == "\xFE\xED\xFA\xCE") { - } else if (Magic == "\xCE\xFA\xED\xFE") { - IsLittleEndian = true; - } else if (Magic == "\xFE\xED\xFA\xCF") { - Is64Bit = true; - } else if (Magic == "\xCF\xFA\xED\xFE") { - IsLittleEndian = true; - Is64Bit = true; - } else { - if (ErrorStr) *ErrorStr = "not a Mach object file (invalid magic)"; - return 0; - } - - // Ensure that the at least the full header is present. - unsigned HeaderSize = Is64Bit ? macho::Header64Size : macho::Header32Size; - if (Buffer->getBufferSize() < HeaderSize) { - if (ErrorStr) *ErrorStr = "not a Mach object file (invalid header)"; - return 0; - } - - OwningPtr<MachOObject> Object(new MachOObject(Buffer, IsLittleEndian, - Is64Bit)); - - // Check for bogus number of load commands. - if (Object->getHeader().NumLoadCommands >= (1 << 20)) { - if (ErrorStr) *ErrorStr = "not a Mach object file (unreasonable header)"; - return 0; - } - - if (ErrorStr) *ErrorStr = ""; - return Object.take(); -} - -StringRef MachOObject::getData(size_t Offset, size_t Size) const { - return Buffer->getBuffer().substr(Offset,Size); -} - -void MachOObject::RegisterStringTable(macho::SymtabLoadCommand &SLC) { - HasStringTable = true; - StringTable = Buffer->getBuffer().substr(SLC.StringTableOffset, - SLC.StringTableSize); -} - -const MachOObject::LoadCommandInfo & -MachOObject::getLoadCommandInfo(unsigned Index) const { - assert(Index < getHeader().NumLoadCommands && "Invalid index!"); - - // Load the command, if necessary. - if (Index >= NumLoadedCommands) { - uint64_t Offset; - if (Index == 0) { - Offset = getHeaderSize(); - } else { - const LoadCommandInfo &Prev = getLoadCommandInfo(Index - 1); - Offset = Prev.Offset + Prev.Command.Size; - } - - LoadCommandInfo &Info = LoadCommands[Index]; - memcpy(&Info.Command, Buffer->getBuffer().data() + Offset, - sizeof(macho::LoadCommand)); - if (IsSwappedEndian) { - SwapValue(Info.Command.Type); - SwapValue(Info.Command.Size); - } - Info.Offset = Offset; - NumLoadedCommands = Index + 1; - } - - return LoadCommands[Index]; -} - -template<> -void SwapStruct(macho::SegmentLoadCommand &Value) { - SwapValue(Value.Type); - SwapValue(Value.Size); - SwapValue(Value.VMAddress); - SwapValue(Value.VMSize); - SwapValue(Value.FileOffset); - SwapValue(Value.FileSize); - SwapValue(Value.MaxVMProtection); - SwapValue(Value.InitialVMProtection); - SwapValue(Value.NumSections); - SwapValue(Value.Flags); -} -void MachOObject::ReadSegmentLoadCommand(const LoadCommandInfo &LCI, - InMemoryStruct<macho::SegmentLoadCommand> &Res) const { - ReadInMemoryStruct(*this, Buffer->getBuffer(), LCI.Offset, Res); -} - -template<> -void SwapStruct(macho::Segment64LoadCommand &Value) { - SwapValue(Value.Type); - SwapValue(Value.Size); - SwapValue(Value.VMAddress); - SwapValue(Value.VMSize); - SwapValue(Value.FileOffset); - SwapValue(Value.FileSize); - SwapValue(Value.MaxVMProtection); - SwapValue(Value.InitialVMProtection); - SwapValue(Value.NumSections); - SwapValue(Value.Flags); -} -void MachOObject::ReadSegment64LoadCommand(const LoadCommandInfo &LCI, - InMemoryStruct<macho::Segment64LoadCommand> &Res) const { - ReadInMemoryStruct(*this, Buffer->getBuffer(), LCI.Offset, Res); -} - -template<> -void SwapStruct(macho::SymtabLoadCommand &Value) { - SwapValue(Value.Type); - SwapValue(Value.Size); - SwapValue(Value.SymbolTableOffset); - SwapValue(Value.NumSymbolTableEntries); - SwapValue(Value.StringTableOffset); - SwapValue(Value.StringTableSize); -} -void MachOObject::ReadSymtabLoadCommand(const LoadCommandInfo &LCI, - InMemoryStruct<macho::SymtabLoadCommand> &Res) const { - ReadInMemoryStruct(*this, Buffer->getBuffer(), LCI.Offset, Res); -} - -template<> -void SwapStruct(macho::DysymtabLoadCommand &Value) { - SwapValue(Value.Type); - SwapValue(Value.Size); - SwapValue(Value.LocalSymbolsIndex); - SwapValue(Value.NumLocalSymbols); - SwapValue(Value.ExternalSymbolsIndex); - SwapValue(Value.NumExternalSymbols); - SwapValue(Value.UndefinedSymbolsIndex); - SwapValue(Value.NumUndefinedSymbols); - SwapValue(Value.TOCOffset); - SwapValue(Value.NumTOCEntries); - SwapValue(Value.ModuleTableOffset); - SwapValue(Value.NumModuleTableEntries); - SwapValue(Value.ReferenceSymbolTableOffset); - SwapValue(Value.NumReferencedSymbolTableEntries); - SwapValue(Value.IndirectSymbolTableOffset); - SwapValue(Value.NumIndirectSymbolTableEntries); - SwapValue(Value.ExternalRelocationTableOffset); - SwapValue(Value.NumExternalRelocationTableEntries); - SwapValue(Value.LocalRelocationTableOffset); - SwapValue(Value.NumLocalRelocationTableEntries); -} -void MachOObject::ReadDysymtabLoadCommand(const LoadCommandInfo &LCI, - InMemoryStruct<macho::DysymtabLoadCommand> &Res) const { - ReadInMemoryStruct(*this, Buffer->getBuffer(), LCI.Offset, Res); -} - -template<> -void SwapStruct(macho::LinkeditDataLoadCommand &Value) { - SwapValue(Value.Type); - SwapValue(Value.Size); - SwapValue(Value.DataOffset); - SwapValue(Value.DataSize); -} -void MachOObject::ReadLinkeditDataLoadCommand(const LoadCommandInfo &LCI, - InMemoryStruct<macho::LinkeditDataLoadCommand> &Res) const { - ReadInMemoryStruct(*this, Buffer->getBuffer(), LCI.Offset, Res); -} - -template<> -void SwapStruct(macho::LinkerOptionsLoadCommand &Value) { - SwapValue(Value.Type); - SwapValue(Value.Size); - SwapValue(Value.Count); -} -void MachOObject::ReadLinkerOptionsLoadCommand(const LoadCommandInfo &LCI, - InMemoryStruct<macho::LinkerOptionsLoadCommand> &Res) const { - ReadInMemoryStruct(*this, Buffer->getBuffer(), LCI.Offset, Res); -} - -template<> -void SwapStruct(macho::IndirectSymbolTableEntry &Value) { - SwapValue(Value.Index); -} -void -MachOObject::ReadIndirectSymbolTableEntry(const macho::DysymtabLoadCommand &DLC, - unsigned Index, - InMemoryStruct<macho::IndirectSymbolTableEntry> &Res) const { - uint64_t Offset = (DLC.IndirectSymbolTableOffset + - Index * sizeof(macho::IndirectSymbolTableEntry)); - ReadInMemoryStruct(*this, Buffer->getBuffer(), Offset, Res); -} - - -template<> -void SwapStruct(macho::Section &Value) { - SwapValue(Value.Address); - SwapValue(Value.Size); - SwapValue(Value.Offset); - SwapValue(Value.Align); - SwapValue(Value.RelocationTableOffset); - SwapValue(Value.NumRelocationTableEntries); - SwapValue(Value.Flags); - SwapValue(Value.Reserved1); - SwapValue(Value.Reserved2); -} -void MachOObject::ReadSection(const LoadCommandInfo &LCI, - unsigned Index, - InMemoryStruct<macho::Section> &Res) const { - assert(LCI.Command.Type == macho::LCT_Segment && - "Unexpected load command info!"); - uint64_t Offset = (LCI.Offset + sizeof(macho::SegmentLoadCommand) + - Index * sizeof(macho::Section)); - ReadInMemoryStruct(*this, Buffer->getBuffer(), Offset, Res); -} - -template<> -void SwapStruct(macho::Section64 &Value) { - SwapValue(Value.Address); - SwapValue(Value.Size); - SwapValue(Value.Offset); - SwapValue(Value.Align); - SwapValue(Value.RelocationTableOffset); - SwapValue(Value.NumRelocationTableEntries); - SwapValue(Value.Flags); - SwapValue(Value.Reserved1); - SwapValue(Value.Reserved2); - SwapValue(Value.Reserved3); -} -void MachOObject::ReadSection64(const LoadCommandInfo &LCI, - unsigned Index, - InMemoryStruct<macho::Section64> &Res) const { - assert(LCI.Command.Type == macho::LCT_Segment64 && - "Unexpected load command info!"); - uint64_t Offset = (LCI.Offset + sizeof(macho::Segment64LoadCommand) + - Index * sizeof(macho::Section64)); - ReadInMemoryStruct(*this, Buffer->getBuffer(), Offset, Res); -} - -template<> -void