binary_size_tool: fix for ambiguous addr2line output

build/android/pylib/symbols/elf_symbolizer.py

 # Copyright 2014 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.
 
 import collections
 import datetime
 import logging
 import multiprocessing
 import os
 import posixpath
@@ skipping 57 matching lines @@
   |max_queue_size| bound, a new addr2line instance is kicked in.
   In the case of a very eager producer (i.e. all |max_concurrent_jobs| instances
   have a backlog of |max_queue_size|), back-pressure is applied on the caller by
   blocking the SymbolizeAsync method.
 
   This module has been deliberately designed to be dependency free (w.r.t. of
   other modules in this project), to allow easy reuse in external projects.
   """
 
   def __init__(self, elf_file_path, addr2line_path, callback, inlines=False,
-      max_concurrent_jobs=None, addr2line_timeout=30, max_queue_size=50):
+      max_concurrent_jobs=None, addr2line_timeout=30, max_queue_size=50,
+      source_root_path=None, strip_base_path=None):
     """Args:
       elf_file_path: path of the elf file to be symbolized.
       addr2line_path: path of the toolchain's addr2line binary.
       callback: a callback which will be invoked for each resolved symbol with
           the two args (sym_info, callback_arg). The former is an instance of
           |ELFSymbolInfo| and contains the symbol information. The latter is an
           embedder-provided argument which is passed to SymbolizeAsync().
       inlines: when True, the ELFSymbolInfo will contain also the details about
           the outer inlining functions. When False, only the innermost function
           will be provided.
       max_concurrent_jobs: Max number of addr2line instances spawned.
           Parallelize responsibly, addr2line is a memory and I/O monster.
       max_queue_size: Max number of outstanding requests per addr2line instance.
       addr2line_timeout: Max time (in seconds) to wait for a addr2line response.
           After the timeout, the instance will be considered hung and respawned.
+      source_root_path: The path to the library source code. Which is used in

[Primiano Tucci (use gerrit), 2014/06/25 09:02:58]

Nit: s/. Which//

+          the disambiguation process. Disambiguation means to resolve ambiguous
+          source paths, for example turn addr2line output "unicode.cc" into a
+          full and absolute path. In some toolchains only the name of the source

[Primiano Tucci (use gerrit), 2014/06/25 09:02:59]

Uh, what is the difference between a "full" and a "absolute" path?
Probably you want just one of the two?

+          file is output, without any path information; disambiguation searches
+          through the source directory specified by |source_root_path| argument
+          for files whose name matches. If there are multiple files with the
+          same name, disambiguation will fail.

[Primiano Tucci (use gerrit), 2014/06/25 09:02:58]

Can we reword this a bit and make it a bit more clear and concise? 
What about:

"path to the source tree on the local machine. It has two purposes: 1. turn
relative file names names into full paths (foo.c -> /source/root/path/foo.c) iff
foo.c is not ambiguous (i.e. unique in the |source_root_path|) to workaround
some toolchains limitations; 2. serves as the replacement base for
|strip_base_path| (below)"

+      strip_base_path: If this base path is found in the beginning of any

[Primiano Tucci (use gerrit), 2014/06/25 09:02:58]

"rebases the symbols source paths onto |source_root_path| (i.e.
s/^strip_base_path/source_root_path/)."

+          result, we strip it and replace it with |source_root_path|.
     """
     assert(os.path.isfile(addr2line_path)), 'Cannot find ' + addr2line_path
     self.elf_file_path = elf_file_path
     self.addr2line_path = addr2line_path
     self.callback = callback
     self.inlines = inlines
     self.max_concurrent_jobs = (max_concurrent_jobs or
                                 min(multiprocessing.cpu_count(), 4))
     self.max_queue_size = max_queue_size
     self.addr2line_timeout = addr2line_timeout
     self.requests_counter = 0  # For generating monotonic request IDs.
     self._a2l_instances = []  # Up to |max_concurrent_jobs| _Addr2Line inst.
 
+    # If necessary, create disambiguation lookup table
+    self.disambiguate = source_root_path is not None
+    self.lookup_table = {}

[Primiano Tucci (use gerrit), 2014/06/25 09:02:59]

What about:
self.disambiguation_table = {}  # 'foo.c' -> '/abs/path/foo.c'

"lookup_table" has the same value of saying:
self.array = []
self.number = 3 :) not really explicative.
Variables names should reflect the role they play, not the data structure used
to implement them :)

+    self.source_root_path = source_root_path
+    self.strip_base_path = strip_base_path
+    if(self.disambiguate):
+      self._CreateDisambiguationTable()
+
     # Create one addr2line instance. More instances will be created on demand
     # (up to |max_concurrent_jobs|) depending on the rate of the requests.
     self._CreateNewA2LInstance()
 
   def SymbolizeAsync(self, addr, callback_arg=None):
     """Requests symbolization of a given address.
 
