Tools/ChangeLog

   1 2012-05-14  David Kilzer  <ddkilzer@apple.com>
   2
   3         webkit-patch apply-attachment is very slow for big patches
   4         https://bugs.webkit.org/show_bug.cgi?id=74929
   5
   6         Add "excludefield=attachmentdata" to the URL everywhere
   7         "ctype=xml" is used to exclude this field from the XML.
   8
   9         Reviewed by Adam Barth.
  10
  11         * Scripts/prepare-ChangeLog:
  12         (fetchBugDescriptionFromURL):
  13         * Scripts/webkitpy/common/config/urls.py:
  14         * Scripts/webkitpy/common/config/urls_unittest.py:
  15         (URLsTest.test_parse_bug_id):
  16         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
  17         (Bugzilla.bug_url_for_bug_id):
  18
  19 2012-05-13  Ryosuke Niwa  <rniwa@webkit.org>
  20
  21         performance tests should be able to measure runs/sec rather than time
  22         https://bugs.webkit.org/show_bug.cgi?id=86021
  23
  24         Reviewed by Ojan Vafai.
  25
  26         Allow " runs/s" or " ms" to appear after numerical values in tests.
  27
  28         * Scripts/webkitpy/performance_tests/perftest.py:
  29         (PerfTest):
  30
  31 2012-05-12  Tim Horton  <timothy_horton@apple.com>
  32
  33         run-safari and co. should support --guard-malloc command line argument
  34         https://bugs.webkit.org/show_bug.cgi?id=86287
  35
  36         Reviewed by Dan Bernstein.
  37
  38         Add setUpGuardMallocIfNeeded(), which consults the argument --guard-malloc on OS X,
  39         setting DYLD_INSERT_LIBRARIES="/usr/lib/libgmalloc.dylib" if it's enabled.
  40
  41         Use setUpGuardMallocIfNeeded() before running a WebKit app (whether standalone or
  42         in the debugger). This will cause gmalloc to be loaded for gdb as well, but this
  43         tends to be harmless.
  44
  45         Change "malloc guard" to the more common "Guard Malloc" in run-webkit-tests.
  46
  47         * Scripts/old-run-webkit-tests:
  48         * Scripts/webkitdirs.pm:
  49         (appendToEnvironmentVariableList):
  50         (setUpGuardMallocIfNeeded):
  51         (printHelpAndExitForRunAndDebugWebKitAppIfNeeded):
  52         (runMacWebKitApp):
  53         (execMacWebKitAppForDebugging):
  54         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
  55         (parse_args):
  56
  57 2012-05-11  Ojan Vafai  <ojan@chromium.org>
  58
  59         Add previous/next buttons the garden-o-matics perf panel
  60         https://bugs.webkit.org/show_bug.cgi?id=86280
  61
  62         Reviewed by Dimitri Glazkov.
  63
  64         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf.js:
  65         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf_unittests.js:
  66         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/perf.css:
  67
  68 2012-05-11  Kevin Ollivier  <kevino@theolliviers.com>
  69
  70         [wx] Unreviewed build fix. Fix memory exhausted errors from 32-bit ld on *nix.
  71
  72         * waf/build/settings.py:
  73         (common_configure):
  74
  75 2012-05-11  Dirk Pranke  <dpranke@chromium.org>
  76
  77         webkit-patch rebaseline should fetch stale expectations :)
  78         https://bugs.webkit.org/show_bug.cgi?id=86231
  79
  80         Unreviewed, build fix.
  81
  82         This change reverts r115492 and goes back to fetching individual
  83         baselines from the layout tests directory, rather than fetching
  84         entire zip files. The performance hit over networks for people
  85         far away from the bots has turned out to be severe, so we need a
  86         better solution to this.
  87
  88         * Scripts/webkitpy/tool/commands/rebaseline.py:
  89         (RebaselineTest._results_url):
  90         (RebaselineTest._file_name_for_expected_result):
  91         (RebaselineTest._rebaseline_test):
  92         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
  93         (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
  94         (test_rebaseline_updates_expectations_file):
  95         (test_rebaseline_test):
  96         (test_rebaseline_and_copy_test):
  97         (test_rebaseline_and_copy_test_no_existing_result):
  98         (test_rebaseline_and_copy_test_with_lion_result):
  99         (test_rebaseline_and_copy_no_overwrite_test):
 100
 101 2012-05-11  Christophe Dumez  <christophe.dumez@intel.com>
 102
 103         [GTK] LayoutTests/svg/batik/text/verticalText.svg fails to display Chinese characters
 104         https://bugs.webkit.org/show_bug.cgi?id=86198
 105
 106         Reviewed by Martin Robinson.
 107
 108         Update webkitgtk-fonts package to v0.0.3 in order to get new Kochi
 109         Mincho / Gothic fonts to display Japanese and Chinese characters.
 110
 111         * gtk/jhbuild.modules:
 112
 113 2012-05-11  Christophe Dumez  <christophe.dumez@intel.com>
 114
 115         [EFL][DRT] SVG Batik test fails to display Chinese characters
 116         https://bugs.webkit.org/show_bug.cgi?id=84461
 117
 118         Reviewed by Martin Robinson.
 119
 120         Update webkitgtk-fonts package from v0.0.1 to v0.0.3 in order to get
 121         new Kochi Mincho / Gothic fonts to display Japanese and Chinese
 122         characters.
 123
 124         * DumpRenderTree/efl/CMakeLists.txt:
 125         * efl/jhbuild.modules:
 126
 127 2012-05-11  Philippe Normand  <pnormand@igalia.com>
 128
 129         Unreviewed, GTK 64-bit Debug bot fix attempt.
 130
 131         * Scripts/webkitpy/layout_tests/port/gtk.py:
 132         (GtkPort._get_gdb_output): Ignore unicode decoding errors when
 133         processing gdb output.
 134
 135 2012-05-11  Gustavo Noronha Silva  <gns@gnome.org>
 136
 137         REGRESSION(r116205): [GTK]: build no longer supports thin archives
 138         https://bugs.webkit.org/show_bug.cgi?id=86207
 139
 140         * GNUmakefile.am: link libWebCoreModules to DRT.
 141
 142 2012-05-10  Ojan Vafai  <ojan@chromium.org>
 143
 144         Add a perf tab to garden-o-matic
 145         https://bugs.webkit.org/show_bug.cgi?id=86169
 146
 147         Reviewed by Dimitri Glazkov.
 148
 149         This is a first stab at making it easier to view the Chromium WebKit canary
 150         perf graphs. For now, it's just a couple select elements and an iframe.
 151
 152         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
 153         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
 154         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
 155         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
 156         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
 157         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
 158         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf.js: Added.
 159         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf_unittests.js: Added.
 160         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
 161         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/perf.css: Added.
 162         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
 163
 164 2012-05-10  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
 165
 166         Move resumeAnimations to use Internals interface
 167         https://bugs.webkit.org/show_bug.cgi?id=86063
 168
 169         Reviewed by Alexey Proskuryakov.
 170
 171         Remove resumeAnimations functions, because it is able to work in the
 172         cross-port way through the Internals interface.
 173
 174         * DumpRenderTree/LayoutTestController.cpp:
 175         (LayoutTestController::staticFunctions):
 176         * DumpRenderTree/LayoutTestController.h:
 177         (LayoutTestController):
 178         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
 179         * DumpRenderTree/chromium/LayoutTestController.cpp:
 180         (LayoutTestController::LayoutTestController):
 181         * DumpRenderTree/chromium/LayoutTestController.h:
 182         (LayoutTestController):
 183         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
 184         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
 185         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
 186         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
 187         * DumpRenderTree/qt/LayoutTestControllerQt.h:
 188         (LayoutTestController):
 189         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
 190         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
 191         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
 192         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
 193         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
 194         (LayoutTestController):
 195
 196 2012-05-10  Alexey Proskuryakov  <ap@apple.com>
 197
 198         Crash in 3rd party WebKit apps that disable cache at a wrong time
 199         https://bugs.webkit.org/show_bug.cgi?id=86027
 200         <rdar://problem/10615880>
 201
 202         Reviewed by Antti Koivisto.
 203
 204         Added a test that's very similar to MemoryCachePruneWithinResourceLoadDelegate,
 205         but for disabling the cache instead of triggering a prune.
 206
 207         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
 208         * TestWebKitAPI/Tests/mac/MemoryCacheDisableWithinResourceLoadDelegate.html: Added.
 209         * TestWebKitAPI/Tests/mac/MemoryCacheDisableWithinResourceLoadDelegate.mm: Added.
 210         (-[MemoryCacheDisableTestResourceLoadDelegate webView:identifierForInitialRequest:fromDataSource:]):
 211         (-[MemoryCacheDisableTestResourceLoadDelegate webView:resource:willSendRequest:redirectResponse:fromDataSource:]):
 212         (-[MemoryCacheDisableTestResourceLoadDelegate webView:resource:didFinishLoadingFromDataSource:]):
 213         (-[MemoryCacheDisableTestResourceLoadDelegate webView:resource:didFailLoadingWithError:fromDataSource:]):
 214         (TestWebKitAPI::TEST):
 215
 216 2012-05-10  Anders Carlsson  <andersca@apple.com>
 217
 218         WebKit2: Add a way to blacklist specific plug-ins/plug-in versions
 219         https://bugs.webkit.org/show_bug.cgi?id=86164
 220         <rdar://problem/9551196>
 221
 222         Reviewed by Sam Weinig.
 223
 224         Update for WK2 API changes.
 225
 226         * MiniBrowser/mac/BrowserWindowController.m:
 227         (-[BrowserWindowController awakeFromNib]):
 228         * WebKitTestRunner/TestController.cpp:
 229         (WTR::TestController::createOtherPage):
 230         (WTR::TestController::initialize):
 231
 232 2012-05-10  Dirk Pranke  <dpranke@chromium.org>
 233
 234         nrwt: [chromium] fix path to asan symbolize script
 235         https://bugs.webkit.org/show_bug.cgi?id=86161
 236
 237         Reviewed by Abhishek Arya.
 238
 239         * Scripts/webkitpy/layout_tests/port/chromium.py:
 240         (ChromiumPort._get_crash_log):
 241
 242 2012-05-10  Lei Zhang  <thestig@chromium.org>
 243
 244         [chromium] Set Valgrind environment variables for layout tests.
 245         https://bugs.webkit.org/show_bug.cgi?id=86156
 246
 247         Reviewed by Dirk Pranke.
 248
 249         * Scripts/webkitpy/layout_tests/port/chromium.py:
 250         (ChromiumPort.setup_environ_for_server):
 251
 252 2012-05-10  Raphael Kubo da Costa  <rakuco@webkit.org>
 253
 254         [webkitpy][EFL] It should be possible to use a wrapper command for ImageDiff.
 255         https://bugs.webkit.org/show_bug.cgi?id=86145
 256
 257         Reviewed by Dirk Pranke.
 258
 259         The GTK+ and EFL ports use jhbuild to build a few base libraries
 260         which are then used when building the port itself. Programs such
 261         as DRT and ImageDiff are then supposed to be run with the
 262         `run-with-jhbuild' wrapper so that the libraries are properly
 263         found and loaded.
 264
 265         This was already the case for DumpRenderTree (EflPort defines the
 266         'wrapper' option for that), but it was not possible to do the same
 267         with ImageDiff so far.
 268
 269         The code which created the ImageDiff command-line call has been
 270         moved to a separate method, _image_diff_command, which is then
 271         overridden by EflPort to add the jhbuild wrapper.
 272
 273         * Scripts/webkitpy/layout_tests/port/efl.py:
 274         (EflPort.__init__):
 275         (EflPort._image_diff_command):
 276         * Scripts/webkitpy/layout_tests/port/webkit.py:
 277         (WebKitPort):
 278         (WebKitPort._image_diff_command):
 279         (WebKitPort._start_image_diff_process):
 280
 281 2012-05-10  Zan Dobersek  <zandobersek@gmail.com>
 282
 283         [GTK] ENABLE_IFRAME_SEAMLESS support
 284         https://bugs.webkit.org/show_bug.cgi?id=85843
 285
 286         Reviewed by Eric Seidel.
 287
 288         Add the iframe-seamless option to the options list. This makes it
 289         possible to enable the iframe seamless support through build-webkit
 290         script. Enabled by default for all the ports.
 291
 292         * Scripts/webkitperl/FeatureList.pm:
 293
 294 2012-05-10  Raphael Kubo da Costa  <rakuco@webkit.org>
 295
 296         [EFL][webkitpy] Do not redefine check_build() in EflPort.
 297         https://bugs.webkit.org/show_bug.cgi?id=86124
 298
 299         Reviewed by Dirk Pranke.
 300
 301         Remove an old FIXME and do not redefine check_build in EflPort;
 302         once build-dumprendertree recognizes --efl, we can use the
 303         check_build implementation in WebKitPort.
 304
 305         * Scripts/build-dumprendertree:
 306         * Scripts/webkitpy/layout_tests/port/efl.py:
 307         (EflPort.check_build):
 308
 309 2012-05-10  Tommy Widenflycht  <tommyw@google.com>
 310
 311         [chromium] MediaStream API: Fix the ExtraData functionality in WebMediaStreamDescriptor
 312         https://bugs.webkit.org/show_bug.cgi?id=86087
 313
 314         Reviewed by Adam Barth.
 315
 316         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp:
 317         (MockExtraData):
 318         (WebKit):
 319         (WebKit::WebUserMediaClientMock::requestUserMedia):
 320
 321 2012-05-10  Varun Jain  <varunjain@google.com>
 322
 323         [chromium] Trigger context menu for long press gesture
 324         https://bugs.webkit.org/show_bug.cgi?id=85919
 325
 326         Reviewed by Adam Barth.
 327
 328         * DumpRenderTree/chromium/EventSender.cpp:
 329         (EventSender::EventSender):
 330         (EventSender::gestureLongPress):
 331         (EventSender::gestureEvent):
 332         * DumpRenderTree/chromium/EventSender.h:
 333         (EventSender):
 334
 335 2012-05-10  Takashi Toyoshima  <toyoshim@chromium.org>
 336
 337         don't use autoinstall to import pywebsocket but check it in WebKit directly.
 338         https://bugs.webkit.org/show_bug.cgi?id=86107
 339
 340         Reviewed by Adam Barth.
 341
 342         This change removes pywebsocket from webkitpy's autoinstall list and
 343         imports pywebsocket 0.7.4 directly into webkitpy/thirparty.
 344
 345         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
 346         (PyWebSocket._prepare_config):
 347         * Scripts/webkitpy/thirdparty/__init__.py:
 348         (AutoinstallImportHook.find_module):
 349         (AutoinstallImportHook._install_irc):
 350         * Scripts/webkitpy/thirdparty/__init___unittest.py:
 351         (ThirdpartyTest.test_import_hook.MockImportHook.__init__):
 352         (ThirdpartyTest.test_import_hook.MockImportHook._install_eliza):
 353         (ThirdpartyTest):
 354         (ThirdpartyTest.test_import_hook):
 355         * Scripts/webkitpy/thirdparty/mod_pywebsocket/COPYING: Added.
 356         * Scripts/webkitpy/thirdparty/mod_pywebsocket/__init__.py: Added.
 357         * Scripts/webkitpy/thirdparty/mod_pywebsocket/_stream_base.py: Added.
 358         (ConnectionTerminatedException):
 359         (InvalidFrameException):
 360         (BadOperationException):
 361         (UnsupportedFrameException):
 362         (InvalidUTF8Exception):
 363         (StreamBase):
 364         (StreamBase.__init__):
 365         (StreamBase._read):
 366         (StreamBase._write):
 367         (StreamBase.receive_bytes):
 368         (StreamBase._read_until):
 369         * Scripts/webkitpy/thirdparty/mod_pywebsocket/_stream_hixie75.py: Added.
 370         (for):
 371         (StreamHixie75):
 372         (StreamHixie75.for):
 373         (StreamHixie75.__init__):
 374         (StreamHixie75.send_message):
 375         (StreamHixie75._read_payload_length_hixie75):
 376         (StreamHixie75.receive_message):
 377         (StreamHixie75._send_closing_handshake):
 378         (StreamHixie75.close_connection):
 379         (StreamHixie75.send_ping):
 380         * Scripts/webkitpy/thirdparty/mod_pywebsocket/_stream_hybi.py: Added.
 381         (Frame):
 382         (Frame.__init__):
 383         (create_length_header):
 384         (create_header):
 385         (_build_frame):
 386         (_filter_and_format_frame_object):
 387         (create_binary_frame):
 388         (create_text_frame):
 389         (FragmentedFrameBuilder):
 390         (FragmentedFrameBuilder.to):
 391         (FragmentedFrameBuilder.__init__):
 392         (FragmentedFrameBuilder.build):
 393         (_create_control_frame):
 394         (create_ping_frame):
 395         (create_pong_frame):
 396         (create_close_frame):
 397         (StreamOptions):
 398         (StreamOptions.__init__):
 399         (Stream):
 400         (Stream.for):
 401         (Stream.__init__):
 402         (Stream._receive_frame):
 403         (Stream._receive_frame_as_frame_object):
 404         (Stream.send_message):
 405         (Stream.receive_message):
 406         (Stream._send_closing_handshake):
 407         (Stream.close_connection):
 408         (Stream.send_ping):
 409         (Stream._send_pong):
 410         (Stream._drain_received_data):
 411         * Scripts/webkitpy/thirdparty/mod_pywebsocket/common.py: Added.
 412         (is_control_opcode):
 413         (ExtensionParameter):
 414         (ExtensionParameter.__init__):
 415         (ExtensionParameter.name):
 416         (ExtensionParameter.add_parameter):
 417         (ExtensionParameter.get_parameters):
 418         (ExtensionParameter.get_parameter_names):
 419         (ExtensionParameter.has_parameter):
 420         (ExtensionParameter.get_parameter_value):
 421         * Scripts/webkitpy/thirdparty/mod_pywebsocket/dispatch.py: Added.
 422         (DispatchException):
 423         (DispatchException.__init__):
 424         (_default_passive_closing_handshake_handler):
 425         (_normalize_path):
 426         (_create_path_to_resource_converter):
 427         (_create_path_to_resource_converter.converter):
 428         (_enumerate_handler_file_paths):
 429         (_HandlerSuite):
 430         (_HandlerSuite.__init__):
 431         (_source_handler_file):
 432         (_extract_handler):
 433         (Dispatcher):
 434         (Dispatcher.maintains):
 435         (Dispatcher.__init__):
 436         (Dispatcher.add_resource_path_alias):
 437         (Dispatcher.source_warnings):
 438         (Dispatcher.do_extra_handshake):
 439         (Dispatcher.transfer_data):
 440         (Dispatcher.passive_closing_handshake):
 441         (Dispatcher.get_handler_suite):
 442         (Dispatcher._source_handler_files_in_dir):
 443         * Scripts/webkitpy/thirdparty/mod_pywebsocket/extensions.py: Added.
