1 2012-08-08 Peter Beverloo <peter@chromium.org>
3 The cr-android EWS should actually be building Chromium for Android
4 https://bugs.webkit.org/show_bug.cgi?id=93489
6 Reviewed by Adam Barth.
8 The cr-android EWS bots should be building Chromium for Android, which
9 requires some additional flags to Chromium's default configuration.
11 * Scripts/webkitdirs.pm:
13 * Scripts/webkitpy/common/config/ports.py:
14 (DeprecatedPort.port):
15 (ChromiumAndroidPort):
16 (ChromiumAndroidPort.update_webkit_command):
17 (ChromiumAndroidPort.build_webkit_command):
18 * Scripts/webkitpy/common/config/ports_unittest.py:
19 (DeprecatedPortTest.test_chromium_android_port):
20 * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
22 * Scripts/webkitpy/tool/steps/update_unittest.py:
23 (UpdateTest.test_update_command_non_interactive):
24 (UpdateTest.test_update_command_interactive):
26 2012-08-08 Alexey Proskuryakov <ap@apple.com>
28 [WK2] MiniBrowser crashes on window.open()
29 https://bugs.webkit.org/show_bug.cgi?id=93413
31 Reviewed by Dan Bernstein.
33 * MiniBrowser/mac/BrowserWindowController.m: (createNewPage): This is a "create"
34 API, so it needs to return a retained result.
36 * MiniBrowser/mac/WebBundle/WebBundleMain.m: (didClearWindowObjectForFrame):
37 When this function is called, URL appears to be poorly defined. Sometimes, it's
38 the new URL, other times it's the old one. An "old" URL is null in a new page.
40 2012-08-08 Loïc Yhuel <loic.yhuel@softathome.com>
42 [Qt] Compile errors with OpenGLES2
43 https://bugs.webkit.org/show_bug.cgi?id=93206
45 Reviewed by Noam Rosenthal.
47 Fix build with OpenGLES2 on Linux desktop
49 * qmake/mkspecs/features/features.prf: Don't use GLX with OpenGLES2
51 2012-08-08 Peter Beverloo <peter@chromium.org>
53 Introduce cr-android bots in the EWS system
54 https://bugs.webkit.org/show_bug.cgi?id=93472
56 Reviewed by Adam Barth.
58 This adds the queue to the EWS scripts. Two builders are ready to be
59 attached to the queue. All webkitpy tests pass with this change applied.
61 * QueueStatusServer/model/queues.py:
63 * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
65 * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
66 (EarlyWarningSytemTest.test_builder_ewses):
68 2012-08-08 Carlos Garcia Campos <cgarcia@igalia.com>
70 Unreviewed. Fix GTK+ build with GTK2 after r121475.
72 * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp: Add missing
73 include required when building with GTK2.
75 2012-08-08 Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com>
77 [WK2] [WTR] Provide Resource Response dumping.
78 https://bugs.webkit.org/show_bug.cgi?id=93454
80 Reviewed by Kenneth Rohde Christiansen.
82 Added missing dumpResourceResponseMIMETypes() method to testRunner. Provided resource response dumping.
84 * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
85 * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
86 (WTR::InjectedBundlePage::didReceiveResponseForResource):
87 * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
88 (WTR::LayoutTestController::LayoutTestController):
89 * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
90 (WTR::LayoutTestController::dumpProgressFinishedCallback):
91 (WTR::LayoutTestController::dumpResourceResponseMIMETypes):
92 (WTR::LayoutTestController::shouldDumpResourceResponseMIMETypes):
93 (LayoutTestController):
95 2012-08-08 Simon Hausmann <simon.hausmann@nokia.com>
97 [Qt][Win] Fix compilation of DumpRenderTree
98 https://bugs.webkit.org/show_bug.cgi?id=93461
100 Reviewed by Tor Arne Vestbø.
102 DRT includes stdint.h, for which we have wrappers in Source/JavaScriptCore/os-win32. Use these not only
103 for libraries but any building template.
105 * qmake/mkspecs/features/default_post.prf:
107 2012-08-08 Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com>
109 [WK2] [WTR] Refactoring: LayoutTestController::shouldDumpProgressFinishedCallback() should be const
110 https://bugs.webkit.org/show_bug.cgi?id=93457
112 Reviewed by Csaba Osztrogonác.
114 * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
115 (WTR::LayoutTestController::shouldDumpProgressFinishedCallback): Added constness.
117 2012-08-07 Csaba Osztrogonác <ossy@webkit.org>
119 Unreviewed typo fix after r124988.
121 * Scripts/webkitpy/layout_tests/models/test_expectations.py:
124 2012-08-07 YoungTaeck Song <youngtaeck.song@samsung.com>
126 [WK2][EFL] Implement accelerated compositing on WK2 Efl port
127 https://bugs.webkit.org/show_bug.cgi?id=89840
129 Reviewed by Noam Rosenthal.
131 Implement accelerated composition with TiledBackingStore on WK2 Efl port.
132 Add OPENGL_LIBRARIES in CMakeList.txt.
134 * MiniBrowser/efl/CMakeLists.txt:
135 * WebKitTestRunner/PlatformEfl.cmake:
137 2012-08-07 Csaba Osztrogonác <ossy@webkit.org>
139 [NRWT] REGRESSION(r124967): New tests without expected results handled as failures
140 https://bugs.webkit.org/show_bug.cgi?id=93434
142 Reviewed by Tony Chang.
144 Change back "missing results" to "no expected result found", because master.cfg's results parser expects it.
146 * Scripts/webkitpy/layout_tests/models/test_expectations.py:
149 2012-08-07 Csaba Osztrogonác <ossy@webkit.org>
151 buildbot: Remove the configuration of hfreyther*
152 https://bugs.webkit.org/show_bug.cgi?id=91849
154 Unreviewed typo fix after r123757.
156 * BuildSlaveSupport/build.webkit.org-config/config.json:
158 2012-08-07 Dirk Pranke <dpranke@chromium.org>
160 REGRESSION: PrettyPatchTest.test_pretty_diff_encodings has been failing on Chromium Windows
161 https://bugs.webkit.org/show_bug.cgi?id=93192
163 Unreviewed, build fix.
165 Disabling the test for now as PrettyPatch is just broken on win32.
167 * Scripts/webkitpy/common/prettypatch_unittest.py:
168 (test_pretty_diff_encodings):
170 2012-08-07 Dirk Pranke <dpranke@chromium.org>
172 [NRWT] Would like an output mode similar to ORWT verbose one
173 https://bugs.webkit.org/show_bug.cgi?id=88702
175 Reviewed by Ryosuke Niwa.
177 Change the --verbose logging for new-run-webkit-tests so that
178 it matches ORWT more; we just print one line per test. Use
179 --debug-rwt-logging to get the full debug stream (aka old ORWT
182 * Scripts/webkitpy/layout_tests/models/test_expectations.py:
184 * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
185 (_set_up_derived_options):
187 * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
188 (MainTest.test_retrying_and_flaky_tests):
189 * Scripts/webkitpy/layout_tests/views/printing.py:
191 (Printer._print_result_summary_entry):
192 (Printer._print_one_line_summary):
193 (Printer._print_test_result):
194 (Printer._print_baseline):
195 (Printer._print_unexpected_results):
197 2012-08-07 Dirk Pranke <dpranke@chromium.org>
199 nrwt: handle errors from image diff better
200 https://bugs.webkit.org/show_bug.cgi?id=92934
202 Reviewed by Ojan Vafai.
204 Re-land the change in r124801 with a fix ... in the case where
205 the ImageDiff is passed a tolerance and passes the fuzzy check,
206 we were returning the wrong value (missing an empty error
207 string) and crashing; this patch fixes that and adds a test for
208 that case (TestImageDiffer.test_image_diff_passed).
210 * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
211 (SingleTestRunner._compare_image):
212 (SingleTestRunner._compare_output_with_reference):
213 * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
215 * Scripts/webkitpy/layout_tests/controllers/test_result_writer_unittest.py:
216 (TestResultWriterTest.test_reftest_diff_image.ImageDiffTestPort.diff_image):
217 (TestResultWriterTest):
218 * Scripts/webkitpy/layout_tests/port/base.py:
220 * Scripts/webkitpy/layout_tests/port/chromium.py:
221 (ChromiumPort.diff_image):
222 * Scripts/webkitpy/layout_tests/port/chromium_port_testcase.py:
223 (ChromiumPortTestCase.test_diff_image_crashed):
224 * Scripts/webkitpy/layout_tests/port/image_diff.py:
225 (ImageDiffer.diff_image):
227 * Scripts/webkitpy/layout_tests/port/image_diff_unittest.py:
228 (TestImageDiffer.test_diff_image):
229 * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
230 (MockDRTPortTest.test_diff_image_crashed):
231 * Scripts/webkitpy/layout_tests/port/port_testcase.py:
232 (PortTestCase.test_diff_image):
233 (PortTestCase.test_diff_image_crashed):
234 (PortTestCase.test_diff_image_crashed.make_proc):
235 * Scripts/webkitpy/layout_tests/port/server_process_mock.py:
236 (MockServerProcess.__init__):
237 * Scripts/webkitpy/layout_tests/port/test.py:
238 (TestPort.diff_image):
239 * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
240 (MainTest.test_tolerance.ImageDiffTestPort.diff_image):
242 2012-08-07 Dirk Pranke <dpranke@chromium.org>
244 nrwt: --no-build isn't working
245 https://bugs.webkit.org/show_bug.cgi?id=93415
247 Reviewed by Ryosuke Niwa.
249 Turns out our optimization to avoid calling
250 webkit-build-directory N times for each worker broke --no-build.
252 Fixing, and adding a test.
254 Also, the gtk port wasn't using the default check_build() logic,
255 but I don't know why not. Removing their custom hook and will
256 verify that this is okay in the review ...
258 * Scripts/webkitpy/layout_tests/port/base.py:
262 * Scripts/webkitpy/layout_tests/port/gtk.py:
263 (GtkPort._path_to_image_diff):
264 * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
265 (MockDRTPortTest.test_check_build):
266 * Scripts/webkitpy/layout_tests/port/port_testcase.py:
267 (test_path_to_apache_config_file):
269 (test_check_build.build_driver_called):
271 2012-08-07 Ryosuke Niwa <rniwa@webkit.org>
273 run-perf-tests should support --no-show-results
274 https://bugs.webkit.org/show_bug.cgi?id=93409
276 Reviewed by Dirk Pranke.
278 Add the support for --no-show-results.
280 Also replace only local paths to jquery and flot instead of embedding them.
281 The results page is still standalone in that it tries to load scripts
282 from both webkit.org and local filesystem.
284 * Scripts/webkitpy/performance_tests/perftestsrunner.py:
285 (PerfTestsRunner._parse_args):
286 (PerfTestsRunner.run):
287 (PerfTestsRunner._generate_and_show_results):
288 (PerfTestsRunner._generate_output_files):
289 * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
290 (create_runner_and_setup_results_template):
291 (test_run_generates_and_show_results_page):
292 (test_run_respects_no_show_results):
294 2012-08-07 Dirk Pranke <dpranke@chromium.org>
296 test-webkitpy hangs in a new checkout on snow leopard
297 https://bugs.webkit.org/show_bug.cgi?id=93301
299 Reviewed by Ryosuke Niwa.
301 This change works around what appears to be a bug in Python
302 2.6.1 (the version that ships on Mac Snow Leopard) that causes
303 the multiprocessing module to hang after we use the
304 autoinstaller; I'm guessing it's some sort of python sockets
305 issue. I was unable to reproduce this with 2.6.5 or newer
308 * Scripts/webkitpy/common/system/autoinstall.py:
309 (AutoInstaller.install):
310 * Scripts/webkitpy/test/main.py:
312 * Scripts/webkitpy/thirdparty/__init__.py:
313 (AutoinstallImportHook._install_mechanize):
314 (AutoinstallImportHook._install_pep8):
315 (AutoinstallImportHook._install_pylint):
316 (AutoinstallImportHook._install_buildbot):
317 (AutoinstallImportHook._install_coverage):
318 (AutoinstallImportHook._install_eliza):
319 (AutoinstallImportHook._install_irc):
320 (AutoinstallImportHook._install_webpagereplay):
321 (AutoinstallImportHook._install):
322 (autoinstall_everything):
324 2012-08-07 Xianzhu Wang <wangxianzhu@chromium.org>
326 [Chromium-Android][NRWT] Fix 2 FIXMEs in chromium_android.py
327 https://bugs.webkit.org/show_bug.cgi?id=93381
329 Reviewed by Adam Barth.
331 * Scripts/webkitpy/layout_tests/port/chromium_android.py:
332 (ChromiumAndroidDriver.__init__):
333 (ChromiumAndroidDriver._push_executable): Remove temporary lines as the chromium change has been landed: http://src.chromium.org/viewvc/chrome?view=rev&revision=150371
334 (ChromiumAndroidDriver._setup_performance): Update all scaling governor files.
335 (ChromiumAndroidDriver._teardown_performance): Update all scaling governor files.
337 2012-08-07 W. James MacLean <wjmaclean@chromium.org>
339 [chromium] Add support to DumpRenderTree [EventSender] for GestureTapDown events.
340 https://bugs.webkit.org/show_bug.cgi?id=93286
342 Reviewed by James Robinson.
344 Adds support to generate GestureTapDown events from EventSender, to facilitate
345 link highlighting layout tests.
347 Added a GestureTapEvent to one existing test, more tests to follow when LinkHighlight CLs start to land.
349 * DumpRenderTree/chromium/TestRunner/EventSender.cpp:
350 (EventSender::EventSender):
351 (EventSender::gestureTapDown):
352 (EventSender::gestureEvent):
353 * DumpRenderTree/chromium/TestRunner/EventSender.h:
356 2012-08-07 Marcelo Lira <marcelo.lira@openbossa.org>
358 [Qt] Add support for the Gamepad API
359 https://bugs.webkit.org/show_bug.cgi?id=90637
361 Reviewed by Alexis Menard.
363 If the libudev library is present, the GAMEPAD flag is
364 turned on by default for the Qt port.
366 * Scripts/webkitperl/FeatureList.pm:
367 * qmake/mkspecs/features/features.prf:
369 2012-08-07 Rob Buis <rbuis@rim.com>
371 [BlackBerry] Turn on CSS Variables
372 https://bugs.webkit.org/show_bug.cgi?id=93376
374 Reviewed by Antonio Gomes.
376 Turn on CSS Variables for BlackBerry port.
378 * Scripts/webkitperl/FeatureList.pm:
380 2012-08-07 No'am Rosenthal <noam.rosenthal@nokia.com>
382 [Qt] Make it possible to build without QtQuick
384 Reviewed by Simon Hausmann.
387 * qmake/mkspecs/features/features.prf:
389 2012-08-02 Simon Hausmann <simon.hausmann@nokia.com>
391 [Qt] Remove Qt 4 specific code paths
392 https://bugs.webkit.org/show_bug.cgi?id=88161
394 Reviewed by Kenneth Rohde Christiansen.
396 * DumpRenderTree/qt/DumpRenderTree.pro:
397 * DumpRenderTree/qt/EventSenderQt.cpp:
398 (EventSender::touchCancel):
399 (EventSender::sendTouchEvent):
400 * DumpRenderTree/qt/EventSenderQt.h:
402 * DumpRenderTree/qt/ImageDiff.pro:
403 * DumpRenderTree/qt/main.cpp:
405 * QtTestBrowser/QtTestBrowser.pro:
406 * QtTestBrowser/cookiejar.cpp:
407 (TestBrowserCookieJar::TestBrowserCookieJar):
408 * QtTestBrowser/launcherwindow.cpp:
409 (LauncherWindow::setDiskCache):
410 * qmake/configure.pri:
411 * qmake/mkspecs/features/default_post.prf:
412 * qmake/mkspecs/features/default_pre.prf:
413 * qmake/mkspecs/features/features.prf:
414 * qmake/mkspecs/features/functions.prf:
415 * qmake/mkspecs/features/qtwebkit.prf: Removed.
416 * qmake/mkspecs/features/unix/default_pre.prf:
418 2012-08-07 Andras Becsi <andras.becsi@nokia.com>
420 [Qt][WK2] MiniBrowser should only synthesize multiple touch points if Ctrl is pressed
421 https://bugs.webkit.org/show_bug.cgi?id=93277
423 Reviewed by Jocelyn Turcotte.
425 MiniBrowser's multi-touch mocking does not check if Ctrl is pressed
426 when registering multiple touch points based on pressed mouse buttons
427 and sends inconsistent touch events to the WebView which triggers an
428 assert in the pinch gesture recognizer.
429 MiniBrowserApplication::updateTouchPoint should only register multiple
430 touch points if the Ctrl button is pressed else it should ignore the
433 * MiniBrowser/qt/MiniBrowserApplication.cpp:
434 (MiniBrowserApplication::updateTouchPoint):
436 2012-08-07 Csaba Osztrogonác <ossy@webkit.org>
438 REGRESSION(r124800): It broke NRWT result parsing of build.webkit.org
439 https://bugs.webkit.org/show_bug.cgi?id=93346
441 Rubber-stamped by Simon Hausmann.
443 Change back "unexpected" to "Unexpected", because master.cfg's results parser expects it.
445 * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
446 (MainTest.test_retrying_and_flaky_tests):
447 * Scripts/webkitpy/layout_tests/views/printing.py:
448 (Printer._print_unexpected_results):
450 2012-08-06 Balazs Kelemen <kbalazs@webkit.org>
452 [NRWT] runs platform specific tests that it shouldn't with --force
453 https://bugs.webkit.org/show_bug.cgi?id=91089
455 Reviewed by Dirk Pranke.
457 Ignore other platform's directories from platform/.
459 * Scripts/webkitpy/layout_tests/port/base.py:
460 (Port._expanded_paths):
461 * Scripts/webkitpy/layout_tests/port/test.py:
462 * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
463 (MainTest.test_platform_tests_are_found):
464 Updated integration test in accordance to the new behavior.
466 2012-08-06 Gyuyoung Kim <gyuyoung.kim@samsung.com>
468 [EFL] Enable link prefetch
469 https://bugs.webkit.org/show_bug.cgi?id=93281
471 Reviewed by Kentaro Hara.
473 * Scripts/webkitperl/FeatureList.pm: Enable LINK_PREFETCH feature.
