summaryrefslogtreecommitdiffstats
path: root/media/webrtc/trunk/build/android/pylib/run_java_tests.py
diff options
context:
space:
mode:
authorMatt A. Tobin <mattatobin@localhost.localdomain>2018-02-02 04:16:08 -0500
committerMatt A. Tobin <mattatobin@localhost.localdomain>2018-02-02 04:16:08 -0500
commit5f8de423f190bbb79a62f804151bc24824fa32d8 (patch)
tree10027f336435511475e392454359edea8e25895d /media/webrtc/trunk/build/android/pylib/run_java_tests.py
parent49ee0794b5d912db1f95dce6eb52d781dc210db5 (diff)
downloadUXP-5f8de423f190bbb79a62f804151bc24824fa32d8.tar
UXP-5f8de423f190bbb79a62f804151bc24824fa32d8.tar.gz
UXP-5f8de423f190bbb79a62f804151bc24824fa32d8.tar.lz
UXP-5f8de423f190bbb79a62f804151bc24824fa32d8.tar.xz
UXP-5f8de423f190bbb79a62f804151bc24824fa32d8.zip
Add m-esr52 at 52.6.0
Diffstat (limited to 'media/webrtc/trunk/build/android/pylib/run_java_tests.py')
-rw-r--r--media/webrtc/trunk/build/android/pylib/run_java_tests.py591
1 files changed, 591 insertions, 0 deletions
diff --git a/media/webrtc/trunk/build/android/pylib/run_java_tests.py b/media/webrtc/trunk/build/android/pylib/run_java_tests.py
new file mode 100644
index 000000000..fc0a13fd8
--- /dev/null
+++ b/media/webrtc/trunk/build/android/pylib/run_java_tests.py
@@ -0,0 +1,591 @@
+# Copyright (c) 2012 The Chromium Authors. All rights reserved.
+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+
+"""Runs the Java tests. See more information on run_instrumentation_tests.py."""
+
+import fnmatch
+import logging
+import os
+import re
+import shutil
+import sys
+import time
+
+import android_commands
+import apk_info
+from base_test_runner import BaseTestRunner
+from base_test_sharder import BaseTestSharder, SetTestsContainer
+import cmd_helper
+import constants
+import errors
+from forwarder import Forwarder
+from json_perf_parser import GetAverageRunInfoFromJSONString
+from perf_tests_helper import PrintPerfResult
+import sharded_tests_queue
+from test_result import SingleTestResult, TestResults
+import valgrind_tools
+
+_PERF_TEST_ANNOTATION = 'PerfTest'
+
+
+class FatalTestException(Exception):
+ """A fatal test exception."""
+ pass
+
+
+def _TestNameToExpectation(test_name):
+ # A test name is a Package.Path.Class#testName; convert to what we use in
+ # the expectation file.
+ return '.'.join(test_name.replace('#', '.').split('.')[-2:])
+
+
+def FilterTests(test_names, pattern_list, inclusive):
+ """Filters |test_names| using a list of patterns.
+
+ Args:
+ test_names: A list of test names.
+ pattern_list: A list of patterns.
+ inclusive: If True, returns the tests that match any pattern. if False,
+ returns the tests that do not match any pattern.
+ Returns:
+ A list of test names.
+ """
+ ret = []
+ for t in test_names:
+ has_match = False
+ for pattern in pattern_list:
+ has_match = has_match or fnmatch.fnmatch(_TestNameToExpectation(t),
+ pattern)
+ if has_match == inclusive:
+ ret += [t]
+ return ret
+
+
+class TestRunner(BaseTestRunner):
+ """Responsible for running a series of tests connected to a single device."""