 444         (ExtensionProcessorInterface):
 445         (ExtensionProcessorInterface.get_extension_response):
 446         (ExtensionProcessorInterface.setup_stream_options):
 447         (DeflateStreamExtensionProcessor):
 448         (DeflateStreamExtensionProcessor.__init__):
 449         (DeflateStreamExtensionProcessor.get_extension_response):
 450         (DeflateStreamExtensionProcessor.setup_stream_options):
 451         (DeflateFrameExtensionProcessor):
 452         (DeflateFrameExtensionProcessor.__init__):
 453         (DeflateFrameExtensionProcessor.get_extension_response):
 454         (DeflateFrameExtensionProcessor.setup_stream_options):
 455         (DeflateFrameExtensionProcessor.setup_stream_options._OutgoingFilter):
 456         (DeflateFrameExtensionProcessor.setup_stream_options._OutgoingFilter.__init__):
 457         (DeflateFrameExtensionProcessor.setup_stream_options._OutgoingFilter.filter):
 458         (DeflateFrameExtensionProcessor.setup_stream_options._IncomingFilter):
 459         (DeflateFrameExtensionProcessor.setup_stream_options._IncomingFilter.__init__):
 460         (DeflateFrameExtensionProcessor.setup_stream_options._IncomingFilter.filter):
 461         (DeflateFrameExtensionProcessor.set_response_window_bits):
 462         (DeflateFrameExtensionProcessor.set_response_no_context_takeover):
 463         (DeflateFrameExtensionProcessor.enable_outgoing_compression):
 464         (DeflateFrameExtensionProcessor.disable_outgoing_compression):
 465         (DeflateFrameExtensionProcessor._outgoing_filter):
 466         (DeflateFrameExtensionProcessor._incoming_filter):
 467         (get_extension_processor):
 468         (get_extension_processor.is):
 469         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/__init__.py: Added.
 470         (try):
 471         (do_handshake):
 472         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/_base.py: Added.
 473         (AbortedByUserException):
 474         (HandshakeException):
 475         (HandshakeException.__init__):
 476         (VersionException):
 477         (VersionException.__init__):
 478         (get_default_port):
 479         (validate_subprotocol):
 480         (parse_host_header):
 481         (format_header):
 482         (build_location):
 483         (get_mandatory_header):
 484         (validate_mandatory_header):
 485         (check_request_line):
 486         (check_header_lines):
 487         (parse_token_list):
 488         (_parse_extension_param):
 489         (_parse_extension):
 490         (parse_extensions):
 491         (format_extensions):
 492         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/draft75.py: Added.
 493         (Handshaker):
 494         (Handshaker.performs):
 495         (Handshaker.__init__):
 496         (Handshaker.do_handshake):
 497         (Handshaker._set_resource):
 498         (Handshaker._set_origin):
 499         (Handshaker._set_location):
 500         (Handshaker._set_subprotocol):
 501         (Handshaker._set_protocol_version):
 502         (Handshaker._sendall):
 503         (Handshaker._send_handshake):
 504         (Handshaker._check_header_lines):
 505         (Handshaker._check_first_lines):
 506         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/hybi.py: Added.
 507         (compute_accept):
 508         (Handshaker):
 509         (Handshaker.__init__):
 510         (Handshaker._validate_connection_header):
 511         (Handshaker.do_handshake):
 512         (Handshaker._get_origin):
 513         (Handshaker._check_version):
 514         (Handshaker._set_protocol):
 515         (Handshaker._parse_extensions):
 516         (Handshaker._validate_key):
 517         (Handshaker._get_key):
 518         (Handshaker._send_handshake):
 519         * Scripts/webkitpy/thirdparty/mod_pywebsocket/handshake/hybi00.py: Added.
 520         (Handshaker):
 521         (Handshaker.__init__):
 522         (Handshaker.do_handshake):
 523         (Handshaker._set_resource):
 524         (Handshaker._set_subprotocol):
 525         (Handshaker._set_location):
 526         (Handshaker._set_origin):
 527         (Handshaker._set_protocol_version):
 528         (Handshaker._set_challenge_response):
 529         (Handshaker._get_key_value):
 530         (Handshaker._get_challenge):
 531         (Handshaker._send_handshake):
 532         * Scripts/webkitpy/thirdparty/mod_pywebsocket/headerparserhandler.py: Added.
 533         (ApacheLogHandler):
 534         (ApacheLogHandler.__init__):
 535         (ApacheLogHandler.emit):
 536         (_configure_logging):
 537         (_parse_option):
 538         (_create_dispatcher):
 539         (headerparserhandler):
 540         * Scripts/webkitpy/thirdparty/mod_pywebsocket/http_header_util.py: Added.
 541         (_is_char):
 542         (_is_ctl):
 543         (ParsingState):
 544         (ParsingState.__init__):
 545         (peek):
 546         (consume):
 547         (consume_string):
 548         (consume_lws):
 549         (consume_lwses):
 550         (consume_token):
 551         (consume_token_or_quoted_string):
 552         (quote_if_necessary):
 553         (parse_uri):
 554         * Scripts/webkitpy/thirdparty/mod_pywebsocket/memorizingfile.py: Added.
 555         (MemorizingFile):
 556         (MemorizingFile.__init__):
 557         (MemorizingFile.__getattribute__):
 558         (MemorizingFile.readline):
 559         (MemorizingFile.get_memorized_lines):
 560         * Scripts/webkitpy/thirdparty/mod_pywebsocket/msgutil.py: Added.
 561         (close_connection):
 562         (send_message):
 563         (receive_message):
 564         (send_ping):
 565         (MessageReceiver):
 566         (MessageReceiver.receives):
 567         (MessageReceiver.provides):
 568         (MessageReceiver.should):
 569         (MessageReceiver.__init__):
 570         (MessageReceiver.run):
 571         (MessageReceiver.receive):
 572         (MessageReceiver.receive_nowait):
 573         (MessageReceiver.stop):
 574         (MessageSender):
 575         (MessageSender.sends):
 576         (MessageSender.provides):
 577         (MessageSender.should):
 578         (MessageSender.__init__):
 579         (MessageSender.run):
 580         (MessageSender.send):
 581         (MessageSender.send_nowait):
 582         * Scripts/webkitpy/thirdparty/mod_pywebsocket/standalone.py: Added.
 583         (_StandaloneConnection):
 584         (_StandaloneConnection.__init__):
 585         (_StandaloneConnection.get_local_addr):
 586         (_StandaloneConnection.get_remote_addr):
 587         (_StandaloneConnection.write):
 588         (_StandaloneConnection.read):
 589         (_StandaloneConnection.get_memorized_lines):
 590         (_StandaloneRequest):
 591         (_StandaloneRequest.__init__):
 592         (_StandaloneRequest.get_uri):
 593         (_StandaloneRequest.get_method):
 594         (_StandaloneRequest.is_https):
 595         (_StandaloneRequest._drain_received_data):
 596         (_StandaloneSSLConnection):
 597         (_StandaloneSSLConnection.for):
 598         (_StandaloneSSLConnection.__init__):
 599         (_StandaloneSSLConnection.__getattribute__):
 600         (_StandaloneSSLConnection.__setattr__):
 601         (_StandaloneSSLConnection.makefile):
 602         (WebSocketServer):
 603         (WebSocketServer.__init__):
 604         (WebSocketServer._create_sockets):
 605         (WebSocketServer.server_bind):
 606         (WebSocketServer.server_activate):
 607         (WebSocketServer.server_close):
 608         (WebSocketServer.fileno):
 609         (WebSocketServer.handle_error):
 610         (WebSocketServer.get_request):
 611         (WebSocketServer.serve_forever):
 612         (WebSocketServer.shutdown):
 613         (WebSocketRequestHandler):
 614         (WebSocketRequestHandler.setup):
 615         (WebSocketRequestHandler.setup.SocketServer):
 616         (WebSocketRequestHandler.__init__):
 617         (WebSocketRequestHandler.parse_request):
 618         (WebSocketRequestHandler.log_request):
 619         (WebSocketRequestHandler.log_error):
 620         (WebSocketRequestHandler.is_cgi):
 621         (_configure_logging):
 622         (_alias_handlers):
 623         (_build_option_parser):
 624         (ThreadMonitor):
 625         (ThreadMonitor.__init__):
 626         (ThreadMonitor.run):
 627         (_parse_args_and_config):
 628         (_main):
 629         (_main.if):
 630         (_main.if.__check_script):
 631         * Scripts/webkitpy/thirdparty/mod_pywebsocket/stream.py: Copied from Tools/Scripts/webkitpy/thirdparty/__init___unittest.py.
 632         * Scripts/webkitpy/thirdparty/mod_pywebsocket/util.py: Added.
 633         (get_stack_trace):
 634         (prepend_message_to_exception):
 635         (__translate_interp):
 636         (get_script_interp):
 637         (wrap_popen3_for_win):
 638         (wrap_popen3_for_win.__wrap_popen3):
 639         (hexify):
 640         (get_class_logger):
 641         (NoopMasker):
 642         (NoopMasker.__init__):
 643         (NoopMasker.mask):
 644         (RepeatedXorMasker):
 645         (RepeatedXorMasker.__init__):
 646         (RepeatedXorMasker.mask):
 647         (DeflateRequest):
 648         (DeflateRequest.for):
 649         (DeflateRequest.__init__):
 650         (DeflateRequest.__getattribute__):
 651         (DeflateRequest.__setattr__):
 652         (_Deflater):
 653         (_Deflater.__init__):
 654         (_Deflater.compress_and_flush):
 655         (_Inflater):
 656         (_Inflater.__init__):
 657         (_Inflater.decompress):
 658         (_Inflater.append):
 659         (_Inflater.reset):
 660         (_RFC1979Deflater):
 661         (_RFC1979Deflater.that):
 662         (_RFC1979Deflater.__init__):
 663         (_RFC1979Deflater.filter):
 664         (_RFC1979Inflater):
 665         (_RFC1979Inflater.for):
 666         (_RFC1979Inflater.__init__):
 667         (_RFC1979Inflater.filter):
 668         (DeflateSocket):
 669         (DeflateSocket.for):
 670         (DeflateSocket.__init__):
 671         (DeflateSocket.recv):
 672         (DeflateSocket.sendall):
 673         (DeflateSocket.send):
 674         (DeflateConnection):
 675         (DeflateConnection.for):
 676         (DeflateConnection.__init__):
 677         (DeflateConnection.get_remote_addr):
 678         (DeflateConnection.put_bytes):
 679         (DeflateConnection.read):
 680         (DeflateConnection.write):
 681         (_is_ewouldblock_errno):
 682         (drain_received_data):
 683
 684 2012-05-10  Raphael Kubo da Costa  <rakuco@webkit.org>
 685
 686         [EFL][webkitpy] Define the 'wrapper' option in EflPort instead of creating a EflDriver.
 687         https://bugs.webkit.org/show_bug.cgi?id=86117
 688
 689         Reviewed by Adam Barth.
 690
 691         Simplify the code used to wrap driver calls with jhbuild: we can
 692         just define the 'wrapper' option instead of having to create a
 693         Driver class just for that.
 694
 695         * Scripts/webkitpy/layout_tests/port/efl.py:
 696         (EflPort.__init__):
 697         (EflPort._port_flag_for_scripts):
 698
 699 2012-05-10  Sheriff Bot  <webkit.review.bot@gmail.com>
 700
 701         Unreviewed, rolling out r116633.
 702         http://trac.webkit.org/changeset/116633
 703         https://bugs.webkit.org/show_bug.cgi?id=86093
 704
 705         Breaks EFL tests (Requested by kov on #webkit).
 706
 707         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
 708         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
 709         * DumpRenderTree/efl/EventSender.cpp:
 710         (keyDownCallback):
 711         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
 712         (LayoutTestController::setTabKeyCyclesThroughElements):
 713
 714 2012-05-10  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
 715
 716         [EFL] Memory leaks found in EFL DRT
 717         https://bugs.webkit.org/show_bug.cgi?id=85889
 718
 719         Reviewed by Gustavo Noronha Silva.
 720
 721         Fix two memory leaks in EFL's DRT by using JSRetainPtr for JSString
 722         created with JSStringCreateWithUTF8CString.
 723
 724         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
 725         (DumpRenderTreeChrome::onWindowObjectCleared):
 726         * DumpRenderTree/efl/EventSender.cpp:
 727         (modifiersFromJSValue):
 728
 729 2012-05-10  Alexander Shalamov  <alexander.shalamov@intel.com>
 730
 731         [EFL] LayoutTestController needs implementation of setTabKeyCyclesThroughElements
 732         https://bugs.webkit.org/show_bug.cgi?id=82864
 733
 734         Reviewed by Gustavo Noronha Silva.
 735
 736         Add missing implementation for setTabKeyCyclesThroughElements to EFL's LayoutTestController
 737         in order to unskip fast/events/keypress-insert-tab.html
 738
 739         * DumpRenderTree/efl/EventSender.cpp:
 740         (keyDownCallback):
 741         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
 742         (LayoutTestController::setTabKeyCyclesThroughElements):
 743
 744 2012-05-10  Alexander Shalamov  <alexander.shalamov@intel.com>
 745
 746         [EFL][DRT] EventSender - set delayed event information if it was not set before
 747         https://bugs.webkit.org/show_bug.cgi?id=85972
 748
 749         Reviewed by Gustavo Noronha Silva.
 750
 751         Reset delayed event information if it is missing.
 752
 753         * DumpRenderTree/efl/EventSender.cpp:
 754         (feedOrQueueMouseEvent):
 755
 756 2012-05-07  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
 757
 758         WebPageProxy::activeURL() should return the pending API request, even when there's no main frame
 759         https://bugs.webkit.org/show_bug.cgi?id=85806
 760
 761         The m_pendingAPIRequestURL member is used (presumably) to mask over the async
 762         nature of WebKit2, so that starting a load of a URL will reflect that URL
 763         immedeatly from activeURL, even if the request has not been passed over to
 764         the web process yet and reflected there.
 765
 766         This works well, except in the case of the initial request, where the main
 767         frame creation happens on the web process side and is notified back to the
 768         UI process. Until we've recived the notification we don't know about the main
 769         frame, and this race condition will potentially give us an empty url instead
 770         of the pending request.
 771
 772         To solve this we always return the pending API request if it's set, even
 773         when there's no mainframe yet (that we known about).
 774
 775         Reviewed by Simon Hausmann.
 776
 777         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
 778
 779 2012-05-10  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
 780
 781         [EFL][DRT] Clear added user style sheets before a new testcase execution.
 782         https://bugs.webkit.org/show_bug.cgi?id=84792
 783
 784         Reviewed Gustavo Noronha Silva.
 785
 786         The added user style sheets from the previous testcases are cleared in order to avoid
 787         flaky results.
 788
 789         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
 790         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
 791
 792 2012-05-09  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
 793
 794         Move suspendAnimations to use Internals interface.
 795         https://bugs.webkit.org/show_bug.cgi?id=85986
 796
 797         Reviewed by Ryosuke Niwa.
 798
 799         Remove suspendAnimations functions, because it is able to work in the
 800         cross-port way through the Internals interface
 801
 802         * DumpRenderTree/LayoutTestController.cpp:
 803         (LayoutTestController::staticFunctions):
 804         * DumpRenderTree/LayoutTestController.h:
 805         (LayoutTestController):
 806         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
 807         * DumpRenderTree/chromium/LayoutTestController.cpp:
 808         (LayoutTestController::LayoutTestController):
 809         * DumpRenderTree/chromium/LayoutTestController.h:
 810         (LayoutTestController):
 811         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
 812         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
 813         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
 814         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
 815         * DumpRenderTree/qt/LayoutTestControllerQt.h:
 816         (LayoutTestController):
 817         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
 818         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
 819
 820 2012-05-09  Kent Tamura  <tkent@chromium.org>
 821
 822         [Chromium] Attempt to fix an Android build error.
 823         https://bugs.webkit.org/show_bug.cgi?id=86045
 824
 825         * DumpRenderTree/DumpRenderTree.gypi:
 826
 827 2012-05-09  Ojan Vafai  <ojan@chromium.org>
 828
 829         Don't show partytime if there are non-layout test failures.
 830         https://bugs.webkit.org/show_bug.cgi?id=86056
 831
 832         Reviewed by Dimitri Glazkov.
 833
 834         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
 835         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers_unittests.js:
 836         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
 837
 838 2012-05-09  Hayato Ito  <hayato@chromium.org>
 839
 840         Improve a warning message for a reftest having an unused expectation file.
 841         https://bugs.webkit.org/show_bug.cgi?id=85975
 842
 843         Reviewed by Adam Barth.
 844
 845         The current warning message is not intuitive. It might be better to update the message
 846         so that it encourage users to remove unused expectation files.
 847
 848         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
 849         (SingleTestRunner.__init__):
 850
 851 2012-05-09  Ojan Vafai  <ojan@chromium.org>
 852
 853         Change "losing test coverage" to cover all non-layout test problems in garden-o-matic
 854         https://bugs.webkit.org/show_bug.cgi?id=86043
 855
 856         Reviewed by Adam Barth.
 857
 858         -Get rid of the gtest iframe.
 859         -Cover all non-layout test problems. Before we wouldn't catch compile failures
 860         or gtest failures since they were not on the webkit testing bots.
 861         -Include the list of failing steps next to the builder name. This is a lot less
 862         cluttered than you'd expect and already helped me find a compile failure early.
 863
 864         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
 865         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
 866         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
 867         Delete unused map.
 868         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
 869         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
 870         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
 871         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js:
 872         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
 873         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
 874         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
 875         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
 876
 877 2012-05-09  Jessie Berlin  <jberlin@apple.com>
 878
 879         Crash using the new WKBundleDOMWindowExtensions APIs.
 880         https://bugs.webkit.org/show_bug.cgi?id=85888
 881
 882         Reviewed by Brady Eidson.
 883
 884         Cached frames can live slightly longer than the page, but most clients unregister themselves
 885         and do other cleanup in the willDestroyPage callback, making them miss the
 886         willDestroyGlobalObjectForDOMWindowExtension callbacks.
 887
 888         The calls to willDestroyGlobalObjectForDOMWindowExtension in the DOMWindowExtensionBasic
 889         test were all being invoked underneath WebPage::close. This is unrealistic. Update that test
 890         to destroy the BundleDOMWindowExtensions in response to the willDestroyPage callback.
 891
 892         Add a test to verify that willDestroyGlobalObjectForDOMWindowExtension is being called for
 893         pages that don't go into the page cache.
 894
 895         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
 896         Add DOMWindowExtensionNoCache.cpp, DOMWindowExtensionNoCache_Bundle.cpp, simple-unload.html
 897         and simple-iframe-unload.html
 898
 899         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic.cpp:
 900         (TestWebKitAPI):
 901         Remove the expected messages for willDestroyGlobalObjectForDOMWindowExtension.
 902         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
 903         Do not bother to keep track of the live extension count - all of them are expected to be
 904         live until the test completes.
 905         (TestWebKitAPI::TEST):
 906         Fix the calls to EXPECT to pass the expected value first, and use EXPECT_WK_STREQ so that
 907         message failures will be clearer.
