Tools/ChangeLog

   1 2013-01-01  KwangYong Choi  <ky0.choi@samsung.com>
   2
   3         [EFL] Enable MHTML feature
   4         https://bugs.webkit.org/show_bug.cgi?id=105815
   5
   6         Reviewed by Laszlo Gombos.
   7
   8         Enable MHTML feature for EFL port by default.
   9
  10         * Scripts/webkitperl/FeatureList.pm:
  11
  12 2013-01-01  Alan Cutter  <alancutter@chromium.org>
  13
  14         sheriff-bot should know who the gardeners/sheriffs are
  15         https://bugs.webkit.org/show_bug.cgi?id=105698
  16
  17         Reviewed by Eric Seidel.
  18
  19         Added a "sheriffs" command to sheriffbot to fetch the current Chromium Webkit sheriffs.
  20
  21         * Scripts/webkitpy/tool/bot/irc_command.py:
  22         (Current):
  23         (Current._retrieve_webkit_sheriffs):
  24         (Current.execute):
  25         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
  26         (IRCCommandTest.test_current):
  27         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_0.js: Added.
  28         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_1.js: Added.
  29         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_2.js: Added.
  30         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_malformed.js: Added.
  31         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_zero.js: Added.
  32
  33 2012-12-31  Carlos Garcia Campos  <cgarcia@igalia.com>
  34
  35         [GTK] Move ImageDiff to the Tools directory root
  36         https://bugs.webkit.org/show_bug.cgi?id=105421
  37
  38         Reviewed by Kenneth Rohde Christiansen.
  39
  40         * DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Removed.
  41         * GNUmakefile.am:
  42         * ImageDiff/gtk/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/gtk/ImageDiff.cpp.
  43
  44 2012-12-30  Mike West  <mkwst@chromium.org>
  45
  46         Add myself to watchlists for CSP and console logs.
  47         https://bugs.webkit.org/show_bug.cgi?id=105874
  48
  49         Reviewed by Dirk Pranke.
  50
  51         Because I really, really, really need more email.
  52
  53         * Scripts/webkitpy/common/config/committers.py:
  54             Add mkwst+watchlist@chromium.org as a contributor.
  55         * Scripts/webkitpy/common/config/watchlist:
  56             Create watchlist categories for 1) the various mechanisms used to
  57             generate console messages, and 2) CSP usage and files.
  58
  59 2012-12-30  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
  60
  61         [EFL] Enable TEMPLATE_ELEMENT feature
  62         https://bugs.webkit.org/show_bug.cgi?id=105865
  63
  64         Reviewed by Laszlo Gombos.
  65
  66         * Scripts/webkitperl/FeatureList.pm: Enable template element feature for EFL port.
  67
  68 2012-12-30  Zan Dobersek  <zandobersek@gmail.com>
  69
  70         Unreviewed GTK gardening of the unit tests.
  71
  72         Skip the TestInspectorServer unit test that's timing out completely.
  73
  74         * Scripts/run-gtk-tests:
  75         (TestRunner):
  76
  77 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
  78
  79         Web Inspector: Native Memory Instrumentation plugin: do not generate 'not instrumented' warning for instrumented mutable members.
  80         https://bugs.webkit.org/show_bug.cgi?id=105855
  81
  82         Reviewed by Vsevolod Vlasov.
  83
  84         Extract MemberExpr from ImplicitCastExpr. It happens when we instrument a mutable member because
  85         addMember expects const T& and the mutable ref to member implicitly converts into const ref to member.
  86
  87         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
  88         (ReportMemoryUsageVisitor):
  89         (clang::AddMemberCallVisitor::VisitCallExpr):
  90         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
  91         (clang::ReportMemoryUsageVisitor::checkMembersCoverage):
  92
  93 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
  94
  95         Web Inspector: Native Memory Instrumentation plugin: move function bodies out of class declarations.
  96         https://bugs.webkit.org/show_bug.cgi?id=105852
  97
  98         Reviewed by Alexander Pavlov.
  99
 100         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
 101         (clang):
 102         (AddMemberCallVisitor):
 103         (ReportMemoryUsageVisitor):
 104         (clang::ReportMemoryUsageAction::ParseArgs):
 105         (clang::AddMemberCallVisitor::VisitCallExpr):
 106         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
 107         (clang::ReportMemoryUsageVisitor::emitWarning):
 108         (clang::ReportMemoryUsageVisitor::findInstrumentationMethod):
 109         (clang::ReportMemoryUsageVisitor::needsToBeInstrumented):
 110         (clang::ReportMemoryUsageVisitor::CheckMembersCoverage):
 111
 112 2012-12-29  Zan Dobersek  <zandobersek@gmail.com>
 113
 114         [webkitpy] Omit webkitpy/thirdparty/BeautifulSoup.py from code coverage checking
 115         https://bugs.webkit.org/show_bug.cgi?id=105835
 116
 117         Reviewed by Dirk Pranke.
 118
 119         The BeautifulSoup module is a third-party import and as such we don't have much control
 120         over its testing. Because of that it should be omitted from code coverage checking.
 121
 122         * Scripts/webkitpy/test/main.py:
 123         (Tester._run_tests):
 124
 125 2012-12-28  Martin Robinson  <mrobinson@igalia.com>
 126
 127         [GTK][WK2] Add support for IME Composition
 128         https://bugs.webkit.org/show_bug.cgi?id=65093
 129
 130         Reviewed by Carlos Garcia Campos.
 131
 132         Update unit tests for GtkInputMethodFilter to note whether or not an event
 133         was a fake event synthesized for composition purposes. This is now used
 134         for WebKit2.
 135
 136         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
 137         (TestWebKitAPI::TestInputMethodFilter::sendSimpleKeyEvent):
 138         (TestWebKitAPI::TestInputMethodFilter::sendKeyEventWithCompositionResults):
 139         (TestWebKitAPI::TEST):
 140         * gtk/generate-gtkdoc:
 141         (get_webkit2_options): Skip new files in the WebKit2 source directory for documentation
 142         generation.
 143
 144 2012-12-27  Ilya Tikhonovsky  <loislo@chromium.org>
 145
 146         Web Inspector: Native Memory Instrumentation. Update clang plugin according to the current state of memory instrumentation code.
 147         https://bugs.webkit.org/show_bug.cgi?id=105800
 148
 149         Reviewed by Yury Semikhatsky.
 150
 151         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
 152         (clang::ReportMemoryUsageConsumer::ReportMemoryUsageConsumer):
 153         (clang):
 154
 155 2012-12-27  Zan Dobersek  <zandobersek@gmail.com>
 156
 157         Create a GTK build system watchlist and add myself to it
 158         https://bugs.webkit.org/show_bug.cgi?id=105769
 159
 160         Reviewed by Dirk Pranke.
 161
 162         Add a GTK build system watchlist, watching the configure.ac, GNUmakefile.am files
 163         and the input GNUmakefile.features.am.in file. The GNUmakefile.list.am files are
 164         omitted on purpose due to relatively high change rate they are exposed to.
 165
 166         The new watchlist would be of use to me as the GTK build system recently went
 167         through a series of changes that included a changed policy on adding configuration
 168         flags to configure.ac and subsequent additions to GNUmakefile.am files. A PSA on the
 169         mailing list about this change will be published later, but I'd also like to be
 170         notified of any pending changes to these files just so I can check the changes are
 171         in compliance with the new policy.
 172
 173         * Scripts/webkitpy/common/config/watchlist:
 174
 175 2012-12-26  Ilya Tikhonovsky  <loislo@chromium.org>
 176
 177         Unreviewed. Replace find("Tools") with rfind("Tools") because base dir of WebKit may have Tools word.
 178         As example ..../DevTools/src/third_party/WebKit
 179
 180         * Scripts/webkitpy/common/webkit_finder.py:
 181         (WebKitFinder.webkit_base):
 182
 183 2012-12-25  Christophe Dumez  <christophe.dumez@intel.com>
 184
 185         [EFL][WK2] Refactor snapshot taking code
 186         https://bugs.webkit.org/show_bug.cgi?id=105687
 187
 188         Reviewed by Kenneth Rohde Christiansen.
 189
 190         Call WKViewCreateSnapshot() instead of WKViewGetSnapshot()
 191         since the function was renamed to follow naming style.
 192
 193         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
 194         (WTR::PlatformWebView::windowSnapshotImage):
 195
 196 2012-12-24  Ilya Tikhonovsky  <loislo@chromium.org>
 197
 198         Web Inspector: Native Memory Instrumentation: propagate member type as edge type to the serialized heap graph.
 199         https://bugs.webkit.org/show_bug.cgi?id=105725
 200
 201         Reviewed by Yury Semikhatsky.
 202
 203         MemberType value names were adjusted according to Style Guide.
 204         Existing tests were extended with link type validation.
 205
 206         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
 207
 208 2012-12-24  Xiaobo Wang  <xbwang@torchmobile.com.cn>
 209
 210         [BlackBerry] DRT - Crashed when running pixels tests
 211         https://bugs.webkit.org/show_bug.cgi?id=105711
 212
 213         RIM PR 270003
 214
 215         Reviewed by George Staikos.
 216
 217         Calling window->post() here will result in buffer swapping which is
 218         only allowed in UI thread now, and I found we don't need to call it
 219         for GL renderer.
 220
 221         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
 222         (createBitmapContextFromWebView):
 223
 224 2012-12-22  Dan Winship  <danw@gnome.org>
 225
 226         Add a patch to fix libxml2 2.9.0 regression
 227         https://bugs.webkit.org/show_bug.cgi?id=104494
 228
 229         Reviewed by Martin Robinson.
 230
 231         * gtk/jhbuild.modules:
 232         * gtk/patches/libxml2-2.9.0-dtd.patch: patch from upstream
 233
 234 2012-12-23  Allan Sandfeld Jensen  <allan.jensen@digia.com>
 235
 236         Unreviewed update of status to reviewer.
 237
 238         * Scripts/webkitpy/common/config/committers.py:
 239
 240 2012-12-23  Philippe Normand  <pnormand@igalia.com>
 241
 242         [GTK][jhbuild] Bump GStreamer to 1.0.4
 243         https://bugs.webkit.org/show_bug.cgi?id=105640
 244
 245         Reviewed by Martin Robinson.
 246
 247         * gtk/jhbuild.modules: Update GStreamer modules to 1.0.4.
 248
 249 2012-12-21  Nico Weber  <thakis@chromium.org>
 250
 251         chromium: webkit-build-directory doesn't handle ninja output dirs
 252         https://bugs.webkit.org/show_bug.cgi?id=105671
 253
 254         Reviewed by Daniel Bates.
 255
 256         Teach webkit-build-directory to detect the ninja build directory.
 257
 258         * Scripts/webkitdirs.pm:
 259         (determineBaseProductDir):
 260         (determineIsChromiumNinja):
 261
 262 2012-12-21  Ryosuke Niwa  <rniwa@webkit.org>
 263
 264         [Mountain Lion] platform/mac/editing/spelling/removing-underline-after-accepting-autocorrection-using-punctuation.html failing
 265         https://bugs.webkit.org/show_bug.cgi?id=95477
 266
 267         Reviewed by Sam Weinig.
 268
 269         Co-authored by Jia Pu.
 270
 271         Specify a fixed autocorrection dictionary in resetDefaultsToConsistentValues() so that
 272         we can have consistent autocorrection behavior when running tests.
 273
 274         * DumpRenderTree/mac/DumpRenderTree.mm:
 275         (resetDefaultsToConsistentValues):
 276         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
 277         (WTR::InjectedBundle::platformInitialize):
 278
 279 2012-12-21  Zan Dobersek  <zandobersek@gmail.com>
 280
 281         [GTK] Add WebKit1-specific TestExpectations file
 282         https://bugs.webkit.org/show_bug.cgi?id=97562
 283
 284         Reviewed by Dirk Pranke.
 285
 286         Implement the expectations_files method inside the GtkPort class that
 287         returns the list of the TestExpectations files the current test configuration
 288         requires. The method is implemented in a similar way as on the EFL port.
 289         A unit test is also added.
 290
 291         * Scripts/webkitpy/layout_tests/port/gtk.py:
 292         (GtkPort._search_paths):
 293         (GtkPort):
 294         (GtkPort.expectations_files):
 295         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
 296         (GtkPortTest.test_expectations_files):
 297
 298 2012-12-21  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
 299
 300         [EFL] Enable feature option for the CSS sticky position
 301         https://bugs.webkit.org/show_bug.cgi?id=105630
 302
 303         Reviewed by Kenneth Rohde Christiansen.
 304
 305         Enable CSS sticky position feature by default for the
 306         EFL port. This feature option was added in r138252.
 307
 308         No new tests. Existing layout tests pass.
 309
 310         * Scripts/webkitperl/FeatureList.pm:
 311
 312 2012-12-21  Jussi Kukkonen  <jussi.kukkonen@intel.com>
 313
 314         [WK2] REGRESSION(r138232): It made 4 tests timeout
 315         https://bugs.webkit.org/show_bug.cgi?id=105623
 316
 317         Reviewed by Kenneth Rohde Christiansen.
 318
 319         r138232 changed functionality so that empty text output
 320         no longer leads to a dump() call, which leads to a timeout.
 321         Instead do a text dump if there is text output or if there
 322         is no audio output.
 323
 324         * WebKitTestRunner/TestInvocation.cpp:
 325         (WTR::TestInvocation::dumpResults):
 326
 327 2012-12-21  Zeno Albisser  <zeno@webkit.org>
 328
 329         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
 330         https://bugs.webkit.org/show_bug.cgi?id=105309
 331
 332         On Windows the library name is Qt5WebKitWidgets as well.
 333         So r138136 should not have adjusted the library name for Windows.
 334
 335         Reviewed by Simon Hausmann.
 336
 337         * Scripts/webkitdirs.pm:
 338         (builtDylibPathForName):
 339
 340 2012-12-21  Zeno Albisser  <zeno@webkit.org>
 341
 342         [Qt] Add new TestExpectations for qt-5.0-mac-wk2.
 343         https://bugs.webkit.org/show_bug.cgi?id=105424
 344
 345         This change will allow us to run a very specific
 346         subset of the available LayoutTests on a mac buildbot.
 347         The new baseline search path for mac will look as follows:
 348         qt-5.0-mac-wk2 -> qt-5.0-wk2 -> qt-5.0 -> qt-mac -> qt
 349
 350         Reviewed by Dirk Pranke.
 351
 352         * Scripts/webkitpy/layout_tests/port/qt.py:
 353         (QtPort._search_paths):
 354         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
 355         (QtPortTest):
 356
 357 2012-12-20  Ryuan Choi  <ryuan.choi@samsung.com>
 358
 359         [EFL] Build break with latest EFL libraries.
 360         https://bugs.webkit.org/show_bug.cgi?id=104827
 361
 362         Reviewed by Laszlo Gombos.
 363
 364         The eo EFL package is introduced and evas and ecore use it since 1.8.
 365         While introducing Eo, EFL changed several structures of Evas and Ecore
 366         from own specific class to Eo.
 367
 368         So, this patch adds FindEo and changes declaration of Evas, Evas_Object,
 369         Ecore_Timer to build with latest EFL libraries.
 370
 371         * DumpRenderTree/efl/CMakeLists.txt: Includes eo EFL library.
 372         * EWebLauncher/CMakeLists.txt: Ditto.
 373         * MiniBrowser/efl/CMakeLists.txt: Ditto.
 374         * ImageDiff/PlatformEfl.cmake: Ditto.
 375         * TestWebKitAPI/PlatformEfl.cmake: Ditto.
 376         * TestWebKitAPI/PlatformWebView.h: Modified declaration of Evas_Object.
