tools/cygprofile/symbol_extractor.py

   1 #!/usr/bin/python
   2 # Copyright 2015 The Chromium Authors. All rights reserved.
   3 # Use of this source code is governed by a BSD-style license that can be
   4 # found in the LICENSE file.
   5
   6 """Utilities to get and manipulate symbols from a binary."""
   7
   8 import collections
   9 import os
  10 import re
  11 import subprocess
  12 import sys
  13
  14 sys.path.insert(
  15     0, os.path.join(os.path.dirname(__file__), os.pardir, os.pardir,
  16                     'third_party', 'android_platform', 'development',
  17                     'scripts'))
  18 import symbol
  19
  20
  21 SymbolInfo = collections.namedtuple('SymbolInfo', ('name', 'offset', 'size',
  22                                                    'section'))
  23
  24 def SetArchitecture(arch):
  25   """Set the architecture for binaries to be symbolized."""
  26   symbol.ARCH = arch
  27
  28
  29 def _FromObjdumpLine(line):
  30   """Create a SymbolInfo by parsing a properly formatted objdump output line.
  31
  32   Args:
  33     line: line from objdump
  34
  35   Returns:
  36     An instance of SymbolInfo if the line represents a symbol, None otherwise.
  37   """
  38   # All of the symbol lines we care about are in the form
  39   # 0000000000  g    F   .text.foo     000000000 [.hidden] foo
  40   # where g (global) might also be l (local) or w (weak).
  41   parts = line.split()
  42   if len(parts) < 6 or parts[2] != 'F':
  43     return None
  44
  45   assert len(parts) == 6 or (len(parts) == 7 and parts[5] == '.hidden')
  46   accepted_scopes = set(['g', 'l', 'w'])
  47   assert parts[1] in accepted_scopes
  48
  49   offset = int(parts[0], 16)
  50   section = parts[3]
  51   size = int(parts[4], 16)
  52   name = parts[-1].rstrip('\n')
  53   assert re.match('^[a-zA-Z0-9_.]+$', name)
  54   return SymbolInfo(name=name, offset=offset, section=section, size=size)
  55
  56
  57 def _SymbolInfosFromStream(objdump_lines):
  58   """Parses the output of objdump, and get all the symbols from a binary.
  59
  60   Args:
  61     objdump_lines: An iterable of lines
  62
  63   Returns:
  64     A list of SymbolInfo.
  65   """
  66   symbol_infos = []
  67   for line in objdump_lines:
  68     symbol_info = _FromObjdumpLine(line)
  69     if symbol_info is not None:
  70       symbol_infos.append(symbol_info)
  71   return symbol_infos
  72
  73
  74 def SymbolInfosFromBinary(binary_filename):
  75   """Runs objdump to get all the symbols from a binary.
  76
  77   Args:
  78     binary_filename: path to the binary.
  79
  80   Returns:
  81     A list of SymbolInfo from the binary.
  82   """
  83   command = (symbol.ToolPath('objdump'), '-t', '-w', binary_filename)
  84   p = subprocess.Popen(command, shell=False, stdout=subprocess.PIPE)
  85   try:
  86     result = _SymbolInfosFromStream(p.stdout)
  87     return result
  88   finally:
  89     p.wait()
  90
  91
  92 def GroupSymbolInfosByOffset(symbol_infos):
  93   """Create a dict {offset: [symbol_info1, ...], ...}.
  94
  95   As several symbols can be at the same offset, this is a 1-to-many
  96   relationship.
  97
  98   Args:
  99     symbol_infos: iterable of SymbolInfo instances
 100
 101   Returns:
 102     a dict {offset: [symbol_info1, ...], ...}
 103   """
 104   offset_to_symbol_infos = collections.defaultdict(list)
 105   for symbol_info in symbol_infos:
 106     offset_to_symbol_infos[symbol_info.offset].append(symbol_info)
 107   return dict(offset_to_symbol_infos)
 108
 109
 110 def CreateNameToSymbolInfo(symbol_infos):
 111   """Create a dict {name: symbol_info, ...}.
 112
 113   Args:
 114     symbol_infos: iterable of SymbolInfo instances
 115
 116   Returns:
 117     a dict {name: symbol_info, ...}
 118   """
 119   return {symbol_info.name: symbol_info for symbol_info in symbol_infos}