 908         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
 909         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyPage):
 910         Clean up the BundleDOMWindowExtensions.
 911         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyGlobalObjectForDOMWindowExtension):
 912         Add an ASSERT_NOT_REACHED.
 913
 914         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache.cpp: Added.
 915         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
 916         Keep track of the messages received so they can be checked at the end of the test.
 917         (TestWebKitAPI::TEST):
 918         Navigate to uncacheable pages and back.
 919         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache_Bundle.cpp: Copied from Tools/TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp.
 920         (DOMWindowExtensionNoCache):
 921         (TestWebKitAPI::DOMWindowExtensionNoCache::DOMWindowExtensionNoCache):
 922         Set up all the states for each BundleDOMWindowExtension.
 923         (TestWebKitAPI::DOMWindowExtensionNoCache::frameLoadFinished):
 924         Tell the UI Process about the states of the BundleDOMWindowExtensions.
 925         (TestWebKitAPI::DOMWindowExtensionNoCache::sendExtensionStateMessage):
 926         (TestWebKitAPI::DOMWindowExtensionNoCache::initialize):
 927         (TestWebKitAPI::DOMWindowExtensionNoCache::didCreatePage):
 928         (TestWebKitAPI::DOMWindowExtensionNoCache::willDestroyPage):
 929         Remvoe the remaining BundleDOMWindowExtensions, send the updated state, and finish the test.
 930         (TestWebKitAPI::DOMWindowExtensionNoCache::updateExtensionStateRecord):
 931         (TestWebKitAPI::DOMWindowExtensionNoCache::sendBundleMessage):
 932         (TestWebKitAPI::DOMWindowExtensionNoCache::globalObjectIsAvailableForFrame):
 933         (TestWebKitAPI::DOMWindowExtensionNoCache::willDisconnectDOMWindowExtensionFromGlobalObject):
 934         ASSERT that these pages not going into the page cache are not getting disconnected to go into
 935         the page cache.
 936         (TestWebKitAPI::DOMWindowExtensionNoCache::didReconnectDOMWindowExtensionToGlobalObject):
 937         Ditto about getting reconnected when coming out of the page cache.
 938         (TestWebKitAPI::DOMWindowExtensionNoCache::willDestroyGlobalObjectForDOMWindowExtension):
 939         Tell the UI Process, update the state, and get rid of the BundleDOMWindowExtension.
 940         (TestWebKitAPI::didFinishLoadForFrameCallback):
 941         (TestWebKitAPI::globalObjectIsAvailableForFrameCallback):
 942         (TestWebKitAPI::willDisconnectDOMWindowExtensionFromGlobalObjectCallback):
 943         (TestWebKitAPI::didReconnectDOMWindowExtensionToGlobalObjectCallback):
 944         (TestWebKitAPI::willDestroyGlobalObjectForDOMWindowExtensionCallback):
 945
 946         * TestWebKitAPI/Tests/WebKit2/simple-iframe-unload.html: Added.
 947         * TestWebKitAPI/Tests/WebKit2/simple-unload.html: Added.
 948
 949 2012-05-09  Adam Barth  <abarth@webkit.org>
 950
 951         [Chromium] Android's DRT doesn't want to build WebUserMediaClientMock
 952         https://bugs.webkit.org/show_bug.cgi?id=86045
 953
 954         Reviewed by Eric Seidel.
 955
 956         * DumpRenderTree/DumpRenderTree.gypi:
 957
 958 2012-05-09  Adam Barth  <abarth@webkit.org>
 959
 960         [Chromium] MockWebSpeechInputController.cpp should be guarded by ENABLE(INPUT_SPEECH)
 961         https://bugs.webkit.org/show_bug.cgi?id=86041
 962
 963         Reviewed by Eric Seidel.
 964
 965         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp:
 966
 967 2012-05-09  Adam Barth  <abarth@webkit.org>
 968
 969         [Chromium] Android needs to swizzle pixel dumps in order to get the same hash as other platforms
 970         https://bugs.webkit.org/show_bug.cgi?id=86040
 971
 972         Reviewed by Kent Tamura.
 973
 974         * DumpRenderTree/chromium/TestShell.cpp:
 975         (TestShell::dumpImage):
 976
 977 2012-05-09  Raphael Kubo da Costa  <rakuco@webkit.org>
 978
 979         [EFL] Build fix after r116464.
 980
 981         `ulong' is a type defined by glibc, so it is not very
 982         portable. Use `unsigned long' instead.
 983
 984         * DumpRenderTree/efl/EventSender.cpp:
 985         (DelayedEvent::DelayedEvent):
 986         (DelayedEvent):
 987         (leapForwardCallback):
 988
 989 2012-05-09  Adam Barth  <abarth@webkit.org>
 990
 991         [Chromium] LayoutTestController should use ENABLE guards for some features
 992         https://bugs.webkit.org/show_bug.cgi?id=86030
 993
 994         Reviewed by Eric Seidel.
 995
 996         These features are disabled on OS(ANDROID) and so shouldn't be compiled
 997         when the feature is off.
 998
 999         * DumpRenderTree/chromium/LayoutTestController.cpp:
1000         (LayoutTestController::LayoutTestController):
1001         * DumpRenderTree/chromium/LayoutTestController.h:
1002         (LayoutTestController):
1003         * DumpRenderTree/chromium/TestShell.h:
1004         (TestShell):
1005
1006 2012-05-09  Tom Zakrajsek  <tomz@codeaurora.org>
1007
1008         Fix overzealous re-opened since blocked by... message
1009         https://bugs.webkit.org/show_bug.cgi?id=86020
1010
1011         Reviewed by Adam Barth.
1012
1013         Changed CreateBug behavior to only re-open a blocked bug if it
1014         was RESOLVED.  Also had to update the test infrastructure slightly
1015         to allow for distinct mock commit revisions, so one could point
1016         to a RESOLVED bug for testing while another pointed to an OPEN bug.
1017
1018         Also moved default created mock bug to be 60001 instead of 50004
1019         which already exists as a reference mock bug.
1020
1021         * Scripts/webkitpy/common/checkout/checkout_mock.py:
1022         (MockCheckout.__init__):
1023         (MockCheckout.commit_info_for_revision):
1024         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
1025         (MockBugzilla.create_bug):
1026         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
1027         * Scripts/webkitpy/tool/commands/download_unittest.py:
1028         (test_create_rollout_resolved):
1029         * Scripts/webkitpy/tool/commands/queries_unittest.py:
1030         (FailureReasonTest.test_blame_line_for_revision):
1031         * Scripts/webkitpy/tool/steps/createbug.py:
1032         (CreateBug.run):
1033
1034 2012-05-09  Hugo Parente Lima  <hugo.lima@openbossa.org>
1035
1036         Use suitable viewport values on XHTML-MP pages.
1037         https://bugs.webkit.org/show_bug.cgi?id=85425
1038
1039         Reviewed by Kenneth Rohde Christiansen.
1040
1041         Enable LEGACY_VIEWPORT_ADAPTION on Qt port.
1042
1043         * qmake/mkspecs/features/features.prf:
1044
1045 2012-05-09  Jochen Eisinger  <jochen@chromium.org>
1046
1047         When creating a new page during a navigation, prime the initial document with the correct referrer policy
1048         https://bugs.webkit.org/show_bug.cgi?id=86001
1049
1050         Reviewed by Adam Barth.
1051
1052         Store the referrer policy in the request, so the network stack can comply with it.
1053
1054         * DumpRenderTree/chromium/WebViewHost.cpp:
1055         (WebViewHost::willSendRequest):
1056
1057 2012-05-09  Vivek Galatage  <vivekgalatage@gmail.com>
1058
1059         Remove CYGWIN=tty from environment variable as its no longer supported
1060         https://bugs.webkit.org/show_bug.cgi?id=85791
1061
1062         The CYGWIN=tty environment variable is no longer supported from Cygwin
1063         version 1.7.10. This fix avoids addition of this environment variable
1064         for Cygwin versions greater than 1.7.9. For supporting older versions
1065         i.e. 1.7.9 or older, the script takes care of retaining the environment
1066         variable.
1067
1068         Reviewed by Adam Roben.
1069
1070         * Scripts/webkitdirs.pm:
1071         (setupAppleWinEnv):
1072
1073 2012-05-09  Dirk Pranke  <dpranke@chromium.org>
1074
1075         nrwt: --skip-failing-tests should also skip flaky tests
1076         https://bugs.webkit.org/show_bug.cgi?id=85941
1077
1078         Reviewed by Ojan Vafai.
1079
1080         This changes --skip-failing-tests so that we skip tests that
1081         are believed to pass sometimes as well as tests that are
1082         believed to always fail.
1083
1084         In addition, in testing this I discovered that we had lousy
1085         test coverage of the details of retrying test failures and
1086         --clobber-old-results, so I wrote some more tests.
1087
1088         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1089         (Manager.prepare_lists_and_print_output):
1090         * Scripts/webkitpy/layout_tests/port/test.py:
1091         (unit_test_list):
1092         (TestPort.__init__):
1093         (TestDriver.run_test):
1094         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1095         (MainTest.test_skip_failing_tests):
1096         (MainTest.test_retrying_and_flaky_tests):
1097
1098 2012-05-09  Christophe Dumez  <christophe.dumez@intel.com>
1099
1100         [EFL][DRT] EFL's DRT needs to support LayoutTestController.dumpIconChanges()
1101         https://bugs.webkit.org/show_bug.cgi?id=84435
1102
1103         Reviewed by Antonio Gomes.
1104
1105         Catch the "icon,changed" signal on the frames in EFL's DumpRenderTree
1106         to support LayoutTestController's dumpIconChanges().
1107
1108         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1109         (DumpRenderTreeChrome::createView):
1110         (DumpRenderTreeChrome::onFrameIconChanged):
1111         (DumpRenderTreeChrome::onFrameCreated):
1112         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
1113         (DumpRenderTreeChrome):
1114
1115 2012-05-09  Dominik Röttsches  <dominik.rottsches@intel.com>
1116
1117         webkit-patch upload Fails Behind Load-balancing Proxies
1118         https://bugs.webkit.org/show_bug.cgi?id=85980
1119
1120         Reviewed by Adam Barth.
1121
1122         Unchecking the login restriction to IP to allow the tool
1123         to function behind load-balancing proxies that use multiple
1124         outgoing IPs.
1125
1126         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
1127         (Bugzilla.authenticate):
1128
1129 2012-05-09  Hao Zheng  <zhenghao@chromium.org>
1130
1131         [Chromium-android] Add unit test for Android layout test script (chromium_android.py)
1132         https://bugs.webkit.org/show_bug.cgi?id=80852
1133
1134         Reviewed by Adam Barth.
1135
1136         Add some tests to ensure that chromium android port works to some
1137         extent. Some methods is hard to test, as they rely on the actual
1138         behavior on device, e.g. pushing files to device.
1139
1140         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1141         (ChromiumAndroidPort.get_last_stacktrace): Check if dir exists.
1142         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1143         (ChromiumAndroidPortTest):
1144         (ChromiumAndroidPortTest.test_attributes):
1145         (ChromiumAndroidPortTest.test_get_last_stacktrace):
1146         (ChromiumAndroidPortTest.test_get_last_stacktrace.mock_run_command_no_dir):
1147         (ChromiumAndroidPortTest.test_get_last_stacktrace.mock_run_command_no_file):
1148         (ChromiumAndroidPortTest.test_get_last_stacktrace.mock_run_command_non_empty):
1149         (ChromiumAndroidDriverTest):
1150         (ChromiumAndroidDriverTest.setUp):
1151         (ChromiumAndroidDriverTest.test_get_drt_return_value):
1152         (ChromiumAndroidDriverTest.test_has_crash_hint):
1153         (ChromiumAndroidDriverTest.test_read_prompt):
1154         (ChromiumAndroidDriverTest.test_test_shell_command):
1155         (ChromiumAndroidDriverTest.test_write_command_and_read_line):
1156
1157 2012-05-09  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
1158
1159         [Qt] ORWT fails with Qt 5 and WK2.
1160         https://bugs.webkit.org/show_bug.cgi?id=85968
1161
1162         Reviewed by Csaba Osztrogonác.
1163
1164         The Qt5 binary does not have a -style argument, so avoid using it.
1165
1166         * Scripts/old-run-webkit-tests:
1167
1168 2012-05-09  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1169
1170         [EFL][DRT] LayoutTestController addUserScript implementation
1171         https://bugs.webkit.org/show_bug.cgi?id=84718
1172
1173         Reviewed by Antonio Gomes.
1174
1175         Added implementation of addUserScript() to LayoutTestController.
1176
1177         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1178         (DumpRenderTreeChrome::resetDefaultsToConsistentValues): clear added user scripts
1179         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1180         (LayoutTestController::addUserScript): added implementation
1181
1182 2012-05-09  Dan Bernstein  <mitz@apple.com>
1183
1184         Made build-webkit copy WebKitLibraries to the right location again after r116257
1185
1186         Reviewed by Mark Rowe.
1187
1188         * Scripts/build-webkit: Pass the built product directory to copy-webkitlibraries-to-product-directory.
1189         * Scripts/copy-webkitlibraries-to-product-directory: Use the first argument, if present, as the
1190         product directory.
1191
1192 2012-05-08  Ojan Vafai  <ojan@chromium.org>
1193
1194         Show the chromium non-layout test bots in garden-o-matic
1195         https://bugs.webkit.org/show_bug.cgi?id=85903
1196
1197         Reviewed by Eric Seidel.
1198
1199         This is just a single-line high-level overview of the chromium non-layout test bots.
1200         This means that Chromium WebKit gardeners only need to keep garden-o-matic open in
1201         order to do their job.
1202
1203         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
1204         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
1205         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
1206         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
1207         (#chromium-gtests iframe):
1208         (#chromium-gtests):
1209         (#gtest-label):
1210
1211 2012-05-08  Kangil Han  <kangil.han@samsung.com>
1212
1213         [EFL][DRT] Implement TextInputController.
1214         https://bugs.webkit.org/show_bug.cgi?id=78559
1215
1216         Reviewed by Martin Robinson.
1217
1218         TextInputController has been created for an input emulation purpose in DRT.
1219         However, not like other ports, EFL doesn't have it.
1220         Therefore, this patch will enable EFL port to have an input emulation in DRT.
1221
1222         * DumpRenderTree/efl/CMakeLists.txt:
1223         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1224         (DumpRenderTreeChrome::onWindowObjectCleared):
1225         * DumpRenderTree/efl/TextInputController.cpp: Added.
1226         (setMarkedTextCallback):
1227         (hasMarkedTextCallback):
1228         (markedRangeCallback):
1229         (insertTextCallback):
1230         (unmarkTextCallback):
1231         (firstRectForCharacterRangeCallback):
1232         (selectedRangeCallback):
1233         (makeTextInputController):
1234         * DumpRenderTree/efl/TextInputController.h: Added.
1235
1236 2012-04-23  Raphael Kubo da Costa  <rakuco@webkit.org>
1237
1238         [EFL] Pass a WTF::String directly from DRT to ewebkit when possible
1239         https://bugs.webkit.org/show_bug.cgi?id=84634
1240
1241         Reviewed by Gustavo Noronha Silva.
1242
1243         Drop unnecessary const char* -> WTF::String conversions by
1244         creating String objects directly in LayoutTestController out of
1245         JSStringRef's and passing them to DumpRenderTreeSupportEfl.
1246
1247         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1248         (LayoutTestController::queueLoad):
1249         (LayoutTestController::setValueForUser):
1250         (LayoutTestController::findString):
1251         (LayoutTestController::addUserStyleSheet):
1252
1253 2012-05-08  Alexander Shalamov  <alexander.shalamov@intel.com>
1254
1255         [EFL][DRT] EventSender needs implementation of leapForward function
1256         https://bugs.webkit.org/show_bug.cgi?id=83486
1257
1258         Reviewed by Eric Seidel.
1259
1260         Implemented leapForward function in Efl's DRT. This function is required
1261         to delay and queue event execution and simulate real user behavior.
1262
1263         * DumpRenderTree/efl/EventSender.cpp:
1264         (DelayedEvent):
1265         (MouseEventInfo::MouseEventInfo):
1266         (MouseEventInfo):
1267         (sendClick):
1268         (mouseDownCallback):
1269         (mouseUpCallback):
1270         (mouseMoveToCallback):
1271         (leapForwardCallback):
1272         (mouseScrollByCallback):
1273         (feedOrQueueMouseEvent):
1274         (feedMouseEvent):
1275         (feedQueuedMouseEvents):
1276
1277 2012-05-08  Ryuan Choi  <ryuan.choi@samsung.com>
1278
1279         [EFL][DRT] WebKit/DRT should not paint webview for the pixel tests.
1280         https://bugs.webkit.org/show_bug.cgi?id=79853
1281
1282         Reviewed by Eric Seidel.
1283
1284         Screen buffer can be different from rendered results.
1285         For example, WebKit/Efl just moves tiles for ewk_view_tiled when scrolling
1286         contents.
1287         It means that pixel tests can't check regression of ewk_view_tiled.
1288
1289         This patch changes a way to provide bitmap from painting webview to capturing
1290         evas.
1291
1292         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
1293         (createBitmapContextFromWebView):
1294
1295 2012-05-08  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1296
1297         [GTK] Memory leak in DRT EventSender.cpp
1298         https://bugs.webkit.org/show_bug.cgi?id=85886
1299
1300         Reviewed by Martin Robinson.
1301
1302         Fix a memory leak in gdkModifersFromJSValue() by using JSRetainPtr for
1303         JSString created with JSStringCreateWithUTF8CString.
1304
1305         * DumpRenderTree/gtk/EventSender.cpp:
1306         (gdkModifersFromJSValue):
1307
1308 2012-05-08  Christophe Dumez  <christophe.dumez@intel.com>
1309
1310         [EFL] glib dependency needs to be updated in jhbuild
1311         https://bugs.webkit.org/show_bug.cgi?id=85878
1312
1313         Reviewed by Martin Robinson.
1314
1315         Bump glib version from 2.31.8 to 2.32.2 in jhbuild.modules. This
1316         addresses flakiness in media tests due to possible deadlock with the
1317         following output on stderr:
1318         GStreamer-WARNING **: wrong STREAM_LOCK count 0
1319
1320         The bug was fixed upstream by:
1321         http://git.gnome.org/browse/glib/commit/?id=265f265c677bc5399dbb1cfa449cb4b915a6459f
1322
1323         * efl/jhbuild.modules:
1324
1325 2012-05-07  Alexis Menard  <alexis.menard@openbossa.org>
1326
1327         [Qt] Unbreak debugging of WebKit2.
1328         https://bugs.webkit.org/show_bug.cgi?id=85839
1329
1330         Reviewed by Simon Hausmann.
1331
1332         When you attach GDB to a running process, it stops it.
1333         http://trac.webkit.org/changeset/115958 introduced a pause()
1334         call to wait the debugger to be attached to then continue
1335         the execution of the WebProcess. Unfortunately the pause()
1336         function does not return unless a signal handler is called.