475 2012-08-06 Sheriff Bot <webkit.review.bot@gmail.com>
477 Unreviewed, rolling out r124801.
478 http://trac.webkit.org/changeset/124801
479 https://bugs.webkit.org/show_bug.cgi?id=93338
481 It broke NRWT (Requested by Ossy on #webkit).
483 * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
484 (SingleTestRunner._compare_image):
485 (SingleTestRunner._compare_output_with_reference):
486 * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
488 * Scripts/webkitpy/layout_tests/controllers/test_result_writer_unittest.py:
489 (TestResultWriterTest.test_reftest_diff_image.ImageDiffTestPort.diff_image):
490 (TestResultWriterTest):
491 * Scripts/webkitpy/layout_tests/port/base.py:
493 * Scripts/webkitpy/layout_tests/port/chromium.py:
494 (ChromiumPort.diff_image):
495 * Scripts/webkitpy/layout_tests/port/chromium_port_testcase.py:
496 (ChromiumPortTestCase.test_diff_image):
497 * Scripts/webkitpy/layout_tests/port/image_diff.py:
498 (ImageDiffer.diff_image):
500 * Scripts/webkitpy/layout_tests/port/image_diff_unittest.py:
501 (TestImageDiffer.test_diff_image):
502 * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
503 (MockDRTPortTest.test_diff_image):
504 * Scripts/webkitpy/layout_tests/port/port_testcase.py:
505 (PortTestCase.test_diff_image):
506 * Scripts/webkitpy/layout_tests/port/server_process_mock.py:
507 (MockServerProcess.__init__):
508 * Scripts/webkitpy/layout_tests/port/test.py:
509 (TestPort.diff_image):
510 * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
511 (MainTest.test_tolerance.ImageDiffTestPort.diff_image):
513 2012-08-06 Wei James <james.wei@intel.com>
515 [Chromium]duplicated command line options in Android LayoutTest
516 https://bugs.webkit.org/show_bug.cgi?id=93233
518 Reviewed by Tony Chang.
520 Duplicated options found in Android layout test command line:
521 --encode-binary and --enable-hardware-gpu.
523 If there are multiple ChromiumAndroidPort instances,
524 these two options will be appended for multiple times.
526 * Scripts/webkitpy/layout_tests/port/base.py:
527 (Port.additional_drt_flag):
528 * Scripts/webkitpy/layout_tests/port/chromium_android.py:
529 (ChromiumAndroidPort.__init__):
530 (ChromiumAndroidPort.additional_drt_flag):
531 * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
532 (ChromiumAndroidTwoPortsTest):
533 (ChromiumAndroidTwoPortsTest.test_options_with_two_ports):
534 * Scripts/webkitpy/layout_tests/port/driver.py:
537 2012-08-03 Brady Eidson <beidson@apple.com>
539 Out-of-process plug-ins should support asynchronous initialization
540 <rdar://problem/10598594> and https://bugs.webkit.org/show_bug.cgi?id=92919
542 Reviewed by Anders Carlsson.
544 Add a plug-in with an NPP_New that takes 550ms (a reasonable trade-off between a solid test and a slow running test)
545 for testing asynchronous plug-in initialization.
547 * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
548 * DumpRenderTree/TestNetscapePlugIn/Tests/SlowNPPNew.cpp: Copied from Source/WebKit2/Shared/Plugins/PluginProcessCreationParameters.h.
550 (SlowNPPNew::SlowNPPNew):
551 (SlowNPPNew::NPP_New):
553 2012-08-06 Xianzhu Wang <wangxianzhu@chromium.org>
555 [Chromium-Android] Virtual test suites fail
556 https://bugs.webkit.org/show_bug.cgi?id=92515
558 Reviewed by Dirk Pranke.
560 The failure is because our bypassing of DriverProxy.
561 Repeat the logic of virtual tests in DriverProxy in ChromiumAndroidDriver
562 and restart DRT when the command line changes.
564 * Scripts/webkitpy/layout_tests/port/chromium_android.py:
565 (ChromiumAndroidDriver.__init__):
566 (ChromiumAndroidDriver.cmd_line): Added comments about why we override this and have _drt_cmd_line().
567 (ChromiumAndroidDriver.run_test): Repeat the logic of virtual tests in DriverProxy.
568 (ChromiumAndroidDriver.start): Restart DRT when the command line changes.
569 (ChromiumAndroidDriver._start_once):
571 2012-08-06 Xianzhu Wang <wangxianzhu@chromium.org>
573 Remove NRWT --shard-ref-tests
574 https://bugs.webkit.org/show_bug.cgi?id=91539
576 This is basically a revert of "[Chromium-Android] Run ref tests together to avoid expensive driver restarts"
577 (https://bugs.webkit.org/show_bug.cgi?id=91533, http://trac.webkit.org/changeset/122914),
578 with some conflicts resolved (because of refactory of Manager/LayoutTestRunner/Sharder classes).
580 Reviewed by Dirk Pranke.
582 * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
583 (LayoutTestRunner.run_tests):
584 (Sharder.shard_tests):
585 (Sharder._shard_in_two):
586 (Sharder._shard_by_directory):
587 * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
589 (SharderTests.get_test_input):
590 (SharderTests.get_shards):
591 (SharderTests.test_shard_by_dir):
592 (SharderTests.test_shard_in_two):
593 * Scripts/webkitpy/layout_tests/controllers/manager.py:
594 (Manager._test_input_for_file):
595 (Manager._test_is_slow):
596 * Scripts/webkitpy/layout_tests/port/chromium_android.py:
597 (ChromiumAndroidPort.__init__):
598 * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
601 2012-08-06 Luciano Wolf <luciano.wolf@openbossa.org>
603 [Qt] Default sizes for input-text and text-area are different when running DRT/WTR
604 https://bugs.webkit.org/show_bug.cgi?id=91990
606 Reviewed by Kenneth Rohde Christiansen.
608 Initializing TestFonts information into WTR binary. Doing it
609 later leads to wrong font names usage (Dejavu Serif instead
610 of Liberation Serif).
612 * WebKitTestRunner/Target.pri:
613 * WebKitTestRunner/qt/main.cpp:
616 2012-08-06 Jeff Timanus <twiz@chromium.org>
618 Add twiz@{chromium|google}.org to the set of non-committer contributors.
619 https://bugs.webkit.org/show_bug.cgi?id=93288
621 Reviewed by Adam Barth.
623 * Scripts/webkitpy/common/config/committers.py:
625 2012-08-06 Adam Barth <abarth@webkit.org>
627 REGRESSION(124789): EWS errors out because --print option is missing
628 https://bugs.webkit.org/show_bug.cgi?id=93299
632 The --print option doesn't exist anymore. Let's try --quiet instead.
634 * Scripts/webkitpy/tool/steps/runtests.py:
636 * Scripts/webkitpy/tool/steps/runtests_unittest.py:
638 2012-08-06 Ryosuke Niwa <rniwa@webkit.org>
640 run-webkit-tests should have ability to add description to its JSON output
641 https://bugs.webkit.org/show_bug.cgi?id=93296
643 Reviewed by Dirk Pranke.
645 Add --description option.
647 * Scripts/webkitpy/performance_tests/perftestsrunner.py:
648 (PerfTestsRunner._parse_args):
649 (PerfTestsRunner._generate_and_show_results):
650 (PerfTestsRunner._generate_results_dict):
651 * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
652 (test_run_with_description):
654 2012-08-06 Dirk Pranke <dpranke@chromium.org>
656 nrwt: handle errors from image diff better
657 https://bugs.webkit.org/show_bug.cgi?id=92934
659 Reviewed by Ojan Vafai.
661 Currently if ImageDiff crashes, returns a weird exit code, or
662 produces any stderr output, it's basically swallowed. This
663 change ensures that we log errors to stderr, and also appends
664 the error to the stderr for the test (so it'll show up in
667 Most importantly, it'll cause diff_image() to fail and we'll
668 report ImageHashMismatch ... this may be kinda untrue, but I
669 think it's better than ignoring the error.
671 * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
672 (SingleTestRunner._compare_image):
673 (SingleTestRunner._compare_output_with_reference):
674 * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
676 * Scripts/webkitpy/layout_tests/controllers/test_result_writer_unittest.py:
677 (TestResultWriterTest.test_reftest_diff_image.ImageDiffTestPort.diff_image):
678 (TestResultWriterTest):
679 * Scripts/webkitpy/layout_tests/port/base.py:
681 * Scripts/webkitpy/layout_tests/port/chromium.py:
682 (ChromiumPort.diff_image):
683 * Scripts/webkitpy/layout_tests/port/chromium_port_testcase.py:
684 (ChromiumPortTestCase.test_diff_image_crashed):
685 * Scripts/webkitpy/layout_tests/port/driver.py:
687 * Scripts/webkitpy/layout_tests/port/image_diff.py:
688 (ImageDiffer.diff_image):
690 * Scripts/webkitpy/layout_tests/port/image_diff_unittest.py:
691 (TestImageDiffer.test_diff_image):
692 * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
693 (MockDRTPortTest.test_diff_image_crashed):
694 * Scripts/webkitpy/layout_tests/port/port_testcase.py:
695 (PortTestCase.test_diff_image):
697 (PortTestCase.test_diff_image_crashed):
698 (PortTestCase.test_diff_image_crashed.make_proc):
699 * Scripts/webkitpy/layout_tests/port/server_process_mock.py:
700 (MockServerProcess.__init__):
701 * Scripts/webkitpy/layout_tests/port/test.py:
702 (TestPort.diff_image):
703 * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
704 (MainTest.test_tolerance.ImageDiffTestPort.diff_image):
706 2012-08-06 Dirk Pranke <dpranke@chromium.org>
708 nrwt: clean up printing.py
709 https://bugs.webkit.org/show_bug.cgi?id=93026
711 Reviewed by Dirk Pranke.
713 This patch cleans up the implementation of printing.py now
714 that we're not using all the original complexity.
716 There should be no changes in functionality and everything
717 should be covered by the existing tests.
719 * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
720 (LayoutTestRunner.run_tests):
721 * Scripts/webkitpy/layout_tests/views/printing.py:
724 (Printer.print_config):
725 (Printer.print_found):
726 (Printer.print_expected):
727 (Printer.print_workers_and_shards):
729 (Printer._print_expected_results_of_type):
730 (Printer.print_results):
731 (Printer._print_timing_statistics):
732 (Printer._print_aggregate_test_statistics):
733 (Printer._print_individual_test_times):
734 (Printer._print_test_list_timing):
735 (Printer._print_directory_timings):
736 (Printer._print_statistics_for_test_timings):
737 (Printer._print_result_summary):
738 (Printer._print_result_summary_entry):
739 (Printer._print_one_line_summary):
740 (Printer.print_finished_test):
741 (Printer._print_test_result):
742 (Printer._print_test_trace):
743 (Printer._print_baseline):
744 (Printer._print_unexpected_test_result):
745 (Printer._print_progress):
746 (Printer._print_unexpected_results):
747 (Printer._print_unexpected_results.add_result):
748 (Printer._print_quiet):
749 (Printer._print_default):
750 (Printer._print_debug):
751 (Printer._print_for_bot):
752 (Printer.write_update):
754 * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
755 (Testprinter.get_printer):
756 (Testprinter.test_print_config):
757 (Testprinter.test_print_one_line_summary):
758 (Testprinter.test_print_unexpected_results):
761 2012-08-06 Dirk Pranke <dpranke@chromium.org>
763 change bots to pass --debug-rwt-logging instead of --verbose to new-run-webkit-tests
764 https://bugs.webkit.org/show_bug.cgi?id=93043
766 Reviewed by Ryosuke Niwa.
768 in preparation for the cleanup of the logging flags in new-run-webkit-tests.
770 * Scripts/run-webkit-tests:
772 2012-08-06 Dirk Pranke <dpranke@chromium.org>
774 nrwt: clean up logging, part 1
775 https://bugs.webkit.org/show_bug.cgi?id=93018
777 Reviewed by Ojan Vafai.
779 remove --print, --help-printing from nrwt, add three new options:
780 --quiet, which just logs warnings,errors, and unexpected results
781 --debug-rwt-logging, a new name for --verbose
782 --details, to get the old --print trace-everything behavior
784 This patch does not implement the new "one line per test"
785 --verbose behavior specified in bug 88702, and there's a bunch
786 of internal cleanup I can do in printing.py that I'll defer to
787 a later patch to make things easier to review.
789 This patch deletes a lot of unit tests that are no longer
790 necessary since there aren't so many logging combinations.
792 * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
795 * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
798 (StreamTestingMixin.assertContains):
799 (MainTest.test_child_processes_2):
800 (MainTest.test_child_processes_min):
801 (MainTest.test_full_results_html):
802 (MainTest.test_no_tests_found):
803 (MainTest.test_no_tests_found_2):
804 (MainTest.test_repeat_each_iterations_num_tests):
805 (MainTest.test_additional_platform_directory):
806 (RebaselineTest.assertBaselines):
807 * Scripts/webkitpy/layout_tests/views/printing.py:
811 (Printer.print_unexpected_results):
812 * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
813 (TestUtilityFunctions.test_print_options):
814 (Testprinter.get_result_summary):
815 (Testprinter.test_configure_and_cleanup):
816 (Testprinter.test_print_config):
817 (Testprinter.test_print_one_line_summary):
818 (Testprinter.test_print_unexpected_results):
824 2012-08-06 Xianzhu Wang <wangxianzhu@chromium.org>
826 [Chromium-Android] Store test executable, data and fonts in /data/local/tmp
827 https://bugs.webkit.org/show_bug.cgi?id=91910
829 /data/local/tmp is a directory with permission 0777 by default.
830 Place DumpRenderTree resources here so that both the native test app and
831 the NRWT script can access them even if adb shell isn't running as root.
833 Reviewed by Adam Barth.
835 * DumpRenderTree/chromium/TestShellAndroid.cpp:
836 (createFIFO): Set fifo mode to 0666 to allow unrooted adb shell to access.
837 * Scripts/webkitpy/layout_tests/port/chromium_android.py:
838 (ChromiumAndroidPort.driver_cmd_line):
839 (ChromiumAndroidDriver.__init__):
840 (ChromiumAndroidDriver._setup_test):
841 (ChromiumAndroidDriver._push_executable):
842 * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
843 (ChromiumAndroidPortTest.make_port):
844 (ChromiumAndroidPortTest.test_expectations_files):
845 (ChromiumAndroidDriverTest.test_drt_cmd_line):
847 2012-08-06 Abhishek Arya <inferno@chromium.org>
849 [Chromium] Re-expose layoutTestController as various fuzzers depend on it
850 https://bugs.webkit.org/show_bug.cgi?id=93282
852 Reviewed by Ryosuke Niwa.
854 Unfortunately, various fuzzers used in the Chromium project still depends on
855 window.layoutTestController. Re-expose the object while they're making the transition.
857 * DumpRenderTree/chromium/TestShell.cpp:
858 (TestShell::bindJSObjectsToWindow):
860 2012-08-06 Balazs Kelemen <kbalazs@webkit.org>
862 WTR should be able to load external resources
863 https://bugs.webkit.org/show_bug.cgi?id=89382
865 Reviewed by Ryosuke Niwa.
867 Allow to load an external resource as the main frame
868 and allow all subsequent external loads for such a main
869 frame. This behavior is necessary for being able to run
870 performance tests (https://bugs.webkit.org/show_bug.cgi?id=84008).
872 * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
875 (WTR::isHTTPOrHTTPSScheme):
876 (WTR::InjectedBundlePage::willSendRequestForFrame):
878 2012-08-06 Balazs Kelemen <kbalazs@webkit.org>
880 [WK2] REGRESSION(124596) lot of web process unresponsiveness appears on Apple bots
881 https://bugs.webkit.org/show_bug.cgi?id=93120
883 Reviewed by Csaba Osztrogonác.
885 Set the short timeout back to 15 seconds and add more time to WebKitTestRunner
886 so it will be able to detect an unresponsive web process.
888 * Scripts/webkitpy/layout_tests/port/base.py:
889 (Port.default_timeout_ms):
890 * WebKitTestRunner/TestController.cpp:
893 2012-08-06 Carlos Garcia Campos <cgarcia@igalia.com>
895 Unreviewed. Fix make distcheck.
897 * GNUmakefile.am: Add jhbuildutils.py to EXTRA_DIST.
899 2012-08-06 Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com>
901 [EFL] EFL's LayoutTestController needs removeAllVisitedLinks implementation
902 https://bugs.webkit.org/show_bug.cgi?id=82724
904 Reviewed by Csaba Osztrogonác.
906 * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
907 (LayoutTestController::removeAllVisitedLinks):
909 2012-08-06 Mario Sanchez Prada <msanchez@igalia.com>
911 [WK2][GTK] Implement a new spell checker API for WebKit2GTK+
912 https://bugs.webkit.org/show_bug.cgi?id=90268
914 Reviewed by Martin Robinson.
916 Ignore WebKitTextChecker.* private files for gtk-doc.
918 * gtk/generate-gtkdoc:
919 (get_webkit2_options):
921 2012-08-06 Gyuyoung Kim <gyuyoung.kim@samsung.com>
923 Unreviewed. Roll out r124728 because of build bot test fail.
925 * Scripts/webkitpy/common/config/watchlist:
927 2012-08-05 Gyuyoung Kim <gyuyoung.kim@samsung.com>
929 Unreviewed. Change my email address for watch list.
931 * Scripts/webkitpy/common/config/watchlist:
933 2012-08-04 Nico Weber <thakis@chromium.org>
935 [Chromium Mac] Tests are exiting early due to failures to load missingImage.png
936 https://bugs.webkit.org/show_bug.cgi?id=93186
938 Reviewed by Ryosuke Niwa.
940 Add missingImage.png to DumpRenderTree.app's Resource folder.
941 Depends on http://crrev.com/150037 rolling into webkit first.
943 * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
945 2012-08-03 Ryosuke Niwa <rniwa@webkit.org>
947 Stop exposing window.layoutTestController in layout tests in favor of window.testRunner
948 https://bugs.webkit.org/show_bug.cgi?id=93173
950 Reviewed by Adam Barth.
952 Don't expose window.layoutTestController since all layout tests use window.testRunner
953 instead of window.layoutTestController as of r124685.
