[wx] Unreviewed build fixes for recent trunk changes and 32-bit ld *nix builds.
[WebKit-https.git] / Tools / ChangeLog
1 2012-05-11  Kevin Ollivier  <kevino@theolliviers.com>
2
3         [wx] Unreviewed build fix. Fix memory exhausted errors from 32-bit ld on *nix.
4
5         * waf/build/settings.py:
6         (common_configure):
7
8 2012-05-11  Dirk Pranke  <dpranke@chromium.org>
9
10         webkit-patch rebaseline should fetch stale expectations :)
11         https://bugs.webkit.org/show_bug.cgi?id=86231
12
13         Unreviewed, build fix.
14
15         This change reverts r115492 and goes back to fetching individual
16         baselines from the layout tests directory, rather than fetching
17         entire zip files. The performance hit over networks for people
18         far away from the bots has turned out to be severe, so we need a
19         better solution to this.
20
21         * Scripts/webkitpy/tool/commands/rebaseline.py:
22         (RebaselineTest._results_url):
23         (RebaselineTest._file_name_for_expected_result):
24         (RebaselineTest._rebaseline_test):
25         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
26         (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
27         (test_rebaseline_updates_expectations_file):
28         (test_rebaseline_test):
29         (test_rebaseline_and_copy_test):
30         (test_rebaseline_and_copy_test_no_existing_result):
31         (test_rebaseline_and_copy_test_with_lion_result):
32         (test_rebaseline_and_copy_no_overwrite_test):
33
34 2012-05-11  Christophe Dumez  <christophe.dumez@intel.com>
35
36         [GTK] LayoutTests/svg/batik/text/verticalText.svg fails to display Chinese characters
37         https://bugs.webkit.org/show_bug.cgi?id=86198
38
39         Reviewed by Martin Robinson.
40
41         Update webkitgtk-fonts package to v0.0.3 in order to get new Kochi
42         Mincho / Gothic fonts to display Japanese and Chinese characters.
43
44         * gtk/jhbuild.modules:
45
46 2012-05-11  Christophe Dumez  <christophe.dumez@intel.com>
47
48         [EFL][DRT] SVG Batik test fails to display Chinese characters
49         https://bugs.webkit.org/show_bug.cgi?id=84461
50
51         Reviewed by Martin Robinson.
52
53         Update webkitgtk-fonts package from v0.0.1 to v0.0.3 in order to get
54         new Kochi Mincho / Gothic fonts to display Japanese and Chinese
55         characters.
56
57         * DumpRenderTree/efl/CMakeLists.txt:
58         * efl/jhbuild.modules:
59
60 2012-05-11  Philippe Normand  <pnormand@igalia.com>
61
62         Unreviewed, GTK 64-bit Debug bot fix attempt.
63
64         * Scripts/webkitpy/layout_tests/port/gtk.py:
65         (GtkPort._get_gdb_output): Ignore unicode decoding errors when
66         processing gdb output.
67
68 2012-05-11  Gustavo Noronha Silva  <gns@gnome.org>
69
70         REGRESSION(r116205): [GTK]: build no longer supports thin archives
71         https://bugs.webkit.org/show_bug.cgi?id=86207
72
73         * GNUmakefile.am: link libWebCoreModules to DRT.
74
75 2012-05-10  Ojan Vafai  <ojan@chromium.org>
76
77         Add a perf tab to garden-o-matic
78         https://bugs.webkit.org/show_bug.cgi?id=86169
79
80         Reviewed by Dimitri Glazkov.
81
82         This is a first stab at making it easier to view the Chromium WebKit canary
83         perf graphs. For now, it's just a couple select elements and an iframe.
84
85         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
86         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
87         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
88         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
89         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
90         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
91         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf.js: Added.
92         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf_unittests.js: Added.
93         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
94         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/perf.css: Added.
95         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
96
97 2012-05-10  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
98
99         Move resumeAnimations to use Internals interface
100         https://bugs.webkit.org/show_bug.cgi?id=86063
101
102         Reviewed by Alexey Proskuryakov.
103
104         Remove resumeAnimations functions, because it is able to work in the 
105         cross-port way through the Internals interface.
106
107         * DumpRenderTree/LayoutTestController.cpp:
108         (LayoutTestController::staticFunctions):
109         * DumpRenderTree/LayoutTestController.h:
110         (LayoutTestController):
111         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
112         * DumpRenderTree/chromium/LayoutTestController.cpp:
113         (LayoutTestController::LayoutTestController):
114         * DumpRenderTree/chromium/LayoutTestController.h:
115         (LayoutTestController):
116         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
117         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
118         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
119         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
120         * DumpRenderTree/qt/LayoutTestControllerQt.h:
121         (LayoutTestController):
122         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
123         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
124         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
125         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
126         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
127         (LayoutTestController):
128
129 2012-05-10  Alexey Proskuryakov  <ap@apple.com>
130
131         Crash in 3rd party WebKit apps that disable cache at a wrong time
132         https://bugs.webkit.org/show_bug.cgi?id=86027
133         <rdar://problem/10615880>
134
135         Reviewed by Antti Koivisto.
136
137         Added a test that's very similar to MemoryCachePruneWithinResourceLoadDelegate,
138         but for disabling the cache instead of triggering a prune.
139
140         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
141         * TestWebKitAPI/Tests/mac/MemoryCacheDisableWithinResourceLoadDelegate.html: Added.
142         * TestWebKitAPI/Tests/mac/MemoryCacheDisableWithinResourceLoadDelegate.mm: Added.
143         (-[MemoryCacheDisableTestResourceLoadDelegate webView:identifierForInitialRequest:fromDataSource:]):
144         (-[MemoryCacheDisableTestResourceLoadDelegate webView:resource:willSendRequest:redirectResponse:fromDataSource:]):
145         (-[MemoryCacheDisableTestResourceLoadDelegate webView:resource:didFinishLoadingFromDataSource:]):
146         (-[MemoryCacheDisableTestResourceLoadDelegate webView:resource:didFailLoadingWithError:fromDataSource:]):
147         (TestWebKitAPI::TEST):
148
149 2012-05-10  Anders Carlsson  <andersca@apple.com>
150
151         WebKit2: Add a way to blacklist specific plug-ins/plug-in versions
152         https://bugs.webkit.org/show_bug.cgi?id=86164
153         <rdar://problem/9551196>
154
155         Reviewed by Sam Weinig.
156
157         Update for WK2 API changes.
158
159         * MiniBrowser/mac/BrowserWindowController.m:
160         (-[BrowserWindowController awakeFromNib]):
161         * WebKitTestRunner/TestController.cpp:
162         (WTR::TestController::createOtherPage):
163         (WTR::TestController::initialize):
164
165 2012-05-10  Dirk Pranke  <dpranke@chromium.org>
166
167         nrwt: [chromium] fix path to asan symbolize script
168         https://bugs.webkit.org/show_bug.cgi?id=86161
169
170         Reviewed by Abhishek Arya.
171
172         * Scripts/webkitpy/layout_tests/port/chromium.py:
173         (ChromiumPort._get_crash_log):
174
175 2012-05-10  Lei Zhang  <thestig@chromium.org>
176
177         [chromium] Set Valgrind environment variables for layout tests.
178         https://bugs.webkit.org/show_bug.cgi?id=86156
179
180         Reviewed by Dirk Pranke.
181
182         * Scripts/webkitpy/layout_tests/port/chromium.py:
183         (ChromiumPort.setup_environ_for_server):
184
185 2012-05-10  Raphael Kubo da Costa  <rakuco@webkit.org>
186
187         [webkitpy][EFL] It should be possible to use a wrapper command for ImageDiff.
188         https://bugs.webkit.org/show_bug.cgi?id=86145
189
190         Reviewed by Dirk Pranke.
191
192         The GTK+ and EFL ports use jhbuild to build a few base libraries
193         which are then used when building the port itself. Programs such
194         as DRT and ImageDiff are then supposed to be run with the
195         `run-with-jhbuild' wrapper so that the libraries are properly
196         found and loaded.
197
198         This was already the case for DumpRenderTree (EflPort defines the
199         'wrapper' option for that), but it was not possible to do the same
200         with ImageDiff so far.
201
202         The code which created the ImageDiff command-line call has been
203         moved to a separate method, _image_diff_command, which is then
204         overridden by EflPort to add the jhbuild wrapper.
205
206         * Scripts/webkitpy/layout_tests/port/efl.py:
207         (EflPort.__init__):
208         (EflPort._image_diff_command):
209         * Scripts/webkitpy/layout_tests/port/webkit.py:
210         (WebKitPort):
211         (WebKitPort._image_diff_command):
212         (WebKitPort._start_image_diff_process):
213
214 2012-05-10  Zan Dobersek  <zandobersek@gmail.com>
215
216         [GTK] ENABLE_IFRAME_SEAMLESS support
217         https://bugs.webkit.org/show_bug.cgi?id=85843
218
219         Reviewed by Eric Seidel.
220
221         Add the iframe-seamless option to the options list. This makes it
222         possible to enable the iframe seamless support through build-webkit
223         script. Enabled by default for all the ports.
224
225         * Scripts/webkitperl/FeatureList.pm:
226
227 2012-05-10  Raphael Kubo da Costa  <rakuco@webkit.org>
228
229         [EFL][webkitpy] Do not redefine check_build() in EflPort.
230         https://bugs.webkit.org/show_bug.cgi?id=86124
231
232         Reviewed by Dirk Pranke.
233
234         Remove an old FIXME and do not redefine check_build in EflPort;
235         once build-dumprendertree recognizes --efl, we can use the
236         check_build implementation in WebKitPort.
237
238         * Scripts/build-dumprendertree:
239         * Scripts/webkitpy/layout_tests/port/efl.py:
240         (EflPort.check_build):
241
242 2012-05-10  Tommy Widenflycht  <tommyw@google.com>
243
244         [chromium] MediaStream API: Fix the ExtraData functionality in WebMediaStreamDescriptor
245         https://bugs.webkit.org/show_bug.cgi?id=86087
246
247         Reviewed by Adam Barth.
248
249         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp:
250         (MockExtraData):
251         (WebKit):
252         (WebKit::WebUserMediaClientMock::requestUserMedia):
253
254 2012-05-10  Varun Jain  <varunjain@google.com>
255
256         [chromium] Trigger context menu for long press gesture
257         https://bugs.webkit.org/show_bug.cgi?id=85919
258
259         Reviewed by Adam Barth.
260
261         * DumpRenderTree/chromium/EventSender.cpp:
262         (EventSender::EventSender):
263         (EventSender::gestureLongPress):
264         (EventSender::gestureEvent):
265         * DumpRenderTree/chromium/EventSender.h:
266         (EventSender):
267
268 2012-05-10  Takashi Toyoshima  <toyoshim@chromium.org>
269
270         don't use autoinstall to import pywebsocket but check it in WebKit directly.
271         https://bugs.webkit.org/show_bug.cgi?id=86107
272
273         Reviewed by Adam Barth.
274
275         This change removes pywebsocket from webkitpy's autoinstall list and
276         imports pywebsocket 0.7.4 directly into webkitpy/thirparty.
277
278         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
279         (PyWebSocket._prepare_config):
280         * Scripts/webkitpy/thirdparty/__init__.py:
281         (AutoinstallImportHook.find_module):
282         (AutoinstallImportHook._install_irc):
283         * Scripts/webkitpy/thirdparty/__init___unittest.py:
284         (ThirdpartyTest.test_import_hook.MockImportHook.__init__):
285         (ThirdpartyTest.test_import_hook.MockImportHook._install_eliza):
286         (ThirdpartyTest):
287         (ThirdpartyTest.test_import_hook):
288         * Scripts/webkitpy/thirdparty/mod_pywebsocket/COPYING: Added.
289         * Scripts/webkitpy/thirdparty/mod_pywebsocket/__init__.py: Added.
290         * Scripts/webkitpy/thirdparty/mod_pywebsocket/_stream_base.py: Added.
291         (ConnectionTerminatedException):
292         (InvalidFrameException):
293         (BadOperationException):
294         (UnsupportedFrameException):
295         (InvalidUTF8Exception):
296         (StreamBase):
297         (StreamBase.__init__):
298         (StreamBase._read):
299         (StreamBase._write):
300         (StreamBase.receive_bytes):
301         (StreamBase._read_until):
302         * Scripts/webkitpy/thirdparty/mod_pywebsocket/_stream_hixie75.py: Added.
303         (for):
304         (StreamHixie75):
305         (StreamHixie75.for):
306         (StreamHixie75.__init__):
307         (StreamHixie75.send_message):
308         (StreamHixie75._read_payload_length_hixie75):
309         (StreamHixie75.receive_message):
310         (StreamHixie75._send_closing_handshake):
311         (StreamHixie75.close_connection):
312         (StreamHixie75.send_ping):
313         * Scripts/webkitpy/thirdparty/mod_pywebsocket/_stream_hybi.py: Added.
314         (Frame):
315         (Frame.__init__):
316         (create_length_header):
317         (create_header):
318         (_build_frame):
319         (_filter_and_format_frame_object):
320         (create_binary_frame):
321         (create_text_frame):
322         (FragmentedFrameBuilder):
323         (FragmentedFrameBuilder.to):
324         (FragmentedFrameBuilder.__init__):
325         (FragmentedFrameBuilder.build):
326         (_create_control_frame):
327         (create_ping_frame):
328         (create_pong_frame):
329         (create_close_frame):
330         (StreamOptions):
331         (StreamOptions.__init__):
332         (Stream):
333         (Stream.for):
334         (Stream.__init__):
335         (Stream._receive_frame):
336         (Stream._receive_frame_as_frame_object):
337         (Stream.send_message):
338         (Stream.receive_message):
339         (Stream._send_closing_handshake):
340         (Stream.close_connection):
341         (Stream.send_ping):
342         (Stream._send_pong):
343         (Stream._drain_received_data):
344         * Scripts/webkitpy/thirdparty/mod_pywebsocket/common.py: Added.
345         (is_control_opcode):
346         (ExtensionParameter):
347         (ExtensionParameter.__init__):
348         (ExtensionParameter.name):
349         (ExtensionParameter.add_parameter):
350         (ExtensionParameter.get_parameters):
351         (ExtensionParameter.get_parameter_names):
352         (ExtensionParameter.has_parameter):
353         (ExtensionParameter.get_parameter_value):
354         * Scripts/webkitpy/thirdparty/mod_pywebsocket/dispatch.py: Added.
355         (DispatchException):
356         (DispatchException.__init__):
357         (_default_passive_closing_handshake_handler):
358         (_normalize_path):
359         (_create_path_to_resource_converter):
360         (_create_path_to_resource_converter.converter):
361         (_enumerate_handler_file_paths):
362         (_HandlerSuite):
363         (_HandlerSuite.__init__):
364         (_source_handler_file):
365         (_extract_handler):
366         (Dispatcher):
367         (Dispatcher.maintains):
368         (Dispatcher.__init__):
369         (Dispatcher.add_resource_path_alias):
370         (Dispatcher.source_warnings):
371         (Dispatcher.do_extra_handshake):
372         (Dispatcher.transfer_data):
373         (Dispatcher.passive_closing_handshake):
374         (Dispatcher.get_handler_suite):
375         (Dispatcher._source_handler_files_in_dir):
376         * Scripts/webkitpy/thirdparty/mod_pywebsocket/extensions.py: Added.
377         (ExtensionProcessorInterface):
378         (ExtensionProcessorInterface.get_extension_response):
379         (ExtensionProcessorInterface.setup_stream_options):
380         (DeflateStreamExtensionProcessor):
381         (DeflateStreamExtensionProcessor.__init__):
382         (DeflateStreamExtensionProcessor.get_extension_response):
383         (DeflateStreamExtensionProcessor.setup_stream_options):
384         (DeflateFrameExtensionProcessor):
385         (DeflateFrameExtensionProcessor.__init__):
386         (DeflateFrameExtensionProcessor.get_extension_response):
387         (DeflateFrameExtensionProcessor.setup_stream_options):
388         (DeflateFrameExtensionProcessor.setup_stream_options._OutgoingFilter):
389         (DeflateFrameExtensionProcessor.setup_stream_options._OutgoingFilter.__init__):
390         (DeflateFrameExtensionProcessor.setup_stream_options._OutgoingFilter.filter):
391         (DeflateFrameExtensionProcessor.setup_stream_options._IncomingFilter):
392         (DeflateFrameExtensionProcessor.setup_stream_options._IncomingFilter.__init__):
393         (DeflateFrameExtensionProcessor.setup_stream_options._IncomingFilter.filter):
394         (DeflateFrameExtensionProcessor.set_response_window_bits):
395         (DeflateFrameExtensionProcessor.set_response_no_context_takeover):
396         (DeflateFrameExtensionProcessor.enable_outgoing_compression):
397         (DeflateFrameExtensionProcessor.disable_outgoing_compression):
398         (DeflateFrameExtensionProcessor._outgoing_filter):
399         (DeflateFrameExtensionProcessor._incoming_filter):
400         (get_extension_processor):
401         (get_extension_processor.is):
402         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/__init__.py: Added.
403         (try):
404         (do_handshake):
405         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/_base.py: Added.
406         (AbortedByUserException):
407         (HandshakeException):
408         (HandshakeException.__init__):
409         (VersionException):
410         (VersionException.__init__):
411         (get_default_port):
412         (validate_subprotocol):
413         (parse_host_header):
414         (format_header):
415         (build_location):
416         (get_mandatory_header):
417         (validate_mandatory_header):
418         (check_request_line):
419         (check_header_lines):
420         (parse_token_list):
421         (_parse_extension_param):
422         (_parse_extension):
423         (parse_extensions):
424         (format_extensions):
425         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/draft75.py: Added.
426         (Handshaker):
427         (Handshaker.performs):
428         (Handshaker.__init__):
429         (Handshaker.do_handshake):
430         (Handshaker._set_resource):
431         (Handshaker._set_origin):
432         (Handshaker._set_location):
433         (Handshaker._set_subprotocol):
434         (Handshaker._set_protocol_version):
435         (Handshaker._sendall):
436         (Handshaker._send_handshake):
437         (Handshaker._check_header_lines):
438         (Handshaker._check_first_lines):
439         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/hybi.py: Added.
440         (compute_accept):
441         (Handshaker):
442         (Handshaker.__init__):
443         (Handshaker._validate_connection_header):
444         (Handshaker.do_handshake):
445         (Handshaker._get_origin):
446         (Handshaker._check_version):
447         (Handshaker._set_protocol):
448         (Handshaker._parse_extensions):
449         (Handshaker._validate_key):
450         (Handshaker._get_key):
451         (Handshaker._send_handshake):
452         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/hybi00.py: Added.