1337         This patch introduce an event handler to exit from the paused
1338         state when the debugger send the signal SIGCONT. The old code
1339         works with older version of GDB (<7.0) but not with newer
1340         versions where the behavior of pause() is correct.
1341
1342         * WebKitTestRunner/qt/main.cpp:
1343         (sigcontHandler):
1344         (main):
1345
1346 2012-05-08  Christophe Dumez  <christophe.dumez@intel.com>
1347
1348         [EFL][DRT] EFL's DRT does not fully support page visibility
1349         https://bugs.webkit.org/show_bug.cgi?id=85032
1350
1351         Reviewed by Antonio Gomes.
1352
1353         * DumpRenderTree/LayoutTestController.cpp:
1354         (resetPageVisibilityCallback): Fix incorrect argumentCount check.
1355         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1356         (DumpRenderTreeChrome::resetDefaultsToConsistentValues): Properly
1357         reset page visibility setting between tests.
1358         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1359         (LayoutTestController::setPageVisibility): Add support for "preview"
1360         visibility state.
1361
1362 2012-05-08  Gavin Peters  <gavinp@chromium.org>
1363
1364         [webkit-patch] Return meaningful results from single-result queries.
1365         https://bugs.webkit.org/show_bug.cgi?id=85367
1366
1367         Reviewed by Dirk Pranke.
1368
1369         Sometimes a quicksearch returns a single result by just redirecting to the page for that single
1370         bug. This patch hacks around that by noticing the redirect target is a single page, and constructing
1371         a result list consisting of that one bug.
1372
1373         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
1374         (BugzillaQueries._fetch_bugs_from_advanced_query):
1375
1376 2012-05-08  Gavin Peters  <gavinp@chromium.org>
1377
1378         [webkit-patch] Add new chrome-channels command to track down current chrome release channels for a committed bug.
1379         https://bugs.webkit.org/show_bug.cgi?id=85368
1380
1381         Reviewed by Eric Seidel.
1382
1383         * Scripts/webkitpy/common/config/urls.py:
1384         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
1385         (MockBugzillaQueries.fetch_bugs_matching_quicksearch):
1386         * Scripts/webkitpy/common/net/omahaproxy.py: Added.
1387         (OmahaProxy):
1388         (OmahaProxy.__init__):
1389         (OmahaProxy.set_url):
1390         (OmahaProxy._json_url):
1391         (OmahaProxy._get_json):
1392         (OmahaProxy.get_revisions):
1393         * Scripts/webkitpy/common/net/omahaproxy_unittest.py: Added.
1394         (MockOmahaProxy):
1395         (MockOmahaProxy.__init__):
1396         (OmahaProxyTest):
1397         (OmahaProxyTest.test_get_revisions):
1398         * Scripts/webkitpy/tool/commands/__init__.py:
1399         * Scripts/webkitpy/tool/commands/bugsearch.py:
1400         (BugSearch):
1401         * Scripts/webkitpy/tool/commands/chromechannels.py: Added.
1402         (ChromeChannels):
1403         (__init__):
1404         (execute):
1405         * Scripts/webkitpy/tool/commands/chromechannels_unittest.py: Added.
1406         (MockOmahaProxy):
1407         (MockOmahaProxy.get_revisions):
1408         (TestableChromeChannels):
1409         (TestableChromeChannels.__init__):
1410         (ChromeChannelsTest):
1411         (ChromeChannelsTest.test_single_bug):
1412         (ChromeChannelsTest.test_with_query):
1413
1414 2012-05-08  Carlos Garcia Campos  <cgarcia@igalia.com>
1415
1416         Unreviewed. Fix GTK+ unit tests.
1417
1418         * gtk/common.py:
1419         (pkg_config_file_variable):
1420
1421 2012-05-08  Carlos Garcia Campos  <cgarcia@igalia.com>
1422
1423         [GTK] Google tests that time out are leaked
1424         https://bugs.webkit.org/show_bug.cgi?id=84973
1425
1426         Reviewed by Philippe Normand.
1427
1428         Move Tools/Scripts/run-gtk-tests to Tools/gtk/run-api-tests
1429         leaving run-gtk-tests as simple wrapper that runs run-api-tests in
1430         a jhbuild environment when jhbuild is available.
1431
1432         * Scripts/run-gtk-tests: Run Tools/gtk/run-api-tests directly or
1433         with jhbuild if present.
1434         * gtk/common.py:
1435         (pkg_config_file_variable): Generic function to get a variable
1436         from a pkg-config file.
1437         (prefix_of_pkg_config_file): Use pkg_config_file_variable with 'prefix'.
1438         * gtk/run-api-tests: Copied from Tools/Scripts/run-gtk-tests.
1439         (TestRunner.__init__): Use common.build_path() to get the programs path.
1440         (TestRunner._lookup_atspi2_binary): Use
1441         common.pkg_config_file_variable() to get the 'exec_prefix'.
1442         (TestRunner._start_accessibility_daemons): Use subprocess.Popen()
1443         directly instead of helper _create_process that has been removed.
1444         (TestRunner._setup_testing_environment): Use subprocess.Popen()
1445         directly instead of helper _create_process that has been
1446         removed. Use common.top_level_path() and common.build_path()
1447         instead of custom methods that have been removed too.
1448         (TestRunner._run_test_command): Use subprocess.Popen() directly
1449         instead of helper _create_process that has been removed.
1450
1451 2012-05-08  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1452
1453         [EFL] [DRT] Clear extra views before setting a blank page
1454         https://bugs.webkit.org/show_bug.cgi?id=85608
1455
1456         Reviewed by Kenneth Rohde Christiansen.
1457
1458         We need to clear the extra views before setting a blank page,
1459         otherwise it will cause more events interactions and sometimes
1460         trigger a unload(). This was causing frame-crash-with-page-cache.html
1461         to leak events to the next test causing any subsequent test to
1462         fail.
1463
1464         * DumpRenderTree/efl/DumpRenderTree.cpp:
1465         (runTest):
1466         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1467         (DumpRenderTreeChrome::clearExtraViews):
1468         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1469         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
1470         (DumpRenderTreeChrome):
1471
1472 2012-05-08  Balazs Kelemen  <kbalazs@webkit.org>
1473
1474         [Qt] X11 plugins need to be reworked for Qt5+WK1
1475         https://bugs.webkit.org/show_bug.cgi?id=80691
1476
1477         Reviewed by Simon Hausmann.
1478
1479         Implement basic windowless plugin support with Qt5.
1480
1481         * qmake/mkspecs/features/features.prf:
1482         Enable NPAPI plugins if X11 libraries are available.
1483         * qmake/mkspecs/features/functions.prf:
1484         Added a convenience function to determine availability
1485         of X11 libraries.
1486
1487 2012-05-07  Raphael Kubo da Costa  <rakuco@webkit.org>
1488
1489         [webkitpy] Remove Python 2.5-specific workaround from http_server_base.py.
1490         https://bugs.webkit.org/show_bug.cgi?id=85852
1491
1492         Reviewed by Dirk Pranke.
1493
1494         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
1495         (HttpServerBase._is_server_running_on_all_ports): Revert the part
1496         of r89412 which was a workaround for an issue with Python 2.5 on
1497         Mac 10.5.
1498
1499 2012-05-07  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
1500
1501         webkitdirs.pm: Make determineNumberOfCPUs work on FreeBSD.
1502         https://bugs.webkit.org/show_bug.cgi?id=81481
1503
1504         Reviewed by Dirk Pranke.
1505
1506         FreeBSD has the same sysctl interface present in Darwin to query
1507         the number of CPUs available, so use that too.
1508
1509         * Scripts/webkitdirs.pm:
1510         (determineNumberOfCPUs): Share the sysctl call with Darwin.
1511         (isFreeBSD): Added.
1512
1513 2012-05-07  Alexis Menard  <alexis.menard@openbossa.org>
1514
1515         Unreviewed Qt build fix in Debug.
1516
1517         We need some system includes here for getpid() to be recognized.
1518
1519         * WebKitTestRunner/qt/main.cpp:
1520
1521 2012-05-07  Ojan Vafai  <ojan@chromium.org>
1522
1523         When embedding the flakiness dashboard, hide popups when the frame is blurred.
1524         https://bugs.webkit.org/show_bug.cgi?id=85684
1525
1526         Reviewed by Dirk Pranke.
1527
1528         If you show a popup, then click elsewhere in garden-o-matic, it will now
1529         properly hide it.
1530
1531         * TestResultServer/static-dashboards/dashboard_base.js:
1532         (showPopup):
1533         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1534         (showPopupForBuild):
1535         (generatePageForIndividualTests.if):
1536         (generatePageForIndividualTests):
1537         * TestResultServer/static-dashboards/flakiness_dashboard_embedded_unittests.js: Added.
1538         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1539         * TestResultServer/static-dashboards/run-embedded-unittests.html: Added.
1540
1541 2012-05-07  Ian Vollick  <vollick@chromium.org>
1542
1543         [chromium] WebViewHost should setCompositorSurfaceReady
1544         https://bugs.webkit.org/show_bug.cgi?id=85418
1545
1546         Reviewed by James Robinson.
1547
1548         * DumpRenderTree/chromium/WebViewHost.cpp:
1549         (WebViewHost::setWebWidget):
1550
1551 2012-05-07  Jon Lee  <jonlee@apple.com>
1552
1553         WebKitTestRunner needs layoutTestController.dumpFrameLoadCallbacks
1554         https://bugs.webkit.org/show_bug.cgi?id=42331
1555         <rdar://problem/8193641>
1556
1557         Reviewed by Darin Adler.
1558
1559         This patch sets up the framework for supporting dumpFrameLoadCallbacks, as well as
1560         outputting some of the required strings. As tests become unskipped the other callbacks
1561         can be supported.
1562
1563         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Add dumpFrameLoadCallbacks()
1564         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1565         (WTR::InjectedBundle::didReceiveMessage): pass in the dictionary containing keys needed to initialize
1566         dump callbacks for the layoutTestController.
1567         (WTR::InjectedBundle::booleanForKey): Helper function to extract boolean value from key in WKDictionaryRef.
1568         Output some warning if we are trying to extract a value that is not a boolean.
1569         (WTR::InjectedBundle::beginTesting): Update to take in the dictionary passed in by the TestController.
1570         We cannot set the bits on the layoutTestController until it has been created in beginTesting().
1571         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
1572         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1573         (WTR::InjectedBundle::didReceiveMessage):
1574         (WTR):
1575         (WTR::InjectedBundle::beginTesting):
1576         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
1577         (InjectedBundle):
1578         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1579         (WTR):
1580         (WTR::dumpFrameDescriptionSuitableForTestResult): Static function similar to
1581         -[WebFrame _drt_descriptionSuitableForTestResult] in DumpRenderTree.
1582         (WTR::InjectedBundlePage::didStartProvisionalLoadForFrame): Output string.
1583         (WTR::InjectedBundlePage::didCommitLoadForFrame): Output string.
1584         (WTR::InjectedBundlePage::didFinishLoadForFrame): Output string.
1585         (WTR::InjectedBundlePage::didReceiveTitleForFrame): Output string.
1586         (WTR::InjectedBundlePage::didCancelClientRedirectForFrame): Output string.
1587         (WTR::InjectedBundlePage::willPerformClientRedirectForFrame): Output string. In WK1 it's
1588         "willPerformClientRedirectToURL" so we will have to use that value here.
1589         (WTR::InjectedBundlePage::didFinishDocumentLoadForFrame): Output string.
1590         (WTR::InjectedBundlePage::didHandleOnloadEventsForFrame): Output string.
1591         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1592         (WTR::LayoutTestController::LayoutTestController): Initialize to not dump frame load callbacks.
1593         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1594         (WTR::LayoutTestController::dumpFrameLoadCallbacks): Change to use setShouldDumpFrameLoadCallbacks.
1595         (WTR::LayoutTestController::setShouldDumpFrameLoadCallbacks): Set bit.
1596         (WTR::LayoutTestController::shouldDumpFrameLoadCallbacks): Return bit.
1597         (LayoutTestController): Added m_dumpFrameLoadCallbacks to track whether to dump the output.
1598         * WebKitTestRunner/TestInvocation.cpp:
1599         (WTR::shouldLogFrameLoadDelegates): Similar to DRT.
1600         (WTR::TestInvocation::invoke): Based on the URL loaded, set the bit in the dictionary that gets passed
1601         to the injected bundle.
1602
1603 2012-05-07  Tommy Widenflycht  <tommyw@google.com>
1604
1605         [chromium] MediaStream API: Enhance WebUserMediaClientMock and add a test for it
1606         https://bugs.webkit.org/show_bug.cgi?id=85618
1607
1608         Reviewed by Adam Barth.
1609
1610         Adding the capability to control which MediaStreamTracks actually get created,
1611         and adding a test that exercises the JS api using that capability.
1612
1613         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp:
1614         (WebKit::WebUserMediaClientMock::requestUserMedia):
1615         * DumpRenderTree/chromium/WebUserMediaClientMock.h:
1616         * DumpRenderTree/chromium/WebViewHost.cpp:
1617         (WebViewHost::testMediaStreamClient):
1618         * DumpRenderTree/chromium/WebViewHost.h:
1619         (WebViewHost):
1620
1621 2012-05-07  Zan Dobersek  <zandobersek@gmail.com>
1622
1623         [Gtk] WebGL feature is not built anymore through build-webkit after r116251
1624         https://bugs.webkit.org/show_bug.cgi?id=85788
1625
1626         Reviewed by Martin Robinson.
1627
1628         Enable WebGL by default for the Gtk port. The feature is already enabled by
1629         default in configure.ac, so there's no reason it shouldn't be enabled when
1630         building through build-webkit.
1631
1632         * Scripts/webkitperl/FeatureList.pm:
1633
1634 2012-05-07  Csaba Osztrogonác  <ossy@webkit.org>
1635
1636         Unittest fix after r116288, because now there are more than 5 hit for the 'Eric' word.
1637         We have only 2 'Gavin', so it is a better choice than 'Eric' for unit testing.
1638
1639         Reviewed by Tor Arne Vestbø.
1640
1641         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1642         (IRCCommandTest.test_whois):
1643
1644 2012-05-07  Sheriff Bot  <webkit.review.bot@gmail.com>
1645
1646         Unreviewed, rolling out r116299, r116301, and r116303.
1647         http://trac.webkit.org/changeset/116299
1648         http://trac.webkit.org/changeset/116301
1649         http://trac.webkit.org/changeset/116303
1650         https://bugs.webkit.org/show_bug.cgi?id=85795
1651
1652         Build is still broken (Requested by Ossy on #webkit).
1653
1654         * DumpRenderTree/qt/DumpRenderTree.pro:
1655         * DumpRenderTree/qt/main.cpp:
1656         (main):
1657         * QtTestPlatformPlugin/QtTestPlatformPlugin.pro: Removed.
1658         * QtTestPlatformPlugin/TestIntegration.cpp: Removed.
1659         * QtTestPlatformPlugin/TestIntegration.h: Removed.
1660         * QtTestPlatformPlugin/mac/TestFontDatabase.h: Removed.
1661         * QtTestPlatformPlugin/mac/TestFontDatabase.mm: Removed.
1662         * QtTestPlatformPlugin/mac/TestIntegrationMac.mm: Removed.
1663         * QtTestPlatformPlugin/main.cpp: Removed.
1664         * QtTestPlatformPlugin/testplatform.json: Removed.
1665         * Tools.pro:
1666         * WebKitTestRunner/Target.pri:
1667         * WebKitTestRunner/qt/main.cpp:
1668         (main):
1669
1670 2012-05-07  Balazs Kelemen  <kbalazs@webkit.org>
1671
1672         Unreviewed Qt build fix.
1673
1674         One more try to learn how to use
1675         const with a pointer array.
1676
1677         * DumpRenderTree/qt/main.cpp:
1678         (initializeTestPlatformPlugin):
1679
1680 2012-05-07  Balazs Kelemen  <kbalazs@webkit.org>
1681
1682         Unreviewed build fix after r116299.
1683
1684         * DumpRenderTree/qt/main.cpp:
1685         (initializeTestPlatformPlugin):
1686
1687 2012-05-07  Balazs Kelemen  <kbalazs@webkit.org>
1688
1689         [Qt] Add test specific platform plugin to achieve unified layout test results
1690         https://bugs.webkit.org/show_bug.cgi?id=80996
1691
1692         Reviewed by Simon Hausmann.
1693
1694         Added QtTestPlatformPlugin as a new project under Tools.
1695         This is a Qt5-ish platform plugin that can be used to tweak the
1696         platform support interfaces in order to unify layout test results.
1697         For now it only overrides the font database on Mac and redirects
1698         everything else to the real platform plugin. The font database it
1699         provides mimics the way how we set up test fonts with fontconfig on Linux.
1700         Make DumpRenderTree and WebKitTestRunner use this platform plugin.
1701
1702         * DumpRenderTree/qt/DumpRenderTree.pro:
1703         * DumpRenderTree/qt/main.cpp:
1704         (initializeTestPlatformPlugin):
1705         (main):
1706         * QtTestPlatformPlugin/QtTestPlatformPlugin.pro: Added.
1707         * QtTestPlatformPlugin/TestIntegration.cpp: Added.
1708         (TestIntegration::TestIntegration):
1709         (TestIntegration::fontDatabase):
1710         * QtTestPlatformPlugin/TestIntegration.h: Added.
1711         (TestIntegration):
1712         (TestIntegration::hasCapability):
1713         (TestIntegration::createPlatformPixmap):
1714         (TestIntegration::createPlatformWindow):
1715         (TestIntegration::createPlatformBackingStore):
1716         (TestIntegration::createPlatformOpenGLContext):
1717         (TestIntegration::createPlatformSharedGraphicsCache):
1718         (TestIntegration::guiThreadEventDispatcher):
1719         (TestIntegration::clipboard):
1720         (TestIntegration::drag):
1721         (TestIntegration::inputContext):
1722         (TestIntegration::accessibility):
1723         (TestIntegration::nativeInterface):
1724         (TestIntegration::services):
1725         (TestIntegration::styleHint):
1726         (TestIntegration::platformTheme):
1727         * QtTestPlatformPlugin/mac/TestFontDatabase.h: Added.
1728         (TestFontDatabase):
1729         * QtTestPlatformPlugin/mac/TestFontDatabase.mm: Added.
1730         (TestFontDatabase::populateFontDatabase):
1731         * QtTestPlatformPlugin/mac/TestIntegrationMac.mm: Added.
1732         (TestIntegration::fontDatabase):
1733         * QtTestPlatformPlugin/main.cpp: Added.
1734         (TestIntegrationPlugin::keys):
1735         (TestIntegrationPlugin::create):
1736         (TestIntegrationPlugin::initialize):
1737         * QtTestPlatformPlugin/testplatform.json: Added.
