performance tests should be able to measure runs/sec rather than time
[WebKit-https.git] / Tools / ChangeLog
1 2012-05-13  Ryosuke Niwa  <rniwa@webkit.org>
2
3         performance tests should be able to measure runs/sec rather than time
4         https://bugs.webkit.org/show_bug.cgi?id=86021
5
6         Reviewed by Ojan Vafai.
7
8         Allow " runs/s" or " ms" to appear after numerical values in tests.
9
10         * Scripts/webkitpy/performance_tests/perftest.py:
11         (PerfTest):
12
13 2012-05-12  Tim Horton  <timothy_horton@apple.com>
14
15         run-safari and co. should support --guard-malloc command line argument
16         https://bugs.webkit.org/show_bug.cgi?id=86287
17
18         Reviewed by Dan Bernstein.
19
20         Add setUpGuardMallocIfNeeded(), which consults the argument --guard-malloc on OS X,
21         setting DYLD_INSERT_LIBRARIES="/usr/lib/libgmalloc.dylib" if it's enabled.
22
23         Use setUpGuardMallocIfNeeded() before running a WebKit app (whether standalone or
24         in the debugger). This will cause gmalloc to be loaded for gdb as well, but this
25         tends to be harmless.
26
27         Change "malloc guard" to the more common "Guard Malloc" in run-webkit-tests.
28
29         * Scripts/old-run-webkit-tests:
30         * Scripts/webkitdirs.pm:
31         (appendToEnvironmentVariableList):
32         (setUpGuardMallocIfNeeded):
33         (printHelpAndExitForRunAndDebugWebKitAppIfNeeded):
34         (runMacWebKitApp):
35         (execMacWebKitAppForDebugging):
36         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
37         (parse_args):
38
39 2012-05-11  Ojan Vafai  <ojan@chromium.org>
40
41         Add previous/next buttons the garden-o-matics perf panel
42         https://bugs.webkit.org/show_bug.cgi?id=86280
43
44         Reviewed by Dimitri Glazkov.
45
46         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf.js:
47         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf_unittests.js:
48         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/perf.css:
49
50 2012-05-11  Kevin Ollivier  <kevino@theolliviers.com>
51
52         [wx] Unreviewed build fix. Fix memory exhausted errors from 32-bit ld on *nix.
53
54         * waf/build/settings.py:
55         (common_configure):
56
57 2012-05-11  Dirk Pranke  <dpranke@chromium.org>
58
59         webkit-patch rebaseline should fetch stale expectations :)
60         https://bugs.webkit.org/show_bug.cgi?id=86231
61
62         Unreviewed, build fix.
63
64         This change reverts r115492 and goes back to fetching individual
65         baselines from the layout tests directory, rather than fetching
66         entire zip files. The performance hit over networks for people
67         far away from the bots has turned out to be severe, so we need a
68         better solution to this.
69
70         * Scripts/webkitpy/tool/commands/rebaseline.py:
71         (RebaselineTest._results_url):
72         (RebaselineTest._file_name_for_expected_result):
73         (RebaselineTest._rebaseline_test):
74         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
75         (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
76         (test_rebaseline_updates_expectations_file):
77         (test_rebaseline_test):
78         (test_rebaseline_and_copy_test):
79         (test_rebaseline_and_copy_test_no_existing_result):
80         (test_rebaseline_and_copy_test_with_lion_result):
81         (test_rebaseline_and_copy_no_overwrite_test):
82
83 2012-05-11  Christophe Dumez  <christophe.dumez@intel.com>
84
85         [GTK] LayoutTests/svg/batik/text/verticalText.svg fails to display Chinese characters
86         https://bugs.webkit.org/show_bug.cgi?id=86198
87
88         Reviewed by Martin Robinson.
89
90         Update webkitgtk-fonts package to v0.0.3 in order to get new Kochi
91         Mincho / Gothic fonts to display Japanese and Chinese characters.
92
93         * gtk/jhbuild.modules:
94
95 2012-05-11  Christophe Dumez  <christophe.dumez@intel.com>
96
97         [EFL][DRT] SVG Batik test fails to display Chinese characters
98         https://bugs.webkit.org/show_bug.cgi?id=84461
99
100         Reviewed by Martin Robinson.
101
102         Update webkitgtk-fonts package from v0.0.1 to v0.0.3 in order to get
103         new Kochi Mincho / Gothic fonts to display Japanese and Chinese
104         characters.
105
106         * DumpRenderTree/efl/CMakeLists.txt:
107         * efl/jhbuild.modules:
108
109 2012-05-11  Philippe Normand  <pnormand@igalia.com>
110
111         Unreviewed, GTK 64-bit Debug bot fix attempt.
112
113         * Scripts/webkitpy/layout_tests/port/gtk.py:
114         (GtkPort._get_gdb_output): Ignore unicode decoding errors when
115         processing gdb output.
116
117 2012-05-11  Gustavo Noronha Silva  <gns@gnome.org>
118
119         REGRESSION(r116205): [GTK]: build no longer supports thin archives
120         https://bugs.webkit.org/show_bug.cgi?id=86207
121
122         * GNUmakefile.am: link libWebCoreModules to DRT.
123
124 2012-05-10  Ojan Vafai  <ojan@chromium.org>
125
126         Add a perf tab to garden-o-matic
127         https://bugs.webkit.org/show_bug.cgi?id=86169
128
129         Reviewed by Dimitri Glazkov.
130
131         This is a first stab at making it easier to view the Chromium WebKit canary
132         perf graphs. For now, it's just a couple select elements and an iframe.
133
134         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
135         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
136         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
137         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
138         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
139         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
140         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf.js: Added.
141         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf_unittests.js: Added.
142         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
143         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/perf.css: Added.
144         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
145
146 2012-05-10  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
147
148         Move resumeAnimations to use Internals interface
149         https://bugs.webkit.org/show_bug.cgi?id=86063
150
151         Reviewed by Alexey Proskuryakov.
152
153         Remove resumeAnimations functions, because it is able to work in the 
154         cross-port way through the Internals interface.
155
156         * DumpRenderTree/LayoutTestController.cpp:
157         (LayoutTestController::staticFunctions):
158         * DumpRenderTree/LayoutTestController.h:
159         (LayoutTestController):
160         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
161         * DumpRenderTree/chromium/LayoutTestController.cpp:
162         (LayoutTestController::LayoutTestController):
163         * DumpRenderTree/chromium/LayoutTestController.h:
164         (LayoutTestController):
165         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
166         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
167         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
168         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
169         * DumpRenderTree/qt/LayoutTestControllerQt.h:
170         (LayoutTestController):
171         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
172         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
173         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
174         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
175         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
176         (LayoutTestController):
177
178 2012-05-10  Alexey Proskuryakov  <ap@apple.com>
179
180         Crash in 3rd party WebKit apps that disable cache at a wrong time
181         https://bugs.webkit.org/show_bug.cgi?id=86027
182         <rdar://problem/10615880>
183
184         Reviewed by Antti Koivisto.
185
186         Added a test that's very similar to MemoryCachePruneWithinResourceLoadDelegate,
187         but for disabling the cache instead of triggering a prune.
188
189         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
190         * TestWebKitAPI/Tests/mac/MemoryCacheDisableWithinResourceLoadDelegate.html: Added.
191         * TestWebKitAPI/Tests/mac/MemoryCacheDisableWithinResourceLoadDelegate.mm: Added.
192         (-[MemoryCacheDisableTestResourceLoadDelegate webView:identifierForInitialRequest:fromDataSource:]):
193         (-[MemoryCacheDisableTestResourceLoadDelegate webView:resource:willSendRequest:redirectResponse:fromDataSource:]):
194         (-[MemoryCacheDisableTestResourceLoadDelegate webView:resource:didFinishLoadingFromDataSource:]):
195         (-[MemoryCacheDisableTestResourceLoadDelegate webView:resource:didFailLoadingWithError:fromDataSource:]):
196         (TestWebKitAPI::TEST):
197
198 2012-05-10  Anders Carlsson  <andersca@apple.com>
199
200         WebKit2: Add a way to blacklist specific plug-ins/plug-in versions
201         https://bugs.webkit.org/show_bug.cgi?id=86164
202         <rdar://problem/9551196>
203
204         Reviewed by Sam Weinig.
205
206         Update for WK2 API changes.
207
208         * MiniBrowser/mac/BrowserWindowController.m:
209         (-[BrowserWindowController awakeFromNib]):
210         * WebKitTestRunner/TestController.cpp:
211         (WTR::TestController::createOtherPage):
212         (WTR::TestController::initialize):
213
214 2012-05-10  Dirk Pranke  <dpranke@chromium.org>
215
216         nrwt: [chromium] fix path to asan symbolize script
217         https://bugs.webkit.org/show_bug.cgi?id=86161
218
219         Reviewed by Abhishek Arya.
220
221         * Scripts/webkitpy/layout_tests/port/chromium.py:
222         (ChromiumPort._get_crash_log):
223
224 2012-05-10  Lei Zhang  <thestig@chromium.org>
225
226         [chromium] Set Valgrind environment variables for layout tests.
227         https://bugs.webkit.org/show_bug.cgi?id=86156
228
229         Reviewed by Dirk Pranke.
230
231         * Scripts/webkitpy/layout_tests/port/chromium.py:
232         (ChromiumPort.setup_environ_for_server):
233
234 2012-05-10  Raphael Kubo da Costa  <rakuco@webkit.org>
235
236         [webkitpy][EFL] It should be possible to use a wrapper command for ImageDiff.
237         https://bugs.webkit.org/show_bug.cgi?id=86145
238
239         Reviewed by Dirk Pranke.
240
241         The GTK+ and EFL ports use jhbuild to build a few base libraries
242         which are then used when building the port itself. Programs such
243         as DRT and ImageDiff are then supposed to be run with the
244         `run-with-jhbuild' wrapper so that the libraries are properly
245         found and loaded.
246
247         This was already the case for DumpRenderTree (EflPort defines the
248         'wrapper' option for that), but it was not possible to do the same
249         with ImageDiff so far.
250
251         The code which created the ImageDiff command-line call has been
252         moved to a separate method, _image_diff_command, which is then
253         overridden by EflPort to add the jhbuild wrapper.
254
255         * Scripts/webkitpy/layout_tests/port/efl.py:
256         (EflPort.__init__):
257         (EflPort._image_diff_command):
258         * Scripts/webkitpy/layout_tests/port/webkit.py:
259         (WebKitPort):
260         (WebKitPort._image_diff_command):
261         (WebKitPort._start_image_diff_process):
262
263 2012-05-10  Zan Dobersek  <zandobersek@gmail.com>
264
265         [GTK] ENABLE_IFRAME_SEAMLESS support
266         https://bugs.webkit.org/show_bug.cgi?id=85843
267
268         Reviewed by Eric Seidel.
269
270         Add the iframe-seamless option to the options list. This makes it
271         possible to enable the iframe seamless support through build-webkit
272         script. Enabled by default for all the ports.
273
274         * Scripts/webkitperl/FeatureList.pm:
275
276 2012-05-10  Raphael Kubo da Costa  <rakuco@webkit.org>
277
278         [EFL][webkitpy] Do not redefine check_build() in EflPort.
279         https://bugs.webkit.org/show_bug.cgi?id=86124
280
281         Reviewed by Dirk Pranke.
282
283         Remove an old FIXME and do not redefine check_build in EflPort;
284         once build-dumprendertree recognizes --efl, we can use the
285         check_build implementation in WebKitPort.
286
287         * Scripts/build-dumprendertree:
288         * Scripts/webkitpy/layout_tests/port/efl.py:
289         (EflPort.check_build):
290
291 2012-05-10  Tommy Widenflycht  <tommyw@google.com>
292
293         [chromium] MediaStream API: Fix the ExtraData functionality in WebMediaStreamDescriptor
294         https://bugs.webkit.org/show_bug.cgi?id=86087
295
296         Reviewed by Adam Barth.
297
298         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp:
299         (MockExtraData):
300         (WebKit):
301         (WebKit::WebUserMediaClientMock::requestUserMedia):
302
303 2012-05-10  Varun Jain  <varunjain@google.com>
304
305         [chromium] Trigger context menu for long press gesture
306         https://bugs.webkit.org/show_bug.cgi?id=85919
307
308         Reviewed by Adam Barth.
309
310         * DumpRenderTree/chromium/EventSender.cpp:
311         (EventSender::EventSender):
312         (EventSender::gestureLongPress):
313         (EventSender::gestureEvent):
314         * DumpRenderTree/chromium/EventSender.h:
315         (EventSender):
316
317 2012-05-10  Takashi Toyoshima  <toyoshim@chromium.org>
318
319         don't use autoinstall to import pywebsocket but check it in WebKit directly.
320         https://bugs.webkit.org/show_bug.cgi?id=86107
321
322         Reviewed by Adam Barth.
323
324         This change removes pywebsocket from webkitpy's autoinstall list and
325         imports pywebsocket 0.7.4 directly into webkitpy/thirparty.
326
327         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
328         (PyWebSocket._prepare_config):
329         * Scripts/webkitpy/thirdparty/__init__.py:
330         (AutoinstallImportHook.find_module):
331         (AutoinstallImportHook._install_irc):
332         * Scripts/webkitpy/thirdparty/__init___unittest.py:
333         (ThirdpartyTest.test_import_hook.MockImportHook.__init__):
334         (ThirdpartyTest.test_import_hook.MockImportHook._install_eliza):
335         (ThirdpartyTest):
336         (ThirdpartyTest.test_import_hook):
337         * Scripts/webkitpy/thirdparty/mod_pywebsocket/COPYING: Added.
338         * Scripts/webkitpy/thirdparty/mod_pywebsocket/__init__.py: Added.
339         * Scripts/webkitpy/thirdparty/mod_pywebsocket/_stream_base.py: Added.
340         (ConnectionTerminatedException):
341         (InvalidFrameException):
342         (BadOperationException):
343         (UnsupportedFrameException):
344         (InvalidUTF8Exception):
345         (StreamBase):
346         (StreamBase.__init__):
347         (StreamBase._read):
348         (StreamBase._write):
349         (StreamBase.receive_bytes):
350         (StreamBase._read_until):
351         * Scripts/webkitpy/thirdparty/mod_pywebsocket/_stream_hixie75.py: Added.
352         (for):
353         (StreamHixie75):
354         (StreamHixie75.for):
355         (StreamHixie75.__init__):
356         (StreamHixie75.send_message):
357         (StreamHixie75._read_payload_length_hixie75):
358         (StreamHixie75.receive_message):
359         (StreamHixie75._send_closing_handshake):
360         (StreamHixie75.close_connection):
361         (StreamHixie75.send_ping):
362         * Scripts/webkitpy/thirdparty/mod_pywebsocket/_stream_hybi.py: Added.
363         (Frame):
364         (Frame.__init__):
365         (create_length_header):
366         (create_header):
367         (_build_frame):
368         (_filter_and_format_frame_object):
369         (create_binary_frame):
370         (create_text_frame):
371         (FragmentedFrameBuilder):
372         (FragmentedFrameBuilder.to):
373         (FragmentedFrameBuilder.__init__):
374         (FragmentedFrameBuilder.build):
375         (_create_control_frame):
376         (create_ping_frame):
377         (create_pong_frame):
378         (create_close_frame):
379         (StreamOptions):
380         (StreamOptions.__init__):
381         (Stream):
382         (Stream.for):
383         (Stream.__init__):
384         (Stream._receive_frame):
385         (Stream._receive_frame_as_frame_object):
386         (Stream.send_message):
387         (Stream.receive_message):
388         (Stream._send_closing_handshake):
389         (Stream.close_connection):
390         (Stream.send_ping):
391         (Stream._send_pong):
392         (Stream._drain_received_data):
393         * Scripts/webkitpy/thirdparty/mod_pywebsocket/common.py: Added.
394         (is_control_opcode):
395         (ExtensionParameter):
396         (ExtensionParameter.__init__):
397         (ExtensionParameter.name):
398         (ExtensionParameter.add_parameter):
399         (ExtensionParameter.get_parameters):
400         (ExtensionParameter.get_parameter_names):
401         (ExtensionParameter.has_parameter):
402         (ExtensionParameter.get_parameter_value):
403         * Scripts/webkitpy/thirdparty/mod_pywebsocket/dispatch.py: Added.
404         (DispatchException):
405         (DispatchException.__init__):
406         (_default_passive_closing_handshake_handler):
407         (_normalize_path):
408         (_create_path_to_resource_converter):
409         (_create_path_to_resource_converter.converter):
410         (_enumerate_handler_file_paths):
411         (_HandlerSuite):
412         (_HandlerSuite.__init__):
413         (_source_handler_file):
414         (_extract_handler):
415         (Dispatcher):
416         (Dispatcher.maintains):
417         (Dispatcher.__init__):
418         (Dispatcher.add_resource_path_alias):
419         (Dispatcher.source_warnings):
420         (Dispatcher.do_extra_handshake):
421         (Dispatcher.transfer_data):
422         (Dispatcher.passive_closing_handshake):
423         (Dispatcher.get_handler_suite):
424         (Dispatcher._source_handler_files_in_dir):
425         * Scripts/webkitpy/thirdparty/mod_pywebsocket/extensions.py: Added.
426         (ExtensionProcessorInterface):
427         (ExtensionProcessorInterface.get_extension_response):
428         (ExtensionProcessorInterface.setup_stream_options):
429         (DeflateStreamExtensionProcessor):
430         (DeflateStreamExtensionProcessor.__init__):
431         (DeflateStreamExtensionProcessor.get_extension_response):
432         (DeflateStreamExtensionProcessor.setup_stream_options):
433         (DeflateFrameExtensionProcessor):
434         (DeflateFrameExtensionProcessor.__init__):
435         (DeflateFrameExtensionProcessor.get_extension_response):
436         (DeflateFrameExtensionProcessor.setup_stream_options):
437         (DeflateFrameExtensionProcessor.setup_stream_options._OutgoingFilter):
438         (DeflateFrameExtensionProcessor.setup_stream_options._OutgoingFilter.__init__):
439         (DeflateFrameExtensionProcessor.setup_stream_options._OutgoingFilter.filter):
440         (DeflateFrameExtensionProcessor.setup_stream_options._IncomingFilter):
441         (DeflateFrameExtensionProcessor.setup_stream_options._IncomingFilter.__init__):
442         (DeflateFrameExtensionProcessor.setup_stream_options._IncomingFilter.filter):
443         (DeflateFrameExtensionProcessor.set_response_window_bits):
444         (DeflateFrameExtensionProcessor.set_response_no_context_takeover):
445         (DeflateFrameExtensionProcessor.enable_outgoing_compression):
446         (DeflateFrameExtensionProcessor.disable_outgoing_compression):
447         (DeflateFrameExtensionProcessor._outgoing_filter):
448         (DeflateFrameExtensionProcessor._incoming_filter):
449         (get_extension_processor):
450         (get_extension_processor.is):
451         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/__init__.py: Added.