 377         * WebKitTestRunner/InjectedBundle/TestRunner.h: Modified declaration of Ecore_Timer.
 378         * WebKitTestRunner/PlatformEfl.cmake: Includes eo EFL library.
 379         * WebKitTestRunner/PlatformWebView.h: Modified declaration of Evas_Object.
 380
 381 2012-12-20  Dirk Pranke  <dpranke@chromium.org>
 382
 383         TestExpectation fallback is broken
 384         https://bugs.webkit.org/show_bug.cgi?id=105583
 385
 386         Reviewed by Ryosuke Niwa.
 387
 388         When I added support for the -wk2 in port names, it looks
 389         like I broke the way we computed the list of TestExpectations
 390         files so that the implementation-version directory was left out :(.
 391         Fixed and cleaned up the tests.
 392
 393         * Scripts/webkitpy/layout_tests/port/base.py:
 394         (Port.expectations_files):
 395         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
 396         (TestWebKitPort.__init__):
 397         (PortTestCase.test_skipped_directories_for_symbols):
 398         (test_skipped_directories_for_features):
 399         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
 400         (test_skipped_tests_for_unsupported_features_empty_test_list):
 401         (test_skipped_layout_tests):
 402         (test_expectations_files):
 403
 404 2012-12-20  Anders Carlsson  <andersca@apple.com>
 405
 406         Don't include WebKit2 headers when building TestWebKitAPI on windows
 407         https://bugs.webkit.org/show_bug.cgi?id=105582
 408
 409         Reviewed by Tim Horton.
 410
 411         * TestWebKitAPI/config.h:
 412         * TestWebKitAPI/win/main.cpp:
 413
 414 2012-12-20  Ryosuke Niwa  <rniwa@webkit.org>
 415
 416         REGRESSION(r133820?): SimplifyMarkupTest API test asserts
 417         https://bugs.webkit.org/show_bug.cgi?id=105370
 418
 419         Reviewed by Simon Fraser.
 420
 421         Re-enable the test.
 422
 423         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
 424         (TestWebKitAPI::TEST):
 425
 426 2012-12-20  Nico Weber  <thakis@chromium.org>
 427
 428         chromium nrwt: Pick the newest binary found in DEFAULT_BUILD_DIRECTORIES, not the first
 429         https://bugs.webkit.org/show_bug.cgi?id=105498
 430
 431         Reviewed by Dirk Pranke.
 432
 433         Use the newest binary available rather than an than always picking one
 434         build directory over another based on iteration order.
 435
 436         * Scripts/webkitpy/layout_tests/port/chromium.py:
 437         (ChromiumPort._static_build_path):
 438         Check for timestamps.
 439         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
 440         (ChromiumMacPortTest.test_build_path_timestamps):
 441         Test that out / xcodebuild selection happens based on timestamps
 442         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
 443         (ChromiumWinPortTest.test_build_path_timestamps):
 444         Test that out / build selection happens based on timestamps
 445
 446 2012-12-19  Simon Fraser  <simon.fraser@apple.com>
 447
 448         Lots of sticky tests failing in WK2
 449         https://bugs.webkit.org/show_bug.cgi?id=105464
 450
 451         Reviewed by Beth Dakin.
 452
 453         WebKitTestRunner had a race between snapshotting in the UI process,
 454         and resettting after the test in the web process. InjectedBundle::done()
 455         was a bad place to call page()->resetAfterTest(), because of this race;
 456         it could reset the scroll position before the UI snapshot had been obtained.
 457
 458         Fix by moving the call to page()->resetAfterTest() into didReceiveMessage(),
 459         for the "Reset" message which will come in before the next test.
 460
 461         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
 462         (WTR::InjectedBundle::didReceiveMessage):
 463         (WTR::InjectedBundle::done):
 464
 465 2012-12-19  Filip Pizlo  <fpizlo@apple.com>
 466
 467         DFG speculation checks that take JumpList should consolidate OSRExits
 468         https://bugs.webkit.org/show_bug.cgi?id=105401
 469
 470         Reviewed by Oliver Hunt.
 471
 472         Update the profiler to understand that an OSR exit may have multiple
 473         candidate exit sites.
 474
 475         * Scripts/display-profiler-output:
 476
 477 2012-12-20  Eric Seidel  <eric@webkit.org>
 478
 479         EWSes do not need to dump build/test logs when things pass
 480         https://bugs.webkit.org/show_bug.cgi?id=105402
 481
 482         Reviewed by Dirk Pranke.
 483
 484         When we were first writing the commit-queue and EWS system
 485         and running them locally, we thought it was important to see the
 486         build/test output as it happened.  Now these queues have been running
 487         for years, and we don't use this feature anymore.  So this patch
 488         makes us stop tee-ing the output directly to stdout, and instead
 489         only print any output from commands when they fail.
 490
 491         I also added some logging when we execute each command
 492         since we are no longer printing the command output itself.
 493         This should make it much easier to see errors with the EWS
 494         system itself, now that all the passing-build output is removed.
 495
 496         This required updating a zillion unittests, since we have a bunch of
 497         "integration" style testing of the queue-bots.  Many of the unittests for the bots were
 498         already approximating this behavior by using a "logging" Executive.
 499         I disabled logging in those cases to avoid redundant logging in those tests.
 500
 501         * Scripts/webkitpy/common/system/executive.py:
 502         (Executive.command_for_printing):
 503         (Executive.run_command):
 504         * Scripts/webkitpy/common/system/executive_mock.py:
 505         (MockExecutive.command_for_printing):
 506         (MockExecutive.run_command):
 507         * Scripts/webkitpy/common/system/executive_unittest.py:
 508         (ExecutiveTest.test_auto_stringify_args):
 509         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
 510         (EarlyWarningSytemTest._default_expected_logs):
 511         (_test_ews):
 512         (_test_ewses):
 513         * Scripts/webkitpy/tool/commands/queues.py:
 514         (AbstractQueue.run_webkit_patch):
 515         * Scripts/webkitpy/tool/commands/queues_unittest.py:
 516         (AbstractQueueTest._assert_run_webkit_patch):
 517         (CommitQueueTest.test_commit_queue):
 518         (test_rollout):
 519         (test_rollout_lands):
 520         (test_manual_reject_during_processing):
 521         (StyleQueueTest.test_style_queue_with_style_exception):
 522         (test_style_queue_with_watch_list_exception):
 523
 524 2012-12-20  Dominic Mazzoni  <dmazzoni@google.com>
 525
 526         AX: support clickPoint in DRT for chromium
 527         https://bugs.webkit.org/show_bug.cgi?id=97359
 528
 529         Reviewed by Chris Fleizach.
 530
 531         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
 532         several more tests can pass.
 533
 534         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
 535         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
 536         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
 537         (WebTestRunner):
 538         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
 539         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
 540         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
 541         (AccessibilityUIElement):
 542
 543 2012-12-20  Dan Bernstein  <mitz@apple.com>
 544
 545         prepare-ChangeLog adds "Additional information of the change such as approach, rationale" where it doesn't make sense
 546         https://bugs.webkit.org/show_bug.cgi?id=99268
 547
 548         Reviewed by Alexey Proskuryakov.
 549
 550         * Scripts/prepare-ChangeLog:
 551         (generateNewChangeLogs): Removed the "Additional information of the change" line.
 552
 553 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
 554
 555         Unreviewed. Skip unit test /webkit/atk/getTextInParagraphAndBodyModerate
 556
 557         * Scripts/run-gtk-tests:
 558         (TestRunner):
 559
 560 2012-12-20  Thiago Marcos P. Santos  <thiago.santos@intel.com>
 561
 562         [Qt] Move ImageDiff to the Tools directory root
 563         https://bugs.webkit.org/show_bug.cgi?id=105420
 564
 565         Reviewed by Tor Arne Vestbø.
 566
 567         Moved ImageDiff from outside DumpRenderTree because it is also a
 568         dependency when running pixel tests for WebKit2.
 569
 570         * ImageDiff/ImageDiff.pro: Renamed from Tools/DumpRenderTree/qt/ImageDiff.pro.
 571         * ImageDiff/qt/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/qt/ImageDiff.cpp.
 572         * Tools.pro:
 573
 574 2012-12-20  Zan Dobersek  <zandobersek@gmail.com>
 575
 576         [GTK] Remove the --enable-unstable-features configuration option
 577         https://bugs.webkit.org/show_bug.cgi?id=105327
 578
 579         Reviewed by Martin Robinson.
 580
 581         Add a missing feature option for the CSS sticky position feature.
 582
 583         * Scripts/webkitperl/FeatureList.pm:
 584
 585 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
 586
 587         Unreviewed. Skip unit test WebKit2.ForceRepaint that times out
 588
 589         * Scripts/run-gtk-tests:
 590         (TestRunner):
 591
 592 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
 593
 594         [EFL] MiniBrowser does not play Infinite Gangnam Style
 595         https://bugs.webkit.org/show_bug.cgi?id=103531
 596
 597         Reviewed by Kenneth Rohde Christiansen.
 598
 599         For EFL we would like this feature to be built by default.
 600
 601         * Scripts/webkitperl/FeatureList.pm:
 602
 603 2012-12-20  Csaba Osztrogonác  <ossy@webkit.org>
 604
 605         [Qt] EWS bots should call qmake always
 606         https://bugs.webkit.org/show_bug.cgi?id=90049
 607
 608         Reviewed by Simon Hausmann.
 609
 610         * Scripts/webkitdirs.pm:
 611         (buildQMakeProjects):
 612
 613 2012-12-20  Jochen Eisinger  <jochen@chromium.org>
 614
 615         [chromium] fix stray printf in TestRunner's WebTestProxy
 616         https://bugs.webkit.org/show_bug.cgi?id=105515
 617
 618         Reviewed by Ryosuke Niwa.
 619
 620         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
 621         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
 622
 623 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
 624
 625         [WK2] WebAudio WKTR support
 626         https://bugs.webkit.org/show_bug.cgi?id=95084
 627
 628         Reviewed by Kenneth Rohde Christiansen.
 629
 630         Implementing audio data dump support for WebKitTestRunner,
 631         by adding the API to the IDL, enabling IPC for transferring
 632         the audio data as bytebuffer and dumping it according
 633         to run-webkit-tests script's expectations.
 634         This enables audio testing on WebKit2 ports.
 635         Based on initial wip-patch from Philippe Normand - thanks.
 636
 637         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Adding setAudioData API.
 638         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Assign audio buffer data to a local WKDataRef after extracting it from WK2 using WKBundleDataFromUint8Array bundle API.
 639         (WTR::TestRunner::setAudioData):
 640         (WTR):
 641         * WebKitTestRunner/InjectedBundle/TestRunner.h: Add a whatToDump state for dumping Audio.
 642         (TestRunner):
 643         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Add audio to the message that's transferred to the WKTR process.
 644         (WTR::InjectedBundle::done):
 645         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: Allow setting and audio result handle to be transferred.
 646         (WTR::InjectedBundle::setAudioResult):
 647         (InjectedBundle):
 648         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
 649         (WTR::InjectedBundlePage::dump): Dummy handling the Audio case in the switch, build fix for Mac.
 650         * WebKitTestRunner/TestInvocation.cpp: Audio dump support, dump audio if we haven't received a text dump from the web process.
 651         (WTR::TestInvocation::dumpResults):
 652         (WTR):
 653         (WTR::TestInvocation::dumpAudio):
 654         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Storing received audio data before dumping.
 655         * WebKitTestRunner/TestInvocation.h:
 656         (TestInvocation):
 657
 658 2012-12-20  Alberto Garcia  <agarcia@igalia.com>
 659
 660         [GTK] Count failed and timed out tests separately
 661         https://bugs.webkit.org/show_bug.cgi?id=105448
 662
 663         Reviewed by Carlos Garcia Campos.
 664
 665         Display the number of tests that fail, time out and are skipped
 666         separately in the buildbot page.
 667
 668         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
 669         (RunGtkAPITests.commandComplete):
 670         (RunGtkAPITests.evaluateCommand):
 671         (RunGtkAPITests.getText2):
 672         * Scripts/run-gtk-tests:
 673         (TestRunner.run_tests):
 674
 675 2012-12-19  Ryosuke Niwa  <rniwa@webkit.org>
 676
 677         PerfTest.parse_output does too much
 678         https://bugs.webkit.org/show_bug.cgi?id=105391
 679
 680         Reviewed by Tony Chang.
 681
 682         Moved the code to filter lines into PerfTest._filter_output, which has been renamed from _filter_stderr.
 683         Also moved the code to output test results into PerfTest._run_with_driver.
 684
 685         * Scripts/webkitpy/performance_tests/perftest.py:
 686         (PerfTest.__init__): Added _description.
 687         (PerfTest.description): Added.
 688         (PerfTest._run_with_driver): Moved the code to output test description and test results.
 689         (PerfTest._filter_output): Moved and renamed from PerfTest._filter_stderr.
 690         (PerfTest.parse_output): Removed the code to output test results. Return test results and description
 691         respectively. Also removed the code that allowed some tests to omit values since all tests report each
 692         iteration now since r136492.
 693         (PerfTest.output_statistics): Removed the code to print test description, now done in _run_with_driver.
 694         (ChromiumStylePerfTest._run_with_driver): Added. Chromium style tests are sufficiently different from
 695         regular PerfTest that it doesn't make much sense to share _run_with_driver. But really, we should just
 696         get rid of this type of test altogether in favor of regular performance tests that uses runner.js.
 697         (ChromiumStylePerfTest.parse_and_log_output): Renamed from parse_output.
 698         (PageLoadingPerfTest._run_with_driver): Removed the explicit '' for the test description.
 699         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
 700         (MainTest.test_parse_output): Removed the expected logs since parse_output no longer prints out results.
 701         Also added a call to _filter_output since parse_output doesn't filter the output text anymore.
 702         (MainTest.test_parse_output_with_failing_line): Added a call to _filter_output. Also added ',' after 'Time:'
 703         so that it's not string-concatenated with the next line.
 704         (MainTest.test_parse_output_with_description): Added; a test for PerfTest.description().
 705         (MainTest.test_parse_output_with_subtests): Removed the expected logs, and added a call to _filter_output.
 706
 707 2012-12-19  Alexis Menard  <alexis@webkit.org>
 708
 709         Implement CSS parsing for CSS transitions unprefixed.
 710         https://bugs.webkit.org/show_bug.cgi?id=104804
 711
 712         Reviewed by Dean Jackson.
 713
 714         Add a new flag ENABLE_CSS_TRANSFORMS_ANIMATIONS_TRANSITIONS_UNPREFIXED
 715         to cover the work of unprefixing Transforms, Animations and
 716         Transitions. It will let the possibility of each ports to turn it off
 717         in their release branches until we're confident that these CSS
 718         properties are ready to be unprefixed.
 719
 720         * Scripts/webkitperl/FeatureList.pm:
 721         * qmake/mkspecs/features/features.pri:
 722
 723 2012-12-19  Julien Chaffraix  <jchaffraix@webkit.org>
 724
 725         2 clampTo unit tests are failing
 726         https://bugs.webkit.org/show_bug.cgi?id=105253
 727
 728         Reviewed by Emil A Eklund.
 729
 730         Removed clampToUnsigned as it was actually testing exactly what
 731         clampToUnsignedLong was doing. Also made clampToIntFloat more robust
 732         to precision loss (fix suggested by Emil).