1738         * Tools.pro:
1739         * WebKitTestRunner/Target.pri:
1740         * WebKitTestRunner/qt/main.cpp:
1741         (main):
1742
1743 2012-05-07  Christophe Dumez  <christophe.dumez@intel.com>
1744
1745         [EFL] media/video-poster-blocked-by-willsendrequest.html is flaky
1746         https://bugs.webkit.org/show_bug.cgi?id=85609
1747
1748         Reviewed by Simon Hausmann.
1749
1750         Flakiness was caused by the fact that
1751         media/crash-closing-page-with-media-as-plugin-fallback.html and
1752         media/video-poster-blocked-by-willsendrequest.html use the same PNG
1753         file as poster attribute in the video tag. The PNG file gets cached
1754         when the first test is executed and this somehow causes the second
1755         test to fail.
1756
1757         The flakiness was addressed by clearing the memory and page caches
1758         between tests.
1759
1760         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1761         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1762
1763 2012-05-07  Eric Seidel  <eric@webkit.org>
1764
1765         Add values for all features to Qt's features.pri
1766         https://bugs.webkit.org/show_bug.cgi?id=85746
1767
1768         Reviewed by Tor Arne Vestbø.
1769
1770         When generate-feature-files generates features.pri it provides a value
1771         for every possible ENABLE_.  In prepration for landing generate-feature-files
1772         it seems appropriate to add a value for all possible ENABLEs to the existing
1773         features.pri.  This lets me make sure that I'm getting them all right!
1774
1775         * qmake/mkspecs/features/features.pri:
1776
1777 2012-05-07  Adam Bergkvist  <adam.bergkvist@ericsson.com>
1778
1779         Added myself as a committer.
1780
1781         Not reviewed.
1782
1783         * Scripts/webkitpy/common/config/committers.py:
1784
1785 2012-05-07  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1786
1787         [EFL][DRT] Implement LayoutTestController::execCommand
1788         https://bugs.webkit.org/show_bug.cgi?id=82286
1789
1790         Reviewed by Antonio Gomes.
1791
1792         Add missing implementation execCommand and isCommandEnabled to EFL's
1793         LayoutTestController.
1794
1795         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1796         (LayoutTestController::execCommand):
1797         (LayoutTestController::isCommandEnabled):
1798
1799 2012-05-06  Hayato Ito  <hayato@chromium.org>
1800
1801         [prepare-ChangeLog] Excludes a reference file used by reftests from test files.
1802         https://bugs.webkit.org/show_bug.cgi?id=74268
1803
1804         Reviewed by Ryosuke Niwa.
1805
1806         * Scripts/prepare-ChangeLog:
1807         (generateFileList):
1808
1809 2012-05-06  Mary Wu  <mary.wu@torchmobile.com.cn>
1810
1811         Add a contributor to committers.py
1812         https://bugs.webkit.org/show_bug.cgi?id=85761
1813
1814         Reviewed by Unreviewed.
1815
1816         Add myself to committers.py contributor.
1817
1818         * Scripts/webkitpy/common/config/committers.py:
1819
1820 2012-05-06  Dan Bernstein  <mitz@apple.com>
1821
1822         Tools part of: Building and debugging WebKit in the Xcode IDE requires a lot of setup
1823         https://bugs.webkit.org/show_bug.cgi?id=85739
1824
1825         Reviewed by Daniel Bates.
1826
1827         * Scripts/build-webkit: Moved code that copies from WebKitLibraries to the product directory
1828         from here...
1829         * Scripts/copy-webkitlibraries-to-product-directory: ...to this new script.
1830
1831 2012-05-06  Andy Estes  <aestes@apple.com>
1832
1833         Fix the build after r116246.
1834
1835         * DumpRenderTree/LayoutTestController.cpp: Remove an unused function.
1836
1837 2012-05-06  Eric Seidel  <eric@webkit.org>
1838
1839         Update several build-webkit options to match the #defines they toggle
1840         https://bugs.webkit.org/show_bug.cgi?id=85744
1841
1842         Reviewed by Adam Barth.
1843
1844         Most notable here is fixing --3d-canvas to be --webgl these days --
1845         the define was renamed 15 months ago!
1846
1847         * Scripts/webkitperl/FeatureList.pm:
1848
1849 2012-05-06  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1850
1851         [EFL] EFL's LayoutTestController setJavaScriptCanAccessClipboard implementation
1852         https://bugs.webkit.org/show_bug.cgi?id=83687
1853
1854         Reviewed by Antonio Gomes.
1855
1856         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1857         (LayoutTestController::setJavaScriptCanAccessClipboard):
1858
1859 2012-05-06  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1860
1861         Convert isPageBoxVisible to use Internals interface.
1862         https://bugs.webkit.org/show_bug.cgi?id=85692
1863
1864         Reviewed by Darin Adler.
1865
1866         Remove isPageBoxVisible functions, because it is able to work in the
1867         cross-port way through the Internals interface.
1868
1869         * DumpRenderTree/LayoutTestController.cpp:
1870         (LayoutTestController::staticFunctions):
1871         * DumpRenderTree/LayoutTestController.h:
1872         (LayoutTestController):
1873         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
1874         * DumpRenderTree/chromium/LayoutTestController.cpp:
1875         (LayoutTestController::LayoutTestController):
1876         * DumpRenderTree/chromium/LayoutTestController.h:
1877         (LayoutTestController):
1878         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1879         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1880         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1881         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1882         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1883         (LayoutTestController):
1884         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1885         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1886
1887 2012-05-06  Tom Zakrajsek  <tomz@codeaurora.org>
1888
1889         sheriffbot isn't reopening patches after it lands rollouts
1890         https://bugs.webkit.org/show_bug.cgi?id=64418
1891
1892         Reviewed by Adam Barth.
1893
1894         * Scripts/webkitpy/tool/commands/download_unittest.py:
1895         * Scripts/webkitpy/tool/steps/createbug.py:
1896         (CreateBug.run):
1897
1898 2012-05-04  Raphael Kubo da Costa  <rakuco@webkit.org>
1899
1900         webkitpy: Use os.pathsep instead of manually finding out the path separator in NRWT.
1901         https://bugs.webkit.org/show_bug.cgi?id=85697
1902
1903         Reviewed by Eric Seidel.
1904
1905         * Scripts/new-run-webkit-tests: Instead of manually detecting whether
1906         the path separator should be ':' or ';' by checking sys.platform, rely
1907         on os.pathsep which has exactly the same information (os.pathsep is
1908         ':' for Cygwin, from what I could check in Python's sources).
1909
1910 2012-05-05  Sheriff Bot  <webkit.review.bot@gmail.com>
1911
1912         Unreviewed, rolling out r116223.
1913         http://trac.webkit.org/changeset/116223
1914         https://bugs.webkit.org/show_bug.cgi?id=85718
1915
1916         The bots need to set XDG_DATA_DIRS (Requested by rakuco on
1917         #webkit).
1918
1919         * efl/jhbuildrc:
1920         * gtk/jhbuildrc:
1921
1922 2012-05-05  Raphael Kubo da Costa  <rakuco@webkit.org>
1923
1924         [jhbuild] Do not set XDG_DATA_DIRS in jhbuildrc.
1925         https://bugs.webkit.org/show_bug.cgi?id=85717
1926
1927         Reviewed by Martin Robinson.
1928
1929         After r116209, the XDG_DATA_DIRS environment variable is now
1930         passed by webkitpy's layout_tests code to run-with-jhbuild, so
1931         instead of setting a few defaults in jhbuildrc we can now rely on
1932         the values set and really used by each system.
1933
1934         * efl/jhbuildrc:
1935         * gtk/jhbuildrc:
1936
1937 2012-05-05  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1938
1939         [Qt] Buildfix for --minimal build after r116107.
1940
1941         Reviewed by Csaba Osztrogonác.
1942
1943         * qmake/mkspecs/features/features.prf:
1944
1945 2012-05-05  Zan Dobersek  <zandobersek@gmail.com>
1946
1947         [Gtk] GtkDriver is not required anymore after r116134
1948         https://bugs.webkit.org/show_bug.cgi?id=85699
1949
1950         Reviewed by Csaba Osztrogonác.
1951
1952         Remove GtkDriver as it was replaced by the XvfbDriver and is
1953         no longer needed.
1954
1955         * Scripts/webkitpy/layout_tests/port/gtk.py:
1956
1957 2012-05-05  Zan Dobersek  <zandobersek@gmail.com>
1958
1959         Increase the display ID for the Xvfb process to use if pixel_tests argument
1960         is true rather than if pixel_tests option is present. This fixes the
1961         XvfbDriver in the same way r115825 fixed the now-redundant GtkDriver.
1962
1963         Rubber-stamped by Csaba Osztrogonác.
1964
1965         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1966         (XvfbDriver._start):
1967
1968 2012-05-04  Raphael Kubo da Costa  <rakuco@webkit.org>
1969
1970         webkitpy: Preserve XDG_DATA_DIRS in the environment passed to ServerProcess.
1971         https://bugs.webkit.org/show_bug.cgi?id=85694
1972
1973         Reviewed by Martin Robinson.
1974
1975         This is sort of a follow-up to r100674: $XDG_DATA_DIRS should be
1976         preserved, as it is particularly useful for Linux/Unix
1977         environments.  As a real use case, if shared-mime-info is
1978         installed into a non-trivial prefix, Glib and other libraries will
1979         have trouble to properly identify mimetypes (in ResourceResponses,
1980         for example) as the base freedesktop.org mimetype file will not be
1981         found.
1982
1983         * Scripts/webkitpy/layout_tests/port/base.py:
1984         (Port.to.setup_environ_for_server):
1985
1986 2012-05-04  Gustavo Noronha Silva  <gns@gnome.org>
1987
1988         [GTK] Simplify how libWebCoreModules is linked in, and fix WebKit2 build
1989         https://bugs.webkit.org/show_bug.cgi?id=85691
1990
1991         * GNUmakefile.am: no longer link libWebCoreModules to DumpRenderTree.
1992
1993 2012-05-04  Eric Seidel  <eric@webkit.org>
1994
1995         Make capitalization and descriptions in FeatureList.pm more consistent to make easier to autogenerate
1996         https://bugs.webkit.org/show_bug.cgi?id=85583
1997
1998         Reviewed by Daniel Bates.
1999
2000         * Scripts/webkitperl/FeatureList.pm:
2001
2002 2012-05-04  Eric Seidel  <eric@webkit.org>
2003
2004         Sort entries in FeaturesList.pm to make them easier to autogenerate
2005         https://bugs.webkit.org/show_bug.cgi?id=85584
2006
2007         Reviewed by Daniel Bates.
2008
2009         No functional changes, only sorting.
2010
2011         * Scripts/webkitperl/FeatureList.pm:
2012
2013 2012-05-04  Andy Estes  <aestes@apple.com>
2014
2015         REGRESSION (r115752): WebKit2.NewFirstVisuallyNonEmptyLayout API test times out
2016         https://bugs.webkit.org/show_bug.cgi?id=85675
2017
2018         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayout.cpp:
2019         (TestWebKitAPI::TEST): Disable the test.
2020
2021 2012-04-30  Jer Noble  <jer.noble@apple.com>
2022
2023         Full screen will exit during a provisional load of a non-ancestor iframe.
2024         https://bugs.webkit.org/show_bug.cgi?id=85230
2025
2026         Reviewed by Maciej Stachowiak .
2027
2028         Add support for the new UIDelegate webView:closeFullScreenWithListener: method:
2029         * DumpRenderTree/mac/UIDelegate.mm:
2030         (-[UIDelegate webView:closeFullScreenWithListener:]):
2031
2032         Add support for the new closeFullScreen injected bundle method:
2033         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2034         (WTR::InjectedBundlePage::InjectedBundlePage):
2035         (WTR::InjectedBundlePage::closeFullScreen):
2036         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2037
2038 2012-05-04  Ojan Vafai  <ojan@chromium.org>
2039
2040         Fix test naming to be camel-case like the methods they are testing.
2041
2042         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2043         (test):
2044
2045 2012-05-04  Ojan Vafai  <ojan@chromium.org>
2046
2047         Show the results links in the embedded flakiness dashboard
2048         https://bugs.webkit.org/show_bug.cgi?id=85660
2049
2050         Reviewed by Adam Barth.
2051
2052         Turns out these links don't add much visual noise and it's annoying
2053         to have to go to the dashboard from garden-o-matic just to view the
2054         results for a test.
2055
2056         This is useful for cases where garden-o-matic doesn't properly show the
2057         results (e.g. missing tests). If we address all those cases, then
2058         we could easily hide the results links again if we wanted to.
2059
2060         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2061         (showUpdateInfoForTest):
2062         (htmlForIndividualTestOnAllBuildersWithResultsLinks):
2063         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2064         (test):
2065
2066 2012-05-04  Dirk Pranke  <dpranke@chromium.org>
2067
2068         Revert to using chromium --test-shell mode on SL in NRWT
2069
2070         https://bugs.webkit.org/show_bug.cgi?id=83076
2071
2072         Unreviewed, build fix.
2073
2074         We seem to be seeing a lot more flakiness on the bot since
2075         I switched NRWT to "DRT" mode. We'll try reverting back and
2076         see if this helps things.
2077
2078         * Scripts/webkitpy/layout_tests/port/chromium.py:
2079         (ChromiumDriver.__init__):
2080
2081 2012-05-04  Christophe Dumez  <christophe.dumez@intel.com>
2082
2083         [EFL] Implement layoutTestController.dumpResourceResponseMIMETypes
2084         https://bugs.webkit.org/show_bug.cgi?id=84941
2085
2086         Reviewed by Martin Robinson.
2087
2088         Print MIME type detection messages in EFL's DumpRenderTree if
2089         LayoutTestController's dumpResourceResponseMIMETypes() returns true.
2090
2091         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2092         (DumpRenderTreeChrome::onResponseReceived):
2093
2094 2012-05-04  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2095
2096         [GTK] DRT needs an implementation of LayoutTestController::setDefersLoading and ::goBack
2097         https://bugs.webkit.org/show_bug.cgi?id=85134
2098
2099         Reviewed by Martin Robinson.
2100
2101         Add missing implementation setDefersLoading and goBack to GTK's
2102         LayoutTestController so that the test case rely on these methods
2103         can be enabled.
2104
2105         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2106         (resetDefaultsToConsistentValues):
2107         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2108         (LayoutTestController::goBack):
2109         (LayoutTestController::setDefersLoading):
2110
2111 2012-05-04  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2112
2113         [Qt][NRWT] Run each DRT in it's own xvfb
2114         https://bugs.webkit.org/show_bug.cgi?id=77335
2115
2116         Reviewed by Dirk Pranke.
2117
2118         * Scripts/webkitpy/layout_tests/port/gtk.py:
2119         (GtkPort._driver_class):
2120         * Scripts/webkitpy/layout_tests/port/qt.py:
2121         (QtPort._driver_class):
2122         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py: Added.
2123         (XvfbDriver):
2124         (XvfbDriver._start):
2125         (XvfbDriver._start.x_filter):
2126         (XvfbDriver.stop):
2127
2128 2012-05-04  Zan Dobersek  <zandobersek@gmail.com>
2129
2130         Use destructuring assingment for the return values of the
2131         port.diff_image() method call.
2132
2133         Rubber-stamped by Dirk Pranke.
2134
2135         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
2136         (write_test_result):
2137
2138 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2139
2140         Store fewer runs by default in the flakiness dashboard
2141         https://bugs.webkit.org/show_bug.cgi?id=85568
2142
2143         Reviewed by Dirk Pranke.
2144
2145         This will greatly shrink the size of the json files we load,
2146         resulting in a much faster initial load. People rarely care
2147         about more than 100 runs. For the cases they do, they can still click
2148         the show all runs checkbox to show 500 runs.
2149
2150         * TestResultServer/model/jsonresults.py:
2151
2152 2012-05-04  Christophe Dumez  <christophe.dumez@intel.com>
2153
2154         [EFL] Frame load callbacks output is missing for redirections
2155         https://bugs.webkit.org/show_bug.cgi?id=85173
2156
2157         Reviewed by Antonio Gomes.
2158
2159         Catch the new ewk_frame signals to print the expected output in case
2160         of redirections, when LayoutTestController's dumpFrameLoadCallbacks()
2161         returns true.
2162
2163         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2164         (DumpRenderTreeChrome::createView):
2165         (DumpRenderTreeChrome::onWillSendRequest):
2166         (DumpRenderTreeChrome::onFrameCreated):
2167         (DumpRenderTreeChrome::onFrameRedirectForProvisionalLoad):
2168         (DumpRenderTreeChrome::onFrameRedirectRequested):
2169         (DumpRenderTreeChrome::onResponseReceived):
2170         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
2171         (DumpRenderTreeChrome):
2172
2173 2012-05-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2174
2175         [Qt] Fix up warning about missing GLib/Gio/GStreamer
2176
2177         We don't actually fall back.
2178
2179         Rubber-stamped by Csaba Osztrogonác.
2180
2181         * qmake/mkspecs/features/features.prf:
2182
2183 2012-05-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2184
2185         [Qt] Clarify warning about missing GLib/Gio/GStreamer for media support
2186
2187         And only print it once.
2188
2189         Reviewed by Csaba Osztrogonác.
2190
2191         * qmake/mkspecs/features/features.prf:
2192
2193 2012-05-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2194
2195         [Qt] Clean up and split features.prf into a static list of defaults
2196
2197         The static list of feature defaults is used as a fallback for any
2198         feature that's not dynamically detected or overriden on the command
2199         line (though build-webkit or passing DEFINES+= to qmake).
2200
2201         The static list is complete, which allows for auto-generation based
2202         on Features.py (see bug https://bugs.webkit.org/show_bug.cgi?id=85456)
2203
2204         https://bugs.webkit.org/show_bug.cgi?id=85611
2205
2206         Reviewed by Simon Hausmann.
2207
2208         * Scripts/build-webkit:
2209         * Tools.pro:
2210         * qmake/mkspecs/features/features.prf:
2211         * qmake/mkspecs/features/features.pri: Added.
2212
2213 2012-05-04  Sheriff Bot  <webkit.review.bot@gmail.com>
2214
2215         Unreviewed, rolling out r116085, r116091, and r116095.
2216         http://trac.webkit.org/changeset/116085
2217         http://trac.webkit.org/changeset/116091
2218         http://trac.webkit.org/changeset/116095
2219         https://bugs.webkit.org/show_bug.cgi?id=85628
2220
2221         We are not ready with dependencies on all platform yet (mac) +
2222         problems with debug builds. (Requested by Zoltan on #webkit).
2223
2224         * qmake/mkspecs/features/features.prf:
2225
2226 2012-05-04  Christophe Dumez  <christophe.dumez@intel.com>
2227
2228         [EFL] LayoutTestController does not support overriding WebKitLoadSiteIconsKey
2229         https://bugs.webkit.org/show_bug.cgi?id=85171
2230
2231         Reviewed by Andreas Kling.
2232
2233         Implement support for overriding "WebKitLoadSiteIconsKey" setting in
2234         EFL's LayoutTestController.
2235
2236         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2237         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
2238         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2239         (LayoutTestController::overridePreference):
2240
2241 2012-05-03  Alexis Menard  <alexis.menard@openbossa.org>
2242
2243         [Qt] Enable fullscreen API for WebKit2.