+
+ _DEVICE_DATA_DIR = 'chrome/test/data'
+ _EMMA_JAR = os.path.join(os.environ.get('ANDROID_BUILD_TOP', ''),
+ 'external/emma/lib/emma.jar')
+ _COVERAGE_MERGED_FILENAME = 'unittest_coverage.es'
+ _COVERAGE_WEB_ROOT_DIR = os.environ.get('EMMA_WEB_ROOTDIR')
+ _COVERAGE_FILENAME = 'coverage.ec'
+ _COVERAGE_RESULT_PATH = ('/data/data/com.google.android.apps.chrome/files/' +
+ _COVERAGE_FILENAME)
+ _COVERAGE_META_INFO_PATH = os.path.join(os.environ.get('ANDROID_BUILD_TOP',
+ ''),
+ 'out/target/common/obj/APPS',
+ 'Chrome_intermediates/coverage.em')
+ _HOSTMACHINE_PERF_OUTPUT_FILE = '/tmp/chrome-profile'
+ _DEVICE_PERF_OUTPUT_SEARCH_PREFIX = (constants.DEVICE_PERF_OUTPUT_DIR +
+ '/chrome-profile*')
+ _DEVICE_HAS_TEST_FILES = {}
+
+ def __init__(self, options, device, tests_iter, coverage, shard_index, apks,
+ ports_to_forward):
+ """Create a new TestRunner.
+
+ Args:
+ options: An options object with the following required attributes:
+ - build_type: 'Release' or 'Debug'.
+ - install_apk: Re-installs the apk if opted.
+ - save_perf_json: Whether or not to save the JSON file from UI perf
+ tests.
+ - screenshot_failures: Take a screenshot for a test failure
+ - tool: Name of the Valgrind tool.
+ - wait_for_debugger: blocks until the debugger is connected.
+ device: Attached android device.
+ tests_iter: A list of tests to be run.
+ coverage: Collects coverage information if opted.
+ shard_index: shard # for this TestRunner, used to create unique port
+ numbers.
+ apks: A list of ApkInfo objects need to be installed. The first element
+ should be the tests apk, the rests could be the apks used in test.
+ The default is ChromeTest.apk.
+ ports_to_forward: A list of port numbers for which to set up forwarders.
+ Can be optionally requested by a test case.
+ Raises:
+ FatalTestException: if coverage metadata is not available.
+ """
+ BaseTestRunner.__init__(
+ self, device, options.tool, shard_index, options.build_type)
+
+ if not apks:
+ apks = [apk_info.ApkInfo(options.test_apk_path,
+ options.test_apk_jar_path)]
+
+ self.build_type = options.build_type
+ self.install_apk = options.install_apk
+ self.save_perf_json = options.save_perf_json
+ self.screenshot_failures = options.screenshot_failures
+ self.wait_for_debugger = options.wait_for_debugger
+
+ self.tests_iter = tests_iter
+ self.coverage = coverage
+ self.apks = apks
+ self.test_apk = apks[0]
+ self.instrumentation_class_path = self.test_apk.GetPackageName()
+ self.ports_to_forward = ports_to_forward
+
+ self.test_results = TestResults()
+ self.forwarder = None
+
+ if self.coverage:
+ if os.path.exists(TestRunner._COVERAGE_MERGED_FILENAME):
+ os.remove(TestRunner._COVERAGE_MERGED_FILENAME)
+ if not os.path.exists(TestRunner._COVERAGE_META_INFO_PATH):
+ raise FatalTestException('FATAL ERROR in ' + sys.argv[0] +
+ ' : Coverage meta info [' +
+ TestRunner._COVERAGE_META_INFO_PATH +
+ '] does not exist.')
+ if (not TestRunner._COVERAGE_WEB_ROOT_DIR or
+ not os.path.exists(TestRunner._COVERAGE_WEB_ROOT_DIR)):
+ raise FatalTestException('FATAL ERROR in ' + sys.argv[0] +
+ ' : Path specified in $EMMA_WEB_ROOTDIR [' +
+ TestRunner._COVERAGE_WEB_ROOT_DIR +
+ '] does not exist.')
+
+ def _GetTestsIter(self):
+ if not self.tests_iter:
+ # multiprocessing.Queue can't be pickled across processes if we have it as
+ # a member set during constructor. Grab one here instead.
+ self.tests_iter = (BaseTestSharder.tests_container)
+ assert self.tests_iter
+ return self.tests_iter
+
+ def CopyTestFilesOnce(self):
+ """Pushes the test data files to the device. Installs the apk if opted."""