 733
 734         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
 735
 736 2012-12-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
 737
 738         [EFL] Move ImageDiff to the Tools directory root
 739         https://bugs.webkit.org/show_bug.cgi?id=105305
 740
 741         Reviewed by Laszlo Gombos.
 742
 743         We previously moved this directory from the DRT directory to
 744         efl/ImageDiff. The major reason for doing so was because ImageDiff does
 745         not depend on DumpRenderTree but for legacy reasons was living inside the
 746         DRT folder, even though it is needed when running pixel tests for WK2.
 747
 748         Now we are moving to Tool directory root as other ports agreed to do the same.
 749
 750         * CMakeLists.txt:
 751         * ImageDiff/CMakeLists.txt: Added.
 752         * ImageDiff/PlatformEfl.cmake: Added.
 753         * ImageDiff/efl/ImageDiff.cpp: Renamed from Tools/efl/ImageDiff/ImageDiff.cpp.
 754         * efl/ImageDiff/CMakeLists.txt: Removed.
 755
 756 2012-12-19  Zeno Albisser  <zeno@webkit.org>
 757
 758         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
 759         https://bugs.webkit.org/show_bug.cgi?id=105309
 760
 761         The version number is only used as a part
 762         of the library name on Linux.
 763
 764         Reviewed by Kenneth Rohde Christiansen.
 765
 766         * Scripts/webkitdirs.pm:
 767         (builtDylibPathForName):
 768
 769 2012-12-19  Alberto Garcia  <agarcia@igalia.com>
 770
 771         [GTK] Tests that time out are not considered as failures.
 772         https://bugs.webkit.org/show_bug.cgi?id=105318
 773
 774         Reviewed by Martin Robinson.
 775
 776         Fix parsing of the GTK API tests log files.
 777
 778         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
 779         (RunGtkAPITests.commandComplete):
 780
 781 2012-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
 782
 783         Unreviewed, rolling out r138061.
 784         https://bugs.webkit.org/show_bug.cgi?id=105396
 785
 786         Original patch:
 787         http://trac.webkit.org/changeset/138061
 788         https://bugs.webkit.org/show_bug.cgi?id=97359
 789
 790         It is crashing on Debug bots
 791
 792         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
 793         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
 794         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
 795         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
 796         (AccessibilityUIElement):
 797
 798 2012-12-18  Alice Liu  <alice.liu@apple.com>
 799
 800         Test for https://bugs.webkit.org/show_bug.cgi?id=105106
 801         Add SPI to WebKit1 WebFrame for hit testing
 802
 803         Reviewed by Dan Bernstein.
 804
 805         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/ElementAtPointInWebFrame.mm
 806         * TestWebKitAPI/Tests/mac/ElementAtPointInWebFrame.mm: Added.
 807         (TestWebKitAPI::TEST): Loads html with two divs positioned in the 2nd and 4th quadrants of the webview.
 808         Then hit-tests at three points, expecting to hit the two divs and body element.
 809
 810 2012-12-18  Alice Liu  <alice.liu@apple.com>
 811
 812         Test for https://bugs.webkit.org/show_bug.cgi?id=105262
 813         Add SPI to WebKit1 WebFrame for node conversion to JSValueRef
 814
 815         Reviewed by Anders Carlsson.
 816
 817         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/JSWrapperForNodeInWebFrame.mm.mm
 818         * TestWebKitAPI/Tests/mac/JSWrapperForNodeInWebFrame.mm.mm: Added.
 819         (TestWebKitAPI::TEST): Tests for the correct JS wrapper for a DOMNode, provided a WebScriptWorld.
 820         Loads html with a single div element. In an isolated world, creates a property on that node.
 821         Also, in the standard world, creates a different property on that node.  Then tests for 4 things:
 822             - Existence of the isolated property in the isolated world.
 823             - Existence of the standard property in the standard world.
 824             - Non-existence of the isolated property in the standard world.
 825             - Non-existence of the standard property in the isolated world.
 826
 827 2012-12-18  Simon Fraser  <simon.fraser@apple.com>
 828
 829         Disable the SimplifyMarkup API test, wich asserts in debug builds.
 830         Tracked by https://bugs.webkit.org/show_bug.cgi?id=105370.
 831
 832         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
 833         (TestWebKitAPI::TEST):
 834
 835 2012-12-18  Joseph Pecoraro  <pecoraro@apple.com>
 836
 837         [Mac] Add Build Phase to Check Headers for Inappropriate Macros (Platform.h macros)
 838         https://bugs.webkit.org/show_bug.cgi?id=104279
 839
 840         Reviewed by David Kilzer.
 841
 842         Add a build phase script that checks a framework's header files for
 843         Platform.h and Compiler.h macros. Also add some tests for the script.
 844
 845         * Scripts/check-for-inappropriate-macros-in-external-headers: Added.
 846         Script to be used in build phases to check for inappropriate macros in headers.
 847
 848         * Scripts/test-webkitruby: Added.
 849         Script that can be used to run all webkitruby tests.
 850
 851         * Scripts/test-webkit-scripts:
 852         Also run the ruby tests.
 853
 854         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/fake-data-failing-expected.txt: Added.
 855         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/pass-expected.txt: Added.
 856         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Fail.h: Added.
 857         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Pass.h: Added.
 858         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/run-test.rb: Added.
 859         Tests for the new script.
 860
 861 2012-12-18  Dominic Mazzoni  <dmazzoni@google.com>
 862
 863         AX: support clickPoint in DRT for chromium
 864         https://bugs.webkit.org/show_bug.cgi?id=97359
 865
 866         Reviewed by Chris Fleizach.
 867
 868         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
 869         several more tests can pass.
 870
 871         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
 872         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
 873         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
 874         (WebTestRunner):
 875         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
 876         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
 877         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
 878         (AccessibilityUIElement):
 879
 880 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
 881
 882         Build fix after r138045.
 883
 884         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
 885         (PerfTestsRunner._collect_tests): Use relative_path. We skip only when this path
 886         is not explicitly specified. The condition was backwards but never matched so we
 887         always ran the test regardless of whether it's listed in the skipped list or not.
 888         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
 889         (test_collect_tests_with_ignored_skipped_list): Renamed so that it actually runs.
 890
 891 2012-12-18  Emil A Eklund  <eae@chromium.org>
 892
 893         TestWebKitAPI/Tests/WebCore/LayoutUnit fails on mac
 894         https://bugs.webkit.org/show_bug.cgi?id=105332
 895
 896         Reviewed by Levi Weintraub.
 897
 898         Create new test suite for LayoutUnit tests as the KURL tests are
 899         of the incompatible TEST_F type.
 900
 901         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
 902         (TestWebKitAPI):
 903         (TestWebKitAPI::TEST_F):
 904
 905 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
 906
 907         Running a skipped test with run-perf-tests could alert the user
 908         https://bugs.webkit.org/show_bug.cgi?id=105219
 909
 910         Reviewed by Eric Seidel.
 911
 912         Force run skipped performance tests explicitly specified in the command line.
 913         Also warn the user when the specified paths don't exist.
 914
 915         * Scripts/webkitpy/layout_tests/port/base.py:
 916         (Port.relative_test_filename):
 917         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
 918         (PerfTestsRunner._parse_args):
 919         (PerfTestsRunner._collect_tests):
 920         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
 921         (test_collect_tests_with_skipped_list_and_files):
 922
 923 2012-12-18  Yaron Friedman  <yfriedman@chromium.org>
 924
 925         Chrome Android java output is too verbose
 926         https://bugs.webkit.org/show_bug.cgi?id=105247
 927
 928         Reviewed by Tony Chang.
 929
 930         Pass -quiet to ant to suppress extraneous output.
 931
 932         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
 933         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
 934
 935 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
 936
 937         Consider removing --pause-before-testing option
 938         https://bugs.webkit.org/show_bug.cgi?id=105248
 939
 940         Reviewed by Eric Seidel.
 941
 942         Remove --pause-before-testing option.
 943
 944         * Tools/Scripts/webkitpy/layout_tests/port/driver.py:
 945         (DriverProxy.start): Removed.
 946         * Scripts/webkitpy/performance_tests/perftest.py:
 947         (PerfTest.__init__):
 948         (PerfTest.test_path): Renamed from path_or_url.
 949         (PerfTest.run):
 950         (PerfTest): Removed "driver" argument.
 951         (PerfTest._run_with_driver):
 952         (PerfTest.run_single):
 953         (ChromiumStylePerfTest.__init__):
 954         (PageLoadingPerfTest.__init__):
 955         (PageLoadingPerfTest.run_single):
 956         (PageLoadingPerfTest._run_with_driver): Renamed from run so that PerfTest.run can create a driver for us.
 957         (ReplayPerfTest.__init__):
 958         (ReplayPerfTest.prepare):
 959         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
 960         (TestPageLoadingPerfTest.test_run):
 961         (TestPageLoadingPerfTest.test_run_with_memory_output):
 962         (TestPageLoadingPerfTest.test_run_with_bad_output):
 963         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
 964         (PerfTestsRunner._parse_args):
 965         (PerfTestsRunner._run_tests_set): Removed the code to create a driver and pause before running the test.
 966         (PerfTestsRunner._run_single_test): Removed "driver" argument.
 967         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
 968         (run_test):
 969         (test_run_test_pause_before_testing): Removed.
 970
 971 2012-12-18  Alberto Garcia  <agarcia@igalia.com>
 972
 973         [GTK] Tests that time out are not considered as failures.
 974         https://bugs.webkit.org/show_bug.cgi?id=105318
 975
 976         Reviewed by Martin Robinson.
 977
 978         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
 979         (RunGtkAPITests.commandComplete):
 980         Consider log lines that end with TIMEOUT as errors.
 981
 982         * Scripts/run-gtk-tests:
 983         (TestRunner.run_tests):
 984         Consider timeouts as failures when returning the number of failed
 985         tests.
 986
 987 2012-12-18  Pravin D  <pravin.d@samsung.com>
 988
 989         Unreviewed, adding secondary email.
 990
 991         * Scripts/webkitpy/common/config/committers.py:
 992
 993 2012-12-18  Pravin D  <pravin.d@samsung.com>
 994
 995         Unreviewed, promoting myself to committer's list.
 996
 997         * Scripts/webkitpy/common/config/committers.py:
 998
 999 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1000
1001         [EFL] ImageDiff does not build on WebKit2 only builds
1002         https://bugs.webkit.org/show_bug.cgi?id=105288
1003
1004         Reviewed by Kenneth Rohde Christiansen.
1005
1006         Moved ImageDiff to the efl/ folder inside Tools/ (so we don't pollute
1007         the folder with things which are EFL-specific.
1008
1009         ImageDiff will be built by default now regardless of WebKit API flavor
1010         you choose.
1011
1012         * CMakeLists.txt:
1013         * DumpRenderTree/efl/CMakeLists.txt:
1014         * efl/ImageDiff/CMakeLists.txt: Added.
1015         * efl/ImageDiff/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/efl/ImageDiff.cpp.
1016
1017 2012-12-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1018
1019         [EFL][WK2] When creating new windows, use the opener's context
1020         https://bugs.webkit.org/show_bug.cgi?id=105283
1021
1022         Reviewed by Laszlo Gombos.
1023
1024         Use the context of the opener if exists.
1025
1026         * MiniBrowser/efl/main.c:
1027         (on_key_down):
1028         (on_window_create):
1029         (window_create):
1030         (elm_main):
1031
1032 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1033
1034         [EFL][WK2] Gardening of TestWebKitAPI tests
1035         https://bugs.webkit.org/show_bug.cgi?id=105268
1036
1037         Reviewed by Kenneth Rohde Christiansen.
1038
1039         Some tests were marked as failing but are passing and the newly added
1040         LayoutUnit should have it's own binary because it conflicts with the
1041         fixture used by KURL and is failing on the bots because of that.
1042
1043         * TestWebKitAPI/CMakeLists.txt:
1044         * TestWebKitAPI/PlatformEfl.cmake:
1045
1046 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
1047
1048         [chromium] move webintent related methods to TestRunner library
1049         https://bugs.webkit.org/show_bug.cgi?id=105164
1050
1051         Reviewed by Darin Fisher.
1052
1053         Also deleted the unused deliveredIntentFailure and
1054         deliveredIntentResult methods.
1055
1056         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1057         (DRTTestRunner::DRTTestRunner):
1058         * DumpRenderTree/chromium/DRTTestRunner.h:
1059         (DRTTestRunner):
1060         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
1061         (WebKit):
1062         (WebTestRunner::WebTestDelegate::setCurrentWebIntentRequest):
1063         (WebTestRunner::WebTestDelegate::currentWebIntentRequest):
1064         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
1065         (WebKit):
1066         (WebTestProxyBase):
1067         (WebTestRunner::WebTestProxy::registerIntentService):
1068         (WebTestRunner::WebTestProxy::dispatchIntent):
1069         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1070         (WebTestRunner::TestRunner::TestRunner):
1071         (WebTestRunner::TestRunner::sendWebIntentResponse):
1072         (WebTestRunner):
1073         (WebTestRunner::TestRunner::deliverWebIntent):
1074         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1075         (TestRunner):
1076         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
1077         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
1078         (WebTestRunner::WebTestProxyBase::registerIntentService):
1079         (WebTestRunner):
1080         (WebTestRunner::WebTestProxyBase::dispatchIntent):
1081         * DumpRenderTree/chromium/WebViewHost.cpp:
1082         (WebViewHost::setCurrentWebIntentRequest):
1083         (WebViewHost::currentWebIntentRequest):
1084         * DumpRenderTree/chromium/WebViewHost.h:
1085         (WebKit):
1086         (WebViewHost):
1087
1088 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
1089
1090         [chromium] move editing callback dumping to the WebTestProxy
1091         https://bugs.webkit.org/show_bug.cgi?id=105013
1092
1093         Reviewed by Darin Fisher.
1094
1095         To decide whether or not to dump the editing callbacks, the WebTestProxy
1096         needs to query the TestRunner. For this purpose, I introduce a
1097         WebTestRunner class that just exposes the shouldDumpEditingCallbacks
1098         method. In contrast to e.g. WebEventSender, the WebTestRunner is just
1099         an interface, and it is not owned by the WebTestInterfaces, but the
1100         embedder of the TestRunner library must provide it. This will change
1101         once all TestRunner methods are in the library.
1102
1103         * DumpRenderTree/DumpRenderTree.gypi:
1104         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1105         (DRTTestRunner::DRTTestRunner):
1106         (DRTTestRunner::reset):
1107         * DumpRenderTree/chromium/DRTTestRunner.h:
1108         (DRTTestRunner):
1109         (DRTTestRunner::taskList):
1110         (DRTTestRunner::WorkQueue::taskList):
1111         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
1112         (WebTestRunner):
1113         (WebTestInterfaces):
1114         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
1115         (WebKit):
1116         (WebTestRunner):
1117         (WebTestProxyBase):
1118         (WebTestRunner::WebTestProxy::shouldBeginEditing):
1119         (WebTestRunner::WebTestProxy::shouldEndEditing):
1120         (WebTestRunner::WebTestProxy::shouldInsertNode):
1121         (WebTestRunner::WebTestProxy::shouldInsertText):
1122         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
1123         (WebTestRunner::WebTestProxy::shouldDeleteRange):
1124         (WebTestRunner::WebTestProxy::shouldApplyStyle):
1125         (WebTestRunner::WebTestProxy::didBeginEditing):
1126         (WebTestRunner::WebTestProxy::didChangeSelection):
1127         (WebTestRunner::WebTestProxy::didChangeContents):
1128         (WebTestRunner::WebTestProxy::didEndEditing):
1129         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h: Copied from Tools/DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h.