955 We can safely rename classes and files to use TestRunner once this patch is landed.
957 * DumpRenderTree/LayoutTestController.cpp:
958 (LayoutTestController::makeWindowObject):
959 * DumpRenderTree/TestNetscapePlugIn/PluginObject.cpp:
960 (notifyTestCompletion):
961 * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
962 (PluginTest::waitUntilDone):
963 (PluginTest::notifyDone):
964 * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:
965 (ConvertPoint::NPP_SetWindow):
966 * DumpRenderTree/TestNetscapePlugIn/main.cpp:
970 * DumpRenderTree/chromium/TestShell.cpp:
971 (TestShell::bindJSObjectsToWindow):
972 * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
973 (WebCore::DumpRenderTree::initJSObjects):
975 * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
976 (WTR::LayoutTestController::makeWindowObject):
978 2012-08-04 Ilya Tikhonovsky <loislo@chromium.org>
980 Unreviewed: Web Inspector: extend instrumenting methods set in ReportMemoryUsage clang plugin
981 Three methods addString, addVectorPtr and addInstrumentedVectorPtr were added to the list of instrumentation methods.
983 * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
984 (clang::ReportMemoryUsageConsumer::ReportMemoryUsageConsumer):
986 2012-08-03 Yaron Friedman <yfriedman@chromium.org>
988 [Chrome-Android] - Prepare apk tests for switch to checked in SDK.
989 https://bugs.webkit.org/show_bug.cgi?id=92931
991 Reviewed by Adam Barth.
993 Pass Android-specific gyp variables to the native test generator,
994 avoiding any dependencies on environment variables during build time.
996 * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
997 * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
999 2012-08-03 Rick Byers <rbyers@chromium.org>
1001 Double tap gesture should send dblclick event
1002 https://bugs.webkit.org/show_bug.cgi?id=92412
1004 Reviewed by Adam Barth.
1006 Enable EventSender to set the tap count on tap gesture events. The
1007 delta X/Y parameters were not actually being used anymore (used to be
1008 used for radius information).
1010 * DumpRenderTree/chromium/TestRunner/EventSender.cpp:
1011 (EventSender::gestureEvent):
1013 2012-08-03 Ryosuke Niwa <rniwa@webkit.org>
1015 run-perf-tests should generate JSON output and results page by default
1016 https://bugs.webkit.org/show_bug.cgi?id=93042
1018 Reviewed by Eric Seidel.
1020 Generate results JSON and page named PerfTestResults.json and PerfTestResults.html by default.
1022 * Scripts/webkitpy/layout_tests/port/base.py:
1023 (Port.perf_results_directory): Added.
1024 * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1025 (PerfTestsRunner): Added _DEFAULT_JSON_FILENAME.
1026 (PerfTestsRunner._parse_args): Added --no-results option in the case a user doens't want to generate
1027 results JSON or results page.
1028 (PerfTestsRunner.run):
1029 (PerfTestsRunner._generate_and_show_results): Extracted from run. Set the default json file path using
1030 port's perf_results_directory and call show_results_html_file at the end if the results page is generated.
1031 (PerfTestsRunner._generate_results_dict): Renamed from _generate_output to disambiguate it from
1032 _generate_and_show_results.
1033 (PerfTestsRunner._generate_output_files): Takes results page's path instead of a boolean indicating
1034 whether results page should be generated or not.
1035 * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1036 (create_runner_and_setup_results_template):
1037 (test_run_respects_results_output): Added.
1038 (test_run_generates_json_by_default): Added.
1039 (test_run_generates_and_show_results_page): Added a check to ensure show_results_html_file is called.
1041 2012-08-03 Sheriff Bot <webkit.review.bot@gmail.com>
1043 Unreviewed, rolling out r124628.
1044 http://trac.webkit.org/changeset/124628
1045 https://bugs.webkit.org/show_bug.cgi?id=93155
1047 Causes random crashes of DRT on Chromium bots (Requested by
1050 * DumpRenderTree/DumpRenderTree.gypi:
1051 * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp:
1052 (MockWebKitPlatformSupport::createMediaStreamCenter):
1053 * DumpRenderTree/chromium/MockWebKitPlatformSupport.h:
1054 (MockWebKitPlatformSupport):
1055 * DumpRenderTree/chromium/MockWebMediaStreamCenter.cpp:
1056 * DumpRenderTree/chromium/MockWebMediaStreamCenter.h:
1057 * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp: Removed.
1058 * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h: Removed.
1060 2012-08-03 Brady Eidson <beidson@apple.com>
1062 Small part of "Out-of-process plug-ins should support asynchronous initialization."
1063 <rdar://problem/10598594> and https://bugs.webkit.org/show_bug.cgi?id=92919
1065 Reviewed by Anders Carlsson.
1067 * WebKitTestRunner/TestController.cpp:
1068 (WTR::TestController::resetStateToConsistentValues): Reset all preferences related to asynchronous plugin initialization.
1070 2012-08-02 Jeffrey Pfau <jpfau@apple.com>
1072 Add API for enabling blanket third-party data blocking
1073 https://bugs.webkit.org/show_bug.cgi?id=93022
1075 Reviewed by Anders Carlsson.
1077 Added test for default setting of new third-party storage blocking API.
1079 * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
1080 (TestWebKitAPI::TEST):
1082 2012-08-03 Arvid Nilsson <anilsson@rim.com>
1084 Unreviewed, adding self to committers.py after becoming committer.
1086 * Scripts/webkitpy/common/config/committers.py:
1088 2012-08-03 Ádám Kallai <kadam@inf.u-szeged.hu>
1090 Limited the number of queried cached build infos in URLs used by Garden-o-matic.
1091 https://bugs.webkit.org/show_bug.cgi?id=92992.
1093 Reviewed by Adam Barth.
1095 I added a limit to the cached builds that are included in the URL to avoid too long URLs in Garden-o-matic.
1097 * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
1099 * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1101 2012-08-03 Tommy Widenflycht <tommyw@google.com>
1103 [chromium] MediaStream API: Add MockWebRTCPeerConnectionHandler
1104 https://bugs.webkit.org/show_bug.cgi?id=93091
1106 Reviewed by Adam Barth.
1108 Add a skeleton MockWebRTCPeerConnectionHandler to DumpRenderTree,
1109 to complete the infrastructure for RTCPeerConnection.
1111 * DumpRenderTree/DumpRenderTree.gypi:
1112 * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp:
1113 (MockWebKitPlatformSupport::createRTCPeerConnectionHandler):
1114 * DumpRenderTree/chromium/MockWebKitPlatformSupport.h:
1115 (MockWebKitPlatformSupport):
1116 * DumpRenderTree/chromium/MockWebMediaStreamCenter.cpp:
1117 * DumpRenderTree/chromium/MockWebMediaStreamCenter.h:
1118 * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp: Copied from Tools/DumpRenderTree/chromium/MockWebKitPlatformSupport.h.
1119 (MockWebRTCPeerConnectionHandler::MockWebRTCPeerConnectionHandler):
1120 (MockWebRTCPeerConnectionHandler::initialize):
1121 * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h: Copied from Tools/DumpRenderTree/chromium/MockWebKitPlatformSupport.h.
1123 (MockWebRTCPeerConnectionHandler):
1124 (MockWebRTCPeerConnectionHandler::MockWebRTCPeerConnectionHandler):
1126 2012-08-03 Hugo Parente Lima <hugo.lima@openbossa.org>
1128 [Qt][WK2] There's no way to test the gesture tap on WTR
1129 https://bugs.webkit.org/show_bug.cgi?id=92895
1131 Reviewed by Kenneth Rohde Christiansen.
1133 WTR doesn't create the QQuickItem from C++, not from QML, so a call
1134 to componentComplete() was added to mimic the QML behaviour.
1136 * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
1137 (WTR::PlatformWebView::PlatformWebView):
1139 2012-08-03 Balazs Kelemen <kbalazs@webkit.org>
1141 Unreviewed, rolling out r124567.
1142 http://trac.webkit.org/changeset/124567
1143 https://bugs.webkit.org/show_bug.cgi?id=89382
1145 Broke some tests with external resources
1147 * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1148 (WTR::InjectedBundlePage::willSendRequestForFrame):
1150 2012-08-03 Philippe Normand <pnormand@igalia.com>
1152 Unreviewed, rolling out r124614.
1153 http://trac.webkit.org/changeset/124614
1154 https://bugs.webkit.org/show_bug.cgi?id=91727
1156 gstreamer core .po files mess up the build again
1158 * Scripts/webkitdirs.pm:
1159 (buildAutotoolsProject):
1160 * gtk/jhbuild.modules:
1162 2012-08-03 Philippe Normand <pnormand@igalia.com>
1164 [GTK][jhbuild] Switch to GStreamer 0.11 build
1165 https://bugs.webkit.org/show_bug.cgi?id=91727
1167 Reviewed by Gustavo Noronha Silva.
1169 Switch build-webkit --gtk to GStreamer 0.11 support and build the
1170 necessary GStreamer git modules from JHBuild.
1172 * Scripts/webkitdirs.pm:
1173 (buildAutotoolsProject): Build WebKit with GStreamer 0.11 support.
1174 * gtk/jhbuild.modules: GStreamer build support.
1176 2012-08-03 Nico Weber <thakis@chromium.org>
1178 [chromium mac] DumpRenderTree compile fails with warning/error in LayoutTestHelper.mm with 10.7sdk
1179 https://bugs.webkit.org/show_bug.cgi?id=92820
1181 Reviewed by Jochen Eisinger.
1183 When building with the 10.7 SDK, use newer functions to switch color
1184 profiles. Note that these newer functions aren't available on 10.6,
1185 but we don't intend to ship DRT to users, and no bots that currently
1186 build with the 10.7 SDK ship their binaries to 10.6 testers.
1188 The new code was copied from Apple's DRT/mac/LayoutTestHelper.m.
1190 * DumpRenderTree/chromium/LayoutTestHelper.mm:
1191 (installLayoutTestColorProfile):
1192 (restoreUserColorProfile):
1193 (saveCurrentColorProfile):
1195 2012-08-03 Benjamin Poulain <benjamin@webkit.org>
1197 StringImpl created from literal should be BufferInternal
1198 https://bugs.webkit.org/show_bug.cgi?id=92940
1200 Reviewed by Anders Carlsson.
1202 Add tests for the contruction of strings from literal.
1204 * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1205 * TestWebKitAPI/Tests/WTF/AtomicString.cpp: Added.
1207 (TestWebKitAPI::TEST):
1208 * TestWebKitAPI/Tests/WTF/StringImpl.cpp: Added.
1210 (TestWebKitAPI::TEST):
1212 2012-08-03 Balazs Kelemen <kbalazs@webkit.org>
1214 [nrwt] fix unit tests after turned pixel testing to be a per test setting
1215 https://bugs.webkit.org/show_bug.cgi?id=93112
1217 Reviewed by Csaba Osztrogonác.
1219 Change unit tests in accordance of per test switching of pixel testing.
1221 * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1222 (ChromiumAndroidDriverTest.test_command_from_driver_input):
1223 * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
1224 (DriverTest.test_no_timeout):
1226 2012-08-03 Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
1228 [Qt] Populate .qmake.cache from the top-level project file
1230 This makes it possible to build webkit without using the perl build
1231 script. The duplicated logic in build-webkit will be removed when we
1232 remove the Qt4 code paths. The build-webkit script will then simply
1233 call 'qmake WebKit.pro' from WEBKITOUTPUTDIR.
1235 Reviewed by Tor Arne Vestbø.
1237 * qmake/mkspecs/features/win32/default_pre.prf:
1239 2012-08-03 Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
1241 [Qt] Move build config default to build-webkit instead of default_pre
1243 Allows the qmake buildsystem have its own default.
1245 Reviewed by Tor Arne Vestbø.
1247 * Scripts/webkitdirs.pm:
1248 (buildQMakeProjects):
1249 * qmake/mkspecs/features/default_pre.prf:
1251 2012-08-03 Mario Sanchez Prada <msanchez@igalia.com>
1253 Unreviewed build fix for GTK after r124581.
1255 * DumpRenderTree/gtk/DumpRenderTree.cpp:
1256 (runTest): Fix wrong variable name in ASSERT.
1258 2012-08-03 Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
1260 [Qt] Always save CONFIG & DEFINES in default_pre
1262 Always doing it is cheaper than checking whether it needs to be done.
1263 apart from that, it would break configuration initiated from another
1266 Reviewed by Tor Arne Vestbø.
1268 * qmake/mkspecs/features/default_pre.prf:
1270 2012-08-03 Joone Hur <joone.hur@intel.com>
1272 [EFL][DRT] fast/loader/stop-provisional-loads.html fails
1273 https://bugs.webkit.org/show_bug.cgi?id=92219
1275 Reviewed by Kenneth Rohde Christiansen.
1277 This patch adds the ability to tell DRT to call ewk_frame_stop
1278 inside of a didStartProvisionalLoadForFrame signal handler.
1279 It allows to pass fast/loader/stop-provisional-loads.html.
1281 * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1282 (DumpRenderTreeChrome::onFrameProvisionalLoad):
1284 2012-08-03 Balazs Kelemen <kbalazs@webkit.org>
1286 [WTR] process unresponsiveness is reported to the wrong test
1287 https://bugs.webkit.org/show_bug.cgi?id=88404
1289 2nd unreviewed buildfix.
1291 * WebKitTestRunner/TestInvocation.cpp:
1293 2012-08-03 Balazs Kelemen <kbalazs@webkit.org>
1295 [WTR] process unresponsiveness is reported to the wrong test
1296 https://bugs.webkit.org/show_bug.cgi?id=88404
1298 Unreviewed buildfix. Also made the change on the array size
1299 suggested on review because I forgot it when landed.
1301 * WebKitTestRunner/TestInvocation.cpp:
1302 (WTR::TestInvocation::invoke):
1304 2012-08-03 Balazs Kelemen <kbalazs@webkit.org>
1306 [WTR] process unresponsiveness is reported to the wrong test
1307 https://bugs.webkit.org/show_bug.cgi?id=88404
1309 Reviewed by Zoltan Herczeg.
1311 Do resetting to consistent state after the finished test,
1312 not before the next. This way we can report unresponsiveness
1313 to the test that triggered it, and we do a restart before the
1314 next test so it will start in a functional state.
1316 * Scripts/webkitpy/layout_tests/port/webkit.py:
1317 (WebKitDriver._check_for_driver_crash):
1318 Print the error line for unresponsive web process, otherwise
1319 it's hard to tell what happened.
1320 * WebKitTestRunner/TestController.cpp:
1322 Decreased the default short timeout so we can detect unresponsiveness
1323 before the timer of the test harness times out. It was unnecessarily
1324 high anyway, we should get answer from the web process in those cases
1325 when the short timeout is used in less than a second.
1326 (WTR::TestController::runTest):
1327 (WTR::TestController::run):
1328 * WebKitTestRunner/TestController.h:
1330 * WebKitTestRunner/TestInvocation.cpp:
1331 (WTR::TestInvocation::invoke):
1332 (WTR::TestInvocation::dump):
1333 (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
1334 * WebKitTestRunner/TestInvocation.h:
1335 (TestInvocation): Make the dump function static as it doesn't need an object.
1337 2012-08-03 Sudarsana Nagineni <sudarsana.nagineni@linux.intel.com>
1339 [EFL] (REGRESSION 124581) : EFL 64-bit Debug bot is red and build is exiting early after 20 test crashes
1340 https://bugs.webkit.org/show_bug.cgi?id=93084
1342 Reviewed by Csaba Osztrogonác.
1344 Fix regression introduced in r124581.
1346 * DumpRenderTree/efl/DumpRenderTree.cpp:
1347 (getFinalTestURL): Convert the path into a full file URL.
1350 2012-08-03 Balazs Kelemen <kbalazs@webkit.org>
1352 All ports should support per test switching of pixel testing
1353 https://bugs.webkit.org/show_bug.cgi?id=92398
1355 Reviewed by Dirk Pranke.
1357 Teach all test drivers to accept a per test control of whether
1358 to dump pixels. Drivers now accept a -p/--pixel-test argument
1359 on the standart input that means that running the current
1360 test as pixel test is allowed (even if the expected hash is
1361 missing). Removed the --pixel-tests command line option since
1362 there is no need for it anymore.
1364 * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1365 * DumpRenderTree/DumpRenderTree.gypi:
1366 * DumpRenderTree/DumpRenderTree.h:
1367 (TestCommand::TestCommand):
1369 * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1370 * DumpRenderTree/DumpRenderTreeCommon.cpp: Added.
1372 (CommandTokenizer::CommandTokenizer):
1373 (CommandTokenizer::pump):
1374 (CommandTokenizer::next):
1375 (CommandTokenizer::hasNext):
1378 Common logic to parse the input line from the standard input (or from the command line
1379 in standalone mode). Made it somewhat general so we can easily add more arguments if
1382 * DumpRenderTree/chromium/DumpRenderTree.cpp:
1386 * DumpRenderTree/chromium/TestRunner/AccessibilityControllerChromium.cpp: Renamed from Tools/DumpRenderTree/chromium/TestRunner/AccessibilityController.cpp.
1387 * DumpRenderTree/chromium/TestRunner/AccessibilityControllerChromium.h: Renamed from Tools/DumpRenderTree/chromium/TestRunner/AccessibilityController.h.
1388 * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp: Renamed from Tools/DumpRenderTree/chromium/TestRunner/AccessibilityUIElement.cpp.
1389 * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.h: Renamed from Tools/DumpRenderTree/chromium/TestRunner/AccessibilityUIElement.h.