452         (try):
453         (do_handshake):
454         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/_base.py: Added.
455         (AbortedByUserException):
456         (HandshakeException):
457         (HandshakeException.__init__):
458         (VersionException):
459         (VersionException.__init__):
460         (get_default_port):
461         (validate_subprotocol):
462         (parse_host_header):
463         (format_header):
464         (build_location):
465         (get_mandatory_header):
466         (validate_mandatory_header):
467         (check_request_line):
468         (check_header_lines):
469         (parse_token_list):
470         (_parse_extension_param):
471         (_parse_extension):
472         (parse_extensions):
473         (format_extensions):
474         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/draft75.py: Added.
475         (Handshaker):
476         (Handshaker.performs):
477         (Handshaker.__init__):
478         (Handshaker.do_handshake):
479         (Handshaker._set_resource):
480         (Handshaker._set_origin):
481         (Handshaker._set_location):
482         (Handshaker._set_subprotocol):
483         (Handshaker._set_protocol_version):
484         (Handshaker._sendall):
485         (Handshaker._send_handshake):
486         (Handshaker._check_header_lines):
487         (Handshaker._check_first_lines):
488         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/hybi.py: Added.
489         (compute_accept):
490         (Handshaker):
491         (Handshaker.__init__):
492         (Handshaker._validate_connection_header):
493         (Handshaker.do_handshake):
494         (Handshaker._get_origin):
495         (Handshaker._check_version):
496         (Handshaker._set_protocol):
497         (Handshaker._parse_extensions):
498         (Handshaker._validate_key):
499         (Handshaker._get_key):
500         (Handshaker._send_handshake):
501         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/hybi00.py: Added.
502         (Handshaker):
503         (Handshaker.__init__):
504         (Handshaker.do_handshake):
505         (Handshaker._set_resource):
506         (Handshaker._set_subprotocol):
507         (Handshaker._set_location):
508         (Handshaker._set_origin):
509         (Handshaker._set_protocol_version):
510         (Handshaker._set_challenge_response):
511         (Handshaker._get_key_value):
512         (Handshaker._get_challenge):
513         (Handshaker._send_handshake):
514         * Scripts/webkitpy/thirdparty/mod_pywebsocket/headerparserhandler.py: Added.
515         (ApacheLogHandler):
516         (ApacheLogHandler.__init__):
517         (ApacheLogHandler.emit):
518         (_configure_logging):
519         (_parse_option):
520         (_create_dispatcher):
521         (headerparserhandler):
522         * Scripts/webkitpy/thirdparty/mod_pywebsocket/http_header_util.py: Added.
523         (_is_char):
524         (_is_ctl):
525         (ParsingState):
526         (ParsingState.__init__):
527         (peek):
528         (consume):
529         (consume_string):
530         (consume_lws):
531         (consume_lwses):
532         (consume_token):
533         (consume_token_or_quoted_string):
534         (quote_if_necessary):
535         (parse_uri):
536         * Scripts/webkitpy/thirdparty/mod_pywebsocket/memorizingfile.py: Added.
537         (MemorizingFile):
538         (MemorizingFile.__init__):
539         (MemorizingFile.__getattribute__):
540         (MemorizingFile.readline):
541         (MemorizingFile.get_memorized_lines):
542         * Scripts/webkitpy/thirdparty/mod_pywebsocket/msgutil.py: Added.
543         (close_connection):
544         (send_message):
545         (receive_message):
546         (send_ping):
547         (MessageReceiver):
548         (MessageReceiver.receives):
549         (MessageReceiver.provides):
550         (MessageReceiver.should):
551         (MessageReceiver.__init__):
552         (MessageReceiver.run):
553         (MessageReceiver.receive):
554         (MessageReceiver.receive_nowait):
555         (MessageReceiver.stop):
556         (MessageSender):
557         (MessageSender.sends):
558         (MessageSender.provides):
559         (MessageSender.should):
560         (MessageSender.__init__):
561         (MessageSender.run):
562         (MessageSender.send):
563         (MessageSender.send_nowait):
564         * Scripts/webkitpy/thirdparty/mod_pywebsocket/standalone.py: Added.
565         (_StandaloneConnection):
566         (_StandaloneConnection.__init__):
567         (_StandaloneConnection.get_local_addr):
568         (_StandaloneConnection.get_remote_addr):
569         (_StandaloneConnection.write):
570         (_StandaloneConnection.read):
571         (_StandaloneConnection.get_memorized_lines):
572         (_StandaloneRequest):
573         (_StandaloneRequest.__init__):
574         (_StandaloneRequest.get_uri):
575         (_StandaloneRequest.get_method):
576         (_StandaloneRequest.is_https):
577         (_StandaloneRequest._drain_received_data):
578         (_StandaloneSSLConnection):
579         (_StandaloneSSLConnection.for):
580         (_StandaloneSSLConnection.__init__):
581         (_StandaloneSSLConnection.__getattribute__):
582         (_StandaloneSSLConnection.__setattr__):
583         (_StandaloneSSLConnection.makefile):
584         (WebSocketServer):
585         (WebSocketServer.__init__):
586         (WebSocketServer._create_sockets):
587         (WebSocketServer.server_bind):
588         (WebSocketServer.server_activate):
589         (WebSocketServer.server_close):
590         (WebSocketServer.fileno):
591         (WebSocketServer.handle_error):
592         (WebSocketServer.get_request):
593         (WebSocketServer.serve_forever):
594         (WebSocketServer.shutdown):
595         (WebSocketRequestHandler):
596         (WebSocketRequestHandler.setup):
597         (WebSocketRequestHandler.setup.SocketServer):
598         (WebSocketRequestHandler.__init__):
599         (WebSocketRequestHandler.parse_request):
600         (WebSocketRequestHandler.log_request):
601         (WebSocketRequestHandler.log_error):
602         (WebSocketRequestHandler.is_cgi):
603         (_configure_logging):
604         (_alias_handlers):
605         (_build_option_parser):
606         (ThreadMonitor):
607         (ThreadMonitor.__init__):
608         (ThreadMonitor.run):
609         (_parse_args_and_config):
610         (_main):
611         (_main.if):
612         (_main.if.__check_script):
613         * Scripts/webkitpy/thirdparty/mod_pywebsocket/stream.py: Copied from Tools/Scripts/webkitpy/thirdparty/__init___unittest.py.
614         * Scripts/webkitpy/thirdparty/mod_pywebsocket/util.py: Added.
615         (get_stack_trace):
616         (prepend_message_to_exception):
617         (__translate_interp):
618         (get_script_interp):
619         (wrap_popen3_for_win):
620         (wrap_popen3_for_win.__wrap_popen3):
621         (hexify):
622         (get_class_logger):
623         (NoopMasker):
624         (NoopMasker.__init__):
625         (NoopMasker.mask):
626         (RepeatedXorMasker):
627         (RepeatedXorMasker.__init__):
628         (RepeatedXorMasker.mask):
629         (DeflateRequest):
630         (DeflateRequest.for):
631         (DeflateRequest.__init__):
632         (DeflateRequest.__getattribute__):
633         (DeflateRequest.__setattr__):
634         (_Deflater):
635         (_Deflater.__init__):
636         (_Deflater.compress_and_flush):
637         (_Inflater):
638         (_Inflater.__init__):
639         (_Inflater.decompress):
640         (_Inflater.append):
641         (_Inflater.reset):
642         (_RFC1979Deflater):
643         (_RFC1979Deflater.that):
644         (_RFC1979Deflater.__init__):
645         (_RFC1979Deflater.filter):
646         (_RFC1979Inflater):
647         (_RFC1979Inflater.for):
648         (_RFC1979Inflater.__init__):
649         (_RFC1979Inflater.filter):
650         (DeflateSocket):
651         (DeflateSocket.for):
652         (DeflateSocket.__init__):
653         (DeflateSocket.recv):
654         (DeflateSocket.sendall):
655         (DeflateSocket.send):
656         (DeflateConnection):
657         (DeflateConnection.for):
658         (DeflateConnection.__init__):
659         (DeflateConnection.get_remote_addr):
660         (DeflateConnection.put_bytes):
661         (DeflateConnection.read):
662         (DeflateConnection.write):
663         (_is_ewouldblock_errno):
664         (drain_received_data):
665
666 2012-05-10  Raphael Kubo da Costa  <rakuco@webkit.org>
667
668         [EFL][webkitpy] Define the 'wrapper' option in EflPort instead of creating a EflDriver.
669         https://bugs.webkit.org/show_bug.cgi?id=86117
670
671         Reviewed by Adam Barth.
672
673         Simplify the code used to wrap driver calls with jhbuild: we can
674         just define the 'wrapper' option instead of having to create a
675         Driver class just for that.
676
677         * Scripts/webkitpy/layout_tests/port/efl.py:
678         (EflPort.__init__):
679         (EflPort._port_flag_for_scripts):
680
681 2012-05-10  Sheriff Bot  <webkit.review.bot@gmail.com>
682
683         Unreviewed, rolling out r116633.
684         http://trac.webkit.org/changeset/116633
685         https://bugs.webkit.org/show_bug.cgi?id=86093
686
687         Breaks EFL tests (Requested by kov on #webkit).
688
689         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
690         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
691         * DumpRenderTree/efl/EventSender.cpp:
692         (keyDownCallback):
693         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
694         (LayoutTestController::setTabKeyCyclesThroughElements):
695
696 2012-05-10  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
697
698         [EFL] Memory leaks found in EFL DRT
699         https://bugs.webkit.org/show_bug.cgi?id=85889
700
701         Reviewed by Gustavo Noronha Silva.
702
703         Fix two memory leaks in EFL's DRT by using JSRetainPtr for JSString
704         created with JSStringCreateWithUTF8CString.
705
706         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
707         (DumpRenderTreeChrome::onWindowObjectCleared):
708         * DumpRenderTree/efl/EventSender.cpp:
709         (modifiersFromJSValue):
710
711 2012-05-10  Alexander Shalamov  <alexander.shalamov@intel.com>
712
713         [EFL] LayoutTestController needs implementation of setTabKeyCyclesThroughElements
714         https://bugs.webkit.org/show_bug.cgi?id=82864
715
716         Reviewed by Gustavo Noronha Silva.
717
718         Add missing implementation for setTabKeyCyclesThroughElements to EFL's LayoutTestController
719         in order to unskip fast/events/keypress-insert-tab.html
720
721         * DumpRenderTree/efl/EventSender.cpp:
722         (keyDownCallback):
723         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
724         (LayoutTestController::setTabKeyCyclesThroughElements):
725
726 2012-05-10  Alexander Shalamov  <alexander.shalamov@intel.com>
727
728         [EFL][DRT] EventSender - set delayed event information if it was not set before
729         https://bugs.webkit.org/show_bug.cgi?id=85972
730
731         Reviewed by Gustavo Noronha Silva.
732
733         Reset delayed event information if it is missing.
734
735         * DumpRenderTree/efl/EventSender.cpp:
736         (feedOrQueueMouseEvent):
737
738 2012-05-07  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
739
740         WebPageProxy::activeURL() should return the pending API request, even when there's no main frame
741         https://bugs.webkit.org/show_bug.cgi?id=85806
742
743         The m_pendingAPIRequestURL member is used (presumably) to mask over the async
744         nature of WebKit2, so that starting a load of a URL will reflect that URL
745         immedeatly from activeURL, even if the request has not been passed over to
746         the web process yet and reflected there.
747
748         This works well, except in the case of the initial request, where the main
749         frame creation happens on the web process side and is notified back to the
750         UI process. Until we've recived the notification we don't know about the main
751         frame, and this race condition will potentially give us an empty url instead
752         of the pending request.
753
754         To solve this we always return the pending API request if it's set, even
755         when there's no mainframe yet (that we known about).
756
757         Reviewed by Simon Hausmann.
758
759         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
760
761 2012-05-10  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
762
763         [EFL][DRT] Clear added user style sheets before a new testcase execution.
764         https://bugs.webkit.org/show_bug.cgi?id=84792
765
766         Reviewed Gustavo Noronha Silva.
767
768         The added user style sheets from the previous testcases are cleared in order to avoid 
769         flaky results.
770
771         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
772         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
773
774 2012-05-09  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
775
776         Move suspendAnimations to use Internals interface.
777         https://bugs.webkit.org/show_bug.cgi?id=85986
778
779         Reviewed by Ryosuke Niwa.
780
781         Remove suspendAnimations functions, because it is able to work in the 
782         cross-port way through the Internals interface
783
784         * DumpRenderTree/LayoutTestController.cpp:
785         (LayoutTestController::staticFunctions):
786         * DumpRenderTree/LayoutTestController.h:
787         (LayoutTestController):
788         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
789         * DumpRenderTree/chromium/LayoutTestController.cpp:
790         (LayoutTestController::LayoutTestController):
791         * DumpRenderTree/chromium/LayoutTestController.h:
792         (LayoutTestController):
793         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
794         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
795         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
796         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
797         * DumpRenderTree/qt/LayoutTestControllerQt.h:
798         (LayoutTestController):
799         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
800         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
801
802 2012-05-09  Kent Tamura  <tkent@chromium.org>
803
804         [Chromium] Attempt to fix an Android build error.
805         https://bugs.webkit.org/show_bug.cgi?id=86045
806
807         * DumpRenderTree/DumpRenderTree.gypi:
808
809 2012-05-09  Ojan Vafai  <ojan@chromium.org>
810
811         Don't show partytime if there are non-layout test failures.
812         https://bugs.webkit.org/show_bug.cgi?id=86056
813
814         Reviewed by Dimitri Glazkov.
815
816         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
817         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers_unittests.js:
818         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
819
820 2012-05-09  Hayato Ito  <hayato@chromium.org>
821
822         Improve a warning message for a reftest having an unused expectation file.
823         https://bugs.webkit.org/show_bug.cgi?id=85975
824
825         Reviewed by Adam Barth.
826
827         The current warning message is not intuitive. It might be better to update the message
828         so that it encourage users to remove unused expectation files.
829
830         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
831         (SingleTestRunner.__init__):
832
833 2012-05-09  Ojan Vafai  <ojan@chromium.org>
834
835         Change "losing test coverage" to cover all non-layout test problems in garden-o-matic
836         https://bugs.webkit.org/show_bug.cgi?id=86043
837
838         Reviewed by Adam Barth.
839
840         -Get rid of the gtest iframe.
841         -Cover all non-layout test problems. Before we wouldn't catch compile failures
842         or gtest failures since they were not on the webkit testing bots.
843         -Include the list of failing steps next to the builder name. This is a lot less
844         cluttered than you'd expect and already helped me find a compile failure early.
845
846         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
847         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
848         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
849         Delete unused map.
850         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
851         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
852         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
853         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js:
854         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
855         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
856         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
857         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
858
859 2012-05-09  Jessie Berlin  <jberlin@apple.com>
860
861         Crash using the new WKBundleDOMWindowExtensions APIs.
862         https://bugs.webkit.org/show_bug.cgi?id=85888
863
864         Reviewed by Brady Eidson.
865
866         Cached frames can live slightly longer than the page, but most clients unregister themselves
867         and do other cleanup in the willDestroyPage callback, making them miss the
868         willDestroyGlobalObjectForDOMWindowExtension callbacks.
869
870         The calls to willDestroyGlobalObjectForDOMWindowExtension in the DOMWindowExtensionBasic
871         test were all being invoked underneath WebPage::close. This is unrealistic. Update that test
872         to destroy the BundleDOMWindowExtensions in response to the willDestroyPage callback.
873
874         Add a test to verify that willDestroyGlobalObjectForDOMWindowExtension is being called for
875         pages that don't go into the page cache.
876
877         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
878         Add DOMWindowExtensionNoCache.cpp, DOMWindowExtensionNoCache_Bundle.cpp, simple-unload.html
879         and simple-iframe-unload.html
880
881         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic.cpp:
882         (TestWebKitAPI):
883         Remove the expected messages for willDestroyGlobalObjectForDOMWindowExtension.
884         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
885         Do not bother to keep track of the live extension count - all of them are expected to be
886         live until the test completes.
887         (TestWebKitAPI::TEST):
888         Fix the calls to EXPECT to pass the expected value first, and use EXPECT_WK_STREQ so that
889         message failures will be clearer.
890         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
891         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyPage):
892         Clean up the BundleDOMWindowExtensions.
893         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyGlobalObjectForDOMWindowExtension):
894         Add an ASSERT_NOT_REACHED.
895
896         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache.cpp: Added.
897         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
898         Keep track of the messages received so they can be checked at the end of the test.
899         (TestWebKitAPI::TEST):
900         Navigate to uncacheable pages and back.
901         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache_Bundle.cpp: Copied from Tools/TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp.
902         (DOMWindowExtensionNoCache):
903         (TestWebKitAPI::DOMWindowExtensionNoCache::DOMWindowExtensionNoCache):
904         Set up all the states for each BundleDOMWindowExtension.
905         (TestWebKitAPI::DOMWindowExtensionNoCache::frameLoadFinished):
906         Tell the UI Process about the states of the BundleDOMWindowExtensions.
907         (TestWebKitAPI::DOMWindowExtensionNoCache::sendExtensionStateMessage):
908         (TestWebKitAPI::DOMWindowExtensionNoCache::initialize):
909         (TestWebKitAPI::DOMWindowExtensionNoCache::didCreatePage):
910         (TestWebKitAPI::DOMWindowExtensionNoCache::willDestroyPage):
911         Remvoe the remaining BundleDOMWindowExtensions, send the updated state, and finish the test.
912         (TestWebKitAPI::DOMWindowExtensionNoCache::updateExtensionStateRecord):
913         (TestWebKitAPI::DOMWindowExtensionNoCache::sendBundleMessage):
914         (TestWebKitAPI::DOMWindowExtensionNoCache::globalObjectIsAvailableForFrame):
915         (TestWebKitAPI::DOMWindowExtensionNoCache::willDisconnectDOMWindowExtensionFromGlobalObject):
916         ASSERT that these pages not going into the page cache are not getting disconnected to go into
917         the page cache.
