diff options
author | Dimitry Andric <dim@FreeBSD.org> | 2018-07-28 11:09:23 +0000 |
---|---|---|
committer | Dimitry Andric <dim@FreeBSD.org> | 2018-07-28 11:09:23 +0000 |
commit | f73363f1dd94996356cefbf24388f561891acf0b (patch) | |
tree | e3c31248bdb36eaec5fd833490d4278162dba2a0 /source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp | |
parent | 160ee69dd7ae18978f4068116777639ea98dc951 (diff) |
Notes
Diffstat (limited to 'source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp')
-rw-r--r-- | source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp | 560 |
1 files changed, 251 insertions, 309 deletions
diff --git a/source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp b/source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp index 36027dde0432..8701908378f1 100644 --- a/source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp +++ b/source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp @@ -113,7 +113,7 @@ const elf_word LLDB_NT_GNU_ABI_OS_SOLARIS = 0x02; //===----------------------------------------------------------------------===// /// @class ELFRelocation -/// @brief Generic wrapper for ELFRel and ELFRela. +/// Generic wrapper for ELFRel and ELFRela. /// /// This helper class allows us to parse both ELFRel and ELFRela relocation /// entries in a generic manner. @@ -240,11 +240,10 @@ bool ELFNote::Parse(const DataExtractor &data, lldb::offset_t *offset) { if (data.GetU32(offset, &n_namesz, 3) == NULL) return false; - // The name field is required to be nul-terminated, and n_namesz - // includes the terminating nul in observed implementations (contrary - // to the ELF-64 spec). A special case is needed for cores generated - // by some older Linux versions, which write a note named "CORE" - // without a nul terminator and n_namesz = 4. + // The name field is required to be nul-terminated, and n_namesz includes the + // terminating nul in observed implementations (contrary to the ELF-64 spec). + // A special case is needed for cores generated by some older Linux versions, + // which write a note named "CORE" without a nul terminator and n_namesz = 4. if (n_namesz == 4) { char buf[4]; if (data.ExtractBytes(*offset, 4, data.GetByteOrder(), buf) != 4) @@ -295,7 +294,8 @@ static uint32_t mipsVariantFromElfFlags (const elf::ELFHeader &header) { uint32_t arch_variant = ArchSpec::eMIPSSubType_unknown; uint32_t fileclass = header.e_ident[EI_CLASS]; - // If there aren't any elf flags available (e.g core elf file) then return default + // If there aren't any elf flags available (e.g core elf file) then return + // default // 32 or 64 bit arch (without any architecture revision) based on object file's class. if (header.e_type == ET_CORE) { switch (fileclass) { @@ -549,8 +549,8 @@ uint32_t ObjectFileELF::CalculateELFNotesSegmentsCRC32( DataExtractor segment_data; if (segment_data.SetData(object_data, ph_offset, ph_size) != ph_size) { - // The ELF program header contained incorrect data, - // probably corefile is incomplete or corrupted. + // The ELF program header contained incorrect data, probably corefile + // is incomplete or corrupted. break; } @@ -595,8 +595,8 @@ static const char *OSABIAsCString(unsigned char osabi_byte) { // // WARNING : This function is being deprecated -// It's functionality has moved to ArchSpec::SetArchitecture -// This function is only being kept to validate the move. +// It's functionality has moved to ArchSpec::SetArchitecture This function is +// only being kept to validate the move. // // TODO : Remove this function static bool GetOsFromOSABI(unsigned char osabi_byte, @@ -674,29 +674,16 @@ size_t ObjectFileELF::GetModuleSpecifications( __FUNCTION__, file.GetPath().c_str()); } - // In case there is header extension in the section #0, the header - // we parsed above could have sentinel values for e_phnum, e_shnum, - // and e_shstrndx. In this case we need to reparse the header - // with a bigger data source to get the actual values. - size_t section_header_end = header.e_shoff + header.e_shentsize; - if (header.HasHeaderExtension() && - section_header_end > data_sp->GetByteSize()) { - data_sp = MapFileData(file, section_header_end, file_offset); - if (data_sp) { - data.SetData(data_sp); - lldb::offset_t header_offset = data_offset; - header.Parse(data, &header_offset); - } - } - - // Try to get the UUID from the section list. Usually that's at the - // end, so map the file in if we don't have it already. - section_header_end = - header.e_shoff + header.e_shnum * header.e_shentsize; - if (section_header_end > data_sp->GetByteSize()) { - data_sp = MapFileData(file, section_header_end, file_offset); - if (data_sp) - data.SetData(data_sp); + data_sp = MapFileData(file, -1, file_offset); + if (data_sp) + data.SetData(data_sp); + // In case there is header extension in the section #0, the header we + // parsed above could have sentinel values for e_phnum, e_shnum, and + // e_shstrndx. In this case we need to reparse the header with a + // bigger data source to get the actual values. + if (header.HasHeaderExtension()) { + lldb::offset_t header_offset = data_offset; + header.Parse(data, &header_offset); } uint32_t gnu_debuglink_crc = 0; @@ -733,51 +720,27 @@ size_t ObjectFileELF::GetModuleSpecifications( // contents crc32 would be too much of luxury. Thus we will need // to fallback to something simpler. if (header.e_type == llvm::ELF::ET_CORE) { - size_t program_headers_end = - header.e_phoff + header.e_phnum * header.e_phentsize; - if (program_headers_end > data_sp->GetByteSize()) { - data_sp = MapFileData(file, program_headers_end, file_offset); - if (data_sp) - data.SetData(data_sp); - } ProgramHeaderColl program_headers; GetProgramHeaderInfo(program_headers, data, header); - size_t