Home | History | Annotate | Download | only in Scripts 
      1 #!/usr/bin/env python
      2 
      3 import struct
      4 import sys
      5 import StringIO
      6 
      7 import common_dump
      8 
      9 class Reader:
     10     def __init__(self, path):
     11         if path == "-":
     12             # Snarf all the data so we can seek.
     13             self.file = StringIO.StringIO(sys.stdin.read())
     14         else:
     15             self.file = open(path, "rb")
     16         self.isLSB = None
     17         self.is64Bit = None
     18 
     19     def seek(self, pos):
     20         self.file.seek(pos)
     21 
     22     def read(self, N):
     23         data = self.file.read(N)
     24         if len(data) != N:
     25             raise ValueError, "Out of data!"
     26         return data
     27 
     28     def read8(self):
     29         return ord(self.read(1))
     30 
     31     def read16(self):
     32         return struct.unpack('><'[self.isLSB] + 'H', self.read(2))[0]
     33 
     34     def read32(self):
     35         return struct.unpack('><'[self.isLSB] + 'I', self.read(4))[0]
     36 
     37     def read32S(self):
     38         return struct.unpack('><'[self.isLSB] + 'i', self.read(4))[0]
     39 
     40     def read64(self):
     41         return struct.unpack('><'[self.isLSB] + 'Q', self.read(8))[0]
     42 
     43     def read64S(self):
     44         return struct.unpack('><'[self.isLSB] + 'q', self.read(8))[0]
     45 
     46     def readWord(self):
     47         if self.is64Bit:
     48             return self.read64()
     49         else:
     50             return self.read32()
     51 
     52     def readWordS(self):
     53         if self.is64Bit:
     54             return self.read64S()
     55         else:
     56             return self.read32S()
     57 
     58 class StringTable:
     59     def __init__(self, strings):
     60        self.string_table = strings
     61 
     62     def __getitem__(self, index):
     63        end = self.string_table.index('\x00', index)
     64        return self.string_table[index:end]
     65 
     66 class Section:
     67     def __init__(self, f):
     68         self.sh_name = f.read32()
     69         self.sh_type = f.read32()
     70         self.sh_flags = f.readWord()
     71         self.sh_addr = f.readWord()
     72         self.sh_offset = f.readWord()
     73         self.sh_size = f.readWord()
     74         self.sh_link = f.read32()
     75         self.sh_info = f.read32()
     76         self.sh_addralign = f.readWord()
     77         self.sh_entsize = f.readWord()
     78 
     79     def dump(self, shstrtab, f, strtab, dumpdata):
     80         print "  (('sh_name', %s)" % common_dump.HexDump(self.sh_name), "# %r" % shstrtab[self.sh_name]
     81         print "   ('sh_type', %s)" % common_dump.HexDump(self.sh_type)
     82         print "   ('sh_flags', %s)" % common_dump.HexDump(self.sh_flags)
     83         print "   ('sh_addr', %s)" % common_dump.HexDump(self.sh_addr)
     84         print "   ('sh_offset', %s)" % common_dump.HexDump(self.sh_offset)
     85         print "   ('sh_size', %s)" % common_dump.HexDump(self.sh_size)
     86         print "   ('sh_link', %s)" % common_dump.HexDump(self.sh_link)
     87         print "   ('sh_info', %s)" % common_dump.HexDump(self.sh_info)
     88         print "   ('sh_addralign', %s)" % common_dump.HexDump(self.sh_addralign)
     89         print "   ('sh_entsize', %s)" % common_dump.HexDump(self.sh_entsize)
     90         if self.sh_type == 2: # SHT_SYMTAB
     91             print "   ('_symbols', ["
     92             dumpSymtab(f, self, strtab)
     93             print "   ])"
     94         elif self.sh_type == 4 or self.sh_type == 9: # SHT_RELA / SHT_REL
     95             print "   ('_relocations', ["
     96             dumpRel(f, self, self.sh_type == 4)
     97             print "   ])"
     98         elif dumpdata:
     99             f.seek(self.sh_offset)
    100             if self.sh_type != 8: # != SHT_NOBITS
    101                 data = f.read(self.sh_size)
    102                 print "   ('_section_data', '%s')" % common_dump.dataToHex(data)
    103             else:
    104                 print "   ('_section_data', '')" 
    105         print "  ),"
    106 
    107 def dumpSymtab(f, section, strtab):
    108     entries = section.sh_size // section.sh_entsize
    109 
    110     for index in range(entries):
    111         f.seek(section.sh_offset + index * section.sh_entsize)
    112         print "    # Symbol %s" % common_dump.HexDump(index)
    113         name = f.read32()
    114         print "    (('st_name', %s)" % common_dump.HexDump(name), "# %r" % strtab[name]
    115         if not f.is64Bit:
    116             print "     ('st_value', %s)" % common_dump.HexDump(f.read32())
    117             print "     ('st_size', %s)" % common_dump.HexDump(f.read32())
    118         st_info = f.read8()
    119         print "     ('st_bind', %s)" % common_dump.HexDump((st_info >> 4))
    120         print "     ('st_type', %s)" % common_dump.HexDump((st_info & 0xf))
    121         print "     ('st_other', %s)" % common_dump.HexDump(f.read8())
