Loosen the parsing of CDB output in windows; it can start with at least 'ChildEBP' or 'Child-SP', a…

tools/telemetry/telemetry/core/backends/chrome/desktop_browser_backend.py

Index: tools/telemetry/telemetry/core/backends/chrome/desktop_browser_backend.py
diff --git a/tools/telemetry/telemetry/core/backends/chrome/desktop_browser_backend.py b/tools/telemetry/telemetry/core/backends/chrome/desktop_browser_backend.py
index 4982261b16f051bc8a6e316f59e982df56fbaa06..154ae102a07c7b78b764e648cafc610e7a6d5921 100644
--- a/tools/telemetry/telemetry/core/backends/chrome/desktop_browser_backend.py
+++ b/tools/telemetry/telemetry/core/backends/chrome/desktop_browser_backend.py
@@ -8,6 +8,7 @@ import heapq
 import logging
 import os
 import os.path
+import re
 import shutil
 import subprocess as subprocess
 import sys
@@ -264,7 +265,13 @@ class DesktopBrowserBackend(chrome_browser_backend.ChromeBrowserBackend):
         return None
       output = subprocess.check_output([cdb, '-y', self._browser_directory,
                                         '-c', '.ecxr;k30;q', '-z', minidump])
-      stack_start = output.find('ChildEBP')
+      # cdb output can start the stack with "ChildEBP", "Child-SP", and possibly
+      # other things we haven't seen yet. If we can't find the start of the
+      # stack, include output from the beginning.
+      stack_start = 0
+      stack_start_match = re.search("^Child(?:EBP|-SP)", output, re.MULTILINE)
+      if stack_start_match:
+        stack_start = stack_start_match.start()
       stack_end = output.find('quit:')
       return output[stack_start:stack_end]