1390 Renamed these files with a Chromium suffix because they were clashing with the common ones used by other ports after adding the root
1391 DumpRenderTree directory to the include path.
1393 * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
1394 * DumpRenderTree/chromium/TestShell.cpp:
1395 (TestShell::TestShell):
1396 (TestShell::runFileTest):
1398 * DumpRenderTree/chromium/TestShell.h:
1399 (TestParams::TestParams):
1401 * DumpRenderTree/efl/CMakeLists.txt:
1402 * DumpRenderTree/efl/DumpRenderTree.cpp:
1403 (parseCommandLineOptions):
1405 (shouldDumpPixelsAndCompareWithExpected):
1407 * DumpRenderTree/gtk/DumpRenderTree.cpp:
1408 (initializeGlobalsFromCommandLineOptions):
1412 * DumpRenderTree/mac/DumpRenderTree.mm:
1413 (initializeGlobalsFromCommandLineOptions):
1417 * DumpRenderTree/qt/DumpRenderTree.pro:
1418 * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1419 (WebCore::DumpRenderTree::DumpRenderTree):
1420 (WebCore::DumpRenderTree::open):
1421 (WebCore::DumpRenderTree::processLine):
1422 (WebCore::DumpRenderTree::dump):
1423 * DumpRenderTree/qt/DumpRenderTreeQt.h:
1425 * DumpRenderTree/qt/main.cpp:
1429 * DumpRenderTree/win/DumpRenderTree.cpp:
1432 (dllLauncherEntryPoint):
1433 * DumpRenderTree/win/DumpRenderTree.vcproj:
1434 * DumpRenderTree/wscript:
1435 * DumpRenderTree/wx/DumpRenderTreeWx.cpp:
1441 * Scripts/old-run-webkit-tests:
1442 * Scripts/old-run-webkit-tests:
1443 Pass --pixel-test before the hash.
1445 * Scripts/webkitpy/layout_tests/port/base.py:
1446 (Port.should_run_as_pixel_test): Removed now that all ports supports it.
1447 * Scripts/webkitpy/layout_tests/port/driver.py:
1448 (Driver.cmd_line): Don't pass --pixel-tests anymore.
1449 (Driver._command_from_driver_input):
1450 * Scripts/webkitpy/layout_tests/port/qt.py:
1451 (QtPort.should_run_as_pixel_test):
1452 * Scripts/webkitpy/layout_tests/port/test.py:
1453 (TestPort.should_run_as_pixel_test):
1455 * WebKitTestRunner/TestController.cpp:
1456 (WTR::TestController::TestController):
1457 (WTR::TestController::initialize):
1458 (WTR::TestCommand::TestCommand):
1462 (WTR::CommandTokenizer::CommandTokenizer):
1463 (WTR::CommandTokenizer::pump):
1464 (WTR::CommandTokenizer::next):
1465 (WTR::CommandTokenizer::hasNext):
1467 (WTR::parseInputLine):
1468 (WTR::TestController::runTest):
1469 * WebKitTestRunner/TestController.h:
1471 Duplicate the logic for parsing the command line because
1472 unfortunately there is no way to share code between
1473 WebKitTestRunner and DumpRenderTree.
1475 2012-08-03 Ilya Tikhonovsky <loislo@chromium.org>
1477 Web Inspector: clang plugin for checking native memory instrumentation coverage.
1478 https://bugs.webkit.org/show_bug.cgi?id=92650
1480 Reviewed by Yury Semikhatsky.
1482 This is the clang plugin for checking native memory instrumentation coverage.
1483 The idea: InspectorMemoryAgent traverses through instrumented objects and collects pointers and their sizes and types.
1484 It is doing that with help of reportMemoryUsage function that needs to be implemented in each, significant
1485 from native memory footprint point of view, WebCore class.
1486 This plugin checks that the list of class members is matching with the list of members reported in reportMemoryUsage function
1487 and makes a warning for the each member that needs to be reported.
1489 * clang/ReportMemoryUsagePlugin/CMakeLists.txt: Added.
1490 * clang/ReportMemoryUsagePlugin/Makefile: Added.
1491 * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp: Added.
1493 (AddMemberCallVisitor):
1494 (clang::AddMemberCallVisitor::VisitCallExpr):
1495 (clang::AddMemberCallVisitor::instrumentedMembers):
1496 (ReportMemoryUsageVisitor):
1497 (clang::ReportMemoryUsageVisitor::ReportMemoryUsageVisitor):
1498 (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
1499 (clang::ReportMemoryUsageVisitor::emitWarning):
1500 (clang::ReportMemoryUsageVisitor::findInstrumentationMethod):
1501 (clang::ReportMemoryUsageVisitor::needsToBeInstrumented):
1502 (clang::ReportMemoryUsageVisitor::CheckMembersCoverage):
1503 (ReportMemoryUsageConsumer):
1504 (clang::ReportMemoryUsageConsumer::ReportMemoryUsageConsumer):
1505 (clang::ReportMemoryUsageConsumer::HandleTranslationUnit):
1506 (ReportMemoryUsageAction):
1507 (clang::ReportMemoryUsageAction::CreateASTConsumer):
1508 (clang::ReportMemoryUsageAction::ParseArgs):
1509 * clang/ReportMemoryUsagePlugin/tests/Source/WebCore/Test.cpp: Added.
1511 (MemoryInstrumentation):
1513 (aNamespace::MemoryClassInfo::MemoryClassInfo):
1514 (aNamespace::MemoryClassInfo::addMember):
1515 (aNamespace::MemoryClassInfo::addInstrumentedMember):
1517 (aNamespace::OwnPtr::OwnPtr):
1519 (aNamespace::RefPtr::RefPtr):
1522 (NotInstrumentedClass):
1523 (InstrumentedClass):
1524 (aNamespace::InstrumentedClass::reportMemoryUsage):
1525 (InstrumentedChildClass):
1526 (aNamespace::InstrumentedChildClass::reportMemoryUsage):
1527 (InstrumentedChildChildClass):
1528 (InstrumentedChildChildClass::reportMemoryUsage):
1531 2012-08-03 Joone Hur <joone.hur@intel.com>
1533 [EFL][DRT] WebKitAnimation API compile-time disabled
1534 https://bugs.webkit.org/show_bug.cgi?id=84593
1536 Reviewed by Laszlo Gombos.
1538 Enable ENABLE_ANIMATION_API by default on the Efl port.
1540 * Scripts/webkitperl/FeatureList.pm:
1542 2012-08-03 Balazs Kelemen <kbalazs@webkit.org>
1544 WTR should be able to load external resources
1545 https://bugs.webkit.org/show_bug.cgi?id=89382
1547 Reviewed by Ryosuke Niwa.
1549 Allow to load an external resource as the main frame
1550 and allow all subsequent external loads for such a main
1551 frame. This behavior is necessary for being able to run
1552 performance tests (wkb.ug/84008).
1554 * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1557 (WTR::isHTTPOrHTTPSScheme):
1558 (WTR::InjectedBundlePage::willSendRequestForFrame):
1560 2012-08-02 Joone Hur <joone.hur@intel.com>
1562 [GTK] Build break when building DumpRenderTree/gtk/EditingCallbacks.cpp
1563 https://bugs.webkit.org/show_bug.cgi?id=93061
1565 Unreviewed, Fix build break for WebKitGtk+.
1567 Use webkit_dom_element_get_class_name instead of webkit_dom_html_element_get_class_name.
1569 * DumpRenderTree/gtk/EditingCallbacks.cpp:
1570 (shouldShowDeleteInterfaceForElement):
1572 2012-08-02 Dirk Pranke <dpranke@chromium.org>
1574 test-webkitpy can fail on a clean checkout
1575 https://bugs.webkit.org/show_bug.cgi?id=93039
1577 Reviewed by Adam Barth.
1579 Now that we run tests in parallel that can cause the installer
1580 to try and install packages in parallel (even the same package),
1581 and that probably won't work. We fix this by autoinstalling
1582 everything up front (serially).
1584 * Scripts/webkitpy/test/main.py:
1585 (Tester._run_tests):
1586 * Scripts/webkitpy/thirdparty/__init__.py:
1587 (autoinstall_everything):
1589 2012-06-15 David Barton <dbarton@mathscribe.com>
1591 MathML: nested square root symbols have varying descenders
1592 https://bugs.webkit.org/show_bug.cgi?id=43819
1594 Reviewed by Eric Seidel.
1596 Allow the STIXGeneral font to be used during layout tests, especially for MathML.
1598 * DumpRenderTree/mac/DumpRenderTree.mm:
1599 (allowedFontFamilySet):
1600 * WebKitTestRunner/InjectedBundle/mac/ActivateFonts.mm:
1601 (WTR::allowedFontFamilySet):
1603 2012-08-02 Dirk Pranke <dpranke@chromium.org>
1605 test-webkitpy: some tests need to run by themselves
1606 https://bugs.webkit.org/show_bug.cgi?id=92926
1608 Reviewed by Ojan Vafai.
1610 Due to timing issues some of the executive tests will collide
1611 and fail if they're run concurrently. This patch adds support
1612 for writing tests that will be executed one at a time
1613 (serially); to get them, add "serial_" to the front of the test
1616 * Scripts/webkitpy/common/system/executive_unittest.py:
1617 (ExecutiveTest.serial_test_kill_process):
1618 (ExecutiveTest.serial_test_kill_all):
1619 (ExecutiveTest.serial_test_check_running_pid):
1620 (ExecutiveTest.serial_test_running_pids):
1621 (ExecutiveTest.serial_test_run_in_parallel):
1622 * Scripts/webkitpy/test/main.py:
1623 (Tester._run_tests):
1624 (Tester._test_names):
1626 2012-08-02 Dirk Pranke <dpranke@chromium.org>
1628 test-webkitpy: integrate proper support for integration tests
1629 https://bugs.webkit.org/show_bug.cgi?id=92925
1631 Reviewed by Ojan Vafai.
1633 This patch merges the custom loader I used for integration tests
1634 into the main test-webkitpy code. Integration tests are not run
1635 by default yet, but at least they can be run.
1637 * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
1638 (MockTestShellTest.test_test_shell_parse_options):
1639 * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1640 (test_path_to_apache_config_file):
1641 * Scripts/webkitpy/layout_tests/servers/http_server_integrationtest.py:
1642 (WebsocketserverTest):
1643 * Scripts/webkitpy/test/main.py:
1644 (Tester._parse_args):
1645 (Tester._run_tests):
1646 (Tester._test_names):
1647 (Tester._log_exception):
1649 (_Loader.getTestCaseNames):
1650 (_Loader.getTestCaseNames.isTestMethod):
1652 2012-08-02 Dirk Pranke <dpranke@chromium.org>
1654 test-webkitpy: clean up runner in preparation for running tests serially as necessary
1655 https://bugs.webkit.org/show_bug.cgi?id=92922
1657 Reviewed by Ojan Vafai.
1659 In order to run some tests by themselves (serially, rather than
1660 in parallel with other tests), we will need to be able to run
1661 multiple test suites; this causes us to move loading the logic
1662 for finding test method names out of the runner and into main.
1664 I'm taking advantage of this to simplify some other stuff from
1665 the runner as well; it is now very simple and doesn't expose its
1666 dependency on unittest.TestResult at all (nor will the Printer
1669 Subsequent patches will move the custom loader from
1670 port_testcase so that we can properly choose whether to run
1671 integration tests and/or serial tests, and then update the
1672 appropriate tests to run only serially.
1674 * Scripts/webkitpy/test/main.py:
1675 (Tester._run_tests):
1677 (Tester._check_imports):
1678 (Tester._test_names):
1679 (Tester._all_test_names):
1680 * Scripts/webkitpy/test/printer.py:
1682 (Printer.write_update):
1684 (Printer.print_finished_test):
1685 (Printer.print_result):
1686 * Scripts/webkitpy/test/runner.py:
1692 * Scripts/webkitpy/test/runner_unittest.py:
1693 (FakeLoader.loadTestsFromName):
1694 (RunnerTest.test_run):
1696 2012-08-02 Adam Barth <abarth@webkit.org>
1698 Turn on tests for the mac-ews, for realz this time.
1700 * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1703 2012-08-02 Xianzhu Wang <wangxianzhu@chromium.org>
1705 [Chromium-Android] Run layout tests on multiple devices in parallel
1706 https://bugs.webkit.org/show_bug.cgi?id=92877
1708 Reviewed by Dirk Pranke.
1710 Moved methods that run command on a particular device from ChromiumAndroidPort to ChromiumAndroidDriver.
1711 The drivers run adb commands with the '-s serial_number' parameter which specifies the device according to the work_number.
1713 * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1714 (ChromiumAndroidPort.__init__):
1715 (ChromiumAndroidPort.default_child_processes): Default to the number of attached devices.
1716 (ChromiumAndroidPort.test_expectations): Removed because it is unused.
1717 (ChromiumAndroidPort.setup_test_run): Most contents moved into ChromiumAndroidDriver._setup_test()
1718 (ChromiumAndroidPort.clean_up_test_run): Now the http server is stopped here.
1719 (ChromiumAndroidPort._get_devices):
1720 (ChromiumAndroidPort._get_device_serial):
1721 (ChromiumAndroidDriver):
1722 (ChromiumAndroidDriver.__init__):
1723 (ChromiumAndroidDriver.__del__):
1724 (ChromiumAndroidDriver._setup_test): Original contents of ChromiumAndroidPort.setup_test_run().
1725 (ChromiumAndroidDriver._push_executable): Moved from ChromiumAndroidPort.
1726 (ChromiumAndroidDriver._push_fonts): Moved from ChromiumAndroidPort.
1727 (ChromiumAndroidDriver._push_test_resources): Moved from ChromiumAndroidPort.
1728 (ChromiumAndroidDriver._synchronize_datetime): Moved from ChromiumAndroidPort.
1729 (ChromiumAndroidDriver._run_adb_command): Moved from ChromiumAndroidPort.
1730 (ChromiumAndroidDriver._teardown_performance): Moved from ChromiumAndroidPort.
1731 (ChromiumAndroidDriver._get_crash_log): Moved from ChromiumAndroidPort.
1732 (ChromiumAndroidDriver.cmd_line):
1733 (ChromiumAndroidDriver._file_exists_on_device):
1734 (ChromiumAndroidDriver._remove_all_pipes):
1735 (ChromiumAndroidDriver._start):
1736 (ChromiumAndroidDriver._start_once):
1737 (ChromiumAndroidDriver.stop):
1738 * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1740 (MockRunCommand.__init__):
1741 (MockRunCommand.mock_run_command_fn):
1742 (MockRunCommand.mock_no_device):
1743 (MockRunCommand.mock_one_device):
1744 (MockRunCommand.mock_two_devices):
1745 (MockRunCommand.mock_no_tombstone_dir):
1746 (MockRunCommand.mock_no_tombstone_file):
1747 (MockRunCommand.mock_ten_tombstones):
1748 (MockRunCommand.mock_logcat):
1749 (ChromiumAndroidPortTest):
1750 (ChromiumAndroidPortTest.make_port):
1751 (ChromiumAndroidPortTest.test_driver_cmd_line):
1752 (ChromiumAndroidPortTest.test_get_devices_no_device):
1753 (ChromiumAndroidPortTest.test_get_devices_one_device):
1754 (ChromiumAndroidPortTest.test_get_devices_two_devices):
1755 (ChromiumAndroidPortTest.test_get_device_serial_no_device):
1756 (ChromiumAndroidPortTest.test_get_device_serial_one_device):
1757 (ChromiumAndroidPortTest.test_get_device_serial_two_devices):
1758 (ChromiumAndroidDriverTest):
1759 (ChromiumAndroidDriverTest.setUp):
1760 (ChromiumAndroidDriverTest.test_get_last_stacktrace):
1761 (ChromiumAndroidDriverTest.test_get_crash_log):
1762 (ChromiumAndroidDriverTest.test_cmd_line):
1763 (ChromiumAndroidDriverTwoDriversTest):
1764 (ChromiumAndroidDriverTwoDriversTest.test_two_drivers):
1765 * Scripts/webkitpy/layout_tests/port/driver.py:
1767 (Driver._get_crash_log): Added to allow subclasses to override.
1768 * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1769 (parse_args): Removed the --adb-args command-line parameter because now we select device automatically. Added --adb-device to specify devices.
1771 2012-08-02 Dinu Jacob <dinu.jacob@nokia.com>
1773 WebKitTestRunner needs layoutTestController.setUserStyleSheetEnabled
1774 https://bugs.webkit.org/show_bug.cgi?id=42679
1776 Reviewed by Eric Seidel.
1778 * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Added
1779 setUserStyleSheetEnabled and setUserStyleSheetLocation.
1780 * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1781 (WTR::LayoutTestController::LayoutTestController): Initialize new members added.
1782 (WTR::LayoutTestController::setUserStyleSheetEnabled): Added.
1783 (WTR::LayoutTestController::setUserStyleSheetLocation): Added.
1784 * WebKitTestRunner/InjectedBundle/LayoutTestController.h: Added members
1785 userStyleSheetEnabled, and m_userStyleSheetLocation and methods
1786 setUserStyleSheetEnabled, and setUserStyleSheetLocation.
1788 2012-08-02 Hans Wennborg <hans@chromium.org>
1790 Speech JavaScript API: Fire speech start event at the same time as sound start event
1791 https://bugs.webkit.org/show_bug.cgi?id=92971
1793 Reviewed by Adam Barth.
1795 Update the MockWebSpeechRecognizer to not fire "speech started" events
1798 * DumpRenderTree/chromium/MockWebSpeechRecognizer.cpp:
1799 (MockWebSpeechRecognizer::start):
1801 2012-08-02 Philippe Normand <pnormand@igalia.com>
1803 [GTK] make sometimes using a single core
1804 https://bugs.webkit.org/show_bug.cgi?id=92998
1806 Reviewed by Martin Robinson.
1808 In the GTK 64-bit Release buildbot some builds use a single core
1809 for the make process. I suspect this is because in those cases
1810 nproc reports a single core available. The proposed solution is to
1811 always rely on all the cores available in the machine.