453         (Handshaker):
454         (Handshaker.__init__):
455         (Handshaker.do_handshake):
456         (Handshaker._set_resource):
457         (Handshaker._set_subprotocol):
458         (Handshaker._set_location):
459         (Handshaker._set_origin):
460         (Handshaker._set_protocol_version):
461         (Handshaker._set_challenge_response):
462         (Handshaker._get_key_value):
463         (Handshaker._get_challenge):
464         (Handshaker._send_handshake):
465         * Scripts/webkitpy/thirdparty/mod_pywebsocket/headerparserhandler.py: Added.
466         (ApacheLogHandler):
467         (ApacheLogHandler.__init__):
468         (ApacheLogHandler.emit):
469         (_configure_logging):
470         (_parse_option):
471         (_create_dispatcher):
472         (headerparserhandler):
473         * Scripts/webkitpy/thirdparty/mod_pywebsocket/http_header_util.py: Added.
474         (_is_char):
475         (_is_ctl):
476         (ParsingState):
477         (ParsingState.__init__):
478         (peek):
479         (consume):
480         (consume_string):
481         (consume_lws):
482         (consume_lwses):
483         (consume_token):
484         (consume_token_or_quoted_string):
485         (quote_if_necessary):
486         (parse_uri):
487         * Scripts/webkitpy/thirdparty/mod_pywebsocket/memorizingfile.py: Added.
488         (MemorizingFile):
489         (MemorizingFile.__init__):
490         (MemorizingFile.__getattribute__):
491         (MemorizingFile.readline):
492         (MemorizingFile.get_memorized_lines):
493         * Scripts/webkitpy/thirdparty/mod_pywebsocket/msgutil.py: Added.
494         (close_connection):
495         (send_message):
496         (receive_message):
497         (send_ping):
498         (MessageReceiver):
499         (MessageReceiver.receives):
500         (MessageReceiver.provides):
501         (MessageReceiver.should):
502         (MessageReceiver.__init__):
503         (MessageReceiver.run):
504         (MessageReceiver.receive):
505         (MessageReceiver.receive_nowait):
506         (MessageReceiver.stop):
507         (MessageSender):
508         (MessageSender.sends):
509         (MessageSender.provides):
510         (MessageSender.should):
511         (MessageSender.__init__):
512         (MessageSender.run):
513         (MessageSender.send):
514         (MessageSender.send_nowait):
515         * Scripts/webkitpy/thirdparty/mod_pywebsocket/standalone.py: Added.
516         (_StandaloneConnection):
517         (_StandaloneConnection.__init__):
518         (_StandaloneConnection.get_local_addr):
519         (_StandaloneConnection.get_remote_addr):
520         (_StandaloneConnection.write):
521         (_StandaloneConnection.read):
522         (_StandaloneConnection.get_memorized_lines):
523         (_StandaloneRequest):
524         (_StandaloneRequest.__init__):
525         (_StandaloneRequest.get_uri):
526         (_StandaloneRequest.get_method):
527         (_StandaloneRequest.is_https):
528         (_StandaloneRequest._drain_received_data):
529         (_StandaloneSSLConnection):
530         (_StandaloneSSLConnection.for):
531         (_StandaloneSSLConnection.__init__):
532         (_StandaloneSSLConnection.__getattribute__):
533         (_StandaloneSSLConnection.__setattr__):
534         (_StandaloneSSLConnection.makefile):
535         (WebSocketServer):
536         (WebSocketServer.__init__):
537         (WebSocketServer._create_sockets):
538         (WebSocketServer.server_bind):
539         (WebSocketServer.server_activate):
540         (WebSocketServer.server_close):
541         (WebSocketServer.fileno):
542         (WebSocketServer.handle_error):
543         (WebSocketServer.get_request):
544         (WebSocketServer.serve_forever):
545         (WebSocketServer.shutdown):
546         (WebSocketRequestHandler):
547         (WebSocketRequestHandler.setup):
548         (WebSocketRequestHandler.setup.SocketServer):
549         (WebSocketRequestHandler.__init__):
550         (WebSocketRequestHandler.parse_request):
551         (WebSocketRequestHandler.log_request):
552         (WebSocketRequestHandler.log_error):
553         (WebSocketRequestHandler.is_cgi):
554         (_configure_logging):
555         (_alias_handlers):
556         (_build_option_parser):
557         (ThreadMonitor):
558         (ThreadMonitor.__init__):
559         (ThreadMonitor.run):
560         (_parse_args_and_config):
561         (_main):
562         (_main.if):
563         (_main.if.__check_script):
564         * Scripts/webkitpy/thirdparty/mod_pywebsocket/stream.py: Copied from Tools/Scripts/webkitpy/thirdparty/__init___unittest.py.
565         * Scripts/webkitpy/thirdparty/mod_pywebsocket/util.py: Added.
566         (get_stack_trace):
567         (prepend_message_to_exception):
568         (__translate_interp):
569         (get_script_interp):
570         (wrap_popen3_for_win):
571         (wrap_popen3_for_win.__wrap_popen3):
572         (hexify):
573         (get_class_logger):
574         (NoopMasker):
575         (NoopMasker.__init__):
576         (NoopMasker.mask):
577         (RepeatedXorMasker):
578         (RepeatedXorMasker.__init__):
579         (RepeatedXorMasker.mask):
580         (DeflateRequest):
581         (DeflateRequest.for):
582         (DeflateRequest.__init__):
583         (DeflateRequest.__getattribute__):
584         (DeflateRequest.__setattr__):
585         (_Deflater):
586         (_Deflater.__init__):
587         (_Deflater.compress_and_flush):
588         (_Inflater):
589         (_Inflater.__init__):
590         (_Inflater.decompress):
591         (_Inflater.append):
592         (_Inflater.reset):
593         (_RFC1979Deflater):
594         (_RFC1979Deflater.that):
595         (_RFC1979Deflater.__init__):
596         (_RFC1979Deflater.filter):
597         (_RFC1979Inflater):
598         (_RFC1979Inflater.for):
599         (_RFC1979Inflater.__init__):
600         (_RFC1979Inflater.filter):
601         (DeflateSocket):
602         (DeflateSocket.for):
603         (DeflateSocket.__init__):
604         (DeflateSocket.recv):
605         (DeflateSocket.sendall):
606         (DeflateSocket.send):
607         (DeflateConnection):
608         (DeflateConnection.for):
609         (DeflateConnection.__init__):
610         (DeflateConnection.get_remote_addr):
611         (DeflateConnection.put_bytes):
612         (DeflateConnection.read):
613         (DeflateConnection.write):
614         (_is_ewouldblock_errno):
615         (drain_received_data):
616
617 2012-05-10  Raphael Kubo da Costa  <rakuco@webkit.org>
618
619         [EFL][webkitpy] Define the 'wrapper' option in EflPort instead of creating a EflDriver.
620         https://bugs.webkit.org/show_bug.cgi?id=86117
621
622         Reviewed by Adam Barth.
623
624         Simplify the code used to wrap driver calls with jhbuild: we can
625         just define the 'wrapper' option instead of having to create a
626         Driver class just for that.
627
628         * Scripts/webkitpy/layout_tests/port/efl.py:
629         (EflPort.__init__):
630         (EflPort._port_flag_for_scripts):
631
632 2012-05-10  Sheriff Bot  <webkit.review.bot@gmail.com>
633
634         Unreviewed, rolling out r116633.
635         http://trac.webkit.org/changeset/116633
636         https://bugs.webkit.org/show_bug.cgi?id=86093
637
638         Breaks EFL tests (Requested by kov on #webkit).
639
640         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
641         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
642         * DumpRenderTree/efl/EventSender.cpp:
643         (keyDownCallback):
644         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
645         (LayoutTestController::setTabKeyCyclesThroughElements):
646
647 2012-05-10  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
648
649         [EFL] Memory leaks found in EFL DRT
650         https://bugs.webkit.org/show_bug.cgi?id=85889
651
652         Reviewed by Gustavo Noronha Silva.
653
654         Fix two memory leaks in EFL's DRT by using JSRetainPtr for JSString
655         created with JSStringCreateWithUTF8CString.
656
657         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
658         (DumpRenderTreeChrome::onWindowObjectCleared):
659         * DumpRenderTree/efl/EventSender.cpp:
660         (modifiersFromJSValue):
661
662 2012-05-10  Alexander Shalamov  <alexander.shalamov@intel.com>
663
664         [EFL] LayoutTestController needs implementation of setTabKeyCyclesThroughElements
665         https://bugs.webkit.org/show_bug.cgi?id=82864
666
667         Reviewed by Gustavo Noronha Silva.
668
669         Add missing implementation for setTabKeyCyclesThroughElements to EFL's LayoutTestController
670         in order to unskip fast/events/keypress-insert-tab.html
671
672         * DumpRenderTree/efl/EventSender.cpp:
673         (keyDownCallback):
674         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
675         (LayoutTestController::setTabKeyCyclesThroughElements):
676
677 2012-05-10  Alexander Shalamov  <alexander.shalamov@intel.com>
678
679         [EFL][DRT] EventSender - set delayed event information if it was not set before
680         https://bugs.webkit.org/show_bug.cgi?id=85972
681
682         Reviewed by Gustavo Noronha Silva.
683
684         Reset delayed event information if it is missing.
685
686         * DumpRenderTree/efl/EventSender.cpp:
687         (feedOrQueueMouseEvent):
688
689 2012-05-07  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
690
691         WebPageProxy::activeURL() should return the pending API request, even when there's no main frame
692         https://bugs.webkit.org/show_bug.cgi?id=85806
693
694         The m_pendingAPIRequestURL member is used (presumably) to mask over the async
695         nature of WebKit2, so that starting a load of a URL will reflect that URL
696         immedeatly from activeURL, even if the request has not been passed over to
697         the web process yet and reflected there.
698
699         This works well, except in the case of the initial request, where the main
700         frame creation happens on the web process side and is notified back to the
701         UI process. Until we've recived the notification we don't know about the main
702         frame, and this race condition will potentially give us an empty url instead
703         of the pending request.
704
705         To solve this we always return the pending API request if it's set, even
706         when there's no mainframe yet (that we known about).
707
708         Reviewed by Simon Hausmann.
709
710         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
711
712 2012-05-10  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
713
714         [EFL][DRT] Clear added user style sheets before a new testcase execution.
715         https://bugs.webkit.org/show_bug.cgi?id=84792
716
717         Reviewed Gustavo Noronha Silva.
718
719         The added user style sheets from the previous testcases are cleared in order to avoid 
720         flaky results.
721
722         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
723         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
724
725 2012-05-09  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
726
727         Move suspendAnimations to use Internals interface.
728         https://bugs.webkit.org/show_bug.cgi?id=85986
729
730         Reviewed by Ryosuke Niwa.
731
732         Remove suspendAnimations functions, because it is able to work in the 
733         cross-port way through the Internals interface
734
735         * DumpRenderTree/LayoutTestController.cpp:
736         (LayoutTestController::staticFunctions):
737         * DumpRenderTree/LayoutTestController.h:
738         (LayoutTestController):
739         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
740         * DumpRenderTree/chromium/LayoutTestController.cpp:
741         (LayoutTestController::LayoutTestController):
742         * DumpRenderTree/chromium/LayoutTestController.h:
743         (LayoutTestController):
744         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
745         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
746         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
747         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
748         * DumpRenderTree/qt/LayoutTestControllerQt.h:
749         (LayoutTestController):
750         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
751         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
752
753 2012-05-09  Kent Tamura  <tkent@chromium.org>
754
755         [Chromium] Attempt to fix an Android build error.
756         https://bugs.webkit.org/show_bug.cgi?id=86045
757
758         * DumpRenderTree/DumpRenderTree.gypi:
759
760 2012-05-09  Ojan Vafai  <ojan@chromium.org>
761
762         Don't show partytime if there are non-layout test failures.
763         https://bugs.webkit.org/show_bug.cgi?id=86056
764
765         Reviewed by Dimitri Glazkov.
766
767         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
768         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers_unittests.js:
769         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
770
771 2012-05-09  Hayato Ito  <hayato@chromium.org>
772
773         Improve a warning message for a reftest having an unused expectation file.
774         https://bugs.webkit.org/show_bug.cgi?id=85975
775
776         Reviewed by Adam Barth.
777
778         The current warning message is not intuitive. It might be better to update the message
779         so that it encourage users to remove unused expectation files.
780
781         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
782         (SingleTestRunner.__init__):
783
784 2012-05-09  Ojan Vafai  <ojan@chromium.org>
785
786         Change "losing test coverage" to cover all non-layout test problems in garden-o-matic
787         https://bugs.webkit.org/show_bug.cgi?id=86043
788
789         Reviewed by Adam Barth.
790
791         -Get rid of the gtest iframe.
792         -Cover all non-layout test problems. Before we wouldn't catch compile failures
793         or gtest failures since they were not on the webkit testing bots.
794         -Include the list of failing steps next to the builder name. This is a lot less
795         cluttered than you'd expect and already helped me find a compile failure early.
796
797         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
798         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
799         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
800         Delete unused map.
801         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
802         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
803         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
804         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js:
805         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
806         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
807         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
808         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
809
810 2012-05-09  Jessie Berlin  <jberlin@apple.com>
811
812         Crash using the new WKBundleDOMWindowExtensions APIs.
813         https://bugs.webkit.org/show_bug.cgi?id=85888
814
815         Reviewed by Brady Eidson.
816
817         Cached frames can live slightly longer than the page, but most clients unregister themselves
818         and do other cleanup in the willDestroyPage callback, making them miss the
819         willDestroyGlobalObjectForDOMWindowExtension callbacks.
820
821         The calls to willDestroyGlobalObjectForDOMWindowExtension in the DOMWindowExtensionBasic
822         test were all being invoked underneath WebPage::close. This is unrealistic. Update that test
823         to destroy the BundleDOMWindowExtensions in response to the willDestroyPage callback.
824
825         Add a test to verify that willDestroyGlobalObjectForDOMWindowExtension is being called for
826         pages that don't go into the page cache.
827
828         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
829         Add DOMWindowExtensionNoCache.cpp, DOMWindowExtensionNoCache_Bundle.cpp, simple-unload.html
830         and simple-iframe-unload.html
831
832         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic.cpp:
833         (TestWebKitAPI):
834         Remove the expected messages for willDestroyGlobalObjectForDOMWindowExtension.
835         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
836         Do not bother to keep track of the live extension count - all of them are expected to be
837         live until the test completes.
838         (TestWebKitAPI::TEST):
839         Fix the calls to EXPECT to pass the expected value first, and use EXPECT_WK_STREQ so that
840         message failures will be clearer.
841         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
842         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyPage):
843         Clean up the BundleDOMWindowExtensions.
844         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyGlobalObjectForDOMWindowExtension):
845         Add an ASSERT_NOT_REACHED.
846
847         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache.cpp: Added.
848         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
849         Keep track of the messages received so they can be checked at the end of the test.
850         (TestWebKitAPI::TEST):
851         Navigate to uncacheable pages and back.
852         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache_Bundle.cpp: Copied from Tools/TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp.
853         (DOMWindowExtensionNoCache):
854         (TestWebKitAPI::DOMWindowExtensionNoCache::DOMWindowExtensionNoCache):
855         Set up all the states for each BundleDOMWindowExtension.
856         (TestWebKitAPI::DOMWindowExtensionNoCache::frameLoadFinished):
857         Tell the UI Process about the states of the BundleDOMWindowExtensions.
858         (TestWebKitAPI::DOMWindowExtensionNoCache::sendExtensionStateMessage):
859         (TestWebKitAPI::DOMWindowExtensionNoCache::initialize):
860         (TestWebKitAPI::DOMWindowExtensionNoCache::didCreatePage):
861         (TestWebKitAPI::DOMWindowExtensionNoCache::willDestroyPage):
862         Remvoe the remaining BundleDOMWindowExtensions, send the updated state, and finish the test.
863         (TestWebKitAPI::DOMWindowExtensionNoCache::updateExtensionStateRecord):
864         (TestWebKitAPI::DOMWindowExtensionNoCache::sendBundleMessage):
865         (TestWebKitAPI::DOMWindowExtensionNoCache::globalObjectIsAvailableForFrame):
866         (TestWebKitAPI::DOMWindowExtensionNoCache::willDisconnectDOMWindowExtensionFromGlobalObject):
867         ASSERT that these pages not going into the page cache are not getting disconnected to go into
868         the page cache.
869         (TestWebKitAPI::DOMWindowExtensionNoCache::didReconnectDOMWindowExtensionToGlobalObject):
870         Ditto about getting reconnected when coming out of the page cache.
871         (TestWebKitAPI::DOMWindowExtensionNoCache::willDestroyGlobalObjectForDOMWindowExtension):
872         Tell the UI Process, update the state, and get rid of the BundleDOMWindowExtension.
873         (TestWebKitAPI::didFinishLoadForFrameCallback):
874         (TestWebKitAPI::globalObjectIsAvailableForFrameCallback):
875         (TestWebKitAPI::willDisconnectDOMWindowExtensionFromGlobalObjectCallback):
876         (TestWebKitAPI::didReconnectDOMWindowExtensionToGlobalObjectCallback):
877         (TestWebKitAPI::willDestroyGlobalObjectForDOMWindowExtensionCallback):
878
879         * TestWebKitAPI/Tests/WebKit2/simple-iframe-unload.html: Added.
880         * TestWebKitAPI/Tests/WebKit2/simple-unload.html: Added.
881
882 2012-05-09  Adam Barth  <abarth@webkit.org>
883
884         [Chromium] Android's DRT doesn't want to build WebUserMediaClientMock
885         https://bugs.webkit.org/show_bug.cgi?id=86045
886
887         Reviewed by Eric Seidel.
888
889         * DumpRenderTree/DumpRenderTree.gypi:
890
891 2012-05-09  Adam Barth  <abarth@webkit.org>
892
893         [Chromium] MockWebSpeechInputController.cpp should be guarded by ENABLE(INPUT_SPEECH)
894         https://bugs.webkit.org/show_bug.cgi?id=86041
895
896         Reviewed by Eric Seidel.
897
898         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp:
899
900 2012-05-09  Adam Barth  <abarth@webkit.org>
901
902         [Chromium] Android needs to swizzle pixel dumps in order to get the same hash as other platforms
903         https://bugs.webkit.org/show_bug.cgi?id=86040
904
905         Reviewed by Kent Tamura.
906
907         * DumpRenderTree/chromium/TestShell.cpp:
908         (TestShell::dumpImage):
909
910 2012-05-09  Raphael Kubo da Costa  <rakuco@webkit.org>
911
912         [EFL] Build fix after r116464.
913
914         `ulong' is a type defined by glibc, so it is not very
915         portable. Use `unsigned long' instead.