+ if TestRunner._DEVICE_HAS_TEST_FILES.get(self.device, False):
+ logging.warning('Already copied test files to device %s, skipping.',
+ self.device)
+ return
+ host_test_files = [
+ ('android_webview/test/data/device_files', 'webview'),
+ ('content/test/data/android/device_files', 'content'),
+ ('chrome/test/data/android/device_files', 'chrome')
+ ]
+ for (host_src, dst_layer) in host_test_files:
+ host_test_files_path = constants.CHROME_DIR + '/' + host_src
+ if os.path.exists(host_test_files_path):
+ self.adb.PushIfNeeded(host_test_files_path,
+ self.adb.GetExternalStorage() + '/' +
+ TestRunner._DEVICE_DATA_DIR + '/' + dst_layer)
+ if self.install_apk:
+ for apk in self.apks:
+ self.adb.ManagedInstall(apk.GetApkPath(),
+ package_name=apk.GetPackageName())
+ self.tool.CopyFiles()
+ TestRunner._DEVICE_HAS_TEST_FILES[self.device] = True
+
+ def SaveCoverageData(self, test):
+ """Saves the Emma coverage data before it's overwritten by the next test.
+
+ Args:
+ test: the test whose coverage data is collected.
+ """
+ if not self.coverage:
+ return
+ if not self.adb.Adb().Pull(TestRunner._COVERAGE_RESULT_PATH,
+ constants.CHROME_DIR):
+ logging.error('ERROR: Unable to find file ' +
+ TestRunner._COVERAGE_RESULT_PATH +
+ ' on the device for test ' + test)
+ pulled_coverage_file = os.path.join(constants.CHROME_DIR,
+ TestRunner._COVERAGE_FILENAME)
+ if os.path.exists(TestRunner._COVERAGE_MERGED_FILENAME):
+ cmd = ['java', '-classpath', TestRunner._EMMA_JAR, 'emma', 'merge',
+ '-in', pulled_coverage_file,
+ '-in', TestRunner._COVERAGE_MERGED_FILENAME,
+ '-out', TestRunner._COVERAGE_MERGED_FILENAME]
+ cmd_helper.RunCmd(cmd)
+ else:
+ shutil.copy(pulled_coverage_file,
+ TestRunner._COVERAGE_MERGED_FILENAME)
+ os.remove(pulled_coverage_file)
+
+ def GenerateCoverageReportIfNeeded(self):
+ """Uses the Emma to generate a coverage report and a html page."""
+ if not self.coverage:
+ return
+ cmd = ['java', '-classpath', TestRunner._EMMA_JAR,
+ 'emma', 'report', '-r', 'html',
+ '-in', TestRunner._COVERAGE_MERGED_FILENAME,
+ '-in', TestRunner._COVERAGE_META_INFO_PATH]
+ cmd_helper.RunCmd(cmd)
+ new_dir = os.path.join(TestRunner._COVERAGE_WEB_ROOT_DIR,
+ time.strftime('Coverage_for_%Y_%m_%d_%a_%H:%M'))
+ shutil.copytree('coverage', new_dir)
+
+ latest_dir = os.path.join(TestRunner._COVERAGE_WEB_ROOT_DIR,
+ 'Latest_Coverage_Run')
+ if os.path.exists(latest_dir):
+ shutil.rmtree(latest_dir)
+ os.mkdir(latest_dir)
+ webserver_new_index = os.path.join(new_dir, 'index.html')
+ webserver_new_files = os.path.join(new_dir, '_files')
+ webserver_latest_index = os.path.join(latest_dir, 'index.html')
+ webserver_latest_files = os.path.join(latest_dir, '_files')
+ # Setup new softlinks to last result.
+ os.symlink(webserver_new_index, webserver_latest_index)
+ os.symlink(webserver_new_files, webserver_latest_files)
+ cmd_helper.RunCmd(['chmod', '755', '-R', latest_dir, new_dir])
+
+ def _GetInstrumentationArgs(self):
+ ret = {}
+ if self.coverage:
+ ret['coverage'] = 'true'
+ if self.wait_for_debugger:
+ ret['debug'] = 'true'
+ return ret
+
+ def _TakeScreenshot(self, test):
+ """Takes a screenshot from the device."""
+ screenshot_tool = os.path.join(constants.CHROME_DIR,
+ 'third_party/android_tools/sdk/tools/monkeyrunner')
+ screenshot_script = os.path.join(constants.CHROME_DIR,
+ 'build/android/monkeyrunner_screenshot.py')
+ screenshot_path = os.path.join(constants.CHROME_DIR,
+ 'out_screenshots')
+ if not os.path.exists(screenshot_path):
+ os.mkdir(screenshot_path)
+ screenshot_name = os.path.join(screenshot_path, test + '.png')
+ logging.info('Taking screenshot named %s', screenshot_name)
+ cmd_helper.RunCmd([screenshot_tool, screenshot_script,
+ '--serial', self.device,
+ '--file', screenshot_name])
+
+ def SetUp(self):
+ """Sets up the test harness and device before all tests are run."""