918         (TestWebKitAPI::DOMWindowExtensionNoCache::didReconnectDOMWindowExtensionToGlobalObject):
919         Ditto about getting reconnected when coming out of the page cache.
920         (TestWebKitAPI::DOMWindowExtensionNoCache::willDestroyGlobalObjectForDOMWindowExtension):
921         Tell the UI Process, update the state, and get rid of the BundleDOMWindowExtension.
922         (TestWebKitAPI::didFinishLoadForFrameCallback):
923         (TestWebKitAPI::globalObjectIsAvailableForFrameCallback):
924         (TestWebKitAPI::willDisconnectDOMWindowExtensionFromGlobalObjectCallback):
925         (TestWebKitAPI::didReconnectDOMWindowExtensionToGlobalObjectCallback):
926         (TestWebKitAPI::willDestroyGlobalObjectForDOMWindowExtensionCallback):
927
928         * TestWebKitAPI/Tests/WebKit2/simple-iframe-unload.html: Added.
929         * TestWebKitAPI/Tests/WebKit2/simple-unload.html: Added.
930
931 2012-05-09  Adam Barth  <abarth@webkit.org>
932
933         [Chromium] Android's DRT doesn't want to build WebUserMediaClientMock
934         https://bugs.webkit.org/show_bug.cgi?id=86045
935
936         Reviewed by Eric Seidel.
937
938         * DumpRenderTree/DumpRenderTree.gypi:
939
940 2012-05-09  Adam Barth  <abarth@webkit.org>
941
942         [Chromium] MockWebSpeechInputController.cpp should be guarded by ENABLE(INPUT_SPEECH)
943         https://bugs.webkit.org/show_bug.cgi?id=86041
944
945         Reviewed by Eric Seidel.
946
947         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp:
948
949 2012-05-09  Adam Barth  <abarth@webkit.org>
950
951         [Chromium] Android needs to swizzle pixel dumps in order to get the same hash as other platforms
952         https://bugs.webkit.org/show_bug.cgi?id=86040
953
954         Reviewed by Kent Tamura.
955
956         * DumpRenderTree/chromium/TestShell.cpp:
957         (TestShell::dumpImage):
958
959 2012-05-09  Raphael Kubo da Costa  <rakuco@webkit.org>
960
961         [EFL] Build fix after r116464.
962
963         `ulong' is a type defined by glibc, so it is not very
964         portable. Use `unsigned long' instead.
965
966         * DumpRenderTree/efl/EventSender.cpp:
967         (DelayedEvent::DelayedEvent):
968         (DelayedEvent):
969         (leapForwardCallback):
970
971 2012-05-09  Adam Barth  <abarth@webkit.org>
972
973         [Chromium] LayoutTestController should use ENABLE guards for some features
974         https://bugs.webkit.org/show_bug.cgi?id=86030
975
976         Reviewed by Eric Seidel.
977
978         These features are disabled on OS(ANDROID) and so shouldn't be compiled
979         when the feature is off.
980
981         * DumpRenderTree/chromium/LayoutTestController.cpp:
982         (LayoutTestController::LayoutTestController):
983         * DumpRenderTree/chromium/LayoutTestController.h:
984         (LayoutTestController):
985         * DumpRenderTree/chromium/TestShell.h:
986         (TestShell):
987
988 2012-05-09  Tom Zakrajsek  <tomz@codeaurora.org>
989
990         Fix overzealous re-opened since blocked by... message
991         https://bugs.webkit.org/show_bug.cgi?id=86020
992
993         Reviewed by Adam Barth.
994
995         Changed CreateBug behavior to only re-open a blocked bug if it
996         was RESOLVED.  Also had to update the test infrastructure slightly
997         to allow for distinct mock commit revisions, so one could point
998         to a RESOLVED bug for testing while another pointed to an OPEN bug.
999
1000         Also moved default created mock bug to be 60001 instead of 50004
1001         which already exists as a reference mock bug.
1002
1003         * Scripts/webkitpy/common/checkout/checkout_mock.py:
1004         (MockCheckout.__init__):
1005         (MockCheckout.commit_info_for_revision):
1006         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
1007         (MockBugzilla.create_bug):
1008         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
1009         * Scripts/webkitpy/tool/commands/download_unittest.py:
1010         (test_create_rollout_resolved):
1011         * Scripts/webkitpy/tool/commands/queries_unittest.py:
1012         (FailureReasonTest.test_blame_line_for_revision):
1013         * Scripts/webkitpy/tool/steps/createbug.py:
1014         (CreateBug.run):
1015
1016 2012-05-09  Hugo Parente Lima  <hugo.lima@openbossa.org>
1017
1018         Use suitable viewport values on XHTML-MP pages.
1019         https://bugs.webkit.org/show_bug.cgi?id=85425
1020
1021         Reviewed by Kenneth Rohde Christiansen.
1022
1023         Enable LEGACY_VIEWPORT_ADAPTION on Qt port.
1024
1025         * qmake/mkspecs/features/features.prf:
1026
1027 2012-05-09  Jochen Eisinger  <jochen@chromium.org>
1028
1029         When creating a new page during a navigation, prime the initial document with the correct referrer policy
1030         https://bugs.webkit.org/show_bug.cgi?id=86001
1031
1032         Reviewed by Adam Barth.
1033
1034         Store the referrer policy in the request, so the network stack can comply with it.
1035
1036         * DumpRenderTree/chromium/WebViewHost.cpp:
1037         (WebViewHost::willSendRequest):
1038
1039 2012-05-09  Vivek Galatage  <vivekgalatage@gmail.com>
1040
1041         Remove CYGWIN=tty from environment variable as its no longer supported
1042         https://bugs.webkit.org/show_bug.cgi?id=85791
1043
1044         The CYGWIN=tty environment variable is no longer supported from Cygwin
1045         version 1.7.10. This fix avoids addition of this environment variable
1046         for Cygwin versions greater than 1.7.9. For supporting older versions
1047         i.e. 1.7.9 or older, the script takes care of retaining the environment
1048         variable.
1049
1050         Reviewed by Adam Roben.
1051
1052         * Scripts/webkitdirs.pm:
1053         (setupAppleWinEnv):
1054
1055 2012-05-09  Dirk Pranke  <dpranke@chromium.org>
1056
1057         nrwt: --skip-failing-tests should also skip flaky tests
1058         https://bugs.webkit.org/show_bug.cgi?id=85941
1059
1060         Reviewed by Ojan Vafai.
1061
1062         This changes --skip-failing-tests so that we skip tests that
1063         are believed to pass sometimes as well as tests that are
1064         believed to always fail.
1065
1066         In addition, in testing this I discovered that we had lousy
1067         test coverage of the details of retrying test failures and
1068         --clobber-old-results, so I wrote some more tests.
1069
1070         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1071         (Manager.prepare_lists_and_print_output):
1072         * Scripts/webkitpy/layout_tests/port/test.py:
1073         (unit_test_list):
1074         (TestPort.__init__):
1075         (TestDriver.run_test):
1076         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1077         (MainTest.test_skip_failing_tests):
1078         (MainTest.test_retrying_and_flaky_tests):
1079
1080 2012-05-09  Christophe Dumez  <christophe.dumez@intel.com>
1081
1082         [EFL][DRT] EFL's DRT needs to support LayoutTestController.dumpIconChanges()
1083         https://bugs.webkit.org/show_bug.cgi?id=84435
1084
1085         Reviewed by Antonio Gomes.
1086
1087         Catch the "icon,changed" signal on the frames in EFL's DumpRenderTree
1088         to support LayoutTestController's dumpIconChanges().
1089
1090         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1091         (DumpRenderTreeChrome::createView):
1092         (DumpRenderTreeChrome::onFrameIconChanged):
1093         (DumpRenderTreeChrome::onFrameCreated):
1094         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
1095         (DumpRenderTreeChrome):
1096
1097 2012-05-09  Dominik Röttsches  <dominik.rottsches@intel.com>
1098
1099         webkit-patch upload Fails Behind Load-balancing Proxies
1100         https://bugs.webkit.org/show_bug.cgi?id=85980
1101
1102         Reviewed by Adam Barth.
1103
1104         Unchecking the login restriction to IP to allow the tool
1105         to function behind load-balancing proxies that use multiple
1106         outgoing IPs.
1107
1108         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
1109         (Bugzilla.authenticate):
1110
1111 2012-05-09  Hao Zheng  <zhenghao@chromium.org>
1112
1113         [Chromium-android] Add unit test for Android layout test script (chromium_android.py)
1114         https://bugs.webkit.org/show_bug.cgi?id=80852
1115
1116         Reviewed by Adam Barth.
1117
1118         Add some tests to ensure that chromium android port works to some
1119         extent. Some methods is hard to test, as they rely on the actual
1120         behavior on device, e.g. pushing files to device.
1121
1122         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1123         (ChromiumAndroidPort.get_last_stacktrace): Check if dir exists.
1124         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1125         (ChromiumAndroidPortTest):
1126         (ChromiumAndroidPortTest.test_attributes):
1127         (ChromiumAndroidPortTest.test_get_last_stacktrace):
1128         (ChromiumAndroidPortTest.test_get_last_stacktrace.mock_run_command_no_dir):
1129         (ChromiumAndroidPortTest.test_get_last_stacktrace.mock_run_command_no_file):
1130         (ChromiumAndroidPortTest.test_get_last_stacktrace.mock_run_command_non_empty):
1131         (ChromiumAndroidDriverTest):
1132         (ChromiumAndroidDriverTest.setUp):
1133         (ChromiumAndroidDriverTest.test_get_drt_return_value):
1134         (ChromiumAndroidDriverTest.test_has_crash_hint):
1135         (ChromiumAndroidDriverTest.test_read_prompt):
1136         (ChromiumAndroidDriverTest.test_test_shell_command):
1137         (ChromiumAndroidDriverTest.test_write_command_and_read_line):
1138
1139 2012-05-09  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
1140
1141         [Qt] ORWT fails with Qt 5 and WK2.
1142         https://bugs.webkit.org/show_bug.cgi?id=85968
1143
1144         Reviewed by Csaba Osztrogonác.
1145
1146         The Qt5 binary does not have a -style argument, so avoid using it.
1147
1148         * Scripts/old-run-webkit-tests:
1149
1150 2012-05-09  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1151
1152         [EFL][DRT] LayoutTestController addUserScript implementation
1153         https://bugs.webkit.org/show_bug.cgi?id=84718
1154
1155         Reviewed by Antonio Gomes.
1156
1157         Added implementation of addUserScript() to LayoutTestController.
1158
1159         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1160         (DumpRenderTreeChrome::resetDefaultsToConsistentValues): clear added user scripts
1161         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1162         (LayoutTestController::addUserScript): added implementation
1163
1164 2012-05-09  Dan Bernstein  <mitz@apple.com>
1165
1166         Made build-webkit copy WebKitLibraries to the right location again after r116257
1167
1168         Reviewed by Mark Rowe.
1169
1170         * Scripts/build-webkit: Pass the built product directory to copy-webkitlibraries-to-product-directory.
1171         * Scripts/copy-webkitlibraries-to-product-directory: Use the first argument, if present, as the
1172         product directory.
1173
1174 2012-05-08  Ojan Vafai  <ojan@chromium.org>
1175
1176         Show the chromium non-layout test bots in garden-o-matic
1177         https://bugs.webkit.org/show_bug.cgi?id=85903
1178
1179         Reviewed by Eric Seidel.
1180
1181         This is just a single-line high-level overview of the chromium non-layout test bots.
1182         This means that Chromium WebKit gardeners only need to keep garden-o-matic open in
1183         order to do their job.
1184
1185         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
1186         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
1187         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
1188         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
1189         (#chromium-gtests iframe):
1190         (#chromium-gtests):
1191         (#gtest-label):
1192
1193 2012-05-08  Kangil Han  <kangil.han@samsung.com>
1194
1195         [EFL][DRT] Implement TextInputController.
1196         https://bugs.webkit.org/show_bug.cgi?id=78559
1197
1198         Reviewed by Martin Robinson.
1199
1200         TextInputController has been created for an input emulation purpose in DRT.
1201         However, not like other ports, EFL doesn't have it.
1202         Therefore, this patch will enable EFL port to have an input emulation in DRT.
1203
1204         * DumpRenderTree/efl/CMakeLists.txt:
1205         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1206         (DumpRenderTreeChrome::onWindowObjectCleared):
1207         * DumpRenderTree/efl/TextInputController.cpp: Added.
1208         (setMarkedTextCallback):
1209         (hasMarkedTextCallback):
1210         (markedRangeCallback):
1211         (insertTextCallback):
1212         (unmarkTextCallback):
1213         (firstRectForCharacterRangeCallback):
1214         (selectedRangeCallback):
1215         (makeTextInputController):
1216         * DumpRenderTree/efl/TextInputController.h: Added.
1217
1218 2012-04-23  Raphael Kubo da Costa  <rakuco@webkit.org>
1219
1220         [EFL] Pass a WTF::String directly from DRT to ewebkit when possible
1221         https://bugs.webkit.org/show_bug.cgi?id=84634
1222
1223         Reviewed by Gustavo Noronha Silva.
1224
1225         Drop unnecessary const char* -> WTF::String conversions by
1226         creating String objects directly in LayoutTestController out of
1227         JSStringRef's and passing them to DumpRenderTreeSupportEfl.
1228
1229         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1230         (LayoutTestController::queueLoad):
1231         (LayoutTestController::setValueForUser):
1232         (LayoutTestController::findString):
1233         (LayoutTestController::addUserStyleSheet):
1234
1235 2012-05-08  Alexander Shalamov  <alexander.shalamov@intel.com>
1236
1237         [EFL][DRT] EventSender needs implementation of leapForward function
1238         https://bugs.webkit.org/show_bug.cgi?id=83486
1239
1240         Reviewed by Eric Seidel.
1241
1242         Implemented leapForward function in Efl's DRT. This function is required
1243         to delay and queue event execution and simulate real user behavior.
1244
1245         * DumpRenderTree/efl/EventSender.cpp:
1246         (DelayedEvent):
1247         (MouseEventInfo::MouseEventInfo):
1248         (MouseEventInfo):
1249         (sendClick):
1250         (mouseDownCallback):
1251         (mouseUpCallback):
1252         (mouseMoveToCallback):
1253         (leapForwardCallback):
1254         (mouseScrollByCallback):
1255         (feedOrQueueMouseEvent):
1256         (feedMouseEvent):
1257         (feedQueuedMouseEvents):
1258
1259 2012-05-08  Ryuan Choi  <ryuan.choi@samsung.com>
1260
1261         [EFL][DRT] WebKit/DRT should not paint webview for the pixel tests.
1262         https://bugs.webkit.org/show_bug.cgi?id=79853
1263
1264         Reviewed by Eric Seidel.
1265
1266         Screen buffer can be different from rendered results.
1267         For example, WebKit/Efl just moves tiles for ewk_view_tiled when scrolling
1268         contents.
1269         It means that pixel tests can't check regression of ewk_view_tiled.
1270
1271         This patch changes a way to provide bitmap from painting webview to capturing
1272         evas.
1273
1274         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
1275         (createBitmapContextFromWebView):
1276
1277 2012-05-08  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1278
1279         [GTK] Memory leak in DRT EventSender.cpp
1280         https://bugs.webkit.org/show_bug.cgi?id=85886
1281
1282         Reviewed by Martin Robinson.
1283
1284         Fix a memory leak in gdkModifersFromJSValue() by using JSRetainPtr for
1285         JSString created with JSStringCreateWithUTF8CString.
1286
1287         * DumpRenderTree/gtk/EventSender.cpp:
1288         (gdkModifersFromJSValue):
1289
1290 2012-05-08  Christophe Dumez  <christophe.dumez@intel.com>
1291
1292         [EFL] glib dependency needs to be updated in jhbuild
1293         https://bugs.webkit.org/show_bug.cgi?id=85878
1294
1295         Reviewed by Martin Robinson.
1296
1297         Bump glib version from 2.31.8 to 2.32.2 in jhbuild.modules. This
1298         addresses flakiness in media tests due to possible deadlock with the
1299         following output on stderr:
1300         GStreamer-WARNING **: wrong STREAM_LOCK count 0
1301
1302         The bug was fixed upstream by:
1303         http://git.gnome.org/browse/glib/commit/?id=265f265c677bc5399dbb1cfa449cb4b915a6459f
1304
1305         * efl/jhbuild.modules:
1306
1307 2012-05-07  Alexis Menard  <alexis.menard@openbossa.org>
1308
1309         [Qt] Unbreak debugging of WebKit2.
1310         https://bugs.webkit.org/show_bug.cgi?id=85839
1311
1312         Reviewed by Simon Hausmann.
1313
1314         When you attach GDB to a running process, it stops it.
1315         http://trac.webkit.org/changeset/115958 introduced a pause() 
1316         call to wait the debugger to be attached to then continue 
1317         the execution of the WebProcess. Unfortunately the pause()
1318         function does not return unless a signal handler is called.
1319         This patch introduce an event handler to exit from the paused
1320         state when the debugger send the signal SIGCONT. The old code
1321         works with older version of GDB (<7.0) but not with newer
1322         versions where the behavior of pause() is correct.
1323
1324         * WebKitTestRunner/qt/main.cpp:
1325         (sigcontHandler):
1326         (main):
1327
1328 2012-05-08  Christophe Dumez  <christophe.dumez@intel.com>
1329
1330         [EFL][DRT] EFL's DRT does not fully support page visibility
1331         https://bugs.webkit.org/show_bug.cgi?id=85032
1332
1333         Reviewed by Antonio Gomes.
1334
1335         * DumpRenderTree/LayoutTestController.cpp:
1336         (resetPageVisibilityCallback): Fix incorrect argumentCount check.
1337         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1338         (DumpRenderTreeChrome::resetDefaultsToConsistentValues): Properly
1339         reset page visibility setting between tests.
1340         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1341         (LayoutTestController::setPageVisibility): Add support for "preview"
1342         visibility state.
1343
1344 2012-05-08  Gavin Peters  <gavinp@chromium.org>
1345
1346         [webkit-patch] Return meaningful results from single-result queries.
1347         https://bugs.webkit.org/show_bug.cgi?id=85367
1348
1349         Reviewed by Dirk Pranke.
1350
1351         Sometimes a quicksearch returns a single result by just redirecting to the page for that single
1352         bug. This patch hacks around that by noticing the redirect target is a single page, and constructing
1353         a result list consisting of that one bug.
1354
1355         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
1356         (BugzillaQueries._fetch_bugs_from_advanced_query):
1357
1358 2012-05-08  Gavin Peters  <gavinp@chromium.org>
1359
1360         [webkit-patch] Add new chrome-channels command to track down current chrome release channels for a committed bug.
1361         https://bugs.webkit.org/show_bug.cgi?id=85368
1362
1363         Reviewed by Eric Seidel.
1364
1365         * Scripts/webkitpy/common/config/urls.py:
1366         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
1367         (MockBugzillaQueries.fetch_bugs_matching_quicksearch):
1368         * Scripts/webkitpy/common/net/omahaproxy.py: Added.