916
917         * DumpRenderTree/efl/EventSender.cpp:
918         (DelayedEvent::DelayedEvent):
919         (DelayedEvent):
920         (leapForwardCallback):
921
922 2012-05-09  Adam Barth  <abarth@webkit.org>
923
924         [Chromium] LayoutTestController should use ENABLE guards for some features
925         https://bugs.webkit.org/show_bug.cgi?id=86030
926
927         Reviewed by Eric Seidel.
928
929         These features are disabled on OS(ANDROID) and so shouldn't be compiled
930         when the feature is off.
931
932         * DumpRenderTree/chromium/LayoutTestController.cpp:
933         (LayoutTestController::LayoutTestController):
934         * DumpRenderTree/chromium/LayoutTestController.h:
935         (LayoutTestController):
936         * DumpRenderTree/chromium/TestShell.h:
937         (TestShell):
938
939 2012-05-09  Tom Zakrajsek  <tomz@codeaurora.org>
940
941         Fix overzealous re-opened since blocked by... message
942         https://bugs.webkit.org/show_bug.cgi?id=86020
943
944         Reviewed by Adam Barth.
945
946         Changed CreateBug behavior to only re-open a blocked bug if it
947         was RESOLVED.  Also had to update the test infrastructure slightly
948         to allow for distinct mock commit revisions, so one could point
949         to a RESOLVED bug for testing while another pointed to an OPEN bug.
950
951         Also moved default created mock bug to be 60001 instead of 50004
952         which already exists as a reference mock bug.
953
954         * Scripts/webkitpy/common/checkout/checkout_mock.py:
955         (MockCheckout.__init__):
956         (MockCheckout.commit_info_for_revision):
957         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
958         (MockBugzilla.create_bug):
959         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
960         * Scripts/webkitpy/tool/commands/download_unittest.py:
961         (test_create_rollout_resolved):
962         * Scripts/webkitpy/tool/commands/queries_unittest.py:
963         (FailureReasonTest.test_blame_line_for_revision):
964         * Scripts/webkitpy/tool/steps/createbug.py:
965         (CreateBug.run):
966
967 2012-05-09  Hugo Parente Lima  <hugo.lima@openbossa.org>
968
969         Use suitable viewport values on XHTML-MP pages.
970         https://bugs.webkit.org/show_bug.cgi?id=85425
971
972         Reviewed by Kenneth Rohde Christiansen.
973
974         Enable LEGACY_VIEWPORT_ADAPTION on Qt port.
975
976         * qmake/mkspecs/features/features.prf:
977
978 2012-05-09  Jochen Eisinger  <jochen@chromium.org>
979
980         When creating a new page during a navigation, prime the initial document with the correct referrer policy
981         https://bugs.webkit.org/show_bug.cgi?id=86001
982
983         Reviewed by Adam Barth.
984
985         Store the referrer policy in the request, so the network stack can comply with it.
986
987         * DumpRenderTree/chromium/WebViewHost.cpp:
988         (WebViewHost::willSendRequest):
989
990 2012-05-09  Vivek Galatage  <vivekgalatage@gmail.com>
991
992         Remove CYGWIN=tty from environment variable as its no longer supported
993         https://bugs.webkit.org/show_bug.cgi?id=85791
994
995         The CYGWIN=tty environment variable is no longer supported from Cygwin
996         version 1.7.10. This fix avoids addition of this environment variable
997         for Cygwin versions greater than 1.7.9. For supporting older versions
998         i.e. 1.7.9 or older, the script takes care of retaining the environment
999         variable.
1000
1001         Reviewed by Adam Roben.
1002
1003         * Scripts/webkitdirs.pm:
1004         (setupAppleWinEnv):
1005
1006 2012-05-09  Dirk Pranke  <dpranke@chromium.org>
1007
1008         nrwt: --skip-failing-tests should also skip flaky tests
1009         https://bugs.webkit.org/show_bug.cgi?id=85941
1010
1011         Reviewed by Ojan Vafai.
1012
1013         This changes --skip-failing-tests so that we skip tests that
1014         are believed to pass sometimes as well as tests that are
1015         believed to always fail.
1016
1017         In addition, in testing this I discovered that we had lousy
1018         test coverage of the details of retrying test failures and
1019         --clobber-old-results, so I wrote some more tests.
1020
1021         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1022         (Manager.prepare_lists_and_print_output):
1023         * Scripts/webkitpy/layout_tests/port/test.py:
1024         (unit_test_list):
1025         (TestPort.__init__):
1026         (TestDriver.run_test):
1027         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1028         (MainTest.test_skip_failing_tests):
1029         (MainTest.test_retrying_and_flaky_tests):
1030
1031 2012-05-09  Christophe Dumez  <christophe.dumez@intel.com>
1032
1033         [EFL][DRT] EFL's DRT needs to support LayoutTestController.dumpIconChanges()
1034         https://bugs.webkit.org/show_bug.cgi?id=84435
1035
1036         Reviewed by Antonio Gomes.
1037
1038         Catch the "icon,changed" signal on the frames in EFL's DumpRenderTree
1039         to support LayoutTestController's dumpIconChanges().
1040
1041         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1042         (DumpRenderTreeChrome::createView):
1043         (DumpRenderTreeChrome::onFrameIconChanged):
1044         (DumpRenderTreeChrome::onFrameCreated):
1045         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
1046         (DumpRenderTreeChrome):
1047
1048 2012-05-09  Dominik Röttsches  <dominik.rottsches@intel.com>
1049
1050         webkit-patch upload Fails Behind Load-balancing Proxies
1051         https://bugs.webkit.org/show_bug.cgi?id=85980
1052
1053         Reviewed by Adam Barth.
1054
1055         Unchecking the login restriction to IP to allow the tool
1056         to function behind load-balancing proxies that use multiple
1057         outgoing IPs.
1058
1059         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
1060         (Bugzilla.authenticate):
1061
1062 2012-05-09  Hao Zheng  <zhenghao@chromium.org>
1063
1064         [Chromium-android] Add unit test for Android layout test script (chromium_android.py)
1065         https://bugs.webkit.org/show_bug.cgi?id=80852
1066
1067         Reviewed by Adam Barth.
1068
1069         Add some tests to ensure that chromium android port works to some
1070         extent. Some methods is hard to test, as they rely on the actual
1071         behavior on device, e.g. pushing files to device.
1072
1073         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1074         (ChromiumAndroidPort.get_last_stacktrace): Check if dir exists.
1075         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1076         (ChromiumAndroidPortTest):
1077         (ChromiumAndroidPortTest.test_attributes):
1078         (ChromiumAndroidPortTest.test_get_last_stacktrace):
1079         (ChromiumAndroidPortTest.test_get_last_stacktrace.mock_run_command_no_dir):
1080         (ChromiumAndroidPortTest.test_get_last_stacktrace.mock_run_command_no_file):
1081         (ChromiumAndroidPortTest.test_get_last_stacktrace.mock_run_command_non_empty):
1082         (ChromiumAndroidDriverTest):
1083         (ChromiumAndroidDriverTest.setUp):
1084         (ChromiumAndroidDriverTest.test_get_drt_return_value):
1085         (ChromiumAndroidDriverTest.test_has_crash_hint):
1086         (ChromiumAndroidDriverTest.test_read_prompt):
1087         (ChromiumAndroidDriverTest.test_test_shell_command):
1088         (ChromiumAndroidDriverTest.test_write_command_and_read_line):
1089
1090 2012-05-09  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
1091
1092         [Qt] ORWT fails with Qt 5 and WK2.
1093         https://bugs.webkit.org/show_bug.cgi?id=85968
1094
1095         Reviewed by Csaba Osztrogonác.
1096
1097         The Qt5 binary does not have a -style argument, so avoid using it.
1098
1099         * Scripts/old-run-webkit-tests:
1100
1101 2012-05-09  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1102
1103         [EFL][DRT] LayoutTestController addUserScript implementation
1104         https://bugs.webkit.org/show_bug.cgi?id=84718
1105
1106         Reviewed by Antonio Gomes.
1107
1108         Added implementation of addUserScript() to LayoutTestController.
1109
1110         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1111         (DumpRenderTreeChrome::resetDefaultsToConsistentValues): clear added user scripts
1112         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1113         (LayoutTestController::addUserScript): added implementation
1114
1115 2012-05-09  Dan Bernstein  <mitz@apple.com>
1116
1117         Made build-webkit copy WebKitLibraries to the right location again after r116257
1118
1119         Reviewed by Mark Rowe.
1120
1121         * Scripts/build-webkit: Pass the built product directory to copy-webkitlibraries-to-product-directory.
1122         * Scripts/copy-webkitlibraries-to-product-directory: Use the first argument, if present, as the
1123         product directory.
1124
1125 2012-05-08  Ojan Vafai  <ojan@chromium.org>
1126
1127         Show the chromium non-layout test bots in garden-o-matic
1128         https://bugs.webkit.org/show_bug.cgi?id=85903
1129
1130         Reviewed by Eric Seidel.
1131
1132         This is just a single-line high-level overview of the chromium non-layout test bots.
1133         This means that Chromium WebKit gardeners only need to keep garden-o-matic open in
1134         order to do their job.
1135
1136         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
1137         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
1138         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
1139         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
1140         (#chromium-gtests iframe):
1141         (#chromium-gtests):
1142         (#gtest-label):
1143
1144 2012-05-08  Kangil Han  <kangil.han@samsung.com>
1145
1146         [EFL][DRT] Implement TextInputController.
1147         https://bugs.webkit.org/show_bug.cgi?id=78559
1148
1149         Reviewed by Martin Robinson.
1150
1151         TextInputController has been created for an input emulation purpose in DRT.
1152         However, not like other ports, EFL doesn't have it.
1153         Therefore, this patch will enable EFL port to have an input emulation in DRT.
1154
1155         * DumpRenderTree/efl/CMakeLists.txt:
1156         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1157         (DumpRenderTreeChrome::onWindowObjectCleared):
1158         * DumpRenderTree/efl/TextInputController.cpp: Added.
1159         (setMarkedTextCallback):
1160         (hasMarkedTextCallback):
1161         (markedRangeCallback):
1162         (insertTextCallback):
1163         (unmarkTextCallback):
1164         (firstRectForCharacterRangeCallback):
1165         (selectedRangeCallback):
1166         (makeTextInputController):
1167         * DumpRenderTree/efl/TextInputController.h: Added.
1168
1169 2012-04-23  Raphael Kubo da Costa  <rakuco@webkit.org>
1170
1171         [EFL] Pass a WTF::String directly from DRT to ewebkit when possible
1172         https://bugs.webkit.org/show_bug.cgi?id=84634
1173
1174         Reviewed by Gustavo Noronha Silva.
1175
1176         Drop unnecessary const char* -> WTF::String conversions by
1177         creating String objects directly in LayoutTestController out of
1178         JSStringRef's and passing them to DumpRenderTreeSupportEfl.
1179
1180         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1181         (LayoutTestController::queueLoad):
1182         (LayoutTestController::setValueForUser):
1183         (LayoutTestController::findString):
1184         (LayoutTestController::addUserStyleSheet):
1185
1186 2012-05-08  Alexander Shalamov  <alexander.shalamov@intel.com>
1187
1188         [EFL][DRT] EventSender needs implementation of leapForward function
1189         https://bugs.webkit.org/show_bug.cgi?id=83486
1190
1191         Reviewed by Eric Seidel.
1192
1193         Implemented leapForward function in Efl's DRT. This function is required
1194         to delay and queue event execution and simulate real user behavior.
1195
1196         * DumpRenderTree/efl/EventSender.cpp:
1197         (DelayedEvent):
1198         (MouseEventInfo::MouseEventInfo):
1199         (MouseEventInfo):
1200         (sendClick):
1201         (mouseDownCallback):
1202         (mouseUpCallback):
1203         (mouseMoveToCallback):
1204         (leapForwardCallback):
1205         (mouseScrollByCallback):
1206         (feedOrQueueMouseEvent):
1207         (feedMouseEvent):
1208         (feedQueuedMouseEvents):
1209
1210 2012-05-08  Ryuan Choi  <ryuan.choi@samsung.com>
1211
1212         [EFL][DRT] WebKit/DRT should not paint webview for the pixel tests.
1213         https://bugs.webkit.org/show_bug.cgi?id=79853
1214
1215         Reviewed by Eric Seidel.
1216
1217         Screen buffer can be different from rendered results.
1218         For example, WebKit/Efl just moves tiles for ewk_view_tiled when scrolling
1219         contents.
1220         It means that pixel tests can't check regression of ewk_view_tiled.
1221
1222         This patch changes a way to provide bitmap from painting webview to capturing
1223         evas.
1224
1225         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
1226         (createBitmapContextFromWebView):
1227
1228 2012-05-08  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1229
1230         [GTK] Memory leak in DRT EventSender.cpp
1231         https://bugs.webkit.org/show_bug.cgi?id=85886
1232
1233         Reviewed by Martin Robinson.
1234
1235         Fix a memory leak in gdkModifersFromJSValue() by using JSRetainPtr for
1236         JSString created with JSStringCreateWithUTF8CString.
1237
1238         * DumpRenderTree/gtk/EventSender.cpp:
1239         (gdkModifersFromJSValue):
1240
1241 2012-05-08  Christophe Dumez  <christophe.dumez@intel.com>
1242
1243         [EFL] glib dependency needs to be updated in jhbuild
1244         https://bugs.webkit.org/show_bug.cgi?id=85878
1245
1246         Reviewed by Martin Robinson.
1247
1248         Bump glib version from 2.31.8 to 2.32.2 in jhbuild.modules. This
1249         addresses flakiness in media tests due to possible deadlock with the
1250         following output on stderr:
1251         GStreamer-WARNING **: wrong STREAM_LOCK count 0
1252
1253         The bug was fixed upstream by:
1254         http://git.gnome.org/browse/glib/commit/?id=265f265c677bc5399dbb1cfa449cb4b915a6459f
1255
1256         * efl/jhbuild.modules:
1257
1258 2012-05-07  Alexis Menard  <alexis.menard@openbossa.org>
1259
1260         [Qt] Unbreak debugging of WebKit2.
1261         https://bugs.webkit.org/show_bug.cgi?id=85839
1262
1263         Reviewed by Simon Hausmann.
1264
1265         When you attach GDB to a running process, it stops it.
1266         http://trac.webkit.org/changeset/115958 introduced a pause() 
1267         call to wait the debugger to be attached to then continue 
1268         the execution of the WebProcess. Unfortunately the pause()
1269         function does not return unless a signal handler is called.
1270         This patch introduce an event handler to exit from the paused
1271         state when the debugger send the signal SIGCONT. The old code
1272         works with older version of GDB (<7.0) but not with newer
1273         versions where the behavior of pause() is correct.
1274
1275         * WebKitTestRunner/qt/main.cpp:
1276         (sigcontHandler):
1277         (main):
1278
1279 2012-05-08  Christophe Dumez  <christophe.dumez@intel.com>
1280
1281         [EFL][DRT] EFL's DRT does not fully support page visibility
1282         https://bugs.webkit.org/show_bug.cgi?id=85032
1283
1284         Reviewed by Antonio Gomes.
1285
1286         * DumpRenderTree/LayoutTestController.cpp:
1287         (resetPageVisibilityCallback): Fix incorrect argumentCount check.
1288         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1289         (DumpRenderTreeChrome::resetDefaultsToConsistentValues): Properly
1290         reset page visibility setting between tests.
1291         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1292         (LayoutTestController::setPageVisibility): Add support for "preview"
1293         visibility state.
1294
1295 2012-05-08  Gavin Peters  <gavinp@chromium.org>
1296
1297         [webkit-patch] Return meaningful results from single-result queries.
1298         https://bugs.webkit.org/show_bug.cgi?id=85367
1299
1300         Reviewed by Dirk Pranke.
1301
1302         Sometimes a quicksearch returns a single result by just redirecting to the page for that single
1303         bug. This patch hacks around that by noticing the redirect target is a single page, and constructing
1304         a result list consisting of that one bug.
1305
1306         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
1307         (BugzillaQueries._fetch_bugs_from_advanced_query):
1308
1309 2012-05-08  Gavin Peters  <gavinp@chromium.org>
1310
1311         [webkit-patch] Add new chrome-channels command to track down current chrome release channels for a committed bug.
1312         https://bugs.webkit.org/show_bug.cgi?id=85368
1313
1314         Reviewed by Eric Seidel.
1315
1316         * Scripts/webkitpy/common/config/urls.py:
1317         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
1318         (MockBugzillaQueries.fetch_bugs_matching_quicksearch):
1319         * Scripts/webkitpy/common/net/omahaproxy.py: Added.
1320         (OmahaProxy):
1321         (OmahaProxy.__init__):
1322         (OmahaProxy.set_url):
1323         (OmahaProxy._json_url):
1324         (OmahaProxy._get_json):
1325         (OmahaProxy.get_revisions):
1326         * Scripts/webkitpy/common/net/omahaproxy_unittest.py: Added.
1327         (MockOmahaProxy):
1328         (MockOmahaProxy.__init__):
1329         (OmahaProxyTest):
1330         (OmahaProxyTest.test_get_revisions):
1331         * Scripts/webkitpy/tool/commands/__init__.py:
1332         * Scripts/webkitpy/tool/commands/bugsearch.py:
1333         (BugSearch):
1334         * Scripts/webkitpy/tool/commands/chromechannels.py: Added.
1335         (ChromeChannels):
1336         (__init__):
1337         (execute):
1338         * Scripts/webkitpy/tool/commands/chromechannels_unittest.py: Added.
1339         (MockOmahaProxy):
1340         (MockOmahaProxy.get_revisions):
1341         (TestableChromeChannels):
1342         (TestableChromeChannels.__init__):
1343         (ChromeChannelsTest):
1344         (ChromeChannelsTest.test_single_bug):
1345         (ChromeChannelsTest.test_with_query):
1346
1347 2012-05-08  Carlos Garcia Campos  <cgarcia@igalia.com>
1348
1349         Unreviewed. Fix GTK+ unit tests.
1350
1351         * gtk/common.py:
1352         (pkg_config_file_variable):
1353
1354 2012-05-08  Carlos Garcia Campos  <cgarcia@igalia.com>
1355
1356         [GTK] Google tests that time out are leaked
1357         https://bugs.webkit.org/show_bug.cgi?id=84973
1358
1359         Reviewed by Philippe Normand.
1360
1361         Move Tools/Scripts/run-gtk-tests to Tools/gtk/run-api-tests
1362         leaving run-gtk-tests as simple wrapper that runs run-api-tests in
1363         a jhbuild environment when jhbuild is available.
1364
1365         * Scripts/run-gtk-tests: Run Tools/gtk/run-api-tests directly or
1366         with jhbuild if present.
