Adding upstream version 115.7.0esr.upstream/115.7.0esr

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 4102 insertions, 0 deletions

diff --git a/testing/mochitest/runtests.py b/testing/mochitest/runtests.py
new file mode 100644
index 0000000000..6605b4ce20
--- /dev/null
+++ b/testing/mochitest/runtests.py
@@ -0,0 +1,4102 @@
+# This Source Code Form is subject to the terms of the Mozilla Public
+# License, v. 2.0. If a copy of the MPL was not distributed with this
+# file, You can obtain one at http://mozilla.org/MPL/2.0/.
+
+"""
+Runs the Mochitest test harness.
+"""
+
+import os
+import sys
+
+SCRIPT_DIR = os.path.abspath(os.path.realpath(os.path.dirname(__file__)))
+sys.path.insert(0, SCRIPT_DIR)
+
+import ctypes
+import glob
+import json
+import numbers
+import platform
+import re
+import shlex
+import shutil
+import signal
+import socket
+import subprocess
+import sys
+import tempfile
+import time
+import traceback
+import uuid
+import zipfile
+from argparse import Namespace
+from collections import defaultdict
+from contextlib import closing
+from ctypes.util import find_library
+from datetime import datetime, timedelta
+from distutils import spawn
+
+import bisection
+import mozcrash
+import mozdebug
+import mozinfo
+import mozprocess
+import mozrunner
+from manifestparser import TestManifest
+from manifestparser.filters import (
+    chunk_by_dir,
+    chunk_by_runtime,
+    chunk_by_slice,
+    failures,
+    pathprefix,
+    subsuite,
+    tags,
+)
+from manifestparser.util import normsep
+from mozgeckoprofiler import symbolicate_profile_json, view_gecko_profile
+from mozserve import DoHServer, Http3Server
+
+try:
+    from marionette_driver.addons import Addons
+    from marionette_driver.marionette import Marionette
+except ImportError as e:  # noqa
+    # Defer ImportError until attempt to use Marionette
+    def reraise(*args, **kwargs):
+        raise (e)  # noqa
+
+    Marionette = reraise
+
+import mozleak
+from leaks import LSANLeaks, ShutdownLeaks
+from mochitest_options import (
+    MochitestArgumentParser,
+    build_obj,
+    get_default_valgrind_suppression_files,
+)
+from mozlog import commandline, get_proxy_logger
+from mozprofile import Profile
+from mozprofile.cli import KeyValueParseError, parse_key_value, parse_preferences
+from mozprofile.permissions import ServerLocations
+from mozrunner.utils import get_stack_fixer_function, test_environment
+from mozscreenshot import dump_screen
+
+HAVE_PSUTIL = False
+try:
+    import psutil
+
+    HAVE_PSUTIL = True
+except ImportError:
+    pass
+
+import six
+from six.moves.urllib.parse import quote_plus as encodeURIComponent
+from six.moves.urllib_request import urlopen
+
+try:
+    from mozbuild.base import MozbuildObject
+
+    build = MozbuildObject.from_environment(cwd=SCRIPT_DIR)
+except ImportError:
+    build = None
+
+here = os.path.abspath(os.path.dirname(__file__))
+
+NO_TESTS_FOUND = """
+No tests were found for flavor '{}' and the following manifest filters:
+{}
+
+Make sure the test paths (if any) are spelt correctly and the corresponding
+--flavor and --subsuite are being used. See `mach mochitest --help` for a
+list of valid flavors.
+""".lstrip()
+
+
+########################################
+# Option for MOZ (former NSPR) logging #
+########################################
+
+# Set the desired log modules you want a log be produced
+# by a try run for, or leave blank to disable the feature.
+# This will be passed to MOZ_LOG environment variable.
+# Try run will then put a download link for a zip archive
+# of all the log files on treeherder.
+MOZ_LOG = ""
+
+########################################
+# Option for web server log            #
+########################################
+
+# If True, debug logging from the web server will be
+# written to mochitest-server-%d.txt artifacts on
+# treeherder.
+MOCHITEST_SERVER_LOGGING = False
+
+#####################
+# Test log handling #
+#####################
+
+# output processing
+TBPL_RETRY = 4  # Defined in mozharness
+
+
+class MessageLogger(object):
+
+    """File-like object for logging messages (structured logs)"""
+
+    BUFFERING_THRESHOLD = 100
+    # This is a delimiter used by the JS side to avoid logs interleaving
+    DELIMITER = "\ue175\uee31\u2c32\uacbf"
+    BUFFERED_ACTIONS = set(["test_status", "log"])
+    VALID_ACTIONS = set(
+        [
+            "suite_start",
+            "suite_end",
+            "test_start",
+            "test_end",
+            "test_status",
+            "log",
+            "assertion_count",
+            "buffering_on",
+            "buffering_off",
+        ]
+    )
+    # Regexes that will be replaced with an empty string if found in a test
+    # name. We do this to normalize test names which may contain URLs and test
+    # package prefixes.
+    TEST_PATH_PREFIXES = [
+        r"^/tests/",
+        r"^\w+://[\w\.]+(:\d+)?(/\w+)?/(tests?|a11y|chrome|browser)/",
+    ]
+
+    def __init__(self, logger, buffering=True, structured=True):
+        self.logger = logger
+        self.structured = structured
+        self.gecko_id = "GECKO"
+        self.is_test_running = False
+
+        # Even if buffering is enabled, we only want to buffer messages between
+        # TEST-START/TEST-END. So it is off to begin, but will be enabled after
+        # a TEST-START comes in.
+        self._buffering = False
+        self.restore_buffering = buffering
+
+        # Guard to ensure we never buffer if this value was initially `False`
+        self._buffering_initially_enabled = buffering
+
+        # Message buffering
+        self.buffered_messages = []
+
+    def validate(self, obj):
+        """Tests whether the given object is a valid structured message
+        (only does a superficial validation)"""
+        if not (
+            isinstance(obj, dict)
+            and "action" in obj
+            and obj["action"] in MessageLogger.VALID_ACTIONS
+        ):
+            raise ValueError
+
+    def _fix_subtest_name(self, message):
+        """Make sure subtest name is a string"""
+        if "subtest" in message and not isinstance(
+            message["subtest"], six.string_types
+        ):
+            message["subtest"] = str(message["subtest"])
+
+    def _fix_test_name(self, message):
+        """Normalize a logged test path to match the relative path from the sourcedir."""
+        if message.get("test") is not None:
+            test = message["test"]
+            for pattern in MessageLogger.TEST_PATH_PREFIXES:
+                test = re.sub(pattern, "", test)
+                if test != message["test"]:
+                    message["test"] = test
+                    break
+
+    def _fix_message_format(self, message):
+        if "message" in message:
+            if isinstance(message["message"], bytes):
+                message["message"] = message["message"].decode("utf-8", "replace")
+            elif not isinstance(message["message"], six.text_type):
+                message["message"] = six.text_type(message["message"])
+
+    def parse_line(self, line):
+        """Takes a given line of input (structured or not) and
+        returns a list of structured messages"""
+        if isinstance(line, six.binary_type):
+            # if line is a sequence of bytes, let's decode it
+            line = line.rstrip().decode("UTF-8", "replace")
+        else:
+            # line is in unicode - so let's use it as it is
+            line = line.rstrip()
+
+        messages = []
+        for fragment in line.split(MessageLogger.DELIMITER):
+            if not fragment:
+                continue
+            try:
+                message = json.loads(fragment)
+                self.validate(message)
+            except ValueError:
+                if self.structured:
+                    message = dict(
+                        action="process_output",
+                        process=self.gecko_id,
+                        data=fragment,
+                    )
+                else:
+                    message = dict(
+                        action="log",
+                        level="info",
+                        message=fragment,
+                    )
+
+            self._fix_subtest_name(message)
+            self._fix_test_name(message)
+            self._fix_message_format(message)
+            messages.append(message)
+
+        return messages
+
+    @property
+    def buffering(self):
+        if not self._buffering_initially_enabled:
+            return False
+        return self._buffering
+
+    @buffering.setter
+    def buffering(self, val):
+        self._buffering = val
+
+    def process_message(self, message):
+        """Processes a structured message. Takes into account buffering, errors, ..."""
+        # Activation/deactivating message buffering from the JS side
+        if message["action"] == "buffering_on":
+            if self.is_test_running:
+                self.buffering = True
+            return
+        if message["action"] == "buffering_off":
+            self.buffering = False
+            return
+
+        # Error detection also supports "raw" errors (in log messages) because some tests
+        # manually dump 'TEST-UNEXPECTED-FAIL'.
+        if "expected" in message or (
+            message["action"] == "log"
+            and message.get("message", "").startswith("TEST-UNEXPECTED")
+        ):
+            self.restore_buffering = self.restore_buffering or self.buffering
+            self.buffering = False
+            if self.buffered_messages:
+                snipped = len(self.buffered_messages) - self.BUFFERING_THRESHOLD
+                if snipped > 0:
+                    self.logger.info(
+                        "<snipped {0} output lines - "
+                        "if you need more context, please use "
+                        "SimpleTest.requestCompleteLog() in your test>".format(snipped)
+                    )
+                # Dumping previously buffered messages
+                self.dump_buffered(limit=True)
+
+            # Logging the error message
+            self.logger.log_raw(message)
+        # Determine if message should be buffered
+        elif (
+            self.buffering
+            and self.structured
+            and message["action"] in self.BUFFERED_ACTIONS
+        ):
+            self.buffered_messages.append(message)
+        # Otherwise log the message directly
+        else:
+            self.logger.log_raw(message)
+
+        # If a test ended, we clean the buffer
+        if message["action"] == "test_end":
+            self.is_test_running = False
+            self.buffered_messages = []
+            self.restore_buffering = self.restore_buffering or self.buffering
+            self.buffering = False
+
+        if message["action"] == "test_start":
+            self.is_test_running = True
+            if self.restore_buffering:
+                self.restore_buffering = False
+                self.buffering = True
+
+    def write(self, line):
+        messages = self.parse_line(line)
+        for message in messages:
+            self.process_message(message)
+        return messages
+
+    def flush(self):
+        sys.stdout.flush()
+
+    def dump_buffered(self, limit=False):
+        if limit:
+            dumped_messages = self.buffered_messages[-self.BUFFERING_THRESHOLD :]
+        else:
+            dumped_messages = self.buffered_messages
+
+        last_timestamp = None
+        for buf in dumped_messages:
+            # pylint --py3k W1619
+            timestamp = datetime.fromtimestamp(buf["time"] / 1000).strftime("%H:%M:%S")
+            if timestamp != last_timestamp:
+                self.logger.info("Buffered messages logged at {}".format(timestamp))
+            last_timestamp = timestamp
+
+            self.logger.log_raw(buf)
+        self.logger.info("Buffered messages finished")
+        # Cleaning the list of buffered messages
+        self.buffered_messages = []
+
+    def finish(self):
+        self.dump_buffered()
+        self.buffering = False
+        self.logger.suite_end()
+
+
+####################
+# PROCESS HANDLING #
+####################
+
+
+def call(*args, **kwargs):
+    """front-end function to mozprocess.ProcessHandler"""
+    # TODO: upstream -> mozprocess
+    # https://bugzilla.mozilla.org/show_bug.cgi?id=791383
+    log = get_proxy_logger("mochitest")
+
+    def on_output(line):
+        log.process_output(
+            process=process.pid,
+            data=line.decode("utf8", "replace"),
+            command=process.commandline,
+        )
+
+    process = mozprocess.ProcessHandlerMixin(
+        *args, processOutputLine=on_output, **kwargs
+    )
+    process.run()
+    return process.wait()
+
+
+def killPid(pid, log):
+    # see also https://bugzilla.mozilla.org/show_bug.cgi?id=911249#c58
+
+    if HAVE_PSUTIL:
+        # Kill a process tree (including grandchildren) with signal.SIGTERM
+        if pid == os.getpid():
+            raise RuntimeError("Error: trying to kill ourselves, not another process")
+        try:
+            parent = psutil.Process(pid)
+            children = parent.children(recursive=True)
+            children.append(parent)
+            for p in children:
+                p.send_signal(signal.SIGTERM)
+            gone, alive = psutil.wait_procs(children, timeout=30)
+            for p in gone:
+                log.info("psutil found pid %s dead" % p.pid)
+            for p in alive:
+                log.info("failed to kill pid %d after 30s" % p.pid)
+        except Exception as e:
+            log.info("Error: Failed to kill process %d: %s" % (pid, str(e)))
+    else:
+        try:
+            os.kill(pid, getattr(signal, "SIGKILL", signal.SIGTERM))
+        except Exception as e:
+            log.info("Failed to kill process %d: %s" % (pid, str(e)))
+
+
+if mozinfo.isWin:
+    import ctypes.wintypes
+
+    def isPidAlive(pid):
+        STILL_ACTIVE = 259
+        PROCESS_QUERY_LIMITED_INFORMATION = 0x1000
+        pHandle = ctypes.windll.kernel32.OpenProcess(
+            PROCESS_QUERY_LIMITED_INFORMATION, 0, pid
+        )
+        if not pHandle:
+            return False
+
+        try:
+            pExitCode = ctypes.wintypes.DWORD()
+            ctypes.windll.kernel32.GetExitCodeProcess(pHandle, ctypes.byref(pExitCode))
+
+            if pExitCode.value != STILL_ACTIVE:
+                return False
+
+            # We have a live process handle.  But Windows aggressively
+            # re-uses pids, so let's attempt to verify that this is
+            # actually Firefox.
+            namesize = 1024
+            pName = ctypes.create_string_buffer(namesize)
+            namelen = ctypes.windll.psapi.GetProcessImageFileNameA(
+                pHandle, pName, namesize
+            )
+            if namelen == 0:
+                # Still an active process, so conservatively assume it's Firefox.
+                return True
+
+            return pName.value.endswith((b"firefox.exe", b"plugin-container.exe"))
+        finally:
+            ctypes.windll.kernel32.CloseHandle(pHandle)
+
+
+else:
+    import errno
+
+    def isPidAlive(pid):
+        try:
+            # kill(pid, 0) checks for a valid PID without actually sending a signal
+            # The method throws OSError if the PID is invalid, which we catch
+            # below.
+            os.kill(pid, 0)
+
+            # Wait on it to see if it's a zombie. This can throw OSError.ECHILD if
+            # the process terminates before we get to this point.
+            wpid, wstatus = os.waitpid(pid, os.WNOHANG)
+            return wpid == 0
+        except OSError as err:
+            # Catch the errors we might expect from os.kill/os.waitpid,
+            # and re-raise any others
+            if err.errno in (errno.ESRCH, errno.ECHILD, errno.EPERM):
+                return False
+            raise
+
+
+# TODO: ^ upstream isPidAlive to mozprocess
+
+#######################
+# HTTP SERVER SUPPORT #
+#######################
+
+
+class MochitestServer(object):
+
+    "Web server used to serve Mochitests, for closer fidelity to the real web."
+
+    instance_count = 0
+
+    def __init__(self, options, logger):
+        if isinstance(options, Namespace):
+            options = vars(options)
+        self._log = logger
+        self._keep_open = bool(options["keep_open"])
+        self._utilityPath = options["utilityPath"]
+        self._xrePath = options["xrePath"]
+        self._profileDir = options["profilePath"]
+        self.webServer = options["webServer"]
+        self.httpPort = options["httpPort"]
+        if options.get("remoteWebServer") == "10.0.2.2":
+            # probably running an Android emulator and 10.0.2.2 will
+            # not be visible from host
+            shutdownServer = "127.0.0.1"
+        else:
+            shutdownServer = self.webServer
+        self.shutdownURL = "http://%(server)s:%(port)s/server/shutdown" % {
+            "server": shutdownServer,
+            "port": self.httpPort,
+        }
+        self.debugURL = "http://%(server)s:%(port)s/server/debug?2" % {
+            "server": shutdownServer,
+            "port": self.httpPort,
+        }
+        self.testPrefix = "undefined"
+
+        if options.get("httpdPath"):
+            self._httpdPath = options["httpdPath"]
+        else:
+            self._httpdPath = SCRIPT_DIR
+        self._httpdPath = os.path.abspath(self._httpdPath)
+
+        MochitestServer.instance_count += 1
+
+    def start(self):
+        "Run the Mochitest server, returning the process ID of the server."
+
+        # get testing environment
+        env = test_environment(xrePath=self._xrePath, log=self._log)
+        env["XPCOM_DEBUG_BREAK"] = "warn"
+        if "LD_LIBRARY_PATH" not in env or env["LD_LIBRARY_PATH"] is None:
+            env["LD_LIBRARY_PATH"] = self._xrePath
+        else:
+            env["LD_LIBRARY_PATH"] = ":".join([self._xrePath, env["LD_LIBRARY_PATH"]])
+
+        # When running with an ASan build, our xpcshell server will also be ASan-enabled,
+        # thus consuming too much resources when running together with the browser on
+        # the test machines. Try to limit the amount of resources by disabling certain
+        # features.
+        env["ASAN_OPTIONS"] = "quarantine_size=1:redzone=32:malloc_context_size=5"
+
+        # Likewise, when running with a TSan build, our xpcshell server will
+        # also be TSan-enabled. Except that in this case, we don't really
+        # care about races in xpcshell. So disable TSan for the server.
+        env["TSAN_OPTIONS"] = "report_bugs=0"
+
+        # Don't use socket process for the xpcshell server.