+ super(TestRunner, self).SetUp()
+ if not self.adb.IsRootEnabled():
+ logging.warning('Unable to enable java asserts for %s, non rooted device',
+ self.device)
+ else:
+ if self.adb.SetJavaAssertsEnabled(enable=True):
+ self.adb.Reboot(full_reboot=False)
+
+ # We give different default value to launch HTTP server based on shard index
+ # because it may have race condition when multiple processes are trying to
+ # launch lighttpd with same port at same time.
+ http_server_ports = self.LaunchTestHttpServer(
+ os.path.join(constants.CHROME_DIR),
+ (constants.LIGHTTPD_RANDOM_PORT_FIRST + self.shard_index))
+ if self.ports_to_forward:
+ port_pairs = [(port, port) for port in self.ports_to_forward]
+ # We need to remember which ports the HTTP server is using, since the
+ # forwarder will stomp on them otherwise.
+ port_pairs.append(http_server_ports)
+ self.forwarder = Forwarder(
+ self.adb, port_pairs, self.tool, '127.0.0.1', self.build_type)
+ self.CopyTestFilesOnce()
+ self.flags.AddFlags(['--enable-test-intents'])
+
+ def TearDown(self):
+ """Cleans up the test harness and saves outstanding data from test run."""
+ if self.forwarder:
+ self.forwarder.Close()
+ self.GenerateCoverageReportIfNeeded()
+ super(TestRunner, self).TearDown()
+
+ def TestSetup(self, test):
+ """Sets up the test harness for running a particular test.
+
+ Args:
+ test: The name of the test that will be run.
+ """
+ self.SetupPerfMonitoringIfNeeded(test)
+ self._SetupIndividualTestTimeoutScale(test)
+ self.tool.SetupEnvironment()
+
+ # Make sure the forwarder is still running.
+ self.RestartHttpServerForwarderIfNecessary()
+
+ def _IsPerfTest(self, test):
+ """Determines whether a test is a performance test.
+
+ Args:
+ test: The name of the test to be checked.
+
+ Returns:
+ Whether the test is annotated as a performance test.
+ """
+ return _PERF_TEST_ANNOTATION in self.test_apk.GetTestAnnotations(test)
+
+ def SetupPerfMonitoringIfNeeded(self, test):
+ """Sets up performance monitoring if the specified test requires it.
+
+ Args:
+ test: The name of the test to be run.
+ """
+ if not self._IsPerfTest(test):
+ return
+ self.adb.Adb().SendCommand('shell rm ' +
+ TestRunner._DEVICE_PERF_OUTPUT_SEARCH_PREFIX)
+ self.adb.StartMonitoringLogcat()
+
+ def TestTeardown(self, test, test_result):
+ """Cleans up the test harness after running a particular test.
+
+ Depending on the options of this TestRunner this might handle coverage
+ tracking or performance tracking. This method will only be called if the
+ test passed.
+
+ Args:
+ test: The name of the test that was just run.
+ test_result: result for this test.
+ """
+
+ self.tool.CleanUpEnvironment()
+
+ # The logic below relies on the test passing.
+ if not test_result or test_result.GetStatusCode():
+ return
+
+ self.TearDownPerfMonitoring(test)
+ self.SaveCoverageData(test)
+
+ def TearDownPerfMonitoring(self, test):
+ """Cleans up performance monitoring if the specified test required it.
+
+ Args:
+ test: The name of the test that was just run.
+ Raises:
+ FatalTestException: if there's anything wrong with the perf data.
+ """
+ if not self._IsPerfTest(test):
+ return
+ raw_test_name = test.split('#')[1]
+
+ # Wait and grab annotation data so we can figure out which traces to parse
+ regex = self.adb.WaitForLogMatch(re.compile('\*\*PERFANNOTATION\(' +
+ raw_test_name +
+ '\)\:(.*)'), None)
+
+ # If the test is set to run on a specific device type only (IE: only
+ # tablet or phone) and it is being run on the wrong device, the test
+ # just quits and does not do anything. The java test harness will still
+ # print the appropriate annotation for us, but will add --NORUN-- for
+ # us so we know to ignore the results.