segment_data_end = 0; - for (ProgramHeaderCollConstIter I = program_headers.begin(); - I != program_headers.end(); ++I) { - segment_data_end = std::max<unsigned long long>( - I->p_offset + I->p_filesz, segment_data_end); - } - - if (segment_data_end > data_sp->GetByteSize()) { - data_sp = MapFileData(file, segment_data_end, file_offset); - if (data_sp) - data.SetData(data_sp); - } - core_notes_crc = CalculateELFNotesSegmentsCRC32(program_headers, data); } else { - // Need to map entire file into memory to calculate the crc. - data_sp = MapFileData(file, -1, file_offset); - if (data_sp) { - data.SetData(data_sp); - gnu_debuglink_crc = calc_gnu_debuglink_crc32( - data.GetDataStart(), data.GetByteSize()); - } + gnu_debuglink_crc = calc_gnu_debuglink_crc32( + data.GetDataStart(), data.GetByteSize()); } } + using u32le = llvm::support::ulittle32_t; if (gnu_debuglink_crc) { // Use 4 bytes of crc from the .gnu_debuglink section. - uint32_t uuidt[4] = {gnu_debuglink_crc, 0, 0, 0}; - uuid.SetBytes(uuidt, sizeof(uuidt)); + u32le data(gnu_debuglink_crc); + uuid = UUID::fromData(&data, sizeof(data)); } else if (core_notes_crc) { // Use 8 bytes - first 4 bytes for *magic* prefix, mainly to make - // it look different form - // .gnu_debuglink crc followed by 4 bytes of note segments crc. - uint32_t uuidt[4] = {g_core_uuid_magic, core_notes_crc, 0, 0}; - uuid.SetBytes(uuidt, sizeof(uuidt)); + // it look different form .gnu_debuglink crc followed by 4 bytes + // of note segments crc. + u32le data[] = {u32le(g_core_uuid_magic), u32le(core_notes_crc)}; + uuid = UUID::fromData(data, sizeof(data)); } } @@ -861,21 +824,19 @@ bool ObjectFileELF::SetLoadAddress(Target &target, lldb::addr_t value, size_t sect_idx = 0; for (sect_idx = 0; sect_idx < num_sections; ++sect_idx) { - // Iterate through the object file sections to find all - // of the sections that have SHF_ALLOC in their flag bits. + // Iterate through the object file sections to find all of the sections + // that have SHF_ALLOC in their flag bits. SectionSP section_sp(section_list->GetSectionAtIndex(sect_idx)); if (section_sp && section_sp->Test(SHF_ALLOC)) { lldb::addr_t load_addr = section_sp->GetFileAddress(); // We don't want to update the load address of a section with type // eSectionTypeAbsoluteAddress as they already have the absolute load - // address - // already specified + // address already specified if (section_sp->GetType() != eSectionTypeAbsoluteAddress) load_addr += value; // On 32-bit systems the load address have to fit into 4 bytes. The - // rest of - // the bytes are the overflow from the addition. + // rest of the bytes are the overflow from the addition. if (GetAddressByteSize() == 4) load_addr &= 0xFFFFFFFF; @@ -905,24 +866,23 @@ uint32_t ObjectFileELF::GetAddressByteSize() const { AddressClass ObjectFileELF::GetAddressClass(addr_t file_addr) { Symtab *symtab = GetSymtab(); if (!symtab) - return eAddressClassUnknown; + return AddressClass::eUnknown; - // The address class is determined based on the symtab. Ask it from the object - // file what - // contains the symtab information. + // The address class is determined based on the symtab. Ask it from the + // object file what contains the symtab information. ObjectFile *symtab_objfile = symtab->GetObjectFile(); if (symtab_objfile != nullptr && symtab_objfile != this) return symtab_objfile->GetAddressClass(file_addr); auto res = ObjectFile::GetAddressClass(file_addr); - if (res != eAddressClassCode) + if (res != AddressClass::eCode) return res; auto ub = m_address_class_map.upper_bound(file_addr); if (ub == m_address_class_map.begin()) { - // No entry in the address class map before the address. Return - // default address class for an address in a code section. - return eAddressClassCode; + // No entry in the address class map before the address. Return default + // address class for an address in a code section. + return AddressClass::eCode; } // Move iterator to the address class entry preceding address @@ -950,6 +910,7 @@ bool ObjectFileELF::GetUUID(lldb_private::UUID *uuid) { if (!ParseSectionHeaders() && GetType() != ObjectFile::eTypeCoreFile) return false; + using u32le = llvm::support::ulittle32_t; if (m_uuid.IsValid()) { // We have the full build id uuid. *uuid = m_uuid; @@ -963,11 +924,11 @@ bool ObjectFileELF::GetUUID(lldb_private::UUID *uuid) { core_notes_crc = CalculateELFNotesSegmentsCRC32(m_program_headers, m_data); if (core_notes_crc) { - // Use 8 bytes - first 4 bytes for *magic* prefix, mainly to make it - // look different form .gnu_debuglink crc - followed by 4 bytes of note + // Use 8 bytes - first 4 bytes for *magic* prefix, mainly to make it look + // different form .gnu_debuglink crc - followed by 4 bytes of note // segments crc. - uint32_t uuidt[4] = {g_core_uuid_magic, core_notes_crc, 0, 0}; - m_uuid.SetBytes(uuidt, sizeof(uuidt)); + u32le data[] = {u32le(g_core_uuid_magic), u32le(core_notes_crc)}; + m_uuid = UUID::fromData(data, sizeof(data)); } } else { if (!m_gnu_debuglink_crc) @@ -975,8 +936,8 @@ bool ObjectFileELF::GetUUID(lldb_private::UUID *uuid) { calc_gnu_debuglink_crc32(m_data.GetDataStart(), m_data.GetByteSize()); if (m_gnu_debuglink_crc) { // Use 4 bytes of crc from the .gnu_debuglink section. - uint32_t uuidt[4] = {m_gnu_debuglink_crc, 0, 0, 0}; - m_uuid.SetBytes(uuidt, sizeof(uuidt)); + u32le data(m_gnu_debuglink_crc); + m_uuid = UUID::fromData(&data, sizeof(data)); } } @@ -1034,8 +995,8 @@ Address ObjectFileELF::GetImageInfoAddress(Target *target) { ELFDynamic &symbol = m_dynamic_symbols[i]; if (symbol.d_tag == DT_DEBUG) { - // Compute the offset as the number of previous entries plus the - // size of d_tag. + // Compute the offset as the number of previous entries plus the size of + // d_tag. addr_t offset = i * dynsym_hdr->sh_entsize + GetAddressByteSize(); return Address(dynsym_section_sp, offset); } @@ -1314,18 +1275,16 @@ ObjectFileELF::RefineModuleDetailsFromNote(lldb_private::DataExtractor &data, // Only bother processing this if we don't already have the uuid set. if (!uuid.IsValid()) { // 16 bytes is UUID|MD5, 20 bytes is SHA1. Other linkers may produce a - // build-id of a different - // length. Accept it as long as it's at least 4 bytes as it will be - // better than our own crc32. - if (note.n_descsz >= 4 && note.n_descsz <= 20) { - uint8_t uuidbuf[20]; - if (data.GetU8(&offset, &uuidbuf, note.n_descsz) == nullptr) { + // build-id of a different length. Accept it as long as it's at least + // 4 bytes as it will be better than our own crc32. + if (note.n_descsz >= 4) { + if (const uint8_t *buf = data.PeekData(offset, note.n_descsz)) { + // Save the build id as the UUID for the module. + uuid = UUID::fromData(buf, note.n_descsz); + } else { error.SetErrorString("failed to read GNU_BUILD_ID note payload"); return error; } - - // Save the build id as the UUID for the module. - uuid.SetBytes(uuidbuf, note.n_descsz); } } break; @@ -1368,8 +1327,8 @@ ObjectFileELF::RefineModuleDetailsFromNote(lldb_private::DataExtractor &data, arch_spec.GetTriple().setVendor(llvm::Triple::VendorType::CSR); // TODO At some point the description string could be processed. - // It could provide a steer towards the kalimba variant which - // this ELF targets. + // It could provide a steer towards the kalimba variant which this ELF + // targets. if (note.n_descsz) { const char *cstr = data.GetCStr(&offset, llvm::alignTo(note.n_descsz, 4)); @@ -1384,36 +1343,28 @@ ObjectFileELF::RefineModuleDetailsFromNote(lldb_private::DataExtractor &data, // register info arch_spec.GetTriple().setOS(llvm::Triple::OSType::Linux); } else if (note.n_name == LLDB_NT_OWNER_CORE) { - // Parse the NT_FILE to look for stuff in paths to shared libraries - // As the contents look like this in a 64 bit ELF core file: - // count = 0x000000000000000a (10) - // page_size = 0x0000000000001000 (4096) - // Index start end file_ofs path - // ===== ------------------ ------------------ ------------------ - // ------------------------------------- - // [ 0] 0x0000000000400000 0x0000000000401000 0x0000000000000000 - // /tmp/a.out - // [ 1] 0x0000000000600000 0x0000000000601000 0x0000000000000000 - // /tmp/a.out - // [ 2] 0x0000000000601000 0x0000000000602000 0x0000000000000001 - // /tmp/a.out + // Parse the NT_FILE to look for stuff in paths to shared libraries As + // the contents look like this in a 64 bit ELF core file: count = + // 0x000000000000000a (10) page_size = 0x0000000000001000 (4096) Index + // start end file_ofs path ===== + // ------------------ ------------------ ------------------ + // ------------------------------------- [ 0] 0x0000000000400000 + // 0x0000000000401000 0x0000000000000000 /tmp/a.out [ 1] + // 0x0000000000600000 0x0000000000601000 0x0000000000000000 /tmp/a.out [ + // 2] 0x0000000000601000 0x0000000000602000 0x0000000000000001 /tmp/a.out // [ 3] 0x00007fa79c9ed000 0x00007fa79cba8000 0x0000000000000000 - // /lib/x86_64-linux-gnu/libc-2.19.so - // [ 4] 0x00007fa79cba8000 0x00007fa79cda7000 0x00000000000001bb - // /lib/x86_64-linux-gnu/libc-2.19.so - // [ 5] 0x00007fa79cda7000 0x00007fa79cdab000 0x00000000000001ba - // /lib/x86_64-linux-gnu/libc-2.19.so - // [ 6] 0x00007fa79cdab000 0x00007fa79cdad000 0x00000000000001be - // /lib/x86_64-linux-gnu/libc-2.19.so - // [ 7] 0x00007fa79cdb2000 0x00007fa79cdd5000 0x0000000000000000 - // /lib/x86_64-linux-gnu/ld-2.19.so - // [ 8] 0x00007fa79cfd4000 0x00007fa79cfd5000 0x0000000000000022 - // /lib/x86_64-linux-gnu/ld-2.19.so - // [ 9] 0x00007fa79cfd5000 0x00007fa79cfd6000 0x0000000000000023 - // /lib/x86_64-linux-gnu/ld-2.19.so - // In the 32 bit ELFs the count, page_size, start, end, file_ofs are - // uint32_t - // For reference: see readelf source code (in binutils). + // /lib/x86_64-linux-gnu/libc-2.19.so [ 4] 0x00007fa79cba8000 + // 0x00007fa79cda7000 0x00000000000001bb /lib/x86_64-linux- + // gnu/libc-2.19.so [ 5] 0x00007fa79cda7000 0x00007fa79cdab000 + // 0x00000000000001ba /lib/x86_64-linux-gnu/libc-2.19.so [ 6] + // 0x00007fa79cdab000 0x00007fa79cdad000 0x00000000000001be /lib/x86_64 + // -linux-gnu/libc-2.19.so [ 7] 0x00007fa79cdb2000 0x00007fa79cdd5000 + // 0x0000000000000000 /lib/x86_64-linux-gnu/ld-2.19.so [ 8] + // 0x00007fa79cfd4000 0x00007fa79cfd5000 0x0000000000000022 /lib/x86_64 + // -linux-gnu/ld-2.19.so [ 9] 0x00007fa79cfd5000 0x00007fa79cfd6000 + // 0x0000000000000023 /lib/x86_64-linux-gnu/ld-2.19.so In the 32 bit ELFs + // the count, page_size, start, end, file_ofs are uint32_t For reference: + // see readelf source code (in binutils). if (note.n_type == NT_FILE) { uint64_t count = data.GetAddress(&offset); const char *cstr; @@ -1437,15 +1388,14 @@ ObjectFileELF::RefineModuleDetailsFromNote(lldb_private::DataExtractor &data, } if (arch_spec.IsMIPS() && arch_spec.GetTriple().getOS() == llvm::Triple::OSType::UnknownOS) - // In case of MIPSR6, the LLDB_NT_OWNER_GNU note is missing - // for some cases (e.g. compile with -nostdlib) - // Hence set OS to Linux + // In case of MIPSR6, the LLDB_NT_OWNER_GNU note is missing for some + // cases (e.g. compile with -nostdlib) Hence set OS to Linux arch_spec.GetTriple().setOS(llvm::Triple::OSType::Linux); } } - // Calculate the offset of the next note just in case "offset" has been used - // to poke at the contents of the note data + // Calculate the offset of the next note just in case "offset" has been + // used to poke at the contents of the note data offset = note_offset + note.GetByteSize(); } @@ -1545,13 +1495,12 @@ size_t