+        env["MOZ_DISABLE_SOCKET_PROCESS"] = "1"
+
+        if mozinfo.isWin:
+            env["PATH"] = env["PATH"] + ";" + str(self._xrePath)
+
+        args = [
+            "-g",
+            self._xrePath,
+            "-f",
+            os.path.join(self._httpdPath, "httpd.js"),
+            "-e",
+            "const _PROFILE_PATH = '%(profile)s'; const _SERVER_PORT = '%(port)s'; "
+            "const _SERVER_ADDR = '%(server)s'; const _TEST_PREFIX = %(testPrefix)s; "
+            "const _DISPLAY_RESULTS = %(displayResults)s;"
+            % {
+                "profile": self._profileDir.replace("\\", "\\\\"),
+                "port": self.httpPort,
+                "server": self.webServer,
+                "testPrefix": self.testPrefix,
+                "displayResults": str(self._keep_open).lower(),
+            },
+            "-f",
+            os.path.join(SCRIPT_DIR, "server.js"),
+        ]
+
+        xpcshell = os.path.join(
+            self._utilityPath, "xpcshell" + mozinfo.info["bin_suffix"]
+        )
+        command = [xpcshell] + args
+        server_logfile = None
+        if MOCHITEST_SERVER_LOGGING and "MOZ_UPLOAD_DIR" in os.environ:
+            server_logfile = os.path.join(
+                os.environ["MOZ_UPLOAD_DIR"],
+                "mochitest-server-%d.txt" % MochitestServer.instance_count,
+            )
+        self._process = mozprocess.ProcessHandler(
+            command, cwd=SCRIPT_DIR, env=env, logfile=server_logfile
+        )
+        self._process.run()
+        self._log.info("%s : launching %s" % (self.__class__.__name__, command))
+        pid = self._process.pid
+        self._log.info("runtests.py | Server pid: %d" % pid)
+        if MOCHITEST_SERVER_LOGGING and "MOZ_UPLOAD_DIR" in os.environ:
+            self._log.info("runtests.py enabling server debugging...")
+            i = 0
+            while i < 5:
+                try:
+                    with closing(urlopen(self.debugURL)) as c:
+                        self._log.info(six.ensure_text(c.read()))
+                    break
+                except Exception as e:
+                    self._log.info("exception when enabling debugging: %s" % str(e))
+                    time.sleep(1)
+                    i += 1
+
+    def ensureReady(self, timeout):
+        assert timeout >= 0
+
+        aliveFile = os.path.join(self._profileDir, "server_alive.txt")
+        i = 0
+        while i < timeout:
+            if os.path.exists(aliveFile):
+                break
+            time.sleep(0.05)
+            i += 0.05
+        else:
+            self._log.error(
+                "TEST-UNEXPECTED-FAIL | runtests.py | Timed out while waiting for server startup."
+            )
+            self.stop()
+            sys.exit(1)
+
+    def stop(self):
+        try:
+            with closing(urlopen(self.shutdownURL)) as c:
+                self._log.info(six.ensure_text(c.read()))
+        except Exception:
+            self._log.info("Failed to stop web server on %s" % self.shutdownURL)
+            traceback.print_exc()
+        finally:
+            if self._process is not None:
+                # Kill the server immediately to avoid logging intermittent
+                # shutdown crashes, sometimes observed on Windows 10.
+                self._process.kill()
+                self._log.info("Web server killed.")
+
+
+class WebSocketServer(object):
+
+    "Class which encapsulates the mod_pywebsocket server"
+
+    def __init__(self, options, scriptdir, logger, debuggerInfo=None):
+        self.port = options.webSocketPort
+        self.debuggerInfo = debuggerInfo
+        self._log = logger
+        self._scriptdir = scriptdir
+
+    def start(self):
+        # Invoke pywebsocket through a wrapper which adds special SIGINT handling.
+        #
+        # If we're in an interactive debugger, the wrapper causes the server to
+        # ignore SIGINT so the server doesn't capture a ctrl+c meant for the
+        # debugger.
+        #
+        # If we're not in an interactive debugger, the wrapper causes the server to
+        # die silently upon receiving a SIGINT.
+        scriptPath = "pywebsocket_wrapper.py"
+        script = os.path.join(self._scriptdir, scriptPath)
+
+        cmd = [sys.executable, script]
+        if self.debuggerInfo and self.debuggerInfo.interactive:
+            cmd += ["--interactive"]
+        cmd += [
+            "-H",
+            "127.0.0.1",
+            "-p",
+            str(self.port),
+            "-w",
+            self._scriptdir,
+            "-l",
+            os.path.join(self._scriptdir, "websock.log"),
+            "--log-level=debug",
+            "--allow-handlers-outside-root-dir",
+        ]
+        env = dict(os.environ)
+        env["PYTHONPATH"] = os.pathsep.join(sys.path)
+        # Start the process. Ignore stderr so that exceptions from the server
+        # are not treated as failures when parsing the test log.
+        self._process = mozprocess.ProcessHandler(
+            cmd,
+            cwd=SCRIPT_DIR,
+            env=env,
+            processStderrLine=lambda _: None,
+        )
+        self._process.run()
+        pid = self._process.pid
+        self._log.info("runtests.py | Websocket server pid: %d" % pid)
+
+    def stop(self):
+        if self._process is not None:
+            self._process.kill()
+
+
+class SSLTunnel:
+    def __init__(self, options, logger):
+        self.log = logger
+        self.process = None
+        self.utilityPath = options.utilityPath
+        self.xrePath = options.xrePath
+        self.certPath = options.certPath
+        self.sslPort = options.sslPort
+        self.httpPort = options.httpPort
+        self.webServer = options.webServer
+        self.webSocketPort = options.webSocketPort
+
+        self.customCertRE = re.compile("^cert=(?P<nickname>[0-9a-zA-Z_ ]+)")
+        self.clientAuthRE = re.compile("^clientauth=(?P<clientauth>[a-z]+)")
+        self.redirRE = re.compile("^redir=(?P<redirhost>[0-9a-zA-Z_ .]+)")
+
+    def writeLocation(self, config, loc):
+        for option in loc.options:
+            match = self.customCertRE.match(option)
+            if match:
+                customcert = match.group("nickname")
+                config.write(
+                    "listen:%s:%s:%s:%s\n"
+                    % (loc.host, loc.port, self.sslPort, customcert)
+                )
+
+            match = self.clientAuthRE.match(option)
+            if match:
+                clientauth = match.group("clientauth")
+                config.write(
+                    "clientauth:%s:%s:%s:%s\n"
+                    % (loc.host, loc.port, self.sslPort, clientauth)
+                )
+
+            match = self.redirRE.match(option)
+            if match:
+                redirhost = match.group("redirhost")
+                config.write(
+                    "redirhost:%s:%s:%s:%s\n"
+                    % (loc.host, loc.port, self.sslPort, redirhost)
+                )
+
+            if option in (
+                "tls1",
+                "tls1_1",
+                "tls1_2",
+                "tls1_3",
+                "ssl3",
+                "3des",
+                "failHandshake",
+            ):
+                config.write(
+                    "%s:%s:%s:%s\n" % (option, loc.host, loc.port, self.sslPort)
+                )
+
+    def buildConfig(self, locations, public=None):
+        """Create the ssltunnel configuration file"""
+        configFd, self.configFile = tempfile.mkstemp(prefix="ssltunnel", suffix=".cfg")
+        with os.fdopen(configFd, "w") as config:
+            config.write("httpproxy:1\n")
+            config.write("certdbdir:%s\n" % self.certPath)
+            config.write("forward:127.0.0.1:%s\n" % self.httpPort)
+            config.write(
+                "websocketserver:%s:%s\n" % (self.webServer, self.webSocketPort)
+            )
+            # Use "*" to tell ssltunnel to listen on the public ip
+            # address instead of the loopback address 127.0.0.1. This
+            # may have the side-effect of causing firewall warnings on
+            # macOS and Windows. Use "127.0.0.1" to listen on the
+            # loopback address.  Remote tests using physical or
+            # emulated Android devices must use the public ip address
+            # in order for the sslproxy to work but Desktop tests
+            # which run on the same host as ssltunnel may use the
+            # loopback address.
+            listen_address = "*" if public else "127.0.0.1"
+            config.write("listen:%s:%s:pgoserver\n" % (listen_address, self.sslPort))
+
+            for loc in locations:
+                if loc.scheme == "https" and "nocert" not in loc.options:
+                    self.writeLocation(config, loc)
+
+    def start(self):
+        """Starts the SSL Tunnel"""
+
+        # start ssltunnel to provide https:// URLs capability
+        ssltunnel = os.path.join(self.utilityPath, "ssltunnel")
+        if os.name == "nt":
+            ssltunnel += ".exe"
+        if not os.path.exists(ssltunnel):
+            self.log.error(
+                "INFO | runtests.py | expected to find ssltunnel at %s" % ssltunnel
+            )
+            exit(1)
+
+        env = test_environment(xrePath=self.xrePath, log=self.log)
+        env["LD_LIBRARY_PATH"] = self.xrePath
+        self.process = mozprocess.ProcessHandler([ssltunnel, self.configFile], env=env)
+        self.process.run()
+        self.log.info("runtests.py | SSL tunnel pid: %d" % self.process.pid)
+
+    def stop(self):
+        """Stops the SSL Tunnel and cleans up"""
+        if self.process is not None:
+            self.process.kill()
+        if os.path.exists(self.configFile):
+            os.remove(self.configFile)
+
+
+def checkAndConfigureV4l2loopback(device):
+    """
+    Determine if a given device path is a v4l2loopback device, and if so
+    toggle a few settings on it via fcntl. Very linux-specific.
+
+    Returns (status, device name) where status is a boolean.
+    """
+    if not mozinfo.isLinux:
+        return False, ""
+
+    libc = ctypes.cdll.LoadLibrary(find_library("c"))
+    O_RDWR = 2
+    # These are from linux/videodev2.h
+
+    class v4l2_capability(ctypes.Structure):
+        _fields_ = [
+            ("driver", ctypes.c_char * 16),
+            ("card", ctypes.c_char * 32),
+            ("bus_info", ctypes.c_char * 32),
+            ("version", ctypes.c_uint32),
+            ("capabilities", ctypes.c_uint32),
+            ("device_caps", ctypes.c_uint32),
+            ("reserved", ctypes.c_uint32 * 3),
+        ]
+
+    VIDIOC_QUERYCAP = 0x80685600
+
+    fd = libc.open(six.ensure_binary(device), O_RDWR)
+    if fd < 0:
+        return False, ""
+
+    vcap = v4l2_capability()
+    if libc.ioctl(fd, VIDIOC_QUERYCAP, ctypes.byref(vcap)) != 0:
+        return False, ""
+
+    if six.ensure_text(vcap.driver) != "v4l2 loopback":
+        return False, ""
+
+    class v4l2_control(ctypes.Structure):
+        _fields_ = [("id", ctypes.c_uint32), ("value", ctypes.c_int32)]
+
+    # These are private v4l2 control IDs, see:
+    # https://github.com/umlaeute/v4l2loopback/blob/fd822cf0faaccdf5f548cddd9a5a3dcebb6d584d/v4l2loopback.c#L131
+    KEEP_FORMAT = 0x8000000
+    SUSTAIN_FRAMERATE = 0x8000001
+    VIDIOC_S_CTRL = 0xC008561C
+
+    control = v4l2_control()
+    control.id = KEEP_FORMAT
+    control.value = 1
+    libc.ioctl(fd, VIDIOC_S_CTRL, ctypes.byref(control))
+
+    control.id = SUSTAIN_FRAMERATE
+    control.value = 1
+    libc.ioctl(fd, VIDIOC_S_CTRL, ctypes.byref(control))
+    libc.close(fd)
+
+    return True, six.ensure_text(vcap.card)
+
+
+def findTestMediaDevices(log):
+    """
+    Find the test media devices configured on this system, and return a dict
+    containing information about them. The dict will have keys for 'audio'
+    and 'video', each containing the name of the media device to use.
+
+    If audio and video devices could not be found, return None.
+
+    This method is only currently implemented for Linux.
+    """
+    if not mozinfo.isLinux:
+        return None
+
+    info = {}
+    # Look for a v4l2loopback device.
+    name = None
+    device = None
+    for dev in sorted(glob.glob("/dev/video*")):
+        result, name_ = checkAndConfigureV4l2loopback(dev)
+        if result:
+            name = name_
+            device = dev
+            break
+
+    if not (name and device):
+        log.error("Couldn't find a v4l2loopback video device")
+        return None
+
+    # Feed it a frame of output so it has something to display
+    gst01 = spawn.find_executable("gst-launch-0.1")
+    gst010 = spawn.find_executable("gst-launch-0.10")
+    gst10 = spawn.find_executable("gst-launch-1.0")
+    if gst01:
+        gst = gst01
+    if gst010:
+        gst = gst010
+    else:
+        gst = gst10
+    process = mozprocess.ProcessHandler(
+        [
+            gst,
+            "--no-fault",
+            "videotestsrc",
+            "pattern=green",
+            "num-buffers=1",
+            "!",
+            "v4l2sink",
+            "device=%s" % device,
+        ]
+    )
+    process.run()
+    info["video"] = {"name": name, "process": process}
+
+    # check if PulseAudio module-null-sink is loaded
+    pactl = spawn.find_executable("pactl")
+
+    if not pactl:
+        log.error("Could not find pactl on system")
+        return None
+
+    try:
+        o = subprocess.check_output([pactl, "list", "short", "modules"])
+    except subprocess.CalledProcessError:
+        log.error("Could not list currently loaded modules")
+        return None
+
+    null_sink = [x for x in o.splitlines() if b"module-null-sink" in x]
+
+    if not null_sink:
+        try:
+            subprocess.check_call([pactl, "load-module", "module-null-sink"])
+        except subprocess.CalledProcessError:
+            log.error("Could not load module-null-sink")
+            return None
+
+    # Hardcode the name since it's always the same.
+    info["audio"] = {"name": "Monitor of Null Output"}
+    return info
+
+
+def create_zip(path):
+    """
+    Takes a `path` on disk and creates a zipfile with its contents. Returns a
+    path to the location of the temporary zip file.
+    """
+    with tempfile.NamedTemporaryFile() as f:
+        # `shutil.make_archive` writes to "{f.name}.zip", so we're really just
+        # using `NamedTemporaryFile` as a way to get a random path.
+        return shutil.make_archive(f.name, "zip", path)
+
+
+def update_mozinfo():
+    """walk up directories to find mozinfo.json update the info"""
+    # TODO: This should go in a more generic place, e.g. mozinfo
+
+    path = SCRIPT_DIR
+    dirs = set()
+    while path != os.path.expanduser("~"):
+        if path in dirs:
+            break
+        dirs.add(path)
+        path = os.path.split(path)[0]
+
+    mozinfo.find_and_update_from_json(*dirs)
+
+
+class MochitestDesktop(object):
+    """
+    Mochitest class for desktop firefox.
+    """
+
+    oldcwd = os.getcwd()
+
+    # Path to the test script on the server
+    TEST_PATH = "tests"
+    CHROME_PATH = "redirect.html"
+
+    certdbNew = False
+    sslTunnel = None
+    DEFAULT_TIMEOUT = 60.0
+    mediaDevices = None
+    mozinfo_variables_shown = False
+
+    patternFiles = {}
+
+    # XXX use automation.py for test name to avoid breaking legacy
+    # TODO: replace this with 'runtests.py' or 'mochitest' or the like
+    test_name = "automation.py"
+
+    def __init__(self, flavor, logger_options, staged_addons=None, quiet=False):
+        update_mozinfo()
+        self.flavor = flavor
+        self.staged_addons = staged_addons
+        self.server = None
+        self.wsserver = None
+        self.websocketProcessBridge = None
+        self.sslTunnel = None
+        self.manifest = None
+        self.tests_by_manifest = defaultdict(list)
+        self.args_by_manifest = defaultdict(set)
+        self.prefs_by_manifest = defaultdict(set)
+        self.env_vars_by_manifest = defaultdict(set)
+        self.tests_dirs_by_manifest = defaultdict(set)
+        self._active_tests = None
+        self.currentTests = None
+        self._locations = None
+        self.browserEnv = None
+
+        self.marionette = None
+        self.start_script = None
+        self.mozLogs = None
+        self.start_script_kwargs = {}
+        self.extraArgs = []
+        self.extraPrefs = {}
+        self.extraEnv = {}
+        self.extraTestsDirs = []
+        self.conditioned_profile_dir = None
+
+        if logger_options.get("log"):
+            self.log = logger_options["log"]
+        else:
+            self.log = commandline.setup_logging(
+                "mochitest", logger_options, {"tbpl": sys.stdout}
+            )
+
+        self.message_logger = MessageLogger(
+            logger=self.log, buffering=quiet, structured=True
+        )
+
+        # Max time in seconds to wait for server startup before tests will fail -- if
+        # this seems big, it's mostly for debug machines where cold startup
+        # (particularly after a build) takes forever.
+        self.SERVER_STARTUP_TIMEOUT = 180 if mozinfo.info.get("debug") else 90
+
+        # metro browser sub process id
+        self.browserProcessId = None
+
+        self.haveDumpedScreen = False
+        # Create variables to count the number of passes, fails, todos.
+        self.countpass = 0
+        self.countfail = 0
+        self.counttodo = 0
+
+        self.expectedError = {}
+        self.result = {}
+
+        self.start_script = os.path.join(here, "start_desktop.js")
+
+        # Used to temporarily serve a performance profile
+        self.profiler_tempdir = None
+
+    def environment(self, **kwargs):
+        kwargs["log"] = self.log
+        return test_environment(**kwargs)
+
+    def getFullPath(self, path):
+        "Get an absolute path relative to self.oldcwd."
+        return os.path.normpath(os.path.join(self.oldcwd, os.path.expanduser(path)))
+
+    def getLogFilePath(self, logFile):
+        """return the log file path relative to the device we are testing on, in most cases
+        it will be the full path on the local system
+        """
+        return self.getFullPath(logFile)
+
+    @property
+    def locations(self):
+        if self._locations is not None:
+            return self._locations
+        locations_file = os.path.join(SCRIPT_DIR, "server-locations.txt")
+        self._locations = ServerLocations(locations_file)
+        return self._locations
+
+    def buildURLOptions(self, options, env):
+        """Add test control options from the command line to the url
+
+        URL parameters to test URL:
+
+        autorun -- kick off tests automatically
+        closeWhenDone -- closes the browser after the tests
+        hideResultsTable -- hides the table of individual test results
+        logFile -- logs test run to an absolute path
+        startAt -- name of test to start at
+        endAt -- name of test to end at
+        timeout -- per-test timeout in seconds
+        repeat -- How many times to repeat the test, ie: repeat=1 will run the test twice.