+ # The --NORUN-- tag is managed by MainActivityTestBase.java
+ if regex.group(1) != '--NORUN--':
+
+ # Obtain the relevant perf data. The data is dumped to a
+ # JSON formatted file.
+ json_string = self.adb.GetFileContents(
+ '/data/data/com.google.android.apps.chrome/files/PerfTestData.txt')
+
+ if json_string:
+ json_string = '\n'.join(json_string)
+ else:
+ raise FatalTestException('Perf file does not exist or is empty')
+
+ if self.save_perf_json:
+ json_local_file = '/tmp/chromium-android-perf-json-' + raw_test_name
+ with open(json_local_file, 'w') as f:
+ f.write(json_string)
+ logging.info('Saving Perf UI JSON from test ' +
+ test + ' to ' + json_local_file)
+
+ raw_perf_data = regex.group(1).split(';')
+
+ for raw_perf_set in raw_perf_data:
+ if raw_perf_set:
+ perf_set = raw_perf_set.split(',')
+ if len(perf_set) != 3:
+ raise FatalTestException('Unexpected number of tokens in '
+ 'perf annotation string: ' + raw_perf_set)
+
+ # Process the performance data
+ result = GetAverageRunInfoFromJSONString(json_string, perf_set[0])
+
+ PrintPerfResult(perf_set[1], perf_set[2],
+ [result['average']], result['units'])
+
+ def _SetupIndividualTestTimeoutScale(self, test):
+ timeout_scale = self._GetIndividualTestTimeoutScale(test)
+ valgrind_tools.SetChromeTimeoutScale(self.adb, timeout_scale)
+
+ def _GetIndividualTestTimeoutScale(self, test):
+ """Returns the timeout scale for the given |test|."""
+ annotations = self.apks[0].GetTestAnnotations(test)
+ timeout_scale = 1
+ if 'TimeoutScale' in annotations:
+ for annotation in annotations:
+ scale_match = re.match('TimeoutScale:([0-9]+)', annotation)
+ if scale_match:
+ timeout_scale = int(scale_match.group(1))
+ if self.wait_for_debugger:
+ timeout_scale *= 100
+ return timeout_scale
+
+ def _GetIndividualTestTimeoutSecs(self, test):
+ """Returns the timeout in seconds for the given |test|."""
+ annotations = self.apks[0].GetTestAnnotations(test)
+ if 'Manual' in annotations:
+ return 600 * 60
+ if 'External' in annotations:
+ return 10 * 60
+ if 'LargeTest' in annotations or _PERF_TEST_ANNOTATION in annotations:
+ return 5 * 60
+ if 'MediumTest' in annotations:
+ return 3 * 60
+ return 1 * 60
+
+ def RunTests(self):
+ """Runs the tests, generating the coverage if needed.
+
+ Returns:
+ A TestResults object.
+ """
+ instrumentation_path = (self.instrumentation_class_path +
+ '/android.test.InstrumentationTestRunner')
+ instrumentation_args = self._GetInstrumentationArgs()
+ for test in self._GetTestsIter():
+ test_result = None
+ start_date_ms = None
+ try:
+ self.TestSetup(test)
+ start_date_ms = int(time.time()) * 1000
+ args_with_filter = dict(instrumentation_args)
+ args_with_filter['class'] = test
+ # |test_results| is a list that should contain
+ # a single TestResult object.
+ logging.warn(args_with_filter)
+ (test_results, _) = self.adb.Adb().StartInstrumentation(
+ instrumentation_path=instrumentation_path,
+ instrumentation_args=args_with_filter,
+ timeout_time=(self._GetIndividualTestTimeoutSecs(test) *
+ self._GetIndividualTestTimeoutScale(test) *
+ self.tool.GetTimeoutScale()))
+ duration_ms = int(time.time()) * 1000 - start_date_ms
+ assert len(test_results) == 1
+ test_result = test_results[0]
+ status_code = test_result.GetStatusCode()
+ if status_code:
+ log = test_result.GetFailureReason()
+ if not log:
+ log = 'No information.'
