--- a/testing/mozharness/mozharness/mozilla/testing/talos.py
+++ b/testing/mozharness/mozharness/mozilla/testing/talos.py
@@ -165,17 +165,22 @@ class Talos(TestingMixin, MercurialScrip
     # We accept some configuration options from the try commit message in the format mozharness: <options>
     # Example try commit message:
     #   mozharness: --spsProfile try: <stuff>
     def query_sps_profile_options(self):
         sps_results = []
         if self.buildbot_config:
             # this is inside automation
             # now let's see if we added spsProfile specs in the commit message
-            junk, junk, opts = self.buildbot_config['sourcestamp']['changes'][-1]['comments'].partition('mozharness:')
+            try:
+                junk, junk, opts = self.buildbot_config['sourcestamp']['changes'][-1]['comments'].partition('mozharness:')
+            except IndexError:
+                # when we don't have comments on changes (bug 1255187)
+                opts = None
+
             if opts:
               opts = re.sub(r'\w+:.*', '', opts).strip().split(' ')
               if "--spsProfile" in opts:
                   # overwrite whatever was set here.
                   self.sps_profile = True
               try:
                     idx = opts.index('--spsProfileInterval')
                     if len(opts) > idx + 1:
@@ -232,16 +237,18 @@ class Talos(TestingMixin, MercurialScrip
         if binary_path.endswith('.exe'):
             binary_path = binary_path[:-4]
         # options overwritten from **kw
         kw_options = {'executablePath': binary_path}
         if 'suite' in self.config:
             kw_options['suite'] = self.config['suite']
         if self.config.get('title'):
             kw_options['title'] = self.config['title']
+            if kw_options['title'].startswith('tst-linux64-spot'):
+                kw_options['framework'] = 'talos-aws'
         if self.config.get('branch'):
             kw_options['branchName'] = self.config['branch']
         if self.symbols_path:
             kw_options['symbolsPath'] = self.symbols_path
         kw_options.update(kw)
         # talos expects tests to be in the format (e.g.) 'ts:tp5:tsvg'
         tests = kw_options.get('activeTests')
         if tests and not isinstance(tests, basestring):

--- a/testing/talos/talos/cmdline.py
+++ b/testing/talos/talos/cmdline.py
@@ -129,16 +129,20 @@ def create_parser(mach_interface=False):
     add_arg('--sourcestamp',
             help='Specify the hg revision or sourcestamp for the changeset'
                  ' we are testing.  This will use the value found in'
                  ' application.ini if it is not specified.')
     add_arg('--repository',
             help='Specify the url for the repository we are testing. '
                  'This will use the value found in application.ini if'
                  ' it is not specified.')
+    add_arg('--framework',
+            help='Will post to the specified framework for Perfherder. '
+                 'Default "talos".  Used primarily for experiments on '
+                 'new platforms')
     add_arg('--print-tests', action=_ListTests,
             help="print available tests")
     add_arg('--print-suites', action=_ListSuite,
             help="list available suites")
 
     add_logging_group(parser)
     return parser
 

--- a/testing/talos/talos/config.py
+++ b/testing/talos/talos/config.py
@@ -409,16 +409,17 @@ def get_browser_config(config):
     required = ('preferences', 'extensions', 'browser_path', 'browser_wait',
                 'extra_args', 'buildid', 'env', 'init_url', 'webserver')
     optional = {'bcontroller_config': '${talos}/bcontroller.json',
                 'branch_name': '',
                 'child_process': 'plugin-container',
                 'develop': False,
                 'e10s': False,
                 'process': '',
+                'framework': 'talos',
                 'repository': None,
                 'sourcestamp': None,
                 'symbols_path': None,
                 'test_name_extension': '',
                 'test_timeout': 1200,
                 'xperf_path': None,
                 'error_filename': None,
                 }

--- a/testing/talos/talos/output.py
+++ b/testing/talos/talos/output.py
@@ -156,17 +156,17 @@ class PerfherderOutput(Output):
             return filter.geometric_mean([i for i, j in vals])
         else:
             return filter.mean([i for i, j in vals])
 
     def __call__(self):
         suites = []
         test_results = {
             'framework': {
-                'name': 'talos',
+                'name': self.results.results[0].framework,
             },
             'suites': suites,
         }
 
         for test in self.results.results:
             # serialize test results
             tsresult = None
             if not test.using_xperf:

--- a/testing/talos/talos/results.py
+++ b/testing/talos/talos/results.py
@@ -79,23 +79,23 @@ class TalosResults(object):
 
         if tbpl_output:
             print "TinderboxPrint: TalosResult: %s" % json.dumps(tbpl_output)
 
 
 class TestResults(object):
     """container object for all test results across cycles"""
 
-    def __init__(self, test_config, global_counters=None, extensions=None):
+    def __init__(self, test_config, global_counters=None, framework=None):
         self.results = []
         self.test_config = test_config
         self.format = None
         self.global_counters = global_counters or {}
         self.all_counter_results = []
-        self.extensions = extensions
+        self.framework = framework
         self.using_xperf = False
 
     def name(self):
         return self.test_config['name']
 
     def extension(self):
         return self.test_config['test_name_extension']
 

--- a/testing/talos/talos/ttest.py
+++ b/testing/talos/talos/ttest.py
@@ -96,17 +96,18 @@ class TTest(object):
             setup.env['MOZ_INSTRUMENT_EVENT_LOOP'] = '1'
             setup.env['MOZ_INSTRUMENT_EVENT_LOOP_THRESHOLD'] = '20'
             setup.env['MOZ_INSTRUMENT_EVENT_LOOP_INTERVAL'] = '10'
             global_counters['responsiveness'] = []
 
         # instantiate an object to hold test results
         test_results = results.TestResults(
             test_config,
-            global_counters
+            global_counters,
+            browser_config.get('framework')
         )
 
         for i in range(test_config['cycles']):
             LOG.info("Running cycle %d/%d for %s test..."
                      % (i+1, test_config['cycles'], test_config['name']))
 
             # remove the browser  error file
             mozfile.remove(browser_config['error_filename'])