+        """
+        self.urlOpts = []
+
+        if not hasattr(options, "logFile"):
+            options.logFile = ""
+        if not hasattr(options, "fileLevel"):
+            options.fileLevel = "INFO"
+
+        # allow relative paths for logFile
+        if options.logFile:
+            options.logFile = self.getLogFilePath(options.logFile)
+
+        if options.flavor in ("a11y", "browser", "chrome"):
+            self.makeTestConfig(options)
+        else:
+            if options.autorun:
+                self.urlOpts.append("autorun=1")
+            if options.timeout:
+                self.urlOpts.append("timeout=%d" % options.timeout)
+            if options.maxTimeouts:
+                self.urlOpts.append("maxTimeouts=%d" % options.maxTimeouts)
+            if not options.keep_open:
+                self.urlOpts.append("closeWhenDone=1")
+            if options.logFile:
+                self.urlOpts.append("logFile=" + encodeURIComponent(options.logFile))
+                self.urlOpts.append(
+                    "fileLevel=" + encodeURIComponent(options.fileLevel)
+                )
+            if options.consoleLevel:
+                self.urlOpts.append(
+                    "consoleLevel=" + encodeURIComponent(options.consoleLevel)
+                )
+            if options.startAt:
+                self.urlOpts.append("startAt=%s" % options.startAt)
+            if options.endAt:
+                self.urlOpts.append("endAt=%s" % options.endAt)
+            if options.shuffle:
+                self.urlOpts.append("shuffle=1")
+            if "MOZ_HIDE_RESULTS_TABLE" in env and env["MOZ_HIDE_RESULTS_TABLE"] == "1":
+                self.urlOpts.append("hideResultsTable=1")
+            if options.runUntilFailure:
+                self.urlOpts.append("runUntilFailure=1")
+            if options.repeat:
+                self.urlOpts.append("repeat=%d" % options.repeat)
+            if len(options.test_paths) == 1 and os.path.isfile(
+                os.path.join(
+                    self.oldcwd,
+                    os.path.dirname(__file__),
+                    self.TEST_PATH,
+                    options.test_paths[0],
+                )
+            ):
+                self.urlOpts.append(
+                    "testname=%s" % "/".join([self.TEST_PATH, options.test_paths[0]])
+                )
+            if options.manifestFile:
+                self.urlOpts.append("manifestFile=%s" % options.manifestFile)
+            if options.failureFile:
+                self.urlOpts.append(
+                    "failureFile=%s" % self.getFullPath(options.failureFile)
+                )
+            if options.runSlower:
+                self.urlOpts.append("runSlower=true")
+            if options.debugOnFailure:
+                self.urlOpts.append("debugOnFailure=true")
+            if options.dumpOutputDirectory:
+                self.urlOpts.append(
+                    "dumpOutputDirectory=%s"
+                    % encodeURIComponent(options.dumpOutputDirectory)
+                )
+            if options.dumpAboutMemoryAfterTest:
+                self.urlOpts.append("dumpAboutMemoryAfterTest=true")
+            if options.dumpDMDAfterTest:
+                self.urlOpts.append("dumpDMDAfterTest=true")
+            if options.debugger:
+                self.urlOpts.append("interactiveDebugger=true")
+            if options.jscov_dir_prefix:
+                self.urlOpts.append("jscovDirPrefix=%s" % options.jscov_dir_prefix)
+            if options.cleanupCrashes:
+                self.urlOpts.append("cleanupCrashes=true")
+            if "MOZ_XORIGIN_MOCHITEST" in env and env["MOZ_XORIGIN_MOCHITEST"] == "1":
+                options.xOriginTests = True
+            if options.xOriginTests:
+                self.urlOpts.append("xOriginTests=true")
+            if options.comparePrefs:
+                self.urlOpts.append("comparePrefs=true")
+            self.urlOpts.append("ignorePrefsFile=ignorePrefs.json")
+
+    def normflavor(self, flavor):
+        """
+        In some places the string 'browser-chrome' is expected instead of
+        'browser' and 'mochitest' instead of 'plain'. Normalize the flavor
+        strings for those instances.
+        """
+        # TODO Use consistent flavor strings everywhere and remove this
+        if flavor == "browser":
+            return "browser-chrome"
+        elif flavor == "plain":
+            return "mochitest"
+        return flavor
+
+    # This check can be removed when bug 983867 is fixed.
+    def isTest(self, options, filename):
+        allow_js_css = False
+        if options.flavor == "browser":
+            allow_js_css = True
+            testPattern = re.compile(r"browser_.+\.js")
+        elif options.flavor in ("a11y", "chrome"):
+            testPattern = re.compile(r"(browser|test)_.+\.(xul|html|js|xhtml)")
+        else:
+            testPattern = re.compile(r"test_")
+
+        if not allow_js_css and (".js" in filename or ".css" in filename):
+            return False
+
+        pathPieces = filename.split("/")
+
+        return testPattern.match(pathPieces[-1]) and not re.search(
+            r"\^headers\^$", filename
+        )
+
+    def setTestRoot(self, options):
+        if options.flavor != "plain":
+            self.testRoot = options.flavor
+        else:
+            self.testRoot = self.TEST_PATH
+        self.testRootAbs = os.path.join(SCRIPT_DIR, self.testRoot)
+
+    def buildTestURL(self, options, scheme="http"):
+        if scheme == "https":
+            testHost = "https://example.com:443"
+        elif options.xOriginTests:
+            testHost = "http://mochi.xorigin-test:8888"
+        else:
+            testHost = "http://mochi.test:8888"
+        testURL = "/".join([testHost, self.TEST_PATH])
+
+        if len(options.test_paths) == 1:
+            if os.path.isfile(
+                os.path.join(
+                    self.oldcwd,
+                    os.path.dirname(__file__),
+                    self.TEST_PATH,
+                    options.test_paths[0],
+                )
+            ):
+                testURL = "/".join([testURL, os.path.dirname(options.test_paths[0])])
+            else:
+                testURL = "/".join([testURL, options.test_paths[0]])
+
+        if options.flavor in ("a11y", "chrome"):
+            testURL = "/".join([testHost, self.CHROME_PATH])
+        elif options.flavor == "browser":
+            testURL = "about:blank"
+        return testURL
+
+    def parseAndCreateTestsDirs(self, m):
+        testsDirs = list(self.tests_dirs_by_manifest[m])[0]
+        self.extraTestsDirs = []
+        if testsDirs:
+            self.extraTestsDirs = testsDirs.strip().split()
+            self.log.info(
+                "The following extra test directories will be created:\n  {}".format(
+                    "\n  ".join(self.extraTestsDirs)
+                )
+            )
+            self.createExtraTestsDirs(self.extraTestsDirs, m)
+
+    def createExtraTestsDirs(self, extraTestsDirs=None, manifest=None):
+        """Take a list of directories that might be needed to exist by the test
+        prior to even the main process be executed, and:
+         - verify it does not already exists
+         - create it if it does
+        Removal of those directories is handled in cleanup()
+        """
+        if type(extraTestsDirs) != list:
+            return
+
+        for d in extraTestsDirs:
+            if os.path.exists(d):
+                raise FileExistsError(
+                    "Directory '{}' already exists. This is a member of "
+                    "test-directories in manifest {}.".format(d, manifest)
+                )
+
+        created = []
+        for d in extraTestsDirs:
+            os.makedirs(d)
+            created += [d]
+
+        if created != extraTestsDirs:
+            raise EnvironmentError(
+                "Not all directories were created: extraTestsDirs={} -- created={}".format(
+                    extraTestsDirs, created
+                )
+            )
+
+    def getTestsByScheme(
+        self, options, testsToFilter=None, disabled=True, manifestToFilter=None
+    ):
+        """Build the url path to the specific test harness and test file or directory
+        Build a manifest of tests to run and write out a json file for the harness to read
+        testsToFilter option is used to filter/keep the tests provided in the list
+
+        disabled -- This allows to add all disabled tests on the build side
+                    and then on the run side to only run the enabled ones
+        """
+
+        tests = self.getActiveTests(options, disabled)
+        paths = []
+        for test in tests:
+            if testsToFilter and (test["path"] not in testsToFilter):
+                continue
+            # If we are running a specific manifest, the previously computed set of active
+            # tests should be filtered out based on the manifest that contains that entry.
+            #
+            # This is especially important when a test file is listed in multiple
+            # manifests (e.g. because the same test runs under a different configuration,
+            # and so it is being included in multiple manifests), without filtering the
+            # active tests based on the current manifest (configuration) that we are
+            # running for each of the N manifests we would be executing the active tests
+            # exactly N times (and so NxN runs instead of the expected N runs, one for each
+            # manifest).
+            if manifestToFilter and (test["manifest"] not in manifestToFilter):
+                continue
+            paths.append(test)
+
+        # Generate test by schemes
+        for (scheme, grouped_tests) in self.groupTestsByScheme(paths).items():
+            # Bug 883865 - add this functionality into manifestparser
+            with open(
+                os.path.join(SCRIPT_DIR, options.testRunManifestFile), "w"
+            ) as manifestFile:
+                manifestFile.write(json.dumps({"tests": grouped_tests}))
+            options.manifestFile = options.testRunManifestFile
+            yield (scheme, grouped_tests)
+
+    def startWebSocketServer(self, options, debuggerInfo):
+        """Launch the websocket server"""
+        self.wsserver = WebSocketServer(options, SCRIPT_DIR, self.log, debuggerInfo)
+        self.wsserver.start()
+
+    def startWebServer(self, options):
+        """Create the webserver and start it up"""
+
+        self.server = MochitestServer(options, self.log)
+        self.server.start()
+
+        if options.pidFile != "":
+            with open(options.pidFile + ".xpcshell.pid", "w") as f:
+                f.write("%s" % self.server._process.pid)
+
+    def startWebsocketProcessBridge(self, options):
+        """Create a websocket server that can launch various processes that
+        JS needs (eg; ICE server for webrtc testing)
+        """
+
+        command = [
+            sys.executable,
+            os.path.join("websocketprocessbridge", "websocketprocessbridge.py"),
+            "--port",
+            options.websocket_process_bridge_port,
+        ]
+        self.websocketProcessBridge = mozprocess.ProcessHandler(command, cwd=SCRIPT_DIR)
+        self.websocketProcessBridge.run()
+        self.log.info(
+            "runtests.py | websocket/process bridge pid: %d"
+            % self.websocketProcessBridge.pid
+        )
+
+        # ensure the server is up, wait for at most ten seconds
+        for i in range(1, 100):
+            if self.websocketProcessBridge.proc.poll() is not None:
+                self.log.error(
+                    "runtests.py | websocket/process bridge failed "
+                    "to launch. Are all the dependencies installed?"
+                )
+                return
+
+            try:
+                sock = socket.create_connection(("127.0.0.1", 8191))
+                sock.close()
+                break
+            except Exception:
+                time.sleep(0.1)
+        else:
+            self.log.error(
+                "runtests.py | Timed out while waiting for "
+                "websocket/process bridge startup."
+            )
+
+    def needsWebsocketProcessBridge(self, options):
+        """
+        Returns a bool indicating if the current test configuration needs
+        to start the websocket process bridge or not. The boils down to if
+        WebRTC tests that need the bridge are present.
+        """
+        tests = self.getActiveTests(options)
+        is_webrtc_tag_present = False
+        for test in tests:
+            if "webrtc" in test.get("tags", ""):
+                is_webrtc_tag_present = True
+                break
+        return is_webrtc_tag_present and options.subsuite in ["media"]
+
+    def startHttp3Server(self, options):
+        """
+        Start a Http3 test server.
+        """
+        http3ServerPath = os.path.join(
+            options.utilityPath, "http3server" + mozinfo.info["bin_suffix"]
+        )
+        serverOptions = {}
+        serverOptions["http3ServerPath"] = http3ServerPath
+        serverOptions["profilePath"] = options.profilePath
+        serverOptions["isMochitest"] = True
+        serverOptions["isWin"] = mozinfo.isWin
+        serverOptions["proxyPort"] = options.http3ServerPort
+        env = test_environment(xrePath=options.xrePath, log=self.log)
+        self.http3Server = Http3Server(serverOptions, env, self.log)
+        self.http3Server.start()
+
+        port = self.http3Server.ports().get("MOZHTTP3_PORT_PROXY")
+        if int(port) != options.http3ServerPort:
+            self.http3Server = None
+            raise RuntimeError("Error: Unable to start Http/3 server")
+
+    def startDoHServer(self, options):
+        """
+        Start a DoH test server.
+        """
+        # We try to find the node executable in the path given to us by the user in
+        # the MOZ_NODE_PATH environment variable
+        nodeBin = os.getenv("MOZ_NODE_PATH", None)
+        self.log.info("Use MOZ_NODE_PATH at %s" % (nodeBin))
+        if not nodeBin and build:
+            nodeBin = build.substs.get("NODEJS")
+            self.log.info("Use build node at %s" % (nodeBin))
+
+        serverOptions = {}
+        serverOptions["serverPath"] = os.path.join(
+            SCRIPT_DIR, "DoHServer", "doh_server.js"
+        )
+        serverOptions["nodeBin"] = nodeBin
+        serverOptions["dstServerPort"] = options.http3ServerPort
+        serverOptions["isWin"] = mozinfo.isWin
+        serverOptions["port"] = options.dohServerPort
+        env = test_environment(xrePath=options.xrePath, log=self.log)
+        self.dohServer = DoHServer(serverOptions, env, self.log)
+        self.dohServer.start()
+
+        port = self.dohServer.port()
+        if port != options.dohServerPort:
+            raise RuntimeError("Error: Unable to start DoH server")
+
+    def startServers(self, options, debuggerInfo, public=None):
+        # start servers and set ports
+        # TODO: pass these values, don't set on `self`
+        self.webServer = options.webServer
+        self.httpPort = options.httpPort
+        self.sslPort = options.sslPort
+        self.webSocketPort = options.webSocketPort
+
+        # httpd-path is specified by standard makefile targets and may be specified
+        # on the command line to select a particular version of httpd.js. If not
+        # specified, try to select the one from hostutils.zip, as required in
+        # bug 882932.
+        if not options.httpdPath:
+            options.httpdPath = os.path.join(options.utilityPath, "components")
+
+        self.startWebServer(options)
+        self.startWebSocketServer(options, debuggerInfo)
+
+        # Only webrtc mochitests in the media suite need the websocketprocessbridge.
+        if self.needsWebsocketProcessBridge(options):
+            self.startWebsocketProcessBridge(options)
+
+        # start SSL pipe
+        self.sslTunnel = SSLTunnel(options, logger=self.log)
+        self.sslTunnel.buildConfig(self.locations, public=public)
+        self.sslTunnel.start()
+
+        # If we're lucky, the server has fully started by now, and all paths are
+        # ready, etc.  However, xpcshell cold start times suck, at least for debug
+        # builds.  We'll try to connect to the server for awhile, and if we fail,
+        # we'll try to kill the server and exit with an error.
+        if self.server is not None:
+            self.server.ensureReady(self.SERVER_STARTUP_TIMEOUT)
+
+        self.log.info("use http3 server: %d" % options.useHttp3Server)
+        self.http3Server = None
+        self.dohServer = None
+        if options.useHttp3Server:
+            self.startHttp3Server(options)
+            self.startDoHServer(options)
+
+    def stopServers(self):
+        """Servers are no longer needed, and perhaps more importantly, anything they
+        might spew to console might confuse things."""
+        if self.server is not None:
+            try:
+                self.log.info("Stopping web server")
+                self.server.stop()
+            except Exception:
+                self.log.critical("Exception when stopping web server")
+
+        if self.wsserver is not None:
+            try:
+                self.log.info("Stopping web socket server")
+                self.wsserver.stop()
+            except Exception:
+                self.log.critical("Exception when stopping web socket server")
+
+        if self.sslTunnel is not None:
+            try:
+                self.log.info("Stopping ssltunnel")
+                self.sslTunnel.stop()
+            except Exception:
+                self.log.critical("Exception stopping ssltunnel")
+
+        if self.websocketProcessBridge is not None:
+            try:
+                self.websocketProcessBridge.kill()
+                self.websocketProcessBridge.wait()
+                self.log.info("Stopping websocket/process bridge")
+            except Exception:
+                self.log.critical("Exception stopping websocket/process bridge")
+        if self.http3Server is not None:
+            try:
+                self.http3Server.stop()
+            except Exception:
+                self.log.critical("Exception stopping http3 server")
+        if self.dohServer is not None:
+            try:
+                self.dohServer.stop()
+            except Exception:
+                self.log.critical("Exception stopping doh server")
+
+        if hasattr(self, "gstForV4l2loopbackProcess"):
+            try:
+                self.gstForV4l2loopbackProcess.kill()
+                self.gstForV4l2loopbackProcess.wait()
+                self.log.info("Stopping gst for v4l2loopback")
+            except Exception:
+                self.log.critical("Exception stopping gst for v4l2loopback")
+
+    def copyExtraFilesToProfile(self, options):
+        "Copy extra files or dirs specified on the command line to the testing profile."
+        for f in options.extraProfileFiles:
+            abspath = self.getFullPath(f)
+            if os.path.isfile(abspath):
+                shutil.copy2(abspath, options.profilePath)
+            elif os.path.isdir(abspath):
+                dest = os.path.join(options.profilePath, os.path.basename(abspath))
+                shutil.copytree(abspath, dest)
+            else:
+                self.log.warning("runtests.py | Failed to copy %s to profile" % abspath)
+
+    def getChromeTestDir(self, options):
+        dir = os.path.join(os.path.abspath("."), SCRIPT_DIR) + "/"
+        if mozinfo.isWin:
+            dir = "file:///" + dir.replace("\\", "/")
+        return dir
+
+    def writeChromeManifest(self, options):
+        manifest = os.path.join(options.profilePath, "tests.manifest")
+        with open(manifest, "w") as manifestFile:
+            # Register chrome directory.