1369         (OmahaProxy):
1370         (OmahaProxy.__init__):
1371         (OmahaProxy.set_url):
1372         (OmahaProxy._json_url):
1373         (OmahaProxy._get_json):
1374         (OmahaProxy.get_revisions):
1375         * Scripts/webkitpy/common/net/omahaproxy_unittest.py: Added.
1376         (MockOmahaProxy):
1377         (MockOmahaProxy.__init__):
1378         (OmahaProxyTest):
1379         (OmahaProxyTest.test_get_revisions):
1380         * Scripts/webkitpy/tool/commands/__init__.py:
1381         * Scripts/webkitpy/tool/commands/bugsearch.py:
1382         (BugSearch):
1383         * Scripts/webkitpy/tool/commands/chromechannels.py: Added.
1384         (ChromeChannels):
1385         (__init__):
1386         (execute):
1387         * Scripts/webkitpy/tool/commands/chromechannels_unittest.py: Added.
1388         (MockOmahaProxy):
1389         (MockOmahaProxy.get_revisions):
1390         (TestableChromeChannels):
1391         (TestableChromeChannels.__init__):
1392         (ChromeChannelsTest):
1393         (ChromeChannelsTest.test_single_bug):
1394         (ChromeChannelsTest.test_with_query):
1395
1396 2012-05-08  Carlos Garcia Campos  <cgarcia@igalia.com>
1397
1398         Unreviewed. Fix GTK+ unit tests.
1399
1400         * gtk/common.py:
1401         (pkg_config_file_variable):
1402
1403 2012-05-08  Carlos Garcia Campos  <cgarcia@igalia.com>
1404
1405         [GTK] Google tests that time out are leaked
1406         https://bugs.webkit.org/show_bug.cgi?id=84973
1407
1408         Reviewed by Philippe Normand.
1409
1410         Move Tools/Scripts/run-gtk-tests to Tools/gtk/run-api-tests
1411         leaving run-gtk-tests as simple wrapper that runs run-api-tests in
1412         a jhbuild environment when jhbuild is available.
1413
1414         * Scripts/run-gtk-tests: Run Tools/gtk/run-api-tests directly or
1415         with jhbuild if present.
1416         * gtk/common.py:
1417         (pkg_config_file_variable): Generic function to get a variable
1418         from a pkg-config file.
1419         (prefix_of_pkg_config_file): Use pkg_config_file_variable with 'prefix'.
1420         * gtk/run-api-tests: Copied from Tools/Scripts/run-gtk-tests.
1421         (TestRunner.__init__): Use common.build_path() to get the programs path.
1422         (TestRunner._lookup_atspi2_binary): Use
1423         common.pkg_config_file_variable() to get the 'exec_prefix'.
1424         (TestRunner._start_accessibility_daemons): Use subprocess.Popen()
1425         directly instead of helper _create_process that has been removed.
1426         (TestRunner._setup_testing_environment): Use subprocess.Popen()
1427         directly instead of helper _create_process that has been
1428         removed. Use common.top_level_path() and common.build_path()
1429         instead of custom methods that have been removed too.
1430         (TestRunner._run_test_command): Use subprocess.Popen() directly
1431         instead of helper _create_process that has been removed.
1432
1433 2012-05-08  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1434
1435         [EFL] [DRT] Clear extra views before setting a blank page
1436         https://bugs.webkit.org/show_bug.cgi?id=85608
1437
1438         Reviewed by Kenneth Rohde Christiansen.
1439
1440         We need to clear the extra views before setting a blank page,
1441         otherwise it will cause more events interactions and sometimes
1442         trigger a unload(). This was causing frame-crash-with-page-cache.html
1443         to leak events to the next test causing any subsequent test to
1444         fail.
1445
1446         * DumpRenderTree/efl/DumpRenderTree.cpp:
1447         (runTest):
1448         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1449         (DumpRenderTreeChrome::clearExtraViews):
1450         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1451         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
1452         (DumpRenderTreeChrome):
1453
1454 2012-05-08  Balazs Kelemen  <kbalazs@webkit.org>
1455
1456         [Qt] X11 plugins need to be reworked for Qt5+WK1
1457         https://bugs.webkit.org/show_bug.cgi?id=80691
1458
1459         Reviewed by Simon Hausmann.
1460
1461         Implement basic windowless plugin support with Qt5.
1462
1463         * qmake/mkspecs/features/features.prf:
1464         Enable NPAPI plugins if X11 libraries are available.
1465         * qmake/mkspecs/features/functions.prf:
1466         Added a convenience function to determine availability
1467         of X11 libraries.
1468
1469 2012-05-07  Raphael Kubo da Costa  <rakuco@webkit.org>
1470
1471         [webkitpy] Remove Python 2.5-specific workaround from http_server_base.py.
1472         https://bugs.webkit.org/show_bug.cgi?id=85852
1473
1474         Reviewed by Dirk Pranke.
1475
1476         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
1477         (HttpServerBase._is_server_running_on_all_ports): Revert the part
1478         of r89412 which was a workaround for an issue with Python 2.5 on
1479         Mac 10.5.
1480
1481 2012-05-07  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
1482
1483         webkitdirs.pm: Make determineNumberOfCPUs work on FreeBSD.
1484         https://bugs.webkit.org/show_bug.cgi?id=81481
1485
1486         Reviewed by Dirk Pranke.
1487
1488         FreeBSD has the same sysctl interface present in Darwin to query
1489         the number of CPUs available, so use that too.
1490
1491         * Scripts/webkitdirs.pm:
1492         (determineNumberOfCPUs): Share the sysctl call with Darwin.
1493         (isFreeBSD): Added.
1494
1495 2012-05-07  Alexis Menard  <alexis.menard@openbossa.org>
1496
1497         Unreviewed Qt build fix in Debug.
1498
1499         We need some system includes here for getpid() to be recognized.
1500
1501         * WebKitTestRunner/qt/main.cpp:
1502
1503 2012-05-07  Ojan Vafai  <ojan@chromium.org>
1504
1505         When embedding the flakiness dashboard, hide popups when the frame is blurred.
1506         https://bugs.webkit.org/show_bug.cgi?id=85684
1507
1508         Reviewed by Dirk Pranke.
1509
1510         If you show a popup, then click elsewhere in garden-o-matic, it will now
1511         properly hide it.
1512
1513         * TestResultServer/static-dashboards/dashboard_base.js:
1514         (showPopup):
1515         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1516         (showPopupForBuild):
1517         (generatePageForIndividualTests.if):
1518         (generatePageForIndividualTests):
1519         * TestResultServer/static-dashboards/flakiness_dashboard_embedded_unittests.js: Added.
1520         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1521         * TestResultServer/static-dashboards/run-embedded-unittests.html: Added.
1522
1523 2012-05-07  Ian Vollick  <vollick@chromium.org>
1524
1525         [chromium] WebViewHost should setCompositorSurfaceReady
1526         https://bugs.webkit.org/show_bug.cgi?id=85418
1527
1528         Reviewed by James Robinson.
1529
1530         * DumpRenderTree/chromium/WebViewHost.cpp:
1531         (WebViewHost::setWebWidget):
1532
1533 2012-05-07  Jon Lee  <jonlee@apple.com>
1534
1535         WebKitTestRunner needs layoutTestController.dumpFrameLoadCallbacks
1536         https://bugs.webkit.org/show_bug.cgi?id=42331
1537         <rdar://problem/8193641>
1538
1539         Reviewed by Darin Adler.
1540
1541         This patch sets up the framework for supporting dumpFrameLoadCallbacks, as well as
1542         outputting some of the required strings. As tests become unskipped the other callbacks
1543         can be supported.
1544
1545         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Add dumpFrameLoadCallbacks()
1546         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1547         (WTR::InjectedBundle::didReceiveMessage): pass in the dictionary containing keys needed to initialize
1548         dump callbacks for the layoutTestController.
1549         (WTR::InjectedBundle::booleanForKey): Helper function to extract boolean value from key in WKDictionaryRef.
1550         Output some warning if we are trying to extract a value that is not a boolean.
1551         (WTR::InjectedBundle::beginTesting): Update to take in the dictionary passed in by the TestController.
1552         We cannot set the bits on the layoutTestController until it has been created in beginTesting().
1553         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
1554         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1555         (WTR::InjectedBundle::didReceiveMessage):
1556         (WTR):
1557         (WTR::InjectedBundle::beginTesting):
1558         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
1559         (InjectedBundle):
1560         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1561         (WTR):
1562         (WTR::dumpFrameDescriptionSuitableForTestResult): Static function similar to
1563         -[WebFrame _drt_descriptionSuitableForTestResult] in DumpRenderTree.
1564         (WTR::InjectedBundlePage::didStartProvisionalLoadForFrame): Output string.
1565         (WTR::InjectedBundlePage::didCommitLoadForFrame): Output string.
1566         (WTR::InjectedBundlePage::didFinishLoadForFrame): Output string.
1567         (WTR::InjectedBundlePage::didReceiveTitleForFrame): Output string.
1568         (WTR::InjectedBundlePage::didCancelClientRedirectForFrame): Output string.
1569         (WTR::InjectedBundlePage::willPerformClientRedirectForFrame): Output string. In WK1 it's
1570         "willPerformClientRedirectToURL" so we will have to use that value here.
1571         (WTR::InjectedBundlePage::didFinishDocumentLoadForFrame): Output string.
1572         (WTR::InjectedBundlePage::didHandleOnloadEventsForFrame): Output string.
1573         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1574         (WTR::LayoutTestController::LayoutTestController): Initialize to not dump frame load callbacks.
1575         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1576         (WTR::LayoutTestController::dumpFrameLoadCallbacks): Change to use setShouldDumpFrameLoadCallbacks.
1577         (WTR::LayoutTestController::setShouldDumpFrameLoadCallbacks): Set bit.
1578         (WTR::LayoutTestController::shouldDumpFrameLoadCallbacks): Return bit.
1579         (LayoutTestController): Added m_dumpFrameLoadCallbacks to track whether to dump the output.
1580         * WebKitTestRunner/TestInvocation.cpp:
1581         (WTR::shouldLogFrameLoadDelegates): Similar to DRT.
1582         (WTR::TestInvocation::invoke): Based on the URL loaded, set the bit in the dictionary that gets passed
1583         to the injected bundle.
1584
1585 2012-05-07  Tommy Widenflycht  <tommyw@google.com>
1586
1587         [chromium] MediaStream API: Enhance WebUserMediaClientMock and add a test for it
1588         https://bugs.webkit.org/show_bug.cgi?id=85618
1589
1590         Reviewed by Adam Barth.
1591
1592         Adding the capability to control which MediaStreamTracks actually get created,
1593         and adding a test that exercises the JS api using that capability.
1594
1595         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp:
1596         (WebKit::WebUserMediaClientMock::requestUserMedia):
1597         * DumpRenderTree/chromium/WebUserMediaClientMock.h:
1598         * DumpRenderTree/chromium/WebViewHost.cpp:
1599         (WebViewHost::testMediaStreamClient):
1600         * DumpRenderTree/chromium/WebViewHost.h:
1601         (WebViewHost):
1602
1603 2012-05-07  Zan Dobersek  <zandobersek@gmail.com>
1604
1605         [Gtk] WebGL feature is not built anymore through build-webkit after r116251
1606         https://bugs.webkit.org/show_bug.cgi?id=85788
1607
1608         Reviewed by Martin Robinson.
1609
1610         Enable WebGL by default for the Gtk port. The feature is already enabled by
1611         default in configure.ac, so there's no reason it shouldn't be enabled when
1612         building through build-webkit.
1613
1614         * Scripts/webkitperl/FeatureList.pm:
1615
1616 2012-05-07  Csaba Osztrogonác  <ossy@webkit.org>
1617
1618         Unittest fix after r116288, because now there are more than 5 hit for the 'Eric' word.
1619         We have only 2 'Gavin', so it is a better choice than 'Eric' for unit testing.
1620
1621         Reviewed by Tor Arne Vestbø.
1622
1623         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1624         (IRCCommandTest.test_whois):
1625
1626 2012-05-07  Sheriff Bot  <webkit.review.bot@gmail.com>
1627
1628         Unreviewed, rolling out r116299, r116301, and r116303.
1629         http://trac.webkit.org/changeset/116299
1630         http://trac.webkit.org/changeset/116301
1631         http://trac.webkit.org/changeset/116303
1632         https://bugs.webkit.org/show_bug.cgi?id=85795
1633
1634         Build is still broken (Requested by Ossy on #webkit).
1635
1636         * DumpRenderTree/qt/DumpRenderTree.pro:
1637         * DumpRenderTree/qt/main.cpp:
1638         (main):
1639         * QtTestPlatformPlugin/QtTestPlatformPlugin.pro: Removed.
1640         * QtTestPlatformPlugin/TestIntegration.cpp: Removed.
1641         * QtTestPlatformPlugin/TestIntegration.h: Removed.
1642         * QtTestPlatformPlugin/mac/TestFontDatabase.h: Removed.
1643         * QtTestPlatformPlugin/mac/TestFontDatabase.mm: Removed.
1644         * QtTestPlatformPlugin/mac/TestIntegrationMac.mm: Removed.
1645         * QtTestPlatformPlugin/main.cpp: Removed.
1646         * QtTestPlatformPlugin/testplatform.json: Removed.
1647         * Tools.pro:
1648         * WebKitTestRunner/Target.pri:
1649         * WebKitTestRunner/qt/main.cpp:
1650         (main):
1651
1652 2012-05-07  Balazs Kelemen  <kbalazs@webkit.org>
1653
1654         Unreviewed Qt build fix.
1655
1656         One more try to learn how to use
1657         const with a pointer array.
1658
1659         * DumpRenderTree/qt/main.cpp:
1660         (initializeTestPlatformPlugin):
1661
1662 2012-05-07  Balazs Kelemen  <kbalazs@webkit.org>
1663
1664         Unreviewed build fix after r116299.
1665
1666         * DumpRenderTree/qt/main.cpp:
1667         (initializeTestPlatformPlugin):
1668
1669 2012-05-07  Balazs Kelemen  <kbalazs@webkit.org>
1670
1671         [Qt] Add test specific platform plugin to achieve unified layout test results
1672         https://bugs.webkit.org/show_bug.cgi?id=80996
1673
1674         Reviewed by Simon Hausmann.
1675
1676         Added QtTestPlatformPlugin as a new project under Tools.
1677         This is a Qt5-ish platform plugin that can be used to tweak the
1678         platform support interfaces in order to unify layout test results.
1679         For now it only overrides the font database on Mac and redirects
1680         everything else to the real platform plugin. The font database it
1681         provides mimics the way how we set up test fonts with fontconfig on Linux.
1682         Make DumpRenderTree and WebKitTestRunner use this platform plugin.
1683
1684         * DumpRenderTree/qt/DumpRenderTree.pro:
1685         * DumpRenderTree/qt/main.cpp:
1686         (initializeTestPlatformPlugin):
1687         (main):
1688         * QtTestPlatformPlugin/QtTestPlatformPlugin.pro: Added.
1689         * QtTestPlatformPlugin/TestIntegration.cpp: Added.
1690         (TestIntegration::TestIntegration):
1691         (TestIntegration::fontDatabase):
1692         * QtTestPlatformPlugin/TestIntegration.h: Added.
1693         (TestIntegration):
1694         (TestIntegration::hasCapability):
1695         (TestIntegration::createPlatformPixmap):
1696         (TestIntegration::createPlatformWindow):
1697         (TestIntegration::createPlatformBackingStore):
1698         (TestIntegration::createPlatformOpenGLContext):
1699         (TestIntegration::createPlatformSharedGraphicsCache):
1700         (TestIntegration::guiThreadEventDispatcher):
1701         (TestIntegration::clipboard):
1702         (TestIntegration::drag):
1703         (TestIntegration::inputContext):
1704         (TestIntegration::accessibility):
1705         (TestIntegration::nativeInterface):
1706         (TestIntegration::services):
1707         (TestIntegration::styleHint):
1708         (TestIntegration::platformTheme):
1709         * QtTestPlatformPlugin/mac/TestFontDatabase.h: Added.
1710         (TestFontDatabase):
1711         * QtTestPlatformPlugin/mac/TestFontDatabase.mm: Added.
1712         (TestFontDatabase::populateFontDatabase):
1713         * QtTestPlatformPlugin/mac/TestIntegrationMac.mm: Added.
1714         (TestIntegration::fontDatabase):
1715         * QtTestPlatformPlugin/main.cpp: Added.
1716         (TestIntegrationPlugin::keys):
1717         (TestIntegrationPlugin::create):
1718         (TestIntegrationPlugin::initialize):
1719         * QtTestPlatformPlugin/testplatform.json: Added.
1720         * Tools.pro:
1721         * WebKitTestRunner/Target.pri:
1722         * WebKitTestRunner/qt/main.cpp:
1723         (main):
1724
1725 2012-05-07  Christophe Dumez  <christophe.dumez@intel.com>
1726
1727         [EFL] media/video-poster-blocked-by-willsendrequest.html is flaky
1728         https://bugs.webkit.org/show_bug.cgi?id=85609
1729
1730         Reviewed by Simon Hausmann.
1731
1732         Flakiness was caused by the fact that
1733         media/crash-closing-page-with-media-as-plugin-fallback.html and
1734         media/video-poster-blocked-by-willsendrequest.html use the same PNG
1735         file as poster attribute in the video tag. The PNG file gets cached
1736         when the first test is executed and this somehow causes the second
1737         test to fail. 
1738
1739         The flakiness was addressed by clearing the memory and page caches
1740         between tests.
1741
1742         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1743         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1744
1745 2012-05-07  Eric Seidel  <eric@webkit.org>
1746
1747         Add values for all features to Qt's features.pri
1748         https://bugs.webkit.org/show_bug.cgi?id=85746
1749
1750         Reviewed by Tor Arne Vestbø.
1751
1752         When generate-feature-files generates features.pri it provides a value
1753         for every possible ENABLE_.  In prepration for landing generate-feature-files
1754         it seems appropriate to add a value for all possible ENABLEs to the existing
1755         features.pri.  This lets me make sure that I'm getting them all right!
1756
1757         * qmake/mkspecs/features/features.pri:
1758
1759 2012-05-07  Adam Bergkvist  <adam.bergkvist@ericsson.com>
1760
1761         Added myself as a committer.
1762
1763         Not reviewed.
1764
1765         * Scripts/webkitpy/common/config/committers.py:
1766
1767 2012-05-07  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1768
1769         [EFL][DRT] Implement LayoutTestController::execCommand
1770         https://bugs.webkit.org/show_bug.cgi?id=82286
1771
1772         Reviewed by Antonio Gomes.
1773
1774         Add missing implementation execCommand and isCommandEnabled to EFL's
1775         LayoutTestController. 
1776
1777         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1778         (LayoutTestController::execCommand):
1779         (LayoutTestController::isCommandEnabled):
1780
1781 2012-05-06  Hayato Ito  <hayato@chromium.org>
1782
1783         [prepare-ChangeLog] Excludes a reference file used by reftests from test files.