2244         https://bugs.webkit.org/show_bug.cgi?id=85498
2245
2246         Reviewed by Simon Hausmann.
2247
2248         Enable the fullscreen API on the MiniBrowser and turn
2249         it default on WebKit2, disable it on WK1.
2250
2251         * MiniBrowser/qt/BrowserWindow.cpp:
2252         (BrowserWindow::BrowserWindow):
2253         * MiniBrowser/qt/qml/BrowserWindow.qml:
2254         * Scripts/webkitperl/FeatureList.pm:
2255         * qmake/mkspecs/features/features.prf:
2256
2257 2012-05-04  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
2258
2259         REGRESSION(r116008) old-run-webkit-tests broken
2260         https://bugs.webkit.org/show_bug.cgi?id=85597
2261
2262         Reviewed by Simon Hausmann.
2263
2264         Reinstate libraryContainsSymbol.
2265
2266         * Scripts/webkitperl/features.pm:
2267         (libraryContainsSymbol):
2268
2269 2012-05-04  Zoltan Horvath  <zoltan@webkit.org>
2270
2271         [Qt] Set WebCore imagedecoders as default and add fallback to QImageDecoder
2272         https://bugs.webkit.org/show_bug.cgi?id=80400
2273
2274         Remove WTF_USE_QT_IMAGE_DECODER macro.
2275
2276         Reviewed by Simon Hausmann.
2277
2278         * qmake/mkspecs/features/features.prf:
2279
2280 2012-05-03  Christophe Dumez  <christophe.dumez@intel.com>
2281
2282         [EFL] DRT should obey layoutTestController's addURLToRedirect()
2283         https://bugs.webkit.org/show_bug.cgi?id=82722
2284
2285         Reviewed by Maciej Stachowiak.
2286
2287         Implement support for layoutTestController's addURLToRedirect()
2288         in EFL's DRT so that it redirects properly when indicated.
2289
2290         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2291         (DumpRenderTreeChrome::onWillSendRequest):
2292
2293 2012-05-03  Christophe Dumez  <christophe.dumez@intel.com>
2294
2295         [EFL] DRT should support LayoutTestController's willSendRequestReturnsNullOnRedirect()
2296         https://bugs.webkit.org/show_bug.cgi?id=82704
2297
2298         Reviewed by Antonio Gomes.
2299
2300         EFL's DRT needs to obey LayoutTestController's
2301         willSendRequestReturnsNullOnRedirect().
2302
2303         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2304         (DumpRenderTreeChrome::onWillSendRequest):
2305
2306 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2307
2308         git-add-reviewer: Do not put the bug title and its URL in the same line in the commit message.
2309         https://bugs.webkit.org/show_bug.cgi?id=85570
2310
2311         Reviewed by Adam Roben.
2312
2313         The standard format of commit messages and ChangeLog entries has a
2314         short summary in the first line, followed by the bug URL in
2315         Bugzilla. Passing "%s" in the pretty format line to git-rev-list
2316         merges these two lines into a single one separated by a space
2317         character, which is not what we want.
2318
2319         Since "%s%n%n%b" is in practice "%B" (raw message), use it
2320         instead, as the first two lines do not end up being merged
2321         together.
2322
2323         * Scripts/git-add-reviewer:
2324         (writeCommitMessageToFile):
2325
2326 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2327
2328         Improve UI for garden-o-matic examine view when there are fewer than 5 tests.
2329         https://bugs.webkit.org/show_bug.cgi?id=85566
2330
2331         Reviewed by Adam Barth.
2332
2333         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2334         (.):
2335         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2336         (.):
2337         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
2338
2339 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2340
2341         webkitpy: Recognize FreeBSD as a valid platform.
2342         https://bugs.webkit.org/show_bug.cgi?id=81467
2343
2344         Reviewed by Adam Barth.
2345
2346         Add PlatformInfo.is_freebsd() so that trying to use webkit-patch
2347         on FreeBSD does not raise an "unrecognized platform" exception.
2348
2349         Most of this patch contains some unit tests for the mentioned
2350         method, as currently platform-specific code is needed only on
2351         Windows.
2352
2353         * Scripts/webkitpy/common/system/platforminfo.py:
2354         (PlatformInfo.__init__):
2355         (PlatformInfo.is_freebsd):
2356         (PlatformInfo._determine_os_name):
2357         * Scripts/webkitpy/common/system/platforminfo_mock.py:
2358         (MockPlatformInfo.is_freebsd):
2359         * Scripts/webkitpy/common/system/platforminfo_unittest.py:
2360         (fake_platform):
2361         (fake_platform.FakePlatformModule.release):
2362         (TestPlatformInfo.test_real_code):
2363         (TestPlatformInfo.test_os_name_and_wrappers):
2364         (TestPlatformInfo.test_os_version):
2365         (TestPlatformInfo.test_display_name):
2366         (TestPlatformInfo.test_total_bytes_memory):
2367         (TestPlatformInfo.test_free_bytes_memory):
2368         * Scripts/webkitpy/layout_tests/port/factory.py:
2369         (PortFactory._default_port):
2370
2371 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2372
2373         Make the table headers white-space:nowrap. This decreases the vertical size,
2374         which makes for more minimal embedding in garden-o-matic.
2375
2376         * TestResultServer/static-dashboards/flakiness_dashboard_tests.css:
2377         (.table-header-content *):
2378
2379 2012-05-03  Eric Seidel  <eric@webkit.org>
2380
2381         Split build-webkit's feature option list off into its own module in preparation for autogeneration
2382         https://bugs.webkit.org/show_bug.cgi?id=85548
2383
2384         Reviewed by Daniel Bates.
2385
2386         My Perl-fu is very weak.  This moves the feature option list off into a separate module
2387         file so that I can easily autogenerate just that file with the fancy new generate-feature-files command.
2388         It's a bit odd to have an array which we return out to build-webkit to modify, but it turns out
2389         not to matter.  Also, build-webkit doesn't actually use any of the variables we're using to
2390         store the results from the option parse, so those end up inaccessible from build-webkit (except by reference)
2391         which is actually cleaner.  Callers can't assume that they would get a new array every time, but someone
2392         with more Perl-fu than I could create something fancier than this if we need in the future.
2393
2394         * Scripts/build-webkit:
2395         * Scripts/webkitperl/FeatureList.pm: Added.
2396         (getFeatureOptionList):
2397
2398 2012-05-03  Dirk Pranke  <dpranke@chromium.org>
2399
2400         REGRESSION: run-webkit-tests failing on Chromium Mac
2401         https://bugs.webkit.org/show_bug.cgi?id=85459
2402
2403         Unreviewed, build fix.
2404
2405         Turns out select doesn't throw IOErrors, it has it's own kind of
2406         error :(. Trap that instead.
2407
2408         * Scripts/webkitpy/layout_tests/port/server_process.py:
2409         (ServerProcess._wait_for_data_and_update_buffers_using_select):
2410
2411 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2412
2413         [CMake] Rewrite FindCairo.cmake.
2414         https://bugs.webkit.org/show_bug.cgi?id=84895
2415
2416         Reviewed by Daniel Bates.
2417
2418         The old approach relied on pkg-config for finding Cairo (which
2419         introduced a dependency on pkg-config that could be avoided), used
2420         the LibFindMacros code that we should probably remove in the
2421         future and did not use the FindPackageHandleStandardArguments
2422         module.
2423
2424         Change all that by rewriting the module.
2425         - Use the pkg-config output optionally instead of requiring it
2426         like LibFindMacros did.
2427         - Remove the implicit dependency on FreeType which often found it
2428         the wrong way via pkg-config and without considering
2429         CMAKE_PREFIX_PATH.
2430         - Retrieve the Cairo version by looking at cairo-version.h instead
2431         of relying on pkg-config. It requires some additional code for
2432         checking if the desired version has been found, but that will not
2433         be needed once we start depending on CMake 2.8.3 or later.
2434
2435         The only downside is that FPHSA sets <UPPERCASED_NAME>_FOUND
2436         instead of <Name>_FOUND, and to keep things consistent
2437         Cairo_LIBRARIES and Cairo_INCLUDE_DIRS have become CAIRO_LIBRARIES
2438         and CAIRO_INCLUDE_DIRS.
2439
2440         * DumpRenderTree/efl/CMakeLists.txt: Use CAIRO_FOO instead of
2441         Cairo_FOO.
2442         * EWebLauncher/CMakeLists.txt: Ditto.
2443
2444 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2445
2446         Fix modifier and bug sorting on the flakiness dashboard
2447         https://bugs.webkit.org/show_bug.cgi?id=85554
2448
2449         Reviewed by Dirk Pranke.
2450
2451         The sorting assumes that the column header name matches a field
2452         in the testResults object and uses that field for sorting.
2453         Filter the bugs out of the modifiers list and put them in a bugs field
2454         and only put the non-bug modifiers in the modifiers list.
2455
2456         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2457         (populateExpectationsData):
2458         (htmlForSingleTestRow):
2459         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2460
2461 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2462
2463         Fix some state resetting when running unittests. HtmlForSingleTestRow
2464         was flaky.
2465
2466         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2467         (htmlForSingleTestRow):
2468         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2469         (resetGlobals):
2470
2471 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2472
2473         webkitpy: Use PlatformInfo wherever possible in the User class.
2474         https://bugs.webkit.org/show_bug.cgi?id=85549
2475
2476         Reviewed by Dirk Pranke.
2477
2478         Instead of relying on sys.platform to detect whether the current
2479         platform is a Mac, use the PlatformInfo class instead.
2480
2481         * Scripts/webkitpy/common/system/user.py:
2482         (User.__init__):
2483         (User.edit_changelog):
2484
2485 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2486
2487         Remove the percent failed column from the flakiness dashboard
2488         https://bugs.webkit.org/show_bug.cgi?id=85552
2489
2490         Reviewed by Dirk Pranke.
2491
2492         This column isn't useful enough to justify the visual noise.
2493
2494         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2495         (processMissingAndExtraExpectations):
2496         (tableHeaders):
2497         (htmlForSingleTestRow):
2498         (sortTests):
2499         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2500         (test):
2501
2502 2012-05-03  Eric Seidel  <eric@webkit.org>
2503
2504         --coverage should not not use the "feature" options system in build-webkit, similarly remove special-casing for WTF_URL
2505         https://bugs.webkit.org/show_bug.cgi?id=85542
2506
2507         Reviewed by Daniel Bates.
2508
2509         Remove these last two odd-men-out, so that we can generate the default feature options
2510         for build-webkit using the new generate-feature-files command.
2511         The Mac build system has some special casing for ENABLE_ defines which gets them
2512         defined as -D on the build command, even though initially they start out as
2513         environment variables ENABLE_FOO = ENABLE_FOO.  The Feature options in build-webkit
2514         depend on this support, but it doesn't work for WTF_USE_ defines like WTF_USE_WTF_URL.
2515         I'm told that WTF_URL is not actively being worked on, so just removing it for now.
2516         It should probably come back to life as an ENABLE_, or build-webkit or the Xcode projects
2517         should be made generically smarter about these USE_WTF defines.
2518         (There is still WTF_USE_TILED_BACKING_STORE in the feature list, but it only seems
2519         used by Qt.  If someone tried build-webkit --tiled-backing-store on Mac today it
2520         would not work!)
2521
2522         * Scripts/build-webkit:
2523
2524 2012-05-03  Eric Seidel  <eric@webkit.org>
2525
2526         Remove the last bits of Leopard support from our perl scripts
2527         https://bugs.webkit.org/show_bug.cgi?id=85545
2528
2529         Reviewed by Daniel Bates.
2530
2531         Chromium still supports Leopard (for a little while yet), but AppleMacWebKit dropped
2532         Leopard support several months ago.  Remove the last bits of isLeopard() branching
2533         in our Perl scripts, since no one is using this (Chromium has never used this Perl code).
2534
2535         * Scripts/build-webkit:
2536         * Scripts/old-run-webkit-tests:
2537         (countAndPrintLeaks):
2538         (captureSavedCrashLog):
2539         * Scripts/webkitdirs.pm:
2540         (determineArchitecture):
2541         (argumentsForRunAndDebugMacWebKitApp):
2542
2543 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2544
2545         [Chromium] Don't show the WebKit Linux ASAN bot on the flakiness dashboard.
2546         https://bugs.webkit.org/show_bug.cgi?id=85551
2547
2548         Reviewed by Dirk Pranke.
2549
2550         It's not uploading results and in the short term we don't plan to have it start.
2551
2552         * TestResultServer/static-dashboards/builders.js:
2553         (isChromiumWebkitTipOfTreeTestRunner):
2554         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2555
2556 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2557
2558         Show the test name in the dashboard if showChrome==false and we're showing multiple tests
2559         https://bugs.webkit.org/show_bug.cgi?id=85243
2560
2561         Reviewed by Dirk Pranke.
2562
2563         This happens for cases where a test name matches multiple tests.
2564         In practice, this only happens for virtual test suites (e.g. the canvas tests).
2565
2566         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2567         (htmlForIndividualTestOnAllBuildersWithChrome):
2568         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2569         (testHtmlForIndividualTestOnAllBuildersWithChromeNonexistant):
2570         (testHtmlForIndividualTestOnAllBuildersWithChrome):
2571         (testHtmlForIndividualTestOnAllBuildersWithChromeWebkitMaster):
2572         (testHtmlForIndividualTests):
2573
2574 2012-05-03  Ojan Vafai  <ojan@chromium.org>
2575
2576         Port the flakiness dashboard unittests to qunit
2577         https://bugs.webkit.org/show_bug.cgi?id=85544
2578
2579         Reviewed by Adam Barth.
2580
2581         * TestResultServer/static-dashboards/dashboard_base.js:
2582         (appendJSONScriptElements):
2583         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2584         (resetGlobals):
2585         (runExpectationsTest):
2586         (test):
2587         * TestResultServer/static-dashboards/run-unittests.html: Added.
2588
2589 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2590
2591         webkitpy: Remove check for readline on Mac.
2592         https://bugs.webkit.org/show_bug.cgi?id=85547
2593
2594         Reviewed by Eric Seidel.
2595
2596         Python 2.5 is not supported by the WebKit project anymore, and
2597         Python's readline documentation does not mention the module not
2598         being available in the supported versions.
2599
2600         * Scripts/webkitpy/common/system/user.py:
2601
2602 2012-05-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2603
2604         [NRWT] Run performance tests with lock
2605         https://bugs.webkit.org/show_bug.cgi?id=78628
2606
2607         Reviewed by Tony Chang.
2608
2609         Locking performance tests (like we do for http tests) will force them
2610         to run in serial. This reduces the load of the machine when running perf
2611         tests and minimizes the chances of the tests to fail, specially
2612         timeouts.
2613
2614         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2615         (Manager.__init__):
2616         (Manager._is_perf_test):
2617         (Manager._test_requires_lock):
2618         * Scripts/webkitpy/layout_tests/port/http_lock.py:
2619
2620 2012-05-03  Dirk Pranke  <dpranke@chromium.org>
2621
2622         Attempt to fix a crash on SL ...
2623
2624         Unreviewed, build fix.
2625
2626         It looks like we may be getting an IOError raised during
2627         select() on chromium mac, and that error isn't being caught.
2628         Pushing it into the try/catch block will hopefully catch this
2629         and allow us to see why the subprocess is crashing.
2630
2631         * Scripts/webkitpy/layout_tests/port/server_process.py:
2632         (ServerProcess._wait_for_data_and_update_buffers_using_select):
2633
2634 2012-05-02  Jon Honeycutt  <jhoneycutt@apple.com>
2635
2636         WebFrameLoaderClient::dispatchWillSendSubmitEvent() needs to be
2637         implemented for WebKit2
2638         https://bugs.webkit.org/show_bug.cgi?id=84304
2639
2640         Reviewed by Jessie Berlin.
2641
2642         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2643         Added new files to project.
2644
2645         * TestWebKitAPI/Tests/WebKit2/WillSendSubmitEvent.cpp: Added.
2646         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
2647         Look for the "DidReceiveWillSendSubmitEvent" message. The message body
2648         will be the dictionary of form control names and values passed to the
2649         injected bundle's willSendSubmitEvent handler; look for the pairs
2650         "textField" -> "text field" and "passwordField" -> "password field".
2651         Check that "hiddenField" is not included.
2652         (TestWebKitAPI::setInjectedBundleClient):
2653         (TestWebKitAPI::TEST):
2654
2655         * TestWebKitAPI/Tests/WebKit2/WillSendSubmitEvent_Bundle.cpp: Added.
2656         (TestWebKitAPI::willSendSubmitEvent):
2657         Post a message to the TestWebKitAPI process with the dictionary of form
2658         control names and values that we received.
2659         (TestWebKitAPI::WillSendSubmitEventTest::WillSendSubmitEventTest):
2660         (TestWebKitAPI::WillSendSubmitEventTest::didCreatePage):
2661         Set up the page's form client.
2662
2663         * TestWebKitAPI/Tests/WebKit2/auto-submitting-form.html: Added.
2664
2665 2012-05-03  Eric Seidel  <eric@webkit.org>
2666
2667         Remove dead code from build-webkit for working around old XCode dependency bug
2668         https://bugs.webkit.org/show_bug.cgi?id=85535
2669
2670         Reviewed by Adam Barth.
2671
2672         Xcode used to have a bug whereby it didn't notice if you changed the #defines
2673         passed on the command line to xcodebuild.  I wrote code to work around this
2674         while bringing up SVG support 5-6 years ago.  I think we can finally remove this
2675         code from build-webkit, since SVG is the only feature flag to ever use it
2676         (modern in-progress features like MathML don't seem to be using the support)
2677         an I suspect the XCode bug has long been fixed.
2678
2679         * Scripts/build-webkit:
2680         * Scripts/webkitperl/features.pm:
2681
2682 2012-05-03  Andy Estes  <aestes@apple.com>
2683
2684         InjectedBundleControllerMac.mm needs to include config.h
2685         https://bugs.webkit.org/show_bug.cgi?id=85531
2686
2687         Reviewed by Andreas Kling.
2688
2689         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm:
2690
2691 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2692
2693         webkitpy: Use PlatformInfo whenever possible in server_process.py.
2694         https://bugs.webkit.org/show_bug.cgi?id=85516
2695
2696         Reviewed by Eric Seidel.
2697
2698         Eliminate direct access to sys.platform wherever possible by using
2699         PlatformInfo.is_win() instead.
2700
2701         * Scripts/webkitpy/layout_tests/port/server_process.py:
2702         (ServerProcess.__init__): Remove the executive parameter as it was
2703         not passed by any caller, obtain a SystemHost from port_obj
2704         instead.
2705         (ServerProcess._log): Style, remove extra empty line.
2706         (ServerProcess._start):
2707         (ServerProcess.stop):
2708         (ServerProcess.kill):
2709         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
2710         (TrivialMockPort.__init__): Create a MockSystemHost used by
2711         ServerProcess.
2712         (TestServerProcess.test_broken_pipe): Test different platforms.