+            chrometestDir = self.getChromeTestDir(options)
+            manifestFile.write(
+                "content mochitests %s contentaccessible=yes\n" % chrometestDir
+            )
+            manifestFile.write(
+                "content mochitests-any %s contentaccessible=yes remoteenabled=yes\n"
+                % chrometestDir
+            )
+            manifestFile.write(
+                "content mochitests-content %s contentaccessible=yes remoterequired=yes\n"
+                % chrometestDir
+            )
+
+            if options.testingModulesDir is not None:
+                manifestFile.write(
+                    "resource testing-common file:///%s\n" % options.testingModulesDir
+                )
+        if options.store_chrome_manifest:
+            shutil.copyfile(manifest, options.store_chrome_manifest)
+        return manifest
+
+    def addChromeToProfile(self, options):
+        "Adds MochiKit chrome tests to the profile."
+
+        # Create (empty) chrome directory.
+        chromedir = os.path.join(options.profilePath, "chrome")
+        os.mkdir(chromedir)
+
+        # Write userChrome.css.
+        chrome = """
+/* set default namespace to XUL */
+@namespace url("http://www.mozilla.org/keymaster/gatekeeper/there.is.only.xul");
+toolbar,
+toolbarpalette {
+  background-color: rgb(235, 235, 235) !important;
+}
+toolbar#nav-bar {
+  background-image: none !important;
+}
+"""
+        with open(
+            os.path.join(options.profilePath, "userChrome.css"), "a"
+        ) as chromeFile:
+            chromeFile.write(chrome)
+
+        manifest = self.writeChromeManifest(options)
+
+        return manifest
+
+    def getExtensionsToInstall(self, options):
+        "Return a list of extensions to install in the profile"
+        extensions = []
+        appDir = (
+            options.app[: options.app.rfind(os.sep)]
+            if options.app
+            else options.utilityPath
+        )
+
+        extensionDirs = [
+            # Extensions distributed with the test harness.
+            os.path.normpath(os.path.join(SCRIPT_DIR, "extensions")),
+        ]
+        if appDir:
+            # Extensions distributed with the application.
+            extensionDirs.append(os.path.join(appDir, "distribution", "extensions"))
+
+        for extensionDir in extensionDirs:
+            if os.path.isdir(extensionDir):
+                for dirEntry in os.listdir(extensionDir):
+                    if dirEntry not in options.extensionsToExclude:
+                        path = os.path.join(extensionDir, dirEntry)
+                        if os.path.isdir(path) or (
+                            os.path.isfile(path) and path.endswith(".xpi")
+                        ):
+                            extensions.append(path)
+        extensions.extend(options.extensionsToInstall)
+        return extensions
+
+    def logPreamble(self, tests):
+        """Logs a suite_start message and test_start/test_end at the beginning of a run."""
+        self.log.suite_start(
+            self.tests_by_manifest, name="mochitest-{}".format(self.flavor)
+        )
+        for test in tests:
+            if "disabled" in test:
+                self.log.test_start(test["path"])
+                self.log.test_end(test["path"], "SKIP", message=test["disabled"])
+
+    def loadFailurePatternFile(self, pat_file):
+        if pat_file in self.patternFiles:
+            return self.patternFiles[pat_file]
+        if not os.path.isfile(pat_file):
+            self.log.warning(
+                "runtests.py | Cannot find failure pattern file " + pat_file
+            )
+            return None
+
+        # Using ":error" to ensure it shows up in the failure summary.
+        self.log.warning(
+            "[runtests.py:error] Using {} to filter failures. If there "
+            "is any number mismatch below, you could have fixed "
+            "something documented in that file. Please reduce the "
+            "failure count appropriately.".format(pat_file)
+        )
+        patternRE = re.compile(
+            r"""
+            ^\s*\*\s*               # list bullet
+            (test_\S+|\.{3})        # test name
+            (?:\s*(`.+?`|asserts))? # failure pattern
+            (?::.+)?                # optional description
+            \s*\[(\d+|\*)\]         # expected count
+            \s*$
+        """,
+            re.X,
+        )
+        patterns = {}
+        with open(pat_file) as f:
+            last_name = None
+            for line in f:
+                match = patternRE.match(line)
+                if not match:
+                    continue
+                name = match.group(1)
+                name = last_name if name == "..." else name
+                last_name = name
+                pat = match.group(2)
+                if pat is not None:
+                    pat = "ASSERTION" if pat == "asserts" else pat[1:-1]
+                count = match.group(3)
+                count = None if count == "*" else int(count)
+                if name not in patterns:
+                    patterns[name] = []
+                patterns[name].append((pat, count))
+        self.patternFiles[pat_file] = patterns
+        return patterns
+
+    def getFailurePatterns(self, pat_file, test_name):
+        patterns = self.loadFailurePatternFile(pat_file)
+        if patterns:
+            return patterns.get(test_name, None)
+
+    def getActiveTests(self, options, disabled=True):
+        """
+        This method is used to parse the manifest and return active filtered tests.
+        """
+        if self._active_tests:
+            return self._active_tests
+
+        tests = []
+        manifest = self.getTestManifest(options)
+        if manifest:
+            if options.extra_mozinfo_json:
+                mozinfo.update(options.extra_mozinfo_json)
+
+            info = mozinfo.info
+
+            filters = [
+                subsuite(options.subsuite),
+            ]
+
+            # Allow for only running tests/manifests which match this tag
+            if options.conditionedProfile:
+                if not options.test_tags:
+                    options.test_tags = []
+                options.test_tags.append("condprof")
+
+            if options.test_tags:
+                filters.append(tags(options.test_tags))
+
+            if options.test_paths:
+                options.test_paths = self.normalize_paths(options.test_paths)
+                filters.append(pathprefix(options.test_paths))
+
+            # Add chunking filters if specified
+            if options.totalChunks:
+                if options.chunkByDir:
+                    filters.append(
+                        chunk_by_dir(
+                            options.thisChunk, options.totalChunks, options.chunkByDir
+                        )
+                    )
+                elif options.chunkByRuntime:
+                    if mozinfo.info["os"] == "android":
+                        platkey = "android"
+                    elif mozinfo.isWin:
+                        platkey = "windows"
+                    else:
+                        platkey = "unix"
+
+                    runtime_file = os.path.join(
+                        SCRIPT_DIR,
+                        "runtimes",
+                        "manifest-runtimes-{}.json".format(platkey),
+                    )
+                    if not os.path.exists(runtime_file):
+                        self.log.error("runtime file %s not found!" % runtime_file)
+                        sys.exit(1)
+
+                    # Given the mochitest flavor, load the runtimes information
+                    # for only that flavor due to manifest runtime format change in Bug 1637463.
+                    with open(runtime_file, "r") as f:
+                        if "suite_name" in options:
+                            runtimes = json.load(f).get(options.suite_name, {})
+                        else:
+                            runtimes = {}
+
+                    filters.append(
+                        chunk_by_runtime(
+                            options.thisChunk, options.totalChunks, runtimes
+                        )
+                    )
+                else:
+                    filters.append(
+                        chunk_by_slice(options.thisChunk, options.totalChunks)
+                    )
+
+            noDefaultFilters = False
+            if options.runFailures:
+                filters.append(failures(options.runFailures))
+                noDefaultFilters = True
+
+            tests = manifest.active_tests(
+                exists=False,
+                disabled=disabled,
+                filters=filters,
+                noDefaultFilters=noDefaultFilters,
+                **info
+            )
+
+            if len(tests) == 0:
+                self.log.error(
+                    NO_TESTS_FOUND.format(options.flavor, manifest.fmt_filters())
+                )
+
+        paths = []
+        for test in tests:
+            if len(tests) == 1 and "disabled" in test:
+                del test["disabled"]
+
+            pathAbs = os.path.abspath(test["path"])
+            assert os.path.normcase(pathAbs).startswith(
+                os.path.normcase(self.testRootAbs)
+            )
+            tp = pathAbs[len(self.testRootAbs) :].replace("\\", "/").strip("/")
+
+            if not self.isTest(options, tp):
+                self.log.warning(
+                    "Warning: %s from manifest %s is not a valid test"
+                    % (test["name"], test["manifest"])
+                )
+                continue
+
+            manifest_key = test["manifest_relpath"]
+            # Ignore ancestor_manifests that live at the root (e.g, don't have a
+            # path separator).
+            if "ancestor_manifest" in test and "/" in normsep(
+                test["ancestor_manifest"]
+            ):
+                manifest_key = "{}:{}".format(test["ancestor_manifest"], manifest_key)
+
+            self.tests_by_manifest[manifest_key.replace("\\", "/")].append(tp)
+            self.args_by_manifest[manifest_key].add(test.get("args"))
+            self.prefs_by_manifest[manifest_key].add(test.get("prefs"))
+            self.env_vars_by_manifest[manifest_key].add(test.get("environment"))
+            self.tests_dirs_by_manifest[manifest_key].add(test.get("test-directories"))
+
+            for key in ["args", "prefs", "environment", "test-directories"]:
+                if key in test and not options.runByManifest and "disabled" not in test:
+                    self.log.error(
+                        "parsing {}: runByManifest mode must be enabled to "
+                        "set the `{}` key".format(test["manifest_relpath"], key)
+                    )
+                    sys.exit(1)
+
+            testob = {"path": tp, "manifest": manifest_key}
+            if "disabled" in test:
+                testob["disabled"] = test["disabled"]
+            if "expected" in test:
+                testob["expected"] = test["expected"]
+            if "https_first_disabled" in test:
+                testob["https_first_disabled"] = test["https_first_disabled"] == "true"
+            if "allow_xul_xbl" in test:
+                testob["allow_xul_xbl"] = test["allow_xul_xbl"] == "true"
+            if "scheme" in test:
+                testob["scheme"] = test["scheme"]
+            if "tags" in test:
+                testob["tags"] = test["tags"]
+            if options.failure_pattern_file:
+                pat_file = os.path.join(
+                    os.path.dirname(test["manifest"]), options.failure_pattern_file
+                )
+                patterns = self.getFailurePatterns(pat_file, test["name"])
+                if patterns:
+                    testob["expected"] = patterns
+            paths.append(testob)
+
+        # The 'args' key needs to be set in the DEFAULT section, unfortunately
+        # we can't tell what comes from DEFAULT or not. So to validate this, we
+        # stash all args from tests in the same manifest into a set. If the
+        # length of the set > 1, then we know 'args' didn't come from DEFAULT.
+        args_not_default = [
+            m for m, p in six.iteritems(self.args_by_manifest) if len(p) > 1
+        ]
+        if args_not_default:
+            self.log.error(
+                "The 'args' key must be set in the DEFAULT section of a "
+                "manifest. Fix the following manifests: {}".format(
+                    "\n".join(args_not_default)
+                )
+            )
+            sys.exit(1)
+
+        # The 'prefs' key needs to be set in the DEFAULT section too.
+        pref_not_default = [
+            m for m, p in six.iteritems(self.prefs_by_manifest) if len(p) > 1
+        ]
+        if pref_not_default:
+            self.log.error(
+                "The 'prefs' key must be set in the DEFAULT section of a "
+                "manifest. Fix the following manifests: {}".format(
+                    "\n".join(pref_not_default)
+                )
+            )
+            sys.exit(1)
+        # The 'environment' key needs to be set in the DEFAULT section too.
+        env_not_default = [
+            m for m, p in six.iteritems(self.env_vars_by_manifest) if len(p) > 1
+        ]
+        if env_not_default:
+            self.log.error(
+                "The 'environment' key must be set in the DEFAULT section of a "
+                "manifest. Fix the following manifests: {}".format(
+                    "\n".join(env_not_default)
+                )
+            )
+            sys.exit(1)
+
+        paths.sort(key=lambda p: p["path"].split("/"))
+        if options.dump_tests:
+            options.dump_tests = os.path.expanduser(options.dump_tests)
+            assert os.path.exists(os.path.dirname(options.dump_tests))
+            with open(options.dump_tests, "w") as dumpFile:
+                dumpFile.write(json.dumps({"active_tests": paths}))
+
+            self.log.info("Dumping active_tests to %s file." % options.dump_tests)
+            sys.exit()
+
+        # Upload a list of test manifests that were executed in this run.
+        if "MOZ_UPLOAD_DIR" in os.environ:
+            artifact = os.path.join(os.environ["MOZ_UPLOAD_DIR"], "manifests.list")
+            with open(artifact, "a") as fh:
+                fh.write("\n".join(sorted(self.tests_by_manifest.keys())))
+
+        self._active_tests = paths
+        return self._active_tests
+
+    def getTestManifest(self, options):
+        if isinstance(options.manifestFile, TestManifest):
+            manifest = options.manifestFile
+        elif options.manifestFile and os.path.isfile(options.manifestFile):
+            manifestFileAbs = os.path.abspath(options.manifestFile)
+            assert manifestFileAbs.startswith(SCRIPT_DIR)
+            manifest = TestManifest([options.manifestFile], strict=False)
+        elif options.manifestFile and os.path.isfile(
+            os.path.join(SCRIPT_DIR, options.manifestFile)
+        ):
+            manifestFileAbs = os.path.abspath(
+                os.path.join(SCRIPT_DIR, options.manifestFile)
+            )
+            assert manifestFileAbs.startswith(SCRIPT_DIR)
+            manifest = TestManifest([manifestFileAbs], strict=False)
+        else:
+            masterName = self.normflavor(options.flavor) + ".ini"
+            masterPath = os.path.join(SCRIPT_DIR, self.testRoot, masterName)
+
+            if os.path.exists(masterPath):
+                manifest = TestManifest([masterPath], strict=False)
+            else:
+                manifest = None
+                self.log.warning(
+                    "TestManifest masterPath %s does not exist" % masterPath
+                )
+
+        return manifest
+
+    def makeTestConfig(self, options):
+        "Creates a test configuration file for customizing test execution."
+        options.logFile = options.logFile.replace("\\", "\\\\")
+
+        if (
+            "MOZ_HIDE_RESULTS_TABLE" in os.environ
+            and os.environ["MOZ_HIDE_RESULTS_TABLE"] == "1"
+        ):
+            options.hideResultsTable = True
+
+        # strip certain unnecessary items to avoid serialization errors in json.dumps()
+        d = dict(
+            (k, v)
+            for k, v in options.__dict__.items()
+            if (v is None) or isinstance(v, (six.string_types, numbers.Number))
+        )
+        d["testRoot"] = self.testRoot
+        if options.jscov_dir_prefix:
+            d["jscovDirPrefix"] = options.jscov_dir_prefix
+        if not options.keep_open:
+            d["closeWhenDone"] = "1"
+
+        d["runFailures"] = False
+        if options.runFailures:
+            d["runFailures"] = True
+        content = json.dumps(d)
+
+        shutil.copy(
+            os.path.join(SCRIPT_DIR, "ignorePrefs.json"),
+            os.path.join(options.profilePath, "ignorePrefs.json"),
+        )
+        d["ignorePrefsFile"] = "ignorePrefs.json"
+
+        with open(os.path.join(options.profilePath, "testConfig.js"), "w") as config:
+            config.write(content)
+
+    def buildBrowserEnv(self, options, debugger=False, env=None):
+        """build the environment variables for the specific test and operating system"""
+        if mozinfo.info["asan"] and mozinfo.isLinux and mozinfo.bits == 64:
+            useLSan = True
+        else:
+            useLSan = False
+
+        browserEnv = self.environment(
+            xrePath=options.xrePath, env=env, debugger=debugger, useLSan=useLSan
+        )
+
+        if hasattr(options, "topsrcdir"):
+            browserEnv["MOZ_DEVELOPER_REPO_DIR"] = options.topsrcdir
+        if hasattr(options, "topobjdir"):
+            browserEnv["MOZ_DEVELOPER_OBJ_DIR"] = options.topobjdir
+
+        if options.headless:
+            browserEnv["MOZ_HEADLESS"] = "1"
+
+        if options.dmd:
+            browserEnv["DMD"] = os.environ.get("DMD", "1")
+
+        # bug 1443327: do not set MOZ_CRASHREPORTER_SHUTDOWN during browser-chrome
+        # tests, since some browser-chrome tests test content process crashes;
+        # also exclude non-e10s since at least one non-e10s mochitest is problematic
+        if (
+            options.flavor == "browser" or not options.e10s
+        ) and "MOZ_CRASHREPORTER_SHUTDOWN" in browserEnv:
+            del browserEnv["MOZ_CRASHREPORTER_SHUTDOWN"]
+
+        try:
+            browserEnv.update(
+                dict(
+                    parse_key_value(
+                        self.extraEnv, context="environment variable in manifest"
+                    )
+                )
+            )
+        except KeyValueParseError as e:
+            self.log.error(str(e))
+            return None
+
+        # These variables are necessary for correct application startup; change
+        # via the commandline at your own risk.
+        browserEnv["XPCOM_DEBUG_BREAK"] = "stack"
+
+        # interpolate environment passed with options
+        try:
+            browserEnv.update(
+                dict(parse_key_value(options.environment, context="--setenv"))
+            )
+        except KeyValueParseError as e:
+            self.log.error(str(e))
+            return None
+
+        if (
+            "MOZ_PROFILER_STARTUP_FEATURES" not in browserEnv
+            or "nativeallocations"
+            not in browserEnv["MOZ_PROFILER_STARTUP_FEATURES"].split(",")
+        ):
+            # Only turn on the bloat log if the profiler's native allocation feature is
+            # not enabled. The two are not compatible.
+            browserEnv["XPCOM_MEM_BLOAT_LOG"] = self.leak_report_file
+
+        # If profiling options are enabled, turn on the gecko profiler by using the
+        # profiler environmental variables.
+        if options.profiler:
+            # The user wants to capture a profile, and automatically view it. The
+            # profile will be saved to a temporary folder, then deleted after
+            # opening in profiler.firefox.com.