1813 * Scripts/webkitdirs.pm:
1814 (determineNumberOfCPUs):
1816 2012-08-02 Adam Barth <abarth@webkit.org>
1818 Re-land http://trac.webkit.org/changeset/94441 now that lforschler is ready.
1820 This patch enables testing on the mac-ews bots.
1822 * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1824 2012-08-02 Adam Barth <abarth@webkit.org>
1826 [Chromium] Merge final nits to DumpRenderTree.gyp for Android
1827 https://bugs.webkit.org/show_bug.cgi?id=90920
1829 Reviewed by Tony Chang.
1831 This patch contains the last few small changes to DumpRenderTree.gyp
1832 from the chromium-android branch. After this change, this file will be
1835 * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1837 2012-08-02 Peter Beverloo <peter@chromium.org>
1839 [Chromium] Toggle the type of ant compile for webkit_unit_tests and TestWebKitAPI
1840 https://bugs.webkit.org/show_bug.cgi?id=92858
1842 Reviewed by Adam Barth.
1844 Now that the sdk_build variable is available, we can remove these two differences
1845 as well. This goes together with Adam's bug 90920.
1847 After this patch, the whole Tools/ directory will be unforked :-).
1849 * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
1851 2012-08-02 Zoltan Arvai <zarvai@inf.u-szeged.hu>
1853 [Qt] MSVC specific buildfix for DRT.
1854 https://bugs.webkit.org/show_bug.cgi?id=92978
1856 Reviewed by Simon Hausmann.
1858 DumpRenderTree/qt subdirectory is missing from generated makefile under MSVC build, need to be added to the pro file.
1860 * DumpRenderTree/qt/DumpRenderTree.pro:
1862 2012-08-02 Alexander Shalamov <alexander.shalamov@intel.com>
1864 [EFL][WK2] WTR is failing when X server is not running
1865 https://bugs.webkit.org/show_bug.cgi?id=92719
1867 Reviewed by Hajime Morita.
1869 EFL's WebKitTestRunner doesn't execute tests when X server is not running.
1870 This patch fixes the problem by checking environment variable before ecore x initialization.
1872 * Scripts/webkitpy/layout_tests/port/efl.py:
1873 (EflPort.setup_environ_for_server):
1874 * WebKitTestRunner/efl/main.cpp:
1877 2012-08-01 Dirk Pranke <dpranke@chromium.org>
1879 REGRESSION(r123893): Reftest mismatches are run through ImageDiff with 0.1 tolerance
1880 https://bugs.webkit.org/show_bug.cgi?id=92847
1882 Reviewed by Ojan Vafai.
1884 I was failing to distinguish between 'None' and 0 :(. Fixed and added tests.
1886 * Scripts/webkitpy/layout_tests/port/base.py:
1888 * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1889 (PortTestCase.test_diff_image.make_proc):
1890 (PortTestCase.test_diff_image):
1892 2012-08-01 Dirk Pranke <dpranke@chromium.org>
1894 test-webkitpy: clean up handling of tests to skip
1895 https://bugs.webkit.org/show_bug.cgi?id=92909
1897 Reviewed by Ryosuke Niwa.
1899 This change moves the handling of tests to skip into main.py
1900 where it is at least slightly more findable and generic.
1902 Also fix a couple of lint nits.
1904 * Scripts/webkitpy/test/finder.py:
1907 (Finder._default_names):
1908 * Scripts/webkitpy/test/main.py:
1911 * Scripts/webkitpy/test/main_unittest.py:
1912 (TesterTest.test_no_tests_found):
1913 * Scripts/webkitpy/test/runner_unittest.py:
1915 2012-08-01 Dirk Pranke <dpranke@chromium.org>
1917 test-webkitpy: remove --skip-integrationtests flag
1918 https://bugs.webkit.org/show_bug.cgi?id=92907
1920 Reviewed by Ryosuke Niwa.
1922 This flag is no longer used since the tests complete quickly.
1924 Also clean up the unused skip_if_parallel arg in finder, and
1925 make a couple of lint fixes.
1927 * Scripts/webkitpy/test/finder.py:
1928 (Finder.find_names):
1929 (Finder._default_names):
1930 * Scripts/webkitpy/test/finder_unittest.py:
1931 (FinderTest.check_names):
1932 (FinderTest.test_default_names):
1933 * Scripts/webkitpy/test/main.py:
1934 (Tester._parse_args):
1936 * Scripts/webkitpy/test/main_unittest.py:
1937 (TesterTest.test_no_tests_found):
1939 2012-08-01 Ryosuke Niwa <rniwa@webkit.org>
1941 Generalize SheriffIRCBot to prepare for PerfBot
1942 https://bugs.webkit.org/show_bug.cgi?id=92912
1944 Reviewed by Adam Barth.
1946 Renamed SheriffIRCBot to IRCBot and generalized to support non-sheriffbot IRC bot in the future.
1947 This will be useful when I add an experimental perf EWS IRC bot.
1949 * Scripts/webkitpy/tool/bot/irc_command.py:
1951 * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1953 * Scripts/webkitpy/tool/bot/ircbot.py: Moved from Tools/Scripts/webkitpy/tool/bot/sheriffircbot.py.
1954 (Eliza): Moved from irc_command.
1959 (IRCBot.irc_delegate):
1960 (IRCBot._parse_command_and_args):
1961 (IRCBot.process_message):
1962 * Scripts/webkitpy/tool/bot/ircbot_unittest.py: Moved from Tools/Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py.
1965 (IRCBotTest.test_eliza): Moved from IRCCommandTest.
1966 (IRCBotTest.test_parse_command_and_args):
1967 (IRCBotTest.test_exception_during_command):
1968 * Scripts/webkitpy/tool/bot/sheriffircbot.py: Removed.
1969 * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py: Removed.
1970 * Scripts/webkitpy/tool/commands/sheriffbot.py:
1971 (SheriffBot.begin_work_queue):
1972 * Scripts/webkitpy/webkitpy.pyproj:
1974 2012-08-01 Dirk Pranke <dpranke@chromium.org>
1976 nrwt: reenable the test for --verbose working in child processes
1977 https://bugs.webkit.org/show_bug.cgi?id=92894
1979 Reviewed by Ryosuke Niwa.
1981 This change rewrites the test to not use outputcapture, meaning
1982 that the MockHost can propagate to Worker properly and things
1985 * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1986 (MainTest.test_verbose_in_child_processes):
1988 2012-08-01 Ryosuke Niwa <rniwa@webkit.org>
1990 run-perf-tests --build-directory doesn't work
1991 https://bugs.webkit.org/show_bug.cgi?id=92051
1993 Reviewed by Dirk Pranke.
1995 The bug was caused by not prepending build_directory even when one is defined.
1998 * Scripts/webkitpy/layout_tests/port/base.py:
2000 * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2001 (PortTest.test_build_path):
2003 2012-08-01 Sheriff Bot <webkit.review.bot@gmail.com>
2005 Unreviewed, rolling out r123865.
2006 http://trac.webkit.org/changeset/123865
2007 https://bugs.webkit.org/show_bug.cgi?id=92891
2009 This patch is causing the style-queue to fall behind
2010 (Requested by abarth on #webkit).
2012 * Scripts/webkitpy/tool/commands/queues.py:
2013 (AbstractReviewQueue.begin_work_queue):
2014 (StyleQueue.__init__):
2015 * Scripts/webkitpy/tool/commands/queues_unittest.py:
2016 (StyleQueueTest.test_style_queue_with_style_exception):
2017 (test_style_queue_with_watch_list_exception):
2019 2012-08-01 Ryosuke Niwa <rniwa@webkit.org>
2021 run-perf-tests throws an exception when the output json is malformed
2022 https://bugs.webkit.org/show_bug.cgi?id=92887
2024 Reviewed by Dirk Pranke.
2026 Catch exceptions and gracefully fail. Also split _generate_json into smaller methods.
2028 * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2030 (PerfTestsRunner.run):
2031 (PerfTestsRunner._generate_output): Extracted from _generate_json.
2032 (PerfTestsRunner._merge_source_json): Ditto; catch all exceptions since they are too many
2033 exceptions to consder here.
2034 (PerfTestsRunner._merge_outputs): Ditto.
2035 (PerfTestsRunner._generate_output_files): Extracted from _generate_json.
2036 * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2037 (_test_run_with_json_output): Don't assert logs when we except an non-zero exit code.
2038 (create_runner_and_setup_results_template): Extracted from test_run_generates_results_page.
2039 (test_run_generates_results_page):
2040 (test_run_with_bad_output_json): Added.
2041 (test_run_with_bad_json_source): Added.
2042 (test_run_with_upload_json):
2044 2012-08-01 Thiago Marcos P. Santos <thiago.santos@intel.com>
2046 Regression(r124135): nrwt: --verbose logging does not work right on windows
2047 https://bugs.webkit.org/show_bug.cgi?id=92845
2049 Reviewed by Dirk Pranke.
2051 Disabled the failing test. The Workers spawned by this test are using Host()
2052 instead of MockHost() and trying to access the real filesystem instead of the
2055 * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2056 (MainTest.disabled_test_verbose):
2058 2012-08-01 Arnaud Renevier <a.renevier@sisa.samsung.com>
2060 keyring.get_password may raise an exception.
2061 https://bugs.webkit.org/show_bug.cgi?id=92876
2063 Reviewed by Dirk Pranke.
2065 keyring.get_password sometimes raises an exception. We wrap this call,
2066 and also set_password to avoid breaking webkit-patch in that case.
2068 * Scripts/webkitpy/common/net/credentials.py:
2069 (Credentials._offer_to_store_credentials_in_keyring):
2070 (Credentials.read_credentials):
2072 2012-08-01 Xianzhu Wang <wangxianzhu@chromium.org>
2074 Layout Test fast/text/descent-clip-in-scaled-page.html is failing on linux since it was added
2075 https://bugs.webkit.org/show_bug.cgi?id=91386
2077 Reviewed by Tony Chang.
2079 Added font mapping from SubpixelPositioningAhem to Ahem on Linux and Android
2080 (used in layout test fast/text/descent-clip-in-scaled-page.html).
2081 Added font mapping from SubpixelPositioning to Times New Roman on Android to match Linux
2082 (used in platform/chromium-linux/fast/text/chromium-linux-text-subpixel-positioning.html)
2084 * DumpRenderTree/chromium/android_main_fonts.xml:
2085 * DumpRenderTree/chromium/fonts.conf:
2087 2012-08-01 Sheriff Bot <webkit.review.bot@gmail.com>
2089 Unreviewed, rolling out r124325.
2090 http://trac.webkit.org/changeset/124325
2091 https://bugs.webkit.org/show_bug.cgi?id=92865
2093 broke android build (Requested by mnaganov on #webkit).
2095 * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2097 2012-08-01 Thiago Marcos P. Santos <thiago.santos@intel.com>
2099 [EFL] Dump a backtrace in case of a crash on the UIProcess
2100 https://bugs.webkit.org/show_bug.cgi?id=92843
2102 Reviewed by Csaba Osztrogonác.
2104 We are already dumping the backtraces when automated tests are
2105 crashing, but only for the WebProcess. This patch adds the same
2106 hooks for the UIProcess.
2108 * TestWebKitAPI/efl/main.cpp:
2110 * WebKitTestRunner/efl/main.cpp:
2113 2012-08-01 Adam Barth <abarth@webkit.org>
2115 [Chromium] Merge final nits to DumpRenderTree.gyp for Android
2116 https://bugs.webkit.org/show_bug.cgi?id=90920
2118 Reviewed by Tony Chang.
2120 This patch contains the last few small changes to DumpRenderTree.gyp
2121 from the chromium-android branch. After this change, this file will be
2124 * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2126 2012-08-01 Csaba Osztrogonác <ossy@webkit.org>
2128 [Qt] ImageDiff should use float division instead of integer
2129 https://bugs.webkit.org/show_bug.cgi?id=92859
2131 Reviewed by Zoltan Herczeg.
2133 * DumpRenderTree/qt/ImageDiff.cpp:
2136 2012-08-01 Sheriff Bot <webkit.review.bot@gmail.com>
2138 Unreviewed, rolling out r124313.
2139 http://trac.webkit.org/changeset/124313
2140 https://bugs.webkit.org/show_bug.cgi?id=92855
2142 Compilation failed on Chromium ports (Requested by yosin on
2145 * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2146 * DumpRenderTree/DumpRenderTree.gypi:
2147 * DumpRenderTree/DumpRenderTree.h:
2148 * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2149 * DumpRenderTree/DumpRenderTreeCommon.cpp: Removed.
2150 * DumpRenderTree/chromium/DumpRenderTree.cpp:
2153 * DumpRenderTree/chromium/TestShell.cpp:
2154 (TestShell::TestShell):
2155 (TestShell::runFileTest):
2157 * DumpRenderTree/chromium/TestShell.h:
2159 (TestParams::TestParams):
2161 * DumpRenderTree/efl/CMakeLists.txt:
2162 * DumpRenderTree/efl/DumpRenderTree.cpp:
2163 (parseCommandLineOptions):
2165 (getExpectedPixelHash):
2167 (shouldDumpPixelsAndCompareWithExpected):
2169 * DumpRenderTree/gtk/DumpRenderTree.cpp:
2170 (initializeGlobalsFromCommandLineOptions):
2174 * DumpRenderTree/mac/DumpRenderTree.mm:
2175 (initializeGlobalsFromCommandLineOptions):
2179 * DumpRenderTree/qt/DumpRenderTree.pro:
2180 * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2181 (WebCore::DumpRenderTree::processLine):
2182 * DumpRenderTree/win/DumpRenderTree.cpp:
2185 (dllLauncherEntryPoint):
2186 * DumpRenderTree/win/DumpRenderTree.vcproj:
2187 * DumpRenderTree/wscript:
2188 * DumpRenderTree/wx/DumpRenderTreeWx.cpp:
2193 * Scripts/old-run-webkit-tests:
2194 * Scripts/webkitpy/layout_tests/port/base.py:
2195 (Port.supports_switching_pixel_tests_per_test):
2197 (Port._supports_switching_pixel_tests_per_test):
2198 (Port.should_run_as_pixel_test):
2199 * Scripts/webkitpy/layout_tests/port/driver.py:
2201 (Driver._command_from_driver_input):
2202 * Scripts/webkitpy/layout_tests/port/qt.py:
2203 (QtPort._supports_switching_pixel_tests_per_test):
2204 * Scripts/webkitpy/layout_tests/port/test.py:
2205 (TestPort.supports_switching_pixel_tests_per_test):
2206 * WebKitTestRunner/TestController.cpp:
2207 (WTR::TestController::TestController):
2208 (WTR::TestController::initialize):
2209 (WTR::TestController::runTest):
2210 * WebKitTestRunner/TestController.h:
2213 2012-08-01 Balazs Kelemen <kbalazs@webkit.org>
2215 All ports should support per test switching of pixel testing
2216 https://bugs.webkit.org/show_bug.cgi?id=92398
2218 Reviewed by Dirk Pranke.
2220 Teach all test drivers to accept a per test control of whether
2221 to dump pixels. Drivers now accept a -p/--pixel-test argument
2222 on the standart input that means that running the current
2223 test as pixel test is allowed (even if the expected hash is
2224 missing). Removed the --pixel-tests command line option since
2225 there is no need for it anymore.
2227 * DumpRenderTree/DumpRenderTree.gypi:
2228 * DumpRenderTree/DumpRenderTree.h:
2229 (TestCommand::TestCommand):
2231 * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2232 * DumpRenderTree/DumpRenderTreeCommon.cpp: Added.
2234 (CommandTokenizer::CommandTokenizer):
2235 (CommandTokenizer::pump):
2236 (CommandTokenizer::next):
2237 (CommandTokenizer::hasNext):
2240 Common logic to parse the input line from the standard input (or from the command line
2241 in standalone mode). Made it somewhat general so we can easily add more arguments if
2243 * DumpRenderTree/chromium/DumpRenderTree.cpp:
2246 * DumpRenderTree/chromium/TestShell.cpp:
2247 (TestShell::TestShell):
2248 (TestShell::runFileTest):
2250 * DumpRenderTree/chromium/TestShell.h:
2251 (TestParams::TestParams):
2253 * DumpRenderTree/efl/CMakeLists.txt:
2254 * DumpRenderTree/efl/DumpRenderTree.cpp:
2255 (parseCommandLineOptions):
2257 (shouldDumpPixelsAndCompareWithExpected):
2259 * DumpRenderTree/gtk/DumpRenderTree.cpp:
2260 (initializeGlobalsFromCommandLineOptions):
2262 * DumpRenderTree/mac/DumpRenderTree.mm:
2263 (initializeGlobalsFromCommandLineOptions):
2265 * DumpRenderTree/qt/DumpRenderTree.pro:
2266 * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2267 (WebCore::DumpRenderTree::processLine):
2268 * DumpRenderTree/win/DumpRenderTree.cpp:
2271 (dllLauncherEntryPoint):
2272 * DumpRenderTree/win/DumpRenderTree.vcproj:
2273 * DumpRenderTree/wscript:
2274 * DumpRenderTree/wx/DumpRenderTreeWx.cpp:
2278 * Scripts/old-run-webkit-tests: Pass --pixel-test before the hash.
2279 Changed the separator form ' to : because it don't need to be escaped
2280 when passing manually on the comand line.
2281 * Scripts/webkitpy/layout_tests/port/base.py:
2282 (Port.should_run_as_pixel_test): Removed now that all ports supports it.
2283 * Scripts/webkitpy/layout_tests/port/driver.py:
2284 (Driver.cmd_line): Don't pass --pixel-tests anymore.
2285 (Driver._command_from_driver_input):
2287 * WebKitTestRunner/TestController.cpp:
2288 (WTR::TestController::TestController):
2289 (WTR::TestController::initialize):
2290 (WTR::TestCommand::TestCommand):
2294 (WTR::CommandTokenizer::CommandTokenizer):
2295 (WTR::CommandTokenizer::pump):
2296 (WTR::CommandTokenizer::next):
2297 (WTR::CommandTokenizer::hasNext):
2299 (WTR::parseInputLine):
2300 (WTR::TestController::runTest):
2301 * WebKitTestRunner/TestController.h:
2303 Duplicate the logic for parsing the command line because
2304 unfortunately there is no way to share code between
2305 WebKitTestRunner and DumpRenderTree.
2307 2012-08-01 Martin Robinson <mrobinson@igalia.com>
2309 Add the jhbuild source directory to the __builtin__ object
2311 Reviewed by Philippe Normand.
2313 It seems that the jhbuild script explicitly looks for the SRCDIR __builtin__ when
2314 initializing. This is necessary for loading the jhbuild from the checkout directory.