1367         * gtk/common.py:
1368         (pkg_config_file_variable): Generic function to get a variable
1369         from a pkg-config file.
1370         (prefix_of_pkg_config_file): Use pkg_config_file_variable with 'prefix'.
1371         * gtk/run-api-tests: Copied from Tools/Scripts/run-gtk-tests.
1372         (TestRunner.__init__): Use common.build_path() to get the programs path.
1373         (TestRunner._lookup_atspi2_binary): Use
1374         common.pkg_config_file_variable() to get the 'exec_prefix'.
1375         (TestRunner._start_accessibility_daemons): Use subprocess.Popen()
1376         directly instead of helper _create_process that has been removed.
1377         (TestRunner._setup_testing_environment): Use subprocess.Popen()
1378         directly instead of helper _create_process that has been
1379         removed. Use common.top_level_path() and common.build_path()
1380         instead of custom methods that have been removed too.
1381         (TestRunner._run_test_command): Use subprocess.Popen() directly
1382         instead of helper _create_process that has been removed.
1383
1384 2012-05-08  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1385
1386         [EFL] [DRT] Clear extra views before setting a blank page
1387         https://bugs.webkit.org/show_bug.cgi?id=85608
1388
1389         Reviewed by Kenneth Rohde Christiansen.
1390
1391         We need to clear the extra views before setting a blank page,
1392         otherwise it will cause more events interactions and sometimes
1393         trigger a unload(). This was causing frame-crash-with-page-cache.html
1394         to leak events to the next test causing any subsequent test to
1395         fail.
1396
1397         * DumpRenderTree/efl/DumpRenderTree.cpp:
1398         (runTest):
1399         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1400         (DumpRenderTreeChrome::clearExtraViews):
1401         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1402         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
1403         (DumpRenderTreeChrome):
1404
1405 2012-05-08  Balazs Kelemen  <kbalazs@webkit.org>
1406
1407         [Qt] X11 plugins need to be reworked for Qt5+WK1
1408         https://bugs.webkit.org/show_bug.cgi?id=80691
1409
1410         Reviewed by Simon Hausmann.
1411
1412         Implement basic windowless plugin support with Qt5.
1413
1414         * qmake/mkspecs/features/features.prf:
1415         Enable NPAPI plugins if X11 libraries are available.
1416         * qmake/mkspecs/features/functions.prf:
1417         Added a convenience function to determine availability
1418         of X11 libraries.
1419
1420 2012-05-07  Raphael Kubo da Costa  <rakuco@webkit.org>
1421
1422         [webkitpy] Remove Python 2.5-specific workaround from http_server_base.py.
1423         https://bugs.webkit.org/show_bug.cgi?id=85852
1424
1425         Reviewed by Dirk Pranke.
1426
1427         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
1428         (HttpServerBase._is_server_running_on_all_ports): Revert the part
1429         of r89412 which was a workaround for an issue with Python 2.5 on
1430         Mac 10.5.
1431
1432 2012-05-07  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
1433
1434         webkitdirs.pm: Make determineNumberOfCPUs work on FreeBSD.
1435         https://bugs.webkit.org/show_bug.cgi?id=81481
1436
1437         Reviewed by Dirk Pranke.
1438
1439         FreeBSD has the same sysctl interface present in Darwin to query
1440         the number of CPUs available, so use that too.
1441
1442         * Scripts/webkitdirs.pm:
1443         (determineNumberOfCPUs): Share the sysctl call with Darwin.
1444         (isFreeBSD): Added.
1445
1446 2012-05-07  Alexis Menard  <alexis.menard@openbossa.org>
1447
1448         Unreviewed Qt build fix in Debug.
1449
1450         We need some system includes here for getpid() to be recognized.
1451
1452         * WebKitTestRunner/qt/main.cpp:
1453
1454 2012-05-07  Ojan Vafai  <ojan@chromium.org>
1455
1456         When embedding the flakiness dashboard, hide popups when the frame is blurred.
1457         https://bugs.webkit.org/show_bug.cgi?id=85684
1458
1459         Reviewed by Dirk Pranke.
1460
1461         If you show a popup, then click elsewhere in garden-o-matic, it will now
1462         properly hide it.
1463
1464         * TestResultServer/static-dashboards/dashboard_base.js:
1465         (showPopup):
1466         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1467         (showPopupForBuild):
1468         (generatePageForIndividualTests.if):
1469         (generatePageForIndividualTests):
1470         * TestResultServer/static-dashboards/flakiness_dashboard_embedded_unittests.js: Added.
1471         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1472         * TestResultServer/static-dashboards/run-embedded-unittests.html: Added.
1473
1474 2012-05-07  Ian Vollick  <vollick@chromium.org>
1475
1476         [chromium] WebViewHost should setCompositorSurfaceReady
1477         https://bugs.webkit.org/show_bug.cgi?id=85418
1478
1479         Reviewed by James Robinson.
1480
1481         * DumpRenderTree/chromium/WebViewHost.cpp:
1482         (WebViewHost::setWebWidget):
1483
1484 2012-05-07  Jon Lee  <jonlee@apple.com>
1485
1486         WebKitTestRunner needs layoutTestController.dumpFrameLoadCallbacks
1487         https://bugs.webkit.org/show_bug.cgi?id=42331
1488         <rdar://problem/8193641>
1489
1490         Reviewed by Darin Adler.
1491
1492         This patch sets up the framework for supporting dumpFrameLoadCallbacks, as well as
1493         outputting some of the required strings. As tests become unskipped the other callbacks
1494         can be supported.
1495
1496         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Add dumpFrameLoadCallbacks()
1497         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1498         (WTR::InjectedBundle::didReceiveMessage): pass in the dictionary containing keys needed to initialize
1499         dump callbacks for the layoutTestController.
1500         (WTR::InjectedBundle::booleanForKey): Helper function to extract boolean value from key in WKDictionaryRef.
1501         Output some warning if we are trying to extract a value that is not a boolean.
1502         (WTR::InjectedBundle::beginTesting): Update to take in the dictionary passed in by the TestController.
1503         We cannot set the bits on the layoutTestController until it has been created in beginTesting().
1504         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
1505         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1506         (WTR::InjectedBundle::didReceiveMessage):
1507         (WTR):
1508         (WTR::InjectedBundle::beginTesting):
1509         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
1510         (InjectedBundle):
1511         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1512         (WTR):
1513         (WTR::dumpFrameDescriptionSuitableForTestResult): Static function similar to
1514         -[WebFrame _drt_descriptionSuitableForTestResult] in DumpRenderTree.
1515         (WTR::InjectedBundlePage::didStartProvisionalLoadForFrame): Output string.
1516         (WTR::InjectedBundlePage::didCommitLoadForFrame): Output string.
1517         (WTR::InjectedBundlePage::didFinishLoadForFrame): Output string.
1518         (WTR::InjectedBundlePage::didReceiveTitleForFrame): Output string.
1519         (WTR::InjectedBundlePage::didCancelClientRedirectForFrame): Output string.
1520         (WTR::InjectedBundlePage::willPerformClientRedirectForFrame): Output string. In WK1 it's
1521         "willPerformClientRedirectToURL" so we will have to use that value here.
1522         (WTR::InjectedBundlePage::didFinishDocumentLoadForFrame): Output string.
1523         (WTR::InjectedBundlePage::didHandleOnloadEventsForFrame): Output string.
1524         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1525         (WTR::LayoutTestController::LayoutTestController): Initialize to not dump frame load callbacks.
1526         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1527         (WTR::LayoutTestController::dumpFrameLoadCallbacks): Change to use setShouldDumpFrameLoadCallbacks.
1528         (WTR::LayoutTestController::setShouldDumpFrameLoadCallbacks): Set bit.
1529         (WTR::LayoutTestController::shouldDumpFrameLoadCallbacks): Return bit.
1530         (LayoutTestController): Added m_dumpFrameLoadCallbacks to track whether to dump the output.
1531         * WebKitTestRunner/TestInvocation.cpp:
1532         (WTR::shouldLogFrameLoadDelegates): Similar to DRT.
1533         (WTR::TestInvocation::invoke): Based on the URL loaded, set the bit in the dictionary that gets passed
1534         to the injected bundle.
1535
1536 2012-05-07  Tommy Widenflycht  <tommyw@google.com>
1537
1538         [chromium] MediaStream API: Enhance WebUserMediaClientMock and add a test for it
1539         https://bugs.webkit.org/show_bug.cgi?id=85618
1540
1541         Reviewed by Adam Barth.
1542
1543         Adding the capability to control which MediaStreamTracks actually get created,
1544         and adding a test that exercises the JS api using that capability.
1545
1546         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp:
1547         (WebKit::WebUserMediaClientMock::requestUserMedia):
1548         * DumpRenderTree/chromium/WebUserMediaClientMock.h:
1549         * DumpRenderTree/chromium/WebViewHost.cpp:
1550         (WebViewHost::testMediaStreamClient):
1551         * DumpRenderTree/chromium/WebViewHost.h:
1552         (WebViewHost):
1553
1554 2012-05-07  Zan Dobersek  <zandobersek@gmail.com>
1555
1556         [Gtk] WebGL feature is not built anymore through build-webkit after r116251
1557         https://bugs.webkit.org/show_bug.cgi?id=85788
1558
1559         Reviewed by Martin Robinson.
1560
1561         Enable WebGL by default for the Gtk port. The feature is already enabled by
1562         default in configure.ac, so there's no reason it shouldn't be enabled when
1563         building through build-webkit.
1564
1565         * Scripts/webkitperl/FeatureList.pm:
1566
1567 2012-05-07  Csaba Osztrogonác  <ossy@webkit.org>
1568
1569         Unittest fix after r116288, because now there are more than 5 hit for the 'Eric' word.
1570         We have only 2 'Gavin', so it is a better choice than 'Eric' for unit testing.
1571
1572         Reviewed by Tor Arne Vestbø.
1573
1574         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1575         (IRCCommandTest.test_whois):
1576
1577 2012-05-07  Sheriff Bot  <webkit.review.bot@gmail.com>
1578
1579         Unreviewed, rolling out r116299, r116301, and r116303.
1580         http://trac.webkit.org/changeset/116299
1581         http://trac.webkit.org/changeset/116301
1582         http://trac.webkit.org/changeset/116303
1583         https://bugs.webkit.org/show_bug.cgi?id=85795
1584
1585         Build is still broken (Requested by Ossy on #webkit).
1586
1587         * DumpRenderTree/qt/DumpRenderTree.pro:
1588         * DumpRenderTree/qt/main.cpp:
1589         (main):
1590         * QtTestPlatformPlugin/QtTestPlatformPlugin.pro: Removed.
1591         * QtTestPlatformPlugin/TestIntegration.cpp: Removed.
1592         * QtTestPlatformPlugin/TestIntegration.h: Removed.
1593         * QtTestPlatformPlugin/mac/TestFontDatabase.h: Removed.
1594         * QtTestPlatformPlugin/mac/TestFontDatabase.mm: Removed.
1595         * QtTestPlatformPlugin/mac/TestIntegrationMac.mm: Removed.
1596         * QtTestPlatformPlugin/main.cpp: Removed.
1597         * QtTestPlatformPlugin/testplatform.json: Removed.
1598         * Tools.pro:
1599         * WebKitTestRunner/Target.pri:
1600         * WebKitTestRunner/qt/main.cpp:
1601         (main):
1602
1603 2012-05-07  Balazs Kelemen  <kbalazs@webkit.org>
1604
1605         Unreviewed Qt build fix.
1606
1607         One more try to learn how to use
1608         const with a pointer array.
1609
1610         * DumpRenderTree/qt/main.cpp:
1611         (initializeTestPlatformPlugin):
1612
1613 2012-05-07  Balazs Kelemen  <kbalazs@webkit.org>
1614
1615         Unreviewed build fix after r116299.
1616
1617         * DumpRenderTree/qt/main.cpp:
1618         (initializeTestPlatformPlugin):
1619
1620 2012-05-07  Balazs Kelemen  <kbalazs@webkit.org>
1621
1622         [Qt] Add test specific platform plugin to achieve unified layout test results
1623         https://bugs.webkit.org/show_bug.cgi?id=80996
1624
1625         Reviewed by Simon Hausmann.
1626
1627         Added QtTestPlatformPlugin as a new project under Tools.
1628         This is a Qt5-ish platform plugin that can be used to tweak the
1629         platform support interfaces in order to unify layout test results.
1630         For now it only overrides the font database on Mac and redirects
1631         everything else to the real platform plugin. The font database it
1632         provides mimics the way how we set up test fonts with fontconfig on Linux.
1633         Make DumpRenderTree and WebKitTestRunner use this platform plugin.
1634
1635         * DumpRenderTree/qt/DumpRenderTree.pro:
1636         * DumpRenderTree/qt/main.cpp:
1637         (initializeTestPlatformPlugin):
1638         (main):
1639         * QtTestPlatformPlugin/QtTestPlatformPlugin.pro: Added.
1640         * QtTestPlatformPlugin/TestIntegration.cpp: Added.
1641         (TestIntegration::TestIntegration):
1642         (TestIntegration::fontDatabase):
1643         * QtTestPlatformPlugin/TestIntegration.h: Added.
1644         (TestIntegration):
1645         (TestIntegration::hasCapability):
1646         (TestIntegration::createPlatformPixmap):
1647         (TestIntegration::createPlatformWindow):
1648         (TestIntegration::createPlatformBackingStore):
1649         (TestIntegration::createPlatformOpenGLContext):
1650         (TestIntegration::createPlatformSharedGraphicsCache):
1651         (TestIntegration::guiThreadEventDispatcher):
1652         (TestIntegration::clipboard):
1653         (TestIntegration::drag):
1654         (TestIntegration::inputContext):
1655         (TestIntegration::accessibility):
1656         (TestIntegration::nativeInterface):
1657         (TestIntegration::services):
1658         (TestIntegration::styleHint):
1659         (TestIntegration::platformTheme):
1660         * QtTestPlatformPlugin/mac/TestFontDatabase.h: Added.
1661         (TestFontDatabase):
1662         * QtTestPlatformPlugin/mac/TestFontDatabase.mm: Added.
1663         (TestFontDatabase::populateFontDatabase):
1664         * QtTestPlatformPlugin/mac/TestIntegrationMac.mm: Added.
1665         (TestIntegration::fontDatabase):
1666         * QtTestPlatformPlugin/main.cpp: Added.
1667         (TestIntegrationPlugin::keys):
1668         (TestIntegrationPlugin::create):
1669         (TestIntegrationPlugin::initialize):
1670         * QtTestPlatformPlugin/testplatform.json: Added.
1671         * Tools.pro:
1672         * WebKitTestRunner/Target.pri:
1673         * WebKitTestRunner/qt/main.cpp:
1674         (main):
1675
1676 2012-05-07  Christophe Dumez  <christophe.dumez@intel.com>
1677
1678         [EFL] media/video-poster-blocked-by-willsendrequest.html is flaky
1679         https://bugs.webkit.org/show_bug.cgi?id=85609
1680
1681         Reviewed by Simon Hausmann.
1682
1683         Flakiness was caused by the fact that
1684         media/crash-closing-page-with-media-as-plugin-fallback.html and
1685         media/video-poster-blocked-by-willsendrequest.html use the same PNG
1686         file as poster attribute in the video tag. The PNG file gets cached
1687         when the first test is executed and this somehow causes the second
1688         test to fail. 
1689
1690         The flakiness was addressed by clearing the memory and page caches
1691         between tests.
1692
1693         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1694         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1695
1696 2012-05-07  Eric Seidel  <eric@webkit.org>
1697
1698         Add values for all features to Qt's features.pri
1699         https://bugs.webkit.org/show_bug.cgi?id=85746
1700
1701         Reviewed by Tor Arne Vestbø.
1702
1703         When generate-feature-files generates features.pri it provides a value
1704         for every possible ENABLE_.  In prepration for landing generate-feature-files
1705         it seems appropriate to add a value for all possible ENABLEs to the existing
1706         features.pri.  This lets me make sure that I'm getting them all right!
1707
1708         * qmake/mkspecs/features/features.pri:
1709
1710 2012-05-07  Adam Bergkvist  <adam.bergkvist@ericsson.com>
1711
1712         Added myself as a committer.
1713
1714         Not reviewed.
1715
1716         * Scripts/webkitpy/common/config/committers.py:
1717
1718 2012-05-07  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1719
1720         [EFL][DRT] Implement LayoutTestController::execCommand
1721         https://bugs.webkit.org/show_bug.cgi?id=82286
1722
1723         Reviewed by Antonio Gomes.
1724
1725         Add missing implementation execCommand and isCommandEnabled to EFL's
1726         LayoutTestController. 
1727
1728         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1729         (LayoutTestController::execCommand):
1730         (LayoutTestController::isCommandEnabled):
1731
1732 2012-05-06  Hayato Ito  <hayato@chromium.org>
1733
1734         [prepare-ChangeLog] Excludes a reference file used by reftests from test files.
1735         https://bugs.webkit.org/show_bug.cgi?id=74268
1736
1737         Reviewed by Ryosuke Niwa.
1738
1739         * Scripts/prepare-ChangeLog:
1740         (generateFileList):
1741
1742 2012-05-06  Mary Wu  <mary.wu@torchmobile.com.cn>
1743
1744         Add a contributor to committers.py
1745         https://bugs.webkit.org/show_bug.cgi?id=85761
1746
1747         Reviewed by Unreviewed.
1748
1749         Add myself to committers.py contributor.
1750
1751         * Scripts/webkitpy/common/config/committers.py:
1752
1753 2012-05-06  Dan Bernstein  <mitz@apple.com>
1754
1755         Tools part of: Building and debugging WebKit in the Xcode IDE requires a lot of setup
1756         https://bugs.webkit.org/show_bug.cgi?id=85739
1757
1758         Reviewed by Daniel Bates.
1759
1760         * Scripts/build-webkit: Moved code that copies from WebKitLibraries to the product directory
1761         from here...
1762         * Scripts/copy-webkitlibraries-to-product-directory: ...to this new script.
1763
1764 2012-05-06  Andy Estes  <aestes@apple.com>
1765
1766         Fix the build after r116246.
1767
1768         * DumpRenderTree/LayoutTestController.cpp: Remove an unused function.
1769
1770 2012-05-06  Eric Seidel  <eric@webkit.org>
1771
1772         Update several build-webkit options to match the #defines they toggle
1773         https://bugs.webkit.org/show_bug.cgi?id=85744
1774
1775         Reviewed by Adam Barth.
1776
1777         Most notable here is fixing --3d-canvas to be --webgl these days --
1778         the define was renamed 15 months ago!