+            self.profiler_tempdir = tempfile.mkdtemp()
+            browserEnv["MOZ_PROFILER_SHUTDOWN"] = os.path.join(
+                self.profiler_tempdir, "mochitest-profile.json"
+            )
+            browserEnv["MOZ_PROFILER_STARTUP"] = "1"
+
+        if options.profilerSaveOnly:
+            # The user wants to capture a profile, but only to save it. This defaults
+            # to the MOZ_UPLOAD_DIR.
+            browserEnv["MOZ_PROFILER_STARTUP"] = "1"
+            if "MOZ_UPLOAD_DIR" in browserEnv:
+                browserEnv["MOZ_PROFILER_SHUTDOWN"] = os.path.join(
+                    browserEnv["MOZ_UPLOAD_DIR"], "mochitest-profile.json"
+                )
+            else:
+                self.log.error(
+                    "--profiler-save-only was specified, but no MOZ_UPLOAD_DIR "
+                    "environment variable was provided. Please set this "
+                    "environment variable to a directory path in order to save "
+                    "a performance profile."
+                )
+                return None
+
+        try:
+            gmp_path = self.getGMPPluginPath(options)
+            if gmp_path is not None:
+                browserEnv["MOZ_GMP_PATH"] = gmp_path
+        except EnvironmentError:
+            self.log.error("Could not find path to gmp-fake plugin!")
+            return None
+
+        if options.fatalAssertions:
+            browserEnv["XPCOM_DEBUG_BREAK"] = "stack-and-abort"
+
+        # Produce a mozlog, if setup (see MOZ_LOG global at the top of
+        # this script).
+        self.mozLogs = MOZ_LOG and "MOZ_UPLOAD_DIR" in os.environ
+        if self.mozLogs:
+            browserEnv["MOZ_LOG"] = MOZ_LOG
+
+        return browserEnv
+
+    def killNamedProc(self, pname, orphans=True):
+        """Kill processes matching the given command name"""
+        self.log.info("Checking for %s processes..." % pname)
+
+        if HAVE_PSUTIL:
+            for proc in psutil.process_iter():
+                try:
+                    if proc.name() == pname:
+                        procd = proc.as_dict(attrs=["pid", "ppid", "name", "username"])
+                        if proc.ppid() == 1 or not orphans:
+                            self.log.info("killing %s" % procd)
+                            killPid(proc.pid, self.log)
+                        else:
+                            self.log.info("NOT killing %s (not an orphan?)" % procd)
+                except Exception as e:
+                    self.log.info(
+                        "Warning: Unable to kill process %s: %s" % (pname, str(e))
+                    )
+                    # may not be able to access process info for all processes
+                    continue
+        else:
+
+            def _psInfo(line):
+                if pname in line:
+                    self.log.info(line)
+
+            process = mozprocess.ProcessHandler(
+                ["ps", "-f"], processOutputLine=_psInfo, universal_newlines=True
+            )
+            process.run()
+            process.wait()
+
+            def _psKill(line):
+                parts = line.split()
+                if len(parts) == 3 and parts[0].isdigit():
+                    pid = int(parts[0])
+                    ppid = int(parts[1])
+                    if parts[2] == pname:
+                        if ppid == 1 or not orphans:
+                            self.log.info("killing %s (pid %d)" % (pname, pid))
+                            killPid(pid, self.log)
+                        else:
+                            self.log.info(
+                                "NOT killing %s (pid %d) (not an orphan?)"
+                                % (pname, pid)
+                            )
+
+            process = mozprocess.ProcessHandler(
+                ["ps", "-o", "pid,ppid,comm"],
+                processOutputLine=_psKill,
+                universal_newlines=True,
+            )
+            process.run()
+            process.wait()
+
+    def execute_start_script(self):
+        if not self.start_script or not self.marionette:
+            return
+
+        if os.path.isfile(self.start_script):
+            with open(self.start_script, "r") as fh:
+                script = fh.read()
+        else:
+            script = self.start_script
+
+        with self.marionette.using_context("chrome"):
+            return self.marionette.execute_script(
+                script, script_args=(self.start_script_kwargs,)
+            )
+
+    def fillCertificateDB(self, options):
+        # TODO: move -> mozprofile:
+        # https://bugzilla.mozilla.org/show_bug.cgi?id=746243#c35
+
+        pwfilePath = os.path.join(options.profilePath, ".crtdbpw")
+        with open(pwfilePath, "w") as pwfile:
+            pwfile.write("\n")
+
+        # Pre-create the certification database for the profile
+        env = self.environment(xrePath=options.xrePath)
+        env["LD_LIBRARY_PATH"] = options.xrePath
+        bin_suffix = mozinfo.info.get("bin_suffix", "")
+        certutil = os.path.join(options.utilityPath, "certutil" + bin_suffix)
+        pk12util = os.path.join(options.utilityPath, "pk12util" + bin_suffix)
+        toolsEnv = env
+        if mozinfo.info["asan"]:
+            # Disable leak checking when running these tools
+            toolsEnv["ASAN_OPTIONS"] = "detect_leaks=0"
+        if mozinfo.info["tsan"]:
+            # Disable race checking when running these tools
+            toolsEnv["TSAN_OPTIONS"] = "report_bugs=0"
+
+        if self.certdbNew:
+            # android uses the new DB formats exclusively
+            certdbPath = "sql:" + options.profilePath
+        else:
+            # desktop seems to use the old
+            certdbPath = options.profilePath
+
+        # certutil.exe depends on some DLLs in the app directory
+        # When running tests against an MSIX-installed Firefox, these DLLs
+        # cannot be used out of the install directory, they must be copied
+        # elsewhere first.
+        if "WindowsApps" in options.app:
+            install_dir = os.path.dirname(options.app)
+            for f in os.listdir(install_dir):
+                if f.endswith(".dll"):
+                    shutil.copy(os.path.join(install_dir, f), options.utilityPath)
+
+        status = call(
+            [certutil, "-N", "-d", certdbPath, "-f", pwfilePath], env=toolsEnv
+        )
+        if status:
+            return status
+
+        # Walk the cert directory and add custom CAs and client certs
+        files = os.listdir(options.certPath)
+        for item in files:
+            root, ext = os.path.splitext(item)
+            if ext == ".ca":
+                trustBits = "CT,,"
+                if root.endswith("-object"):
+                    trustBits = "CT,,CT"
+                call(
+                    [
+                        certutil,
+                        "-A",
+                        "-i",
+                        os.path.join(options.certPath, item),
+                        "-d",
+                        certdbPath,
+                        "-f",
+                        pwfilePath,
+                        "-n",
+                        root,
+                        "-t",
+                        trustBits,
+                    ],
+                    env=toolsEnv,
+                )
+            elif ext == ".client":
+                call(
+                    [
+                        pk12util,
+                        "-i",
+                        os.path.join(options.certPath, item),
+                        "-w",
+                        pwfilePath,
+                        "-d",
+                        certdbPath,
+                    ],
+                    env=toolsEnv,
+                )
+
+        os.unlink(pwfilePath)
+        return 0
+
+    def findFreePort(self, type):
+        with closing(socket.socket(socket.AF_INET, type)) as s:
+            s.bind(("127.0.0.1", 0))
+            s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+            return s.getsockname()[1]
+
+    def proxy(self, options):
+        # proxy
+        # use SSL port for legacy compatibility; see
+        # - https://bugzilla.mozilla.org/show_bug.cgi?id=688667#c66
+        # - https://bugzilla.mozilla.org/show_bug.cgi?id=899221
+        # - https://github.com/mozilla/mozbase/commit/43f9510e3d58bfed32790c82a57edac5f928474d
+        #             'ws': str(self.webSocketPort)
+        proxyOptions = {
+            "remote": options.webServer,
+            "http": options.httpPort,
+            "https": options.sslPort,
+            "ws": options.sslPort,
+        }
+
+        if options.useHttp3Server:
+            options.dohServerPort = self.findFreePort(socket.SOCK_STREAM)
+            options.http3ServerPort = self.findFreePort(socket.SOCK_DGRAM)
+            proxyOptions["dohServerPort"] = options.dohServerPort
+            self.log.info("use doh server at port: %d" % options.dohServerPort)
+            self.log.info("use http3 server at port: %d" % options.http3ServerPort)
+        return proxyOptions
+
+    def merge_base_profiles(self, options, category):
+        """Merge extra profile data from testing/profiles."""
+
+        # In test packages used in CI, the profile_data directory is installed
+        # in the SCRIPT_DIR.
+        profile_data_dir = os.path.join(SCRIPT_DIR, "profile_data")
+        # If possible, read profile data from topsrcdir. This prevents us from
+        # requiring a re-build to pick up newly added extensions in the
+        # <profile>/extensions directory.
+        if build_obj:
+            path = os.path.join(build_obj.topsrcdir, "testing", "profiles")
+            if os.path.isdir(path):
+                profile_data_dir = path
+        # Still not found? Look for testing/profiles relative to testing/mochitest.
+        if not os.path.isdir(profile_data_dir):
+            path = os.path.abspath(os.path.join(SCRIPT_DIR, "..", "profiles"))
+            if os.path.isdir(path):
+                profile_data_dir = path
+
+        with open(os.path.join(profile_data_dir, "profiles.json"), "r") as fh:
+            base_profiles = json.load(fh)[category]
+
+        # values to use when interpolating preferences
+        interpolation = {
+            "server": "%s:%s" % (options.webServer, options.httpPort),
+        }
+
+        for profile in base_profiles:
+            path = os.path.join(profile_data_dir, profile)
+            self.profile.merge(path, interpolation=interpolation)
+
+    @property
+    def conditioned_profile_copy(self):
+        """Returns a copy of the original conditioned profile that was created."""
+
+        condprof_copy = os.path.join(tempfile.mkdtemp(), "profile")
+        shutil.copytree(
+            self.conditioned_profile_dir,
+            condprof_copy,
+            ignore=shutil.ignore_patterns("lock"),
+        )
+        self.log.info("Created a conditioned-profile copy: %s" % condprof_copy)
+        return condprof_copy
+
+    def downloadConditionedProfile(self, profile_scenario, app):
+        from condprof.client import get_profile
+        from condprof.util import get_current_platform, get_version
+
+        if self.conditioned_profile_dir:
+            # We already have a directory, so provide a copy that
+            # will get deleted after it's done with
+            return self.conditioned_profile_copy
+
+        temp_download_dir = tempfile.mkdtemp()
+
+        # Call condprof's client API to yield our platform-specific
+        # conditioned-profile binary
+        platform = get_current_platform()
+
+        if not profile_scenario:
+            profile_scenario = "settled"
+
+        version = get_version(app)
+        try:
+            cond_prof_target_dir = get_profile(
+                temp_download_dir,
+                platform,
+                profile_scenario,
+                repo="mozilla-central",
+                version=version,
+                retries=2,  # quicker failure
+            )
+        except Exception:
+            if version is None:
+                # any other error is a showstopper
+                self.log.critical("Could not get the conditioned profile")
+                traceback.print_exc()
+                raise
+            version = None
+            try:
+                self.log.info("retrying a profile with no version specified")
+                cond_prof_target_dir = get_profile(
+                    temp_download_dir,
+                    platform,
+                    profile_scenario,
+                    repo="mozilla-central",
+                    version=version,
+                )
+            except Exception:
+                self.log.critical("Could not get the conditioned profile")
+                traceback.print_exc()
+                raise
+
+        # Now get the full directory path to our fetched conditioned profile
+        self.conditioned_profile_dir = os.path.join(
+            temp_download_dir, cond_prof_target_dir
+        )
+        if not os.path.exists(cond_prof_target_dir):
+            self.log.critical(
+                "Can't find target_dir {}, from get_profile()"
+                "temp_download_dir {}, platform {}, scenario {}".format(
+                    cond_prof_target_dir, temp_download_dir, platform, profile_scenario
+                )
+            )
+            raise OSError
+
+        self.log.info(
+            "Original self.conditioned_profile_dir is now set: {}".format(
+                self.conditioned_profile_dir
+            )
+        )
+        return self.conditioned_profile_copy
+
+    def buildProfile(self, options):
+        """create the profile and add optional chrome bits and files if requested"""
+        # get extensions to install
+        extensions = self.getExtensionsToInstall(options)
+
+        # Whitelist the _tests directory (../..) so that TESTING_JS_MODULES work
+        tests_dir = os.path.dirname(os.path.dirname(SCRIPT_DIR))
+        sandbox_whitelist_paths = [tests_dir] + options.sandboxReadWhitelist
+        if platform.system() == "Linux" or platform.system() in (
+            "Windows",
+            "Microsoft",
+        ):
+            # Trailing slashes are needed to indicate directories on Linux and Windows
+            sandbox_whitelist_paths = [
+                os.path.join(p, "") for p in sandbox_whitelist_paths
+            ]
+
+        if options.conditionedProfile:
+            if options.profilePath and os.path.exists(options.profilePath):
+                shutil.rmtree(options.profilePath, ignore_errors=True)
+            options.profilePath = self.downloadConditionedProfile("full", options.app)
+
+            # This is causing `certutil -N -d -f`` to not use -f (pwd file)
+            try:
+                os.remove(os.path.join(options.profilePath, "key4.db"))
+            except Exception as e:
+                self.log.info(
+                    "Caught exception while removing key4.db"
+                    "during setup of conditioned profile: %s" % e
+                )
+
+        # Create the profile
+        self.profile = Profile(
+            profile=options.profilePath,
+            addons=extensions,
+            locations=self.locations,
+            proxy=self.proxy(options),
+            whitelistpaths=sandbox_whitelist_paths,
+        )
+
+        # Fix options.profilePath for legacy consumers.
+        options.profilePath = self.profile.profile
+
+        manifest = self.addChromeToProfile(options)
+        self.copyExtraFilesToProfile(options)
+
+        # create certificate database for the profile
+        # TODO: this should really be upstreamed somewhere, maybe mozprofile
+        certificateStatus = self.fillCertificateDB(options)
+        if certificateStatus:
+            self.log.error(
+                "TEST-UNEXPECTED-FAIL | runtests.py | Certificate integration failed"
+            )
+            return None
+
+        # Set preferences in the following order (latter overrides former):
+        # 1) Preferences from base profile (e.g from testing/profiles)
+        # 2) Prefs hardcoded in this function
+        # 3) Prefs from --setpref
+
+        # Prefs from base profiles
+        self.merge_base_profiles(options, "mochitest")
+
+        # Hardcoded prefs (TODO move these into a base profile)
+        prefs = {
+            "browser.tabs.remote.autostart": options.e10s,
+            # Enable tracing output for detailed failures in case of
+            # failing connection attempts, and hangs (bug 1397201)
+            "remote.log.level": "Trace",
+            # Disable async font fallback, because the unpredictable
+            # extra reflow it can trigger (potentially affecting a later
+            # test) results in spurious intermittent failures.
+            "gfx.font_rendering.fallback.async": False,
+        }
+
+        test_timeout = None
+        if options.flavor == "browser" and options.timeout:
+            test_timeout = options.timeout
+
+        # browser-chrome tests use a fairly short default timeout of 45 seconds;
+        # this is sometimes too short on asan and debug, where we expect reduced
+        # performance.
+        if (
+            (mozinfo.info["asan"] or mozinfo.info["debug"])
+            and options.flavor == "browser"
+            and options.timeout is None
+        ):
+            self.log.info("Increasing default timeout to 90 seconds (asan or debug)")
+            test_timeout = 90
+
+        # tsan builds need even more time
+        if (
+            mozinfo.info["tsan"]
+            and options.flavor == "browser"
+            and options.timeout is None
+        ):
+            self.log.info("Increasing default timeout to 120 seconds (tsan)")
+            test_timeout = 120
+
+        if mozinfo.info["os"] == "win" and mozinfo.info["processor"] == "aarch64":
+            test_timeout = self.DEFAULT_TIMEOUT * 4
+            self.log.info(
+                "Increasing default timeout to {} seconds (win aarch64)".format(
+                    test_timeout
+                )
+            )
+
+        if "MOZ_CHAOSMODE=0xfb" in options.environment and test_timeout:
+            test_timeout *= 2
+            self.log.info(
+                "Increasing default timeout to {} seconds (MOZ_CHAOSMODE)".format(
+                    test_timeout
+                )
+            )
+
+        if test_timeout:
+            prefs["testing.browserTestHarness.timeout"] = test_timeout
+
+        if getattr(self, "testRootAbs", None):
+            prefs["mochitest.testRoot"] = self.testRootAbs
+
+        # See if we should use fake media devices.
+        if options.useTestMediaDevices:
+            prefs["media.audio_loopback_dev"] = self.mediaDevices["audio"]["name"]
+            prefs["media.video_loopback_dev"] = self.mediaDevices["video"]["name"]
+            prefs["media.cubeb.output_device"] = "Null Output"
+            prefs["media.volume_scale"] = "1.0"
+            self.gstForV4l2loopbackProcess = self.mediaDevices["video"]["process"]
+
+        self.profile.set_preferences(prefs)
+
+        # Extra prefs from --setpref
+        self.profile.set_preferences(self.extraPrefs)
+        return manifest
+
+    def getGMPPluginPath(self, options):
+        if options.gmp_path:
+            return options.gmp_path
+
+        gmp_parentdirs = [
+            # For local builds, GMP plugins will be under dist/bin.
+            options.xrePath,
+            # For packaged builds, GMP plugins will get copied under
+            # $profile/plugins.
+            os.path.join(self.profile.profile, "plugins"),
+        ]
+
+        gmp_subdirs = [
+            os.path.join("gmp-fake", "1.0"),
+            os.path.join("gmp-fakeopenh264", "1.0"),
+            os.path.join("gmp-clearkey", "0.1"),
+        ]
+
+        gmp_paths = [
+            os.path.join(parent, sub)
+            for parent in gmp_parentdirs
+            for sub in gmp_subdirs
+            if os.path.isdir(os.path.join(parent, sub))
+        ]
+
+        if not gmp_paths:
+            # This is fatal for desktop environments.