ObjectFileELF::GetSectionHeaderInfo(SectionHeaderColl §ion_headers, arch_spec.SetArchitecture(eArchTypeELF, header.e_machine, sub_type, header.e_ident[EI_OSABI]); - // Validate if it is ok to remove GetOsFromOSABI. - // Note, that now the OS is determined based on EI_OSABI flag and - // the info extracted from ELF notes (see RefineModuleDetailsFromNote). - // However in some cases that still might be not enough: for example - // a shared library might not have any notes at all - // and have EI_OSABI flag set to System V, - // as result the OS will be set to UnknownOS. + // Validate if it is ok to remove GetOsFromOSABI. Note, that now the OS is + // determined based on EI_OSABI flag and the info extracted from ELF notes + // (see RefineModuleDetailsFromNote). However in some cases that still + // might be not enough: for example a shared library might not have any + // notes at all and have EI_OSABI flag set to System V, as result the OS + // will be set to UnknownOS. GetOsFromOSABI(header.e_ident[EI_OSABI], ostype); spec_ostype = arch_spec.GetTriple().getOS(); assert(spec_ostype == ostype); @@ -1842,6 +1791,7 @@ void ObjectFileELF::CreateSections(SectionList &unified_section_list) { static ConstString g_sect_name_dwarf_debug_loc(".debug_loc"); static ConstString g_sect_name_dwarf_debug_macinfo(".debug_macinfo"); static ConstString g_sect_name_dwarf_debug_macro(".debug_macro"); + static ConstString g_sect_name_dwarf_debug_names(".debug_names"); static ConstString g_sect_name_dwarf_debug_pubnames(".debug_pubnames"); static ConstString g_sect_name_dwarf_debug_pubtypes(".debug_pubtypes"); static ConstString g_sect_name_dwarf_debug_ranges(".debug_ranges"); @@ -1857,10 +1807,12 @@ void ObjectFileELF::CreateSections(SectionList &unified_section_list) { static ConstString g_sect_name_dwarf_debug_str_dwo(".debug_str.dwo"); static ConstString g_sect_name_dwarf_debug_str_offsets_dwo( ".debug_str_offsets.dwo"); + static ConstString g_sect_name_dwarf_debug_types(".debug_types"); static ConstString g_sect_name_eh_frame(".eh_frame"); static ConstString g_sect_name_arm_exidx(".ARM.exidx"); static ConstString g_sect_name_arm_extab(".ARM.extab"); static ConstString g_sect_name_go_symtab(".gosymtab"); + static ConstString g_sect_name_dwarf_gnu_debugaltlink(".gnu_debugaltlink"); SectionType sect_type = eSectionTypeOther; @@ -1881,23 +1833,19 @@ void ObjectFileELF::CreateSections(SectionList &unified_section_list) { } // .debug_abbrev – Abbreviations used in the .debug_info section // .debug_aranges – Lookup table for mapping addresses to compilation - // units - // .debug_frame – Call frame information - // .debug_info – The core DWARF information section - // .debug_line – Line number information + // units .debug_frame – Call frame information .debug_info – The core + // DWARF information section .debug_line – Line number information // .debug_loc – Location lists used in DW_AT_location attributes - // .debug_macinfo – Macro information - // .debug_pubnames – Lookup table for mapping object and function names to - // compilation units + // .debug_macinfo – Macro information .debug_pubnames – Lookup table + // for mapping object and function names to compilation units // .debug_pubtypes – Lookup table for mapping type names to compilation - // units - // .debug_ranges – Address ranges used in DW_AT_ranges attributes - // .debug_str – String table used in .debug_info - // MISSING? .gnu_debugdata - "mini debuginfo / MiniDebugInfo" section, - // http://sourceware.org/gdb/onlinedocs/gdb/MiniDebugInfo.html - // MISSING? .debug-index - - // http://src.chromium.org/viewvc/chrome/trunk/src/build/gdb-add-index?pathrev=144644 - // MISSING? .debug_types - Type descriptions from DWARF 4? See + // units .debug_ranges – Address ranges used in DW_AT_ranges attributes + // .debug_str – String table used in .debug_info MISSING? + // .gnu_debugdata - "mini debuginfo / MiniDebugInfo" section, + // http://sourceware.org/gdb/onlinedocs/gdb/MiniDebugInfo.html MISSING? + // .debug-index - http://src.chromium.org/viewvc/chrome/trunk/src/build + // /gdb-add-index?pathrev=144644 MISSING? .debug_types - Type + // descriptions from DWARF 4? See // http://gcc.gnu.org/wiki/DwarfSeparateTypeInfo else if (name == g_sect_name_dwarf_debug_abbrev) sect_type = eSectionTypeDWARFDebugAbbrev; @@ -1919,6 +1867,8 @@ void ObjectFileELF::CreateSections(SectionList &unified_section_list) { sect_type = eSectionTypeDWARFDebugMacInfo; else if (name == g_sect_name_dwarf_debug_macro) sect_type = eSectionTypeDWARFDebugMacro; + else if (name == g_sect_name_dwarf_debug_names) + sect_type = eSectionTypeDWARFDebugNames; else if (name == g_sect_name_dwarf_debug_pubnames) sect_type = eSectionTypeDWARFDebugPubNames; else if (name == g_sect_name_dwarf_debug_pubtypes) @@ -1927,6 +1877,8 @@ void ObjectFileELF::CreateSections(SectionList &unified_section_list) { sect_type = eSectionTypeDWARFDebugRanges; else if (name == g_sect_name_dwarf_debug_str) sect_type = eSectionTypeDWARFDebugStr; + else if (name == g_sect_name_dwarf_debug_types) + sect_type = eSectionTypeDWARFDebugTypes; else if (name == g_sect_name_dwarf_debug_str_offsets) sect_type = eSectionTypeDWARFDebugStrOffsets; else if (name == g_sect_name_dwarf_debug_abbrev_dwo) @@ -1951,6 +1903,8 @@ void ObjectFileELF::CreateSections(SectionList &unified_section_list) { sect_type = eSectionTypeARMextab; else if (name == g_sect_name_go_symtab) sect_type = eSectionTypeGoSymtab; + else if (name == g_sect_name_dwarf_gnu_debugaltlink) + sect_type = eSectionTypeDWARFGNUDebugAltLink; const uint32_t permissions = ((header.sh_flags & SHF_ALLOC) ? ePermissionsReadable : 0u) | @@ -1978,13 +1932,21 @@ void ObjectFileELF::CreateSections(SectionList &unified_section_list) { if (eSectionTypeOther == sect_type) { // the kalimba toolchain assumes that ELF section names are free-form. - // It does - // support linkscripts which (can) give rise to various arbitrarily - // named - // sections being "Code" or "Data". + // It does support linkscripts which (can) give rise to various + // arbitrarily named sections being "Code" or "Data". sect_type = kalimbaSectionType(m_header, header); } + // In common case ELF code section can have arbitrary name (for example, + // we can specify it using section attribute for particular function) so + // assume that section is a code section if it has SHF_EXECINSTR flag set + // and has SHT_PROGBITS type. + if (eSectionTypeOther == sect_type && + llvm::ELF::SHT_PROGBITS == header.sh_type && + (header.sh_flags & SHF_EXECINSTR)) { + sect_type = eSectionTypeCode; + } + const uint32_t target_bytes_size = (eSectionTypeData == sect_type || eSectionTypeZeroFill == sect_type) ? m_arch_spec.GetDataByteSize() @@ -2024,46 +1986,16 @@ void ObjectFileELF::CreateSections(SectionList &unified_section_list) { } } - if (m_sections_ap.get()) { - if (GetType() == eTypeDebugInfo) { - static const SectionType g_sections[] = { - eSectionTypeDWARFDebugAbbrev, eSectionTypeDWARFDebugAddr, - eSectionTypeDWARFDebugAranges, eSectionTypeDWARFDebugCuIndex, - eSectionTypeDWARFDebugFrame, eSectionTypeDWARFDebugInfo, - eSectionTypeDWARFDebugLine, eSectionTypeDWARFDebugLoc, - eSectionTypeDWARFDebugMacInfo, eSectionTypeDWARFDebugPubNames, - eSectionTypeDWARFDebugPubTypes, eSectionTypeDWARFDebugRanges, - eSectionTypeDWARFDebugStr, eSectionTypeDWARFDebugStrOffsets, - eSectionTypeELFSymbolTable, - }; - SectionList *elf_section_list = m_sections_ap.get(); - for (size_t idx = 0; idx < sizeof(g_sections) / sizeof(g_sections[0]); - ++idx) { - SectionType section_type = g_sections[idx]; - SectionSP section_sp( - elf_section_list->FindSectionByType(section_type, true)); - if (section_sp) { - SectionSP module_section_sp( - unified_section_list.FindSectionByType(section_type, true)); - if (module_section_sp) - unified_section_list.ReplaceSection(module_section_sp->GetID(), - section_sp); - else - unified_section_list.AddSection(section_sp); - } - } - } else { - unified_section_list = *m_sections_ap; - } - } + // For eTypeDebugInfo files, the Symbol Vendor will take care of updating the + // unified section list. + if (GetType() != eTypeDebugInfo) + unified_section_list = *m_sections_ap; } // Find the arm/aarch64 mapping symbol character in the given symbol name. -// Mapping symbols have the -// form of "$<char>[.<any>]*". Additionally we recognize cases when the mapping -// symbol prefixed by -// an arbitrary string because if a symbol prefix added to each symbol in the -// object file with +// Mapping symbols have the form of "$<char>[.<any>]*". Additionally we +// recognize cases when the mapping symbol prefixed by an arbitrary string +// because if a symbol prefix added to each symbol in the object file with // objcopy then the mapping symbols are also prefixed. static char FindArmAarch64MappingSymbol(const char *symbol_name) { if (!symbol_name) @@ -2105,22 +2037,18 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, static ConstString opd_section_name(".opd"); // For ppc64 // On Android the oatdata and the oatexec symbols in the oat and odex files - // covers the full - // .text section what causes issues with displaying unusable symbol name to - // the user and very - // slow unwinding speed because the instruction emulation based unwind plans - // try to emulate all - // instructions in these symbols. Don't add these symbols to the symbol list - // as they have no - // use for the debugger and they are causing a lot of trouble. - // Filtering can't be restricted to Android because this special object file - // don't contain the - // note section specifying the environment to Android but the custom extension - // and file name - // makes it highly unlikely that this will collide with anything else. + // covers the full .text section what causes issues with displaying unusable + // symbol name to the user and very slow unwinding speed because the + // instruction emulation based unwind plans try to emulate all instructions + // in these symbols. Don't add these symbols to the symbol list as they have + // no use for the debugger and they are causing a lot of trouble. Filtering + // can't be restricted to Android because this special object file don't + // contain the note section specifying the environment to Android but the + // custom extension and file name makes it highly unlikely that this will + // collide with anything else. ConstString file_extension = m_file.GetFileNameExtension(); - bool skip_oatdata_oatexec = file_extension == ConstString("oat") || - file_extension == ConstString("odex"); + bool skip_oatdata_oatexec = file_extension == ConstString(".oat") || + file_extension == ConstString(".odex"); ArchSpec arch; GetArchitecture(arch); @@ -2129,8 +2057,8 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, module_sp ? module_sp->GetSectionList() : nullptr; // Local cache to avoid doing a FindSectionByName for each symbol. The "const - // char*" key must - // came from a ConstString object so they can be compared by pointer + // char*" key must came from a ConstString object so they can be compared by + // pointer std::unordered_map<const char *, lldb::SectionSP> section_name_to_section; unsigned i; @@ -2148,8 +2076,7 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, continue; // Skipping oatdata and oatexec sections if it is requested. See details - // above the - // definition of skip_oatdata_oatexec for the reasons. + // above the definition of skip_oatdata_oatexec for the reasons. if (skip_oatdata_oatexec && (::strcmp(symbol_name, "oatdata") == 0 || ::strcmp(symbol_name, "oatexec") == 0)) continue; @@ -2180,8 +2107,8 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, break; case STT_OBJECT: - // The symbol is associated with a data object, such as a variable, - // an array, etc. + // The symbol is associated with a data object, such as a variable, an + // array, etc. symbol_type = eSymbolTypeData; break; @@ -2192,13 +2119,13 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, case STT_SECTION: // The symbol is associated with a section. Symbol table entries of - // this type exist primarily for relocation and normally have - // STB_LOCAL binding. + // this type exist primarily for relocation and normally have STB_LOCAL + // binding. break; case STT_FILE: - // Conventionally, the symbol's name gives the name of the source - // file associated with the object file. A file symbol has STB_LOCAL + // Conventionally, the symbol's name gives the name of the source file + // associated with the object file. A file symbol has STB_LOCAL // binding, its section index is SHN_ABS, and it precedes the other // STB_LOCAL symbols for the file, if it is present. symbol_type = eSymbolTypeSourceFile; @@ -2240,18 +2167,18 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, switch (mapping_symbol) { case 'a': // $a[.<any>]* - marks an ARM instruction sequence - m_address_class_map[symbol.st_value] = eAddressClassCode; + m_address_class_map[symbol.st_value] = AddressClass::eCode; break; case 'b': case 't': // $b[.<any>]* - marks a THUMB BL instruction sequence // $t[.<any>]* - marks a THUMB instruction sequence m_address_class_map[symbol.st_value] = - eAddressClassCodeAlternateISA; + AddressClass::eCodeAlternateISA; break; case 'd': // $d[.<any>]* - marks a data item sequence (e.g. lit pool) - m_address_class_map[symbol.st_value] = eAddressClassData; + m_address_class_map[symbol.st_value] = AddressClass::eData; break; } } @@ -2265,11 +2192,11 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, switch (mapping_symbol) { case 'x': // $x[.<any>]* - marks an A64 instruction sequence - m_address_class_map[symbol.st_value] = eAddressClassCode; + m_address_class_map[symbol.st_value] = AddressClass::eCode; break; case 'd': // $d[.<any>]* - marks a data item sequence (e.g. lit pool) - m_address_class_map[symbol.st_value] = eAddressClassData; + m_address_class_map[symbol.st_value] = AddressClass::eData; break; } } @@ -2281,18 +2208,17 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, if (arch.GetMachine() == llvm::Triple::arm) { if (symbol_type == eSymbolTypeCode) { if (symbol.st_value & 1) { - // Subtracting 1 from the address effectively unsets - // the low order bit, which results in the address - // actually pointing to the beginning of the symbol. - // This delta will be used below in conjunction with - // symbol.st_value to produce the final symbol_value - // that we store in the symtab. + // Subtracting 1 from the address effectively unsets the low order + // bit, which results in the address actually pointing to the + // beginning of the symbol. This delta will be used below in + // conjunction with symbol.st_value to produce the final + // symbol_value that we store in the symtab. symbol_value_offset = -1; m_address_class_map[symbol.st_value ^ 1] = - eAddressClassCodeAlternateISA; + AddressClass::eCodeAlternateISA; } else { // This address is ARM - m_address_class_map[symbol.st_value] = eAddressClassCode; + m_address_class_map[symbol.st_value] = AddressClass::eCode; } } } @@ -2317,36 +2243,32 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, llvm_arch == llvm::Triple::mips64 || llvm_arch == llvm::Triple::mips64el) { if (IS_MICROMIPS(symbol.st_other)) - m_address_class_map[symbol.st_value] = eAddressClassCodeAlternateISA; + m_address_class_map[symbol.st_value] = AddressClass::eCodeAlternateISA; else if ((symbol.st_value & 1) && (symbol_type == eSymbolTypeCode)) { symbol.st_value = symbol.st_value & (~1ull); - m_address_class_map[symbol.st_value] = eAddressClassCodeAlternateISA; + m_address_class_map[symbol.st_value] = AddressClass::eCodeAlternateISA; } else { if (symbol_type == eSymbolTypeCode) - m_address_class_map[symbol.st_value] = eAddressClassCode; + m_address_class_map[symbol.st_value] = AddressClass::eCode; else if (symbol_type == eSymbolTypeData) - m_address_class_map[symbol.st_value] = eAddressClassData; + m_address_class_map[symbol.st_value] = AddressClass::eData; else - m_address_class_map[symbol.st_value] = eAddressClassUnknown; + m_address_class_map[symbol.st_value] = AddressClass::eUnknown; } } } // symbol_value_offset may contain 0 for ARM symbols or -1 for THUMB - // symbols. See above for - // more details. + // symbols. See above for more details. uint64_t symbol_value = symbol.st_value + symbol_value_offset; if (symbol_section_sp == nullptr && section_idx == SHN_ABS && symbol.st_size != 0) { // We don't have a section for a symbol with non-zero size. Create a new - // section for it - // so the address range covered by the symbol is also covered by the - // module (represented - // through the section list). It is needed so module lookup for the - // addresses covered - // by this symbol will be successfull. This case happens for absolute - // symbols. + // section for it so the address range covered by the symbol is also + // covered by the module (represented through the section list). It is + // needed so module lookup for the addresses covered by this symbol will + // be successfull. This case happens for absolute symbols. ConstString fake_section_name(std::string(".absolute.") + symbol_name); symbol_section_sp = std::make_shared<Section>(module_sp, this, SHN_ABS, fake_section_name, @@ -2389,8 +2311,7 