SwapStruct(macho::RelocationEntry &Value) { - SwapValue(Value.Word0); - SwapValue(Value.Word1); -} -void MachOObject::ReadRelocationEntry(uint64_t RelocationTableOffset, - unsigned Index, - InMemoryStruct<macho::RelocationEntry> &Res) const { - uint64_t Offset = (RelocationTableOffset + - Index * sizeof(macho::RelocationEntry)); - ReadInMemoryStruct(*this, Buffer->getBuffer(), Offset, Res); -} - -template<> -void SwapStruct(macho::SymbolTableEntry &Value) { - SwapValue(Value.StringIndex); - SwapValue(Value.Flags); - SwapValue(Value.Value); -} -void MachOObject::ReadSymbolTableEntry(uint64_t SymbolTableOffset, - unsigned Index, - InMemoryStruct<macho::SymbolTableEntry> &Res) const { - uint64_t Offset = (SymbolTableOffset + - Index * sizeof(macho::SymbolTableEntry)); - ReadInMemoryStruct(*this, Buffer->getBuffer(), Offset, Res); -} - -template<> -void SwapStruct(macho::Symbol64TableEntry &Value) { - SwapValue(Value.StringIndex); - SwapValue(Value.Flags); - SwapValue(Value.Value); -} -void MachOObject::ReadSymbol64TableEntry(uint64_t SymbolTableOffset, - unsigned Index, - InMemoryStruct<macho::Symbol64TableEntry> &Res) const { - uint64_t Offset = (SymbolTableOffset + - Index * sizeof(macho::Symbol64TableEntry)); - ReadInMemoryStruct(*this, Buffer->getBuffer(), Offset, Res); -} - -template<> -void SwapStruct(macho::DataInCodeTableEntry &Value) { - SwapValue(Value.Offset); - SwapValue(Value.Length); - SwapValue(Value.Kind); -} -void MachOObject::ReadDataInCodeTableEntry(uint64_t TableOffset, - unsigned Index, - InMemoryStruct<macho::DataInCodeTableEntry> &Res) const { - uint64_t Offset = (TableOffset + - Index * sizeof(macho::DataInCodeTableEntry)); - ReadInMemoryStruct(*this, Buffer->getBuffer(), Offset, Res); -} - -void MachOObject::ReadULEB128s(uint64_t Index, - SmallVectorImpl<uint64_t> &Out) const { - DataExtractor extractor(Buffer->getBuffer(), true, 0); - - uint32_t offset = Index; - uint64_t data = 0; - while (uint64_t delta = extractor.getULEB128(&offset)) { - data += delta; - Out.push_back(data); - } -} - -/* ** */ -// Object Dumping Facilities -void MachOObject::dump() const { print(dbgs()); dbgs() << '\n'; } -void MachOObject::dumpHeader() const { printHeader(dbgs()); dbgs() << '\n'; } - -void MachOObject::printHeader(raw_ostream &O) const { - O << "('cputype', " << Header.CPUType << ")\n"; - O << "('cpusubtype', " << Header.CPUSubtype << ")\n"; - O << "('filetype', " << Header.FileType << ")\n"; - O << "('num_load_commands', " << Header.NumLoadCommands << ")\n"; - O << "('load_commands_size', " << Header.SizeOfLoadCommands << ")\n"; - O << "('flag', " << Header.Flags << ")\n"; - - // Print extended header if 64-bit. - if (is64Bit()) - O << "('reserved', " << Header64Ext.Reserved << ")\n"; -} - -void MachOObject::print(raw_ostream &O) const { - O << "Header:\n"; - printHeader(O); - O << "Load Commands:\n"; - - O << "Buffer:\n"; -} diff --git a/lib/Object/MachOObjectFile.cpp b/lib/Object/MachOObjectFile.cpp index 38198b1793..bb6ca93176 100644 --- a/lib/Object/MachOObjectFile.cpp +++ b/lib/Object/MachOObjectFile.cpp @@ -120,6 +120,11 @@ void SwapStruct(macho::Header &H) { } template<> +void SwapStruct(macho::Header64Ext &E) { + SwapValue(E.Reserved); +} + +template<> void SwapStruct(macho::SymtabLoadCommand &C) { SwapValue(C.Type); SwapValue(C.Size); @@ -130,6 +135,30 @@ void SwapStruct(macho::SymtabLoadCommand &C) { } template<> +void SwapStruct(macho::DysymtabLoadCommand &C) { + SwapValue(C.Type); + SwapValue(C.Size); + SwapValue(C.LocalSymbolsIndex); + SwapValue(C.NumLocalSymbols); + SwapValue(C.ExternalSymbolsIndex); + SwapValue(C.NumExternalSymbols); + SwapValue(C.UndefinedSymbolsIndex); + SwapValue(C.NumUndefinedSymbols); + SwapValue(C.TOCOffset); + SwapValue(C.NumTOCEntries); + SwapValue(C.ModuleTableOffset); + SwapValue(C.NumModuleTableEntries); + SwapValue(C.ReferenceSymbolTableOffset); + SwapValue(C.NumReferencedSymbolTableEntries); + SwapValue(C.IndirectSymbolTableOffset); + SwapValue(C.NumIndirectSymbolTableEntries); + SwapValue(C.ExternalRelocationTableOffset); + SwapValue(C.NumExternalRelocationTableEntries); + SwapValue(C.LocalRelocationTableOffset); + SwapValue(C.NumLocalRelocationTableEntries); +} + +template<> void SwapStruct(macho::LinkeditDataLoadCommand &C) { SwapValue(C.Type); SwapValue(C.Size); @@ -165,6 +194,25 @@ void SwapStruct(macho::Segment64LoadCommand &C) { SwapValue(C.Flags); } +template<> +void SwapStruct(macho::IndirectSymbolTableEntry &C) { + SwapValue(C.Index); +} + +template<> +void SwapStruct(macho::LinkerOptionsLoadCommand &C) { + SwapValue(C.Type); + SwapValue(C.Size); + SwapValue(C.Count); +} + +template<> +void SwapStruct(macho::DataInCodeTableEntry &C) { + SwapValue(C.Offset); + SwapValue(C.Length); + SwapValue(C.Kind); +} + template<typename T> T getStruct(const MachOObjectFile *O, const char *P) { T Cmd; @@ -174,32 +222,20 @@ T getStruct(const MachOObjectFile *O, const char *P) { return Cmd; } -static macho::SegmentLoadCommand -getSegmentLoadCommand(const MachOObjectFile *O, - const MachOObjectFile::LoadCommandInfo &L) { - return getStruct<macho::SegmentLoadCommand>(O, L.Ptr); -} - -static macho::Segment64LoadCommand -getSegment64LoadCommand(const MachOObjectFile *O, - const MachOObjectFile::LoadCommandInfo &L) { - return getStruct<macho::Segment64LoadCommand>(O, L.Ptr); -} - static uint32_t getSegmentLoadCommandNumSections(const MachOObjectFile *O, const MachOObjectFile::LoadCommandInfo &L) { if (O->is64Bit()) { - macho::Segment64LoadCommand S = getSegment64LoadCommand(O, L); + macho::Segment64LoadCommand S = O->getSegment64LoadCommand(L); return S.NumSections; } - macho::SegmentLoadCommand S = getSegmentLoadCommand(O, L); + macho::SegmentLoadCommand S = O->getSegmentLoadCommand(L); return S.NumSections; } -static const SectionBase * -getSectionBase(const MachOObjectFile *O, MachOObjectFile::LoadCommandInfo L, - unsigned Sec) { +static const char * +getSectionPtr(const MachOObjectFile *O, MachOObjectFile::LoadCommandInfo L, + unsigned Sec) { uintptr_t CommandAddr = reinterpret_cast<uintptr_t>(L.Ptr); bool Is64 = O->is64Bit(); @@ -209,7 +245,7 @@ getSectionBase(const MachOObjectFile *O, MachOObjectFile::LoadCommandInfo L, sizeof(macho::Section); uintptr_t SectionAddr = CommandAddr + SegmentLoadSize + Sec * SectionSize; - return reinterpret_cast<const SectionBase*>(SectionAddr); + return reinterpret_cast<const char*>(SectionAddr); } static const char *getPtr(const MachOObjectFile *O, size_t Offset) { @@ -377,7 +413,7 @@ MachOObjectFile::MachOObjectFile(MemoryBuffer *Object, bool IsLittleEndian, bool Is64bits, error_code &ec) : ObjectFile(getMachOType(IsLittleEndian, Is64bits), Object), - SymtabLoadCmd(NULL) { + SymtabLoadCmd(NULL), DysymtabLoadCmd(NULL) { uint32_t LoadCommandCount = this->getHeader().NumLoadCommands; macho::LoadCommandType SegmentLoadType = is64Bit() ? macho::LCT_Segment64 : macho::LCT_Segment; @@ -387,13 +423,14 @@ MachOObjectFile::MachOObjectFile(MemoryBuffer *Object, if (Load.C.Type == macho::LCT_Symtab) { assert(!SymtabLoadCmd && "Multiple symbol tables"); SymtabLoadCmd = Load.Ptr; - } - - if (Load.C.Type == SegmentLoadType) { + } else if (Load.C.Type == macho::LCT_Dysymtab) { + assert(!DysymtabLoadCmd && "Multiple dynamic symbol tables"); + DysymtabLoadCmd = Load.Ptr; + } else if (Load.C.Type == SegmentLoadType) { uint32_t NumSections = getSegmentLoadCommandNumSections(this, Load); for (unsigned J = 0; J < NumSections; ++J) { - |