+            raise EnvironmentError("Could not find test gmp plugins")
+
+        return os.pathsep.join(gmp_paths)
+
+    def cleanup(self, options, final=False):
+        """remove temporary files, profile and virtual audio input device"""
+        if hasattr(self, "manifest") and self.manifest is not None:
+            if os.path.exists(self.manifest):
+                os.remove(self.manifest)
+        if hasattr(self, "profile"):
+            del self.profile
+        if hasattr(self, "extraTestsDirs"):
+            for d in self.extraTestsDirs:
+                if os.path.exists(d):
+                    shutil.rmtree(d)
+        if options.pidFile != "" and os.path.exists(options.pidFile):
+            try:
+                os.remove(options.pidFile)
+                if os.path.exists(options.pidFile + ".xpcshell.pid"):
+                    os.remove(options.pidFile + ".xpcshell.pid")
+            except Exception:
+                self.log.warning(
+                    "cleaning up pidfile '%s' was unsuccessful from the test harness"
+                    % options.pidFile
+                )
+        options.manifestFile = None
+
+        if hasattr(self, "virtualInputDeviceIdList"):
+            pactl = spawn.find_executable("pactl")
+
+            if not pactl:
+                self.log.error("Could not find pactl on system")
+                return None
+
+            for id in self.virtualInputDeviceIdList:
+                try:
+                    subprocess.check_call([pactl, "unload-module", str(id)])
+                except subprocess.CalledProcessError:
+                    self.log.error(
+                        "Could not remove pulse module with id {}".format(id)
+                    )
+                    return None
+
+            self.virtualInputDeviceIdList = []
+
+    def dumpScreen(self, utilityPath):
+        if self.haveDumpedScreen:
+            self.log.info(
+                "Not taking screenshot here: see the one that was previously logged"
+            )
+            return
+        self.haveDumpedScreen = True
+        dump_screen(utilityPath, self.log)
+
+    def killAndGetStack(self, processPID, utilityPath, debuggerInfo, dump_screen=False):
+        """
+        Kill the process, preferrably in a way that gets us a stack trace.
+        Also attempts to obtain a screenshot before killing the process
+        if specified.
+        """
+        self.log.info("Killing process: %s" % processPID)
+        if dump_screen:
+            self.dumpScreen(utilityPath)
+
+        if mozinfo.info.get("crashreporter", True) and not debuggerInfo:
+            try:
+                minidump_path = os.path.join(self.profile.profile, "minidumps")
+                mozcrash.kill_and_get_minidump(processPID, minidump_path, utilityPath)
+            except OSError:
+                # https://bugzilla.mozilla.org/show_bug.cgi?id=921509
+                self.log.info("Can't trigger Breakpad, process no longer exists")
+            return
+        self.log.info("Can't trigger Breakpad, just killing process")
+        killPid(processPID, self.log)
+
+    def extract_child_pids(self, process_log, parent_pid=None):
+        """Parses the given log file for the pids of any processes launched by
+        the main process and returns them as a list.
+        If parent_pid is provided, and psutil is available, returns children of
+        parent_pid according to psutil.
+        """
+        rv = []
+        if parent_pid and HAVE_PSUTIL:
+            self.log.info("Determining child pids from psutil...")
+            try:
+                rv = [p.pid for p in psutil.Process(parent_pid).children()]
+                self.log.info(str(rv))
+            except psutil.NoSuchProcess:
+                self.log.warning("Failed to lookup children of pid %d" % parent_pid)
+
+        rv = set(rv)
+        pid_re = re.compile(r"==> process \d+ launched child process (\d+)")
+        with open(process_log) as fd:
+            for line in fd:
+                self.log.info(line.rstrip())
+                m = pid_re.search(line)
+                if m:
+                    rv.add(int(m.group(1)))
+        return rv
+
+    def checkForZombies(self, processLog, utilityPath, debuggerInfo):
+        """Look for hung processes"""
+
+        if not os.path.exists(processLog):
+            self.log.info("Automation Error: PID log not found: %s" % processLog)
+            # Whilst no hung process was found, the run should still display as
+            # a failure
+            return True
+
+        # scan processLog for zombies
+        self.log.info("zombiecheck | Reading PID log: %s" % processLog)
+        processList = self.extract_child_pids(processLog)
+        # kill zombies
+        foundZombie = False
+        for processPID in processList:
+            self.log.info(
+                "zombiecheck | Checking for orphan process with PID: %d" % processPID
+            )
+            if isPidAlive(processPID):
+                foundZombie = True
+                self.log.error(
+                    "TEST-UNEXPECTED-FAIL | zombiecheck | child process "
+                    "%d still alive after shutdown" % processPID
+                )
+                self.killAndGetStack(
+                    processPID, utilityPath, debuggerInfo, dump_screen=not debuggerInfo
+                )
+
+        return foundZombie
+
+    def checkForRunningBrowsers(self):
+        firefoxes = ""
+        if HAVE_PSUTIL:
+            attrs = ["pid", "ppid", "name", "cmdline", "username"]
+            for proc in psutil.process_iter():
+                try:
+                    if "firefox" in proc.name():
+                        firefoxes = "%s%s\n" % (firefoxes, proc.as_dict(attrs=attrs))
+                except Exception:
+                    # may not be able to access process info for all processes
+                    continue
+        if len(firefoxes) > 0:
+            # In automation, this warning is unexpected and should be investigated.
+            # In local testing, this is probably okay, as long as the browser is not
+            # running a marionette server.
+            self.log.warning("Found 'firefox' running before starting test browser!")
+            self.log.warning(firefoxes)
+
+    def runApp(
+        self,
+        testUrl,
+        env,
+        app,
+        profile,
+        extraArgs,
+        utilityPath,
+        debuggerInfo=None,
+        valgrindPath=None,
+        valgrindArgs=None,
+        valgrindSuppFiles=None,
+        symbolsPath=None,
+        timeout=-1,
+        detectShutdownLeaks=False,
+        screenshotOnFail=False,
+        bisectChunk=None,
+        marionette_args=None,
+        e10s=True,
+        runFailures=False,
+        crashAsPass=False,
+        currentManifest=None,
+    ):
+        """
+        Run the app, log the duration it took to execute, return the status code.
+        Kills the app if it runs for longer than |maxTime| seconds, or outputs nothing
+        for |timeout| seconds.
+        """
+        # It can't be the case that both a with-debugger and an
+        # on-Valgrind run have been requested.  doTests() should have
+        # already excluded this possibility.
+        assert not (valgrindPath and debuggerInfo)
+
+        # debugger information
+        interactive = False
+        debug_args = None
+        if debuggerInfo:
+            interactive = debuggerInfo.interactive
+            debug_args = [debuggerInfo.path] + debuggerInfo.args
+
+        # Set up Valgrind arguments.
+        if valgrindPath:
+            interactive = False
+            valgrindArgs_split = (
+                [] if valgrindArgs is None else shlex.split(valgrindArgs)
+            )
+
+            valgrindSuppFiles_final = []
+            if valgrindSuppFiles is not None:
+                valgrindSuppFiles_final = [
+                    "--suppressions=" + path for path in valgrindSuppFiles.split(",")
+                ]
+
+            debug_args = (
+                [valgrindPath]
+                + mozdebug.get_default_valgrind_args()
+                + valgrindArgs_split
+                + valgrindSuppFiles_final
+            )
+
+        # fix default timeout
+        if timeout == -1:
+            timeout = self.DEFAULT_TIMEOUT
+
+        # Note in the log if running on Valgrind
+        if valgrindPath:
+            self.log.info(
+                "runtests.py | Running on Valgrind.  "
+                + "Using timeout of %d seconds." % timeout
+            )
+
+        # copy env so we don't munge the caller's environment
+        env = env.copy()
+
+        # Used to defer a possible IOError exception from Marionette
+        marionette_exception = None
+
+        temp_file_paths = []
+
+        # make sure we clean up after ourselves.
+        try:
+            # set process log environment variable
+            tmpfd, processLog = tempfile.mkstemp(suffix="pidlog")
+            os.close(tmpfd)
+            env["MOZ_PROCESS_LOG"] = processLog
+
+            if debuggerInfo:
+                # If a debugger is attached, don't use timeouts, and don't
+                # capture ctrl-c.
+                timeout = None
+                signal.signal(signal.SIGINT, lambda sigid, frame: None)
+
+            # build command line
+            cmd = os.path.abspath(app)
+            args = list(extraArgs)
+            args.append("-marionette")
+            # TODO: mozrunner should use -foreground at least for mac
+            # https://bugzilla.mozilla.org/show_bug.cgi?id=916512
+            args.append("-foreground")
+            self.start_script_kwargs["testUrl"] = testUrl or "about:blank"
+
+            if detectShutdownLeaks:
+                env["MOZ_LOG"] = (
+                    env["MOZ_LOG"] + "," if env["MOZ_LOG"] else ""
+                ) + "DocShellAndDOMWindowLeak:3"
+                shutdownLeaks = ShutdownLeaks(self.log)
+            else:
+                shutdownLeaks = None
+
+            if mozinfo.info["asan"] and mozinfo.isLinux and mozinfo.bits == 64:
+                lsanLeaks = LSANLeaks(self.log)
+            else:
+                lsanLeaks = None
+
+            # create an instance to process the output
+            outputHandler = self.OutputHandler(
+                harness=self,
+                utilityPath=utilityPath,
+                symbolsPath=symbolsPath,
+                dump_screen_on_timeout=not debuggerInfo,
+                dump_screen_on_fail=screenshotOnFail,
+                shutdownLeaks=shutdownLeaks,
+                lsanLeaks=lsanLeaks,
+                bisectChunk=bisectChunk,
+            )
+
+            def timeoutHandler():
+                browserProcessId = outputHandler.browserProcessId
+                self.handleTimeout(
+                    timeout,
+                    proc,
+                    utilityPath,
+                    debuggerInfo,
+                    browserProcessId,
+                    processLog,
+                )
+
+            kp_kwargs = {
+                "kill_on_timeout": False,
+                "cwd": SCRIPT_DIR,
+                "onTimeout": [timeoutHandler],
+            }
+            kp_kwargs["processOutputLine"] = [outputHandler]
+
+            self.checkForRunningBrowsers()
+
+            # create mozrunner instance and start the system under test process
+            self.lastTestSeen = self.test_name
+            startTime = datetime.now()
+
+            runner_cls = mozrunner.runners.get(
+                mozinfo.info.get("appname", "firefox"), mozrunner.Runner
+            )
+            runner = runner_cls(
+                profile=self.profile,
+                binary=cmd,
+                cmdargs=args,
+                env=env,
+                process_class=mozprocess.ProcessHandlerMixin,
+                process_args=kp_kwargs,
+            )
+
+            # start the runner
+            try:
+                runner.start(
+                    debug_args=debug_args,
+                    interactive=interactive,
+                    outputTimeout=timeout,
+                )
+                proc = runner.process_handler
+                self.log.info("runtests.py | Application pid: %d" % proc.pid)
+
+                gecko_id = "GECKO(%d)" % proc.pid
+                self.log.process_start(gecko_id)
+                self.message_logger.gecko_id = gecko_id
+            except PermissionError:
+                # treat machine as bad, return
+                return TBPL_RETRY, "Failure to launch browser"
+            except Exception as e:
+                raise e  # unknown error
+
+            try:
+                # start marionette and kick off the tests
+                marionette_args = marionette_args or {}
+                self.marionette = Marionette(**marionette_args)
+                self.marionette.start_session()
+
+                # install specialpowers and mochikit addons
+                addons = Addons(self.marionette)
+
+                if self.staged_addons:
+                    for addon_path in self.staged_addons:
+                        if not os.path.isdir(addon_path):
+                            self.log.error(
+                                "TEST-UNEXPECTED-FAIL | invalid setup: missing extension at %s"
+                                % addon_path
+                            )
+                            return 1, self.lastTestSeen
+                        temp_addon_path = create_zip(addon_path)
+                        temp_file_paths.append(temp_addon_path)
+                        addons.install(temp_addon_path)
+
+                self.execute_start_script()
+
+                # an open marionette session interacts badly with mochitest,
+                # delete it until we figure out why.
+                self.marionette.delete_session()
+                del self.marionette
+
+            except IOError:
+                # Any IOError as thrown by Marionette means that something is
+                # wrong with the process, like a crash or the socket is no
+                # longer open. We defer raising this specific error so that
+                # post-test checks for leaks and crashes are performed and
+                # reported first.
+                marionette_exception = sys.exc_info()
+
+            # wait until app is finished
+            # XXX copy functionality from
+            # https://github.com/mozilla/mozbase/blob/master/mozrunner/mozrunner/runner.py#L61
+            # until bug 913970 is fixed regarding mozrunner `wait` not returning status
+            # see https://bugzilla.mozilla.org/show_bug.cgi?id=913970
+            self.log.info("runtests.py | Waiting for browser...")
+            status = proc.wait()
+            if status is None:
+                self.log.warning(
+                    "runtests.py | Failed to get app exit code - running/crashed?"
+                )
+                # must report an integer to process_exit()
+                status = 0
+            self.log.process_exit("Main app process", status)
+            runner.process_handler = None
+
+            # finalize output handler
+            outputHandler.finish()
+
+            # record post-test information
+            if status:
+                # no need to keep return code 137, 245, etc.
+                status = 1
+                self.message_logger.dump_buffered()
+                msg = "application terminated with exit code %s" % status
+                # self.message_logger.is_test_running indicates we need to send a test_end
+                if crashAsPass and self.message_logger.is_test_running:
+                    # this works for browser-chrome, mochitest-plain has status=0
+                    message = {
+                        "action": "test_end",
+                        "status": "CRASH",
+                        "expected": "CRASH",
+                        "thread": None,
+                        "pid": None,
+                        "source": "mochitest",
+                        "time": int(time.time()) * 1000,
+                        "test": self.lastTestSeen,
+                        "message": msg,
+                    }
+                    # need to send a test_end in order to have mozharness process messages properly
+                    # this requires a custom message vs log.error/log.warning/etc.
+                    self.message_logger.process_message(message)
+            else:
+                self.lastTestSeen = (
+                    currentManifest or "Main app process exited normally"
+                )
+
+            self.log.info(
+                "runtests.py | Application ran for: %s"
+                % str(datetime.now() - startTime)
+            )
+
+            # Do a final check for zombie child processes.
+            zombieProcesses = self.checkForZombies(
+                processLog, utilityPath, debuggerInfo
+            )
+
+            # check for crashes
+            quiet = False
+            if crashAsPass:
+                quiet = True
+
+            minidump_path = os.path.join(self.profile.profile, "minidumps")
+            crash_count = mozcrash.log_crashes(
+                self.log,
+                minidump_path,
+                symbolsPath,
+                test=self.lastTestSeen,
+                quiet=quiet,
+            )
+
+            expected = None
+            if crashAsPass or crash_count > 0:
+                # self.message_logger.is_test_running indicates we need a test_end message
+                if self.message_logger.is_test_running:
+                    # this works for browser-chrome, mochitest-plain has status=0
+                    expected = "CRASH"
+                if crashAsPass:
+                    status = 0
+            elif crash_count or zombieProcesses:
+                if self.message_logger.is_test_running:
+                    expected = "PASS"
+                status = 1
+
+            if expected:
+                # send this out so we always wrap up the test-end message
+                message = {
+                    "action": "test_end",
+                    "status": "CRASH",
+                    "expected": expected,
+                    "thread": None,
+                    "pid": None,
+                    "source": "mochitest",
+                    "time": int(time.time()) * 1000,
+                    "test": self.lastTestSeen,
+                    "message": "application terminated with exit code %s" % status,
+                }
+                # need to send a test_end in order to have mozharness process messages properly
+                # this requires a custom message vs log.error/log.warning/etc.
+                self.message_logger.process_message(message)
+        finally:
+            # cleanup
+            if os.path.exists(processLog):
+                os.remove(processLog)
+            for p in temp_file_paths:
+                os.remove(p)
+
+        if marionette_exception is not None:
+            exc, value, tb = marionette_exception
+            six.reraise(exc, value, tb)
+
+        return status, self.lastTestSeen
+
+    def initializeLooping(self, options):
+        """
+        This method is used to clear the contents before each run of for loop.
+        This method is used for --run-by-dir and --bisect-chunk.
+        """
+        if options.conditionedProfile:
+            if options.profilePath and os.path.exists(options.profilePath):
+                shutil.rmtree(options.profilePath, ignore_errors=True)
+                if options.manifestFile and os.path.exists(options.manifestFile):
+                    os.remove(options.manifestFile)
+
+        self.expectedError.clear()
+        self.result.clear()
+        options.manifestFile = None
+        options.profilePath = None
+
+    def initializeVirtualInputDevices(self):
+        """
+        Configure the system to have a number of virtual audio input devices, that
+        each produce a tone at a particular frequency.
+
+        This method is only currently implemented for Linux.
+        """
+        if not mozinfo.isLinux:
+            return
+
+        pactl = spawn.find_executable("pactl")
+
+        if not pactl:
+            self.log.error("Could not find pactl on system")
+            return
+
+        DEVICES_COUNT = 4
+        DEVICES_BASE_FREQUENCY = 110  # Hz
+        self.virtualInputDeviceIdList = []
+        # If the device are already present, find their id and return early
+        o = subprocess.check_output([pactl, "list", "modules", "short"])
+        found_devices = 0
+        for input in o.splitlines():
+            device = input.decode().split("\t")
+            if device[1] == "module-sine-source":
+                self.virtualInputDeviceIdList.append(int(device[0]))
+                found_devices += 1
+
+        if found_devices == DEVICES_COUNT:
+            return
+        elif found_devices != 0:
+            # Remove all devices and reinitialize them properly
+            for id in self.virtualInputDeviceIdList:
+                try:
+                    subprocess.check_call([pactl, "unload-module", str(id)])
+                except subprocess.CalledProcessError:
+                    log.error("Could not remove pulse module with id {}".format(id))
+                    return None
+
+        # We want quite a number of input devices, each with a different tone
+        # frequency and device name so that we can recognize them easily during
+        # testing.
