Enable vdiff test on validator bots

SConstruct

 #! -*- python -*-
 # Copyright (c) 2012 The Native Client 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 atexit
 import os
 import platform
 import subprocess
 import sys
@@ skipping 318 matching lines @@
     desc='Use the zero-address-based x86-64 sandbox model instead of '
       'the r15-based model.')
 
   #########################################################################
   # EXPERIMENTAL
   # This is for generating a testing library for use within private test
   # enuminsts, where we want to compare and test different validators.
   #
   BitFromArgument(env, 'ncval_testing', default=False,
     desc='EXPERIMENTAL: Compile validator code for testing within enuminsts')
+  # Diff test comparing old validator with new R-DFA validator
+  BitFromArgument(env, 'validator_diff_tests', default=False,

[bradn, 2012/08/28 21:14:10]

So I think you should be able to drop this.
This is adding a boolean argument like:
validator_diff_tests=0 etc.
Whereas your actual target is just a normal test target (which should work fine
as you've added it to a new test suite).

[Brad Chen, 2012/08/28 22:09:42]

Done.

+    desc='Compare validator implementations')
 
   # PNaCl sanity checks
   if ((env.Bit('pnacl_generate_pexe') or env.Bit('use_sandboxed_translator'))
       and not env.Bit('bitcode')):
     raise ValueError("pnacl_generate_pexe and use_sandboxed_translator"
                         "don't make sense without bitcode")
 
 
 def CheckArguments():
   for key in ARGUMENTS:
@@ skipping 270 matching lines @@
   'pepper_browser_tests',
   # Lightweight browser tests
   'chrome_browser_tests',
   # Tests written using Chrome's PyAuto test jig
   'pyauto_tests',
   'huge_tests',
   'memcheck_bot_tests',
   'tsan_bot_tests',
   # Special testing environment for testing comparing x86 validators.
   'ncval_testing',
+  # Environment for validator difference testing
+  'validator_diff_tests',
 ])
 
 # These are the test suites we know exist, but aren't run on a regular basis.
 # These test suites are essentially shortcuts that run a specific subset of the
 # test cases.
 ACCEPTABLE_TEST_SUITES = set([
   'barebones_tests',
   'dynamic_load_tests',
   'exception_tests',
   'exit_status_tests',
@@ skipping 889 matching lines @@
 
 pre_base_env.AddMethod(CommandSelLdrTestNacl)
 
 
 TEST_EXTRA_ARGS = ['stdin', 'log_file',
                    'stdout_golden', 'stderr_golden', 'log_golden',
                    'filter_regex', 'filter_inverse', 'filter_group_only',
                    'osenv', 'arch', 'subarch', 'exit_status', 'track_cmdtime',
                    'num_runs', 'process_output_single',
                    'process_output_combined', 'using_nacl_signal_handler',
-                   'declares_exit_status']
+                   'declares_exit_status', 'time_warning', 'time_error']
 
 TEST_TIME_THRESHOLD = {
     'small':   2,
     'medium': 10,
     'large':  60,
     'huge': 1800,
     }
 
 # Valgrind handles SIGSEGV in a way our testing tools do not expect.
 UNSUPPORTED_VALGRIND_EXIT_STATUS = ['sigabrt',
@@ skipping 1788 matching lines @@
   nacl_env.ValidateSdk()
 
 if BROKEN_TEST_COUNT > 0:
   msg = "There are %d broken tests." % BROKEN_TEST_COUNT
   if GetOption('brief_comstr'):
     msg += " Add --verbose to the command line for more information."
   print msg
 
 # separate warnings from actual build output
 Banner('B U I L D - O U T P U T:')