1784         https://bugs.webkit.org/show_bug.cgi?id=74268
1785
1786         Reviewed by Ryosuke Niwa.
1787
1788         * Scripts/prepare-ChangeLog:
1789         (generateFileList):
1790
1791 2012-05-06  Mary Wu  <mary.wu@torchmobile.com.cn>
1792
1793         Add a contributor to committers.py
1794         https://bugs.webkit.org/show_bug.cgi?id=85761
1795
1796         Reviewed by Unreviewed.
1797
1798         Add myself to committers.py contributor.
1799
1800         * Scripts/webkitpy/common/config/committers.py:
1801
1802 2012-05-06  Dan Bernstein  <mitz@apple.com>
1803
1804         Tools part of: Building and debugging WebKit in the Xcode IDE requires a lot of setup
1805         https://bugs.webkit.org/show_bug.cgi?id=85739
1806
1807         Reviewed by Daniel Bates.
1808
1809         * Scripts/build-webkit: Moved code that copies from WebKitLibraries to the product directory
1810         from here...
1811         * Scripts/copy-webkitlibraries-to-product-directory: ...to this new script.
1812
1813 2012-05-06  Andy Estes  <aestes@apple.com>
1814
1815         Fix the build after r116246.
1816
1817         * DumpRenderTree/LayoutTestController.cpp: Remove an unused function.
1818
1819 2012-05-06  Eric Seidel  <eric@webkit.org>
1820
1821         Update several build-webkit options to match the #defines they toggle
1822         https://bugs.webkit.org/show_bug.cgi?id=85744
1823
1824         Reviewed by Adam Barth.
1825
1826         Most notable here is fixing --3d-canvas to be --webgl these days --
1827         the define was renamed 15 months ago!
1828
1829         * Scripts/webkitperl/FeatureList.pm:
1830
1831 2012-05-06  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1832
1833         [EFL] EFL's LayoutTestController setJavaScriptCanAccessClipboard implementation
1834         https://bugs.webkit.org/show_bug.cgi?id=83687
1835
1836         Reviewed by Antonio Gomes.
1837
1838         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1839         (LayoutTestController::setJavaScriptCanAccessClipboard):
1840
1841 2012-05-06  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1842
1843         Convert isPageBoxVisible to use Internals interface.
1844         https://bugs.webkit.org/show_bug.cgi?id=85692
1845
1846         Reviewed by Darin Adler.
1847
1848         Remove isPageBoxVisible functions, because it is able to work in the
1849         cross-port way through the Internals interface.
1850
1851         * DumpRenderTree/LayoutTestController.cpp:
1852         (LayoutTestController::staticFunctions):
1853         * DumpRenderTree/LayoutTestController.h:
1854         (LayoutTestController):
1855         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
1856         * DumpRenderTree/chromium/LayoutTestController.cpp:
1857         (LayoutTestController::LayoutTestController):
1858         * DumpRenderTree/chromium/LayoutTestController.h:
1859         (LayoutTestController):
1860         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1861         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1862         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1863         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1864         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1865         (LayoutTestController):
1866         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1867         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1868
1869 2012-05-06  Tom Zakrajsek  <tomz@codeaurora.org>
1870
1871         sheriffbot isn't reopening patches after it lands rollouts
1872         https://bugs.webkit.org/show_bug.cgi?id=64418
1873
1874         Reviewed by Adam Barth.
1875
1876         * Scripts/webkitpy/tool/commands/download_unittest.py:
1877         * Scripts/webkitpy/tool/steps/createbug.py:
1878         (CreateBug.run):
1879
1880 2012-05-04  Raphael Kubo da Costa  <rakuco@webkit.org>
1881
1882         webkitpy: Use os.pathsep instead of manually finding out the path separator in NRWT.
1883         https://bugs.webkit.org/show_bug.cgi?id=85697
1884
1885         Reviewed by Eric Seidel.
1886
1887         * Scripts/new-run-webkit-tests: Instead of manually detecting whether
1888         the path separator should be ':' or ';' by checking sys.platform, rely
1889         on os.pathsep which has exactly the same information (os.pathsep is
1890         ':' for Cygwin, from what I could check in Python's sources).
1891
1892 2012-05-05  Sheriff Bot  <webkit.review.bot@gmail.com>
1893
1894         Unreviewed, rolling out r116223.
1895         http://trac.webkit.org/changeset/116223
1896         https://bugs.webkit.org/show_bug.cgi?id=85718
1897
1898         The bots need to set XDG_DATA_DIRS (Requested by rakuco on
1899         #webkit).
1900
1901         * efl/jhbuildrc:
1902         * gtk/jhbuildrc:
1903
1904 2012-05-05  Raphael Kubo da Costa  <rakuco@webkit.org>
1905
1906         [jhbuild] Do not set XDG_DATA_DIRS in jhbuildrc.
1907         https://bugs.webkit.org/show_bug.cgi?id=85717
1908
1909         Reviewed by Martin Robinson.
1910
1911         After r116209, the XDG_DATA_DIRS environment variable is now
1912         passed by webkitpy's layout_tests code to run-with-jhbuild, so
1913         instead of setting a few defaults in jhbuildrc we can now rely on
1914         the values set and really used by each system.
1915
1916         * efl/jhbuildrc:
1917         * gtk/jhbuildrc:
1918
1919 2012-05-05  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1920
1921         [Qt] Buildfix for --minimal build after r116107.
1922
1923         Reviewed by Csaba Osztrogonác.
1924
1925         * qmake/mkspecs/features/features.prf:
1926
1927 2012-05-05  Zan Dobersek  <zandobersek@gmail.com>
1928
1929         [Gtk] GtkDriver is not required anymore after r116134
1930         https://bugs.webkit.org/show_bug.cgi?id=85699
1931
1932         Reviewed by Csaba Osztrogonác.
1933
1934         Remove GtkDriver as it was replaced by the XvfbDriver and is
1935         no longer needed.
1936
1937         * Scripts/webkitpy/layout_tests/port/gtk.py:
1938
1939 2012-05-05  Zan Dobersek  <zandobersek@gmail.com>
1940
1941         Increase the display ID for the Xvfb process to use if pixel_tests argument
1942         is true rather than if pixel_tests option is present. This fixes the
1943         XvfbDriver in the same way r115825 fixed the now-redundant GtkDriver.
1944
1945         Rubber-stamped by Csaba Osztrogonác.
1946
1947         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1948         (XvfbDriver._start):
1949
1950 2012-05-04  Raphael Kubo da Costa  <rakuco@webkit.org>
1951
1952         webkitpy: Preserve XDG_DATA_DIRS in the environment passed to ServerProcess.
1953         https://bugs.webkit.org/show_bug.cgi?id=85694
1954
1955         Reviewed by Martin Robinson.
1956
1957         This is sort of a follow-up to r100674: $XDG_DATA_DIRS should be
1958         preserved, as it is particularly useful for Linux/Unix
1959         environments.  As a real use case, if shared-mime-info is
1960         installed into a non-trivial prefix, Glib and other libraries will
1961         have trouble to properly identify mimetypes (in ResourceResponses,
1962         for example) as the base freedesktop.org mimetype file will not be
1963         found.
1964
1965         * Scripts/webkitpy/layout_tests/port/base.py:
1966         (Port.to.setup_environ_for_server):
1967
1968 2012-05-04  Gustavo Noronha Silva  <gns@gnome.org>
1969
1970         [GTK] Simplify how libWebCoreModules is linked in, and fix WebKit2 build
1971         https://bugs.webkit.org/show_bug.cgi?id=85691
1972
1973         * GNUmakefile.am: no longer link libWebCoreModules to DumpRenderTree.
1974
1975 2012-05-04  Eric Seidel  <eric@webkit.org>
1976
1977         Make capitalization and descriptions in FeatureList.pm more consistent to make easier to autogenerate
1978         https://bugs.webkit.org/show_bug.cgi?id=85583
1979
1980         Reviewed by Daniel Bates.
1981
1982         * Scripts/webkitperl/FeatureList.pm:
1983
1984 2012-05-04  Eric Seidel  <eric@webkit.org>
1985
1986         Sort entries in FeaturesList.pm to make them easier to autogenerate
1987         https://bugs.webkit.org/show_bug.cgi?id=85584
1988
1989         Reviewed by Daniel Bates.
1990
1991         No functional changes, only sorting.
1992
1993         * Scripts/webkitperl/FeatureList.pm:
1994
1995 2012-05-04  Andy Estes  <aestes@apple.com>
1996
1997         REGRESSION (r115752): WebKit2.NewFirstVisuallyNonEmptyLayout API test times out
1998         https://bugs.webkit.org/show_bug.cgi?id=85675
1999
2000         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayout.cpp:
2001         (TestWebKitAPI::TEST): Disable the test.
2002
2003 2012-04-30  Jer Noble  <jer.noble@apple.com>
2004
2005         Full screen will exit during a provisional load of a non-ancestor iframe.
2006         https://bugs.webkit.org/show_bug.cgi?id=85230
2007
2008         Reviewed by Maciej Stachowiak .
2009
2010         Add support for the new UIDelegate webView:closeFullScreenWithListener: method:
2011         * DumpRenderTree/mac/UIDelegate.mm:
2012         (-[UIDelegate webView:closeFullScreenWithListener:]):
2013
2014         Add support for the new closeFullScreen injected bundle method:
2015         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2016         (WTR::InjectedBundlePage::InjectedBundlePage):
2017         (WTR::InjectedBundlePage::closeFullScreen):
2018         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2019
2020 2012-05-04  Ojan Vafai  <ojan@chromium.org>
2021
2022         Fix test naming to be camel-case like the methods they are testing.
2023
2024         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2025         (test):
2026
2027 2012-05-04  Ojan Vafai  <ojan@chromium.org>
2028
2029         Show the results links in the embedded flakiness dashboard
2030         https://bugs.webkit.org/show_bug.cgi?id=85660
2031
2032         Reviewed by Adam Barth.
2033
2034         Turns out these links don't add much visual noise and it's annoying
2035         to have to go to the dashboard from garden-o-matic just to view the
2036         results for a test.
2037
2038         This is useful for cases where garden-o-matic doesn't properly show the
2039         results (e.g. missing tests). If we address all those cases, then
2040         we could easily hide the results links again if we wanted to.
2041
2042         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2043         (showUpdateInfoForTest):
2044         (htmlForIndividualTestOnAllBuildersWithResultsLinks):
2045         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2046         (test):
2047
2048 2012-05-04  Dirk Pranke  <dpranke@chromium.org>
2049
2050         Revert to using chromium --test-shell mode on SL in NRWT
2051
2052         https://bugs.webkit.org/show_bug.cgi?id=83076
2053
2054         Unreviewed, build fix.
2055
2056         We seem to be seeing a lot more flakiness on the bot since
2057         I switched NRWT to "DRT" mode. We'll try reverting back and
2058         see if this helps things.
2059
2060         * Scripts/webkitpy/layout_tests/port/chromium.py:
2061         (ChromiumDriver.__init__):
2062
2063 2012-05-04  Christophe Dumez  <christophe.dumez@intel.com>
2064
2065         [EFL] Implement layoutTestController.dumpResourceResponseMIMETypes
2066         https://bugs.webkit.org/show_bug.cgi?id=84941
2067
2068         Reviewed by Martin Robinson.
2069
2070         Print MIME type detection messages in EFL's DumpRenderTree if
2071         LayoutTestController's dumpResourceResponseMIMETypes() returns true.
2072
2073         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2074         (DumpRenderTreeChrome::onResponseReceived):
2075
2076 2012-05-04  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2077
2078         [GTK] DRT needs an implementation of LayoutTestController::setDefersLoading and ::goBack
2079         https://bugs.webkit.org/show_bug.cgi?id=85134
2080
2081         Reviewed by Martin Robinson.
2082
2083         Add missing implementation setDefersLoading and goBack to GTK's
2084         LayoutTestController so that the test case rely on these methods
2085         can be enabled.
2086
2087         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2088         (resetDefaultsToConsistentValues):
2089         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2090         (LayoutTestController::goBack):
2091         (LayoutTestController::setDefersLoading):
2092
2093 2012-05-04  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2094
2095         [Qt][NRWT] Run each DRT in it's own xvfb
2096         https://bugs.webkit.org/show_bug.cgi?id=77335
2097
2098         Reviewed by Dirk Pranke.
2099
2100         * Scripts/webkitpy/layout_tests/port/gtk.py:
2101         (GtkPort._driver_class):
2102         * Scripts/webkitpy/layout_tests/port/qt.py:
2103         (QtPort._driver_class):
2104         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py: Added.
2105         (XvfbDriver):
2106         (XvfbDriver._start):
2107         (XvfbDriver._start.x_filter):
2108         (XvfbDriver.stop):
2109
2110 2012-05-04  Zan Dobersek  <zandobersek@gmail.com>
2111
2112         Use destructuring assingment for the return values of the
2113         port.diff_image() method call.
2114
2115         Rubber-stamped by Dirk Pranke.
2116
2117         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
2118         (write_test_result):
2119
2120 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2121
2122         Store fewer runs by default in the flakiness dashboard
2123         https://bugs.webkit.org/show_bug.cgi?id=85568
2124
2125         Reviewed by Dirk Pranke.
2126
2127         This will greatly shrink the size of the json files we load,
2128         resulting in a much faster initial load. People rarely care
2129         about more than 100 runs. For the cases they do, they can still click
2130         the show all runs checkbox to show 500 runs.
2131
2132         * TestResultServer/model/jsonresults.py:
2133
2134 2012-05-04  Christophe Dumez  <christophe.dumez@intel.com>
2135
2136         [EFL] Frame load callbacks output is missing for redirections
2137         https://bugs.webkit.org/show_bug.cgi?id=85173
2138
2139         Reviewed by Antonio Gomes.
2140
2141         Catch the new ewk_frame signals to print the expected output in case
2142         of redirections, when LayoutTestController's dumpFrameLoadCallbacks()
2143         returns true.
2144
2145         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2146         (DumpRenderTreeChrome::createView):
2147         (DumpRenderTreeChrome::onWillSendRequest):
2148         (DumpRenderTreeChrome::onFrameCreated):
2149         (DumpRenderTreeChrome::onFrameRedirectForProvisionalLoad):
2150         (DumpRenderTreeChrome::onFrameRedirectRequested):
2151         (DumpRenderTreeChrome::onResponseReceived):
2152         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
2153         (DumpRenderTreeChrome):
2154
2155 2012-05-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2156
2157         [Qt] Fix up warning about missing GLib/Gio/GStreamer
2158
2159         We don't actually fall back.
2160
2161         Rubber-stamped by Csaba Osztrogonác.
2162
2163         * qmake/mkspecs/features/features.prf:
2164
2165 2012-05-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2166
2167         [Qt] Clarify warning about missing GLib/Gio/GStreamer for media support
2168
2169         And only print it once.
2170
2171         Reviewed by Csaba Osztrogonác.
2172
2173         * qmake/mkspecs/features/features.prf:
2174
2175 2012-05-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2176
2177         [Qt] Clean up and split features.prf into a static list of defaults
2178
2179         The static list of feature defaults is used as a fallback for any
2180         feature that's not dynamically detected or overriden on the command
2181         line (though build-webkit or passing DEFINES+= to qmake).
2182
2183         The static list is complete, which allows for auto-generation based
2184         on Features.py (see bug https://bugs.webkit.org/show_bug.cgi?id=85456)
2185
2186         https://bugs.webkit.org/show_bug.cgi?id=85611
2187
2188         Reviewed by Simon Hausmann.
2189
2190         * Scripts/build-webkit:
2191         * Tools.pro:
2192         * qmake/mkspecs/features/features.prf:
2193         * qmake/mkspecs/features/features.pri: Added.
2194
2195 2012-05-04  Sheriff Bot  <webkit.review.bot@gmail.com>
2196
2197         Unreviewed, rolling out r116085, r116091, and r116095.
2198         http://trac.webkit.org/changeset/116085
2199         http://trac.webkit.org/changeset/116091
2200         http://trac.webkit.org/changeset/116095
2201         https://bugs.webkit.org/show_bug.cgi?id=85628
2202
2203         We are not ready with dependencies on all platform yet (mac) +
2204         problems with debug builds. (Requested by Zoltan on #webkit).
2205
2206         * qmake/mkspecs/features/features.prf:
2207
2208 2012-05-04  Christophe Dumez  <christophe.dumez@intel.com>
2209
2210         [EFL] LayoutTestController does not support overriding WebKitLoadSiteIconsKey
2211         https://bugs.webkit.org/show_bug.cgi?id=85171
2212
2213         Reviewed by Andreas Kling.
2214
2215         Implement support for overriding "WebKitLoadSiteIconsKey" setting in
2216         EFL's LayoutTestController.
2217
2218         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2219         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
2220         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2221         (LayoutTestController::overridePreference):
2222
2223 2012-05-03  Alexis Menard  <alexis.menard@openbossa.org>
2224
2225         [Qt] Enable fullscreen API for WebKit2.
2226         https://bugs.webkit.org/show_bug.cgi?id=85498
2227
2228         Reviewed by Simon Hausmann.
2229
2230         Enable the fullscreen API on the MiniBrowser and turn
2231         it default on WebKit2, disable it on WK1.
2232
2233         * MiniBrowser/qt/BrowserWindow.cpp:
2234         (BrowserWindow::BrowserWindow):
2235         * MiniBrowser/qt/qml/BrowserWindow.qml:
2236         * Scripts/webkitperl/FeatureList.pm:
2237         * qmake/mkspecs/features/features.prf:
2238
2239 2012-05-04  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
2240
2241         REGRESSION(r116008) old-run-webkit-tests broken
2242         https://bugs.webkit.org/show_bug.cgi?id=85597
2243
2244         Reviewed by Simon Hausmann.
2245
2246         Reinstate libraryContainsSymbol.
2247
2248         * Scripts/webkitperl/features.pm:
2249         (libraryContainsSymbol):
2250
2251 2012-05-04  Zoltan Horvath  <zoltan@webkit.org>
2252
2253         [Qt] Set WebCore imagedecoders as default and add fallback to QImageDecoder
2254         https://bugs.webkit.org/show_bug.cgi?id=80400
2255
2256         Remove WTF_USE_QT_IMAGE_DECODER macro.
2257
2258         Reviewed by Simon Hausmann.
2259
2260         * qmake/mkspecs/features/features.prf:
2261
2262 2012-05-03  Christophe Dumez  <christophe.dumez@intel.com>
2263
2264         [EFL] DRT should obey layoutTestController's addURLToRedirect()
2265         https://bugs.webkit.org/show_bug.cgi?id=82722
2266
2267         Reviewed by Maciej Stachowiak.
2268
2269         Implement support for layoutTestController's addURLToRedirect()
2270         in EFL's DRT so that it redirects properly when indicated.