2316 * jhbuild/jhbuildutils.py:
2317 (enter_jhbuild_environment_if_available): Set SRCDIR to the source directory.
2319 2012-07-31 Dirk Pranke <dpranke@chromium.org>
2321 nrwt: move actual test-running code into layout_test_runner.py
2322 https://bugs.webkit.org/show_bug.cgi?id=92806
2324 Reviewed by Ryosuke Niwa.
2326 This is the final patch in manager-refactoring series (for now).
2327 This moves all of the logic to actually run the tests into
2328 layout_test_runner (in a new LayoutTestRunner) class. Now
2329 the mechanics of actually executing all of the tests are
2330 contained in one file (of course the actual work of talking to
2331 DRT/WTR and diff'ing the results is still in single_test_runner).
2333 Now manager is left with setting up and tearing down the test
2334 environment, finding the tests to run, and processing the
2335 results of the tests, which is a pretty managable amount of code
2338 The unit tests for manager were split in two and all of the
2339 runner-specific tests moved into layout_test_runner_unittest.
2340 They were significantly cleaned up to be easier to read and
2341 maintain as a part of this.
2343 * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2344 (TestRunInterruptedException):
2345 (TestRunInterruptedException.__init__):
2346 (TestRunInterruptedException.__reduce__):
2348 (LayoutTestRunner.__init__):
2349 (LayoutTestRunner.test_key):
2350 (LayoutTestRunner.run_tests):
2351 (LayoutTestRunner.run_tests.worker_factory):
2352 (LayoutTestRunner.run_tests.instead):
2353 (LayoutTestRunner._mark_interrupted_tests_as_skipped):
2354 (LayoutTestRunner._interrupt_if_at_failure_limits):
2355 (LayoutTestRunner._interrupt_if_at_failure_limits.interrupt_if_at_failure_limit):
2356 (LayoutTestRunner._update_summary_with_result):
2357 (LayoutTestRunner.start_servers_with_lock):
2358 (LayoutTestRunner.stop_servers_with_lock):
2359 (LayoutTestRunner.handle):
2360 (LayoutTestRunner._handle_started_test):
2361 (LayoutTestRunner._handle_finished_test_list):
2362 (LayoutTestRunner._handle_finished_test_list.find):
2363 (LayoutTestRunner._handle_finished_test):
2364 * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
2366 (FakePrinter.print_workers_and_shards):
2367 (FakePrinter.print_finished_test):
2368 (FakePrinter.write):
2369 (FakePrinter.write_update):
2370 (FakePrinter.flush):
2371 (LockCheckingRunner):
2372 (LockCheckingRunner.__init__):
2373 (LockCheckingRunner.handle_finished_list):
2374 (LayoutTestRunnerTests):
2375 (LayoutTestRunnerTests._runner):
2376 (LayoutTestRunnerTests._result_summary):
2377 (LayoutTestRunnerTests._run_tests):
2378 (LayoutTestRunnerTests.test_http_locking):
2379 (LayoutTestRunnerTests.test_perf_locking):
2380 (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
2381 (LayoutTestRunnerTests.test_update_summary_with_result):
2382 (LayoutTestRunnerTests.test_servers_started):
2383 (LayoutTestRunnerTests.test_servers_started.start_http_server):
2384 (LayoutTestRunnerTests.test_servers_started.start_websocket_server):
2385 (LayoutTestRunnerTests.test_servers_started.stop_http_server):
2386 (LayoutTestRunnerTests.test_servers_started.stop_websocket_server):
2387 * Scripts/webkitpy/layout_tests/controllers/manager.py:
2388 (summarize_results):
2390 (Manager._is_http_test):
2392 (Manager._is_websocket_test):
2393 (Manager._websocket_tests):
2394 (Manager._prepare_lists):
2395 (Manager._is_ref_test):
2397 (Manager._run_tests):
2398 (Manager._look_for_new_crash_logs):
2399 (Manager._show_results_html_file):
2400 * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2401 (ManagerTest.test_look_for_new_crash_logs):
2402 (ResultSummaryTest.summarized_results):
2403 (ResultSummaryTest.test_summarized_results_wontfix):
2405 2012-07-31 Dirk Pranke <dpranke@chromium.org>
2407 nrwt: move sharding logic into layout_test_runner.py
2408 https://bugs.webkit.org/show_bug.cgi?id=92805
2410 Reviewed by Ryosuke Niwa.
2412 The sharding logic is specific to how we actually run the tests,
2413 so it belongs in this file instead of manager.py or finder.py.
2415 * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2416 (Worker._run_single_test):
2418 (TestShard.__init__):
2419 (TestShard.__repr__):
2423 (Sharder.shard_tests):
2424 (Sharder.shard_tests.or):
2425 (Sharder._shard_in_two):
2426 (Sharder._shard_every_file):
2427 (Sharder._shard_by_directory):
2428 (Sharder._resize_shards):
2429 (Sharder._resize_shards.divide_and_round_up):
2430 (Sharder._resize_shards.extract_and_flatten):
2431 (Sharder._resize_shards.split_at):
2432 (Sharder._dir_for_test_input):
2434 (Sharder.natural_sort_key):
2435 (Sharder.natural_sort_key.tryint):
2436 * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py: Added.
2438 (SharderTests.get_test_input):
2439 (SharderTests.get_shards):
2440 (SharderTests.get_shards.split):
2441 (SharderTests.assert_shards):
2442 (SharderTests.test_shard_by_dir):
2443 (SharderTests.test_shard_by_dir_sharding_ref_tests):
2444 (SharderTests.test_shard_every_file):
2445 (SharderTests.test_shard_in_two):
2446 (SharderTests.test_shard_in_two_sharding_ref_tests):
2447 (SharderTests.test_shard_in_two_has_no_locked_shards):
2448 (SharderTests.test_shard_in_two_has_no_unlocked_shards):
2449 (SharderTests.test_multiple_locked_shards):
2450 (NaturalCompareTest):
2451 (NaturalCompareTest.assert_cmp):
2452 (NaturalCompareTest.test_natural_compare):
2454 (KeyCompareTest.setUp):
2455 (KeyCompareTest.setUp.split):
2456 (KeyCompareTest.assert_cmp):
2457 (KeyCompareTest.test_test_key):
2458 * Scripts/webkitpy/layout_tests/controllers/manager.py:
2460 (Manager._prepare_lists):
2461 (Manager._is_ref_test):
2462 (Manager._run_tests):
2463 (Manager._handle_finished_test):
2464 * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2465 (ManagerTest.get_options):
2466 (ManagerTest.test_interrupt_if_at_failure_limits):
2467 (ManagerTest.test_update_summary_with_result):
2468 (ManagerTest.test_needs_servers.get_manager_with_tests):
2469 (ManagerTest.integration_test_needs_servers.get_manager_with_tests):
2470 (ManagerTest.test_look_for_new_crash_logs.get_manager_with_tests):
2471 (ManagerTest.test_servers_started):
2472 * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2475 2012-07-31 Dirk Pranke <dpranke@chromium.org>
2477 nrwt: rename worker.py to layout_test_runner.py
2478 https://bugs.webkit.org/show_bug.cgi?id=92804
2480 Reviewed by Ojan Vafai.
2482 Home stretch of this round of manager refactoring ... I will be
2483 moving all of the actual test-running code (which includes the
2484 code that shards the tests for the workers) into a different
2485 module, and it makes sense for that module to contain the actual
2486 worker code, so I'm renaming worker.py to layout_test_runner.py.
2488 * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py: Renamed from Tools/Scripts/webkitpy/layout_tests/controllers/worker.py.
2494 (Worker._update_test_input):
2498 (Worker._kill_driver):
2499 (Worker._run_test_with_timeout):
2500 (Worker._clean_up_after_test):
2501 (Worker._run_test_in_another_thread):
2502 (Worker._run_test_in_another_thread.SingleTestThread):
2503 (Worker._run_test_in_another_thread.SingleTestThread.__init__):
2504 (Worker._run_test_in_another_thread.SingleTestThread.run):
2505 (Worker._run_test_in_this_thread):
2506 (Worker._run_single_test):
2507 * Scripts/webkitpy/layout_tests/controllers/manager.py:
2508 (Manager._run_tests.worker_factory):
2510 2012-07-31 Dirk Pranke <dpranke@chromium.org>
2512 nrwt: clean up names in sharding code
2513 https://bugs.webkit.org/show_bug.cgi?id=92785
2515 Reviewed by Ryosuke Niwa.
2517 More refactoring ... this makes the methods use TestInputs
2518 consistently (and updates the names accordingly) and improves
2519 encapsulation a bit. The sharding code is now pretty
2522 This change adds no new functionality and is covered by the
2523 existing (updated) tests.
2525 * Scripts/webkitpy/layout_tests/controllers/manager.py:
2526 (TestShard.visible.__init__):
2527 (TestShard.visible.__repr__):
2528 (Manager._dir_for_test_input):
2529 (Manager._shard_tests):
2530 (Manager._shard_in_two):
2531 (Manager._shard_every_file):
2532 (Manager._shard_by_directory):
2533 (Manager._run_tests):
2534 * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2535 (ManagerWrapper._test_input_for_file):
2536 (ShardingTests.get_shards):
2537 * Scripts/webkitpy/layout_tests/models/test_input.py:
2538 (TestInput.__init__):
2539 (TestInput.__repr__):
2541 2012-07-31 Dirk Pranke <dpranke@chromium.org>
2543 nrwt: clean up TestInputs in preparation for cleaning up sharding
2544 https://bugs.webkit.org/show_bug.cgi?id=92784
2546 Reviewed by Ryosuke Niwa.
2548 Currently, in order to shard the tests you need to refer to
2549 state in the manager as well as the state in the TestInputs;
2550 this change embeds the necessary state into the TestInputs so
2551 sharding them can be a standalone operation.
2553 The actual clean up of the sharding will follow in a subsequent patch.
2555 Covered by existing tests; no new functionality. However, I did
2556 rework the sharding tests to be less dependent on the test
2557 scaffolding and easier to follow.
2559 * Scripts/webkitpy/layout_tests/controllers/manager.py:
2560 (Manager._test_input_for_file):
2561 (Manager._shard_in_two):
2562 (Manager._shard_every_file):
2563 (Manager._shard_by_directory):
2564 * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2565 (ManagerWrapper._test_input_for_file):
2566 (ShardingTests.assert_shards):
2567 (ShardingTests.test_shard_by_dir):
2568 (ShardingTests.test_shard_by_dir_sharding_ref_tests):
2569 (ShardingTests.test_shard_every_file):
2570 (ShardingTests.test_shard_in_two):
2571 (ShardingTests.test_shard_in_two_sharding_ref_tests):
2572 (ShardingTests.test_shard_in_two_has_no_unlocked_shards):
2573 (ShardingTests.test_multiple_locked_shards):
2574 * Scripts/webkitpy/layout_tests/models/test_input.py:
2575 (TestInput.__init__):
2576 (TestInput.__repr__):
2578 2012-07-31 Thiago Marcos P. Santos <thiago.santos@intel.com>
2580 [EFL] Dump a backtrace in case of a crash
2581 https://bugs.webkit.org/show_bug.cgi?id=92489
2583 Reviewed by Csaba Osztrogonác.
2585 Dump a stack trace in case of a unexpected signal. This should
2586 provide a better report at the build bots when WTR crashes.
2588 * DumpRenderTree/efl/DumpRenderTree.cpp:
2590 * TestWebKitAPI/efl/InjectedBundleController.cpp:
2591 (TestWebKitAPI::InjectedBundleController::platformInitialize):
2592 * WebKitTestRunner/InjectedBundle/efl/InjectedBundleEfl.cpp:
2593 (WTR::InjectedBundle::platformInitialize):
2595 2012-07-31 Joshua Netterfield <jnetterfield@rim.com>
2597 [BlackBerry] Enable CSS Filter Effects
2598 https://bugs.webkit.org/show_bug.cgi?id=92685
2600 Reviewed by Rob Buis.
2602 Enable CSS filter effects, with the exception of custom effects (CSS shaders) and reference effects (SVG effects)
2604 Internally reviewed by Arvid Nilsson <anilsson@rim.com> and Antonio Gomes <agomes@rim.com>.
2606 * Scripts/webkitperl/FeatureList.pm: Acknowledge CSS filter effects.
2608 2012-07-31 Dirk Pranke <dpranke@chromium.org>
2610 nrwt: clean up prepare_lists_and_print_output, run, set_up_run a bit
2611 https://bugs.webkit.org/show_bug.cgi?id=92781
2613 Reviewed by Ryosuke Niwa.
2615 More refactoring ... rename prepare_lists_and_print_output to
2616 just prepare_lists so that it only has a single purpose, and
2617 clean up the surrounding code a bit as well.
2619 * Scripts/webkitpy/layout_tests/controllers/manager.py:
2620 (Manager._collect_tests):
2621 (Manager._prepare_lists):
2622 (Manager._set_up_run):
2625 2012-07-31 Dirk Pranke <dpranke@chromium.org>
2627 nrwt: clean up self._test_files_list vs. self._test_files, other nits
2628 https://bugs.webkit.org/show_bug.cgi?id=92702
2630 Reviewed by Ojan Vafai.
2632 Get rid of self._test_files, rename self._test_files_list to
2633 self._test_names (removing the confusion between the two fields).
2635 Also inline the one call to _parse_expectations, remove the one
2636 use of more_tests_to_skip, and fix a couple of minor nits.
2638 * Scripts/webkitpy/layout_tests/controllers/finder.py:
2639 (LayoutTestFinder.split_into_chunks):
2640 * Scripts/webkitpy/layout_tests/controllers/manager.py:
2642 (Manager._collect_tests):
2643 (Manager._http_tests):
2644 (Manager._websocket_tests):
2645 (Manager._is_perf_test):
2646 (Manager.prepare_lists_and_print_output):
2647 (Manager.needs_servers):
2649 (Manager._mark_interrupted_tests_as_skipped):
2650 (Manager._update_summary_with_result):
2651 (Manager._upload_json_files):
2652 * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2653 (ManagerTest.test_interrupt_if_at_failure_limits):
2654 (ManagerTest.test_needs_servers.get_manager_with_tests):
2655 (ManagerTest.test_servers_started):
2657 2012-07-31 Dirk Pranke <dpranke@chromium.org>
2659 nrwt: --verbose logging does not work right on windows
2660 https://bugs.webkit.org/show_bug.cgi?id=92673
2662 Reviewed by Ojan Vafai.
2664 Add a test for r124090.
2666 * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2667 (MainTest.test_verbose):
2669 2012-07-31 Dirk Pranke <dpranke@chromium.org>
2671 REGRESSION(124116): Number of skipped tests isn't printed anymore
2672 https://bugs.webkit.org/show_bug.cgi?id=92736
2674 Reviewed by Tony Chang.
2676 Add the skipped number back in; it was useful ... also clean up
2677 the formatting a little bit.
2679 * Scripts/webkitpy/layout_tests/views/printing.py:
2680 (Printer.print_found):
2682 2012-07-31 Hans Wennborg <hans@chromium.org>
2684 Speech JavaScript API: Throw exception for start() when already started
2685 https://bugs.webkit.org/show_bug.cgi?id=92756
2687 Reviewed by Adam Barth.
2689 Fix the task queue in MockWebSpeechRecognizer.
2690 It's important that we remove the task from the queue before we run
2691 it, as running the task could alter the queue's state.
2693 * DumpRenderTree/chromium/MockWebSpeechRecognizer.cpp:
2694 (MockWebSpeechRecognizer::abort):
2695 (MockWebSpeechRecognizer::setError):
2696 (MockWebSpeechRecognizer::clearTaskQueue):
2697 (MockWebSpeechRecognizer::StepTask::runIfValid):
2698 * DumpRenderTree/chromium/MockWebSpeechRecognizer.h:
2699 (MockWebSpeechRecognizer):
2701 2012-07-31 Dirk Pranke <dpranke@chromium.org>
2703 nrwt: move handling the initial list of tests to skip to finder
2704 https://bugs.webkit.org/show_bug.cgi?id=92701
2706 Reviewed by Ryosuke Niwa.
2708 More refactoring of manager ...
2710 * Scripts/webkitpy/layout_tests/controllers/finder.py:
2711 (LayoutTestFinder.skip_tests):
2712 * Scripts/webkitpy/layout_tests/controllers/manager.py:
2713 (Manager._collect_tests):
2714 (Manager.prepare_lists_and_print_output):
2716 2012-07-31 Alexey Proskuryakov <ap@apple.com>
2718 [WK2] Kill the concept of secondary shared process
2719 https://bugs.webkit.org/show_bug.cgi?id=92676
2721 Reviewed by Sam Weinig.
2723 * MiniBrowser/win/BrowserView.cpp: (BrowserView::create): Update Windows build fix.
2724 Don't create a new context for every view.
2726 2012-07-31 Peter Beverloo <peter@chromium.org>
2728 [Chromium] Don't use ninja for building Chromium for Android
2729 https://bugs.webkit.org/show_bug.cgi?id=92764
2731 Reviewed by Adam Barth.
2733 When Ninja projects files are generated (which will be done for Android),
2734 make sure that the Android builder doesn't unintentionally switch to
2735 using Ninja. We should make that switch at a predetermined time.
2737 * Scripts/webkitdirs.pm:
2740 2012-07-31 Thiago Marcos P. Santos <thiago.santos@intel.com>
2742 Add a mechanism to dump the stack trace in case of a crash
2743 https://bugs.webkit.org/show_bug.cgi?id=92666
2745 Reviewed by Csaba Osztrogonác.
2747 Move crash signal handlers to WTFInstallReportBacktraceOnCrashHook()
2748 and eliminate some duplicated code.
2750 * DumpRenderTree/qt/main.cpp:
2752 * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
2753 (WTR::InjectedBundle::platformInitialize):
2755 2012-07-31 Jochen Eisinger <jochen@chromium.org>
2757 [chromium] move EventSender into TestRunner.a
2758 https://bugs.webkit.org/show_bug.cgi?id=92277
2760 Reviewed by Adam Barth.
2762 * DumpRenderTree/DumpRenderTree.gypi:
2763 * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2764 * DumpRenderTree/chromium/TestRunner/EventSender.cpp: Renamed from Tools/DumpRenderTree/chromium/EventSender.cpp.
