examples/dwarf_decode_address.py

   1 #-------------------------------------------------------------------------------
   2 # elftools example: dwarf_decode_address.py
   3 #
   4 # Decode an address in an ELF file to find out which function it belongs to
   5 # and from which filename/line it comes in the original source file.
   6 #
   7 # Eli Bendersky (eliben@gmail.com)
   8 # This code is in the public domain
   9 #-------------------------------------------------------------------------------
  10 from __future__ import print_function
  11 import sys
  12
  13 # If pyelftools is not installed, the example can also run from the root or
  14 # examples/ dir of the source distribution.
  15 sys.path[0:0] = ['.', '..']
  16
  17 from elftools.common.py3compat import maxint, bytes2str
  18 from elftools.elf.elffile import ELFFile
  19
  20
  21 def process_file(filename, address):
  22     print('Processing file:', filename)
  23     with open(filename, 'rb') as f:
  24         elffile = ELFFile(f)
  25
  26         if not elffile.has_dwarf_info():
  27             print('  file has no DWARF info')
  28             return
  29
  30         # get_dwarf_info returns a DWARFInfo context object, which is the
  31         # starting point for all DWARF-based processing in pyelftools.
  32         dwarfinfo = elffile.get_dwarf_info()
  33
  34         funcname = decode_funcname(dwarfinfo, address)
  35         file, line = decode_file_line(dwarfinfo, address)
  36
  37         print('Function:', bytes2str(funcname))
  38         print('File:', bytes2str(file))
  39         print('Line:', line)
  40
  41
  42 def decode_funcname(dwarfinfo, address):
  43     # Go over all DIEs in the DWARF information, looking for a subprogram
  44     # entry with an address range that includes the given address. Note that
  45     # this simplifies things by disregarding subprograms that may have
  46     # split address ranges.
  47     for CU in dwarfinfo.iter_CUs():
  48         for DIE in CU.iter_DIEs():
  49             try:
  50                 if DIE.tag == 'DW_TAG_subprogram':
  51                     lowpc = DIE.attributes['DW_AT_low_pc'].value
  52                     highpc = DIE.attributes['DW_AT_high_pc'].value
  53                     if lowpc <= address <= highpc:
  54                         return DIE.attributes['DW_AT_name'].value
  55             except KeyError:
  56                 continue
  57     return None
  58
  59
  60 def decode_file_line(dwarfinfo, address):
  61     # Go over all the line programs in the DWARF information, looking for
  62     # one that describes the given address.
  63     for CU in dwarfinfo.iter_CUs():
  64         # First, look at line programs to find the file/line for the address
  65         lineprog = dwarfinfo.line_program_for_CU(CU)
  66         prevstate = None
  67         for entry in lineprog.get_entries():
  68             # We're interested in those entries where a new state is assigned
  69             if entry.state is None or entry.state.end_sequence:
  70                 continue
  71             # Looking for a range of addresses in two consecutive states that
  72             # contain the required address.
  73             if prevstate and prevstate.address <= address < entry.state.address:
  74                 filename = lineprog['file_entry'][prevstate.file - 1].name
  75                 line = prevstate.line
  76                 return filename, line
  77             prevstate = entry.state
  78     return None, None
  79
  80
  81 if __name__ == '__main__':
  82     if sys.argv[1] == '--test':
  83         process_file(sys.argv[2], 0x400503)
  84         sys.exit(0)
  85
  86     if len(sys.argv) < 3:
  87         print('Expected usage: {0} <address> <executable>'.format(sys.argv[0]))
  88         sys.exit(1)
  89     addr = int(sys.argv[1], 0)
  90     process_file(sys.argv[2], addr)