A tool for mapping runtime addresses to symbol names.

tools/find_runtime_symbols/prepare_symbol_info.py

Index: tools/find_runtime_symbols/prepare_symbol_info.py
diff --git a/tools/find_runtime_symbols/prepare_symbol_info.py b/tools/find_runtime_symbols/prepare_symbol_info.py
new file mode 100755
index 0000000000000000000000000000000000000000..310568c129745c1a8e91a837fc0c029f322bed9e
--- /dev/null
+++ b/tools/find_runtime_symbols/prepare_symbol_info.py
@@ -0,0 +1,114 @@
+#!/usr/bin/env python
+
+import json
+import logging
+import os
+import re
+import shutil
+import subprocess
+import sys
+import tempfile
+
+from parse_proc_maps import parse_proc_maps
+
+
+def _executable_condition(entry):

[M-A Ruel, 2012/07/19 15:21:15]

Why duplicate this function?

[Dai Mikurube (NOT FULLTIME), 2012/07/20 04:32:39]

Moved it to util.py, and shared it.

+  return (entry.executable == 'x' and re.match(
+      '\S+(\.(so|dll|dylib|bundle)|chrome)((\.\d+)+\w*(\.\d+){0,3})?',
+      entry.name))
+
+
+def prepare_symbol_info(maps_path, output_dir_path=None, loglevel=logging.WARN):
+  log = logging.getLogger('prepare_symbol_info')
+  log.setLevel(loglevel)
+  handler = logging.StreamHandler()
+  handler.setLevel(loglevel)
+  formatter = logging.Formatter('%(message)s')
+  handler.setFormatter(formatter)
+  log.addHandler(handler)
+
+  if not output_dir_path:
+    matched = re.match('^(.*)\.maps$', os.path.basename(maps_path))
+    if matched:
+      output_dir_path = matched.group(1) + '.pre'
+  if not output_dir_path:
+    matched = re.match('^/proc/(.*)/maps$', os.path.realpath(maps_path))

[M-A Ruel, 2012/07/19 15:21:15]

You should make this clear that these files are linux (?) specific, have the
main refuse to run on other OSes.

[Dai Mikurube (NOT FULLTIME), 2012/07/20 04:32:39]

Done.

+    if matched:
+      output_dir_path = matched.group(1) + '.pre'
+  if not output_dir_path:
+    output_dir_prefix = os.path.basename(maps_path) + '.pre'
+  # TODO(dmikurube): Find another candidate for output_dir_path.
+
+  log.info('Data for profiling will be collected in "%s".' % output_dir_path)
+  output_dir_path_exists = False
+  if os.path.exists(output_dir_path):
+    if os.path.isdir(output_dir_path) and not os.listdir(output_dir_path):
+      log.warn('Using an empty directory existing at "%s".' % output_dir_path)
+    else:
+      log.warn('A file or a directory exists at "%s".' % output_dir_path)
+      output_dir_path_exists = True
+  else:
+    log.info('Creating a new directory at "%s".' % output_dir_path)
+    os.mkdir(output_dir_path)
+
+  if output_dir_path_exists:
+    return 1
+
+  shutil.copyfile(maps_path, os.path.join(output_dir_path, 'maps'))
+
+  with open(maps_path, mode='r') as FILE:
+    maps = parse_proc_maps(FILE)
+
+  log.debug('Listing up symbols.')
+  nm_files = {}
+  for entry in maps.iter(_executable_condition):
+    log.debug('  %016x-%016x +%06x %s' % (
+        entry.begin, entry.end, entry.offset, entry.name))
+    with tempfile.NamedTemporaryFile(
+        prefix=os.path.basename(entry.name) + '.',
+        suffix='.nm', delete=False, mode='w+', dir=output_dir_path) as FILE:
+      nm_filename = os.path.realpath(FILE.name)
+      nm_succeeded = False
+      cppfilt_succeeded = False
+      p_nm = subprocess.Popen(
+          'nm -n --format bsd %s' % entry.name, shell=True,
+          stdout=subprocess.PIPE, stderr=subprocess.PIPE)
+      p_cppfilt = subprocess.Popen(
+          'c++filt', shell=True,
+          stdin=p_nm.stdout, stdout=FILE, stderr=subprocess.PIPE)
+      if p_nm.wait() == 0:

[M-A Ruel, 2012/07/19 15:21:15]

See http://docs.python.org/library/subprocess.html#subprocess.Popen.wait

[Dai Mikurube (NOT FULLTIME), 2012/07/20 04:32:39]

The problem is that it's giving p_nm.stdout to c++filt's stdin. 
Popen.communicate() intercepts a part of p_nm.stdout, and c++filt fails with
p_nm.communicate().

I'd like to avoid reading contents from p_nm.stdout.  p_nm.stdout will be very
big...  Do you have any ideas for such a case?
(fyi, I can drop stderr.  It's just providing error messages for users.)

+        nm_succeeded = True
+      for line in p_nm.stderr:
+        log.debug(line.rstrip())
+      if p_cppfilt.wait() == 0:
+        cppfilt_succeeded = True
+      for line in p_cppfilt.stderr:
+        log.debug(line.rstrip())
+
+    if nm_succeeded and cppfilt_succeeded:
+      nm_files[entry.name] = {
+        'file': os.path.basename(os.path.basename(nm_filename)),

[M-A Ruel, 2012/07/19 15:21:15]

Did you mean dirname?

[Dai Mikurube (NOT FULLTIME), 2012/07/20 04:32:39]

Ah, one of basename is not required.

+        'format': 'bsd',
+        'mangled': False,
+        }

[M-A Ruel, 2012/07/19 15:21:15]

funny alignment

[Dai Mikurube (NOT FULLTIME), 2012/07/20 04:32:39]

Done.

+    else:
+      os.remove(nm_filename)
+
+  with open(os.path.join(output_dir_path, 'nm.json'), 'w+') as FILE:
+    json.dump(nm_files, FILE, indent=2, sort_keys=True)

[M-A Ruel, 2012/07/19 15:21:15]

json.dump in a w+ opened file? It's going to be invalid.

[Dai Mikurube (NOT FULLTIME), 2012/07/20 04:32:39]

Done.

+
+  log.info('Collected symbol information at "%s".' % output_dir_path)
+  return 0
+
+
+if __name__ == '__main__':
+  if len(sys.argv) < 2:
+    sys.stderr.write("""Usage:
+%s /path/to/maps [/path/to/output_data_dir/]
+""" % sys.argv[0])
+    sys.exit(1)
+  elif len(sys.argv) == 2:
+    sys.exit(prepare_symbol_info(sys.argv[1], loglevel=logging.INFO))
+  else:
+    sys.exit(prepare_symbol_info(sys.argv[1], sys.argv[2],
+                                 loglevel=logging.INFO))