2765 * DumpRenderTree/chromium/TestRunner/EventSender.h: Renamed from Tools/DumpRenderTree/chromium/EventSender.h.
2766 * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
2767 (TestInterfaces::Internal::eventSender):
2768 (TestInterfaces::Internal):
2769 (TestInterfaces::Internal::Internal):
2770 (TestInterfaces::Internal::~Internal):
2771 (TestInterfaces::Internal::setWebView):
2772 (TestInterfaces::Internal::setDelegate):
2773 (TestInterfaces::Internal::bindTo):
2774 (TestInterfaces::Internal::resetAll):
2775 (TestInterfaces::eventSender):
2776 * DumpRenderTree/chromium/TestRunner/TestInterfaces.h:
2778 * DumpRenderTree/chromium/TestShell.cpp:
2779 (TestShell::initialize):
2780 (TestShell::createMainWindow):
2781 (TestShell::~TestShell):
2782 (TestShell::resetTestController):
2783 (TestShell::bindJSObjectsToWindow):
2784 * DumpRenderTree/chromium/TestShell.h:
2785 (TestShell::eventSender):
2787 * DumpRenderTree/chromium/WebViewHost.cpp:
2789 2012-07-31 Jochen Eisinger <jochen@chromium.org>
2791 Unreviewed. Update watchlist
2793 * Scripts/webkitpy/common/config/watchlist: Turned out to be less useful than I had hoped
2795 2012-07-31 Kent Tamura <tkent@chromium.org>
2797 Unreviewed, update watchlist
2799 * Scripts/webkitpy/common/config/watchlist:
2800 - Improve "Forms" filename pattern.
2801 - tkent stops watching Chromium DRT. It was not so helpful.
2803 2012-07-31 Martin Robinson <mrobinson@igalia.com>
2805 [jhbuild] Also try to load jhbuild Python module from the source checkout
2807 Reviewed by Philippe Normand.
2809 JHBuild also seems to have an installation mode where it prefers to load its
2810 Python modules from the source checkout. This seems to be used particularly
2811 on the GTK+ 64-bit release bot. We should fall back to this mode.
2813 * jhbuild/jhbuildutils.py:
2814 (enter_jhbuild_environment_if_available): Use the source checkout if there
2815 is no installation path.
2817 2012-07-30 Ryosuke Niwa <rniwa@webkit.org>
2819 Revert r124153 and land the "right" build fix per ap's comment on the bug 92676.
2821 * MiniBrowser/win/BrowserView.cpp:
2822 (BrowserView::create):
2824 2012-07-30 Sam Weinig <sam@webkit.org>
2826 Rename WebKit2.h to WebKit2_C.h
2827 https://bugs.webkit.org/show_bug.cgi?id=92704
2829 Reviewed by Dan Bernstein.
2831 * MiniBrowser/mac/MiniBrowser_Prefix.pch:
2832 * MiniBrowser/win/BrowserView.h:
2833 * TestWebKitAPI/config.h:
2834 * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2835 * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
2836 * WebKitTestRunner/WebKitTestRunnerPrefix.h:
2837 * WebKitTestRunner/config.h:
2838 Update for changed header file name.
2840 2012-07-30 Dirk Pranke <dpranke@chromium.org>
2842 chromium win compile broken after removing webkit.py
2843 https://bugs.webkit.org/show_bug.cgi?id=92549
2845 Reviewed by Ryosuke Niwa.
2847 Remove the stub file for webkit.py now that it's no longer
2850 * Scripts/webkitpy/layout_tests/port/webkit.py: Removed.
2852 2012-07-30 Ryosuke Niwa <rniwa@webkit.org>
2854 Windows build fix attempt after r124092.
2856 * MiniBrowser/win/BrowserView.cpp:
2857 (BrowserView::create):
2859 2012-07-30 Ryosuke Niwa <rniwa@webkit.org>
2861 Remove erroneously committed debugging print call from a test.
2863 * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2864 (test_run_generates_results_page):
2866 2012-07-30 Arnaud Renevier <a.renevier@sisa.samsung.com>
2868 webkit-patch: system keyring is not used to read my password
2869 https://bugs.webkit.org/show_bug.cgi?id=92532
2871 Reviewed by Dirk Pranke.
2873 In case no username can be fetched from environment, git or keychain,
2874 prompt for username, and try to get password from keyring associated
2877 * Scripts/webkitpy/common/net/credentials.py:
2878 (Credentials.read_credentials):
2879 * Scripts/webkitpy/common/net/credentials_unittest.py:
2880 (test_keyring_without_git_repo_nor_keychain):
2881 (test_keyring_without_git_repo_nor_keychain.MockKeyring):
2882 (test_keyring_without_git_repo_nor_keychain.MockKeyring.get_password):
2883 (test_keyring_without_git_repo_nor_keychain.FakeCredentials):
2884 (test_keyring_without_git_repo_nor_keychain.FakeCredentials._credentials_from_keychain):
2885 (test_keyring_without_git_repo_nor_keychain.FakeCredentials._credentials_from_environment):
2886 (test_keyring_without_git_repo_nor_keychain.FakeUser):
2887 (test_keyring_without_git_repo_nor_keychain.FakeUser.prompt):
2888 (test_keyring_without_git_repo_nor_keychain.FakeUser.prompt_password):
2890 2012-07-30 Dirk Pranke <dpranke@chromium.org>
2892 nrwt: move the code that identifies the chunk of tests to run into finder
2893 https://bugs.webkit.org/show_bug.cgi?id=92694
2895 Reviewed by Ryosuke Niwa.
2897 Another patch to make manager.py smaller and clearer; this patch moves
2898 the handling of breaking the list of tests into chunks (--run-part,
2899 --run-chunk) from manager.py to finder.py.
2901 There are no functional changes and this is covered by existing tests.
2903 * Scripts/webkitpy/layout_tests/controllers/finder.py:
2905 (Finder._read_test_files):
2907 (Finder.split_into_chunks_if_necessary):
2908 * Scripts/webkitpy/layout_tests/controllers/manager.py:
2910 (Manager._collect_tests):
2911 (Manager._parse_expectations):
2912 (Manager.prepare_lists_and_print_output):
2914 2012-07-30 Dirk Pranke <dpranke@chromium.org>
2916 nrwt: split test-finding code out from manager.py
2917 https://bugs.webkit.org/show_bug.cgi?id=92693
2919 Reviewed by Ryosuke Niwa.
2921 In the interest of making manager.py smaller, this patch moves
2922 the code that actually takes the command line arguments and
2923 --test-file lists of tests to run and expands them into an
2924 actual list of tests out into a separate module.
2926 * Scripts/webkitpy/layout_tests/controllers/finder.py: Added.
2928 (LayoutTestFinder.__init__):
2929 (LayoutTestFinder.find_tests):
2930 (LayoutTestFinder._strip_test_dir_prefixes):
2931 (LayoutTestFinder._strip_test_dir_prefix):
2932 (LayoutTestFinder._read_test_names_from_file):
2933 * Scripts/webkitpy/layout_tests/controllers/manager.py:
2934 (Manager._collect_tests):
2935 (Manager._handle_finished_test):
2936 * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2939 2012-07-30 Adam Barth <abarth@webkit.org>
2941 commit-queue is corrupting ChangeLogs
2942 https://bugs.webkit.org/show_bug.cgi?id=92681
2944 Reviewed by Tony Chang.
2946 Some commits made with the commit-queue are ending up with duplicate
2947 ChangeLog entries. I've hot-patched the servers to not use a special
2948 merge driver for ChangeLogs, which should fix this issue. This patch
2949 removes the merge driver from our cold-boot.sh script so that we won't
2950 use it on new commit-queue instances.
2952 * EWSTools/cold-boot.sh:
2954 2012-07-28 Sam Weinig <sam@webkit.org>
2956 Add ability to load from a string to the ObjC WK API
2957 https://bugs.webkit.org/show_bug.cgi?id=92590
2959 Reviewed by Dan Bernstein.
2961 Add tests for [WKBrowsingContextController loadHTMLString:baseURL:]. We will be able
2962 to greatly improve these tests (to test more than just not crashing) when methods to
2963 access page content are added (soon!).
2966 Test: WKBrowsingContextLoadDelegateTest_SimpleLoadOfHTMLString
2967 Test: WKBrowsingContextLoadDelegateTest_SimpleLoadOfHTMLString_NilBaseURL
2968 Test: WKBrowsingContextLoadDelegateTest_SimpleLoadOfHTMLString_NilHTMLStringAndBaseURL
2970 * TestWebKitAPI/Tests/WebKit2ObjC/WKBrowsingContextLoadDelegateTest.mm:
2971 Adds tests and refactors delegates to not use global state.
2973 2012-07-30 Thiago Marcos P. Santos <thiago.santos@intel.com>
2975 [CMake] TestWebKitAPI bundle should link with WTF
2976 https://bugs.webkit.org/show_bug.cgi?id=92616
2978 Reviewed by Antonio Gomes.
2980 Fixes build when shared core is disabled.
2982 * TestWebKitAPI/CMakeLists.txt:
2984 2012-07-30 Dirk Pranke <dpranke@chromium.org>
2986 nrwt: clean up handling of 'expected' stats
2987 https://bugs.webkit.org/show_bug.cgi?id=92527
2989 Reviewed by Tony Chang.
2991 This patch alters the way we compute and log the "expected"
2992 results and how we treat skipped tests; we will now log the
2993 number of skipped tests separately from the categories, e.g.:
2995 Found 31607 tests; running 24464.
2996 Expect: 23496 passes (23496 now, 0 wontfix)
2997 Expect: 548 failures ( 543 now, 5 wontfix)
2998 Expect: 420 flaky ( 245 now, 175 wontfix)
3000 (so that the "expect" totals add up to the "running" totals);
3001 in addition, the totals in the one-line-progress reflect the
3002 number of tests we will actually run. If --iterations or
3003 --repeat-each are specified, the number of tests we run are
3004 multiplied as appropriate, but the "expect" numbers are
3005 unchanged, since we don't count multiple invocations of the same
3006 test multiple times. In addition, if we are using --run-part or
3007 --run-chunk, the tests we don't run are treated as skipped
3008 for consistency. We will also log the values for --iterations
3009 and --repeat each as part of the found/running line.
3011 Previously the code had parsed and re-parsed the
3012 TestExpectations files several times in an attempt to come up
3013 with some sane statistics, but this was expensive and lead to
3014 confusing layer; treating files as skipped in the way described
3015 above is more consistent and cleaner.
3017 * Scripts/webkitpy/layout_tests/controllers/manager.py:
3018 (Manager._split_into_chunks_if_necessary):
3019 (Manager.prepare_lists_and_print_output):
3021 * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3022 (ManagerTest.test_interrupt_if_at_failure_limits):
3023 (ManagerTest.test_update_summary_with_result):
3024 (ManagerTest.test_look_for_new_crash_logs):
3025 (ResultSummaryTest.get_result_summary):
3026 * Scripts/webkitpy/layout_tests/models/result_summary.py:
3027 (ResultSummary.__init__):
3028 * Scripts/webkitpy/layout_tests/models/test_expectations.py:
3029 (TestExpectationParser.expectation_for_skipped_test):
3030 (TestExpectations.__init__):
3031 (TestExpectations.add_skipped_tests):
3032 Here we make add_skipped_tests() public, so that we can update
3033 the expectations for tests that we are skipping due to
3034 --run-part or --run-chunk; we use the wontfix flag so that
3035 the tests that are intentionally skipped aren't considered
3037 * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
3038 (SkippedTests.check):
3039 * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3041 * Scripts/webkitpy/layout_tests/views/printing.py:
3042 (Printer.print_found):
3044 (Printer.print_expected):
3045 (Printer._print_result_summary):
3046 (Printer._print_result_summary_entry):
3047 Here we split out printing the number of tests found and run
3048 from the expected results, to be clearer and so that we don't
3049 have to reparse the expectations to update the stats.
3050 * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
3051 (Testprinter.get_result_summary):
3053 2012-07-30 Sadrul Habib Chowdhury <sadrul@chromium.org>
3055 Propagate gesture events to plugins.
3056 https://bugs.webkit.org/show_bug.cgi?id=92281
3058 Reviewed by Adam Barth.
3060 Update TestWebPlugin to receive events, and output logs for the events.
3062 * DumpRenderTree/chromium/TestWebPlugin.cpp:
3063 (TestWebPlugin::handleInputEvent):
3064 * DumpRenderTree/chromium/TestWebPlugin.h:
3065 (TestWebPlugin::acceptsInputEvents):
3068 2012-07-30 Raphael Kubo da Costa <rakuco@webkit.org>
3070 [jhbuild] gnutls 2.12.14 does not build with glibc 2.16.0
3071 https://bugs.webkit.org/show_bug.cgi?id=90643
3073 Reviewed by Gustavo Noronha Silva.
3075 The 2.x series of gnutls has a version of Gnulib that is
3076 incompatible with glibc 2.16.0, which now does not define gets()
3079 After talking to kov, mrobinson and philn, it looks like it makes
3080 sense to simply remove gnutls from jhbuild.modules and rely on the
3081 version installed on the system being recent enough.
3083 This means at least version 2.12.8 for PKCS11 support, or at the
3084 very least 2.11.0 (these are the minimum versions enforced by
3085 glib-networking's configure.ac).
3087 * efl/jhbuild.modules: Do not build gnutls anymore, make
3088 glib-networking depend on glib instead, and move the libgcrypt
3089 dependency to eet, which is what currently needs it.
3090 * gtk/jhbuild.modules: Do not build gnutls anymore.
3092 2012-07-30 Dirk Pranke <dpranke@chromium.org>
3094 nrwt: --verbose logging does not work right on windows
3095 https://bugs.webkit.org/show_bug.cgi?id=92673
3097 Unreviewed, build fix.
3099 The --verbose log level is not being propagated to the worker
3100 processes on windows properly; this has been broken ever since I
3101 refactored logging into the message pool directly :(.
3103 Also, fixing this revealed that outputcapture wasn't saving and
3104 restoring log levels correctly, so I had to fix that as well.
3106 * Scripts/webkitpy/common/message_pool.py:
3107 (_MessagePool._start_workers):
3108 (_MessagePool._worker_log_level):
3110 (_Worker._set_up_logging):
3111 (_WorkerLogHandler.__init__):
3112 * Scripts/webkitpy/common/system/outputcapture.py:
3113 (OutputCapture.capture_output):
3114 (OutputCapture.restore_output):
3116 2012-07-30 Dirk Pranke <dpranke@chromium.org>
3118 After r123895, new-run-webkit-tests can fail with TypeError when a test crashes
3119 https://bugs.webkit.org/show_bug.cgi?id=92664
3121 Reviewed by Michael Saboff.
3123 Updated to handle crash_logs being returned as a tuple of (stderr, crash_log).
3125 * Scripts/webkitpy/layout_tests/port/mac.py:
3126 (MacPort.look_for_new_crash_logs):
3128 2012-07-30 Stephen White <senorblanco@chromium.org>
3130 Added a SkiaGraphics definition to the watchlist file; cc'ed myself
3131 to that group. Unreviewed.
3133 * Scripts/webkitpy/common/config/watchlist:
3135 2012-07-30 Sheriff Bot <webkit.review.bot@gmail.com>
3137 Unreviewed, rolling out r123966 and r123967.
3138 http://trac.webkit.org/changeset/123966
3139 http://trac.webkit.org/changeset/123967
3140 https://bugs.webkit.org/show_bug.cgi?id=92656
3142 This patch is causing assertion failures on the debug bot
3143 (also rolling out a dependent patch) (Requested by mrobinson
3146 * gtk/generate-gtkdoc:
3147 (get_webkit2_options):
3149 2012-07-30 Martin Robinson <mrobinson@igalia.com>
3151 [GTK] Add a non-subprocess jhbuild environment and use it for run-gtk-tests
3152 https://bugs.webkit.org/show_bug.cgi?id=92626
3154 Reviewed by Gustavo Noronha Silva.
3156 Instead of always shelling out to enter a jhbuild environment, add a Python
3157 hook for entering a jhbuild environment directly. This avoids requiring a
3158 script to be wrapped in jhbuild.
3160 * GNUmakefile.am: Remove gtk/run-api-tests from the distribution list.
3161 * Scripts/run-gtk-tests: Move most of the logic from run-api-tests here
3162 with additional code for entering a jhbuild environment directly.
3163 * gtk/run-api-tests: Removed.
3164 * jhbuild/jhbuild-wrapper: Use the new helper methods defined in jhbuildutils.py.
3165 * jhbuild/jhbuildutils.py: Add some new helper methods here including
3166 a method which can transform the current environment into a jhbuild-ified one.
3168 2012-07-30 Alexander Pavlov <apavlov@chromium.org>
3170 Unreviewed, upgrade apavlov to reviewer.
3171 http://lists.webkit.org/mailman/private/webkit-committers/2012-July/000184.html
3173 * Scripts/webkitpy/common/config/committers.py:
3175 2012-07-30 Robin Cao <robin.cao@torchmobile.com.cn>
3177 [BlackBerry] Adapt to changes in the SharedArray platform API
3178 https://bugs.webkit.org/show_bug.cgi?id=92631
3180 Reviewed by Rob Buis.
3182 Adapt to changes in the SharedArray platform API. No behavioural change.
3184 * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
3185 (LayoutTestController::webHistoryItemCount):
3187 2012-07-30 Balazs Kelemen <kbalazs@webkit.org>
3189 [Qt][NRWT] REGRESSION(123729): Forcing pixel tests with -p doesn't work
3190 https://bugs.webkit.org/show_bug.cgi?id=92627
3192 Reviewed by Noam Rosenthal.
3194 Remove the default behavior I added in r123729 so that we run all tests as pixel
3195 test with -p. It was intentional but it seems like we don't have consensus on
3198 * Scripts/webkitpy/layout_tests/port/qt.py:
3199 (QtPort._supports_switching_pixel_tests_per_test):
3201 2012-07-30 Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com>
3203 [EFL][WTR] Enforce SHIFT modifier to upper case keys in event sender.