1779
1780         * Scripts/webkitperl/FeatureList.pm:
1781
1782 2012-05-06  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1783
1784         [EFL] EFL's LayoutTestController setJavaScriptCanAccessClipboard implementation
1785         https://bugs.webkit.org/show_bug.cgi?id=83687
1786
1787         Reviewed by Antonio Gomes.
1788
1789         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1790         (LayoutTestController::setJavaScriptCanAccessClipboard):
1791
1792 2012-05-06  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1793
1794         Convert isPageBoxVisible to use Internals interface.
1795         https://bugs.webkit.org/show_bug.cgi?id=85692
1796
1797         Reviewed by Darin Adler.
1798
1799         Remove isPageBoxVisible functions, because it is able to work in the
1800         cross-port way through the Internals interface.
1801
1802         * DumpRenderTree/LayoutTestController.cpp:
1803         (LayoutTestController::staticFunctions):
1804         * DumpRenderTree/LayoutTestController.h:
1805         (LayoutTestController):
1806         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
1807         * DumpRenderTree/chromium/LayoutTestController.cpp:
1808         (LayoutTestController::LayoutTestController):
1809         * DumpRenderTree/chromium/LayoutTestController.h:
1810         (LayoutTestController):
1811         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1812         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1813         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1814         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1815         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1816         (LayoutTestController):
1817         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1818         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1819
1820 2012-05-06  Tom Zakrajsek  <tomz@codeaurora.org>
1821
1822         sheriffbot isn't reopening patches after it lands rollouts
1823         https://bugs.webkit.org/show_bug.cgi?id=64418
1824
1825         Reviewed by Adam Barth.
1826
1827         * Scripts/webkitpy/tool/commands/download_unittest.py:
1828         * Scripts/webkitpy/tool/steps/createbug.py:
1829         (CreateBug.run):
1830
1831 2012-05-04  Raphael Kubo da Costa  <rakuco@webkit.org>
1832
1833         webkitpy: Use os.pathsep instead of manually finding out the path separator in NRWT.
1834         https://bugs.webkit.org/show_bug.cgi?id=85697
1835
1836         Reviewed by Eric Seidel.
1837
1838         * Scripts/new-run-webkit-tests: Instead of manually detecting whether
1839         the path separator should be ':' or ';' by checking sys.platform, rely
1840         on os.pathsep which has exactly the same information (os.pathsep is
1841         ':' for Cygwin, from what I could check in Python's sources).
1842
1843 2012-05-05  Sheriff Bot  <webkit.review.bot@gmail.com>
1844
1845         Unreviewed, rolling out r116223.
1846         http://trac.webkit.org/changeset/116223
1847         https://bugs.webkit.org/show_bug.cgi?id=85718
1848
1849         The bots need to set XDG_DATA_DIRS (Requested by rakuco on
1850         #webkit).
1851
1852         * efl/jhbuildrc:
1853         * gtk/jhbuildrc:
1854
1855 2012-05-05  Raphael Kubo da Costa  <rakuco@webkit.org>
1856
1857         [jhbuild] Do not set XDG_DATA_DIRS in jhbuildrc.
1858         https://bugs.webkit.org/show_bug.cgi?id=85717
1859
1860         Reviewed by Martin Robinson.
1861
1862         After r116209, the XDG_DATA_DIRS environment variable is now
1863         passed by webkitpy's layout_tests code to run-with-jhbuild, so
1864         instead of setting a few defaults in jhbuildrc we can now rely on
1865         the values set and really used by each system.
1866
1867         * efl/jhbuildrc:
1868         * gtk/jhbuildrc:
1869
1870 2012-05-05  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1871
1872         [Qt] Buildfix for --minimal build after r116107.
1873
1874         Reviewed by Csaba Osztrogonác.
1875
1876         * qmake/mkspecs/features/features.prf:
1877
1878 2012-05-05  Zan Dobersek  <zandobersek@gmail.com>
1879
1880         [Gtk] GtkDriver is not required anymore after r116134
1881         https://bugs.webkit.org/show_bug.cgi?id=85699
1882
1883         Reviewed by Csaba Osztrogonác.
1884
1885         Remove GtkDriver as it was replaced by the XvfbDriver and is
1886         no longer needed.
1887
1888         * Scripts/webkitpy/layout_tests/port/gtk.py:
1889
1890 2012-05-05  Zan Dobersek  <zandobersek@gmail.com>
1891
1892         Increase the display ID for the Xvfb process to use if pixel_tests argument
1893         is true rather than if pixel_tests option is present. This fixes the
1894         XvfbDriver in the same way r115825 fixed the now-redundant GtkDriver.
1895
1896         Rubber-stamped by Csaba Osztrogonác.
1897
1898         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1899         (XvfbDriver._start):
1900
1901 2012-05-04  Raphael Kubo da Costa  <rakuco@webkit.org>
1902
1903         webkitpy: Preserve XDG_DATA_DIRS in the environment passed to ServerProcess.
1904         https://bugs.webkit.org/show_bug.cgi?id=85694
1905
1906         Reviewed by Martin Robinson.
1907
1908         This is sort of a follow-up to r100674: $XDG_DATA_DIRS should be
1909         preserved, as it is particularly useful for Linux/Unix
1910         environments.  As a real use case, if shared-mime-info is
1911         installed into a non-trivial prefix, Glib and other libraries will
1912         have trouble to properly identify mimetypes (in ResourceResponses,
1913         for example) as the base freedesktop.org mimetype file will not be
1914         found.
1915
1916         * Scripts/webkitpy/layout_tests/port/base.py:
1917         (Port.to.setup_environ_for_server):
1918
1919 2012-05-04  Gustavo Noronha Silva  <gns@gnome.org>
1920
1921         [GTK] Simplify how libWebCoreModules is linked in, and fix WebKit2 build
1922         https://bugs.webkit.org/show_bug.cgi?id=85691
1923
1924         * GNUmakefile.am: no longer link libWebCoreModules to DumpRenderTree.
1925
1926 2012-05-04  Eric Seidel  <eric@webkit.org>
1927
1928         Make capitalization and descriptions in FeatureList.pm more consistent to make easier to autogenerate
1929         https://bugs.webkit.org/show_bug.cgi?id=85583
1930
1931         Reviewed by Daniel Bates.
1932
1933         * Scripts/webkitperl/FeatureList.pm:
1934
1935 2012-05-04  Eric Seidel  <eric@webkit.org>
1936
1937         Sort entries in FeaturesList.pm to make them easier to autogenerate
1938         https://bugs.webkit.org/show_bug.cgi?id=85584
1939
1940         Reviewed by Daniel Bates.
1941
1942         No functional changes, only sorting.
1943
1944         * Scripts/webkitperl/FeatureList.pm:
1945
1946 2012-05-04  Andy Estes  <aestes@apple.com>
1947
1948         REGRESSION (r115752): WebKit2.NewFirstVisuallyNonEmptyLayout API test times out
1949         https://bugs.webkit.org/show_bug.cgi?id=85675
1950
1951         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayout.cpp:
1952         (TestWebKitAPI::TEST): Disable the test.
1953
1954 2012-04-30  Jer Noble  <jer.noble@apple.com>
1955
1956         Full screen will exit during a provisional load of a non-ancestor iframe.
1957         https://bugs.webkit.org/show_bug.cgi?id=85230
1958
1959         Reviewed by Maciej Stachowiak .
1960
1961         Add support for the new UIDelegate webView:closeFullScreenWithListener: method:
1962         * DumpRenderTree/mac/UIDelegate.mm:
1963         (-[UIDelegate webView:closeFullScreenWithListener:]):
1964
1965         Add support for the new closeFullScreen injected bundle method:
1966         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1967         (WTR::InjectedBundlePage::InjectedBundlePage):
1968         (WTR::InjectedBundlePage::closeFullScreen):
1969         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
1970
1971 2012-05-04  Ojan Vafai  <ojan@chromium.org>
1972
1973         Fix test naming to be camel-case like the methods they are testing.
1974
1975         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1976         (test):
1977
1978 2012-05-04  Ojan Vafai  <ojan@chromium.org>
1979
1980         Show the results links in the embedded flakiness dashboard
1981         https://bugs.webkit.org/show_bug.cgi?id=85660
1982
1983         Reviewed by Adam Barth.
1984
1985         Turns out these links don't add much visual noise and it's annoying
1986         to have to go to the dashboard from garden-o-matic just to view the
1987         results for a test.
1988
1989         This is useful for cases where garden-o-matic doesn't properly show the
1990         results (e.g. missing tests). If we address all those cases, then
1991         we could easily hide the results links again if we wanted to.
1992
1993         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1994         (showUpdateInfoForTest):
1995         (htmlForIndividualTestOnAllBuildersWithResultsLinks):
1996         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1997         (test):
1998
1999 2012-05-04  Dirk Pranke  <dpranke@chromium.org>
2000
2001         Revert to using chromium --test-shell mode on SL in NRWT
2002
2003         https://bugs.webkit.org/show_bug.cgi?id=83076
2004
2005         Unreviewed, build fix.
2006
2007         We seem to be seeing a lot more flakiness on the bot since
2008         I switched NRWT to "DRT" mode. We'll try reverting back and
2009         see if this helps things.
2010
2011         * Scripts/webkitpy/layout_tests/port/chromium.py:
2012         (ChromiumDriver.__init__):
2013
2014 2012-05-04  Christophe Dumez  <christophe.dumez@intel.com>
2015
2016         [EFL] Implement layoutTestController.dumpResourceResponseMIMETypes
2017         https://bugs.webkit.org/show_bug.cgi?id=84941
2018
2019         Reviewed by Martin Robinson.
2020
2021         Print MIME type detection messages in EFL's DumpRenderTree if
2022         LayoutTestController's dumpResourceResponseMIMETypes() returns true.
2023
2024         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2025         (DumpRenderTreeChrome::onResponseReceived):
2026
2027 2012-05-04  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2028
2029         [GTK] DRT needs an implementation of LayoutTestController::setDefersLoading and ::goBack
2030         https://bugs.webkit.org/show_bug.cgi?id=85134
2031
2032         Reviewed by Martin Robinson.
2033
2034         Add missing implementation setDefersLoading and goBack to GTK's
2035         LayoutTestController so that the test case rely on these methods
2036         can be enabled.
2037
2038         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2039         (resetDefaultsToConsistentValues):
2040         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2041         (LayoutTestController::goBack):
2042         (LayoutTestController::setDefersLoading):
2043
2044 2012-05-04  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2045
2046         [Qt][NRWT] Run each DRT in it's own xvfb
2047         https://bugs.webkit.org/show_bug.cgi?id=77335
2048
2049         Reviewed by Dirk Pranke.
2050
2051         * Scripts/webkitpy/layout_tests/port/gtk.py:
2052         (GtkPort._driver_class):
2053         * Scripts/webkitpy/layout_tests/port/qt.py:
2054         (QtPort._driver_class):
2055         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py: Added.
2056         (XvfbDriver):
2057         (XvfbDriver._start):
2058         (XvfbDriver._start.x_filter):
2059         (XvfbDriver.stop):
2060
2061 2012-05-04  Zan Dobersek  <zandobersek@gmail.com>
2062
2063         Use destructuring assingment for the return values of the
2064         port.diff_image() method call.
2065
2066         Rubber-stamped by Dirk Pranke.
2067
2068         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
2069         (write_test_result):
2070
2071 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2072
2073         Store fewer runs by default in the flakiness dashboard
2074         https://bugs.webkit.org/show_bug.cgi?id=85568
2075
2076         Reviewed by Dirk Pranke.
2077
2078         This will greatly shrink the size of the json files we load,
2079         resulting in a much faster initial load. People rarely care
2080         about more than 100 runs. For the cases they do, they can still click
2081         the show all runs checkbox to show 500 runs.
2082
2083         * TestResultServer/model/jsonresults.py:
2084
2085 2012-05-04  Christophe Dumez  <christophe.dumez@intel.com>
2086
2087         [EFL] Frame load callbacks output is missing for redirections
2088         https://bugs.webkit.org/show_bug.cgi?id=85173
2089
2090         Reviewed by Antonio Gomes.
2091
2092         Catch the new ewk_frame signals to print the expected output in case
2093         of redirections, when LayoutTestController's dumpFrameLoadCallbacks()
2094         returns true.
2095
2096         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2097         (DumpRenderTreeChrome::createView):
2098         (DumpRenderTreeChrome::onWillSendRequest):
2099         (DumpRenderTreeChrome::onFrameCreated):
2100         (DumpRenderTreeChrome::onFrameRedirectForProvisionalLoad):
2101         (DumpRenderTreeChrome::onFrameRedirectRequested):
2102         (DumpRenderTreeChrome::onResponseReceived):
2103         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
2104         (DumpRenderTreeChrome):
2105
2106 2012-05-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2107
2108         [Qt] Fix up warning about missing GLib/Gio/GStreamer
2109
2110         We don't actually fall back.
2111
2112         Rubber-stamped by Csaba Osztrogonác.
2113
2114         * qmake/mkspecs/features/features.prf:
2115
2116 2012-05-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2117
2118         [Qt] Clarify warning about missing GLib/Gio/GStreamer for media support
2119
2120         And only print it once.
2121
2122         Reviewed by Csaba Osztrogonác.
2123
2124         * qmake/mkspecs/features/features.prf:
2125
2126 2012-05-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2127
2128         [Qt] Clean up and split features.prf into a static list of defaults
2129
2130         The static list of feature defaults is used as a fallback for any
2131         feature that's not dynamically detected or overriden on the command
2132         line (though build-webkit or passing DEFINES+= to qmake).
2133
2134         The static list is complete, which allows for auto-generation based
2135         on Features.py (see bug https://bugs.webkit.org/show_bug.cgi?id=85456)
2136
2137         https://bugs.webkit.org/show_bug.cgi?id=85611
2138
2139         Reviewed by Simon Hausmann.
2140
2141         * Scripts/build-webkit:
2142         * Tools.pro:
2143         * qmake/mkspecs/features/features.prf:
2144         * qmake/mkspecs/features/features.pri: Added.
2145
2146 2012-05-04  Sheriff Bot  <webkit.review.bot@gmail.com>
2147
2148         Unreviewed, rolling out r116085, r116091, and r116095.
2149         http://trac.webkit.org/changeset/116085
2150         http://trac.webkit.org/changeset/116091
2151         http://trac.webkit.org/changeset/116095
2152         https://bugs.webkit.org/show_bug.cgi?id=85628
2153
2154         We are not ready with dependencies on all platform yet (mac) +
2155         problems with debug builds. (Requested by Zoltan on #webkit).
2156
2157         * qmake/mkspecs/features/features.prf:
2158
2159 2012-05-04  Christophe Dumez  <christophe.dumez@intel.com>
2160
2161         [EFL] LayoutTestController does not support overriding WebKitLoadSiteIconsKey
2162         https://bugs.webkit.org/show_bug.cgi?id=85171
2163
2164         Reviewed by Andreas Kling.
2165
2166         Implement support for overriding "WebKitLoadSiteIconsKey" setting in
2167         EFL's LayoutTestController.
2168
2169         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2170         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
2171         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2172         (LayoutTestController::overridePreference):
2173
2174 2012-05-03  Alexis Menard  <alexis.menard@openbossa.org>
2175
2176         [Qt] Enable fullscreen API for WebKit2.
2177         https://bugs.webkit.org/show_bug.cgi?id=85498
2178
2179         Reviewed by Simon Hausmann.
2180
2181         Enable the fullscreen API on the MiniBrowser and turn
2182         it default on WebKit2, disable it on WK1.
2183
2184         * MiniBrowser/qt/BrowserWindow.cpp:
2185         (BrowserWindow::BrowserWindow):
2186         * MiniBrowser/qt/qml/BrowserWindow.qml:
2187         * Scripts/webkitperl/FeatureList.pm:
2188         * qmake/mkspecs/features/features.prf:
2189
2190 2012-05-04  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
2191
2192         REGRESSION(r116008) old-run-webkit-tests broken
2193         https://bugs.webkit.org/show_bug.cgi?id=85597
2194
2195         Reviewed by Simon Hausmann.
2196
2197         Reinstate libraryContainsSymbol.
2198
2199         * Scripts/webkitperl/features.pm:
2200         (libraryContainsSymbol):
2201
2202 2012-05-04  Zoltan Horvath  <zoltan@webkit.org>
2203
2204         [Qt] Set WebCore imagedecoders as default and add fallback to QImageDecoder
2205         https://bugs.webkit.org/show_bug.cgi?id=80400
2206
2207         Remove WTF_USE_QT_IMAGE_DECODER macro.
2208
2209         Reviewed by Simon Hausmann.
2210
2211         * qmake/mkspecs/features/features.prf:
2212
2213 2012-05-03  Christophe Dumez  <christophe.dumez@intel.com>
2214
2215         [EFL] DRT should obey layoutTestController's addURLToRedirect()
2216         https://bugs.webkit.org/show_bug.cgi?id=82722
2217
2218         Reviewed by Maciej Stachowiak.
2219
2220         Implement support for layoutTestController's addURLToRedirect()
2221         in EFL's DRT so that it redirects properly when indicated.
2222
2223         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2224         (DumpRenderTreeChrome::onWillSendRequest):
2225
2226 2012-05-03  Christophe Dumez  <christophe.dumez@intel.com>
2227
2228         [EFL] DRT should support LayoutTestController's willSendRequestReturnsNullOnRedirect()
2229         https://bugs.webkit.org/show_bug.cgi?id=82704
2230
2231         Reviewed by Antonio Gomes.
2232
2233         EFL's DRT needs to obey LayoutTestController's
2234         willSendRequestReturnsNullOnRedirect().
2235
2236         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2237         (DumpRenderTreeChrome::onWillSendRequest):
2238
2239 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2240
2241         git-add-reviewer: Do not put the bug title and its URL in the same line in the commit message.
2242         https://bugs.webkit.org/show_bug.cgi?id=85570
2243
2244         Reviewed by Adam Roben.
2245
2246         The standard format of commit messages and ChangeLog entries has a
2247         short summary in the first line, followed by the bug URL in
2248         Bugzilla. Passing "%s" in the pretty format line to git-rev-list
2249         merges these two lines into a single one separated by a space
2250         character, which is not what we want.
2251
2252         Since "%s%n%n%b" is in practice "%B" (raw message), use it
2253         instead, as the first two lines do not end up being merged
2254         together.
2255
2256         * Scripts/git-add-reviewer:
2257         (writeCommitMessageToFile):
2258
2259 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2260
2261         Improve UI for garden-o-matic examine view when there are fewer than 5 tests.
2262         https://bugs.webkit.org/show_bug.cgi?id=85566
2263
2264         Reviewed by Adam Barth.
2265
2266         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2267         (.):
2268         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2269         (.):
2270         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
2271
2272 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2273
2274         webkitpy: Recognize FreeBSD as a valid platform.