1130         (WebTestRunner):
1131         (WebTestRunner::WebTestRunner::shouldDumpEditingCallbacks):
1132         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1133         (WebTestRunner::TestRunner::TestRunner):
1134         (WebTestRunner::TestRunner::~TestRunner):
1135         (WebTestRunner):
1136         (WebTestRunner::TestRunner::reset):
1137         (WebTestRunner::TestRunner::shouldDumpEditingCallbacks):
1138         (WebTestRunner::TestRunner::dumpEditingCallbacks):
1139         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1140         (TestRunner):
1141         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
1142         (WebTestRunner::WebTestInterfaces::Internal::testRunner):
1143         (WebTestRunner::WebTestInterfaces::Internal::setTestRunner):
1144         (WebTestInterfaces::Internal):
1145         (WebTestRunner::WebTestInterfaces::Internal::Internal):
1146         (WebTestRunner::WebTestInterfaces::testRunner):
1147         (WebTestRunner):
1148         (WebTestRunner::WebTestInterfaces::setTestRunner):
1149         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
1150         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
1151         (WebTestRunner):
1152         (WebTestRunner::WebTestProxyBase::shouldBeginEditing):
1153         (WebTestRunner::WebTestProxyBase::shouldEndEditing):
1154         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
1155         (WebTestRunner::WebTestProxyBase::shouldInsertText):
1156         (WebTestRunner::WebTestProxyBase::shouldChangeSelectedRange):
1157         (WebTestRunner::WebTestProxyBase::shouldDeleteRange):
1158         (WebTestRunner::WebTestProxyBase::shouldApplyStyle):
1159         (WebTestRunner::WebTestProxyBase::didBeginEditing):
1160         (WebTestRunner::WebTestProxyBase::didChangeSelection):
1161         (WebTestRunner::WebTestProxyBase::didChangeContents):
1162         (WebTestRunner::WebTestProxyBase::didEndEditing):
1163         * DumpRenderTree/chromium/TestShell.cpp:
1164         (TestShell::initialize):
1165         (TestShell::createNewWindow):
1166         * DumpRenderTree/chromium/WebViewHost.cpp:
1167         (WebViewHost::shouldBeginEditing):
1168         (WebViewHost::shouldEndEditing):
1169         (WebViewHost::shouldInsertNode):
1170         (WebViewHost::shouldInsertText):
1171         (WebViewHost::shouldChangeSelectedRange):
1172         (WebViewHost::shouldDeleteRange):
1173         (WebViewHost::shouldApplyStyle):
1174         * DumpRenderTree/chromium/WebViewHost.h:
1175         (WebViewHost):
1176
1177 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
1178
1179         [chromium] Fix methods in TestRunner that directly modify the settings instead of going through WebPreferences
1180         https://bugs.webkit.org/show_bug.cgi?id=105154
1181
1182         Reviewed by Tony Chang.
1183
1184         If the settings are modified directly, the embedder can't keep track of
1185         them and might accidentially override the changed settings.
1186
1187         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1188         (WebPreferences):
1189         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1190         (WebTestRunner::TestRunner::TestRunner):
1191         (WebTestRunner::TestRunner::setAsynchronousSpellCheckingEnabled):
1192         (WebTestRunner):
1193         (WebTestRunner::TestRunner::setMinimumTimerInterval):
1194         (WebTestRunner::TestRunner::setTouchDragDropEnabled):
1195         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1196         (TestRunner):
1197         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1198         (WebTestRunner::WebPreferences::reset):
1199         (WebTestRunner::WebPreferences::applyTo):
1200
1201 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
1202
1203         Unreviewed unit test skipping.
1204
1205         Filed https://bugs.webkit.org/show_bug.cgi?id=105253 to re-enable them.
1206
1207         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
1208         (TestWebKitAPI):
1209         (TestWebKitAPI::TEST):
1210
1211 2012-12-17  Dean Jackson  <dino@apple.com>
1212
1213         Unreviewed. Add Silvia Pfeiffer to contributor list.
1214
1215         * Scripts/webkitpy/common/config/committers.py:
1216
1217 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1218
1219         build.webkit.org - add more google windows bots
1220         https://bugs.webkit.org/show_bug.cgi?id=105225
1221
1222         Reviewed by Eric Seidel.
1223
1224         Adds two more windows bots (google-windows-3, a builder, and
1225         google-windows-4, a tester) for the Chromium configs.
1226
1227         * BuildSlaveSupport/build.webkit.org-config/config.json:
1228
1229 2012-12-17  Eric Seidel  <eric@webkit.org>
1230
1231         Cleanup --profiler=perf code a little and allow up to 1000 report files (instead of 100)
1232         https://bugs.webkit.org/show_bug.cgi?id=105217
1233
1234         Reviewed by Dirk Pranke.
1235
1236         Previously --profile would fail strangely if find_unused_filename returned None
1237         (which would happen as soon as we had 100 report files).
1238         Now we'll search through 1000 filenames before giving up (and assert if we failed to find one).
1239         I also made some minor cleanups to the output of perf, and tried to make the
1240         commands more self-documenting by using long-names for args.
1241
1242         * Scripts/webkitpy/common/system/profiler.py:
1243         (SingleFileOutputProfiler.__init__):
1244         (Perf.attach_to_pid):
1245         (Perf.profile_after_exit):
1246
1247 2012-12-17  Sheriff Bot  <webkit.review.bot@gmail.com>
1248
1249         Unreviewed, rolling out r137198.
1250         http://trac.webkit.org/changeset/137198
1251         https://bugs.webkit.org/show_bug.cgi?id=105212
1252
1253         This patch is causing API behavior compatibility problems
1254         (Requested by zdobersek on #webkit).
1255
1256         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1257         (willSendRequestCallback):
1258
1259 2012-12-17  Elliott Sprehn  <esprehn@chromium.org>
1260
1261         Unreviewed. Adding myself to committers list.
1262
1263         * Scripts/webkitpy/common/config/committers.py:
1264
1265 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1266
1267         nrwt: shuffle code around for cleanup in run_webkit_tests.py
1268         https://bugs.webkit.org/show_bug.cgi?id=105078
1269
1270         Reviewed by Ojan Vafai.
1271
1272         This patch reorders functions so that run_webkit_tests can be
1273         understood in a top-down matter better, and to make things slightly
1274         more sensible when I merge manager.py into it.
1275
1276         Also, this adds tests for the actual main() routine, which had several
1277         bugs get through testing a week or two ago.
1278
1279         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1280         (main):
1281         (parse_args):
1282         (_set_up_derived_options):
1283         (run):
1284         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1285         (RunTest):
1286         (RunTest.test_no_http_tests):
1287         (PortTest.disabled_test_mac_lion):
1288         (MainTest):
1289         (MainTest.test_exception_handling):
1290         (MainTest.test_exception_handling.interrupting_run):
1291         (MainTest.test_exception_handling.successful_run):
1292         (MainTest.test_exception_handling.successful_run.FakeRunDetails):
1293         (MainTest.test_exception_handling.exception_raising_run):
1294
1295 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1296
1297         webkitpy: move --lint-test-files code into its own module
1298         https://bugs.webkit.org/show_bug.cgi?id=105077
1299
1300         Reviewed by Eric Seidel.
1301
1302         It really didn't need to be in run_webkit_tests.py anymore.
1303         This patch adds a new lint-test-expectations scripts to Tools/Scripts;
1304         rwt --lint-test-files is still supported for backwards compatibility
1305         for now.
1306
1307         * Scripts/lint-test-expectations: Added.
1308         * Scripts/webkitpy/layout_tests/lint_test_expectations.py: Added.
1309         (lint):
1310         (main):
1311         * Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py: Added.
1312         (FakePort):
1313         (FakePort.__init__):
1314         (FakePort.test_configuration):
1315         (FakePort.expectations_dict):
1316         (FakePort.skipped_layout_tests):
1317         (FakePort.all_test_configurations):
1318         (FakePort.configuration_specifier_macros):
1319         (FakePort.get_option):
1320         (FakeFactory):
1321         (FakeFactory.__init__):
1322         (FakeFactory.get):
1323         (FakeFactory.all_port_names):
1324         (LintTest):
1325         (LintTest.test_all_configurations):
1326         (LintTest.test_lint_test_files):
1327         (LintTest.test_lint_test_files__errors):
1328         (MainTest):
1329         (MainTest.test_success):
1330         (MainTest.test_success.interrupting_lint):
1331         (MainTest.test_success.successful_lint):
1332         (MainTest.test_success.exception_raising_lint):
1333         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1334         (main):
1335         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1336         (StreamTestingMixin.assertNotEmpty):
1337
1338 2012-12-17  Adam Klein  <adamk@chromium.org>
1339
1340         build-webkit: rename --template-tag to --template-element to match ENABLE #define name
1341         https://bugs.webkit.org/show_bug.cgi?id=105072
1342
1343         Reviewed by Laszlo Gombos.
1344
1345         * Scripts/webkitperl/FeatureList.pm:
1346
1347 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
1348
1349         Add some unit testing for WTF::clampTo* functions
1350         https://bugs.webkit.org/show_bug.cgi?id=105060
1351
1352         Reviewed by Emil A Eklund.
1353
1354         The new tests are targeted at clamping to int / unsigned and float as those are
1355         the one explicitely exposed through more dedicated functions (like clampToInteger,
1356         clampToPositiveInteger and clampToFloat).
1357
1358         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
1359
1360 2012-12-17  Eric Seidel  <eric@webkit.org>
1361
1362         rpt --profile --chromium-android throws exception
1363         https://bugs.webkit.org/show_bug.cgi?id=105061
1364
1365         Reviewed by Daniel Bates.
1366
1367         Somehow popen(args) got turned into popen(*args) at some point,
1368         which caused my new stringify code to fail when args was passed by name.
1369         Fixed and tested.
1370
1371         * Scripts/webkitpy/common/system/executive.py:
1372         (Executive._stringify_args):
1373         (Executive):
1374         (Executive.popen):
1375         * Scripts/webkitpy/common/system/executive_unittest.py:
1376         (ExecutiveTest.test_popen_args):
1377
1378 2012-12-17  Mark Pilgrim  <pilgrim@chromium.org>
1379
1380         Use Platform::current() instead of webKitPlatformSupport() in DumpRenderTree
1381         https://bugs.webkit.org/show_bug.cgi?id=105054
1382
1383         Reviewed by Jochen Eisinger.
1384
1385         Part of a refactoring series; see tracking bug 82948.
1386
1387         * DumpRenderTree/chromium/NotificationPresenter.cpp:
1388         (NotificationPresenter::show):
1389         * DumpRenderTree/chromium/Task.cpp:
1390         (postTask):
1391         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
1392
1393 2012-12-17  Emil A Eklund  <eae@chromium.org>
1394
1395         Clamp values in LayoutUnit::operator/ when SATURATED_LAYOUT_ARITHMETIC is enabled
1396         https://bugs.webkit.org/show_bug.cgi?id=104955
1397
1398         Reviewed by Julien Chaffraix.
1399
1400         Add tests for LayoutUnit.
1401
1402         * TestWebKitAPI/CMakeLists.txt:
1403         * TestWebKitAPI/ForwardingHeaders: Added.
1404         * TestWebKitAPI/ForwardingHeaders/WebCore: Added.
1405         * TestWebKitAPI/ForwardingHeaders/WebCore/LayoutUnit.h: Added.
1406         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
1407         * TestWebKitAPI/TestWebKitAPI.gypi:
1408         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1409         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp: Added.
1410         (TestWebKitAPI):
1411         (TestWebKitAPI::TEST):
1412         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
1413
1414 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
1415
1416         [chromium] removed unused preferences from testRunner.overridePreference
1417         https://bugs.webkit.org/show_bug.cgi?id=105169
1418
1419         Reviewed by Tony Chang.
1420
1421         Most notably this removes the code related to font family maps which is
1422         now handled via window.internals.settings
1423
1424         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1425         (WebPreferences):
1426         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1427         (WebTestRunner::TestRunner::overridePreference):
1428         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1429         (TestRunner):
1430         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1431         (WebTestRunner::WebPreferences::applyTo):
1432
1433 2012-12-17  James Robinson  <jamesr@chromium.org>
1434
1435         [chromium] Move creation of WebLayerTreeView to WebWidgetClient
1436         https://bugs.webkit.org/show_bug.cgi?id=105071
1437
1438         Reviewed by Adrienne Walker.
1439
1440         Update WebViewHost for the new WebWidgetClient interface.
1441
1442         * DumpRenderTree/chromium/WebViewHost.cpp:
1443         (WebViewHost::initializeLayerTreeView):
1444         (WebViewHost::layerTreeView):
1445         (WebViewHost::~WebViewHost):
1446         * DumpRenderTree/chromium/WebViewHost.h:
1447         (WebViewHost):
1448
1449 2012-12-17  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1450
1451         [EFL][WK2] window_create doesn't receive the url
1452         https://bugs.webkit.org/show_bug.cgi?id=105184
1453
1454         Reviewed by Alexis Menard.
1455
1456         * MiniBrowser/efl/main.c: Use the url supplied to window_create
1457         (on_window_create):
1458
1459 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
1460
1461         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
1462         https://bugs.webkit.org/show_bug.cgi?id=105026
1463
1464         Reviewed by Yury Semikhatsky.
1465
1466         Old SFINAE test was replaced with new one based on this article:
1467         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
1468
1469         New test which covers this problem was added.
1470
1471         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1472
1473 2012-12-17  Anthony Scian  <ascian@rim.com>
1474
1475         [Blackberry] Static code analysis warning fixes
1476         https://bugs.webkit.org/show_bug.cgi?id=105129
1477
1478         Reviewed by Rob Buis.
1479
1480         Added null checks for fopen calls, no recovery attempted; access fault prevention only.
1481
1482         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
1483         (BlackBerry::WebKit::createFile):
1484         (BlackBerry::WebKit::DumpRenderTree::runTest):
1485
1486 2012-12-17  John Mellor  <johnme@chromium.org>
1487
1488         Add Anton Vayvod and Tim Volodine as contributors in committers.py
1489         https://bugs.webkit.org/show_bug.cgi?id=105025
1490
1491         Unreviewed. Adds two contributors to committers.py.
1492         They are initially contributing to the Text Autosizing feature (master bug: 84186).
1493
1494         * Scripts/webkitpy/common/config/committers.py:
1495
1496 2012-12-17  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1497
1498         Unreviewed, add myself to committers.py.
1499
1500         * Scripts/webkitpy/common/config/committers.py:
1501
1502 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
1503
1504         [Qt] remove some unnecessary CONFIG additions
1505
1506         Reviewed by Simon Hausmann.
1507
1508         qt is already added by spec_pre.prf, warn_on and depend_includepath by
1509         default_pre.prf.
1510
1511         * qmake/mkspecs/features/default_post.prf:
1512
1513 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
1514
1515         Unreviewed, rolling out r137892.
1516         http://trac.webkit.org/changeset/137892
1517         https://bugs.webkit.org/show_bug.cgi?id=105026
1518
1519         it broke compilation on windows
1520
1521         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1522
1523 2012-12-14  Ilya Tikhonovsky  <loislo@chromium.org>
1524
1525         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
1526         https://bugs.webkit.org/show_bug.cgi?id=105026
1527
1528         Reviewed by Yury Semikhatsky.
1529
1530         Old SFINAE test was replaced with new one based on this article:
1531         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
1532
1533         New test which covers this problem was added.
1534
1535         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1536
1537 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
1538
1539         [Qt] rely on automatic output directory setup
1540
1541         Reviewed by Simon Hausmann.
1542
1543         Now being a proper qt module (by virtue of having load(qt_build_config)
1544         in .qmake.conf), webkit gets the path setup goodies for free.
1545
1546         This also fixes Makefile.api.Debug/Release trying to generate qrc_WebKit.cpp
1547         at the same time, instead of the debug and release version ending up in different
1548         directories as they should.