+ if self.screenshot_failures or log.find('INJECT_EVENTS perm') >= 0:
+ self._TakeScreenshot(test)
+ self.test_results.failed += [SingleTestResult(test, start_date_ms,
+ duration_ms, log)]
+ else:
+ result = [SingleTestResult(test, start_date_ms, duration_ms)]
+ self.test_results.ok += result
+ # Catch exceptions thrown by StartInstrumentation().
+ # See ../../third_party/android/testrunner/adb_interface.py
+ except (errors.WaitForResponseTimedOutError,
+ errors.DeviceUnresponsiveError,
+ errors.InstrumentationError), e:
+ if start_date_ms:
+ duration_ms = int(time.time()) * 1000 - start_date_ms
+ else:
+ start_date_ms = int(time.time()) * 1000
+ duration_ms = 0
+ message = str(e)
+ if not message:
+ message = 'No information.'
+ self.test_results.crashed += [SingleTestResult(test, start_date_ms,
+ duration_ms,
+ message)]
+ test_result = None
+ self.TestTeardown(test, test_result)
+ return self.test_results
+
+
+class TestSharder(BaseTestSharder):
+ """Responsible for sharding the tests on the connected devices."""
+
+ def __init__(self, attached_devices, options, tests, apks):
+ BaseTestSharder.__init__(self, attached_devices)
+ self.options = options
+ self.tests = tests
+ self.apks = apks
+
+ def SetupSharding(self, tests):
+ """Called before starting the shards."""
+ SetTestsContainer(sharded_tests_queue.ShardedTestsQueue(
+ len(self.attached_devices), tests))
+
+ def CreateShardedTestRunner(self, device, index):
+ """Creates a sharded test runner.
+
+ Args:
+ device: Device serial where this shard will run.
+ index: Index of this device in the pool.
+
+ Returns:
+ A TestRunner object.
+ """
+ return TestRunner(self.options, device, None, False, index, self.apks, [])
+
+
+def DispatchJavaTests(options, apks):
+ """Dispatches Java tests onto connected device(s).
+
+ If possible, this method will attempt to shard the tests to
+ all connected devices. Otherwise, dispatch and run tests on one device.
+
+ Args:
+ options: Command line options.
+ apks: list of APKs to use.
+
+ Returns:
+ A TestResults object holding the results of the Java tests.
+
+ Raises:
+ FatalTestException: when there's no attached the devices.
+ """
+ test_apk = apks[0]
+ if options.annotation:
+ available_tests = test_apk.GetAnnotatedTests(options.annotation)
+ if len(options.annotation) == 1 and options.annotation[0] == 'SmallTest':
+ tests_without_annotation = [
+ m for m in
+ test_apk.GetTestMethods()
+ if not test_apk.GetTestAnnotations(m) and
+ not apk_info.ApkInfo.IsPythonDrivenTest(m)]
+ if tests_without_annotation:
+ tests_without_annotation.sort()
+ logging.warning('The following tests do not contain any annotation. '
+ 'Assuming "SmallTest":\n%s',
+ '\n'.join(tests_without_annotation))
+ available_tests += tests_without_annotation
+ else:
+ available_tests = [m for m in test_apk.GetTestMethods()
+ if not apk_info.ApkInfo.IsPythonDrivenTest(m)]
+ coverage = os.environ.get('EMMA_INSTRUMENT') == 'true'
+
+ tests = []
+ if options.test_filter:
+ # |available_tests| are in adb instrument format: package.path.class#test.
+ filter_without_hash = options.test_filter.replace('#', '.')
+ tests = [t for t in available_tests
+ if filter_without_hash in t.replace('#', '.')]
+ else:
+ tests = available_tests
+
+ if not tests:
+ logging.warning('No Java tests to run with current args.')
+ return TestResults()
+
+ tests *= options.number_of_runs
+
+ attached_devices = android_commands.GetAttachedDevices()
+ test_results = TestResults()
+
+ if not attached_devices:
+ raise FatalTestException('You have no devices attached or visible!')
+ if options.device:
+ attached_devices = [options.device]
+
+ logging.info('Will run: %s', str(tests))
+
+ if len(attached_devices) > 1 and (coverage or options.wait_for_debugger):
+ logging.warning('Coverage / debugger can not be sharded, '
+ 'using first available device')
+ attached_devices = attached_devices[:1]
+ sharder = TestSharder(attached_devices, options, tests, apks)
+ test_results = sharder.RunShardedTests()
+ return test_results