2275         https://bugs.webkit.org/show_bug.cgi?id=81467
2276
2277         Reviewed by Adam Barth.
2278
2279         Add PlatformInfo.is_freebsd() so that trying to use webkit-patch
2280         on FreeBSD does not raise an "unrecognized platform" exception.
2281
2282         Most of this patch contains some unit tests for the mentioned
2283         method, as currently platform-specific code is needed only on
2284         Windows.
2285
2286         * Scripts/webkitpy/common/system/platforminfo.py:
2287         (PlatformInfo.__init__):
2288         (PlatformInfo.is_freebsd):
2289         (PlatformInfo._determine_os_name):
2290         * Scripts/webkitpy/common/system/platforminfo_mock.py:
2291         (MockPlatformInfo.is_freebsd):
2292         * Scripts/webkitpy/common/system/platforminfo_unittest.py:
2293         (fake_platform):
2294         (fake_platform.FakePlatformModule.release):
2295         (TestPlatformInfo.test_real_code):
2296         (TestPlatformInfo.test_os_name_and_wrappers):
2297         (TestPlatformInfo.test_os_version):
2298         (TestPlatformInfo.test_display_name):
2299         (TestPlatformInfo.test_total_bytes_memory):
2300         (TestPlatformInfo.test_free_bytes_memory):
2301         * Scripts/webkitpy/layout_tests/port/factory.py:
2302         (PortFactory._default_port):
2303
2304 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2305
2306         Make the table headers white-space:nowrap. This decreases the vertical size,
2307         which makes for more minimal embedding in garden-o-matic.
2308
2309         * TestResultServer/static-dashboards/flakiness_dashboard_tests.css:
2310         (.table-header-content *):
2311
2312 2012-05-03  Eric Seidel  <eric@webkit.org>
2313
2314         Split build-webkit's feature option list off into its own module in preparation for autogeneration
2315         https://bugs.webkit.org/show_bug.cgi?id=85548
2316
2317         Reviewed by Daniel Bates.
2318
2319         My Perl-fu is very weak.  This moves the feature option list off into a separate module
2320         file so that I can easily autogenerate just that file with the fancy new generate-feature-files command.
2321         It's a bit odd to have an array which we return out to build-webkit to modify, but it turns out
2322         not to matter.  Also, build-webkit doesn't actually use any of the variables we're using to
2323         store the results from the option parse, so those end up inaccessible from build-webkit (except by reference)
2324         which is actually cleaner.  Callers can't assume that they would get a new array every time, but someone
2325         with more Perl-fu than I could create something fancier than this if we need in the future.
2326
2327         * Scripts/build-webkit:
2328         * Scripts/webkitperl/FeatureList.pm: Added.
2329         (getFeatureOptionList):
2330
2331 2012-05-03  Dirk Pranke  <dpranke@chromium.org>
2332
2333         REGRESSION: run-webkit-tests failing on Chromium Mac
2334         https://bugs.webkit.org/show_bug.cgi?id=85459
2335
2336         Unreviewed, build fix.
2337
2338         Turns out select doesn't throw IOErrors, it has it's own kind of
2339         error :(. Trap that instead.
2340
2341         * Scripts/webkitpy/layout_tests/port/server_process.py:
2342         (ServerProcess._wait_for_data_and_update_buffers_using_select):
2343
2344 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2345
2346         [CMake] Rewrite FindCairo.cmake.
2347         https://bugs.webkit.org/show_bug.cgi?id=84895
2348
2349         Reviewed by Daniel Bates.
2350
2351         The old approach relied on pkg-config for finding Cairo (which
2352         introduced a dependency on pkg-config that could be avoided), used
2353         the LibFindMacros code that we should probably remove in the
2354         future and did not use the FindPackageHandleStandardArguments
2355         module.
2356
2357         Change all that by rewriting the module.
2358         - Use the pkg-config output optionally instead of requiring it
2359         like LibFindMacros did.
2360         - Remove the implicit dependency on FreeType which often found it
2361         the wrong way via pkg-config and without considering
2362         CMAKE_PREFIX_PATH.
2363         - Retrieve the Cairo version by looking at cairo-version.h instead
2364         of relying on pkg-config. It requires some additional code for
2365         checking if the desired version has been found, but that will not
2366         be needed once we start depending on CMake 2.8.3 or later.
2367
2368         The only downside is that FPHSA sets <UPPERCASED_NAME>_FOUND
2369         instead of <Name>_FOUND, and to keep things consistent
2370         Cairo_LIBRARIES and Cairo_INCLUDE_DIRS have become CAIRO_LIBRARIES
2371         and CAIRO_INCLUDE_DIRS.
2372
2373         * DumpRenderTree/efl/CMakeLists.txt: Use CAIRO_FOO instead of
2374         Cairo_FOO.
2375         * EWebLauncher/CMakeLists.txt: Ditto.
2376
2377 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2378
2379         Fix modifier and bug sorting on the flakiness dashboard
2380         https://bugs.webkit.org/show_bug.cgi?id=85554
2381
2382         Reviewed by Dirk Pranke.
2383
2384         The sorting assumes that the column header name matches a field
2385         in the testResults object and uses that field for sorting.
2386         Filter the bugs out of the modifiers list and put them in a bugs field
2387         and only put the non-bug modifiers in the modifiers list.
2388
2389         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2390         (populateExpectationsData):
2391         (htmlForSingleTestRow):
2392         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2393
2394 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2395
2396         Fix some state resetting when running unittests. HtmlForSingleTestRow
2397         was flaky.
2398
2399         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2400         (htmlForSingleTestRow):
2401         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2402         (resetGlobals):
2403
2404 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2405
2406         webkitpy: Use PlatformInfo wherever possible in the User class.
2407         https://bugs.webkit.org/show_bug.cgi?id=85549
2408
2409         Reviewed by Dirk Pranke.
2410
2411         Instead of relying on sys.platform to detect whether the current
2412         platform is a Mac, use the PlatformInfo class instead.
2413
2414         * Scripts/webkitpy/common/system/user.py:
2415         (User.__init__):
2416         (User.edit_changelog):
2417
2418 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2419
2420         Remove the percent failed column from the flakiness dashboard
2421         https://bugs.webkit.org/show_bug.cgi?id=85552
2422
2423         Reviewed by Dirk Pranke.
2424
2425         This column isn't useful enough to justify the visual noise.
2426
2427         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2428         (processMissingAndExtraExpectations):
2429         (tableHeaders):
2430         (htmlForSingleTestRow):
2431         (sortTests):
2432         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2433         (test):
2434
2435 2012-05-03  Eric Seidel  <eric@webkit.org>
2436
2437         --coverage should not not use the "feature" options system in build-webkit, similarly remove special-casing for WTF_URL
2438         https://bugs.webkit.org/show_bug.cgi?id=85542
2439
2440         Reviewed by Daniel Bates.
2441
2442         Remove these last two odd-men-out, so that we can generate the default feature options
2443         for build-webkit using the new generate-feature-files command.
2444         The Mac build system has some special casing for ENABLE_ defines which gets them
2445         defined as -D on the build command, even though initially they start out as
2446         environment variables ENABLE_FOO = ENABLE_FOO.  The Feature options in build-webkit
2447         depend on this support, but it doesn't work for WTF_USE_ defines like WTF_USE_WTF_URL.
2448         I'm told that WTF_URL is not actively being worked on, so just removing it for now.
2449         It should probably come back to life as an ENABLE_, or build-webkit or the Xcode projects
2450         should be made generically smarter about these USE_WTF defines.
2451         (There is still WTF_USE_TILED_BACKING_STORE in the feature list, but it only seems
2452         used by Qt.  If someone tried build-webkit --tiled-backing-store on Mac today it
2453         would not work!)
2454
2455         * Scripts/build-webkit:
2456
2457 2012-05-03  Eric Seidel  <eric@webkit.org>
2458
2459         Remove the last bits of Leopard support from our perl scripts
2460         https://bugs.webkit.org/show_bug.cgi?id=85545
2461
2462         Reviewed by Daniel Bates.
2463
2464         Chromium still supports Leopard (for a little while yet), but AppleMacWebKit dropped
2465         Leopard support several months ago.  Remove the last bits of isLeopard() branching
2466         in our Perl scripts, since no one is using this (Chromium has never used this Perl code).
2467
2468         * Scripts/build-webkit:
2469         * Scripts/old-run-webkit-tests:
2470         (countAndPrintLeaks):
2471         (captureSavedCrashLog):
2472         * Scripts/webkitdirs.pm:
2473         (determineArchitecture):
2474         (argumentsForRunAndDebugMacWebKitApp):
2475
2476 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2477
2478         [Chromium] Don't show the WebKit Linux ASAN bot on the flakiness dashboard.
2479         https://bugs.webkit.org/show_bug.cgi?id=85551
2480
2481         Reviewed by Dirk Pranke.
2482
2483         It's not uploading results and in the short term we don't plan to have it start.
2484
2485         * TestResultServer/static-dashboards/builders.js:
2486         (isChromiumWebkitTipOfTreeTestRunner):
2487         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2488
2489 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2490
2491         Show the test name in the dashboard if showChrome==false and we're showing multiple tests
2492         https://bugs.webkit.org/show_bug.cgi?id=85243
2493
2494         Reviewed by Dirk Pranke.
2495
2496         This happens for cases where a test name matches multiple tests.
2497         In practice, this only happens for virtual test suites (e.g. the canvas tests).
2498
2499         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2500         (htmlForIndividualTestOnAllBuildersWithChrome):
2501         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2502         (testHtmlForIndividualTestOnAllBuildersWithChromeNonexistant):
2503         (testHtmlForIndividualTestOnAllBuildersWithChrome):
2504         (testHtmlForIndividualTestOnAllBuildersWithChromeWebkitMaster):
2505         (testHtmlForIndividualTests):
2506
2507 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2508
2509         Port the flakiness dashboard unittests to qunit
2510         https://bugs.webkit.org/show_bug.cgi?id=85544
2511
2512         Reviewed by Adam Barth.
2513
2514         * TestResultServer/static-dashboards/dashboard_base.js:
2515         (appendJSONScriptElements):
2516         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2517         (resetGlobals):
2518         (runExpectationsTest):
2519         (test):
2520         * TestResultServer/static-dashboards/run-unittests.html: Added.
2521
2522 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2523
2524         webkitpy: Remove check for readline on Mac.
2525         https://bugs.webkit.org/show_bug.cgi?id=85547
2526
2527         Reviewed by Eric Seidel.
2528
2529         Python 2.5 is not supported by the WebKit project anymore, and
2530         Python's readline documentation does not mention the module not
2531         being available in the supported versions.
2532
2533         * Scripts/webkitpy/common/system/user.py:
2534
2535 2012-05-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2536
2537         [NRWT] Run performance tests with lock
2538         https://bugs.webkit.org/show_bug.cgi?id=78628
2539
2540         Reviewed by Tony Chang.
2541
2542         Locking performance tests (like we do for http tests) will force them
2543         to run in serial. This reduces the load of the machine when running perf
2544         tests and minimizes the chances of the tests to fail, specially
2545         timeouts.
2546
2547         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2548         (Manager.__init__):
2549         (Manager._is_perf_test):
2550         (Manager._test_requires_lock):
2551         * Scripts/webkitpy/layout_tests/port/http_lock.py:
2552
2553 2012-05-03  Dirk Pranke  <dpranke@chromium.org>
2554
2555         Attempt to fix a crash on SL ...
2556
2557         Unreviewed, build fix.
2558
2559         It looks like we may be getting an IOError raised during
2560         select() on chromium mac, and that error isn't being caught.
2561         Pushing it into the try/catch block will hopefully catch this
2562         and allow us to see why the subprocess is crashing.
2563
2564         * Scripts/webkitpy/layout_tests/port/server_process.py:
2565         (ServerProcess._wait_for_data_and_update_buffers_using_select):
2566
2567 2012-05-02  Jon Honeycutt  <jhoneycutt@apple.com>
2568
2569         WebFrameLoaderClient::dispatchWillSendSubmitEvent() needs to be
2570         implemented for WebKit2
2571         https://bugs.webkit.org/show_bug.cgi?id=84304
2572
2573         Reviewed by Jessie Berlin.
2574
2575         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2576         Added new files to project.
2577
2578         * TestWebKitAPI/Tests/WebKit2/WillSendSubmitEvent.cpp: Added.
2579         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
2580         Look for the "DidReceiveWillSendSubmitEvent" message. The message body
2581         will be the dictionary of form control names and values passed to the
2582         injected bundle's willSendSubmitEvent handler; look for the pairs
2583         "textField" -> "text field" and "passwordField" -> "password field".
2584         Check that "hiddenField" is not included.
2585         (TestWebKitAPI::setInjectedBundleClient):
2586         (TestWebKitAPI::TEST):
2587
2588         * TestWebKitAPI/Tests/WebKit2/WillSendSubmitEvent_Bundle.cpp: Added.
2589         (TestWebKitAPI::willSendSubmitEvent):
2590         Post a message to the TestWebKitAPI process with the dictionary of form
2591         control names and values that we received.
2592         (TestWebKitAPI::WillSendSubmitEventTest::WillSendSubmitEventTest):
2593         (TestWebKitAPI::WillSendSubmitEventTest::didCreatePage):
2594         Set up the page's form client.
2595
2596         * TestWebKitAPI/Tests/WebKit2/auto-submitting-form.html: Added.
2597
2598 2012-05-03  Eric Seidel  <eric@webkit.org>
2599
2600         Remove dead code from build-webkit for working around old XCode dependency bug
2601         https://bugs.webkit.org/show_bug.cgi?id=85535
2602
2603         Reviewed by Adam Barth.
2604
2605         Xcode used to have a bug whereby it didn't notice if you changed the #defines
2606         passed on the command line to xcodebuild.  I wrote code to work around this
2607         while bringing up SVG support 5-6 years ago.  I think we can finally remove this
2608         code from build-webkit, since SVG is the only feature flag to ever use it
2609         (modern in-progress features like MathML don't seem to be using the support)
2610         an I suspect the XCode bug has long been fixed.
2611
2612         * Scripts/build-webkit:
2613         * Scripts/webkitperl/features.pm:
2614
2615 2012-05-03  Andy Estes  <aestes@apple.com>
2616
2617         InjectedBundleControllerMac.mm needs to include config.h
2618         https://bugs.webkit.org/show_bug.cgi?id=85531
2619
2620         Reviewed by Andreas Kling.
2621
2622         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm:
2623
2624 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2625
2626         webkitpy: Use PlatformInfo whenever possible in server_process.py.
2627         https://bugs.webkit.org/show_bug.cgi?id=85516
2628
2629         Reviewed by Eric Seidel.
2630
2631         Eliminate direct access to sys.platform wherever possible by using
2632         PlatformInfo.is_win() instead.
2633
2634         * Scripts/webkitpy/layout_tests/port/server_process.py:
2635         (ServerProcess.__init__): Remove the executive parameter as it was
2636         not passed by any caller, obtain a SystemHost from port_obj
2637         instead.
2638         (ServerProcess._log): Style, remove extra empty line.
2639         (ServerProcess._start):
2640         (ServerProcess.stop):
2641         (ServerProcess.kill):
2642         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
2643         (TrivialMockPort.__init__): Create a MockSystemHost used by
2644         ServerProcess.
2645         (TestServerProcess.test_broken_pipe): Test different platforms.
2646
2647 2012-05-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2648
2649         webkitpy: Style checker broken when committing a PNG
2650         https://bugs.webkit.org/show_bug.cgi?id=85504
2651
2652         Reviewed by Dirk Pranke.
2653
2654         * Scripts/webkitpy/style/checkers/png.py:
2655         (PNGChecker.__init__):
2656
2657 2012-05-03  Philippe Normand  <pnormand@igalia.com>
2658
2659         Unreviewed, fix another call to join(), similar to the ones landed
2660         in r115975.
2661
2662         * Scripts/webkitdirs.pm:
2663         (jhbuildConfigurationChanged):
2664
2665 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2666
2667         webkitpy: Remove sys.platform accesses in CrashLogs.
2668         https://bugs.webkit.org/show_bug.cgi?id=85436
2669
2670         Reviewed by Dirk Pranke.
2671
2672         Instead of directly checking for sys.platform == 'darwin', use a
2673         SystemHost and call PlatformInfo.is_mac().
2674
2675         * Scripts/webkitpy/common/system/crashlogs.py:
2676         (CrashLogs.__init__):
2677         (CrashLogs.find_newest_log):
2678         (CrashLogs._log_directory_darwin):
2679         (CrashLogs._find_newest_log_darwin):
2680         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
2681         (CrashLogsTest.test_find_log_darwin):
2682         * Scripts/webkitpy/layout_tests/port/mac.py:
2683         (MacPort._get_crash_log):
2684         * Scripts/webkitpy/layout_tests/port/test.py:
2685         (TestDriver.run_test):
2686         * Scripts/webkitpy/tool/commands/queries.py:
2687         (execute):
2688
2689 2012-05-03  Philippe Normand  <pnormand@igalia.com>
2690
2691         [GTK] join() is wrongly used in webkitdirs.pm since r115532
2692         https://bugs.webkit.org/show_bug.cgi?id=85501
2693
2694         Reviewed by Martin Robinson.
2695
2696         * Scripts/webkitdirs.pm:
2697         (getJhbuildPath):
2698         (buildAutotoolsProject):
2699
2700 2012-05-03  Zan Dobersek  <zandobersek@gmail.com>
2701
2702         [Gtk] ImageDiff crashes when it can't open a display
2703         https://bugs.webkit.org/show_bug.cgi?id=85476
2704
2705         Reviewed by Martin Robinson.
2706
2707         Replace gdk_init() initialization call with the g_type_init()
2708         call. GdkPixbufLoader doesn't actually need the current call to be
2709         made and this change will also remove crashes when ImageDiff
2710         is started in an environment without a DISPLAY env.
2711
2712         * DumpRenderTree/gtk/ImageDiff.cpp:
2713         (main):
2714
2715 2012-05-03  Zan Dobersek  <zandobersek@gmail.com>
2716
2717         ImageDiff should be run inside a properly established environment
2718         https://bugs.webkit.org/show_bug.cgi?id=85292
2719
2720         Reviewed by Martin Robinson.
2721
2722         When creating the ImageDiff server process, pass along a freshly-set-up environment
2723         in which the process should be executed in.
2724
2725         * Scripts/webkitpy/layout_tests/port/webkit.py:
2726         (WebKitPort._start_image_diff_process):
2727
2728 2012-05-03  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2729
2730         [Qt] Allow the web process and WTR to be paused on startup
2731
2732         Makes it easier to debug the web process or run-webkit-tests -2, as you
2733         have ample time to attach gdb to the process.