2271
2272         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2273         (DumpRenderTreeChrome::onWillSendRequest):
2274
2275 2012-05-03  Christophe Dumez  <christophe.dumez@intel.com>
2276
2277         [EFL] DRT should support LayoutTestController's willSendRequestReturnsNullOnRedirect()
2278         https://bugs.webkit.org/show_bug.cgi?id=82704
2279
2280         Reviewed by Antonio Gomes.
2281
2282         EFL's DRT needs to obey LayoutTestController's
2283         willSendRequestReturnsNullOnRedirect().
2284
2285         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2286         (DumpRenderTreeChrome::onWillSendRequest):
2287
2288 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2289
2290         git-add-reviewer: Do not put the bug title and its URL in the same line in the commit message.
2291         https://bugs.webkit.org/show_bug.cgi?id=85570
2292
2293         Reviewed by Adam Roben.
2294
2295         The standard format of commit messages and ChangeLog entries has a
2296         short summary in the first line, followed by the bug URL in
2297         Bugzilla. Passing "%s" in the pretty format line to git-rev-list
2298         merges these two lines into a single one separated by a space
2299         character, which is not what we want.
2300
2301         Since "%s%n%n%b" is in practice "%B" (raw message), use it
2302         instead, as the first two lines do not end up being merged
2303         together.
2304
2305         * Scripts/git-add-reviewer:
2306         (writeCommitMessageToFile):
2307
2308 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2309
2310         Improve UI for garden-o-matic examine view when there are fewer than 5 tests.
2311         https://bugs.webkit.org/show_bug.cgi?id=85566
2312
2313         Reviewed by Adam Barth.
2314
2315         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2316         (.):
2317         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2318         (.):
2319         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
2320
2321 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2322
2323         webkitpy: Recognize FreeBSD as a valid platform.
2324         https://bugs.webkit.org/show_bug.cgi?id=81467
2325
2326         Reviewed by Adam Barth.
2327
2328         Add PlatformInfo.is_freebsd() so that trying to use webkit-patch
2329         on FreeBSD does not raise an "unrecognized platform" exception.
2330
2331         Most of this patch contains some unit tests for the mentioned
2332         method, as currently platform-specific code is needed only on
2333         Windows.
2334
2335         * Scripts/webkitpy/common/system/platforminfo.py:
2336         (PlatformInfo.__init__):
2337         (PlatformInfo.is_freebsd):
2338         (PlatformInfo._determine_os_name):
2339         * Scripts/webkitpy/common/system/platforminfo_mock.py:
2340         (MockPlatformInfo.is_freebsd):
2341         * Scripts/webkitpy/common/system/platforminfo_unittest.py:
2342         (fake_platform):
2343         (fake_platform.FakePlatformModule.release):
2344         (TestPlatformInfo.test_real_code):
2345         (TestPlatformInfo.test_os_name_and_wrappers):
2346         (TestPlatformInfo.test_os_version):
2347         (TestPlatformInfo.test_display_name):
2348         (TestPlatformInfo.test_total_bytes_memory):
2349         (TestPlatformInfo.test_free_bytes_memory):
2350         * Scripts/webkitpy/layout_tests/port/factory.py:
2351         (PortFactory._default_port):
2352
2353 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2354
2355         Make the table headers white-space:nowrap. This decreases the vertical size,
2356         which makes for more minimal embedding in garden-o-matic.
2357
2358         * TestResultServer/static-dashboards/flakiness_dashboard_tests.css:
2359         (.table-header-content *):
2360
2361 2012-05-03  Eric Seidel  <eric@webkit.org>
2362
2363         Split build-webkit's feature option list off into its own module in preparation for autogeneration
2364         https://bugs.webkit.org/show_bug.cgi?id=85548
2365
2366         Reviewed by Daniel Bates.
2367
2368         My Perl-fu is very weak.  This moves the feature option list off into a separate module
2369         file so that I can easily autogenerate just that file with the fancy new generate-feature-files command.
2370         It's a bit odd to have an array which we return out to build-webkit to modify, but it turns out
2371         not to matter.  Also, build-webkit doesn't actually use any of the variables we're using to
2372         store the results from the option parse, so those end up inaccessible from build-webkit (except by reference)
2373         which is actually cleaner.  Callers can't assume that they would get a new array every time, but someone
2374         with more Perl-fu than I could create something fancier than this if we need in the future.
2375
2376         * Scripts/build-webkit:
2377         * Scripts/webkitperl/FeatureList.pm: Added.
2378         (getFeatureOptionList):
2379
2380 2012-05-03  Dirk Pranke  <dpranke@chromium.org>
2381
2382         REGRESSION: run-webkit-tests failing on Chromium Mac
2383         https://bugs.webkit.org/show_bug.cgi?id=85459
2384
2385         Unreviewed, build fix.
2386
2387         Turns out select doesn't throw IOErrors, it has it's own kind of
2388         error :(. Trap that instead.
2389
2390         * Scripts/webkitpy/layout_tests/port/server_process.py:
2391         (ServerProcess._wait_for_data_and_update_buffers_using_select):
2392
2393 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2394
2395         [CMake] Rewrite FindCairo.cmake.
2396         https://bugs.webkit.org/show_bug.cgi?id=84895
2397
2398         Reviewed by Daniel Bates.
2399
2400         The old approach relied on pkg-config for finding Cairo (which
2401         introduced a dependency on pkg-config that could be avoided), used
2402         the LibFindMacros code that we should probably remove in the
2403         future and did not use the FindPackageHandleStandardArguments
2404         module.
2405
2406         Change all that by rewriting the module.
2407         - Use the pkg-config output optionally instead of requiring it
2408         like LibFindMacros did.
2409         - Remove the implicit dependency on FreeType which often found it
2410         the wrong way via pkg-config and without considering
2411         CMAKE_PREFIX_PATH.
2412         - Retrieve the Cairo version by looking at cairo-version.h instead
2413         of relying on pkg-config. It requires some additional code for
2414         checking if the desired version has been found, but that will not
2415         be needed once we start depending on CMake 2.8.3 or later.
2416
2417         The only downside is that FPHSA sets <UPPERCASED_NAME>_FOUND
2418         instead of <Name>_FOUND, and to keep things consistent
2419         Cairo_LIBRARIES and Cairo_INCLUDE_DIRS have become CAIRO_LIBRARIES
2420         and CAIRO_INCLUDE_DIRS.
2421
2422         * DumpRenderTree/efl/CMakeLists.txt: Use CAIRO_FOO instead of
2423         Cairo_FOO.
2424         * EWebLauncher/CMakeLists.txt: Ditto.
2425
2426 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2427
2428         Fix modifier and bug sorting on the flakiness dashboard
2429         https://bugs.webkit.org/show_bug.cgi?id=85554
2430
2431         Reviewed by Dirk Pranke.
2432
2433         The sorting assumes that the column header name matches a field
2434         in the testResults object and uses that field for sorting.
2435         Filter the bugs out of the modifiers list and put them in a bugs field
2436         and only put the non-bug modifiers in the modifiers list.
2437
2438         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2439         (populateExpectationsData):
2440         (htmlForSingleTestRow):
2441         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2442
2443 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2444
2445         Fix some state resetting when running unittests. HtmlForSingleTestRow
2446         was flaky.
2447
2448         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2449         (htmlForSingleTestRow):
2450         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2451         (resetGlobals):
2452
2453 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2454
2455         webkitpy: Use PlatformInfo wherever possible in the User class.
2456         https://bugs.webkit.org/show_bug.cgi?id=85549
2457
2458         Reviewed by Dirk Pranke.
2459
2460         Instead of relying on sys.platform to detect whether the current
2461         platform is a Mac, use the PlatformInfo class instead.
2462
2463         * Scripts/webkitpy/common/system/user.py:
2464         (User.__init__):
2465         (User.edit_changelog):
2466
2467 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2468
2469         Remove the percent failed column from the flakiness dashboard
2470         https://bugs.webkit.org/show_bug.cgi?id=85552
2471
2472         Reviewed by Dirk Pranke.
2473
2474         This column isn't useful enough to justify the visual noise.
2475
2476         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2477         (processMissingAndExtraExpectations):
2478         (tableHeaders):
2479         (htmlForSingleTestRow):
2480         (sortTests):
2481         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2482         (test):
2483
2484 2012-05-03  Eric Seidel  <eric@webkit.org>
2485
2486         --coverage should not not use the "feature" options system in build-webkit, similarly remove special-casing for WTF_URL
2487         https://bugs.webkit.org/show_bug.cgi?id=85542
2488
2489         Reviewed by Daniel Bates.
2490
2491         Remove these last two odd-men-out, so that we can generate the default feature options
2492         for build-webkit using the new generate-feature-files command.
2493         The Mac build system has some special casing for ENABLE_ defines which gets them
2494         defined as -D on the build command, even though initially they start out as
2495         environment variables ENABLE_FOO = ENABLE_FOO.  The Feature options in build-webkit
2496         depend on this support, but it doesn't work for WTF_USE_ defines like WTF_USE_WTF_URL.
2497         I'm told that WTF_URL is not actively being worked on, so just removing it for now.
2498         It should probably come back to life as an ENABLE_, or build-webkit or the Xcode projects
2499         should be made generically smarter about these USE_WTF defines.
2500         (There is still WTF_USE_TILED_BACKING_STORE in the feature list, but it only seems
2501         used by Qt.  If someone tried build-webkit --tiled-backing-store on Mac today it
2502         would not work!)
2503
2504         * Scripts/build-webkit:
2505
2506 2012-05-03  Eric Seidel  <eric@webkit.org>
2507
2508         Remove the last bits of Leopard support from our perl scripts
2509         https://bugs.webkit.org/show_bug.cgi?id=85545
2510
2511         Reviewed by Daniel Bates.
2512
2513         Chromium still supports Leopard (for a little while yet), but AppleMacWebKit dropped
2514         Leopard support several months ago.  Remove the last bits of isLeopard() branching
2515         in our Perl scripts, since no one is using this (Chromium has never used this Perl code).
2516
2517         * Scripts/build-webkit:
2518         * Scripts/old-run-webkit-tests:
2519         (countAndPrintLeaks):
2520         (captureSavedCrashLog):
2521         * Scripts/webkitdirs.pm:
2522         (determineArchitecture):
2523         (argumentsForRunAndDebugMacWebKitApp):
2524
2525 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2526
2527         [Chromium] Don't show the WebKit Linux ASAN bot on the flakiness dashboard.
2528         https://bugs.webkit.org/show_bug.cgi?id=85551
2529
2530         Reviewed by Dirk Pranke.
2531
2532         It's not uploading results and in the short term we don't plan to have it start.
2533
2534         * TestResultServer/static-dashboards/builders.js:
2535         (isChromiumWebkitTipOfTreeTestRunner):
2536         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2537
2538 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2539
2540         Show the test name in the dashboard if showChrome==false and we're showing multiple tests
2541         https://bugs.webkit.org/show_bug.cgi?id=85243
2542
2543         Reviewed by Dirk Pranke.
2544
2545         This happens for cases where a test name matches multiple tests.
2546         In practice, this only happens for virtual test suites (e.g. the canvas tests).
2547
2548         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2549         (htmlForIndividualTestOnAllBuildersWithChrome):
2550         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2551         (testHtmlForIndividualTestOnAllBuildersWithChromeNonexistant):
2552         (testHtmlForIndividualTestOnAllBuildersWithChrome):
2553         (testHtmlForIndividualTestOnAllBuildersWithChromeWebkitMaster):
2554         (testHtmlForIndividualTests):
2555
2556 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2557
2558         Port the flakiness dashboard unittests to qunit
2559         https://bugs.webkit.org/show_bug.cgi?id=85544
2560
2561         Reviewed by Adam Barth.
2562
2563         * TestResultServer/static-dashboards/dashboard_base.js:
2564         (appendJSONScriptElements):
2565         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2566         (resetGlobals):
2567         (runExpectationsTest):
2568         (test):
2569         * TestResultServer/static-dashboards/run-unittests.html: Added.
2570
2571 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2572
2573         webkitpy: Remove check for readline on Mac.
2574         https://bugs.webkit.org/show_bug.cgi?id=85547
2575
2576         Reviewed by Eric Seidel.
2577
2578         Python 2.5 is not supported by the WebKit project anymore, and
2579         Python's readline documentation does not mention the module not
2580         being available in the supported versions.
2581
2582         * Scripts/webkitpy/common/system/user.py:
2583
2584 2012-05-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2585
2586         [NRWT] Run performance tests with lock
2587         https://bugs.webkit.org/show_bug.cgi?id=78628
2588
2589         Reviewed by Tony Chang.
2590
2591         Locking performance tests (like we do for http tests) will force them
2592         to run in serial. This reduces the load of the machine when running perf
2593         tests and minimizes the chances of the tests to fail, specially
2594         timeouts.
2595
2596         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2597         (Manager.__init__):
2598         (Manager._is_perf_test):
2599         (Manager._test_requires_lock):
2600         * Scripts/webkitpy/layout_tests/port/http_lock.py:
2601
2602 2012-05-03  Dirk Pranke  <dpranke@chromium.org>
2603
2604         Attempt to fix a crash on SL ...
2605
2606         Unreviewed, build fix.
2607
2608         It looks like we may be getting an IOError raised during
2609         select() on chromium mac, and that error isn't being caught.
2610         Pushing it into the try/catch block will hopefully catch this
2611         and allow us to see why the subprocess is crashing.
2612
2613         * Scripts/webkitpy/layout_tests/port/server_process.py:
2614         (ServerProcess._wait_for_data_and_update_buffers_using_select):
2615
2616 2012-05-02  Jon Honeycutt  <jhoneycutt@apple.com>
2617
2618         WebFrameLoaderClient::dispatchWillSendSubmitEvent() needs to be
2619         implemented for WebKit2
2620         https://bugs.webkit.org/show_bug.cgi?id=84304
2621
2622         Reviewed by Jessie Berlin.
2623
2624         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2625         Added new files to project.
2626
2627         * TestWebKitAPI/Tests/WebKit2/WillSendSubmitEvent.cpp: Added.
2628         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
2629         Look for the "DidReceiveWillSendSubmitEvent" message. The message body
2630         will be the dictionary of form control names and values passed to the
2631         injected bundle's willSendSubmitEvent handler; look for the pairs
2632         "textField" -> "text field" and "passwordField" -> "password field".
2633         Check that "hiddenField" is not included.
2634         (TestWebKitAPI::setInjectedBundleClient):
2635         (TestWebKitAPI::TEST):
2636
2637         * TestWebKitAPI/Tests/WebKit2/WillSendSubmitEvent_Bundle.cpp: Added.
2638         (TestWebKitAPI::willSendSubmitEvent):
2639         Post a message to the TestWebKitAPI process with the dictionary of form
2640         control names and values that we received.
2641         (TestWebKitAPI::WillSendSubmitEventTest::WillSendSubmitEventTest):
2642         (TestWebKitAPI::WillSendSubmitEventTest::didCreatePage):
2643         Set up the page's form client.
2644
2645         * TestWebKitAPI/Tests/WebKit2/auto-submitting-form.html: Added.
2646
2647 2012-05-03  Eric Seidel  <eric@webkit.org>
2648
2649         Remove dead code from build-webkit for working around old XCode dependency bug
2650         https://bugs.webkit.org/show_bug.cgi?id=85535
2651
2652         Reviewed by Adam Barth.
2653
2654         Xcode used to have a bug whereby it didn't notice if you changed the #defines
2655         passed on the command line to xcodebuild.  I wrote code to work around this
2656         while bringing up SVG support 5-6 years ago.  I think we can finally remove this
2657         code from build-webkit, since SVG is the only feature flag to ever use it
2658         (modern in-progress features like MathML don't seem to be using the support)
2659         an I suspect the XCode bug has long been fixed.
2660
2661         * Scripts/build-webkit:
2662         * Scripts/webkitperl/features.pm:
2663
2664 2012-05-03  Andy Estes  <aestes@apple.com>
2665
2666         InjectedBundleControllerMac.mm needs to include config.h
2667         https://bugs.webkit.org/show_bug.cgi?id=85531
2668
2669         Reviewed by Andreas Kling.
2670
2671         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm:
2672
2673 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2674
2675         webkitpy: Use PlatformInfo whenever possible in server_process.py.
2676         https://bugs.webkit.org/show_bug.cgi?id=85516
2677
2678         Reviewed by Eric Seidel.
2679
2680         Eliminate direct access to sys.platform wherever possible by using
2681         PlatformInfo.is_win() instead.
2682
2683         * Scripts/webkitpy/layout_tests/port/server_process.py:
2684         (ServerProcess.__init__): Remove the executive parameter as it was
2685         not passed by any caller, obtain a SystemHost from port_obj
2686         instead.
2687         (ServerProcess._log): Style, remove extra empty line.
2688         (ServerProcess._start):
2689         (ServerProcess.stop):
2690         (ServerProcess.kill):
2691         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
2692         (TrivialMockPort.__init__): Create a MockSystemHost used by
2693         ServerProcess.
2694         (TestServerProcess.test_broken_pipe): Test different platforms.
2695
2696 2012-05-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2697
2698         webkitpy: Style checker broken when committing a PNG
2699         https://bugs.webkit.org/show_bug.cgi?id=85504
2700
2701         Reviewed by Dirk Pranke.
2702
2703         * Scripts/webkitpy/style/checkers/png.py:
2704         (PNGChecker.__init__):
2705
2706 2012-05-03  Philippe Normand  <pnormand@igalia.com>
2707
2708         Unreviewed, fix another call to join(), similar to the ones landed
2709         in r115975.
2710
2711         * Scripts/webkitdirs.pm:
2712         (jhbuildConfigurationChanged):
2713
2714 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2715
2716         webkitpy: Remove sys.platform accesses in CrashLogs.