1549
1550         * qmake/mkspecs/features/default_pre.prf:
1551
1552 2012-12-17  Charles Wei  <charles.wei@torchmobile.com.cn>
1553
1554         [BlackBerry] Need to re-enable video track
1555         https://bugs.webkit.org/show_bug.cgi?id=105155
1556
1557         Reviewed by George Staikos.
1558
1559         Just enable Video Track for BlackBerry.
1560
1561         * Scripts/webkitperl/FeatureList.pm:
1562
1563 2012-12-17  János Badics  <jbadics@inf.u-szeged.hu>
1564
1565         Unreviewed. Added myself to committers.py
1566
1567         * Scripts/webkitpy/common/config/committers.py:
1568
1569 2012-12-17  Jonathan Liu  <net147@gmail.com>
1570
1571         Fix linking of libraries with -Wl,-whole-archive
1572         https://bugs.webkit.org/show_bug.cgi?id=104436
1573
1574         Reviewed by Simon Hausmann.
1575
1576         If there is more than one library in LIBS linked with
1577         -Wl,-whole-archive -l... -Wl,-no-whole-archive, qmake tries to
1578         merge the duplicate flags in LIBS which causes linking to fail.
1579
1580         Add no_smart_library_merge to CONFIG to prevent the duplicate
1581         flags from being removed when using -Wl,-whole-archive.
1582
1583         * qmake/mkspecs/features/functions.prf:
1584
1585 2012-12-16  Arvid Nilsson  <anilsson@rim.com>
1586
1587         [BlackBerry] Adapt to new BlackBerry::Platform::TouchPoint API
1588         https://bugs.webkit.org/show_bug.cgi?id=105143
1589         RIM PR 171941
1590
1591         Reviewed by Rob Buis.
1592         Internally reviewed by George Staikos.
1593
1594         Adapt to new method names and encapsulation of TouchPoint data members.
1595
1596         * DumpRenderTree/blackberry/EventSender.cpp:
1597         (addTouchPointCallback):
1598         (updateTouchPointCallback):
1599         (touchEndCallback):
1600         (releaseTouchPointCallback):
1601         (sendTouchEvent):
1602
1603 2012-12-16  Maciej Stachowiak  <mjs@apple.com>
1604
1605         Add a script to compute the mean and 95% confidence interval (using two-sided t-test) of a sample set
1606         https://bugs.webkit.org/show_bug.cgi?id=105148
1607
1608         Reviewed by Filip Pizlo.
1609
1610         Usage examples:
1611
1612         $ sampstat --help
1613         Usage: sampstat [options] [FILES]
1614           Compute the mean and 95% confidence interval of a sample set.
1615           Standard input or files must contain two or more decimal numbers, one per line.
1616
1617         Options:
1618           -h, --help            show this help message and exit
1619           -u UNIT, --unit=UNIT  assume values are in units of UNIT
1620           -v, --verbose         print all values (with units)
1621
1622         $ sampstat -u MB memresults.txt
1623         2356.90 MB +/- 101.34 MB (4.3%)
1624
1625         $ sampstat -v -u MB memresults.txt
1626               2318.21 MB
1627               2399.56 MB
1628               2352.93 MB
1629         ----------------
1630         Mean: 2356.90 MB +/- 101.34 MB (4.3%)
1631
1632
1633         * Scripts/sampstat: Added.
1634         (sum): Helper function to compute the sum of a list.
1635         (arithmeticMean): Compute the meam of a list.
1636         (standardDeviation): Compute the sample standard deviation (unbiased estimator).
1637         (standardError): Compute the sample standard error.
1638         (tDist): Compute t(.025, n-1), the t-value for atwo-sided 95% confidence interval.
1639         (twoSidedConfidenceInterval): Compute the two-sided confidence interval range about the mean.
1640
1641 2012-12-15  Nima Ghanavatian  <nghanavatian@rim.com>
1642
1643         [BlackBerry] Pass key modifiers with touch and mouse events
1644         https://bugs.webkit.org/show_bug.cgi?id=105108
1645
1646         Reviewed by Rob Buis.
1647
1648         Update DRT to new api.
1649
1650         Internally reviewed by Gen Mak.
1651
1652         * DumpRenderTree/blackberry/EventSender.cpp:
1653         (mouseDownCallback):
1654         (mouseUpCallback):
1655         (mouseMoveToCallback):
1656
1657 2012-12-15  Zan Dobersek  <zandobersek@gmail.com>
1658
1659         Stop the XvfbDriver before proceeding with its starting setup
1660         https://bugs.webkit.org/show_bug.cgi?id=105006
1661
1662         Reviewed by Philippe Normand.
1663
1664         Just as with the base Driver implementation, stop the XvfbDriver
1665         before proceeding with setting up lock files and a new Xvfb instance.
1666         This forces any driver or Xvfb instances that are currently under
1667         the driver's control to be closed/killed and cleaned up after.
1668
1669         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1670         (XvfbDriver._start):
1671
1672 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
1673
1674         webkitpy: make port_name a required parameter to the non-test ports
1675         https://bugs.webkit.org/show_bug.cgi?id=105057
1676
1677         Reviewed by Eric Seidel.
1678
1679         This is just minor code cleanup along the path to making Port
1680         constructors simpler and pushing more logic into the port factory.
1681
1682         * Scripts/webkitpy/layout_tests/port/base.py:
1683         (Port.__init__):
1684         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1685         (PortTest.make_port):
1686         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1687         (TestWebKitPort.__init__):
1688         * Scripts/webkitpy/layout_tests/port/test.py:
1689         (TestPort):
1690         (TestPort.determine_full_port_name):
1691         (TestPort.__init__):
1692         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py:
1693         (XvfbDriverTest.make_driver):
1694         (XvfbDriverTest.test_stop):
1695
1696 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
1697
1698         garden-o-matic doesn't know about reftests
1699         https://bugs.webkit.org/show_bug.cgi?id=101976
1700
1701         Reviewed by Eric Seidel.
1702
1703         Attempt to make garden-o-matic properly aware of reftests again.
1704
1705         This is re-landing r137407 with a change to
1706         rebaselineWithStatusUpdates() to actually pass the filtered
1707         failureInofLost to checkout.rebaseline().
1708
1709         Unfortunately, the testing coverage for this whole module is almost
1710         non-existent, and it's not obvious that there's a good way to
1711         stub out checkout.rebaseline() to test that the right thing
1712         happens. I verified the change manually.
1713
1714         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1715         (.):
1716         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1717         (.):
1718         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
1719
1720 2012-12-14  Ryosuke Niwa  <rniwa@webkit.org>
1721
1722         "Running 1 DumpRenderTree over X shards" is not a helpful output
1723         https://bugs.webkit.org/show_bug.cgi?id=104858
1724
1725         Reviewed by Dirk Pranke.
1726
1727         The word "shard" is not typically used in test frameworks. This word is confusing for many WebKit
1728         contributors who typically don't have much experience on distributed computations and databases.
1729         So only print this message out in the debug mode.
1730
1731         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1732         * Scripts/webkitpy/layout_tests/views/printing.py:
1733         (Printer.print_workers_and_shards):
1734
1735 2012-12-13  Eric Seidel  <eric@webkit.org>
1736
1737         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
1738         https://bugs.webkit.org/show_bug.cgi?id=104891
1739
1740         Reviewed by Ryosuke Niwa.
1741
1742         I also implemented a very simple "Sample" Profiler using
1743         Mac OS X's /usr/bin/sample command line tool.
1744
1745         The real reason for this abstraction is to make it easy
1746         to support both perf and pprof on linux which seem to
1747         be about equally popular among those I ask in the Chrome team.
1748
1749         * Scripts/webkitpy/common/system/profiler.py:
1750         (ProfilerFactory.create_profiler):
1751         (ProfilerFactory):
1752         (ProfilerFactory.available_profilers_by_name):
1753         (ProfilerFactory.default_profiler_name):
1754         (Sample):
1755         (Sample.__init__):
1756         (Sample.attach_to_pid):
1757         (Sample.profile_after_exit):
1758         (IProfiler.attach_to_pid):
1759         * Scripts/webkitpy/layout_tests/port/driver.py:
1760         (Driver.__init__):
1761         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1762         (PerfTestsRunner._parse_args):
1763
1764 2012-12-13  Eric Seidel  <eric@webkit.org>
1765
1766         Unreviewed, fixing typo in python unittest.
1767
1768         I changed the default profiler on linux from pprof to perf in
1769         bug 104971.  I failed to update the unittest results at that time.
1770
1771         * Scripts/webkitpy/common/system/profiler_unittest.py:
1772         (ProfilerFactoryTest.test_basic):
1773
1774 2012-12-13  Ian Vollick  <vollick@chromium.org>
1775
1776         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
1777         https://bugs.webkit.org/show_bug.cgi?id=104911
1778
1779         Reviewed by Dirk Pranke.
1780
1781         Adds two virtual tests suites to run the tests in
1782         compositing/overflow/ and scrollbars/ with the flag
1783         --enable-accelerated-overflow-scroll, which I've also plumbed through
1784         in the usual way.
1785
1786         * DumpRenderTree/chromium/DumpRenderTree.cpp:
1787         (main):
1788         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1789         (WebPreferences):
1790         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1791         (WebTestRunner::WebPreferences::reset):
1792         (WebTestRunner::WebPreferences::applyTo):
1793         * DumpRenderTree/chromium/TestShell.cpp:
1794         (TestShell::TestShell):
1795         (TestShell::resetWebSettings):
1796         * DumpRenderTree/chromium/TestShell.h:
1797         (TestShell::setAcceleratedCompositingForOverflowScrollEnabled):
1798         (TestShell):
1799         * Scripts/webkitpy/layout_tests/port/chromium.py:
1800         (ChromiumPort.virtual_test_suites):
1801
1802 2012-12-13  Eric Seidel  <eric@webkit.org>
1803
1804         Use 'perf' to profile on linux instead of google-pprof by default
1805         https://bugs.webkit.org/show_bug.cgi?id=104971
1806
1807         Reviewed by Dirk Pranke.
1808
1809         This makes Chromium Linux match Chromium Android and use the perf
1810         tool by default.  Once https://bugs.webkit.org/show_bug.cgi?id=104891
1811         lands it will be possible to use pprof again on Linux.
1812
1813         This is slightly more advanced than the AndroidPerf profiler as
1814         instead of using a timeout on "perf record" I instead watch
1815         for the termination of the target process and then control-C
1816         the 'perf record' process.  This required me to add two new
1817         methods to Executive, one to have a limited-time wait() and
1818         the second to be able to send a control-C.  I chose to add
1819         these to Executive to make them easier to mock/fix-for-win32
1820         at a later time if needed.
1821
1822         * Scripts/webkitpy/common/system/executive.py:
1823         (Executive.wait_limited):
1824         (Executive.interrupt):
1825         * Scripts/webkitpy/common/system/profiler.py:
1826         (ProfilerFactory.create_profiler):
1827         (Perf):
1828         (Perf.__init__):
1829         (Perf._perf_path):
1830         (Perf.attach_to_pid):
1831         (Perf._first_ten_lines_of_profile):
1832         (Perf.profile_after_exit):
1833
1834 2012-12-13  Julie Parent  <jparent@chromium.org>
1835
1836         Dashboard cleanup: remove usage of global g_builders.
1837         https://bugs.webkit.org/show_bug.cgi?id=104941
1838
1839         Reviewed by Dirk Pranke.
1840
1841         The dashboards use a lot of global state, which makes hacking on them
1842         complicated. This change removes the use of one such global: g_builders.
1843         In most cases, we can just use currentBuilderGroup().builders instead,
1844         which is now currentBuilders().
1845         Surprisingly, the most changes were required to the unit tests, since
1846         they were even bigger offenders of bad hygiene, relying on global state
1847         set by other tests, randomly clobbering global variables in ways the
1848         real code doesn't, etc.
1849
1850         * TestResultServer/static-dashboards/builders.js:
1851         (BuilderGroup.prototype.setup):
1852         * TestResultServer/static-dashboards/dashboard_base.js:
1853         (.switch.return):
1854         (htmlForTestTypeSwitcher):
1855         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1856         (generatePage):
1857         (getAllTestsTrie):
1858         (processTestRunsForAllBuilders):
1859         (showPopupForBuild):
1860         (generatePageForExpectationsUpdate):
1861         (loadExpectationsLayoutTests):
1862         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1863         (resetGlobals):
1864         (stubResultsByBuilder):
1865         (test):
1866         * TestResultServer/static-dashboards/loader.js:
1867         (.):
1868         * TestResultServer/static-dashboards/loader_unittests.js:
1869
1870 2012-12-13  Eric Seidel  <eric@webkit.org>
1871
1872         Callers should not have to stringify args before calling Executive run_command/popen
1873         https://bugs.webkit.org/show_bug.cgi?id=104975
1874
1875         Reviewed by Dirk Pranke.
1876
1877         One could argue that we should match the python call syntax here,
1878         but I think it's a more friendly API if we automagically handle
1879         stringification of args in run_command, etc.
1880         This removes map(unicode, args) from several callsites.
1881
1882         When I first tried to land this change, I didn't realize that
1883         Executive._command_for_printing depended on this behavior
1884         having been applied to args in run_command.  The fix is to
1885         call _stringify_args in both run_command and popen.
1886         This is slightly redundant, but given how short args have to be
1887         (due to shell limits), I don't think the double-encode check
1888         matters in practice.
1889
1890         This is slightly complicated by the fact that apache_http_server.py
1891         is the one caller in our codebase which uses shell=True.
1892         shell=True is a well-documented trail-of-tears:
1893         http://stackoverflow.com/questions/3172470/actual-meaning-of-shell-true-in-subprocess
1894         but to support this legacy (windows-only) code (which I can't easily test)
1895         I've added an if-hack to avoid stringifying the the popen(shell=True) case.
1896
1897         * Scripts/webkitpy/common/system/executive.py:
1898         (Executive.run_command):
1899         (Executive._stringify_args):
1900         (Executive.popen):
1901         * Scripts/webkitpy/common/system/executive_unittest.py:
1902         (ExecutiveTest.test_auto_stringify_args):
1903         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1904         (attach_to_pid):
1905
1906 2012-12-13  Adrienne Walker  <enne@chromium.org>
1907
1908         Unreviewed, rolling out r137645, r137646, and r137667.
1909         http://trac.webkit.org/changeset/137645
1910         http://trac.webkit.org/changeset/137646
1911         http://trac.webkit.org/changeset/137667
1912         https://bugs.webkit.org/show_bug.cgi?id=104911
1913
1914         Breaks some overflow layout tests
1915
1916         * Scripts/webkitpy/layout_tests/port/chromium.py:
1917         (ChromiumPort.virtual_test_suites):
1918
1919 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1920
1921         Unreviewed, rolling out r137661.
1922         http://trac.webkit.org/changeset/137661
1923         https://bugs.webkit.org/show_bug.cgi?id=104891
1924
1925         broke unit tests, run-webkit-tests
1926
1927         * Scripts/webkitpy/common/system/executive.py:
1928         (Executive._run_command_with_teed_output):
1929         (Executive.run_command):
1930         (Executive.popen):
1931         * Scripts/webkitpy/common/system/profiler.py:
1932         (ProfilerFactory.create_profiler):
1933         (GooglePProf.profile_after_exit):
1934         (IProfiler.attach_to_pid):
1935         * Scripts/webkitpy/layout_tests/port/driver.py:
1936         (Driver.__init__):
1937         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1938         (PerfTestsRunner._parse_args):
1939
1940 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1941
1942         nrwt: move rundetails, summarize_results into test_run_results
1943         https://bugs.webkit.org/show_bug.cgi?id=104963
1944
1945         Reviewed by Eric Seidel.
1946
1947         More cleanup; this the plain-old-data functions out of manager.py
1948         and into test_run_results next to the other aggregate data structures.