2734
2735         Reviewed by Simon Hausmann.
2736
2737         * Scripts/webkitpy/layout_tests/port/qt.py:
2738         (QtPort.setup_environ_for_server):
2739         * WebKitTestRunner/qt/main.cpp:
2740         (main):
2741
2742 2012-05-03  Csaba Osztrogonác  <ossy@webkit.org>
2743
2744         Password protect "Stop build" button
2745         https://bugs.webkit.org/show_bug.cgi?id=81982
2746
2747         Reviewed by Ryosuke Niwa.
2748
2749         Only authenticated users should be able to use stopBuild, stopAllBuilds, cancelPendingBuild
2750         features to prevent killing builds on build.webkit.org by SPAM bots.
2751
2752         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2753
2754 2012-05-03  Dongwoo Im  <dw.im@samsung.com>
2755
2756         [DRT] Remove the name of parameter, visibility, from the setPageVisibility stub implementations.
2757         https://bugs.webkit.org/show_bug.cgi?id=85468
2758
2759         Reviewed by Nikolas Zimmermann.
2760
2761         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
2762         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2763         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2764         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2765         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2766         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2767         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2768         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2769         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2770         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2771
2772 2012-05-03  Stephanie Lewis  <slewis@apple.com>
2773
2774         https://bugs.webkit.org/show_bug.cgi?id=85450 unbounded growth of JSDOMWindowShells loading pages in the same window
2775         <rdar://problem/11320059> REGRESSION (r115083): PLT3 shows linear memory growth and gets slower with each run
2776
2777         Reviewed by Brady Eidson.
2778
2779         The API added for DOMWindowExtension, didCreateGlobalObjectForFrame, would create a global object
2780         for every world, even those that did not need the callback.  This had the side effect of creating a
2781         JSDOMWindowShell that the associated world didn't necessarily know to clean up.  Instead of creating
2782         unnecessary objects change the API to globalObjectIsAvailableForFrame and do not pass the global object
2783         in the API.  The object can be accessed later by those worlds which require it.
2784
2785         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic.cpp:
2786         (TestWebKitAPI):
2787         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
2788         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
2789         (TestWebKitAPI):
2790         (DOMWindowExtensionBasic):
2791         (TestWebKitAPI::DOMWindowExtensionBasic::didCreatePage):
2792         (TestWebKitAPI::DOMWindowExtensionBasic::globalObjectIsAvailableForFrame):
2793         (TestWebKitAPI::globalObjectIsAvailableForFrameCallback):
2794
2795 2012-05-03  Nikolas Zimmermann  <nzimmermann@rim.com>
2796
2797         Not reviewed. Fix mac build, it was missing setPageVisibility/resetPageVisibility stub implementations.
2798
2799         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2800         (LayoutTestController::resetPageVisibility):
2801         (LayoutTestController::setPageVisibility):
2802
2803 2012-05-02  Alexander Færøy  <ahf@0x90.dk>
2804
2805         Rename deviceDPI to devicePixelRatio
2806         https://bugs.webkit.org/show_bug.cgi?id=85049
2807
2808         Reviewed by Kenneth Rohde Christiansen.
2809
2810         Use 1.5 as device pixel ratio for the Qt MiniBrowser.
2811
2812         * MiniBrowser/qt/qml/BrowserWindow.qml:
2813
2814 2012-05-03  Dongwoo Im  <dw.im@samsung.com>
2815
2816         [EFL][DRT] Implement the LayoutTestController's methods related to the Page Visibility API.
2817         https://bugs.webkit.org/show_bug.cgi?id=85347
2818
2819         Reviewed by Nikolas Zimmermann.
2820
2821         * DumpRenderTree/LayoutTestController.h:
2822         (LayoutTestController):
2823         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
2824         (LayoutTestController::resetPageVisibility): Not implemented.
2825         (LayoutTestController::setPageVisibility): Not implemented.
2826         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2827         (LayoutTestController::resetPageVisibility): Reset the page visibility.
2828         (LayoutTestController::setPageVisibility): Set the page visibility.
2829         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2830         (LayoutTestController::resetPageVisibility): Not implemented.
2831         (LayoutTestController::setPageVisibility): Not implemented.
2832         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2833         (LayoutTestController::resetPageVisibility): Not implemented.
2834         (LayoutTestController::setPageVisibility): Not implemented.
2835         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2836         (LayoutTestController::resetPageVisibility): Not implemented.
2837         (LayoutTestController::setPageVisibility): Not implemented.
2838         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2839         (LayoutTestController::resetPageVisibility): Not implemented.
2840         (LayoutTestController::setPageVisibility): Not implemented.
2841
2842 2012-05-02  Sheriff Bot  <webkit.review.bot@gmail.com>
2843
2844         Unreviewed, rolling out r115918.
2845         http://trac.webkit.org/changeset/115918
2846         https://bugs.webkit.org/show_bug.cgi?id=85452
2847
2848         "Broke test-webkitpy on Apple Lion bots" (Requested by estes
2849         on #webkit).
2850
2851         * Scripts/webkitpy/common/system/crashlogs.py:
2852         (CrashLogs.__init__):
2853         (CrashLogs.find_newest_log):
2854         (CrashLogs._log_directory_darwin):
2855         (CrashLogs._find_newest_log_darwin):
2856         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
2857         (CrashLogsTest.test_find_log_darwin):
2858         * Scripts/webkitpy/layout_tests/port/mac.py:
2859         (MacPort._get_crash_log):
2860         * Scripts/webkitpy/layout_tests/port/test.py:
2861         (TestDriver.run_test):
2862         * Scripts/webkitpy/tool/commands/queries.py:
2863         (execute):
2864
2865 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
2866
2867         webkitpy: Remove sys.platform accesses in CrashLogs.
2868         https://bugs.webkit.org/show_bug.cgi?id=85436
2869
2870         Reviewed by Dirk Pranke.
2871
2872         Instead of directly checking for sys.platform == 'darwin', use a
2873         SystemHost and call PlatformInfo.is_mac().
2874
2875         * Scripts/webkitpy/common/system/crashlogs.py:
2876         (CrashLogs.__init__):
2877         (CrashLogs.find_newest_log):
2878         (CrashLogs._log_directory_darwin):
2879         (CrashLogs._find_newest_log_darwin):
2880         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
2881         (CrashLogsTest.test_find_log_darwin):
2882         * Scripts/webkitpy/layout_tests/port/mac.py:
2883         (MacPort._get_crash_log):
2884         * Scripts/webkitpy/layout_tests/port/test.py:
2885         (TestDriver.run_test):
2886         * Scripts/webkitpy/tool/commands/queries.py:
2887         (execute):
2888
2889 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
2890
2891         webkitpy: Special-case win32 calls in file_lock and executive.
2892         https://bugs.webkit.org/show_bug.cgi?id=85431
2893
2894         Reviewed by Dirk Pranke.
2895
2896         Simplify accesses to sys.platform by only checking if sys.platform
2897         is "win32" and removing explicit checks for the else case.
2898
2899         * Scripts/webkitpy/common/system/executive.py:
2900         (Executive.check_running_pid):
2901         * Scripts/webkitpy/common/system/file_lock.py:
2902         (FileLock._create_lock):
2903         (FileLock._remove_lock):
2904
2905 2012-05-02  Ojan Vafai  <ojan@chromium.org>
2906
2907         Remove pink from flakiness dashboard
2908         https://bugs.webkit.org/show_bug.cgi?id=85422
2909
2910         Reviewed by Dirk Pranke.
2911
2912         The pink indicates that the test doesn't match it's listing in test_expectations.txt.
2913         In practice, this just added noise and confused people.
2914
2915         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2916         (htmlForSingleTestRow):
2917         (showLegend):
2918         * TestResultServer/static-dashboards/flakiness_dashboard_tests.css:
2919         (#legend-contents *):
2920         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2921         (testHtmlForSingleTestRow):
2922
2923 2012-05-02  Gustavo Noronha Silva  <gns@gnome.org>
2924
2925         Unreviewed. Fix mistake I committed when applying review comments.
2926
2927         * jhbuild/jhbuild-wrapper: dependencies_path was getting 'Root'
2928         appended to its path, so jhbuild itself was being cloned and
2929         installed in the wrong place when using WEBKITOUTPUTDIR.
2930
2931 2012-05-02  Dirk Pranke  <dpranke@chromium.org>
2932
2933         nrwt: make ServerProcess work on chromium win
2934         https://bugs.webkit.org/show_bug.cgi?id=85333
2935
2936         Reviewed by Ojan Vafai.
2937
2938         This change implements a poor man's select() that will
2939         slow-spin doing non-blocking reads on the stdout and stderr
2940         named pipes connecting the worker to the driver. Seems to work
2941         and I have yet to see much overhead or ill effects but it
2942         probably needs more testing.
2943
2944         * Scripts/webkitpy/layout_tests/port/chromium.py:
2945         (ChromiumDriver.__init__):
2946         * Scripts/webkitpy/layout_tests/port/server_process.py:
2947         (ServerProcess._start):
2948         (ServerProcess._wait_for_data_and_update_buffers_using_select):
2949         (ServerProcess._wait_for_data_and_update_buffers_using_win32_apis):
2950         (ServerProcess):
2951         (ServerProcess._non_blocking_read_win32):
2952         (ServerProcess._read):
2953         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
2954         (TestServerProcess.test_basic): Added.
2955
2956 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
2957
2958         webkitpy: Use a SystemHost instead of PlatformInfo in the PNG checker.
2959         https://bugs.webkit.org/show_bug.cgi?id=85426
2960
2961         Reviewed by Dirk Pranke.
2962
2963         r115880 replaced the direct use of sys.platform with a
2964         PlatformInfo object. Dirk Pranke later told me SystemHost objects
2965         should be preferred, as nothing out of webkitpy.common.system
2966         should create its own PlatformInfo instances.
2967
2968         * Scripts/webkitpy/common/system/systemhost_mock.py:
2969         (MockSystemHost.__init__): Make it possible to specify a custom
2970         filesystem object.
2971         * Scripts/webkitpy/style/checkers/png.py:
2972         (PNGChecker.__init__): Accept a host object instead of a
2973         filesystem one, which is now obtained from host.
2974         (PNGChecker._config_file_path): Retrieve PlatformInfo from
2975         SystemHost.
2976         * Scripts/webkitpy/style/checkers/png_unittest.py:
2977         (PNGCheckerTest.test_init): Adjust calls to PNGChecker to its new
2978         constructor signature.
2979         (PNGCheckerTest.test_check): Ditto.
2980
2981 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
2982
2983         webkitpy: Remove direct checks for sys.platform == 'darwin' in the layout test classes.
2984         https://bugs.webkit.org/show_bug.cgi?id=85424
2985
2986         Reviewed by Tony Chang.
2987
2988         We should reduce direct usage of sys.platform in webkitpy and use
2989         PlatformInfo instead.
2990
2991         * Scripts/webkitpy/layout_tests/servers/http_server.py:
2992         (Lighttpd._prepare_config): Call PlatformInfo.is_mac() instead of
2993         checking for 'darwin' directly.
2994         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
2995         (HttpServerBase.__init__): Ditto.
2996
2997 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
2998
2999         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
3000         https://bugs.webkit.org/show_bug.cgi?id=85419
3001
3002         Reviewed by Dirk Pranke.
3003
3004         As pointed out by Dirk Pranke, a PlatformInfo is not supposed to
3005         be created directly, we should should a SystemHost instead.
3006
3007         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3008         (MainTest.setUp): Create a SystemHost and obtain the PlatformInfo
3009         object from it instead of creating PlatformInfo directly.
3010
3011 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3012
3013         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
3014         https://bugs.webkit.org/show_bug.cgi?id=85419
3015
3016         Rubber-stamped by Dirk Pranke.
3017
3018         Follow-up patch to r115887 which fixes some capitalization problems
3019         pointed out in comments 2 and 4.
3020
3021         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3022         (MainTest.setUp):
3023         (MainTest.test_child_processes_2):
3024         (MainTest.test_child_processes_min):
3025         (MainTest.test_exception_raised):
3026         (MainTest.test_keyboard_interrupt):
3027
3028 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3029
3030         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
3031         https://bugs.webkit.org/show_bug.cgi?id=85419
3032
3033         Reviewed by Dirk Pranke.
3034
3035         Another part of the work towards replacing direct sys.platform
3036         usage with PlatformInfo.
3037
3038         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3039         (MainTest.setUp): Move the setting of SHOULD_TEST_PROCESSES
3040         here. Even though it is costlier, unittest.setUpClass was
3041         introduced in Python 2.7.
3042         (MainTest.test_child_processes_2): Use self.SHOULD_TEST_PROCESSES.
3043         (MainTest.test_child_processes_min): Ditto.
3044         (MainTest.test_exception_raised): Ditto.
3045         (MainTest.test_keyboard_interrupt): Ditto.
3046         (MainTest.test_crash_log): Check for PlatformInfo.is_mac().
3047         (MainTest.test_web_process_crash_log): Ditto, and remove duplicate
3048         check at the end.
3049         (MainTest.test_no_http_tests): Remove check for (cygwin && Python
3050         2.5.2), as we only support Python 2.6+.
3051
3052 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3053
3054         webkitpy: Remove direct sys.platform usage in the PNG checker.
3055         https://bugs.webkit.org/show_bug.cgi?id=85407
3056
3057         Reviewed by Tony Chang.
3058
3059         Work towards replacing direct sys.platform usage with
3060         PlatformInfo.
3061
3062         * Scripts/webkitpy/common/system/platforminfo_mock.py:
3063         (MockPlatformInfo.__init__): Accept os_name and os_version as
3064         arguments.
3065         * Scripts/webkitpy/style/checkers/png.py:
3066         (PNGChecker.__init__): Create a PlatformInfo object instead of
3067         using sys.platform when needed.
3068         (PNGChecker._config_file_path): Check for the current platform via
3069         PlatformInfo.
3070         * Scripts/webkitpy/style/checkers/png_unittest.py:
3071         (PNGCheckerTest.test_check): Update tests after the changes above.
3072
3073 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3074
3075         [watchlist] Update watchlist rules for EFL-related patches.
3076         http://bugs.webkit.org/show_bug.cgi?id=85160
3077
3078         Reviewed by Eric Seidel.
3079
3080         * Scripts/webkitpy/common/config/watchlist: Monitor
3081         Source/WTF/wtf/efl and LayoutTests/platform/efl as well.
3082
3083 2012-05-02  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3084
3085         Revert r115191 - "Make the web view's url property follow the active url"
3086
3087         It uncovered/caused issues in the icon-implementation that can't be
3088         easily fixed, so rolling out instead.
3089
3090         https://bugs.webkit.org/show_bug.cgi?id=77554
3091
3092         Rubber-stamped by Simon Hausmann.
3093
3094         * MiniBrowser/qt/qml/BrowserWindow.qml:
3095         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
3096         (TestWebKitAPI::TEST):
3097
3098 2012-05-02  Zan Dobersek  <zandobersek@gmail.com>
3099
3100         Reftest mismatches are providing undefined diff percent in results.html
3101         https://bugs.webkit.org/show_bug.cgi?id=85283
3102
3103         Reviewed by Ojan Vafai.
3104
3105         When writing test results for a reftest mismatch, also take into account the
3106         difference percent that is produced when generating image diff. This value is
3107         then used when creating test results in the form of dictionaries, stored under
3108         the 'image_diff_percent' key. It ultimately ends up in JSON results data,
3109         providing additional information about the reftest difference.
3110
3111         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3112         (interpret_test_failures): Store the diff percent for reftest failure under the
3113         'image_diff_percent' key.
3114         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
3115         (write_test_result): Also set the new diff_percent attribute of the reftest
3116         mismatch failure to the value of the difference percent ImageDiff outputs.
3117         * Scripts/webkitpy/layout_tests/models/test_failures.py:
3118         (FailureReftestMismatch.__init__): Add the new diff_percent attribute with the
3119         default value of None.
3120         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3121         (EndToEndTest.test_reftest_with_two_notrefs): Update the test case expected results.
3122
3123 2012-05-02  Terry Anderson  <tdanderson@chromium.org>
3124
3125         Addition of a contributor to committers.py
3126         https://bugs.webkit.org/show_bug.cgi?id=85324
3127
3128         Reviewed by Unreviewed.
3129
3130         * Scripts/webkitpy/common/config/committers.py:
3131
3132 2012-05-02  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3133
3134         [EFL] EFL's LayoutTestController overridePreference implementation
3135         https://bugs.webkit.org/show_bug.cgi?id=83007
3136
3137         Reviewed by Antonio Gomes.
3138
3139         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3140         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
3141         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3142         (toBool): aux function converting JSStringRef to bool
3143         (toInt): aux function converting JSStringRef to int
3144         (LayoutTestController::overridePreference): added implementation
3145
3146 2012-05-01  Raphael Kubo da Costa  <rakuco@webkit.org>
3147
3148         webkitpy: Fix sys.platform check in find_files_unittest.
3149         https://bugs.webkit.org/show_bug.cgi?id=85339
3150
3151         Reviewed by Ryosuke Niwa.
3152
3153         Follow-up to r101618 -- 'win' is not a valid sys.platform value,
3154         'win32' should be used instead. The check always returned false
3155         due to this, so assert_normalize() was never called.
3156
3157         * Scripts/webkitpy/common/find_files_unittest.py:
3158         (TestWinNormalize.test_win): Add the missing import for the
3159         FileSystem class and fix the sys.platform check.
3160
3161 2012-05-02  Zan Dobersek  <zandobersek@gmail.com>
3162
3163         [GTK] drag&drop related failures
3164         https://bugs.webkit.org/show_bug.cgi?id=84675
3165
3166         Reviewed by Philippe Normand.
3167
3168         Rather than checking options to see whether we're running pixel tests
3169         the pixel_tests argument passed to GtkDriver._start should be used as
3170         it also covers cases when pixel tests should be run because of testing
3171         reftests. This again creates different Xvfb processes with proper
3172         display IDs for a worker that's performing both ordinary and pixel
3173         tests, as was the case before changes made in r114788.
3174
3175         * Scripts/webkitpy/layout_tests/port/gtk.py:
3176         (GtkDriver._start):
3177
3178 2012-05-02  Simon Hausmann  <simon.hausmann@nokia.com>
3179
3180         [Qt] Enable device orientation/motion with Qt 5
3181
3182         Reviewed by Tor Arne Vestbø.
3183
3184         Enable device orientation/motion features with Qt 5 if Qt Sensors is available.
3185
3186         * qmake/mkspecs/features/features.prf:
3187
3188 2012-05-01  Raphael Kubo da Costa  <rakuco@webkit.org>
3189
3190         webkitpy: Remove unused platform imports.
