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