1949
1950         Also this cleans up a bunch of test code and reduces duplication.
1951
1952         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1953         (Manager.run):
1954         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1955         (ManagerTest.test_look_for_new_crash_logs):
1956         * Scripts/webkitpy/layout_tests/models/test_run_results.py:
1957         (TestRunResults.__init__):
1958         (TestRunResults.add):
1959         (RunDetails):
1960         (RunDetails.__init__):
1961         (_interpret_test_failures):
1962         (summarize_results):
1963         * Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py: Added.
1964         (get_result):
1965         (run_results):
1966         (summarized_results):
1967         (InterpretTestFailuresTest):
1968         (InterpretTestFailuresTest.setUp):
1969         (InterpretTestFailuresTest.test_interpret_test_failures):
1970         (SummarizedResultsTest):
1971         (SummarizedResultsTest.setUp):
1972         (SummarizedResultsTest.test_no_svn_revision):
1973         (SummarizedResultsTest.test_svn_revision):
1974         (SummarizedResultsTest.test_summarized_results_wontfix):
1975         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
1976         (BuildBotPrinterTests.get_printer):
1977         (BuildBotPrinterTests.test_print_unexpected_results):
1978         (BuildBotPrinterTests.test_print_results):
1979
1980 2012-12-13  Eric Seidel  <eric@webkit.org>
1981
1982         Make --profile output easier to understand for pprof on linux
1983         https://bugs.webkit.org/show_bug.cgi?id=104957
1984
1985         Reviewed by Dirk Pranke.
1986
1987         pprof is kinda mysterious, so we print some information
1988         on how to use it, and how to learn how to read the sample dump.
1989         This is similar to what we've done for perf output for Chromium Android.
1990
1991         * Scripts/webkitpy/common/system/profiler.py:
1992         (GooglePProf._pprof_path):
1993         (GooglePProf.profile_after_exit):
1994
1995 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1996
1997         Revert r137407 - it didn't actually work right.
1998         https://bugs.webkit.org/show_bug.cgi?id=101976
1999
2000         Unreviewed, build fix.
2001
2002         Turns out that rebaselining was completely broken with this change :(
2003         We're passing in a list of test names rather than the list of
2004         failureInfo objects.
2005
2006         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2007         (.):
2008         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2009         (.):
2010         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2011
2012 2012-12-13  Antoine Labour  <piman@chromium.org>
2013
2014         [chromium] Small fixes for WebPluginContainer::setWebLayer
2015         https://bugs.webkit.org/show_bug.cgi?id=104953
2016
2017         Reviewed by James Robinson.
2018
2019         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Clean up
2020         dangling pointers, force correct invalidation of the layer.
2021
2022 2012-12-13  Eric Seidel  <eric@webkit.org>
2023
2024         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
2025         https://bugs.webkit.org/show_bug.cgi?id=104891
2026
2027         Reviewed by Dirk Pranke.
2028
2029         I also implemented a very simple "Sample" Profiler using
2030         Mac OS X's /usr/bin/sample command line tool.
2031
2032         I also moved the map(unicode, args) call into Executive.popen
2033         so that all callers don't have to do it themselves.
2034
2035         The real reason for this abstraction is to make it easy
2036         to support both perf and pprof on linux which seem to
2037         be about equally popular among those I ask in the Chrome team.
2038
2039         * Scripts/webkitpy/common/system/executive.py:
2040         (Executive.popen):
2041         * Scripts/webkitpy/common/system/profiler.py:
2042         (ProfilerFactory.create_profiler):
2043         (ProfilerFactory):
2044         (ProfilerFactory.available_profilers_by_name):
2045         (ProfilerFactory.default_profiler_name):
2046         (Sample):
2047         (Sample.__init__):
2048         (Sample.attach_to_pid):
2049         (Sample.profile_after_exit):
2050         (IProfiler.attach_to_pid):
2051         * Scripts/webkitpy/layout_tests/port/driver.py:
2052         (Driver.__init__):
2053         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2054         (PerfTestsRunner._parse_args):
2055
2056 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2057
2058         support -wk2 port names properly in webkitpy.layout_tests.port
2059         https://bugs.webkit.org/show_bug.cgi?id=104761
2060
2061         Reviewed by Eric Seidel.
2062
2063         Our support for WK2 port names was inconsistent at best; this patch
2064         modifies the code so that <port_name>-wk2 is supported for all ports that
2065         have WK2 implementations and is equivalent to specifying <port_name>
2066         and -2/--webkit-test-runner. In addition, this modifies
2067         builders.all_port_names() to include the wk2 variants.
2068
2069         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
2070         * Scripts/webkitpy/layout_tests/port/apple.py:
2071         (ApplePort.determine_full_port_name):
2072         * Scripts/webkitpy/layout_tests/port/base.py:
2073         (Port.determine_full_port_name):
2074         (Port.__init__):
2075         * Scripts/webkitpy/layout_tests/port/builders.py:
2076         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
2077         (DriverTest.make_port):
2078         * Scripts/webkitpy/layout_tests/port/mac.py:
2079         (MacPort.default_baseline_search_path):
2080         * Scripts/webkitpy/layout_tests/port/win.py:
2081         (WinPort.default_baseline_search_path):
2082         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2083         (TestRebaselineExpectations.test_rebaseline_expectations):
2084
2085 2012-12-13  Ian Vollick  <vollick@chromium.org>
2086
2087         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
2088         https://bugs.webkit.org/show_bug.cgi?id=104911
2089
2090         Reviewed by Adrienne Walker.
2091
2092         Adds two virtual tests suites to run the tests in
2093         compositing/overflow/ and scrollbars/ with the flag
2094         --enable-accelerated-compositing-for-overflow-scroll
2095
2096         * Scripts/webkitpy/layout_tests/port/chromium.py:
2097         (ChromiumPort.virtual_test_suites):
2098
2099 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2100
2101         delete extraneous self.fail in mastercfg_unittest.py
2102         https://bugs.webkit.org/show_bug.cgi?id=104940
2103
2104         Reviewed by Eric Seidel.
2105
2106         This was accidentally introduced in r137404.
2107
2108         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2109
2110 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2111
2112         Fix a mis-named variable after r137621.
2113         https://bugs.webkit.org/show_bug.cgi?id=104871
2114
2115         Unreviewed, build fix.
2116
2117         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
2118         (BuildBotPrinter.print_results):
2119         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2120
2121 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2122
2123         webkitpy: rename ResultSummary
2124         https://bugs.webkit.org/show_bug.cgi?id=104871
2125
2126         Reviewed by Eric Seidel.
2127
2128         This patch changes the names of ResultSummary and related modules,
2129         classes, and variables as a part of the larger clean up in bug 103824.
2130
2131         This change is basically a bunch of search and replaces:
2132         - result_summary.py -> test_run_results.py
2133         - class ResultSummary -> class TestRunResults
2134         - "result_summary" or "rs" as a generic variable name -> "run_results",
2135           when it doesn't matter (or we don't know) if this refers to the
2136           initial pass over the tests or the retry pass.
2137         - "result_summary" as a variable containing the results of the initial,
2138           complete test run -> "initial_results"
2139         - "retry_summary" as a variable containing the results of retried
2140           results -> "retry_results"
2141         - ResultSummary.results (the property) -> TestRunResults.results_by_name
2142         - ResultSummary.failures (the property) -> TestRunResults.failures_by_name
2143         - RunDetails.result_summary -> RunDetails.initial_results
2144         - RunDetails.retry_summary -> RunDetails.retry_results
2145
2146         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2147         (LayoutTestRunner.__init__):
2148         (LayoutTestRunner.run_tests):
2149         (LayoutTestRunner._mark_interrupted_tests_as_skipped):
2150         (LayoutTestRunner._interrupt_if_at_failure_limits):
2151         (LayoutTestRunner._interrupt_if_at_failure_limits.interrupt_if_at_failure_limit):
2152         (LayoutTestRunner._update_summary_with_result):
2153         (LayoutTestRunner._handle_finished_test):
2154         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
2155         (FakePrinter.print_expected):
2156         (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
2157         (LayoutTestRunnerTests.test_update_summary_with_result):
2158         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2159         (RunDetails.__init__):
2160         (summarize_results):
2161         (Manager.run):
2162         (Manager._look_for_new_crash_logs):
2163         (Manager._tests_to_retry):
2164         (Manager._upload_json_files):
2165         (Manager._stats_trie):
2166         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2167         (ManagerTest.test_look_for_new_crash_logs):
2168         (SummarizeResultsTest):
2169         (SummarizeResultsTest.get_run_results):
2170         (SummarizeResultsTest.summarized_results):
2171         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
2172         (JSONLayoutResultsGenerator.__init__):
2173         (JSONLayoutResultsGenerator._get_modifier_char):
2174         (JSONLayoutResultsGenerator._insert_failure_summaries):
2175         (JSONLayoutResultsGenerator._get_failure_summary_entry):
2176         * Scripts/webkitpy/layout_tests/models/test_run_results.py: Renamed from Tools/Scripts/webkitpy/layout_tests/models/result_summary.py.
2177         (TestRunResults):
2178         (TestRunResults.__init__):
2179         (TestRunResults.add):
2180         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2181         (get_test_results):
2182         (MainTest.test_basic):
2183         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
2184         (BuildBotPrinter.print_results):
2185         (BuildBotPrinter.print_run_results):
2186         (BuildBotPrinter._print_run_results_entry):
2187         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2188         (BuildBotPrinterTests.get_run_results):
2189         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
2190         (BuildBotPrinterTests):
2191         * Scripts/webkitpy/layout_tests/views/printing.py:
2192         (Printer.print_expected):
2193         (Printer._print_expected_results_of_type):
2194         (Printer.print_results):
2195         (Printer._print_timing_statistics):
2196         (Printer._print_worker_statistics):
2197         (Printer._print_aggregate_test_statistics):
2198         (Printer._print_individual_test_times):
2199         (Printer._print_directory_timings):
2200         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
2201
2202 2012-12-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2203
2204         [EFL] The WebKit2 bots are building and executing WebKit1 tests
2205         https://bugs.webkit.org/show_bug.cgi?id=104690
2206
2207         Reviewed by Csaba Osztrogonác.
2208
2209         Pass --no-webkit1 option to the EFL WK2 bots when building. This
2210         will disable WK1 API and thus, the WK1 API tests.
2211
2212         * BuildSlaveSupport/build.webkit.org-config/config.json:
2213         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2214         (CompileWebKit2Only):
2215         (BuildAndTestFactory):
2216         (BuildAndTestFactory.__init__):
2217         (BuildAndTestWebKit2OnlyFactory):
2218         Now bots using BuildAndTestWebKit2OnlyFactory (for instance only EFL) will
2219         get --no-webkit1 as a parameter on the build-webkit script.
2220         * Scripts/build-webkit:
2221         Added support for -no-webkit1 for the Qt and EFL port.
2222
2223 2012-12-13  Eric Seidel  <eric@webkit.org>
2224
2225         build-webkit --chromium-android forces Make even when update-webkit used --ninja
2226         https://bugs.webkit.org/show_bug.cgi?id=104839
2227
2228         Reviewed by Tony Chang.
2229
2230         This was recommended by Peter as part of:
2231         https://bugs.webkit.org/show_bug.cgi?id=104434#c4
2232         Fixing this makes build-webkit --chromium-android work for the common (ninja) case.
2233
2234         * Scripts/webkitdirs.pm:
2235         (buildChromium):
2236
2237 2012-12-13  Julie Parent  <jparent@chromium.org>
2238
2239         Loader_unittests should reset state between runs.
2240         https://bugs.webkit.org/show_bug.cgi?id=104862
2241
2242         Reviewed by Dirk Pranke.
2243
2244         The loader_unittests were not resetting state between runs, leaving
2245         a lot of data behind in global objects that the dashboard and tests
2246         rely on.  A handful were resetting just a few objects.  Use the common
2247         resetGlobals() method instead.
2248
2249         * TestResultServer/static-dashboards/loader_unittests.js:
2250
2251 2012-12-13  Andras Becsi  <andras.becsi@digia.com>
2252
2253         [Qt][WK2] Fix painting on Mac with retina display
2254         https://bugs.webkit.org/show_bug.cgi?id=104574
2255
2256         Reviewed by Kenneth Rohde Christiansen.
2257
2258         Remove setting the devicePixelRatio experimental property
2259         since the value is now automatically picked up from Qt.
2260
2261         * MiniBrowser/qt/qml/BrowserWindow.qml:
2262
2263 2012-12-13  Jussi Kukkonen  <jussi.kukkonen@intel.com>
2264
2265         [EFL][GTK] Don't call deprecated g_type_init when glib => 2.35
2266         https://bugs.webkit.org/show_bug.cgi?id=103209
2267
2268         Reviewed by Kenneth Rohde Christiansen.
2269
2270         g_type_init() is deprecated from 2.35.0 onwards. Don't call it
2271         in that case to avoid warnings and/or build failure.
2272
2273         * DumpRenderTree/gtk/ImageDiff.cpp:
2274         (main):
2275
2276 2012-12-13  Claudio Saavedra  <csaavedra@igalia.com>
2277
2278         [Tools] Add exception for gtk_style_context_get()
2279         https://bugs.webkit.org/show_bug.cgi?id=104896
2280
2281         Reviewed by Martin Robinson.
2282
2283         Add new exception for gtk_style_context_get().
2284
2285         * Scripts/webkitpy/style/checkers/cpp.py:
2286         (check_for_null): Add exception for gtk_style_context_get().
2287         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2288         (WebKitStyleTest.test_null_false_zero): Add a unit
2289         test for the new exception.
2290
2291 2012-12-13  Joaquim Rocha  <jrocha@igalia.com>
2292
2293         REGRESSION (r137432): The /webkit2/WebKitCookieManager/accept-policy unit test is failing
2294         https://bugs.webkit.org/show_bug.cgi?id=104790
2295
2296         Reviewed by Carlos Garcia Campos.
2297
2298         Unskip the GTK accept policy test.
2299
2300         * Scripts/run-gtk-tests:
2301         (TestRunner):
2302
2303 2012-12-13  Xingnan Wang  <xingnan.wang@intel.com>
2304
2305         Unreviewed. Adding myself to committers list.
2306
2307         * Scripts/webkitpy/common/config/committers.py:
2308
2309 2012-12-12  Roger Fong  <roger_fong@apple.com>
2310
2311         Unreviewed. Fix typo in NRWT.
2312
2313         * Scripts/webkitpy/layout_tests/port/base.py:
2314         (Port.is_reference_html_file):
2315
2316 2012-12-12  Roger Fong  <roger_fong@apple.com>
2317
2318         Fix detection of ref test files in old and new run webkit tests.
2319         https://bugs.webkit.org/show_bug.cgi?id=104850.
2320
2321         Reviewed by Timothy Horton.
2322
2323         * Scripts/old-run-webkit-tests:
2324         (isUsedInReftest):
2325         Add notref and ref prefix and suffix cases to isUsedinReftest.
2326         * Scripts/webkitpy/layout_tests/port/base.py:
2327         (Port):
2328         (Port.is_reference_html_file):
2329         Typo in NRWT.
2330
2331 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2332
2333         rework ChromiumAndroidDriver.start() et. al to avoid funky layering
2334         https://bugs.webkit.org/show_bug.cgi?id=104856
2335
2336         Reviewed by Eric Seidel.
2337
2338         r137523 / bug 10357 introduced some back-and-forth hopping during
2339         driver.start() in the chromium android port. This patch cleans it
2340         up a bit.