3191         https://bugs.webkit.org/show_bug.cgi?id=85341
3192
3193         Reviewed by Eric Seidel.
3194
3195         Remove a few uses of `import platform' which were either rendered
3196         unneeded after r105931 or were never needed at all.
3197
3198         * Scripts/webkitpy/common/system/executive.py:
3199         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
3200
3201 2012-05-01  Terry Anderson  <tdanderson@chromium.org>
3202
3203         [chromium] Accept four parameters when dispatching a WebInputEvent::GestureTap in chromium DRT eventSender
3204         https://bugs.webkit.org/show_bug.cgi?id=85289
3205
3206         Reviewed by Eric Seidel.
3207
3208         * DumpRenderTree/chromium/EventSender.cpp:
3209         (EventSender::gestureEvent):
3210             This change will allow for testing of WK85101. Currently
3211             no tests use the extra parameters deltaX and deltaY. Added
3212             a check for the size of |arguments|.
3213
3214 2012-05-01  Eric Seidel  <eric@webkit.org>
3215
3216         Add myself to the watchlist to watch all rendering changes, as well as sort the watchers in the list
3217         https://bugs.webkit.org/show_bug.cgi?id=85305
3218
3219         Reviewed by Adam Barth.
3220
3221         * Scripts/webkitpy/common/config/watchlist:
3222
3223 2012-05-01  Sheriff Bot  <webkit.review.bot@gmail.com>
3224
3225         Unreviewed, rolling out r115735.
3226         http://trac.webkit.org/changeset/115735
3227         https://bugs.webkit.org/show_bug.cgi?id=85314
3228
3229         cause two tests to crash due to an ASSERTION failure
3230         (Requested by zhenyao on #webkit).
3231
3232         * DumpRenderTree/chromium/EventSender.cpp:
3233         (EventSender::gestureEvent):
3234
3235 2012-05-01  Terry Anderson  <tdanderson@chromium.org>
3236
3237         [chromium] Accept four parameters when dispatching a WebInputEvent::GestureTap in chromium DRT eventSender
3238         https://bugs.webkit.org/show_bug.cgi?id=85289
3239
3240         Reviewed by Eric Seidel.
3241
3242         * DumpRenderTree/chromium/EventSender.cpp:
3243         (EventSender::gestureEvent):
3244             This change will allow for testing of WK85101. Currently no tests use the extra
3245             parameters deltaX and deltaY.
3246
3247 2012-05-01  Zan Dobersek  <zandobersek@gmail.com>
3248
3249         Print out stderr output of ImageDiff if it is present
3250         https://bugs.webkit.org/show_bug.cgi?id=85285
3251
3252         Reviewed by Philippe Normand.
3253
3254         Log the stderr output of ImageDiff as a warning if present. This is
3255         especially helpful when ImageDiff crashes and stderr data might contain
3256         relevant information about what caused the crash.
3257
3258         * Scripts/webkitpy/layout_tests/port/webkit.py:
3259         (WebKitPort._read_image_diff):
3260
3261 2012-05-01  Kenneth Rohde Christiansen  <kenneth@webkit.org>
3262
3263         [Qt] Add an experimental extension to set the min. contents width
3264         https://bugs.webkit.org/show_bug.cgi?id=85281
3265
3266         Reviewed by Antonio Gomes.
3267
3268         * MiniBrowser/qt/qml/BrowserWindow.qml:
3269
3270         Set the value of preferredMinimumContentsWidth to 980.
3271
3272 2012-04-30  Ojan Vafai  <ojan@chromium.org>
3273
3274         Fix tests to not depend on the actual list of builders at build.chromium.org.
3275
3276         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
3277         (testHtmlForIndividualTestOnAllBuildersWithChrome):
3278         (testHtmlForIndividualTestOnAllBuildersWithChromeWebkitMaster):
3279
3280 2012-04-29  Sam Weinig  <sam@webkit.org>
3281
3282         Remove BlobBuilder
3283         https://bugs.webkit.org/show_bug.cgi?id=84036
3284
3285         Reviewed by Anders Carlsson.
3286
3287         * Scripts/build-webkit:
3288         * qmake/mkspecs/features/features.prf:
3289         Make exposing the WebKitBlobBuilder JS object conditional on a new
3290         ENABLE_LEGACY_WEBKIT_BLOB_BUILDER flag. Don't enable this for the Mac,
3291         but do for everyone else.
3292
3293 2012-04-30  Amruth Raj Padmanabhuni  <amruthraj@motorola.com>
3294
3295         [GTK] Add glib as a dependency to build glib-networking
3296         https://bugs.webkit.org/show_bug.cgi?id=85181
3297
3298         Reviewed by Martin Robinson.
3299
3300         * gtk/jhbuild.modules:
3301
3302 2012-04-30  Alexis Menard  <alexis.menard@openbossa.org>
3303
3304         Move myself to reviewers.
3305
3306         Reviewed by Andreas Kling.
3307
3308         * Scripts/webkitpy/common/config/committers.py:
3309
3310 2012-04-30  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
3311
3312         Unreviewed, move myself to reviewers
3313
3314         * Scripts/webkitpy/common/config/committers.py:
3315
3316 2012-04-30  Carlos Garcia Campos  <cgarcia@igalia.com>
3317
3318         Unreviewed. Fix make distcheck.
3319
3320         * TestWebKitAPI/GNUmakefile.am: Add missing headers to compilation
3321         and resources to EXTRA_DIST.
3322
3323 2012-04-30  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
3324
3325         [GTK] DRT support for layoutTestController.removeOriginAccessWhitelistEntry
3326         https://bugs.webkit.org/show_bug.cgi?id=85156
3327
3328         Reviewed by Martin Robinson.
3329
3330         Add missing implementation removeOriginAccessWhitelistEntry to
3331         GTK's LayoutTestController.
3332
3333         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3334         (LayoutTestController::removeOriginAccessWhitelistEntry):
3335
3336 2012-04-29  Maciej Stachowiak  <mjs@apple.com>
3337
3338         REGRESSION: On Lion, run-webkit-tests changes the display color profile even when not running pixel tests
3339         https://bugs.webkit.org/show_bug.cgi?id=81729
3340
3341         Reviewed by Dan Bernstein.
3342
3343         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3344         (Manager._set_up_run): Only launch LayoutTestHelper when running pixel tests.
3345
3346 2012-04-29  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
3347
3348         [GTK] DRT needs an implementation of layoutTestController.setDomainRelaxationForbiddenForURLScheme
3349         https://bugs.webkit.org/show_bug.cgi?id=85131
3350
3351         Reviewed by Martin Robinson.
3352
3353         Add missing implementation setDomainRelaxationForbiddenForURLScheme to
3354         GTK's LayoutTestController.
3355
3356         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3357         (LayoutTestController::setDomainRelaxationForbiddenForURLScheme):
3358
3359 2012-04-28  Sam Weinig  <sam@webkit.org>
3360
3361         Smooth scrolling needs a new key
3362         <rdar://problem/11331632>
3363
3364         Reviewed by Dan Bernstein.
3365
3366         * DumpRenderTree/mac/DumpRenderTree.mm:
3367         (resetDefaultsToConsistentValues):
3368         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm:
3369         (TestWebKitAPI::InjectedBundleController::platformInitialize):
3370         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
3371         (WTR::InjectedBundle::platformInitialize):
3372         Update for new key.
3373
3374 2012-04-28  Zan Dobersek  <zandobersek@gmail.com>
3375
3376         [WK2] http/tests/navigation/anchor-frames-gbk.html fails
3377         https://bugs.webkit.org/show_bug.cgi?id=76896
3378
3379         Reviewed by Darin Adler.
3380
3381         Only dump as text if currently dumping render tree. This ensures that
3382         calling dumpAsText in a test with dumpChildFramesAsText already being
3383         called doesn't override the first decision, possibly making tests fail.
3384
3385         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3386         (WTR::LayoutTestController::dumpAsText):
3387
3388 2012-04-28  Dan Winship  <danw@gnome.org>
3389
3390         Unreviewed; Add me to the SoupNetwork watchlist
3391
3392         * Scripts/webkitpy/common/config/watchlist:
3393
3394 2012-04-27  Gustavo Noronha Silva  <gns@gnome.org>
3395
3396         [GTK] jhbuild cloning is not following WEBKITOUTPUTDIR.
3397         https://bugs.webkit.org/show_bug.cgi?id=76161
3398
3399         Reviewed by Martin Robinson.
3400
3401         * Scripts/webkitdirs.pm:
3402         (getJhbuildPath): New method to obtain the jhbuild base directory,
3403         using the product base directory
3404         (jhbuildConfigurationChanged): Use the new method
3405         (buildAutotoolsProject): Ditto.
3406         * efl/jhbuildrc: Use WEBKITOUTPUTDIR when calculating the path.
3407         * gtk/jhbuildrc: Ditto.
3408         * jhbuild/jhbuild-wrapper: Ditto.
3409
3410 2012-04-27  Scott Graham  <scottmg@chromium.org>
3411
3412         [chromium] make case of npTestNetscapePlugIn.dll match case in copy rule
3413         https://bugs.webkit.org/show_bug.cgi?id=85083
3414
3415         Reviewed by Dirk Pranke.
3416
3417         Make output product_name for npTestNetscapePlugIn.dll match the case
3418         of the copy_TestNetscapePlugIn rule. This is required for ninja, which
3419         is more particular about case matching in rules than the VS build.
3420
3421         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
3422
3423 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3424
3425         [chromium] use "drt-style" output, not "test-shell-style" output, on mac and linux DRT
3426         https://bugs.webkit.org/show_bug.cgi?id=84917
3427
3428         Unreviewed, build fix.
3429
3430         Reland the change in r115453 with a fix for chromium win.
3431         The logic in ChromiumDriver was busted on windows, causing DRT
3432         to get launched without the --test-shell flag.
3433
3434         * Scripts/webkitpy/layout_tests/port/chromium.py:
3435         (ChromiumDriver):
3436         (ChromiumDriver.__init__):
3437         (ChromiumDriver._wrapper_options):
3438         (ChromiumDriver.cmd_line):
3439         (ChromiumDriver._start):
3440         (ChromiumDriver.has_crashed):
3441         (ChromiumDriver.run_test):
3442         (ChromiumDriver.stop):
3443         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3444         (ChromiumDriverTest.setUp):
3445         (ChromiumDriverTest.test_stop):
3446         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
3447         (ChromiumDriverTest.test_two_drivers):
3448
3449 2012-04-27  Peter Beverloo  <peter@chromium.org>
3450
3451         Add John Grabowski as a non-committer to committers.py
3452         https://bugs.webkit.org/show_bug.cgi?id=85080
3453
3454         Reviewed by Dirk Pranke.
3455
3456         He's being CC'ed often enough because of his involvement in the Chromium
3457         side of Chrome for Android upstreaming.
3458
3459         * Scripts/webkitpy/common/config/committers.py:
3460           add jrg@chromium.org
3461
3462 2012-04-27  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
3463
3464         [Qt] Fix minimal build.
3465         https://bugs.webkit.org/show_bug.cgi?id=85045
3466
3467         Reviewed by Tor Arne Vestbø.
3468
3469         Do not set conflicting values of ENABLE_XSLT.
3470
3471         * qmake/mkspecs/features/features.prf:
3472
3473 2012-04-27  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
3474
3475         [EFL][DRT] EFL should sanitize pulseaudio module configuration too
3476         https://bugs.webkit.org/show_bug.cgi?id=85047
3477
3478         Reviewed by Dirk Pranke.
3479
3480         PulseAudio test setup sanitization shared between EFL and GTK.
3481
3482         * Scripts/webkitpy/layout_tests/port/efl.py:
3483         (EflPort):
3484         (EflPort.setup_test_run):
3485         (EflPort.clean_up_test_run):
3486         * Scripts/webkitpy/layout_tests/port/gtk.py:
3487         (GtkPort): Moving pulseaudio unload/restore out to shared class.
3488         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer.py: Added.
3489         (PulseAudioSanitizer):
3490         (PulseAudioSanitizer._unload_pulseaudio_module):
3491         (PulseAudioSanitizer._restore_pulseaudio_module):
3492
3493 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3494
3495         Unreviewed, revert r115453 - broke chromium win bots
3496
3497         * Scripts/webkitpy/layout_tests/port/chromium.py:
3498         (ChromiumPort._path_to_image_diff):
3499         (ChromiumDriver):
3500         (ChromiumDriver.__init__):
3501         (ChromiumDriver._wrapper_options):
3502         (ChromiumDriver.cmd_line):
3503         (ChromiumDriver.stop):
3504         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3505         (ChromiumDriverTest.setUp):
3506         (ChromiumDriverTest.test_stop):
3507         (ChromiumDriverTest.test_two_drivers):
3508         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
3509
3510 2012-04-27  Ryosuke Niwa  <rniwa@webkit.org>
3511
3512         Chromium Windows Perf bots timeout due to not output
3513         https://bugs.webkit.org/show_bug.cgi?id=84940
3514
3515         Reviewed by Dirk Pranke.
3516
3517         Dirk and I investigated the issue on the bot but we couldn't figure out what is going wrong.
3518         Since we run-perf-tests don't need any of fancy feature printer provides, just use python's
3519         built-in logging module instead. Printing out to stdout and stderr seem to work so hopefully
3520         this will fix the issue on the bot.
3521
3522         * Scripts/run-perf-tests:
3523         * Scripts/webkitpy/performance_tests/perftest.py:
3524         (PerfTest.run):
3525         (PerfTest.run_failed):
3526         (PerfTest.parse_output):
3527         (PerfTest.output_statistics):
3528         (ChromiumStylePerfTest.parse_output):
3529         (PageLoadingPerfTest.run):
3530         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
3531         (MainTest.test_parse_output):
3532         (MainTest.test_parse_output_with_failing_line):
3533         (TestPageLoadingPerfTest):
3534         (TestPageLoadingPerfTest.test_run):
3535         (TestPageLoadingPerfTest.test_run_with_bad_output):
3536         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3537         (PerfTestsRunner.__init__):
3538         (PerfTestsRunner._parse_args):
3539         (PerfTestsRunner.run):
3540         (PerfTestsRunner._upload_json):
3541         (PerfTestsRunner._print_status):
3542         (PerfTestsRunner._run_tests_set):
3543         (PerfTestsRunner._run_single_test):
3544         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3545         (create_runner):
3546         (test_run_test_set):
3547         (test_run_test_set_kills_drt_per_run):
3548         (test_run_test_pause_before_testing):
3549         (test_run_test_set_for_parser_tests):
3550         (test_run_test_set_with_json_output):
3551         (test_run_test_set_with_json_source):
3552         (test_run_test_set_with_multiple_repositories):
3553         (test_upload_json):
3554         (test_parse_args):
3555
3556 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3557
3558         [chromium] use "drt-style" output, not "test-shell-style" output, on mac and linux DRT
3559         https://bugs.webkit.org/show_bug.cgi?id=84917
3560
3561         Reviewed by Ojan Vafai.
3562
3563         This change changes the chromium new-run-webkit-tests
3564         implementation to use the code in WebKitDriver and ServerProcess
3565         by default on Mac and Linux, instead of using the old-style
3566         Chromium/test_shell style of output.
3567
3568         We will still use the test_shell style on Windows for now
3569         and you can still get the old style of behavior with
3570         new-run-webkit-tests --additional-drt-flag --test-shell
3571
3572         * Scripts/webkitpy/layout_tests/port/chromium.py:
3573         (ChromiumDriver):
3574         (ChromiumDriver.__init__):
3575         (ChromiumDriver._wrapper_options):
3576         (ChromiumDriver.cmd_line):
3577         (ChromiumDriver._start):
3578         (ChromiumDriver.has_crashed):
3579         (ChromiumDriver.run_test):
3580         (ChromiumDriver.stop):
3581         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3582         (ChromiumDriverTest.setUp):
3583         (ChromiumDriverTest.test_stop):
3584         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
3585         (ChromiumDriverTest.test_two_drivers):
3586
3587 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3588
3589         webkitpy: fix race in unit tests for manager_worker_broker
3590         https://bugs.webkit.org/show_bug.cgi?id=84925
3591
3592         Reviewed by Ojan Vafai.
3593
3594         I have been intermittently seeing hangs when running
3595         test-webkitpy on Mac SL; it appears that there's a bug (?) in
3596         the multiprocessing module where queue references are getting
3597         leaked or not cleaned up somehow. By explicitly closing the
3598         queues, things seem to be happy, so this patch adds a cleanup()
3599         method to the message broker and modifies the manager (and the
3600         unit tests) to call it. This may get rid of the intermittent
3601         error on shutdown on Linux I've seen as well.
3602
3603         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3604         (Manager._run_tests):
3605         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
3606         (_Broker.__del__):
3607         (_Broker):
3608         (_Broker.cleanup):
3609         (_BrokerConnection.cleanup):
3610         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
3611         (_TestsMixin.test_name):
3612         (_TestsMixin.test_cancel):
3613         (_TestsMixin.test_done):
3614         (_TestsMixin.test_unknown_message):
3615
3616 2012-04-27  Gavin Peters  <gavinp@chromium.org>
3617
3618         Add new ENABLE_LINK_PRERENDER define to control the Prerendering API
3619         https://bugs.webkit.org/show_bug.cgi?id=84871
3620
3621         Reviewed by Adam Barth.
3622
3623         Prerendering is currently covered by the ENABLE_LINK_PREFETCH macro, but the new Prerendering
3624         API separates it from prefetching.  Having separate include guards lets ports enable prefetching,
3625         a relatively easy change, without needing to build the infrastructure for prerendering, which
3626         is considerably more complicated.
3627
3628 2012-04-27  Philippe Normand  <pnormand@igalia.com>
3629
3630         Unreviewed, GTK gardening.
3631
3632         * Scripts/run-gtk-tests:
3633         (TestRunner): Skip flaky WebKit2 API test.
3634
3635 2012-04-27  Philippe Normand  <pnormand@igalia.com>
3636
3637         Unreviewed, GTK gardening.
3638
3639         * Scripts/run-gtk-tests:
3640         (TestRunner): Fix another typo. So easy to get confused by a Test
3641         name and the actual name of the executable...
3642
3643 2012-04-27  Philippe Normand  <pnormand@igalia.com>
3644
3645         Unreviewed, GTK gardening.
3646
3647         * Scripts/run-gtk-tests:
3648         (TestRunner): Skip failing WebKit2 API test.
3649
3650 2012-04-27  Christophe Dumez  <christophe.dumez@intel.com>
3651
3652         [EFL] DRT should support LayoutTestController's dumpResourceLoadCallbacks()
3653         https://bugs.webkit.org/show_bug.cgi?id=81891