Only include unique lines when patching orderfile.

tools/cygprofile/patch_orderfile.py

 #!/usr/bin/python
 # Copyright 2013 The Chromium Authors. All rights reserved.
 # Use of this source code is governed by a BSD-style license that can be
 # found in the LICENSE file.
 
 """Patch an orderfile.
 
 Starting with a list of symbols in a binary and an orderfile (ordered list of
 symbols), matches the symbols in the orderfile and augments each symbol with the
 symbols residing at the same address (due to having identical code).
@@ skipping 84 matching lines @@
   """
   command = 'nm -S -n %s | egrep "( t )|( W )|( T )"' % binary_filename
   p = subprocess.Popen(command, shell=True, stdout=subprocess.PIPE)
   try:
     result = _GetSymbolInfosFromStream(p.stdout)
     return result
   finally:
     p.wait()
 
 
+def _StripPrefixes(line):

[pasko, 2015/01/22 17:37:14]

it strips only one prefix, so maybe name it _StripPrefix?

[azarchs, 2015/01/22 17:44:30]

Done.

+  """Get the symbol from a line with a linker section name

[pasko, 2015/01/22 17:37:14]

nit: full stop at the end of the sentence. (here and below)

[azarchs, 2015/01/22 17:44:30]

Done.

+
+  Args:
+    line: a line from an orderfile, usually in the form
+    .text.SymbolName

[pasko, 2015/01/22 17:37:14]

nit: 4 chars shift to the right

[azarchs, 2015/01/22 17:44:30]

Done.

+
+  Returns:
+    The symbol, SymbolName in the example above.
+  """
+  line = line.rstrip('\n')
+  for prefix in _PREFIXES:
+    if line.startswith(prefix):
+      return line[len(prefix):]
+  return line  # Unprefixed case

[Benoit L, 2015/01/22 16:49:53]

nit: You can also use the

for prefix in _PREFIXES:
  ...
else:
  return line

idiom to remove the comment. Up to your personal preference.

[azarchs, 2015/01/22 17:10:12]

Maybe it's because I'm used to C, C++, Java, C#, Basic, and Visual Basic, but
that idiom just looks weird to me.

+
+
 def _GetSymbolsFromStream(lines):
   """Get the symbols from an iterable of lines.
 
   Args:
     lines: iterable of lines from an orderfile.
 
   Returns:
     Same as GetSymbolsFromOrderfile
   """
   # TODO(lizeb): Retain the prefixes later in the processing stages.
   symbols = []
+  unique_symbols = set()
   for line in lines:
-    line = line.rstrip('\n')
-    for prefix in _PREFIXES:
-      if line.startswith(prefix):
-        line = line[len(prefix):]
-        break
+    line = _StripPrefixes(line)
     name = _RemoveClone(line)
-    if name == '':
+    if name == '' or name == '*' or name == '.text':

[pasko, 2015/01/22 17:37:14]

please enhance testGetSymbolsFromStream to verify that lines ".text" and 
".text.*" are filtered out from the input.

Also would be nice to mention that a prefixed start expansion would be filtered
out.

[azarchs, 2015/01/22 17:44:30]

Done.

       continue
-    symbols.append(line)
+    if not line in unique_symbols:
+      symbols.append(line)
+      unique_symbols.add(line)
   return symbols
 
 
 def _GetSymbolsFromOrderfile(filename):
   """Return the symbols from an orderfile.
 
   Args:
     filename: The name of the orderfile.
 
   Returns:
@@ skipping 48 matching lines @@
   for symbol in symbol_infos:
     if symbol.offset not in seen_offsets:
       seen_offsets.add(symbol.offset)
       matching_symbols += [
           s.name for s in offset_to_symbol_infos[symbol.offset]]
   return matching_symbols
 
 
 def _PrintSymbolsWithPrefixes(symbol_names, output_file):
   """For each symbol, outputs it to output_file with the prefixes."""
+  unique_outputs = set()
   for name in symbol_names:
-    output_file.write('\n'.join(prefix + name for prefix in _PREFIXES) + '\n')
+    for prefix in _PREFIXES:
+      linker_section = prefix + name
+      if not linker_section in unique_outputs:
+        output_file.write(linker_section + '\n')
+        unique_outputs.add(linker_section)
 
 
 def main(argv):
   if len(argv) != 3:
     print 'Usage: %s <unpatched_orderfile> <libchrome.so>' % argv[0]
     return 1
   orderfile_filename = argv[1]
   binary_filename = argv[2]
   (offset_to_symbol_infos, name_to_symbol_infos) = _GetSymbolInfosFromBinary(
       binary_filename)
   profiled_symbols = _GetSymbolsFromOrderfile(orderfile_filename)
   matched_symbols = _MatchProfiledSymbols(
       profiled_symbols, name_to_symbol_infos)
   symbols_by_offset = _ExpandSymbolsWithDupsFromSameOffset(
       matched_symbols, offset_to_symbol_infos)
   _PrintSymbolsWithPrefixes(symbols_by_offset, sys.stdout)
   # The following is needed otherwise Gold only applies a partial sort.
   print '.text'    # gets methods not in a section, such as assembly
   print '.text.*'  # gets everything else
   return 0
 
 
 if __name__ == '__main__':
   logging.basicConfig(level=logging.INFO)
   sys.exit(main(sys.argv))