2717         https://bugs.webkit.org/show_bug.cgi?id=85436
2718
2719         Reviewed by Dirk Pranke.
2720
2721         Instead of directly checking for sys.platform == 'darwin', use a
2722         SystemHost and call PlatformInfo.is_mac().
2723
2724         * Scripts/webkitpy/common/system/crashlogs.py:
2725         (CrashLogs.__init__):
2726         (CrashLogs.find_newest_log):
2727         (CrashLogs._log_directory_darwin):
2728         (CrashLogs._find_newest_log_darwin):
2729         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
2730         (CrashLogsTest.test_find_log_darwin):
2731         * Scripts/webkitpy/layout_tests/port/mac.py:
2732         (MacPort._get_crash_log):
2733         * Scripts/webkitpy/layout_tests/port/test.py:
2734         (TestDriver.run_test):
2735         * Scripts/webkitpy/tool/commands/queries.py:
2736         (execute):
2737
2738 2012-05-03  Philippe Normand  <pnormand@igalia.com>
2739
2740         [GTK] join() is wrongly used in webkitdirs.pm since r115532
2741         https://bugs.webkit.org/show_bug.cgi?id=85501
2742
2743         Reviewed by Martin Robinson.
2744
2745         * Scripts/webkitdirs.pm:
2746         (getJhbuildPath):
2747         (buildAutotoolsProject):
2748
2749 2012-05-03  Zan Dobersek  <zandobersek@gmail.com>
2750
2751         [Gtk] ImageDiff crashes when it can't open a display
2752         https://bugs.webkit.org/show_bug.cgi?id=85476
2753
2754         Reviewed by Martin Robinson.
2755
2756         Replace gdk_init() initialization call with the g_type_init()
2757         call. GdkPixbufLoader doesn't actually need the current call to be
2758         made and this change will also remove crashes when ImageDiff
2759         is started in an environment without a DISPLAY env.
2760
2761         * DumpRenderTree/gtk/ImageDiff.cpp:
2762         (main):
2763
2764 2012-05-03  Zan Dobersek  <zandobersek@gmail.com>
2765
2766         ImageDiff should be run inside a properly established environment
2767         https://bugs.webkit.org/show_bug.cgi?id=85292
2768
2769         Reviewed by Martin Robinson.
2770
2771         When creating the ImageDiff server process, pass along a freshly-set-up environment
2772         in which the process should be executed in.
2773
2774         * Scripts/webkitpy/layout_tests/port/webkit.py:
2775         (WebKitPort._start_image_diff_process):
2776
2777 2012-05-03  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2778
2779         [Qt] Allow the web process and WTR to be paused on startup
2780
2781         Makes it easier to debug the web process or run-webkit-tests -2, as you
2782         have ample time to attach gdb to the process.
2783
2784         Reviewed by Simon Hausmann.
2785
2786         * Scripts/webkitpy/layout_tests/port/qt.py:
2787         (QtPort.setup_environ_for_server):
2788         * WebKitTestRunner/qt/main.cpp:
2789         (main):
2790
2791 2012-05-03  Csaba Osztrogonác  <ossy@webkit.org>
2792
2793         Password protect "Stop build" button
2794         https://bugs.webkit.org/show_bug.cgi?id=81982
2795
2796         Reviewed by Ryosuke Niwa.
2797
2798         Only authenticated users should be able to use stopBuild, stopAllBuilds, cancelPendingBuild
2799         features to prevent killing builds on build.webkit.org by SPAM bots.
2800
2801         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2802
2803 2012-05-03  Dongwoo Im  <dw.im@samsung.com>
2804
2805         [DRT] Remove the name of parameter, visibility, from the setPageVisibility stub implementations.
2806         https://bugs.webkit.org/show_bug.cgi?id=85468
2807
2808         Reviewed by Nikolas Zimmermann.
2809
2810         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
2811         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2812         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2813         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2814         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2815         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2816         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2817         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2818         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2819         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2820
2821 2012-05-03  Stephanie Lewis  <slewis@apple.com>
2822
2823         https://bugs.webkit.org/show_bug.cgi?id=85450 unbounded growth of JSDOMWindowShells loading pages in the same window
2824         <rdar://problem/11320059> REGRESSION (r115083): PLT3 shows linear memory growth and gets slower with each run
2825
2826         Reviewed by Brady Eidson.
2827
2828         The API added for DOMWindowExtension, didCreateGlobalObjectForFrame, would create a global object
2829         for every world, even those that did not need the callback.  This had the side effect of creating a
2830         JSDOMWindowShell that the associated world didn't necessarily know to clean up.  Instead of creating
2831         unnecessary objects change the API to globalObjectIsAvailableForFrame and do not pass the global object
2832         in the API.  The object can be accessed later by those worlds which require it.
2833
2834         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic.cpp:
2835         (TestWebKitAPI):
2836         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
2837         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
2838         (TestWebKitAPI):
2839         (DOMWindowExtensionBasic):
2840         (TestWebKitAPI::DOMWindowExtensionBasic::didCreatePage):
2841         (TestWebKitAPI::DOMWindowExtensionBasic::globalObjectIsAvailableForFrame):
2842         (TestWebKitAPI::globalObjectIsAvailableForFrameCallback):
2843
2844 2012-05-03  Nikolas Zimmermann  <nzimmermann@rim.com>
2845
2846         Not reviewed. Fix mac build, it was missing setPageVisibility/resetPageVisibility stub implementations.
2847
2848         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2849         (LayoutTestController::resetPageVisibility):
2850         (LayoutTestController::setPageVisibility):
2851
2852 2012-05-02  Alexander Færøy  <ahf@0x90.dk>
2853
2854         Rename deviceDPI to devicePixelRatio
2855         https://bugs.webkit.org/show_bug.cgi?id=85049
2856
2857         Reviewed by Kenneth Rohde Christiansen.
2858
2859         Use 1.5 as device pixel ratio for the Qt MiniBrowser.
2860
2861         * MiniBrowser/qt/qml/BrowserWindow.qml:
2862
2863 2012-05-03  Dongwoo Im  <dw.im@samsung.com>
2864
2865         [EFL][DRT] Implement the LayoutTestController's methods related to the Page Visibility API.
2866         https://bugs.webkit.org/show_bug.cgi?id=85347
2867
2868         Reviewed by Nikolas Zimmermann.
2869
2870         * DumpRenderTree/LayoutTestController.h:
2871         (LayoutTestController):
2872         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
2873         (LayoutTestController::resetPageVisibility): Not implemented.
2874         (LayoutTestController::setPageVisibility): Not implemented.
2875         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2876         (LayoutTestController::resetPageVisibility): Reset the page visibility.
2877         (LayoutTestController::setPageVisibility): Set the page visibility.
2878         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2879         (LayoutTestController::resetPageVisibility): Not implemented.
2880         (LayoutTestController::setPageVisibility): Not implemented.
2881         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2882         (LayoutTestController::resetPageVisibility): Not implemented.
2883         (LayoutTestController::setPageVisibility): Not implemented.
2884         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2885         (LayoutTestController::resetPageVisibility): Not implemented.
2886         (LayoutTestController::setPageVisibility): Not implemented.
2887         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2888         (LayoutTestController::resetPageVisibility): Not implemented.
2889         (LayoutTestController::setPageVisibility): Not implemented.
2890
2891 2012-05-02  Sheriff Bot  <webkit.review.bot@gmail.com>
2892
2893         Unreviewed, rolling out r115918.
2894         http://trac.webkit.org/changeset/115918
2895         https://bugs.webkit.org/show_bug.cgi?id=85452
2896
2897         "Broke test-webkitpy on Apple Lion bots" (Requested by estes
2898         on #webkit).
2899
2900         * Scripts/webkitpy/common/system/crashlogs.py:
2901         (CrashLogs.__init__):
2902         (CrashLogs.find_newest_log):
2903         (CrashLogs._log_directory_darwin):
2904         (CrashLogs._find_newest_log_darwin):
2905         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
2906         (CrashLogsTest.test_find_log_darwin):
2907         * Scripts/webkitpy/layout_tests/port/mac.py:
2908         (MacPort._get_crash_log):
2909         * Scripts/webkitpy/layout_tests/port/test.py:
2910         (TestDriver.run_test):
2911         * Scripts/webkitpy/tool/commands/queries.py:
2912         (execute):
2913
2914 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
2915
2916         webkitpy: Remove sys.platform accesses in CrashLogs.
2917         https://bugs.webkit.org/show_bug.cgi?id=85436
2918
2919         Reviewed by Dirk Pranke.
2920
2921         Instead of directly checking for sys.platform == 'darwin', use a
2922         SystemHost and call PlatformInfo.is_mac().
2923
2924         * Scripts/webkitpy/common/system/crashlogs.py:
2925         (CrashLogs.__init__):
2926         (CrashLogs.find_newest_log):
2927         (CrashLogs._log_directory_darwin):
2928         (CrashLogs._find_newest_log_darwin):
2929         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
2930         (CrashLogsTest.test_find_log_darwin):
2931         * Scripts/webkitpy/layout_tests/port/mac.py:
2932         (MacPort._get_crash_log):
2933         * Scripts/webkitpy/layout_tests/port/test.py:
2934         (TestDriver.run_test):
2935         * Scripts/webkitpy/tool/commands/queries.py:
2936         (execute):
2937
2938 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
2939
2940         webkitpy: Special-case win32 calls in file_lock and executive.
2941         https://bugs.webkit.org/show_bug.cgi?id=85431
2942
2943         Reviewed by Dirk Pranke.
2944
2945         Simplify accesses to sys.platform by only checking if sys.platform
2946         is "win32" and removing explicit checks for the else case.
2947
2948         * Scripts/webkitpy/common/system/executive.py:
2949         (Executive.check_running_pid):
2950         * Scripts/webkitpy/common/system/file_lock.py:
2951         (FileLock._create_lock):
2952         (FileLock._remove_lock):
2953
2954 2012-05-02  Ojan Vafai  <ojan@chromium.org>
2955
2956         Remove pink from flakiness dashboard
2957         https://bugs.webkit.org/show_bug.cgi?id=85422
2958
2959         Reviewed by Dirk Pranke.
2960
2961         The pink indicates that the test doesn't match it's listing in test_expectations.txt.
2962         In practice, this just added noise and confused people.
2963
2964         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2965         (htmlForSingleTestRow):
2966         (showLegend):
2967         * TestResultServer/static-dashboards/flakiness_dashboard_tests.css:
2968         (#legend-contents *):
2969         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2970         (testHtmlForSingleTestRow):
2971
2972 2012-05-02  Gustavo Noronha Silva  <gns@gnome.org>
2973
2974         Unreviewed. Fix mistake I committed when applying review comments.
2975
2976         * jhbuild/jhbuild-wrapper: dependencies_path was getting 'Root'
2977         appended to its path, so jhbuild itself was being cloned and
2978         installed in the wrong place when using WEBKITOUTPUTDIR.
2979
2980 2012-05-02  Dirk Pranke  <dpranke@chromium.org>
2981
2982         nrwt: make ServerProcess work on chromium win
2983         https://bugs.webkit.org/show_bug.cgi?id=85333
2984
2985         Reviewed by Ojan Vafai.
2986
2987         This change implements a poor man's select() that will
2988         slow-spin doing non-blocking reads on the stdout and stderr
2989         named pipes connecting the worker to the driver. Seems to work
2990         and I have yet to see much overhead or ill effects but it
2991         probably needs more testing.
2992
2993         * Scripts/webkitpy/layout_tests/port/chromium.py:
2994         (ChromiumDriver.__init__):
2995         * Scripts/webkitpy/layout_tests/port/server_process.py:
2996         (ServerProcess._start):
2997         (ServerProcess._wait_for_data_and_update_buffers_using_select):
2998         (ServerProcess._wait_for_data_and_update_buffers_using_win32_apis):
2999         (ServerProcess):
3000         (ServerProcess._non_blocking_read_win32):
3001         (ServerProcess._read):
3002         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
3003         (TestServerProcess.test_basic): Added.
3004
3005 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3006
3007         webkitpy: Use a SystemHost instead of PlatformInfo in the PNG checker.
3008         https://bugs.webkit.org/show_bug.cgi?id=85426
3009
3010         Reviewed by Dirk Pranke.
3011
3012         r115880 replaced the direct use of sys.platform with a
3013         PlatformInfo object. Dirk Pranke later told me SystemHost objects
3014         should be preferred, as nothing out of webkitpy.common.system
3015         should create its own PlatformInfo instances.
3016
3017         * Scripts/webkitpy/common/system/systemhost_mock.py:
3018         (MockSystemHost.__init__): Make it possible to specify a custom
3019         filesystem object.
3020         * Scripts/webkitpy/style/checkers/png.py:
3021         (PNGChecker.__init__): Accept a host object instead of a
3022         filesystem one, which is now obtained from host.
3023         (PNGChecker._config_file_path): Retrieve PlatformInfo from
3024         SystemHost.
3025         * Scripts/webkitpy/style/checkers/png_unittest.py:
3026         (PNGCheckerTest.test_init): Adjust calls to PNGChecker to its new
3027         constructor signature.
3028         (PNGCheckerTest.test_check): Ditto.
3029
3030 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3031
3032         webkitpy: Remove direct checks for sys.platform == 'darwin' in the layout test classes.
3033         https://bugs.webkit.org/show_bug.cgi?id=85424
3034
3035         Reviewed by Tony Chang.
3036
3037         We should reduce direct usage of sys.platform in webkitpy and use
3038         PlatformInfo instead.
3039
3040         * Scripts/webkitpy/layout_tests/servers/http_server.py:
3041         (Lighttpd._prepare_config): Call PlatformInfo.is_mac() instead of
3042         checking for 'darwin' directly.
3043         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
3044         (HttpServerBase.__init__): Ditto.
3045
3046 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3047
3048         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
3049         https://bugs.webkit.org/show_bug.cgi?id=85419
3050
3051         Reviewed by Dirk Pranke.
3052
3053         As pointed out by Dirk Pranke, a PlatformInfo is not supposed to
3054         be created directly, we should should a SystemHost instead.
3055
3056         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3057         (MainTest.setUp): Create a SystemHost and obtain the PlatformInfo
3058         object from it instead of creating PlatformInfo directly.
3059
3060 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3061
3062         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
3063         https://bugs.webkit.org/show_bug.cgi?id=85419
3064
3065         Rubber-stamped by Dirk Pranke.
3066
3067         Follow-up patch to r115887 which fixes some capitalization problems
3068         pointed out in comments 2 and 4.
3069
3070         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3071         (MainTest.setUp):
3072         (MainTest.test_child_processes_2):
3073         (MainTest.test_child_processes_min):
3074         (MainTest.test_exception_raised):
3075         (MainTest.test_keyboard_interrupt):
3076
3077 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3078
3079         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
3080         https://bugs.webkit.org/show_bug.cgi?id=85419
3081
3082         Reviewed by Dirk Pranke.
3083
3084         Another part of the work towards replacing direct sys.platform
3085         usage with PlatformInfo.
3086
3087         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3088         (MainTest.setUp): Move the setting of SHOULD_TEST_PROCESSES
3089         here. Even though it is costlier, unittest.setUpClass was
3090         introduced in Python 2.7.
3091         (MainTest.test_child_processes_2): Use self.SHOULD_TEST_PROCESSES.
3092         (MainTest.test_child_processes_min): Ditto.
3093         (MainTest.test_exception_raised): Ditto.
3094         (MainTest.test_keyboard_interrupt): Ditto.
3095         (MainTest.test_crash_log): Check for PlatformInfo.is_mac().
3096         (MainTest.test_web_process_crash_log): Ditto, and remove duplicate
3097         check at the end.
3098         (MainTest.test_no_http_tests): Remove check for (cygwin && Python
3099         2.5.2), as we only support Python 2.6+.
3100
3101 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3102
3103         webkitpy: Remove direct sys.platform usage in the PNG checker.
3104         https://bugs.webkit.org/show_bug.cgi?id=85407
3105
3106         Reviewed by Tony Chang.
3107
3108         Work towards replacing direct sys.platform usage with
3109         PlatformInfo.
3110
3111         * Scripts/webkitpy/common/system/platforminfo_mock.py:
3112         (MockPlatformInfo.__init__): Accept os_name and os_version as
3113         arguments.
3114         * Scripts/webkitpy/style/checkers/png.py:
3115         (PNGChecker.__init__): Create a PlatformInfo object instead of
3116         using sys.platform when needed.
3117         (PNGChecker._config_file_path): Check for the current platform via
3118         PlatformInfo.
3119         * Scripts/webkitpy/style/checkers/png_unittest.py:
3120         (PNGCheckerTest.test_check): Update tests after the changes above.
3121
3122 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3123
3124         [watchlist] Update watchlist rules for EFL-related patches.
3125         http://bugs.webkit.org/show_bug.cgi?id=85160
3126
3127         Reviewed by Eric Seidel.
3128
3129         * Scripts/webkitpy/common/config/watchlist: Monitor
3130         Source/WTF/wtf/efl and LayoutTests/platform/efl as well.
3131
3132 2012-05-02  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3133
3134         Revert r115191 - "Make the web view's url property follow the active url"
3135
3136         It uncovered/caused issues in the icon-implementation that can't be
3137         easily fixed, so rolling out instead.
3138
3139         https://bugs.webkit.org/show_bug.cgi?id=77554
3140
3141         Rubber-stamped by Simon Hausmann.
3142
3143         * MiniBrowser/qt/qml/BrowserWindow.qml:
3144         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
3145         (TestWebKitAPI::TEST):
3146
3147 2012-05-02  Zan Dobersek  <zandobersek@gmail.com>
3148
3149         Reftest mismatches are providing undefined diff percent in results.html
3150         https://bugs.webkit.org/show_bug.cgi?id=85283
3151
3152         Reviewed by Ojan Vafai.
3153
3154         When writing test results for a reftest mismatch, also take into account the
3155         difference percent that is produced when generating image diff. This value is
3156         then used when creating test results in the form of dictionaries, stored under
3157         the 'image_diff_percent' key. It ultimately ends up in JSON results data,
3158         providing additional information about the reftest difference.
3159
3160         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3161         (interpret_test_failures): Store the diff percent for reftest failure under the
3162         'image_diff_percent' key.
3163         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
3164         (write_test_result): Also set the new diff_percent attribute of the reftest
3165         mismatch failure to the value of the difference percent ImageDiff outputs.
3166         * Scripts/webkitpy/layout_tests/models/test_failures.py:
3167         (FailureReftestMismatch.__init__): Add the new diff_percent attribute with the
3168         default value of None.
3169         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3170         (EndToEndTest.test_reftest_with_two_notrefs): Update the test case expected results.
3171
3172 2012-05-02  Terry Anderson  <tdanderson@chromium.org>
3173
3174         Addition of a contributor to committers.py
3175         https://bugs.webkit.org/show_bug.cgi?id=85324
3176
3177         Reviewed by Unreviewed.
3178
3179         * Scripts/webkitpy/common/config/committers.py:
3180
3181 2012-05-02  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3182
3183         [EFL] EFL's LayoutTestController overridePreference implementation
3184         https://bugs.webkit.org/show_bug.cgi?id=83007
3185
3186         Reviewed by Antonio Gomes.
3187
3188         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3189         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
3190         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3191         (toBool): aux function converting JSStringRef to bool
3192         (toInt): aux function converting JSStringRef to int
3193         (LayoutTestController::overridePreference): added implementation
3194
3195 2012-05-01  Raphael Kubo da Costa  <rakuco@webkit.org>
3196
3197         webkitpy: Fix sys.platform check in find_files_unittest.