3204 https://bugs.webkit.org/show_bug.cgi?id=92366
3206 Reviewed by Antonio Gomes.
3208 Add SHIFT modifier for capital letters as VK_
3209 codes do not distinguish between lows and caps and also to be consistent
3210 with other ports which send caps as lowercase letters with the shift key down.
3212 * WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
3213 (WTR::EventSenderProxy::keyDown):
3215 2012-07-29 Raphael Kubo da Costa <rakuco@webkit.org>
3217 [EFL][jhbuild] Make run-with-jhbuild become the jhbuild process itself.
3218 https://bugs.webkit.org/show_bug.cgi?id=92592
3220 Reviewed by Eric Seidel.
3222 Adapt to the changes made to GTK+'s run-with-jhbuild in r123979
3223 and make EFL's run-with-jhbuild call os.execve() instead of
3224 launching a subprocess as well.
3226 This should make it possible to interrupt execution of jhbuild
3229 * efl/run-with-jhbuild:
3231 2012-07-28 Ryosuke Niwa <rniwa@webkit.org>
3233 run-perf-tests should generate a results page
3234 https://bugs.webkit.org/show_bug.cgi?id=92575
3236 Reviewed by Eric Seidel.
3238 Added the ability to generate a results page that summarizes performance test results from
3239 multiple runs of run-perf-tests when --output-json-path is specified and --test-results-server
3240 is not specified. We cannot generate a results page when --test-results-server is specified
3241 because perf-o-matic cannot parse the new JSON format.
3243 The new JSON format is simply an array of the old JSON output. This JSON is then merged into
3244 results-template.html along with jQuery and copied as a HTML file of the same filename as
3245 the output JSON file with a .html extension.
3247 We merge all scripts and the JSON output into the results page as opposed to including them
3248 with the src content attribute to make it standalone so that we may post it on Bugzilla or
3249 e-mail it to someone else without breaking it.
3251 * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3252 (PerfTestsRunner.run):
3253 (PerfTestsRunner._generate_json): Merge "contents" with the existing JSON file when generating
3254 a results page, and generate the results page from results-template.html merged with jQuery and
3255 the generated JSON output.
3256 * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3257 (_test_run_with_json_output): Extracted from test_run_with_upload_json to be shared with
3259 (_test_run_with_json_output.mock_upload_json): Ditto.
3260 (test_run_with_json_output): Refactored to use _test_run_with_json_output.
3261 (test_run_generates_results_page): Added. Tests the new feature.
3262 (test_run_with_json_source): Refactored to use _test_run_with_json_output.
3263 (test_run_with_multiple_repositories): Ditto.
3264 (test_run_with_upload_json): Ditto.
3266 2012-07-29 Rik Cabanier <cabanier@adobe.com>
3268 Add ENABLE_CSS_COMPOSITING flag
3269 https://bugs.webkit.org/show_bug.cgi?id=92553
3271 Reviewed by Dirk Schulze.
3273 Adds compiler flag CSS_COMPOSITING to build systems to enable CSS blending and compositing. See spec https://dvcs.w3.org/hg/FXTF/rawfile/tip/compositing/index.html
3275 * Scripts/webkitperl/FeatureList.pm:
3277 2012-07-29 Gustavo Noronha Silva <gns@gnome.org>
3279 Reviewed by Martin Robinson.
3281 Make the jhbuild wrapper scripts become the jhbuild process
3282 instead of starting it as a child. This should make it possible to
3283 interrupt execution of jhbuild cleanly.
3285 * gtk/run-with-jhbuild:
3286 * jhbuild/jhbuild-wrapper:
3288 2012-07-29 Mike West <mkwst@chromium.org>
3290 Extend `application/x-webkit-test-netscape` plugins to better support multiple frames.
3291 https://bugs.webkit.org/show_bug.cgi?id=92478
3293 Reviewed by Adam Barth.
3295 DumpRenderTree currently understands
3296 `<object src="data:application/x-webkit-test-netscape,alertwhenloaded">`.
3297 It would be useful if it understood `logifloaded`, which could give
3298 more context regarding test expectations, and can be written in such
3299 as way as to support plugins loaded into multiple iframes in a single
3302 * DumpRenderTree/TestNetscapePlugIn/main.cpp:
3304 Adding support for `logifloaded`.
3306 2012-07-28 Peter Kasting <pkasting@google.com>
3308 Fix Chromium/Win compile.
3309 https://bugs.webkit.org/show_bug.cgi?id=88787
3311 Unreviewed, build fix.
3313 Chromium/Win needs _USE_MATH_DEFINES but doesn't define PLATFORM(WIN).
3314 The other files in WebKit that define this macro (both in
3315 WebCore/platform/graphics/cg) do it by unconditionally #defining it to 1
3316 above the #include of "config.h", so copy that pattern here.
3318 * TestWebKitAPI/Tests/WTF/MediaTime.cpp:
3320 2012-07-28 Mario Sanchez Prada <msanchez@igalia.com>
3322 [WK2][GTK] Implement a new spell checker API for WebKit2GTK+
3323 https://bugs.webkit.org/show_bug.cgi?id=90268
3325 Reviewed by Martin Robinson.
3327 Ignore WebKitTextChecker.* private files for gtk-doc.
3329 * gtk/generate-gtkdoc:
3330 (get_webkit2_options):
3332 2012-07-28 Philippe Normand <pnormand@igalia.com>
3334 [GTK] Transparent Media controls timeline and panel
3335 https://bugs.webkit.org/show_bug.cgi?id=85279
3337 Reviewed by Martin Robinson.
3339 * gtk/jhbuild.modules: Provide gnome-themes-standard build
3340 support, this is necessary so the media controls panel colors are
3341 properly initialized from the GTK+ theme.
3343 2012-07-27 Kihong Kwon <kihong.kwon@samsung.com>
3345 [EFL] Support for HTML media capture
3346 https://bugs.webkit.org/show_bug.cgi?id=91842
3348 Reviewed by Ryosuke Niwa.
3350 Add HTML media capture feature to the feature list.
3351 In addition, enable that on the Efl port.
3353 * Scripts/webkitperl/FeatureList.pm:
3355 2012-07-27 Dirk Pranke <dpranke@chromium.org>
3357 "webkit-patch rebaseline <test>" fails with exit_code 1
3358 https://bugs.webkit.org/show_bug.cgi?id=91560
3360 Reviewed by Adam Barth.
3362 webkit-patch rebaseline-json crashes in an svn checkout if it
3363 tries to add files and there are no files to add; arguably
3364 svn.add() should be robust against this, but rebaseline-json
3365 shouldn't be calling it, either. This patch fixes the latter and
3366 adds better debugging info to garden-o-matic so we can see what
3369 * Scripts/webkitpy/tool/commands/rebaseline.py:
3370 (AbstractParallelRebaselineCommand._rebaseline):
3371 * Scripts/webkitpy/tool/servers/gardeningserver.py:
3372 (GardeningHTTPRequestHandler.rebaselineall):
3373 (GardeningHTTPRequestHandler.rebaselineall.error_handler):
3374 (GardeningHTTPRequestHandler):
3376 2012-07-27 Brady Eidson <beidson@apple.com>
3378 Plugins should not be allowed to override standard properties/attributes in non-standard worlds
3379 <rdar://problem/11975252> and https://bugs.webkit.org/show_bug.cgi?id=92519
3381 Reviewed by Anders Carlsson.
3383 Add a test plugin which overrides all properties it is asked about:
3384 * DumpRenderTree/TestNetscapePlugIn/Tests/PluginScriptableObjectOverridesAllProperties.cpp: Added.
3385 (PluginScriptableObjectOverridesAllProperties):
3386 (PluginScriptableObjectOverridesAllProperties::PluginScriptableObjectOverridesAllProperties):
3388 (PluginScriptableObjectOverridesAllProperties::PluginObject::PluginObject):
3389 (PluginScriptableObjectOverridesAllProperties::PluginObject::~PluginObject):
3390 (PluginScriptableObjectOverridesAllProperties::PluginObject::hasProperty):
3391 (PluginScriptableObjectOverridesAllProperties::PluginObject::getProperty):
3392 (PluginScriptableObjectOverridesAllProperties::NPP_GetValue):
3394 Expose NPN_MemAlloc to the test plugin:
3395 * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
3396 (PluginTest::NPN_MemAlloc):
3397 * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
3400 * DumpRenderTree/DumpRenderTree.gypi:
3401 * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
3402 * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
3403 * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
3405 2012-07-27 Rafael Brandao <rafael.lobo@openbossa.org>
3407 "webkit-patch upload" should strip colors from patch if necessary
3408 https://bugs.webkit.org/show_bug.cgi?id=92306
3410 Reviewed by Adam Barth.
3412 * Scripts/webkitpy/common/checkout/scm/git.py: Explicity remove colors from diff.
3414 2012-07-27 Eric Seidel <eric@webkit.org>
3416 clean-pending-commit throws exception when encountering Ossy
3417 https://bugs.webkit.org/show_bug.cgi?id=92534
3419 Reviewed by Adam Barth.
3421 Mark a couple strings as unicode so that reviewers with unicode names do not cause
3422 clean-pending-commit to barf.
3424 * Scripts/webkitpy/tool/commands/upload.py:
3425 (CleanPendingCommit._flags_to_clear_on_patch):
3426 (CleanPendingCommit.execute):
3427 (AssignToCommitter._assign_bug_to_last_patch_attacher):
3429 2012-07-27 Dirk Pranke <dpranke@chromium.org>
3431 chromium win compile broken after removing webkit.py
3432 https://bugs.webkit.org/show_bug.cgi?id=92549
3434 Unreviewed, build fix.
3436 It turns out that this file is listed in the chromium downstream
3437 'browser_tests.isolate' file, and so deleting it broke gyp.
3438 Adding a stub file back in until we can remove the entry from
3441 * Scripts/webkitpy/layout_tests/port/webkit.py: Added.
3443 2012-07-27 Dirk Pranke <dpranke@chromium.org>
3445 nrwt: remove 'webkit.py' and the 'WebKitPort'
3446 https://bugs.webkit.org/show_bug.cgi?id=92450
3448 Reviewed by Ryosuke Niwa.
3450 This patch finishes the work of merging WebKitPort into Port.
3451 Now Port is way too big :).
3453 This is all cutting and pasting; no new functionality and
3454 no changes in the tests.
3456 * Scripts/webkitpy/layout_tests/port/apple.py:
3458 (ApplePort.__init__):
3459 * Scripts/webkitpy/layout_tests/port/base.py:
3460 (Port.skipped_layout_tests):
3462 (Port._port_flag_for_scripts):
3463 (Port._arguments_for_configuration):
3465 (Port._build_driver):
3466 (Port._build_driver_flags):
3467 (Port._tests_for_other_platforms):
3468 (Port._runtime_feature_list):
3470 (Port._modules_to_search_for_symbols):
3471 (Port._symbols_string):
3472 (Port._missing_feature_to_skipped_tests):
3473 (Port._missing_symbol_to_skipped_tests):
3474 (Port._has_test_in_directories):
3475 (Port._skipped_tests_for_unsupported_features):
3476 (Port._wk2_port_name):
3477 (Port._skipped_file_search_paths):
3478 * Scripts/webkitpy/layout_tests/port/chromium.py:
3480 * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3481 * Scripts/webkitpy/layout_tests/port/chromium_port_testcase.py:
3482 (ChromiumPortTestCase.test_check_build):
3483 * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
3484 * Scripts/webkitpy/layout_tests/port/efl.py:
3487 * Scripts/webkitpy/layout_tests/port/gtk.py:
3489 (GtkPort.setup_environ_for_server):
3490 * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3492 (TestWebKitPort.__init__):
3493 (TestWebKitPort.all_test_configurations):
3494 (TestWebKitPort._symbols_string):
3495 (TestWebKitPort._tests_for_other_platforms):
3496 (TestWebKitPort._tests_for_disabled_features):
3498 (PortTestCase.test_diff_image):
3499 (PortTestCase.test_path_to_test_expectations_file):
3500 (PortTestCase.test_skipped_directories_for_symbols):
3501 (test_skipped_directories_for_features):
3502 (test_skipped_directories_for_features_no_matching_tests_in_test_list):
3503 (test_skipped_tests_for_unsupported_features_empty_test_list):
3504 (test_skipped_layout_tests):
3505 (test_skipped_file_search_paths):
3507 (test_test_expectations):
3508 (test_build_driver):
3509 (_assert_config_file_for_platform):
3510 (test_linux_distro_detection):
3511 (test_apache_config_file_name_for_platform):
3512 (test_path_to_apache_config_file):
3513 * Scripts/webkitpy/layout_tests/port/qt.py:
3516 (QtPort.setup_environ_for_server):
3517 * Scripts/webkitpy/layout_tests/port/webkit.py: Removed.
3518 * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Removed.
3519 * Scripts/webkitpy/tool/servers/rebaselineserver.py:
3520 (get_test_baselines.AllPlatformsPort):
3521 (get_test_baselines.AllPlatformsPort.__init__):
3522 * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
3523 (get_test_config.TestMacPort):
3525 2012-07-27 Christophe Dumez <christophe.dumez@intel.com>
3527 [WK2][WTR] LayoutTestController.deliverWebIntent() needs to be implemented
3528 https://bugs.webkit.org/show_bug.cgi?id=92228
3530 Reviewed by Anders Carlsson.
3532 Implement LayoutTestController.deliverWebIntent() to
3533 allow test cases to deliver a Web intent and test this
3536 This functionality is needed by:
3537 - webintents/web-intents-delivery.html
3539 * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
3540 * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3541 (WTR::LayoutTestController::deliverWebIntent):
3543 * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
3544 (LayoutTestController):
3546 2012-07-27 Jer Noble <jer.noble@apple.com>
3548 Unreviewed build fix.
3550 Wrap MSVC-specific fixes is a COMPILER() check, not a PLATFORM() check.
3552 * TestWebKitAPI/Tests/WTF/MediaTime.cpp:
3554 2012-07-27 Xianzhu Wang <wangxianzhu@chromium.org>
3556 [Chromium-Android] NRWT better handling of DRT deadlocks and crashes
3557 https://bugs.webkit.org/show_bug.cgi?id=92299
3559 Reviewed by Dirk Pranke.
3561 1. Changed the deadlock detector thread target from a method of driver to a function to avoid its reference to the
3562 current driver (see the bug for the problem of the reference.
3563 2. Let the deadlock detector thread exit early when the normal_startup_event is set to avoid it from blocking the main thread when the main thread is exiting.
3564 3. Extracted common loop until timeout logic into ChromiumAndroidDriver._loop_with_timeout().
3565 4. Use the "for i in range(n)" style suggested by dpranke in https://bugs.webkit.org/show_bug.cgi?id=89124 in several functions.
3567 * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3568 (ChromiumAndroidPort.__init__):
3569 (ChromiumAndroidPort._run_adb_command): Reduced the verbose output.
3570 (ChromiumAndroidDriver.__init__):
3571 (ChromiumAndroidDriver._loop_with_timeout): Extracted common loop until timeout logic here.
3572 (ChromiumAndroidDriver._all_pipes_created):
3573 (ChromiumAndroidDriver._remove_all_pipes):
3574 (ChromiumAndroidDriver._start): Changed the retry loop style.
3575 (ChromiumAndroidDriver._start_once): Changed the loop style and code related to deadlock detector.
3576 (ChromiumAndroidDriver._start_once.deadlock_detector): Moved the original ChromiumAndroidDriver._deadlock_detector() here.
3577 (ChromiumAndroidDriver.stop): Changed the loop style.
3579 2012-07-27 Dirk Pranke <dpranke@chromium.org>
3581 nrwt: move collect_tests(), parse_expectations() into manager
3582 https://bugs.webkit.org/show_bug.cgi?id=92423
3584 Reviewed by Ojan Vafai.
3586 This is the first in a series of patches refactoring
3587 run_webkit_tests.py and manager.py to make them more
3588 maintainable; all this patch does is move a couple of functions
3589 into manager.run_tests() so manager's public interface is
3592 * Scripts/webkitpy/layout_tests/controllers/manager.py:
3593 (Manager._collect_tests):
3594 (Manager._parse_expectations):
3595 (Manager._split_into_chunks_if_necessary):
3598 * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3599 (ManagerTest.test_http_locking):
3600 (ManagerTest.test_perf_locking):
3601 (ManagerTest.integration_test_needs_servers.get_manager_with_tests):
3602 (ManagerTest.test_look_for_new_crash_logs.get_manager_with_tests):
3603 * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3606 2012-07-27 Dirk Pranke <dpranke@chromium.org>
3608 nrwt: print filtered stderr after a crash correctly
3609 https://bugs.webkit.org/show_bug.cgi?id=92428
3611 Reviewed by Adam Barth.
3613 Some ports will run the stderr output from DRT/WTR through a
3614 filter after a crash in order to print a more useful stack
3615 trace; while that gets written to the crash log correctly, we
3616 don't log it to the debug output properly. This change fixes
3617 that and makes it clearer that we're logging stderr, not the
3620 * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
3621 (SingleTestRunner._handle_error):
3622 * Scripts/webkitpy/layout_tests/port/base.py:
3623 (Port._get_crash_log):
3624 * Scripts/webkitpy/layout_tests/port/chromium.py:
3625 (ChromiumPort._get_crash_log):
3626 * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
3627 (ChromiumAndroidPortTest.test_get_crash_log):
3628 * Scripts/webkitpy/layout_tests/port/driver.py:
3630 * Scripts/webkitpy/layout_tests/port/gtk.py:
3631 * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
3632 * Scripts/webkitpy/layout_tests/port/mac.py:
3633 (MacPort._get_crash_log):
3634 * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3635 (PortTestCase.test_get_crash_log):
3637 2012-07-27 Dirk Pranke <dpranke@chromium.org>
3639 nrwt: move image diffing code to a separate module
3640 https://bugs.webkit.org/show_bug.cgi?id=92447
3642 Reviewed by Ryosuke Niwa.
3644 This patch moves the code to&nb