+        command = [pactl, "load-module", "module-sine-source", "rate=44100"]
+        for i in range(1, DEVICES_COUNT + 1):
+            freq = i * DEVICES_BASE_FREQUENCY
+            complete_command = command + [
+                "source_name=sine-{}".format(freq),
+                "frequency={}".format(freq),
+            ]
+            try:
+                o = subprocess.check_output(complete_command)
+                self.virtualInputDeviceIdList.append(o)
+
+            except subprocess.CalledProcessError:
+                self.log.error(
+                    "Could not create device with module-sine-source"
+                    " (freq={})".format(freq)
+                )
+
+    def normalize_paths(self, paths):
+        # Normalize test paths so they are relative to test root
+        norm_paths = []
+        for p in paths:
+            abspath = os.path.abspath(os.path.join(self.oldcwd, p))
+            if abspath.startswith(self.testRootAbs):
+                norm_paths.append(os.path.relpath(abspath, self.testRootAbs))
+            else:
+                norm_paths.append(p)
+        return norm_paths
+
+    def runMochitests(self, options, testsToRun, manifestToFilter=None):
+        "This is a base method for calling other methods in this class for --bisect-chunk."
+        # Making an instance of bisect class for --bisect-chunk option.
+        bisect = bisection.Bisect(self)
+        finished = False
+        status = 0
+        bisection_log = 0
+        while not finished:
+            if options.bisectChunk:
+                testsToRun = bisect.pre_test(options, testsToRun, status)
+                # To inform that we are in the process of bisection, and to
+                # look for bleedthrough
+                if options.bisectChunk != "default" and not bisection_log:
+                    self.log.error(
+                        "TEST-UNEXPECTED-FAIL | Bisection | Please ignore repeats "
+                        "and look for 'Bleedthrough' (if any) at the end of "
+                        "the failure list"
+                    )
+                    bisection_log = 1
+
+            result = self.doTests(options, testsToRun, manifestToFilter)
+            if result == TBPL_RETRY:  # terminate task
+                return result
+
+            if options.bisectChunk:
+                status = bisect.post_test(options, self.expectedError, self.result)
+            else:
+                status = -1
+
+            if status == -1:
+                finished = True
+
+        # We need to print the summary only if options.bisectChunk has a value.
+        # Also we need to make sure that we do not print the summary in between
+        # running tests via --run-by-dir.
+        if options.bisectChunk and options.bisectChunk in self.result:
+            bisect.print_summary()
+
+        return result
+
+    def groupTestsByScheme(self, tests):
+        """
+        split tests into groups by schemes. test is classified as http if
+        no scheme specified
+        """
+        httpTests = []
+        httpsTests = []
+        for test in tests:
+            if not test.get("scheme") or test.get("scheme") == "http":
+                httpTests.append(test)
+            elif test.get("scheme") == "https":
+                httpsTests.append(test)
+        return {"http": httpTests, "https": httpsTests}
+
+    def verifyTests(self, options):
+        """
+        Support --verify mode: Run test(s) many times in a variety of
+        configurations/environments in an effort to find intermittent
+        failures.
+        """
+
+        # Number of times to repeat test(s) when running with --repeat
+        VERIFY_REPEAT = 10
+        # Number of times to repeat test(s) when running test in
+        VERIFY_REPEAT_SINGLE_BROWSER = 5
+
+        def step1():
+            options.repeat = VERIFY_REPEAT
+            options.keep_open = False
+            options.runUntilFailure = True
+            options.profilePath = None
+            options.comparePrefs = True
+            result = self.runTests(options)
+            result = result or (-2 if self.countfail > 0 else 0)
+            self.message_logger.finish()
+            return result
+
+        def step2():
+            options.repeat = 0
+            options.keep_open = False
+            options.runUntilFailure = False
+            for i in range(VERIFY_REPEAT_SINGLE_BROWSER):
+                options.profilePath = None
+                result = self.runTests(options)
+                result = result or (-2 if self.countfail > 0 else 0)
+                self.message_logger.finish()
+                if result != 0:
+                    break
+            return result
+
+        def step3():
+            options.repeat = VERIFY_REPEAT
+            options.keep_open = False
+            options.runUntilFailure = True
+            options.environment.append("MOZ_CHAOSMODE=0xfb")
+            options.profilePath = None
+            result = self.runTests(options)
+            options.environment.remove("MOZ_CHAOSMODE=0xfb")
+            result = result or (-2 if self.countfail > 0 else 0)
+            self.message_logger.finish()
+            return result
+
+        def step4():
+            options.repeat = 0
+            options.keep_open = False
+            options.runUntilFailure = False
+            options.environment.append("MOZ_CHAOSMODE=0xfb")
+            for i in range(VERIFY_REPEAT_SINGLE_BROWSER):
+                options.profilePath = None
+                result = self.runTests(options)
+                result = result or (-2 if self.countfail > 0 else 0)
+                self.message_logger.finish()
+                if result != 0:
+                    break
+            options.environment.remove("MOZ_CHAOSMODE=0xfb")
+            return result
+
+        def fission_step(fission_pref):
+            if fission_pref not in options.extraPrefs:
+                options.extraPrefs.append(fission_pref)
+            options.keep_open = False
+            options.runUntilFailure = True
+            options.profilePath = None
+            result = self.runTests(options)
+            result = result or (-2 if self.countfail > 0 else 0)
+            self.message_logger.finish()
+            return result
+
+        def fission_step1():
+            return fission_step("fission.autostart=false")
+
+        def fission_step2():
+            return fission_step("fission.autostart=true")
+
+        if options.verify_fission:
+            steps = [
+                ("1. Run each test without fission.", fission_step1),
+                ("2. Run each test with fission.", fission_step2),
+            ]
+        else:
+            steps = [
+                ("1. Run each test %d times in one browser." % VERIFY_REPEAT, step1),
+                (
+                    "2. Run each test %d times in a new browser each time."
+                    % VERIFY_REPEAT_SINGLE_BROWSER,
+                    step2,
+                ),
+                (
+                    "3. Run each test %d times in one browser, in chaos mode."
+                    % VERIFY_REPEAT,
+                    step3,
+                ),
+                (
+                    "4. Run each test %d times in a new browser each time, "
+                    "in chaos mode." % VERIFY_REPEAT_SINGLE_BROWSER,
+                    step4,
+                ),
+            ]
+
+        stepResults = {}
+        for (descr, step) in steps:
+            stepResults[descr] = "not run / incomplete"
+
+        startTime = datetime.now()
+        maxTime = timedelta(seconds=options.verify_max_time)
+        finalResult = "PASSED"
+        for (descr, step) in steps:
+            if (datetime.now() - startTime) > maxTime:
+                self.log.info("::: Test verification is taking too long: Giving up!")
+                self.log.info(
+                    "::: So far, all checks passed, but not all checks were run."
+                )
+                break
+            self.log.info(":::")
+            self.log.info('::: Running test verification step "%s"...' % descr)
+            self.log.info(":::")
+            result = step()
+            if result != 0:
+                stepResults[descr] = "FAIL"
+                finalResult = "FAILED!"
+                break
+            stepResults[descr] = "Pass"
+
+        self.logPreamble([])
+
+        self.log.info(":::")
+        self.log.info("::: Test verification summary for:")
+        self.log.info(":::")
+        tests = self.getActiveTests(options)
+        for test in tests:
+            self.log.info("::: " + test["path"])
+        self.log.info(":::")
+        for descr in sorted(stepResults.keys()):
+            self.log.info("::: %s : %s" % (descr, stepResults[descr]))
+        self.log.info(":::")
+        self.log.info("::: Test verification %s" % finalResult)
+        self.log.info(":::")
+
+        return 0
+
+    def runTests(self, options):
+        """Prepare, configure, run tests and cleanup"""
+        self.extraPrefs = parse_preferences(options.extraPrefs)
+        self.extraPrefs["fission.autostart"] = not options.disable_fission
+
+        # for test manifest parsing.
+        mozinfo.update(
+            {
+                "a11y_checks": options.a11y_checks,
+                "e10s": options.e10s,
+                "fission": not options.disable_fission,
+                "headless": options.headless,
+                "http3": options.useHttp3Server,
+                # Until the test harness can understand default pref values,
+                # (https://bugzilla.mozilla.org/show_bug.cgi?id=1577912) this value
+                # should by synchronized with the default pref value indicated in
+                # StaticPrefList.yaml.
+                #
+                # Currently for automation, the pref defaults to true (but can be
+                # overridden with --setpref).
+                "serviceworker_e10s": True,
+                "sessionHistoryInParent": not options.disable_fission
+                or not self.extraPrefs.get(
+                    "fission.disableSessionHistoryInParent",
+                    mozinfo.info["os"] == "android",
+                ),
+                "socketprocess_e10s": self.extraPrefs.get(
+                    "network.process.enabled", False
+                ),
+                "socketprocess_networking": self.extraPrefs.get(
+                    "network.http.network_access_on_socket_process.enabled", False
+                ),
+                "swgl": self.extraPrefs.get("gfx.webrender.software", False),
+                "verify": options.verify,
+                "verify_fission": options.verify_fission,
+                "webgl_ipc": self.extraPrefs.get("webgl.out-of-process", False),
+                "wmfme": (
+                    self.extraPrefs.get("media.wmf.media-engine.enabled", False)
+                    and self.extraPrefs.get(
+                        "media.wmf.media-engine.channel-decoder.enabled", False
+                    )
+                ),
+                "xorigin": options.xOriginTests,
+                "condprof": options.conditionedProfile,
+                "msix": "WindowsApps" in options.app,
+            }
+        )
+
+        if not self.mozinfo_variables_shown:
+            self.mozinfo_variables_shown = True
+            self.log.info(
+                "These variables are available in the mozinfo environment and "
+                "can be used to skip tests conditionally:"
+            )
+            for info in sorted(mozinfo.info.items(), key=lambda item: item[0]):
+                self.log.info("    {key}: {value}".format(key=info[0], value=info[1]))
+        self.setTestRoot(options)
+
+        # Despite our efforts to clean up servers started by this script, in practice
+        # we still see infrequent cases where a process is orphaned and interferes
+        # with future tests, typically because the old server is keeping the port in use.
+        # Try to avoid those failures by checking for and killing servers before
+        # trying to start new ones.
+        self.killNamedProc("ssltunnel")
+        self.killNamedProc("xpcshell")
+
+        if options.cleanupCrashes:
+            mozcrash.cleanup_pending_crash_reports()
+
+        tests = self.getActiveTests(options)
+        self.logPreamble(tests)
+
+        if mozinfo.info["fission"] and not mozinfo.info["e10s"]:
+            # Make sure this is logged *after* suite_start so it gets associated with the
+            # current suite in the summary formatters.
+            self.log.error("Fission is not supported without e10s.")
+            return 1
+
+        tests = [t for t in tests if "disabled" not in t]
+
+        # Until we have all green, this does not run on a11y (for perf reasons)
+        if not options.runByManifest:
+            result = self.runMochitests(options, [t["path"] for t in tests])
+            self.handleShutdownProfile(options)
+            return result
+
+        # code for --run-by-manifest
+        manifests = set(t["manifest"] for t in tests)
+        result = 0
+
+        origPrefs = self.extraPrefs.copy()
+        for m in sorted(manifests):
+            self.log.info("Running manifest: {}".format(m))
+
+            args = list(self.args_by_manifest[m])[0]
+            self.extraArgs = []
+            if args:
+                for arg in args.strip().split():
+                    # Split off the argument value if available so that both
+                    # name and value will be set individually
+                    self.extraArgs.extend(arg.split("="))
+
+                self.log.info(
+                    "The following arguments will be set:\n  {}".format(
+                        "\n  ".join(self.extraArgs)
+                    )
+                )
+
+            prefs = list(self.prefs_by_manifest[m])[0]
+            self.extraPrefs = origPrefs.copy()
+            if prefs:
+                prefs = prefs.strip().split()
+                self.log.info(
+                    "The following extra prefs will be set:\n  {}".format(
+                        "\n  ".join(prefs)
+                    )
+                )
+                self.extraPrefs.update(parse_preferences(prefs))
+
+            envVars = list(self.env_vars_by_manifest[m])[0]
+            self.extraEnv = {}
+            if envVars:
+                self.extraEnv = envVars.strip().split()
+                self.log.info(
+                    "The following extra environment variables will be set:\n  {}".format(
+                        "\n  ".join(self.extraEnv)
+                    )
+                )
+
+            self.parseAndCreateTestsDirs(m)
+
+            # If we are using --run-by-manifest, we should not use the profile path (if) provided
+            # by the user, since we need to create a new directory for each run. We would face
+            # problems if we use the directory provided by the user.
+            tests_in_manifest = [t["path"] for t in tests if t["manifest"] == m]
+            res = self.runMochitests(options, tests_in_manifest, manifestToFilter=m)
+            if res == TBPL_RETRY:  # terminate task
+                return res
+            result = result or res
+
+            # Dump the logging buffer
+            self.message_logger.dump_buffered()
+
+            if res == -1:
+                break
+
+        if self.manifest is not None:
+            self.cleanup(options, True)
+
+        e10s_mode = "e10s" if options.e10s else "non-e10s"
+
+        # for failure mode: where browser window has crashed and we have no reported results
+        if (
+            self.countpass == self.countfail == self.counttodo == 0
+            and options.crashAsPass
+        ):
+            self.countpass = 1
+            self.result = 0
+
+        # printing total number of tests
+        if options.flavor == "browser":
+            print("TEST-INFO | checking window state")
+            print("Browser Chrome Test Summary")
+            print("\tPassed: %s" % self.countpass)
+            print("\tFailed: %s" % self.countfail)
+            print("\tTodo: %s" % self.counttodo)
+            print("\tMode: %s" % e10s_mode)
+            print("*** End BrowserChrome Test Results ***")
+        else:
+            print("0 INFO TEST-START | Shutdown")
+            print("1 INFO Passed:  %s" % self.countpass)
+            print("2 INFO Failed:  %s" % self.countfail)
+            print("3 INFO Todo:    %s" % self.counttodo)
+            print("4 INFO Mode:    %s" % e10s_mode)
+            print("5 INFO SimpleTest FINISHED")
+
+        self.handleShutdownProfile(options)
+
+        if not result:
+            if self.countfail or not (self.countpass or self.counttodo):
+                # at least one test failed, or
+                # no tests passed, and no tests failed (possibly a crash)
+                result = 1
+
+        return result
+
+    def handleShutdownProfile(self, options):
+        # If shutdown profiling was enabled, then the user will want to access the
+        # performance profile. The following code will display helpful log messages
+        # and automatically open the profile if it is requested.
+        if self.browserEnv and "MOZ_PROFILER_SHUTDOWN" in self.browserEnv:
+            profile_path = self.browserEnv["MOZ_PROFILER_SHUTDOWN"]
+
+            profiler_logger = get_proxy_logger("profiler")
+            profiler_logger.info("Shutdown performance profiling was enabled")
+            profiler_logger.info("Profile saved locally to: %s" % profile_path)
+
+            if options.profilerSaveOnly or options.profiler:
+                # Only do the extra work of symbolicating and viewing the profile if
+                # officially requested through a command line flag. The MOZ_PROFILER_*
+                # flags can be set by a user.
+                symbolicate_profile_json(profile_path, options.topobjdir)
+                view_gecko_profile_from_mochitest(
+                    profile_path, options, profiler_logger
+                )
+            else:
+                profiler_logger.info(
+                    "The profiler was enabled outside of the mochitests. "
+                    "Use --profiler instead of MOZ_PROFILER_SHUTDOWN to "
+                    "symbolicate and open the profile automatically."
+                )
+
+            # Clean up the temporary file if it exists.
+            if self.profiler_tempdir:
+                shutil.rmtree(self.profiler_tempdir)
+
+    def doTests(self, options, testsToFilter=None, manifestToFilter=None):
+        # A call to initializeLooping method is required in case of --run-by-dir or --bisect-chunk
+        # since we need to initialize variables for each loop.
+        if options.bisectChunk or options.runByManifest:
+            self.initializeLooping(options)
+
+        # get debugger info, a dict of:
+        # {'path': path to the debugger (string),
+        #  'interactive': whether the debugger is interactive or not (bool)
+        #  'args': arguments to the debugger (list)
+        # TODO: use mozrunner.local.debugger_arguments:
+        # https://github.com/mozilla/mozbase/blob/master/mozrunner/mozrunner/local.py#L42
+
+        debuggerInfo = None
+        if options.debugger:
+            debuggerInfo = mozdebug.get_debugger_info(
+                options.debugger, options.debuggerArgs, options.debuggerInteractive
+            )
+
+        if options.useTestMediaDevices:
+            devices = findTestMediaDevices(self.log)
+            if not devices:
+                self.log.error("Could not find test media devices to use")
+                return 1
+            self.mediaDevices = devices
+            self.initializeVirtualInputDevices()
+
+        # See if we were asked to run on Valgrind
+        valgrindPath = None
+        valgrindArgs = None
+        valgrindSuppFiles = None
+        if options.valgrind:
+            valgrindPath = options.valgrind
+        if options.valgrindArgs:
+            valgrindArgs = options.valgrindArgs
+        if options.valgrindSuppFiles:
+            valgrindSuppFiles = options.valgrindSuppFiles
+
+        if (valgrindArgs or valgrindSuppFiles) and not valgrindPath:
+            self.log.error(
+                "Specified --valgrind-args or --valgrind-supp-files,"
+                " but not --valgrind"
+            )
+            return 1
+
+        if valgrindPath and debuggerInfo:
+            self.log.error("Can't use both --debugger and --valgrind together")
+            return 1
+
+        if valgrindPath and not valgrindSuppFiles:
+            valgrindSuppFiles = ",".join(get_default_valgrind_suppression_files())
+
+        # buildProfile sets self.profile .