     This method is not guaranteed to return immediately. It generally does, but
     in some scenarios (e.g. all addr2line instances have full queues) it can
     block to create back-pressure.
@@ skipping 37 matching lines @@
     for a2l in self._a2l_instances:
       a2l.WaitForIdle()
       a2l.Terminate()
 
   def _CreateNewA2LInstance(self):
     assert(len(self._a2l_instances) < self.max_concurrent_jobs)
     a2l = ELFSymbolizer.Addr2Line(self)
     self._a2l_instances.append(a2l)
     return a2l
 
+  def _CreateDisambiguationTable(self):

[Primiano Tucci (use gerrit), 2014/06/25 09:02:59]

Just add a one line comment please:
"""Non-unique file names will result in None entries"""

+    self.lookup_table = {}
+    source_root_path = os.path.abspath(self.source_root_path)

[Primiano Tucci (use gerrit), 2014/06/25 09:02:59]

I'd do this on line 121 and save one line:
self.source_root_path = os.path.abspath(source_root_path)

+
+    for root, _, filenames in os.walk(source_root_path):
+      for f in filenames:
+        self.lookup_table[f] = os.path.join(root, f) if (f not in
+                                self.lookup_table) else None
+
 
   class Addr2Line(object):
     """A python wrapper around an addr2line instance.
 
     The communication with the addr2line process looks as follows:
       [STDIN]         [STDOUT]  (from addr2line's viewpoint)
     > f001111
     > f002222
                     < Symbol::Name(foo, bar) for f001111
                     < /path/to/source/file.c:line_number
@@ skipping 131 matching lines @@
         source_line = None
         m = ELFSymbolizer.Addr2Line.SYM_ADDR_RE.match(line2)
         if m:
           if not m.group(1).startswith('?'):
             source_path = m.group(1)
             if not m.group(2).startswith('?'):
               source_line = int(m.group(2))
         else:
           logging.warning('Got invalid symbol path from addr2line: %s' % line2)
 
-        sym_info = ELFSymbolInfo(name, source_path, source_line)
+        # In case disambiguation is on, and needed
+        was_ambiguous = False
+        disambiguated = False
+        if self._symbolizer.disambiguate:
+          if source_path and not source_path.startswith('/'):

[Primiano Tucci (use gerrit), 2014/06/25 09:02:58]

Probably you can just use: posixpath.isabs(source_path) here

+            source_path = self._symbolizer.lookup_table.get(source_path)
+            was_ambiguous = True
+            disambiguated = source_path is not None

[Primiano Tucci (use gerrit), 2014/06/25 09:02:59]

Just a doubt: at this point if a2l here returns 'foo.c' and you don't have foo.c
in your disambiguation table, source_path becomes none.
In essence, looks like you're trying to enrich the output here, but if you
happen to fail, you end up reducing the quality of the output ('foo.c' -> None.)
Don't you probably want to do something like:

full_path = self._symbolizer.disambiguation_table.get(source_path)
if full_path:
  source_path = full_path
  disambiguated = True
?

+
+        if source_path and self._symbolizer.strip_base_path:
+          # Strip the base path
+          source_path = re.sub('^' + self._symbolizer.strip_base_path,
+                   self._symbolizer.source_root_path or '', source_path)

[Primiano Tucci (use gerrit), 2014/06/25 09:02:58]

Nit: indentation should be 4 spaces here, or inline. i.e.
either

source_path = re.sub('^' + self._symbolizer.strip_base_path,
    self._symbolizer.source_root_path or '', source_path)

or

source_path = re.sub('^' + self._symbolizer.strip_base_path,
                     self._symbolizer.source_root_path or '', 
                     source_path)

+
+        sym_info = ELFSymbolInfo(name, source_path, source_line, was_ambiguous,
+                                 disambiguated)
         if prev_sym_info:
           prev_sym_info.inlined_by = sym_info
         if not innermost_sym_info:
           innermost_sym_info = sym_info
 
       self._processed_symbols_count += 1
       self._symbolizer.callback(innermost_sym_info, callback_arg)
 
     def _RestartAddr2LineProcess(self):
       if self._proc:
@@ skipping 60 matching lines @@
 
     @property
     def first_request_id(self):
       """Returns the request_id of the oldest pending request in the queue."""
       return self._request_queue[0][2] if self._request_queue else 0
 
 
 class ELFSymbolInfo(object):
   """The result of the symbolization passed as first arg. of each callback."""
 
-  def __init__(self, name, source_path, source_line):
+  def __init__(self, name, source_path, source_line, was_ambiguous=False,
+               disambiguated=False):
     """All the fields here can be None (if addr2line replies with '??')."""
     self.name = name
-    self.source_path = source_path
+    self.source_path = (None if source_path is None else

[Primiano Tucci (use gerrit), 2014/06/25 09:02:58]

why do you need this logic (the abs path) here? Didn't you cover that above? If
the client didn't ask for disambiguation, he's fine to get relative paths.
Also, this is conceptually wrong: if you have 'foo.c' (suppose you didn't chose
disambiguation), abspath('foo.c') = PWD + 'foo.c', which is probably not what
you want.

+                       os.path.abspath(source_path))
     self.source_line = source_line
     # In the case of |inlines|=True, the |inlined_by| points to the outer
     # function inlining the current one (and so on, to form a chain).
     self.inlined_by = None
+    self.disambiguated = disambiguated
+    self.was_ambiguous = was_ambiguous
 
   def __str__(self):
     return '%s [%s:%d]' % (
         self.name or '??', self.source_path or '??', self.source_line or 0)