3198         https://bugs.webkit.org/show_bug.cgi?id=85339
3199
3200         Reviewed by Ryosuke Niwa.
3201
3202         Follow-up to r101618 -- 'win' is not a valid sys.platform value,
3203         'win32' should be used instead. The check always returned false
3204         due to this, so assert_normalize() was never called.
3205
3206         * Scripts/webkitpy/common/find_files_unittest.py:
3207         (TestWinNormalize.test_win): Add the missing import for the
3208         FileSystem class and fix the sys.platform check.
3209
3210 2012-05-02  Zan Dobersek  <zandobersek@gmail.com>
3211
3212         [GTK] drag&drop related failures
3213         https://bugs.webkit.org/show_bug.cgi?id=84675
3214
3215         Reviewed by Philippe Normand.
3216
3217         Rather than checking options to see whether we're running pixel tests
3218         the pixel_tests argument passed to GtkDriver._start should be used as
3219         it also covers cases when pixel tests should be run because of testing
3220         reftests. This again creates different Xvfb processes with proper
3221         display IDs for a worker that's performing both ordinary and pixel
3222         tests, as was the case before changes made in r114788.
3223
3224         * Scripts/webkitpy/layout_tests/port/gtk.py:
3225         (GtkDriver._start):
3226
3227 2012-05-02  Simon Hausmann  <simon.hausmann@nokia.com>
3228
3229         [Qt] Enable device orientation/motion with Qt 5
3230
3231         Reviewed by Tor Arne Vestbø.
3232
3233         Enable device orientation/motion features with Qt 5 if Qt Sensors is available.
3234
3235         * qmake/mkspecs/features/features.prf:
3236
3237 2012-05-01  Raphael Kubo da Costa  <rakuco@webkit.org>
3238
3239         webkitpy: Remove unused platform imports.
3240         https://bugs.webkit.org/show_bug.cgi?id=85341
3241
3242         Reviewed by Eric Seidel.
3243
3244         Remove a few uses of `import platform' which were either rendered
3245         unneeded after r105931 or were never needed at all.
3246
3247         * Scripts/webkitpy/common/system/executive.py:
3248         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
3249
3250 2012-05-01  Terry Anderson  <tdanderson@chromium.org>
3251
3252         [chromium] Accept four parameters when dispatching a WebInputEvent::GestureTap in chromium DRT eventSender
3253         https://bugs.webkit.org/show_bug.cgi?id=85289
3254
3255         Reviewed by Eric Seidel.
3256
3257         * DumpRenderTree/chromium/EventSender.cpp:
3258         (EventSender::gestureEvent):
3259             This change will allow for testing of WK85101. Currently
3260             no tests use the extra parameters deltaX and deltaY. Added
3261             a check for the size of |arguments|.
3262
3263 2012-05-01  Eric Seidel  <eric@webkit.org>
3264
3265         Add myself to the watchlist to watch all rendering changes, as well as sort the watchers in the list
3266         https://bugs.webkit.org/show_bug.cgi?id=85305
3267
3268         Reviewed by Adam Barth.
3269
3270         * Scripts/webkitpy/common/config/watchlist:
3271
3272 2012-05-01  Sheriff Bot  <webkit.review.bot@gmail.com>
3273
3274         Unreviewed, rolling out r115735.
3275         http://trac.webkit.org/changeset/115735
3276         https://bugs.webkit.org/show_bug.cgi?id=85314
3277
3278         cause two tests to crash due to an ASSERTION failure
3279         (Requested by zhenyao on #webkit).
3280
3281         * DumpRenderTree/chromium/EventSender.cpp:
3282         (EventSender::gestureEvent):
3283
3284 2012-05-01  Terry Anderson  <tdanderson@chromium.org>
3285
3286         [chromium] Accept four parameters when dispatching a WebInputEvent::GestureTap in chromium DRT eventSender
3287         https://bugs.webkit.org/show_bug.cgi?id=85289
3288
3289         Reviewed by Eric Seidel.
3290
3291         * DumpRenderTree/chromium/EventSender.cpp:
3292         (EventSender::gestureEvent):
3293             This change will allow for testing of WK85101. Currently no tests use the extra
3294             parameters deltaX and deltaY.
3295
3296 2012-05-01  Zan Dobersek  <zandobersek@gmail.com>
3297
3298         Print out stderr output of ImageDiff if it is present
3299         https://bugs.webkit.org/show_bug.cgi?id=85285
3300
3301         Reviewed by Philippe Normand.
3302
3303         Log the stderr output of ImageDiff as a warning if present. This is
3304         especially helpful when ImageDiff crashes and stderr data might contain
3305         relevant information about what caused the crash.
3306
3307         * Scripts/webkitpy/layout_tests/port/webkit.py:
3308         (WebKitPort._read_image_diff):
3309
3310 2012-05-01  Kenneth Rohde Christiansen  <kenneth@webkit.org>
3311
3312         [Qt] Add an experimental extension to set the min. contents width
3313         https://bugs.webkit.org/show_bug.cgi?id=85281
3314
3315         Reviewed by Antonio Gomes.
3316
3317         * MiniBrowser/qt/qml/BrowserWindow.qml:
3318
3319         Set the value of preferredMinimumContentsWidth to 980.
3320
3321 2012-04-30  Ojan Vafai  <ojan@chromium.org>
3322
3323         Fix tests to not depend on the actual list of builders at build.chromium.org.
3324
3325         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
3326         (testHtmlForIndividualTestOnAllBuildersWithChrome):
3327         (testHtmlForIndividualTestOnAllBuildersWithChromeWebkitMaster):
3328
3329 2012-04-29  Sam Weinig  <sam@webkit.org>
3330
3331         Remove BlobBuilder
3332         https://bugs.webkit.org/show_bug.cgi?id=84036
3333
3334         Reviewed by Anders Carlsson.
3335
3336         * Scripts/build-webkit:
3337         * qmake/mkspecs/features/features.prf:
3338         Make exposing the WebKitBlobBuilder JS object conditional on a new
3339         ENABLE_LEGACY_WEBKIT_BLOB_BUILDER flag. Don't enable this for the Mac,
3340         but do for everyone else.
3341
3342 2012-04-30  Amruth Raj Padmanabhuni  <amruthraj@motorola.com>
3343
3344         [GTK] Add glib as a dependency to build glib-networking
3345         https://bugs.webkit.org/show_bug.cgi?id=85181
3346
3347         Reviewed by Martin Robinson.
3348
3349         * gtk/jhbuild.modules:
3350
3351 2012-04-30  Alexis Menard  <alexis.menard@openbossa.org>
3352
3353         Move myself to reviewers.
3354
3355         Reviewed by Andreas Kling.
3356
3357         * Scripts/webkitpy/common/config/committers.py:
3358
3359 2012-04-30  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
3360
3361         Unreviewed, move myself to reviewers
3362
3363         * Scripts/webkitpy/common/config/committers.py:
3364
3365 2012-04-30  Carlos Garcia Campos  <cgarcia@igalia.com>
3366
3367         Unreviewed. Fix make distcheck.
3368
3369         * TestWebKitAPI/GNUmakefile.am: Add missing headers to compilation
3370         and resources to EXTRA_DIST.
3371
3372 2012-04-30  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
3373
3374         [GTK] DRT support for layoutTestController.removeOriginAccessWhitelistEntry
3375         https://bugs.webkit.org/show_bug.cgi?id=85156
3376
3377         Reviewed by Martin Robinson.
3378
3379         Add missing implementation removeOriginAccessWhitelistEntry to
3380         GTK's LayoutTestController.
3381
3382         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3383         (LayoutTestController::removeOriginAccessWhitelistEntry):
3384
3385 2012-04-29  Maciej Stachowiak  <mjs@apple.com>
3386
3387         REGRESSION: On Lion, run-webkit-tests changes the display color profile even when not running pixel tests
3388         https://bugs.webkit.org/show_bug.cgi?id=81729
3389
3390         Reviewed by Dan Bernstein.
3391
3392         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3393         (Manager._set_up_run): Only launch LayoutTestHelper when running pixel tests.
3394
3395 2012-04-29  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
3396
3397         [GTK] DRT needs an implementation of layoutTestController.setDomainRelaxationForbiddenForURLScheme
3398         https://bugs.webkit.org/show_bug.cgi?id=85131
3399
3400         Reviewed by Martin Robinson.
3401
3402         Add missing implementation setDomainRelaxationForbiddenForURLScheme to
3403         GTK's LayoutTestController.
3404
3405         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3406         (LayoutTestController::setDomainRelaxationForbiddenForURLScheme):
3407
3408 2012-04-28  Sam Weinig  <sam@webkit.org>
3409
3410         Smooth scrolling needs a new key
3411         <rdar://problem/11331632>
3412
3413         Reviewed by Dan Bernstein.
3414
3415         * DumpRenderTree/mac/DumpRenderTree.mm:
3416         (resetDefaultsToConsistentValues):
3417         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm:
3418         (TestWebKitAPI::InjectedBundleController::platformInitialize):
3419         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
3420         (WTR::InjectedBundle::platformInitialize):
3421         Update for new key.
3422
3423 2012-04-28  Zan Dobersek  <zandobersek@gmail.com>
3424
3425         [WK2] http/tests/navigation/anchor-frames-gbk.html fails
3426         https://bugs.webkit.org/show_bug.cgi?id=76896
3427
3428         Reviewed by Darin Adler.
3429
3430         Only dump as text if currently dumping render tree. This ensures that
3431         calling dumpAsText in a test with dumpChildFramesAsText already being
3432         called doesn't override the first decision, possibly making tests fail.
3433
3434         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3435         (WTR::LayoutTestController::dumpAsText):
3436
3437 2012-04-28  Dan Winship  <danw@gnome.org>
3438
3439         Unreviewed; Add me to the SoupNetwork watchlist
3440
3441         * Scripts/webkitpy/common/config/watchlist:
3442
3443 2012-04-27  Gustavo Noronha Silva  <gns@gnome.org>
3444
3445         [GTK] jhbuild cloning is not following WEBKITOUTPUTDIR.
3446         https://bugs.webkit.org/show_bug.cgi?id=76161
3447
3448         Reviewed by Martin Robinson.
3449
3450         * Scripts/webkitdirs.pm:
3451         (getJhbuildPath): New method to obtain the jhbuild base directory,
3452         using the product base directory
3453         (jhbuildConfigurationChanged): Use the new method
3454         (buildAutotoolsProject): Ditto.
3455         * efl/jhbuildrc: Use WEBKITOUTPUTDIR when calculating the path.
3456         * gtk/jhbuildrc: Ditto.
3457         * jhbuild/jhbuild-wrapper: Ditto.
3458
3459 2012-04-27  Scott Graham  <scottmg@chromium.org>
3460
3461         [chromium] make case of npTestNetscapePlugIn.dll match case in copy rule
3462         https://bugs.webkit.org/show_bug.cgi?id=85083
3463
3464         Reviewed by Dirk Pranke.
3465
3466         Make output product_name for npTestNetscapePlugIn.dll match the case
3467         of the copy_TestNetscapePlugIn rule. This is required for ninja, which
3468         is more particular about case matching in rules than the VS build.
3469
3470         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
3471
3472 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3473
3474         [chromium] use "drt-style" output, not "test-shell-style" output, on mac and linux DRT
3475         https://bugs.webkit.org/show_bug.cgi?id=84917
3476
3477         Unreviewed, build fix.
3478
3479         Reland the change in r115453 with a fix for chromium win.
3480         The logic in ChromiumDriver was busted on windows, causing DRT
3481         to get launched without the --test-shell flag.
3482
3483         * Scripts/webkitpy/layout_tests/port/chromium.py:
3484         (ChromiumDriver):
3485         (ChromiumDriver.__init__):
3486         (ChromiumDriver._wrapper_options):
3487         (ChromiumDriver.cmd_line):
3488         (ChromiumDriver._start):
3489         (ChromiumDriver.has_crashed):
3490         (ChromiumDriver.run_test):
3491         (ChromiumDriver.stop):
3492         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3493         (ChromiumDriverTest.setUp):
3494         (ChromiumDriverTest.test_stop):
3495         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
3496         (ChromiumDriverTest.test_two_drivers):
3497
3498 2012-04-27  Peter Beverloo  <peter@chromium.org>
3499
3500         Add John Grabowski as a non-committer to committers.py
3501         https://bugs.webkit.org/show_bug.cgi?id=85080
3502
3503         Reviewed by Dirk Pranke.
3504
3505         He's being CC'ed often enough because of his involvement in the Chromium
3506         side of Chrome for Android upstreaming.
3507
3508         * Scripts/webkitpy/common/config/committers.py:
3509           add jrg@chromium.org
3510
3511 2012-04-27  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
3512
3513         [Qt] Fix minimal build.
3514         https://bugs.webkit.org/show_bug.cgi?id=85045
3515
3516         Reviewed by Tor Arne Vestbø.
3517
3518         Do not set conflicting values of ENABLE_XSLT.
3519
3520         * qmake/mkspecs/features/features.prf:
3521
3522 2012-04-27  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
3523
3524         [EFL][DRT] EFL should sanitize pulseaudio module configuration too
3525         https://bugs.webkit.org/show_bug.cgi?id=85047
3526
3527         Reviewed by Dirk Pranke.
3528
3529         PulseAudio test setup sanitization shared between EFL and GTK.
3530
3531         * Scripts/webkitpy/layout_tests/port/efl.py:
3532         (EflPort):
3533         (EflPort.setup_test_run):
3534         (EflPort.clean_up_test_run):
3535         * Scripts/webkitpy/layout_tests/port/gtk.py:
3536         (GtkPort): Moving pulseaudio unload/restore out to shared class.
3537         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer.py: Added.
3538         (PulseAudioSanitizer):
3539         (PulseAudioSanitizer._unload_pulseaudio_module):
3540         (PulseAudioSanitizer._restore_pulseaudio_module):
3541
3542 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3543
3544         Unreviewed, revert r115453 - broke chromium win bots
3545
3546         * Scripts/webkitpy/layout_tests/port/chromium.py:
3547         (ChromiumPort._path_to_image_diff):
3548         (ChromiumDriver):
3549         (ChromiumDriver.__init__):
3550         (ChromiumDriver._wrapper_options):
3551         (ChromiumDriver.cmd_line):
3552         (ChromiumDriver.stop):
3553         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3554         (ChromiumDriverTest.setUp):
3555         (ChromiumDriverTest.test_stop):
3556         (ChromiumDriverTest.test_two_drivers):
3557         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
3558
3559 2012-04-27  Ryosuke Niwa  <rniwa@webkit.org>
3560
3561         Chromium Windows Perf bots timeout due to not output
3562         https://bugs.webkit.org/show_bug.cgi?id=84940
3563
3564         Reviewed by Dirk Pranke.
3565
3566         Dirk and I investigated the issue on the bot but we couldn't figure out what is going wrong.
3567         Since we run-perf-tests don't need any of fancy feature printer provides, just use python's
3568         built-in logging module instead. Printing out to stdout and stderr seem to work so hopefully
3569         this will fix the issue on the bot.
3570
3571         * Scripts/run-perf-tests:
3572         * Scripts/webkitpy/performance_tests/perftest.py:
3573         (PerfTest.run):
3574         (PerfTest.run_failed):
3575         (PerfTest.parse_output):
3576         (PerfTest.output_statistics):
3577         (ChromiumStylePerfTest.parse_output):
3578         (PageLoadingPerfTest.run):
3579         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
3580         (MainTest.test_parse_output):
3581         (MainTest.test_parse_output_with_failing_line):
3582         (TestPageLoadingPerfTest):
3583         (TestPageLoadingPerfTest.test_run):
3584         (TestPageLoadingPerfTest.test_run_with_bad_output):
3585         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3586         (PerfTestsRunner.__init__):
3587         (PerfTestsRunner._parse_args):
3588         (PerfTestsRunner.run):
3589         (PerfTestsRunner._upload_json):
3590         (PerfTestsRunner._print_status):
3591         (PerfTestsRunner._run_tests_set):
3592         (PerfTestsRunner._run_single_test):
3593         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3594         (create_runner):
3595         (test_run_test_set):
3596         (test_run_test_set_kills_drt_per_run):
3597         (test_run_test_pause_before_testing):
3598         (test_run_test_set_for_parser_tests):
3599         (test_run_test_set_with_json_output):
3600         (test_run_test_set_with_json_source):
3601         (test_run_test_set_with_multiple_repositories):
3602         (test_upload_json):
3603         (test_parse_args):
3604
3605 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3606
3607         [chromium] use "drt-style" output, not "test-shell-style" output, on mac and linux DRT
3608         https://bugs.webkit.org/show_bug.cgi?id=84917
3609
3610         Reviewed by Ojan Vafai.
3611
3612         This change changes the chromium new-run-webkit-tests
3613         implementation to use the code in WebKitDriver and ServerProcess
3614         by default on Mac and Linux, instead of using the old-style
3615         Chromium/test_shell style of output.
3616
3617         We will still use the test_shell style on Windows for now
3618         and you can still get the old style of behavior with
3619         new-run-webkit-tests --additional-drt-flag --test-shell
3620
3621         * Scripts/webkitpy/layout_tests/port/chromium.py:
3622         (ChromiumDriver):
3623         (ChromiumDriver.__init__):
3624         (ChromiumDriver._wrapper_options):
3625         (ChromiumDriver.cmd_line):
3626         (ChromiumDriver._start):
3627         (ChromiumDriver.has_crashed):
3628         (ChromiumDriver.run_test):
3629         (ChromiumDriver.stop):
3630         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3631         (ChromiumDriverTest.setUp):
3632         (ChromiumDriverTest.test_stop):
3633         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
3634         (ChromiumDriverTest.test_two_drivers):
3635
3636 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3637
3638         webkitpy: fix race in unit tests for manager_worker_broker
3639         https://bugs.webkit.org/show_bug.cgi?id=84925
3640
3641         Reviewed by Ojan Vafai.
3642
3643         I have been intermittently seeing hangs when running
3644         test-webkitpy on Mac SL; it appears that there's a bug (?) in
3645         the multiprocessing module where queue references are getting
3646         leaked or not cleaned up somehow. By explicitly closing the
3647         queues, things seem to be happy, so this patch adds a cleanup()
3648         method to the message broker and modifies the manager (and the
3649         unit tests) to call it. This may get rid of the intermittent
3650         error on shutdown on Linux I've seen as well.
3651
3652         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3653