+        # This relies on sideeffects and isn't very stateful:
+        # https://bugzilla.mozilla.org/show_bug.cgi?id=919300
+        self.manifest = self.buildProfile(options)
+        if self.manifest is None:
+            return 1
+
+        self.leak_report_file = os.path.join(options.profilePath, "runtests_leaks.log")
+
+        self.browserEnv = self.buildBrowserEnv(options, debuggerInfo is not None)
+
+        if self.browserEnv is None:
+            return 1
+
+        if self.mozLogs:
+            self.browserEnv["MOZ_LOG_FILE"] = "{}/moz-pid=%PID-uid={}.log".format(
+                self.browserEnv["MOZ_UPLOAD_DIR"], str(uuid.uuid4())
+            )
+
+        status = 0
+        try:
+            self.startServers(options, debuggerInfo)
+
+            if options.jsconsole:
+                options.browserArgs.extend(["--jsconsole"])
+
+            if options.jsdebugger:
+                options.browserArgs.extend(["-wait-for-jsdebugger", "-jsdebugger"])
+
+            # -jsdebugger takes a binary path as an optional argument.
+            # Append jsdebuggerPath right after `-jsdebugger`.
+            if options.jsdebuggerPath:
+                options.browserArgs.extend([options.jsdebuggerPath])
+
+            # Remove the leak detection file so it can't "leak" to the tests run.
+            # The file is not there if leak logging was not enabled in the
+            # application build.
+            if os.path.exists(self.leak_report_file):
+                os.remove(self.leak_report_file)
+
+            # then again to actually run mochitest
+            if options.timeout:
+                timeout = options.timeout + 30
+            elif options.debugger or not options.autorun:
+                timeout = None
+            else:
+                # We generally want the JS harness or marionette to handle
+                # timeouts if they can.
+                # The default JS harness timeout is currently 300 seconds.
+                # The default Marionette socket timeout is currently 360 seconds.
+                # Wait a little (10 seconds) more before timing out here.
+                # See bug 479518 and bug 1414063.
+                timeout = 370.0
+
+            if "MOZ_CHAOSMODE=0xfb" in options.environment and timeout:
+                timeout *= 2
+
+            # Detect shutdown leaks for m-bc runs if
+            # code coverage is not enabled.
+            detectShutdownLeaks = False
+            if options.jscov_dir_prefix is None:
+                detectShutdownLeaks = (
+                    mozinfo.info["debug"]
+                    and options.flavor == "browser"
+                    and options.subsuite != "thunderbird"
+                    and not options.crashAsPass
+                )
+
+            self.start_script_kwargs["flavor"] = self.normflavor(options.flavor)
+            marionette_args = {
+                "symbols_path": options.symbolsPath,
+                "socket_timeout": options.marionette_socket_timeout,
+                "startup_timeout": options.marionette_startup_timeout,
+            }
+
+            if options.marionette:
+                host, port = options.marionette.split(":")
+                marionette_args["host"] = host
+                marionette_args["port"] = int(port)
+
+            # testsToFilter parameter is used to filter out the test list that
+            # is sent to getTestsByScheme
+            for (scheme, tests) in self.getTestsByScheme(
+                options, testsToFilter, True, manifestToFilter
+            ):
+                # read the number of tests here, if we are not going to run any,
+                # terminate early
+                if not tests:
+                    continue
+
+                self.currentTests = [t["path"] for t in tests]
+                testURL = self.buildTestURL(options, scheme=scheme)
+
+                self.buildURLOptions(options, self.browserEnv)
+                if self.urlOpts:
+                    testURL += "?" + "&".join(self.urlOpts)
+
+                if options.runFailures:
+                    testURL += "&runFailures=true"
+
+                if options.timeoutAsPass:
+                    testURL += "&timeoutAsPass=true"
+
+                if options.conditionedProfile:
+                    testURL += "&conditionedProfile=true"
+
+                self.log.info("runtests.py | Running with scheme: {}".format(scheme))
+                self.log.info(
+                    "runtests.py | Running with e10s: {}".format(options.e10s)
+                )
+                self.log.info(
+                    "runtests.py | Running with fission: {}".format(
+                        mozinfo.info.get("fission", True)
+                    )
+                )
+                self.log.info(
+                    "runtests.py | Running with cross-origin iframes: {}".format(
+                        mozinfo.info.get("xorigin", False)
+                    )
+                )
+                self.log.info(
+                    "runtests.py | Running with serviceworker_e10s: {}".format(
+                        mozinfo.info.get("serviceworker_e10s", False)
+                    )
+                )
+                self.log.info(
+                    "runtests.py | Running with socketprocess_e10s: {}".format(
+                        mozinfo.info.get("socketprocess_e10s", False)
+                    )
+                )
+                self.log.info("runtests.py | Running tests: start.\n")
+                ret, _ = self.runApp(
+                    testURL,
+                    self.browserEnv,
+                    options.app,
+                    profile=self.profile,
+                    extraArgs=options.browserArgs + self.extraArgs,
+                    utilityPath=options.utilityPath,
+                    debuggerInfo=debuggerInfo,
+                    valgrindPath=valgrindPath,
+                    valgrindArgs=valgrindArgs,
+                    valgrindSuppFiles=valgrindSuppFiles,
+                    symbolsPath=options.symbolsPath,
+                    timeout=timeout,
+                    detectShutdownLeaks=detectShutdownLeaks,
+                    screenshotOnFail=options.screenshotOnFail,
+                    bisectChunk=options.bisectChunk,
+                    marionette_args=marionette_args,
+                    e10s=options.e10s,
+                    runFailures=options.runFailures,
+                    crashAsPass=options.crashAsPass,
+                    currentManifest=manifestToFilter,
+                )
+                status = ret or status
+        except KeyboardInterrupt:
+            self.log.info("runtests.py | Received keyboard interrupt.\n")
+            status = -1
+        except Exception as e:
+            traceback.print_exc()
+            self.log.error(
+                "Automation Error: Received unexpected exception while running application\n"
+            )
+            if "ADBTimeoutError" in repr(e):
+                self.log.info("runtests.py | Device disconnected. Aborting test.\n")
+                raise
+            status = 1
+        finally:
+            self.stopServers()
+
+        ignoreMissingLeaks = options.ignoreMissingLeaks
+        leakThresholds = options.leakThresholds
+
+        if options.crashAsPass:
+            ignoreMissingLeaks.append("tab")
+            ignoreMissingLeaks.append("socket")
+
+        # Provide a floor for Windows chrome leak detection, because we know
+        # we have some Windows-specific shutdown hangs that we avoid by timing
+        # out and leaking memory.
+        if options.flavor == "chrome" and mozinfo.isWin:
+            leakThresholds["default"] += 1296
+
+        # Stop leak detection if m-bc code coverage is enabled
+        # by maxing out the leak threshold for all processes.
+        if options.jscov_dir_prefix:
+            for processType in leakThresholds:
+                ignoreMissingLeaks.append(processType)
+                leakThresholds[processType] = sys.maxsize
+
+        utilityPath = options.utilityPath or options.xrePath
+        if status == 0:
+            # ignore leak checks for crashes
+            mozleak.process_leak_log(
+                self.leak_report_file,
+                leak_thresholds=leakThresholds,
+                ignore_missing_leaks=ignoreMissingLeaks,
+                log=self.log,
+                stack_fixer=get_stack_fixer_function(utilityPath, options.symbolsPath),
+            )
+
+        self.log.info("runtests.py | Running tests: end.")
+
+        if self.manifest is not None:
+            self.cleanup(options, False)
+
+        return status
+
+    def handleTimeout(
+        self, timeout, proc, utilityPath, debuggerInfo, browser_pid, processLog
+    ):
+        """handle process output timeout"""
+        # TODO: bug 913975 : _processOutput should call self.processOutputLine
+        # one more time one timeout (I think)
+        error_message = (
+            "TEST-UNEXPECTED-TIMEOUT | %s | application timed out after "
+            "%d seconds with no output"
+        ) % (self.lastTestSeen, int(timeout))
+        self.message_logger.dump_buffered()
+        self.message_logger.buffering = False
+        self.log.info(error_message)
+        self.log.warning("Force-terminating active process(es).")
+
+        browser_pid = browser_pid or proc.pid
+        child_pids = self.extract_child_pids(processLog, browser_pid)
+        self.log.info("Found child pids: %s" % child_pids)
+
+        if HAVE_PSUTIL:
+            try:
+                browser_proc = [psutil.Process(browser_pid)]
+            except Exception:
+                self.log.info("Failed to get proc for pid %d" % browser_pid)
+                browser_proc = []
+            try:
+                child_procs = [psutil.Process(pid) for pid in child_pids]
+            except Exception:
+                self.log.info("Failed to get child procs")
+                child_procs = []
+            for pid in child_pids:
+                self.killAndGetStack(
+                    pid, utilityPath, debuggerInfo, dump_screen=not debuggerInfo
+                )
+            gone, alive = psutil.wait_procs(child_procs, timeout=30)
+            for p in gone:
+                self.log.info("psutil found pid %s dead" % p.pid)
+            for p in alive:
+                self.log.warning("failed to kill pid %d after 30s" % p.pid)
+            self.killAndGetStack(
+                browser_pid, utilityPath, debuggerInfo, dump_screen=not debuggerInfo
+            )
+            gone, alive = psutil.wait_procs(browser_proc, timeout=30)
+            for p in gone:
+                self.log.info("psutil found pid %s dead" % p.pid)
+            for p in alive:
+                self.log.warning("failed to kill pid %d after 30s" % p.pid)
+        else:
+            self.log.error(
+                "psutil not available! Will wait 30s before "
+                "attempting to kill parent process. This should "
+                "not occur in mozilla automation. See bug 1143547."
+            )
+            for pid in child_pids:
+                self.killAndGetStack(
+                    pid, utilityPath, debuggerInfo, dump_screen=not debuggerInfo
+                )
+            if child_pids:
+                time.sleep(30)
+
+            self.killAndGetStack(
+                browser_pid, utilityPath, debuggerInfo, dump_screen=not debuggerInfo
+            )
+
+    def archiveMozLogs(self):
+        if self.mozLogs:
+            with zipfile.ZipFile(
+                "{}/mozLogs.zip".format(os.environ["MOZ_UPLOAD_DIR"]),
+                "w",
+                zipfile.ZIP_DEFLATED,
+            ) as logzip:
+                for logfile in glob.glob(
+                    "{}/moz*.log*".format(os.environ["MOZ_UPLOAD_DIR"])
+                ):
+                    logzip.write(logfile, os.path.basename(logfile))
+                    os.remove(logfile)
+                logzip.close()
+
+    class OutputHandler(object):
+
+        """line output handler for mozrunner"""
+
+        def __init__(
+            self,
+            harness,
+            utilityPath,
+            symbolsPath=None,
+            dump_screen_on_timeout=True,
+            dump_screen_on_fail=False,
+            shutdownLeaks=None,
+            lsanLeaks=None,
+            bisectChunk=None,
+        ):
+            """
+            harness -- harness instance
+            dump_screen_on_timeout -- whether to dump the screen on timeout
+            """
+            self.harness = harness
+            self.utilityPath = utilityPath
+            self.symbolsPath = symbolsPath
+            self.dump_screen_on_timeout = dump_screen_on_timeout
+            self.dump_screen_on_fail = dump_screen_on_fail
+            self.shutdownLeaks = shutdownLeaks
+            self.lsanLeaks = lsanLeaks
+            self.bisectChunk = bisectChunk
+            self.browserProcessId = None
+            self.stackFixerFunction = self.stackFixer()
+
+        def processOutputLine(self, line):
+            """per line handler of output for mozprocess"""
+            # Parsing the line (by the structured messages logger).
+            messages = self.harness.message_logger.parse_line(line)
+
+            for message in messages:
+                # Passing the message to the handlers
+                for handler in self.outputHandlers():
+                    message = handler(message)
+
+                # Processing the message by the logger
+                self.harness.message_logger.process_message(message)
+
+        __call__ = processOutputLine
+
+        def outputHandlers(self):
+            """returns ordered list of output handlers"""
+            handlers = [
+                self.fix_stack,
+                self.record_last_test,
+                self.dumpScreenOnTimeout,
+                self.dumpScreenOnFail,
+                self.trackShutdownLeaks,
+                self.trackLSANLeaks,
+                self.countline,
+            ]
+            if self.bisectChunk:
+                handlers.append(self.record_result)
+                handlers.append(self.first_error)
+
+            return handlers
+
+        def stackFixer(self):
+            """
+            return get_stack_fixer_function, if any, to use on the output lines
+            """
+            return get_stack_fixer_function(self.utilityPath, self.symbolsPath)
+
+        def finish(self):
+            if self.shutdownLeaks:
+                self.harness.countfail += self.shutdownLeaks.process()
+
+            if self.lsanLeaks:
+                self.harness.countfail += self.lsanLeaks.process()
+
+        # output message handlers:
+        # these take a message and return a message
+
+        def record_result(self, message):
+            # by default make the result key equal to pass.
+            if message["action"] == "test_start":
+                key = message["test"].split("/")[-1].strip()
+                self.harness.result[key] = "PASS"
+            elif message["action"] == "test_status":
+                if "expected" in message:
+                    key = message["test"].split("/")[-1].strip()
+                    self.harness.result[key] = "FAIL"
+                elif message["status"] == "FAIL":
+                    key = message["test"].split("/")[-1].strip()
+                    self.harness.result[key] = "TODO"
+            return message
+
+        def first_error(self, message):
+            if (
+                message["action"] == "test_status"
+                and "expected" in message
+                and message["status"] == "FAIL"
+            ):
+                key = message["test"].split("/")[-1].strip()
+                if key not in self.harness.expectedError:
+                    self.harness.expectedError[key] = message.get(
+                        "message", message["subtest"]
+                    ).strip()
+            return message
+
+        def countline(self, message):
+            if message["action"] == "log":
+                line = message.get("message", "")
+            elif message["action"] == "process_output":
+                line = message.get("data", "")
+            else:
+                return message
+            val = 0
+            try:
+                val = int(line.split(":")[-1].strip())
+            except (AttributeError, ValueError):
+                return message
+
+            if "Passed:" in line:
+                self.harness.countpass += val
+            elif "Failed:" in line:
+                self.harness.countfail += val
+            elif "Todo:" in line:
+                self.harness.counttodo += val
+            return message
+
+        def fix_stack(self, message):
+            if self.stackFixerFunction:
+                if message["action"] == "log":
+                    message["message"] = self.stackFixerFunction(message["message"])
+                elif message["action"] == "process_output":
+                    message["data"] = self.stackFixerFunction(message["data"])
+            return message
+
+        def record_last_test(self, message):
+            """record last test on harness"""
+            if message["action"] == "test_start":
+                self.harness.lastTestSeen = message["test"]
+            elif message["action"] == "test_end":
+                self.harness.lastTestSeen = "{} (finished)".format(message["test"])
+            return message
+
+        def dumpScreenOnTimeout(self, message):
+            if (
+                not self.dump_screen_on_fail
+                and self.dump_screen_on_timeout
+                and message["action"] == "test_status"
+                and "expected" in message
+                and "Test timed out" in message["subtest"]
+            ):
+                self.harness.dumpScreen(self.utilityPath)
+            return message
+
+        def dumpScreenOnFail(self, message):
+            if (
+                self.dump_screen_on_fail
+                and "expected" in message
+                and message["status"] == "FAIL"
+            ):
+                self.harness.dumpScreen(self.utilityPath)
+            return message
+
+        def trackLSANLeaks(self, message):
+            if self.lsanLeaks and message["action"] in ("log", "process_output"):
+                line = (
+                    message.get("message", "")
+                    if message["action"] == "log"
+                    else message["data"]
+                )
+                self.lsanLeaks.log(line)
+            return message
+
+        def trackShutdownLeaks(self, message):
+            if self.shutdownLeaks:
+                self.shutdownLeaks.log(message)
+            return message
+
+
+def view_gecko_profile_from_mochitest(profile_path, options, profiler_logger):
+    """Getting shutdown performance profiles from just the command line arguments is
+    difficult. This function makes the developer ergonomics a bit easier by taking the
+    generated Gecko profile, and automatically serving it to profiler.firefox.com. The
+    Gecko profile during shutdown is dumped to disk at:
+
+    {objdir}/_tests/testing/mochitest/{profilename}
+
+    This function takes that file, and launches a local webserver, and then points
+    a browser to profiler.firefox.com to view it. From there it's easy to publish
+    or save the profile.
+    """
+
+    if options.profilerSaveOnly:
+        # The user did not want this to automatically open, only share the location.
+        return
+
+    if not os.path.exists(profile_path):
+        profiler_logger.error(
+            "No profile was found at the profile path, cannot "
+            "launch profiler.firefox.com."
+        )
+        return
+
+    profiler_logger.info("Loading this profile in the Firefox Profiler")
+
+    view_gecko_profile(profile_path)
+
+
+def run_test_harness(parser, options):
+    parser.validate(options)
+
+    logger_options = {
+        key: value
+        for key, value in six.iteritems(vars(options))
+        if key.startswith("log") or key == "valgrind"
+    }
+
+    runner = MochitestDesktop(
+        options.flavor, logger_options, options.stagedAddons, quiet=options.quiet
+    )
+
+    options.runByManifest = False
+    if options.flavor in ("plain", "browser", "chrome"):
+        options.runByManifest = True
+
+    if options.verify or options.verify_fission:
+        result = runner.verifyTests(options)
+    else:
+        result = runner.runTests(options)
+
+    runner.archiveMozLogs()
+    runner.message_logger.finish()
+    return result
+
+
+def cli(args=sys.argv[1:]):
+    # parse command line options
+    parser = MochitestArgumentParser(app="generic")
+    options = parser.parse_args(args)
+    if options is None:
+        # parsing error
+        sys.exit(1)
+
+    return run_test_harness(parser, options)
+
+
+if __name__ == "__main__":
+    sys.exit(cli())