2341
2342         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2343         (ChromiumAndroidDriver.__init__):
2344         (ChromiumAndroidDriver._start_once):
2345         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2346         * Scripts/webkitpy/layout_tests/port/driver.py:
2347         (Driver.start):
2348         (Driver._start):
2349
2350 2012-12-12  Eric Seidel  <eric@webkit.org>
2351
2352         Add support for run-perf-tests --chromium-android --profile
2353         https://bugs.webkit.org/show_bug.cgi?id=103587
2354
2355         Reviewed by Dirk Pranke.
2356
2357         This makes it possible to profile PerformanceTests on an Android device
2358         using run-perf-tests --chromium-android --profile.
2359         CPU profiling on Android is sadly non-trivial, but this goes a long
2360         ways towards making it more accessible to non-Android developers.
2361
2362         Although this script requires your device to be flashed with a
2363         userdebug build (an annoyingly non-trivial endeavor for external contributors)
2364         it endeavors to detect and explain the common errors encountered
2365         while trying to setup a device for perf-based profiling.
2366
2367         * Scripts/webkitpy/common/system/profiler.py:
2368         (GooglePProf.profile_after_exit): Discovered that pprof will just hang if passed a missing file.
2369         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2370         (AndroidPerf):
2371         (AndroidPerf.__init__):
2372         (AndroidPerf.check_configuration):
2373         (AndroidPerf.print_setup_instructions):
2374         (_file_exists_on_device):
2375         (_run_adb_command):
2376         (attach_to_pid):
2377         (_perf_version_string):
2378         (_find_perfhost_binary):
2379         (_perfhost_path):
2380         (_first_ten_lines_of_profile):
2381         (profile_after_exit):
2382         (ChromiumAndroidDriver.__init__):
2383         (ChromiumAndroidDriver.__init__.methods):
2384         (ChromiumAndroidDriver._find_or_create_symfs):
2385         (ChromiumAndroidDriver._start_once):
2386         (ChromiumAndroidDriver):
2387         (ChromiumAndroidDriver._run_post_start_tasks):
2388         (ChromiumAndroidDriver._pid_from_android_ps_output):
2389         (ChromiumAndroidDriver._pid_on_target):
2390         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2391         (ChromiumAndroidDriverTest.test_pid_from_android_ps_output):
2392         (AndroidPerfTest):
2393         (AndroidPerfTest.test_perf_output_regexp):
2394         * Scripts/webkitpy/layout_tests/port/driver.py:
2395         (Driver._setup_environ_for_driver):
2396         (Driver):
2397         (Driver._start):
2398         (Driver._run_post_start_tasks):
2399         (Driver._pid_on_target):
2400
2401 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2402
2403         nrwt: remove deprecated ORWT compatibility command line flags
2404         https://bugs.webkit.org/show_bug.cgi?id=104301
2405
2406         Reviewed by Eric Seidel.
2407
2408         The command line flags for --noshow-results, --[no-]launch-safari,
2409         and --use-remote-links-to-tests were for compatibility with
2410         old-run-webkit-tests, but the compatibility is not needed any more
2411         since NRWT is basically default and the scripts that passed the
2412         old flags have been updated.
2413
2414         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2415         (_set_up_derived_options):
2416         (parse_args):
2417
2418 2012-12-12  John Griggs  <jgriggs@rim.com>
2419
2420         2012-12-12  John Griggs  <jgriggs@rim.com>
2421
2422         [BlackBerry] Update Media Controls
2423         https://bugs.webkit.org/show_bug.cgi?id=104443
2424         https://przilla.ott.qnx.com/bugzilla/show_bug.cgi?id=204748
2425
2426         Update Media Controls for BlackBerry platform to allow audio, embedded video and fullscreen video controls to be positioned and styled differently.
2427
2428         Reviewed by Rob Buis.
2429
2430         BlackBerry-specific update to Media Controls - No new tests required.
2431
2432         * Scripts/webkitperl/FeatureList.pm:
2433
2434 2012-12-12  Justin Novosad  <junov@google.com>
2435
2436         Adding myself to SkiaGraphics watchlist
2437
2438         Unreviewed
2439
2440         * Scripts/webkitpy/common/config/watchlist:
2441
2442 2012-12-12  Martin Robinson  <mrobinson@igalia.com>
2443
2444         [GTK] Add authentication support to DRT and fix exposed issues in the libsoup backend
2445         https://bugs.webkit.org/show_bug.cgi?id=104285
2446
2447         Reviewed by Gustavo Noronha Silva.
2448
2449         Add support to DumpRenderTree for running authentication tests. Since the DRT
2450         expects an authentication callback, we add one to DRTSupport to avoid #ifdefs
2451         in platform-independent code for GTK+.
2452
2453         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2454         (resetDefaultsToConsistentValues): Reset the authentication password and username.
2455         (authenticationCallback): Added.
2456         (createWebView): Attach the authentiation callback during startup.
2457         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
2458         (soupURIToKURL): soup_uri_to_string does not preserve passwords embedded
2459         in the URL so we add a somewhat messy method of re-adding them when they exist.
2460         It would be nice to use soupURIToKURL here, but it seems we cannot use KURL without
2461         pulling in lots of WebCore code so we use string search and replace.
2462         (TestRunner::queueLoad): Use the new helper.
2463
2464 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2465
2466         garden-o-matic should prefer efl/ over efl-wk1/ and efl-wk2/ when rebaselining
2467         https://bugs.webkit.org/show_bug.cgi?id=102389
2468
2469         Reviewed by Tony Chang.
2470
2471         The code wasn't aware of both the efl-wk1 and efl-wk2 variants,
2472         because all_port_names() by itself doesn't tell you about the
2473         variants. Work around this for now, and filed bug 104761 to
2474         address the underlying problem.
2475
2476         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
2477         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
2478         (BaselineOptimizerTest.test_efl):
2479
2480 2012-12-12  Krzysztof Czech  <k.czech@samsung.com>
2481
2482         [EFL] Possibility to turn off accessibility feature for WebKit-EFL.
2483         https://bugs.webkit.org/show_bug.cgi?id=103036
2484
2485         Reviewed by Laszlo Gombos.
2486
2487         Guard dependencies for accessibility (ATK), so that they can be turned off.
2488
2489         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
2490         (AccessibilityUIElement):
2491         * WebKitTestRunner/PlatformEfl.cmake:
2492
2493 2012-12-12  Simon Fraser  <simon.fraser@apple.com>
2494
2495         Add some entries in the watchlist, and register me for them.
2496
2497         * Scripts/webkitpy/common/config/watchlist:
2498
2499 2012-12-12  Ilya Tikhonovsky  <loislo@chromium.org>
2500
2501         Web Inspector: Native Memory Instrumentation: remove fake root MemoryObjectInfo.
2502         https://bugs.webkit.org/show_bug.cgi?id=104796
2503
2504         Reviewed by Yury Semikhatsky.
2505
2506         It was not a good idea to introduce a fake root MemoryObjectInfo.
2507         It makes a problem when we visit an object without its own MemoryObjectType.
2508
2509         Example: RenderBox has a global pointer to a hash map.
2510         HashMap doesn't have its own object type because it is a generic container.
2511         It will inherit object type from the fake root memory object info.
2512         The same could happen for another container in another class with other MemoryObjectType.
2513
2514         This fact forces me to create custom process method for root objects
2515         because they need to have their own MemoryObjectInfo with customisable memory object type.
2516
2517         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2518
2519         Drive by fix: InstrumentedPointer* was replaced with Wrapper* because actually it is using
2520         for instrumented and not instrumented object classes.
2521
2522 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2523
2524         [GTK] Remove the last of unnecessary configuration options in configure.ac
2525         https://bugs.webkit.org/show_bug.cgi?id=104793
2526
2527         Reviewed by Martin Robinson.
2528
2529         Removing the filters and indexed database features from the list of features
2530         being configurable through the configure script. This is not the case anymore
2531         as the corresponding configuration options are being removed.
2532
2533         * Scripts/webkitdirs.pm:
2534         (buildAutotoolsProject):
2535
2536 2012-12-12  Mario Sanchez Prada  <mario@webkit.org>
2537
2538         [GTK][WK2] New API to detect display/execution of insecure content
2539         https://bugs.webkit.org/show_bug.cgi?id=104578
2540
2541         Reviewed by Martin Robinson.
2542
2543         Add coding style exception for 'readability/enum_casing' in GTK.
2544
2545         The GTK+ API use upper case, underscore separated, words in
2546         certain types of enums (e.g. signals, properties).
2547
2548         * Scripts/webkitpy/style/checker.py: Added exception, both for
2549         WebKitGTK+ and WebKit2GTK+.
2550
2551 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
2552
2553         [chromium] Fix build on chromium-win
2554
2555         Unreviewed build-fix.
2556
2557         MSVS indeed doesn't have snprintf, adding wtf/StringExtras.h include
2558         to work around that.
2559
2560         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
2561
2562 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
2563
2564         [Qt] Fix build on Mac
2565
2566         Reviewed by Csaba Osztrogonác.
2567
2568         On Mac OS X we have sqlite as a system library available.
2569
2570         * qmake/mkspecs/features/features.prf:
2571
2572 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2573
2574         Unreviewed GTK gardening.
2575
2576         Skipping another unit test that regressed in r137432.
2577
2578         * Scripts/run-gtk-tests:
2579         (TestRunner):
2580
2581 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
2582
2583         [Qt] Fix linkage against SQLite
2584         https://bugs.webkit.org/show_bug.cgi?id=104781
2585
2586         Reviewed by Csaba Osztrogonác.
2587
2588         Detect SQLite either through pkg-config, from the SQLITE3SRCDIR
2589         environment variable or from qtbase via a qt5.git build. Bail out early
2590         with an error message if neither option applies.
2591
2592         This removes the usage of $$QT.core.sources and the dependency on Qt's
2593         system-sqlite configure setting. We should always favour a system
2594         library over a copy in Qt.
2595
2596         * qmake/mkspecs/features/features.prf:
2597
2598 2012-12-12  Tommy Widenflycht  <tommyw@google.com>
2599
2600         MediaStream API: Change the data channel descriptor pattern to a handler pattern
2601         https://bugs.webkit.org/show_bug.cgi?id=104543
2602
2603         Reviewed by Kent Tamura.
2604
2605         Adding mocks to be able to test this refactoring.
2606
2607         * DumpRenderTree/DumpRenderTree.gypi:
2608         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.cpp: Added.
2609         (DataChannelReadyStateTask):
2610         (DataChannelReadyStateTask::DataChannelReadyStateTask):
2611         (MockWebRTCDataChannelHandler::MockWebRTCDataChannelHandler):
2612         (MockWebRTCDataChannelHandler::setClient):
2613         (MockWebRTCDataChannelHandler::bufferedAmount):
2614         (MockWebRTCDataChannelHandler::sendStringData):
2615         (MockWebRTCDataChannelHandler::sendRawData):
2616         (MockWebRTCDataChannelHandler::close):
2617         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.h: Added.
2618         (MockWebRTCDataChannelHandler):
2619         (MockWebRTCDataChannelHandler::taskList):
2620         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
2621         (RTCPeerConnectionReadyStateTask::RTCPeerConnectionReadyStateTask):
2622         (RemoteDataChannelTask::RemoteDataChannelTask):
2623         (MockWebRTCPeerConnectionHandler::createDataChannel):
2624         (MockWebRTCPeerConnectionHandler::stop):
2625         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
2626         (MockWebRTCPeerConnectionHandler):
2627
2628 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
2629
2630         [chromium] WebTestPlugin must not depend on WTF::String
2631         https://bugs.webkit.org/show_bug.cgi?id=104783
2632
2633         Reviewed by Kent Tamura.
2634
2635         In a component build, we can't access the WTF that is part of the
2636         WebKit component. Accessing WTF instead will pull in a separate copy
2637         into the TestRunner library. Instead of WTF::String::number use
2638         snprintf() to print numbers.
2639
2640         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
2641
2642 2012-12-12  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
2643
2644         [Qt] Do not automatically enable force_static_libs_as_shared when using debug_and_release
2645         https://bugs.webkit.org/show_bug.cgi?id=104562
2646
2647         Reviewed by Tor Arne Vestbø.
2648
2649         force_static_libs_as_shared creates issues on Mac if QtWebKit is built as a framework.
2650         QtWebProcess, for example, will link to the framework's release library and will rely
2651         on DYLD_IMAGE_SUFFIX to switch to the debug build. Since QtWebProcess will also links
2652         directly to the debug build of WebCore, and that QtWebKit release will link in the
2653         release build of WebCore, both the debug and release version will be loaded.
2654
2655         To fix this properly we should make sure that applications always link to the release
2656         build of dependent libraries by bypassing resolveFinalLibraryName.
2657         Disable it for now as force_static_libs_as_shared is mainly for development while
2658         debug_and_release is mainly for production builds.
2659
2660         * qmake/mkspecs/features/default_post.prf:
2661
2662 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2663
2664         Unreviewed GTK gardening of the unit tests.
2665
2666         Skip two unit tests that regressed during the hackfest.
2667
2668         * Scripts/run-gtk-tests:
2669         (TestRunner):
2670
2671 2012-12-11  Ryosuke Niwa  <rniwa@webkit.org>
2672
2673         Add Koji Ishii to the list of contributors.
2674
2675         * Scripts/webkitpy/common/config/committers.py:
2676
2677 2012-12-11  Dominic Mazzoni  <dmazzoni@google.com>
2678
2679         AX: Make isActionSupported cross-platform.
2680         https://bugs.webkit.org/show_bug.cgi?id=104681
2681
2682         Reviewed by Chris Fleizach.
2683
2684         Replace isActionSupported, which takes a platform-specific
2685         argument, with three cross-platform methods.
2686
2687         * DumpRenderTree/AccessibilityUIElement.cpp:
2688         (isPressActionSupportedCallback):
2689         (isIncrementActionSupportedCallback):
2690         (isDecrementActionSupportedCallback):
2691         (AccessibilityUIElement::getJSClass):
2692         * DumpRenderTree/AccessibilityUIElement.h:
2693         (AccessibilityUIElement):
2694         * DumpRenderTree/blackberry/AccessibilityUIElementBlackBerry.cpp:
2695         (AccessibilityUIElement::isPressActionSupported):
2696         (AccessibilityUIElement::isIncrementActionSupported):
2697         (AccessibilityUIElement::isDecrementActionSupported):
2698         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2699         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2700         (WebTestRunner::AccessibilityUIElement::isPressActionSupportedCallback):
2701         (WebTestRunner):
2702         (WebTestRunner::AccessibilityUIElement::isIncrementActionSupportedCallback):
2703         (WebTestRunner::AccessibilityUIElement::isDecrementActionSupportedCallback):
2704         (WebTestRunner::AccessibilityUIElement::incrementCallback):
2705         (WebTestRunner::AccessibilityUIElement::decrementCallback):
2706         (WebTestRunner::AccessibilityUIElement::pressCallback):
2707         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2708         (AccessibilityUIElement):
2709         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2710         (AccessibilityUIElement::isPressActionSupported):
2711         (AccessibilityUIElement::isIncrementActionSupported):
2712         (AccessibilityUIElement::isDecrementActionSupported):
2713         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2714         (AccessibilityUIElement::isPressActionSupported):
2715         (AccessibilityUIElement::isIncrementActionSupported):
2716         (AccessibilityUIElement::isDecrementActionSupported):
2717         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2718         (AccessibilityUIElement::isPressActionSupported):
2719         (AccessibilityUIElement::isIncrementActionSupported):
2720         (AccessibilityUIElement::isDecrementActionSupported):
2721         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
2722         (WTR::AccessibilityUIElement::isPressActionSupported):
2723         (WTR::AccessibilityUIElement::isIncrementActionSupported):
2724         (WTR::AccessibilityUIElement::isDecrementActionSupported):
2725         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
2726         (AccessibilityUIElement):
2727         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
2728         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
2729         (WTR::AccessibilityUIElement::isPressActionSupported):
2730         (WTR):
2731         (WTR::AccessibilityUIElement::isIncrementActionSupported):
2732         (WTR::AccessibilityUIElement::isDecrementActionSupported):
2733
2734 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2735
2736         nrwt: remove result_summary.total_tests_by_expectation
2737         https://bugs.webkit.org/show_bug.cgi?id=104756
2738
2739         Reviewed by Eric Seidel.
2740
2741         This field wasn't being used by anything.
2742
2743         * Scripts/webkitpy/layout_tests/models/result_summary.py:
2744         (ResultSummary.__init__):
2745
2746 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2747
2748         garden-o-matic doesn't know about reftests
2749         https://bugs.webkit.org/show_bug.cgi?id=101976
2750
2751         Reviewed by Ojan Vafai.
2752
2753         This is a first attempt at fixing this, sort of a "stop-the-bleeding"
2754         fix. We will now look at the list of tests we're asked to rebaseline,
2755         and log which ones are reftests into the status area, and rebaseline
2756         the rest.