    122         print "     ('st_shndx', %s)" % common_dump.HexDump(f.read16())
    123         if f.is64Bit:
    124             print "     ('st_value', %s)" % common_dump.HexDump(f.read64(), 64)
    125             print "     ('st_size', %s)" % common_dump.HexDump(f.read64(), 64)
    126         print "    ),"
    127 
    128 def dumpRel(f, section, dumprela = False):
    129     entries = section.sh_size // section.sh_entsize
    130 
    131     for index in range(entries):
    132         f.seek(section.sh_offset + index * section.sh_entsize)
    133         print "    # Relocation %s" % common_dump.HexDump(index)
    134         print "    (('r_offset', %s)" % common_dump.HexDump(f.readWord())
    135         r_info = f.readWord()
    136         if f.is64Bit:
    137             print "     ('r_sym', %s)" % common_dump.HexDump((r_info >> 32))
    138             print "     ('r_type', %s)" % common_dump.HexDump((r_info & 0xffffffff))
    139         else:
    140             print "     ('r_sym', %s)" % common_dump.HexDump((r_info >> 8))
    141             print "     ('r_type', %s)" % common_dump.HexDump((r_info & 0xff))
    142         if dumprela:
    143             print "     ('r_addend', %s)" % common_dump.HexDump(f.readWordS())
    144         print "    ),"
    145 
    146 def dumpELF(path, opts):
    147     f = Reader(path)
    148 
    149     magic = f.read(4)
    150     assert magic == '\x7FELF'
    151 
    152     fileclass = f.read8()
    153     if fileclass == 1: # ELFCLASS32
    154         f.is64Bit = False
    155     elif fileclass == 2: # ELFCLASS64
    156         f.is64Bit = True
    157     else:
    158         raise ValueError, "Unknown file class %s" % common_dump.HexDump(fileclass)
    159     print "('e_indent[EI_CLASS]', %s)" % common_dump.HexDump(fileclass)
    160 
    161     byteordering = f.read8()
    162     if byteordering == 1: # ELFDATA2LSB
    163         f.isLSB = True
    164     elif byteordering == 2: # ELFDATA2MSB
    165         f.isLSB = False
    166     else:
    167         raise ValueError, "Unknown byte ordering %s" % common_dump.HexDump(byteordering)
    168     print "('e_indent[EI_DATA]', %s)" % common_dump.HexDump(byteordering)
    169 
    170     print "('e_indent[EI_VERSION]', %s)" % common_dump.HexDump(f.read8())
    171     print "('e_indent[EI_OSABI]', %s)" % common_dump.HexDump(f.read8())
    172     print "('e_indent[EI_ABIVERSION]', %s)" % common_dump.HexDump(f.read8())
    173 
    174     f.seek(16) # Seek to end of e_ident.
    175 
    176     print "('e_type', %s)" % common_dump.HexDump(f.read16())
    177     print "('e_machine', %s)" % common_dump.HexDump(f.read16())
    178     print "('e_version', %s)" % common_dump.HexDump(f.read32())
    179     print "('e_entry', %s)" % common_dump.HexDump(f.readWord())
    180     print "('e_phoff', %s)" % common_dump.HexDump(f.readWord())
    181     e_shoff = f.readWord()
    182     print "('e_shoff', %s)" % common_dump.HexDump(e_shoff)
    183     print "('e_flags', %s)" % common_dump.HexDump(f.read32())
    184     print "('e_ehsize', %s)" % common_dump.HexDump(f.read16())
    185     print "('e_phentsize', %s)" % common_dump.HexDump(f.read16())
    186     print "('e_phnum', %s)" % common_dump.HexDump(f.read16())
    187     e_shentsize = f.read16()
    188     print "('e_shentsize', %s)" % common_dump.HexDump(e_shentsize)
    189     e_shnum = f.read16()
    190     print "('e_shnum', %s)" % common_dump.HexDump(e_shnum)
    191     e_shstrndx = f.read16()
    192     print "('e_shstrndx', %s)" % common_dump.HexDump(e_shstrndx)
    193 
    194     # Read all section headers
    195     sections = []
    196     for index in range(e_shnum):
    197         f.seek(e_shoff + index * e_shentsize)
    198         s = Section(f)
    199         sections.append(s)
    200 
    201     # Read .shstrtab so we can resolve section names
    202     f.seek(sections[e_shstrndx].sh_offset)
    203     shstrtab = StringTable(f.read(sections[e_shstrndx].sh_size))
    204 
    205     # Get the symbol string table
    206     strtab = None
    207     for section in sections:
    208         if shstrtab[section.sh_name] == ".strtab":
    209             f.seek(section.sh_offset)
    210             strtab = StringTable(f.read(section.sh_size))
    211             break
    212 
    213     print "('_sections', ["
    214     for index in range(e_shnum):
    215         print "  # Section %s" % common_dump.HexDump(index)
    216         sections[index].dump(shstrtab, f, strtab, opts.dumpSectionData)
    217     print "])"
    218 
    219 if __name__ == "__main__":
    220     from optparse import OptionParser, OptionGroup
    221     parser = OptionParser("usage: %prog [options] {files}")
    222     parser.add_option("", "--dump-section-data", dest="dumpSectionData",
    223                       help="Dump the contents of sections",
    224                       action="store_true", default=False)
    225     (opts, args) = parser.parse_args()
    226 
    227     if not args:
    228         args.append('-')
    229 
    230     for arg in args:
    231         dumpELF(arg, opts)
    232