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, Mangled mangled(ConstString(symbol_bare), is_mangled); // Now append the suffix back to mangled and unmangled names. Only do it if - // the - // demangling was successful (string is not empty). + // the demangling was successful (string is not empty). if (has_suffix) { llvm::StringRef suffix = symbol_ref.substr(version_pos); @@ -2406,12 +2327,10 @@ unsigned ObjectFileELF::ParseSymbols(Symtab *symtab, user_id_t start_id, } // In ELF all symbol should have a valid size but it is not true for some - // function symbols - // coming from hand written assembly. As none of the function symbol should - // have 0 size we - // try to calculate the size for these symbols in the symtab with saying - // that their original - // size is not valid. + // function symbols coming from hand written assembly. As none of the + // function symbol should have 0 size we try to calculate the size for + // these symbols in the symtab with saying that their original size is not + // valid. bool symbol_size_valid = symbol.st_size != 0 || symbol.getType() != STT_FUNC; @@ -2440,8 +2359,7 @@ unsigned ObjectFileELF::ParseSymbolTable(Symtab *symbol_table, lldb_private::Section *symtab) { if (symtab->GetObjectFile() != this) { // If the symbol table section is owned by a different object file, have it - // do the - // parsing. + // do the parsing. ObjectFileELF *obj_file_elf = static_cast<ObjectFileELF *>(symtab->GetObjectFile()); return obj_file_elf->ParseSymbolTable(symbol_table, start_id, symtab); @@ -2457,8 +2375,8 @@ unsigned ObjectFileELF::ParseSymbolTable(Symtab *symbol_table, assert(symtab_hdr->sh_type == SHT_SYMTAB || symtab_hdr->sh_type == SHT_DYNSYM); - // sh_link: section header index of associated string table. - // Section ID's are ones based. + // sh_link: section header index of associated string table. Section ID's are + // ones based. user_id_t strtab_id = symtab_hdr->sh_link + 1; Section *strtab = section_list->FindSectionByID(strtab_id).get(); @@ -2543,19 +2461,17 @@ unsigned ObjectFileELF::PLTRelocationType() { return 0; } -// Returns the size of the normal plt entries and the offset of the first normal -// plt entry. The -// 0th entry in the plt table is usually a resolution entry which have different -// size in some -// architectures then the rest of the plt entries. +// Returns the size of the normal plt entries and the offset of the first +// normal plt entry. The 0th entry in the plt table is usually a resolution +// entry which have different size in some architectures then the rest of the +// plt entries. static std::pair<uint64_t, uint64_t> GetPltEntrySizeAndOffset(const ELFSectionHeader *rel_hdr, const ELFSectionHeader *plt_hdr) { const elf_xword num_relocations = rel_hdr->sh_size / rel_hdr->sh_entsize; - // Clang 3.3 sets entsize to 4 for 32-bit binaries, but the plt entries are 16 - // bytes. - // So round the entsize up by the alignment if addralign is set. + // Clang 3.3 sets entsize to 4 for 32-bit binaries, but the plt entries are + // 16 bytes. So round the entsize up by the alignment if addralign is set. elf_xword plt_entsize = plt_hdr->sh_addralign ? llvm::alignTo(plt_hdr->sh_entsize, plt_hdr->sh_addralign) @@ -2567,12 +2483,10 @@ GetPltEntrySizeAndOffset(const ELFSectionHeader *rel_hdr, // just in case. if (plt_entsize <= 4) { // The linker haven't set the plt_hdr->sh_entsize field. Try to guess the - // size of the plt - // entries based on the number of entries and the size of the plt section - // with the - // assumption that the size of the 0th entry is at least as big as the size - // of the normal - // entries and it isn't much bigger then that. + // size of the plt entries based on the number of entries and the size of + // the plt section with the assumption that the size of the 0th entry is at + // least as big as the size of the normal entries and it isn't much bigger + // then that. if (plt_hdr->sh_addralign) plt_entsize = plt_hdr->sh_size / plt_hdr->sh_addralign / (num_relocations + 1) * plt_hdr->sh_addralign; @@ -2865,8 +2779,7 @@ Symtab *ObjectFileELF::GetSymtab() { return NULL; // We always want to use the main object file so we (hopefully) only have one - // cached copy - // of our symtab, dynamic sections, etc. + // cached copy of our symtab, dynamic sections, etc. ObjectFile *module_obj_file = module_sp->GetObjectFile(); if (module_obj_file && module_obj_file != this) return module_obj_file->GetSymtab(); @@ -2881,18 +2794,15 @@ Symtab *ObjectFileELF::GetSymtab() { // Sharable objects and dynamic executables usually have 2 distinct symbol // tables, one named ".symtab", and the other ".dynsym". The dynsym is a - // smaller - // version of the symtab that only contains global symbols. The information - // found - // in the dynsym is therefore also found in the symtab, while the reverse is - // not - // necessarily true. + // smaller version of the symtab that only contains global symbols. The + // information found in the dynsym is therefore also found in the symtab, + // while the reverse is not necessarily true. Section *symtab = section_list->FindSectionByType(eSectionTypeELFSymbolTable, true).get(); if (!symtab) { // The symtab section is non-allocable and can be stripped, so if it - // doesn't exist - // then use the dynsym section which should always be there. + // doesn't exist then use the dynsym section which should always be + // there. symtab = section_list->FindSectionByType(eSectionTypeELFDynamicSymbols, true) .get(); @@ -2939,8 +2849,7 @@ Symtab *ObjectFileELF::GetSymtab() { } // If we still don't have any symtab then create an empty instance to avoid - // do the section - // lookup next time. + // do the section lookup next time. if (m_symtab_ap == nullptr) m_symtab_ap.reset(new Symtab(this)); @@ -2954,8 +2863,8 @@ void ObjectFileELF::RelocateSection(lldb_private::Section *section) { static const char *debug_prefix = ".debug"; - // Set relocated bit so we stop getting called, regardless of - // whether we actually relocate. + // Set relocated bit so we stop getting called, regardless of whether we + // actually relocate. section->SetIsRelocated(true); // We only relocate in ELF relocatable files @@ -2998,12 +2907,10 @@ void ObjectFileELF::ParseUnwindSymbols(Symtab *symbol_table, return; // First we save the new symbols into a separate list and add them to the - // symbol table after - // we colleced all symbols we want to add. This is neccessary because adding a - // new symbol - // invalidates the internal index of the symtab what causing the next lookup - // to be slow because - // it have to recalculate the index first. + // symbol table after we colleced all symbols we want to add. This is + // neccessary because adding a new symbol invalidates the internal index of + // the symtab what causing the next lookup to be slow because it have to + // recalculate the index first. std::vector<Symbol> new_symbols; eh_frame->ForEachFDEEntries([this, symbol_table, section_list, &new_symbols]( @@ -3197,8 +3104,8 @@ void ObjectFileELF::DumpELFProgramHeader(Stream *s, //---------------------------------------------------------------------- // DumpELFProgramHeader_p_type // -// Dump an token value for the ELF program header member p_type which -// describes the type of the program header +// Dump an token value for the ELF program header member p_type which describes +// the type of the program header // ---------------------------------------------------------------------- void ObjectFileELF::DumpELFProgramHeader_p_type(Stream *s, elf_word p_type) { const int kStrWidth = 15; @@ -3369,8 +3276,7 @@ bool ObjectFileELF::GetArchitecture(ArchSpec &arch) { if (CalculateType() == eTypeCoreFile && m_arch_spec.TripleOSIsUnspecifiedUnknown()) { // Core files don't have section headers yet they have PT_NOTE program - // headers - // that might shed more light on the architecture + // headers that might shed more light on the architecture if (ParseProgramHeaders()) { for (size_t i = 1, count = GetProgramHeaderCount(); i <= count; ++i) { const elf::ELFProgramHeader *header = GetProgramHeaderByIndex(i); @@ -3431,22 +3337,22 @@ ObjectFile::Strata ObjectFileELF::CalculateStrata() { case llvm::ELF::ET_EXEC: // 2 - Executable file // TODO: is there any way to detect that an executable is a kernel - // related executable by inspecting the program headers, section - // headers, symbols, or any other flag bits??? + // related executable by inspecting the program headers, section headers, + // symbols, or any other flag bits??? return eStrataUser; case llvm::ELF::ET_DYN: // 3 - Shared object file // TODO: is there any way to detect that an shared library is a kernel - // related executable by inspecting the program headers, section - // headers, symbols, or any other flag bits??? + // related executable by inspecting the program headers, section headers, + // symbols, or any other flag bits??? return eStrataUnknown; case ET_CORE: // 4 - Core file // TODO: is there any way to detect that an core file is a kernel - // related executable by inspecting the program headers, section - // headers, symbols, or any other flag bits??? + // related executable by inspecting the program headers, section headers, + // symbols, or any other flag bits??? return eStrataUnknown; default: @@ -3491,8 +3397,9 @@ size_t ObjectFileELF::ReadSectionData(Section *section, size_t(section_data.GetByteSize())}, GetByteOrder() == eByteOrderLittle, GetAddressByteSize() == 8); if (!Decompressor) { - LLDB_LOG(log, "Unable to initialize decompressor for section {0}: {1}", - section->GetName(), llvm::toString(Decompressor.takeError())); + LLDB_LOG_ERROR(log, Decompressor.takeError(), + "Unable to initialize decompressor for section {0}", + section->GetName()); return result; } auto buffer_sp = @@ -3500,10 +3407,45 @@ size_t ObjectFileELF::ReadSectionData(Section *section, if (auto Error = Decompressor->decompress( {reinterpret_cast<char *>(buffer_sp->GetBytes()), size_t(buffer_sp->GetByteSize())})) { - LLDB_LOG(log, "Decompression of section {0} failed: {1}", - section->GetName(), llvm::toString(std::move(Error))); + LLDB_LOG_ERROR(log, std::move(Error), "Decompression of section {0} failed", + section->GetName()); return result; } section_data.SetData(buffer_sp); return buffer_sp->GetByteSize(); } + +bool ObjectFileELF::AnySegmentHasPhysicalAddress() { + size_t header_count = ParseProgramHeaders(); + for (size_t i = 1; i <= header_count; ++i) { + auto header = GetProgramHeaderByIndex(i); + if (header->p_paddr != 0) + return true; + } + return false; +} + +std::vector<ObjectFile::LoadableData> +ObjectFileELF::GetLoadableData(Target &target) { + // Create a list of loadable data from loadable segments, using physical + // addresses if they aren't all null + std::vector<LoadableData> loadables; + size_t header_count = ParseProgramHeaders(); + bool should_use_paddr = AnySegmentHasPhysicalAddress(); + for (size_t i = 1; i <= header_count; ++i) { + LoadableData loadable; + auto header = GetProgramHeaderByIndex(i); + if (header->p_type != llvm::ELF::PT_LOAD) + continue; + loadable.Dest = should_use_paddr ? header->p_paddr : header->p_vaddr; + if (loadable.Dest == LLDB_INVALID_ADDRESS) + continue; + if (header->p_filesz == 0) + continue; + auto segment_data = GetSegmentDataByIndex(i); + loadable.Contents = llvm::ArrayRef<uint8_t>(segment_data.GetDataStart(), + segment_data.GetByteSize()); + loadables.push_back(loadable); + } + return loadables; +} |