2713
2714 2012-05-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2715
2716         webkitpy: Style checker broken when committing a PNG
2717         https://bugs.webkit.org/show_bug.cgi?id=85504
2718
2719         Reviewed by Dirk Pranke.
2720
2721         * Scripts/webkitpy/style/checkers/png.py:
2722         (PNGChecker.__init__):
2723
2724 2012-05-03  Philippe Normand  <pnormand@igalia.com>
2725
2726         Unreviewed, fix another call to join(), similar to the ones landed
2727         in r115975.
2728
2729         * Scripts/webkitdirs.pm:
2730         (jhbuildConfigurationChanged):
2731
2732 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2733
2734         webkitpy: Remove sys.platform accesses in CrashLogs.
2735         https://bugs.webkit.org/show_bug.cgi?id=85436
2736
2737         Reviewed by Dirk Pranke.
2738
2739         Instead of directly checking for sys.platform == 'darwin', use a
2740         SystemHost and call PlatformInfo.is_mac().
2741
2742         * Scripts/webkitpy/common/system/crashlogs.py:
2743         (CrashLogs.__init__):
2744         (CrashLogs.find_newest_log):
2745         (CrashLogs._log_directory_darwin):
2746         (CrashLogs._find_newest_log_darwin):
2747         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
2748         (CrashLogsTest.test_find_log_darwin):
2749         * Scripts/webkitpy/layout_tests/port/mac.py:
2750         (MacPort._get_crash_log):
2751         * Scripts/webkitpy/layout_tests/port/test.py:
2752         (TestDriver.run_test):
2753         * Scripts/webkitpy/tool/commands/queries.py:
2754         (execute):
2755
2756 2012-05-03  Philippe Normand  <pnormand@igalia.com>
2757
2758         [GTK] join() is wrongly used in webkitdirs.pm since r115532
2759         https://bugs.webkit.org/show_bug.cgi?id=85501
2760
2761         Reviewed by Martin Robinson.
2762
2763         * Scripts/webkitdirs.pm:
2764         (getJhbuildPath):
2765         (buildAutotoolsProject):
2766
2767 2012-05-03  Zan Dobersek  <zandobersek@gmail.com>
2768
2769         [Gtk] ImageDiff crashes when it can't open a display
2770         https://bugs.webkit.org/show_bug.cgi?id=85476
2771
2772         Reviewed by Martin Robinson.
2773
2774         Replace gdk_init() initialization call with the g_type_init()
2775         call. GdkPixbufLoader doesn't actually need the current call to be
2776         made and this change will also remove crashes when ImageDiff
2777         is started in an environment without a DISPLAY env.
2778
2779         * DumpRenderTree/gtk/ImageDiff.cpp:
2780         (main):
2781
2782 2012-05-03  Zan Dobersek  <zandobersek@gmail.com>
2783
2784         ImageDiff should be run inside a properly established environment
2785         https://bugs.webkit.org/show_bug.cgi?id=85292
2786
2787         Reviewed by Martin Robinson.
2788
2789         When creating the ImageDiff server process, pass along a freshly-set-up environment
2790         in which the process should be executed in.
2791
2792         * Scripts/webkitpy/layout_tests/port/webkit.py:
2793         (WebKitPort._start_image_diff_process):
2794
2795 2012-05-03  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2796
2797         [Qt] Allow the web process and WTR to be paused on startup
2798
2799         Makes it easier to debug the web process or run-webkit-tests -2, as you
2800         have ample time to attach gdb to the process.
2801
2802         Reviewed by Simon Hausmann.
2803
2804         * Scripts/webkitpy/layout_tests/port/qt.py:
2805         (QtPort.setup_environ_for_server):
2806         * WebKitTestRunner/qt/main.cpp:
2807         (main):
2808
2809 2012-05-03  Csaba Osztrogonác  <ossy@webkit.org>
2810
2811         Password protect "Stop build" button
2812         https://bugs.webkit.org/show_bug.cgi?id=81982
2813
2814         Reviewed by Ryosuke Niwa.
2815
2816         Only authenticated users should be able to use stopBuild, stopAllBuilds, cancelPendingBuild
2817         features to prevent killing builds on build.webkit.org by SPAM bots.
2818
2819         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2820
2821 2012-05-03  Dongwoo Im  <dw.im@samsung.com>
2822
2823         [DRT] Remove the name of parameter, visibility, from the setPageVisibility stub implementations.
2824         https://bugs.webkit.org/show_bug.cgi?id=85468
2825
2826         Reviewed by Nikolas Zimmermann.
2827
2828         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
2829         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2830         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2831         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2832         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2833         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2834         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2835         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2836         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2837         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
2838
2839 2012-05-03  Stephanie Lewis  <slewis@apple.com>
2840
2841         https://bugs.webkit.org/show_bug.cgi?id=85450 unbounded growth of JSDOMWindowShells loading pages in the same window
2842         <rdar://problem/11320059> REGRESSION (r115083): PLT3 shows linear memory growth and gets slower with each run
2843
2844         Reviewed by Brady Eidson.
2845
2846         The API added for DOMWindowExtension, didCreateGlobalObjectForFrame, would create a global object
2847         for every world, even those that did not need the callback.  This had the side effect of creating a
2848         JSDOMWindowShell that the associated world didn't necessarily know to clean up.  Instead of creating
2849         unnecessary objects change the API to globalObjectIsAvailableForFrame and do not pass the global object
2850         in the API.  The object can be accessed later by those worlds which require it.
2851
2852         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic.cpp:
2853         (TestWebKitAPI):
2854         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
2855         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
2856         (TestWebKitAPI):
2857         (DOMWindowExtensionBasic):
2858         (TestWebKitAPI::DOMWindowExtensionBasic::didCreatePage):
2859         (TestWebKitAPI::DOMWindowExtensionBasic::globalObjectIsAvailableForFrame):
2860         (TestWebKitAPI::globalObjectIsAvailableForFrameCallback):
2861
2862 2012-05-03  Nikolas Zimmermann  <nzimmermann@rim.com>
2863
2864         Not reviewed. Fix mac build, it was missing setPageVisibility/resetPageVisibility stub implementations.
2865
2866         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2867         (LayoutTestController::resetPageVisibility):
2868         (LayoutTestController::setPageVisibility):
2869
2870 2012-05-02  Alexander Færøy  <ahf@0x90.dk>
2871
2872         Rename deviceDPI to devicePixelRatio
2873         https://bugs.webkit.org/show_bug.cgi?id=85049
2874
2875         Reviewed by Kenneth Rohde Christiansen.
2876
2877         Use 1.5 as device pixel ratio for the Qt MiniBrowser.
2878
2879         * MiniBrowser/qt/qml/BrowserWindow.qml:
2880
2881 2012-05-03  Dongwoo Im  <dw.im@samsung.com>
2882
2883         [EFL][DRT] Implement the LayoutTestController's methods related to the Page Visibility API.
2884         https://bugs.webkit.org/show_bug.cgi?id=85347
2885
2886         Reviewed by Nikolas Zimmermann.
2887
2888         * DumpRenderTree/LayoutTestController.h:
2889         (LayoutTestController):
2890         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
2891         (LayoutTestController::resetPageVisibility): Not implemented.
2892         (LayoutTestController::setPageVisibility): Not implemented.
2893         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2894         (LayoutTestController::resetPageVisibility): Reset the page visibility.
2895         (LayoutTestController::setPageVisibility): Set the page visibility.
2896         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2897         (LayoutTestController::resetPageVisibility): Not implemented.
2898         (LayoutTestController::setPageVisibility): Not implemented.
2899         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2900         (LayoutTestController::resetPageVisibility): Not implemented.
2901         (LayoutTestController::setPageVisibility): Not implemented.
2902         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2903         (LayoutTestController::resetPageVisibility): Not implemented.
2904         (LayoutTestController::setPageVisibility): Not implemented.
2905         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2906         (LayoutTestController::resetPageVisibility): Not implemented.
2907         (LayoutTestController::setPageVisibility): Not implemented.
2908
2909 2012-05-02  Sheriff Bot  <webkit.review.bot@gmail.com>
2910
2911         Unreviewed, rolling out r115918.
2912         http://trac.webkit.org/changeset/115918
2913         https://bugs.webkit.org/show_bug.cgi?id=85452
2914
2915         "Broke test-webkitpy on Apple Lion bots" (Requested by estes
2916         on #webkit).
2917
2918         * Scripts/webkitpy/common/system/crashlogs.py:
2919         (CrashLogs.__init__):
2920         (CrashLogs.find_newest_log):
2921         (CrashLogs._log_directory_darwin):
2922         (CrashLogs._find_newest_log_darwin):
2923         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
2924         (CrashLogsTest.test_find_log_darwin):
2925         * Scripts/webkitpy/layout_tests/port/mac.py:
2926         (MacPort._get_crash_log):
2927         * Scripts/webkitpy/layout_tests/port/test.py:
2928         (TestDriver.run_test):
2929         * Scripts/webkitpy/tool/commands/queries.py:
2930         (execute):
2931
2932 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
2933
2934         webkitpy: Remove sys.platform accesses in CrashLogs.
2935         https://bugs.webkit.org/show_bug.cgi?id=85436
2936
2937         Reviewed by Dirk Pranke.
2938
2939         Instead of directly checking for sys.platform == 'darwin', use a
2940         SystemHost and call PlatformInfo.is_mac().
2941
2942         * Scripts/webkitpy/common/system/crashlogs.py:
2943         (CrashLogs.__init__):
2944         (CrashLogs.find_newest_log):
2945         (CrashLogs._log_directory_darwin):
2946         (CrashLogs._find_newest_log_darwin):
2947         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
2948         (CrashLogsTest.test_find_log_darwin):
2949         * Scripts/webkitpy/layout_tests/port/mac.py:
2950         (MacPort._get_crash_log):
2951         * Scripts/webkitpy/layout_tests/port/test.py:
2952         (TestDriver.run_test):
2953         * Scripts/webkitpy/tool/commands/queries.py:
2954         (execute):
2955
2956 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
2957
2958         webkitpy: Special-case win32 calls in file_lock and executive.
2959         https://bugs.webkit.org/show_bug.cgi?id=85431
2960
2961         Reviewed by Dirk Pranke.
2962
2963         Simplify accesses to sys.platform by only checking if sys.platform
2964         is "win32" and removing explicit checks for the else case.
2965
2966         * Scripts/webkitpy/common/system/executive.py:
2967         (Executive.check_running_pid):
2968         * Scripts/webkitpy/common/system/file_lock.py:
2969         (FileLock._create_lock):
2970         (FileLock._remove_lock):
2971
2972 2012-05-02  Ojan Vafai  <ojan@chromium.org>
2973
2974         Remove pink from flakiness dashboard
2975         https://bugs.webkit.org/show_bug.cgi?id=85422
2976
2977         Reviewed by Dirk Pranke.
2978
2979         The pink indicates that the test doesn't match it's listing in test_expectations.txt.
2980         In practice, this just added noise and confused people.
2981
2982         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2983         (htmlForSingleTestRow):
2984         (showLegend):
2985         * TestResultServer/static-dashboards/flakiness_dashboard_tests.css:
2986         (#legend-contents *):
2987         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2988         (testHtmlForSingleTestRow):
2989
2990 2012-05-02  Gustavo Noronha Silva  <gns@gnome.org>
2991
2992         Unreviewed. Fix mistake I committed when applying review comments.
2993
2994         * jhbuild/jhbuild-wrapper: dependencies_path was getting 'Root'
2995         appended to its path, so jhbuild itself was being cloned and
2996         installed in the wrong place when using WEBKITOUTPUTDIR.
2997
2998 2012-05-02  Dirk Pranke  <dpranke@chromium.org>
2999
3000         nrwt: make ServerProcess work on chromium win
3001         https://bugs.webkit.org/show_bug.cgi?id=85333
3002
3003         Reviewed by Ojan Vafai.
3004
3005         This change implements a poor man's select() that will
3006         slow-spin doing non-blocking reads on the stdout and stderr
3007         named pipes connecting the worker to the driver. Seems to work
3008         and I have yet to see much overhead or ill effects but it
3009         probably needs more testing.
3010
3011         * Scripts/webkitpy/layout_tests/port/chromium.py:
3012         (ChromiumDriver.__init__):
3013         * Scripts/webkitpy/layout_tests/port/server_process.py:
3014         (ServerProcess._start):
3015         (ServerProcess._wait_for_data_and_update_buffers_using_select):
3016         (ServerProcess._wait_for_data_and_update_buffers_using_win32_apis):
3017         (ServerProcess):
3018         (ServerProcess._non_blocking_read_win32):
3019         (ServerProcess._read):
3020         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
3021         (TestServerProcess.test_basic): Added.
3022
3023 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3024
3025         webkitpy: Use a SystemHost instead of PlatformInfo in the PNG checker.
3026         https://bugs.webkit.org/show_bug.cgi?id=85426
3027
3028         Reviewed by Dirk Pranke.
3029
3030         r115880 replaced the direct use of sys.platform with a
3031         PlatformInfo object. Dirk Pranke later told me SystemHost objects
3032         should be preferred, as nothing out of webkitpy.common.system
3033         should create its own PlatformInfo instances.
3034
3035         * Scripts/webkitpy/common/system/systemhost_mock.py:
3036         (MockSystemHost.__init__): Make it possible to specify a custom
3037         filesystem object.
3038         * Scripts/webkitpy/style/checkers/png.py:
3039         (PNGChecker.__init__): Accept a host object instead of a
3040         filesystem one, which is now obtained from host.
3041         (PNGChecker._config_file_path): Retrieve PlatformInfo from
3042         SystemHost.
3043         * Scripts/webkitpy/style/checkers/png_unittest.py:
3044         (PNGCheckerTest.test_init): Adjust calls to PNGChecker to its new
3045         constructor signature.
3046         (PNGCheckerTest.test_check): Ditto.
3047
3048 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3049
3050         webkitpy: Remove direct checks for sys.platform == 'darwin' in the layout test classes.
3051         https://bugs.webkit.org/show_bug.cgi?id=85424
3052
3053         Reviewed by Tony Chang.
3054
3055         We should reduce direct usage of sys.platform in webkitpy and use
3056         PlatformInfo instead.
3057
3058         * Scripts/webkitpy/layout_tests/servers/http_server.py:
3059         (Lighttpd._prepare_config): Call PlatformInfo.is_mac() instead of
3060         checking for 'darwin' directly.
3061         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
3062         (HttpServerBase.__init__): Ditto.
3063
3064 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3065
3066         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
3067         https://bugs.webkit.org/show_bug.cgi?id=85419
3068
3069         Reviewed by Dirk Pranke.
3070
3071         As pointed out by Dirk Pranke, a PlatformInfo is not supposed to
3072         be created directly, we should should a SystemHost instead.
3073
3074         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3075         (MainTest.setUp): Create a SystemHost and obtain the PlatformInfo
3076         object from it instead of creating PlatformInfo directly.
3077
3078 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3079
3080         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
3081         https://bugs.webkit.org/show_bug.cgi?id=85419
3082
3083         Rubber-stamped by Dirk Pranke.
3084
3085         Follow-up patch to r115887 which fixes some capitalization problems
3086         pointed out in comments 2 and 4.
3087
3088         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3089         (MainTest.setUp):
3090         (MainTest.test_child_processes_2):
3091         (MainTest.test_child_processes_min):
3092         (MainTest.test_exception_raised):
3093         (MainTest.test_keyboard_interrupt):
3094
3095 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3096
3097         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
3098         https://bugs.webkit.org/show_bug.cgi?id=85419
3099
3100         Reviewed by Dirk Pranke.
3101
3102         Another part of the work towards replacing direct sys.platform
3103         usage with PlatformInfo.
3104
3105         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3106         (MainTest.setUp): Move the setting of SHOULD_TEST_PROCESSES
3107         here. Even though it is costlier, unittest.setUpClass was
3108         introduced in Python 2.7.
3109         (MainTest.test_child_processes_2): Use self.SHOULD_TEST_PROCESSES.
3110         (MainTest.test_child_processes_min): Ditto.
3111         (MainTest.test_exception_raised): Ditto.
3112         (MainTest.test_keyboard_interrupt): Ditto.
3113         (MainTest.test_crash_log): Check for PlatformInfo.is_mac().
3114         (MainTest.test_web_process_crash_log): Ditto, and remove duplicate
3115         check at the end.
3116         (MainTest.test_no_http_tests): Remove check for (cygwin && Python
3117         2.5.2), as we only support Python 2.6+.
3118
3119 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3120
3121         webkitpy: Remove direct sys.platform usage in the PNG checker.
3122         https://bugs.webkit.org/show_bug.cgi?id=85407
3123
3124         Reviewed by Tony Chang.
3125
3126         Work towards replacing direct sys.platform usage with
3127         PlatformInfo.
3128
3129         * Scripts/webkitpy/common/system/platforminfo_mock.py:
3130         (MockPlatformInfo.__init__): Accept os_name and os_version as
3131         arguments.
3132         * Scripts/webkitpy/style/checkers/png.py:
3133         (PNGChecker.__init__): Create a PlatformInfo object instead of
3134         using sys.platform when needed.
3135         (PNGChecker._config_file_path): Check for the current platform via
3136         PlatformInfo.
3137         * Scripts/webkitpy/style/checkers/png_unittest.py:
3138         (PNGCheckerTest.test_check): Update tests after the changes above.
3139
3140 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
3141
3142         [watchlist] Update watchlist rules for EFL-related patches.
3143         http://bugs.webkit.org/show_bug.cgi?id=85160
3144
3145         Reviewed by Eric Seidel.
3146
3147         * Scripts/webkitpy/common/config/watchlist: Monitor
3148         Source/WTF/wtf/efl and LayoutTests/platform/efl as well.
3149
3150 2012-05-02  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3151
3152         Revert r115191 - "Make the web view's url property follow the active url"
3153
3154         It uncovered/caused issues in the icon-implementation that can't be
3155         easily fixed, so rolling out instead.
3156
3157         https://bugs.webkit.org/show_bug.cgi?id=77554
3158
3159         Rubber-stamped by Simon Hausmann.
3160
3161         * MiniBrowser/qt/qml/BrowserWindow.qml:
3162         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
3163         (TestWebKitAPI::TEST):
3164
3165 2012-05-02  Zan Dobersek  <zandobersek@gmail.com>
3166
3167         Reftest mismatches are providing undefined diff percent in results.html
3168         https://bugs.webkit.org/show_bug.cgi?id=85283
3169
3170         Reviewed by Ojan Vafai.
3171
3172         When writing test results for a reftest mismatch, also take into account the
3173         difference percent that is produced when generating image diff. This value is
3174         then used when creating test results in the form of dictionaries, stored under
3175         the 'image_diff_percent' key. It ultimately ends up in JSON results data,
3176         providing additional information about the reftest difference.
3177
3178         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3179         (interpret_test_failures): Store the diff percent for reftest failure under the
3180         'image_diff_percent' key.
3181         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
3182         (write_test_result): Also set the new diff_percent attribute of the reftest
3183         mismatch failure to the value of the difference percent ImageDiff outputs.
3184         * Scripts/webkitpy/layout_tests/models/test_failures.py:
3185         (FailureReftestMismatch.__init__): Add the new diff_percent attribute with the
3186         default value of None.
3187         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3188         (EndToEndTest.test_reftest_with_two_notrefs): Update the test case expected results.
3189
3190 2012-05-02  Terry Anderson  <tdanderson@chromium.org>
3191
3192         Addition of a contributor to committers.py
3193         https://bugs.webkit.org/show_bug.cgi?id=85324
3194
3195         Reviewed by Unreviewed.
3196
3197         * Scripts/webkitpy/common/config/committers.py:
3198
3199 2012-05-02  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3200
3201         [EFL] EFL's LayoutTestController overridePreference implementation
3202         https://bugs.webkit.org/show_bug.cgi?id=83007
3203
3204         Reviewed by Antonio Gomes.
3205
3206         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3207         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
3208         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3209         (toBool): aux function converting JSStringRef to bool
3210         (toInt): aux function converting JSStringRef to int
3211         (LayoutTestController::overridePreference): added implementation
3212
3213 2012-05-01  Raphael Kubo da Costa  <rakuco@webkit.org>
3214
3215         webkitpy: Fix sys.platform check in find_files_unittest.
3216         https://bugs.webkit.org/show_bug.cgi?id=85339
3217
3218         Reviewed by Ryosuke Niwa.
3219
3220         Follow-up to r101618 -- 'win' is not a valid sys.platform value,
3221         'win32' should be used instead. The check always returned false
3222         due to this, so assert_normalize() was never called.
3223
3224         * Scripts/webkitpy/common/find_files_unittest.py:
3225         (TestWinNormalize.test_win): Add the missing import for the
3226         FileSystem class and fix the sys.platform check.
3227
3228 2012-05-02  Zan Dobersek  <zandobersek@gmail.com>
3229
3230         [GTK] drag&drop related failures
3231         https://bugs.webkit.org/show_bug.cgi?id=84675
3232
3233         Reviewed by Philippe Normand.
3234
3235         Rather than checking options to see whether we're running pixel tests
3236         the pixel_tests argument passed to GtkDriver._start should be used as
3237         it also covers cases when pixel tests should be run because of testing
3238         reftests. This again creates different Xvfb processes with proper
3239         display IDs for a worker that's performing both ordinary and pixel
3240         tests, as was the case before changes made in r114788.
3241
3242         * Scripts/webkitpy/layout_tests/port/gtk.py:
3243         (GtkDriver._start):
3244
3245 2012-05-02  Simon Hausmann  <simon.hausmann@nokia.com>
3246
3247         [Qt] Enable device orientation/motion with Qt 5
3248
3249         Reviewed by Tor Arne Vestbø.
3250
3251         Enable device orientation/motion features with Qt 5 if Qt Sensors is available.
3252
3253         * qmake/mkspecs/features/features.prf:
3254
3255 2012-05-01  Raphael Kubo da Costa  <rakuco@webkit.org>
3256
3257         webkitpy: Remove unused platform imports.
3258         https://bugs.webkit.org/show_bug.cgi?id=85341
3259
3260         Reviewed by Eric Seidel.
3261
3262         Remove a few uses of `import platform' which were either rendered
3263         unneeded after r105931 or were never needed at all.
3264
3265         * Scripts/webkitpy/common/system/executive.py:
3266         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
3267
3268 2012-05-01  Terry Anderson  <tdanderson@chromium.org>
3269
3270         [chromium] Accept four parameters when dispatching a WebInputEvent::GestureTap in chromium DRT eventSender
3271         https://bugs.webkit.org/show_bug.cgi?id=85289
3272
3273         Reviewed by Eric Seidel.
3274
3275         * DumpRenderTree/chromium/EventSender.cpp:
3276         (EventSender::gestureEvent):
3277             This change will allow for testing of WK85101. Currently
3278             no tests use the extra parameters deltaX and deltaY. Added
3279             a check for the size of |arguments|.
3280
3281 2012-05-01  Eric Seidel  <eric@webkit.org>
3282
3283         Add myself to the watchlist to watch all rendering changes, as well as sort the watchers in the list
3284         https://bugs.webkit.org/show_bug.cgi?id=85305
3285
3286         Reviewed by Adam Barth.
3287
3288         * Scripts/webkitpy/common/config/watchlist:
3289
3290 2012-05-01  Sheriff Bot  <webkit.review.bot@gmail.com>
3291
3292         Unreviewed, rolling out r115735.
3293         http://trac.webkit.org/changeset/115735
3294         https://bugs.webkit.org/show_bug.cgi?id=85314
3295
3296         cause two tests to crash due to an ASSERTION failure
3297         (Requested by zhenyao on #webkit).
3298
3299         * DumpRenderTree/chromium/EventSender.cpp:
3300         (EventSender::gestureEvent):
3301
3302 2012-05-01  Terry Anderson  <tdanderson@chromium.org>
3303
3304         [chromium] Accept four parameters when dispatching a WebInputEvent::GestureTap in chromium DRT eventSender
3305         https://bugs.webkit.org/show_bug.cgi?id=85289
3306
3307         Reviewed by Eric Seidel.
3308
3309         * DumpRenderTree/chromium/EventSender.cpp:
3310         (EventSender::gestureEvent):
3311             This change will allow for testing of WK85101. Currently no tests use the extra
3312             parameters deltaX and deltaY.
3313
3314 2012-05-01  Zan Dobersek  <zandobersek@gmail.com>
3315
3316         Print out stderr output of ImageDiff if it is present
3317         https://bugs.webkit.org/show_bug.cgi?id=85285
3318
3319         Reviewed by Philippe Normand.
3320
3321         Log the stderr output of ImageDiff as a warning if present. This is
3322         especially helpful when ImageDiff crashes and stderr data might contain
3323         relevant information about what caused the crash.
3324
3325         * Scripts/webkitpy/layout_tests/port/webkit.py:
3326         (WebKitPort._read_image_diff):
3327
3328 2012-05-01  Kenneth Rohde Christiansen  <kenneth@webkit.org>
3329
3330         [Qt] Add an experimental extension to set the min. contents width
3331         https://bugs.webkit.org/show_bug.cgi?id=85281
3332
3333         Reviewed by Antonio Gomes.
3334
3335         * MiniBrowser/qt/qml/BrowserWindow.qml:
3336
3337         Set the value of preferredMinimumContentsWidth to 980.
3338
3339 2012-04-30  Ojan Vafai  <ojan@chromium.org>
3340
3341         Fix tests to not depend on the actual list of builders at build.chromium.org.
3342
3343         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
3344         (testHtmlForIndividualTestOnAllBuildersWithChrome):
3345         (testHtmlForIndividualTestOnAllBuildersWithChromeWebkitMaster):
3346
3347 2012-04-29  Sam Weinig  <sam@webkit.org>
3348
3349         Remove BlobBuilder
3350         https://bugs.webkit.org/show_bug.cgi?id=84036
3351
3352         Reviewed by Anders Carlsson.
3353
3354         * Scripts/build-webkit:
3355         * qmake/mkspecs/features/features.prf:
3356         Make exposing the WebKitBlobBuilder JS object conditional on a new
3357         ENABLE_LEGACY_WEBKIT_BLOB_BUILDER flag. Don't enable this for the Mac,
3358         but do for everyone else.
3359
3360 2012-04-30  Amruth Raj Padmanabhuni  <amruthraj@motorola.com>
3361
3362         [GTK] Add glib as a dependency to build glib-networking
3363         https://bugs.webkit.org/show_bug.cgi?id=85181
3364
3365         Reviewed by Martin Robinson.
3366
3367         * gtk/jhbuild.modules:
3368
3369 2012-04-30  Alexis Menard  <alexis.menard@openbossa.org>
3370
3371         Move myself to reviewers.
3372
3373         Reviewed by Andreas Kling.
3374
3375         * Scripts/webkitpy/common/config/committers.py:
3376
3377 2012-04-30  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
3378
3379         Unreviewed, move myself to reviewers
3380
3381         * Scripts/webkitpy/common/config/committers.py:
3382
3383 2012-04-30  Carlos Garcia Campos  <cgarcia@igalia.com>
3384
3385         Unreviewed. Fix make distcheck.
3386
3387         * TestWebKitAPI/GNUmakefile.am: Add missing headers to compilation
3388         and resources to EXTRA_DIST.
3389
3390 2012-04-30  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
3391
3392         [GTK] DRT support for layoutTestController.removeOriginAccessWhitelistEntry
3393         https://bugs.webkit.org/show_bug.cgi?id=85156
3394
3395         Reviewed by Martin Robinson.
3396
3397         Add missing implementation removeOriginAccessWhitelistEntry to
3398         GTK's LayoutTestController.
3399
3400         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3401         (LayoutTestController::removeOriginAccessWhitelistEntry):
3402
3403 2012-04-29  Maciej Stachowiak  <mjs@apple.com>
3404
3405         REGRESSION: On Lion, run-webkit-tests changes the display color profile even when not running pixel tests
3406         https://bugs.webkit.org/show_bug.cgi?id=81729
3407
3408         Reviewed by Dan Bernstein.
3409
3410         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3411         (Manager._set_up_run): Only launch LayoutTestHelper when running pixel tests.
3412
3413 2012-04-29  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
3414
3415         [GTK] DRT needs an implementation of layoutTestController.setDomainRelaxationForbiddenForURLScheme
3416         https://bugs.webkit.org/show_bug.cgi?id=85131
3417
3418         Reviewed by Martin Robinson.
3419
3420         Add missing implementation setDomainRelaxationForbiddenForURLScheme to
3421         GTK's LayoutTestController.
3422
3423         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3424         (LayoutTestController::setDomainRelaxationForbiddenForURLScheme):
3425
3426 2012-04-28  Sam Weinig  <sam@webkit.org>
3427
3428         Smooth scrolling needs a new key
3429         <rdar://problem/11331632>
3430
3431         Reviewed by Dan Bernstein.
3432
3433         * DumpRenderTree/mac/DumpRenderTree.mm:
3434         (resetDefaultsToConsistentValues):
3435         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm:
3436         (TestWebKitAPI::InjectedBundleController::platformInitialize):
3437         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
3438         (WTR::InjectedBundle::platformInitialize):
3439         Update for new key.
3440
3441 2012-04-28  Zan Dobersek  <zandobersek@gmail.com>
3442
3443         [WK2] http/tests/navigation/anchor-frames-gbk.html fails
3444         https://bugs.webkit.org/show_bug.cgi?id=76896
3445
3446         Reviewed by Darin Adler.
3447
3448         Only dump as text if currently dumping render tree. This ensures that
3449         calling dumpAsText in a test with dumpChildFramesAsText already being
3450         called doesn't override the first decision, possibly making tests fail.
3451
3452         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3453         (WTR::LayoutTestController::dumpAsText):
3454
3455 2012-04-28  Dan Winship  <danw@gnome.org>
3456
3457         Unreviewed; Add me to the SoupNetwork watchlist
3458
3459         * Scripts/webkitpy/common/config/watchlist:
3460
3461 2012-04-27  Gustavo Noronha Silva  <gns@gnome.org>
3462
3463         [GTK] jhbuild cloning is not following WEBKITOUTPUTDIR.
3464         https://bugs.webkit.org/show_bug.cgi?id=76161
3465
3466         Reviewed by Martin Robinson.
3467
3468         * Scripts/webkitdirs.pm:
3469         (getJhbuildPath): New method to obtain the jhbuild base directory,
3470         using the product base directory
3471         (jhbuildConfigurationChanged): Use the new method
3472         (buildAutotoolsProject): Ditto.
3473         * efl/jhbuildrc: Use WEBKITOUTPUTDIR when calculating the path.
3474         * gtk/jhbuildrc: Ditto.
3475         * jhbuild/jhbuild-wrapper: Ditto.
3476
3477 2012-04-27  Scott Graham  <scottmg@chromium.org>
3478
3479         [chromium] make case of npTestNetscapePlugIn.dll match case in copy rule
3480         https://bugs.webkit.org/show_bug.cgi?id=85083
3481
3482         Reviewed by Dirk Pranke.
3483
3484         Make output product_name for npTestNetscapePlugIn.dll match the case
3485         of the copy_TestNetscapePlugIn rule. This is required for ninja, which
3486         is more particular about case matching in rules than the VS build.
3487
3488         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
3489
3490 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3491
3492         [chromium] use "drt-style" output, not "test-shell-style" output, on mac and linux DRT
3493         https://bugs.webkit.org/show_bug.cgi?id=84917
3494
3495         Unreviewed, build fix.
3496
3497         Reland the change in r115453 with a fix for chromium win.
3498         The logic in ChromiumDriver was busted on windows, causing DRT
3499         to get launched without the --test-shell flag.
3500
3501         * Scripts/webkitpy/layout_tests/port/chromium.py:
3502         (ChromiumDriver):
3503         (ChromiumDriver.__init__):
3504         (ChromiumDriver._wrapper_options):
3505         (ChromiumDriver.cmd_line):
3506         (ChromiumDriver._start):
3507         (ChromiumDriver.has_crashed):
3508         (ChromiumDriver.run_test):
3509         (ChromiumDriver.stop):
3510         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3511         (ChromiumDriverTest.setUp):
3512         (ChromiumDriverTest.test_stop):
3513         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
3514         (ChromiumDriverTest.test_two_drivers):
3515
3516 2012-04-27  Peter Beverloo  <peter@chromium.org>
3517
3518         Add John Grabowski as a non-committer to committers.py
3519         https://bugs.webkit.org/show_bug.cgi?id=85080
3520
3521         Reviewed by Dirk Pranke.
3522
3523         He's being CC'ed often enough because of his involvement in the Chromium
3524         side of Chrome for Android upstreaming.
3525
3526         * Scripts/webkitpy/common/config/committers.py:
3527           add jrg@chromium.org
3528
3529 2012-04-27  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
3530
3531         [Qt] Fix minimal build.
3532         https://bugs.webkit.org/show_bug.cgi?id=85045
3533
3534         Reviewed by Tor Arne Vestbø.
3535
3536         Do not set conflicting values of ENABLE_XSLT.
3537
3538         * qmake/mkspecs/features/features.prf:
3539
3540 2012-04-27  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
3541
3542         [EFL][DRT] EFL should sanitize pulseaudio module configuration too
3543         https://bugs.webkit.org/show_bug.cgi?id=85047
3544
3545         Reviewed by Dirk Pranke.
3546
3547         PulseAudio test setup sanitization shared between EFL and GTK.
3548
3549         * Scripts/webkitpy/layout_tests/port/efl.py:
3550         (EflPort):
3551         (EflPort.setup_test_run):
3552         (EflPort.clean_up_test_run):
3553         * Scripts/webkitpy/layout_tests/port/gtk.py:
3554         (GtkPort): Moving pulseaudio unload/restore out to shared class.
3555         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer.py: Added.
3556         (PulseAudioSanitizer):
3557         (PulseAudioSanitizer._unload_pulseaudio_module):
3558         (PulseAudioSanitizer._restore_pulseaudio_module):
3559
3560 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3561
3562         Unreviewed, revert r115453 - broke chromium win bots
3563
3564         * Scripts/webkitpy/layout_tests/port/chromium.py:
3565         (ChromiumPort._path_to_image_diff):
3566         (ChromiumDriver):
3567         (ChromiumDriver.__init__):
3568         (ChromiumDriver._wrapper_options):
3569         (ChromiumDriver.cmd_line):
3570         (ChromiumDriver.stop):
3571         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3572         (ChromiumDriverTest.setUp):
3573         (ChromiumDriverTest.test_stop):
3574         (ChromiumDriverTest.test_two_drivers):
3575         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
3576
3577 2012-04-27  Ryosuke Niwa  <rniwa@webkit.org>
3578
3579         Chromium Windows Perf bots timeout due to not output
3580         https://bugs.webkit.org/show_bug.cgi?id=84940
3581
3582         Reviewed by Dirk Pranke.
3583
3584         Dirk and I investigated the issue on the bot but we couldn't figure out what is going wrong.
3585         Since we run-perf-tests don't need any of fancy feature printer provides, just use python's
3586         built-in logging module instead. Printing out to stdout and stderr seem to work so hopefully
3587         this will fix the issue on the bot.
3588
3589         * Scripts/run-perf-tests:
3590         * Scripts/webkitpy/performance_tests/perftest.py:
3591         (PerfTest.run):
3592         (PerfTest.run_failed):
3593         (PerfTest.parse_output):
3594         (PerfTest.output_statistics):
3595         (ChromiumStylePerfTest.parse_output):
3596         (PageLoadingPerfTest.run):
3597         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
3598         (MainTest.test_parse_output):
3599         (MainTest.test_parse_output_with_failing_line):
3600         (TestPageLoadingPerfTest):
3601         (TestPageLoadingPerfTest.test_run):
3602         (TestPageLoadingPerfTest.test_run_with_bad_output):
3603         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3604         (PerfTestsRunner.__init__):
3605         (PerfTestsRunner._parse_args):
3606         (PerfTestsRunner.run):
3607         (PerfTestsRunner._upload_json):
3608         (PerfTestsRunner._print_status):
3609         (PerfTestsRunner._run_tests_set):
3610         (PerfTestsRunner._run_single_test):
3611         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3612         (create_runner):
3613         (test_run_test_set):
3614         (test_run_test_set_kills_drt_per_run):
3615         (test_run_test_pause_before_testing):
3616         (test_run_test_set_for_parser_tests):
3617         (test_run_test_set_with_json_output):
3618         (test_run_test_set_with_json_source):
3619         (test_run_test_set_with_multiple_repositories):
3620         (test_upload_json):
3621         (test_parse_args):
3622
3623 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3624
3625         [chromium] use "drt-style" output, not "test-shell-style" output, on mac and linux DRT
3626         https://bugs.webkit.org/show_bug.cgi?id=84917
3627
3628         Reviewed by Ojan Vafai.
3629
3630         This change changes the chromium new-run-webkit-tests
3631         implementation to use the code in WebKitDriver and ServerProcess
3632         by default on Mac and Linux, instead of using the old-style
3633         Chromium/test_shell style of output.
3634
3635         We will still use the test_shell style on Windows for now
3636         and you can still get the old style of behavior with
3637         new-run-webkit-tests --additional-drt-flag --test-shell
3638
3639         * Scripts/webkitpy/layout_tests/port/chromium.py:
3640         (ChromiumDriver):
3641         (ChromiumDriver.__init__):
3642         (ChromiumDriver._wrapper_options):
3643         (ChromiumDriver.cmd_line):
3644         (ChromiumDriver._start):
3645         (ChromiumDriver.has_crashed):
3646         (ChromiumDriver.run_test):
3647         (ChromiumDriver.stop):
3648         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3649         (ChromiumDriverTest.setUp):
3650         (ChromiumDriverTest.test_stop):
3651         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
3652         (ChromiumDriverTest.test_two_drivers):
3653
3654 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
3655