2757
2758         Longer-term, we should display feedback in the results area based on
2759         which tests we're looking at. If the test list contains no reftests,
2760         we should behave as today; if the list is all reftest, we should
2761         probably disable the button, and if the list is a mix, we should
2762         probably display an alert next to the button or something. It would
2763         also be nice to annotate which tests were reftests, maybe with a
2764         "(reftest)" next to the test name or something.
2765
2766         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2767         (.):
2768         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2769         (.):
2770         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2771
2772 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2773
2774         REGRESSION(r133380): new tests without expected file reported as failing tests on the bots
2775         https://bugs.webkit.org/show_bug.cgi?id=101469
2776
2777         Reviewed by Ryosuke Niwa.
2778
2779         Fix parsing of NRWT's output to identify missing results separately
2780         from other regressions.
2781
2782         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2783         (RunWebKitTests._parseNewRunWebKitTestsOutput):
2784         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2785         (MasterCfgTest.test_nrwt_leaks_parsing):
2786         (test_nrwt_missing_results):
2787
2788 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2789
2790         buildbot: change run-webkit-tests --no-launch-safari to --no-show-results
2791         https://bugs.webkit.org/show_bug.cgi?id=104299
2792
2793         Reviewed by Ryosuke Niwa.
2794
2795         Since --no-show-results is the more generically-named flag that NRWT uses.
2796
2797         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2798         (RunWebKitTests):
2799
2800 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2801
2802         remove --use-remote-links-to-tests from buildbot configs (except apple win)
2803         https://bugs.webkit.org/show_bug.cgi?id=104291
2804
2805         Reviewed by Csaba Osztrogonác.
2806
2807         This flag is ignored by NRWT and was just implemented for compatibility.
2808         ORWT still uses it, though.
2809
2810         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2811         (RunWebKitTests):
2812         (RunWebKitTests.start):
2813
2814 2012-12-11  Anders Carlsson  <andersca@apple.com>
2815
2816         ASSERT running platform/mac/plugins/convert-point.html on the WK2 bots
2817         https://bugs.webkit.org/show_bug.cgi?id=104735
2818
2819         Reviewed by Jon Honeycutt.
2820
2821         Do the testing from within NPP_New instead of NPN_SetWindow. Plug-ins aren't supposed to do any real work inside NPP_SetWindow,
2822         and doing so will more easily trigger the assertion failure. Ideally we should get rid of the initial sync painting pass, but this
2823         gets the test passing again!
2824
2825         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:
2826         (ConvertPoint::NPP_New):
2827
2828 2012-12-11  Julie Parent  <jparent@chromium.org>
2829
2830         Dashboard unit tests should not make xhrs.
2831         https://bugs.webkit.org/show_bug.cgi?id=104598
2832
2833         Reviewed by Dirk Pranke.
2834
2835         The unit test were requesting 25 files that they did not use.
2836         Stub out loader.request to avoid making the xhrs.
2837         Result: order of magnitude speed improvement in test run time (from ~1600ms to 160ms on my machine).
2838
2839         * TestResultServer/static-dashboards/run-embedded-unittests.html:
2840         * TestResultServer/static-dashboards/run-unittests.html:
2841
2842 2012-12-11  Filip Pizlo  <fpizlo@apple.com>
2843
2844         Profiler should show bytecode dumps as they would have been visible to the JITs, including the profiling data that the JITs would see
2845         https://bugs.webkit.org/show_bug.cgi?id=104647
2846
2847         Reviewed by Oliver Hunt.
2848
2849         Added a "profiling" (or "p") command to show the profiling data that the JITs saw
2850         for each JIT compilation of a code block.
2851
2852         Also added instruction counts in the "full" display and made the "full" display the
2853         default thing you see.
2854
2855         * Scripts/display-profiler-output:
2856
2857 2012-12-11  Eric Seidel  <eric@webkit.org>
2858
2859         Unreviewed, rolling out r137371.
2860         http://trac.webkit.org/changeset/137371
2861         https://bugs.webkit.org/show_bug.cgi?id=104434
2862
2863         Various scripts are not ready for out/ as the build directory,
2864         this can't work as written.
2865
2866         * Scripts/update-webkit:
2867         * Scripts/webkitdirs.pm:
2868         (buildChromium):
2869
2870 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
2871
2872         Unreviewed GTK build functionality fix.
2873
2874         Turns out the CSP feature was not enabled, that's why all the tests were failing.
2875
2876         * Scripts/webkitperl/FeatureList.pm:
2877
2878 2012-12-11  Eric Seidel  <eric@webkit.org>
2879
2880         Make Ninja the default build system for build-webkit --chromium
2881         https://bugs.webkit.org/show_bug.cgi?id=104434
2882
2883         Reviewed by Daniel Bates.
2884
2885         As far as I can tell, most Chromium developers are now using Ninja
2886         as their default build system instead of the native Xcode/Visual Studio/Make.
2887         This change makes Ninja the default for developers as well as all Chromium webkit bots.
2888
2889         Developers can pass --no-ninja to get the platform-native buildsystem instead of Ninja.
2890         It's possible we may wish to make this a per-platform default, and have
2891         Ninja only be the default for Mac/Linux.
2892
2893         There was a thread on webkit-dev discussing this change:
2894         http://lists.webkit.org/pipermail/webkit-dev/2012-December/023113.html
2895
2896         * Scripts/update-webkit:
2897
2898 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2899
2900         nrwt: record the order and run times of all of the tests in a single file
2901         https://bugs.webkit.org/show_bug.cgi?id=104184
2902
2903         Reviewed by Ryosuke Niwa.
2904
2905         Replace the N tests_run*.txt files that we output into
2906         layout-test-results with a single packed JSON trie that contains
2907         the test ordering and run statistics. This can eventually be uploaded
2908         to the TestResultsServer to replace the times.json file we currently
2909         upload as well, and will let us replay the results of a previous
2910         run fairly deterministically.
2911
2912         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2913         (Worker.__init__):
2914         (Worker.start):
2915         (Worker._run_test):
2916         (Worker.stop):
2917         (Worker._clean_up_after_test):
2918         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2919         (Manager._upload_json_files):
2920         (Manager._copy_results_html_file):
2921         (Manager):
2922         (Manager._stats_trie):
2923         (Manager._stats_trie._worker_number):
2924         * Scripts/webkitpy/layout_tests/models/test_results.py:
2925         (TestResult.__init__):
2926         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2927         (MainTest.test_missing_and_unexpected_results):
2928         (MainTest.test_retrying_and_flaky_tests):
2929         (EndToEndTest.test_end_to_end):
2930         (EndToEndTest.test_reftest_with_two_notrefs):
2931         (RebaselineTest.test_reset_results):
2932         (RebaselineTest.test_missing_results):
2933         (RebaselineTest.test_new_baseline):
2934
2935 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2936
2937         check-webkit-style complaining about TestExpectations lines a patch didn't modify
2938         https://bugs.webkit.org/show_bug.cgi?id=104687
2939
2940         Reviewed by Ojan Vafai.
2941
2942         Way back in bug 60466 / r86165, we changed the style checker
2943         for TestExpectations to ensure that the entire file linted, not
2944         just the lines modified by a patch, in order to ensure that a
2945         new line didn't cause errors elsewhere in the file (e.g., by
2946         introducing duplicate lines). At the time, errors in the
2947         expectations files were fatal and we didn't have a separate step
2948         on the bots that would lint the files to ensure things were still
2949         okay, so this was probably the right thing to do, but it could
2950         cause some patches to fail the style check even though the patches
2951         themselves were fine.
2952
2953         However, now errors aren't always fatal, and we have a separate build
2954         step to lint the files. So, we should probably not punish the
2955         patches and the style checker should filter out "unrelated" errors.
2956
2957         In theory we should be comparing the output pre- and post-patch
2958         to filter out pre-existing errors, but I don't think the style
2959         checker has that ability.
2960
2961         * Scripts/webkitpy/style/checkers/test_expectations.py:
2962         (TestExpectationsChecker.__init__):
2963         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2964         (TestExpectationsTestCase.assert_lines_lint):
2965
2966 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2967
2968         nrwt isn't rejecting unrecognized expectations
2969         https://bugs.webkit.org/show_bug.cgi?id=104701
2970
2971         Reviewed by Ojan Vafai.
2972
2973         NRWT is silently ignoring unrecognized expectations like "Text"
2974         and "Audio" (largely because I haven't fully removed support for
2975         the old syntax and keywords and restructured the code to do the
2976         proper checking). Temporarily change the tokenizer to reject
2977         unrecognized expectations until I can make those changes.
2978
2979         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2980         (TestExpectationParser._tokenize_line):
2981         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2982         (ExpectationSyntaxTests.test_unknown_expectation):
2983         (TestExpectationSerializationTests.disabled_test_string_whitespace_stripping):
2984
2985 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2986
2987         Fix a typo in my previous commit that is causing test failures.
2988
2989         Reviewed by Gustavo Noronha Silva.
2990
2991         There was an extra bit of statement appended after the end of a conditional. Somehow
2992         this compiled and caused test failures.
2993
2994         * DumpRenderTree/gtk/EventSender.cpp:
2995         (continuousMouseScrollByCallback):
2996
2997 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2998
2999         [GTK] Spurious stderr output from the event sender
3000         https://bugs.webkit.org/show_bug.cgi?id=104678
3001
3002         Reviewed by Gustavo Noronha Silva.
3003
3004         Eliminate some spurious stderr output from the EventSender by using an early return
3005         instead of an assertion that prints a warning to the console.
3006
3007         * DumpRenderTree/gtk/EventSender.cpp:
3008         (continuousMouseScrollByCallback): We don't support continuous scroll events that
3009         move by page, but we should just fail silently instead of printing an error.
3010
3011 2012-12-11  Joone Hur  <joone.hur@intel.com>
3012
3013         [GTK] Add GraphicsLayerActor
3014         https://bugs.webkit.org/show_bug.cgi?id=91940
3015
3016         Reviewed by Gustavo Noronha Silva.
3017
3018         Add GraphicsLayerActor to the list of classes that contain GObjects
3019         in the style checker script so that it does not complain about GObject
3020         coding style.
3021
3022         * Scripts/webkitpy/style/checker.py:
3023
3024 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3025
3026         [Qt] remove hack which avoids unused .obj directories all over the place
3027
3028         Reviewed by Simon Hausmann.
3029
3030         The feature is subject to upstreaming in Qt itself in
3031         https://codereview.qt-project.org/#change,42011
3032
3033         * qmake/mkspecs/features/default_post.prf:
3034
3035 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3036
3037         [Qt] move SQLITE3SRCDIR assignment to only use site
3038
3039         Reviewed by Simon Hausmann.
3040
3041         Target.pri includes WebCore.pri the almost first thing, so that is
3042         covered as well.
3043
3044         * qmake/mkspecs/features/default_pre.prf:
3045
3046 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3047
3048         [Qt] delete dead include paths
3049
3050         Reviewed by Simon Hausmann.
3051
3052         followup to https://bugs.webkit.org/show_bug.cgi?id=93446
3053
3054         * DumpRenderTree/qt/DumpRenderTree.pro:
3055
3056 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3057
3058         [GTK] Implement some missing FrameLoaderClient message in DRT
3059         https://bugs.webkit.org/show_bug.cgi?id=104444
3060
3061         Reviewed by Gustavo Noronha Silva.
3062
3063         Install a FrameLoadEvent callback that prints the necessary output for some
3064         missing FrameLoader events.
3065
3066         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3067         (frameLoadEventCallback): Added.
3068         (createWebView): Install the FrameLoader event callback.
3069
3070 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3071
3072         [Qt] change setup of QMAKE_MOC
3073
3074         Reviewed by Simon Hausmann.
3075
3076         don't load(moc) explicitly - turns out that this breaks
3077         debug_and_release builds (the feature does not like being loaded before
3078         default_post has been loaded).
3079         so instead just do a minimal call to find the moc executable.
3080
3081         as it's in fact only two files which need it, so de-centralize this call
3082         for the sake of saving some cpu cycles spent in qmake.
3083
3084         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
3085         * qmake/mkspecs/features/default_pre.prf:
3086
3087 2012-12-11  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
3088
3089         [Qt] Fix the linking of debug builds on Mac with force_static_libs_as_shared
3090         https://bugs.webkit.org/show_bug.cgi?id=104559
3091
3092         Reviewed by Tor Arne Vestbø.
3093
3094         "contains(libraries, $$TARGET): return(true)" fails for intermediate libraries since
3095         at this point TARGET may contain the "_debug" suffix.
3096
3097         Fix it by moving the TARGET adjustment later in the file.
3098
3099         * qmake/mkspecs/features/default_post.prf:
3100
3101 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3102
3103         [Coverity] [GTK] Remove some redundant null checks in DumpRenderTree
3104         https://bugs.webkit.org/show_bug.cgi?id=104570
3105
3106         Reviewed by Daniel Bates.
3107
3108         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3109         (descriptionSuitableForTestResult): Remove a redundant null check.
3110
3111 2012-12-11  Alberto Garcia  <agarcia@igalia.com>
3112
3113         Remove unused GErrors
3114         https://bugs.webkit.org/show_bug.cgi?id=104657
3115
3116         Reviewed by Martin Robinson.
3117
3118         Replace unused GErrors with null pointers.
3119
3120         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3121         (initializeFonts):
3122         * DumpRenderTree/gtk/EditingCallbacks.cpp:
3123         (dumpRange):
3124         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
3125         (WTR::inititializeFontConfigSetting):
3126
3127 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3128
3129         [GTK] Bump the jhbuild libsoup version to fix an authentication issue
3130         https://bugs.webkit.org/show_bug.cgi?id=104448
3131
3132         Reviewed by Gustavo Noronha Silva.
3133
3134         Bump the libsoup version to one from the git repository. This requires updating
3135         GLib as well which in turn requires adjusting the CFLAGS we pass when compiling
3136         the scanobj step of the gtkdoc generation.
3137
3138         * gtk/generate-gtkdoc:
3139         * gtk/jhbuild.modules:
3140
3141 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
3142
3143         Unreviewed functionality fix for the GTK port.
3144
3145         Enable the MicroData feature support for the GTK port in the FeatureList module.
3146         This feature listing now again affects which features are enabled and which disabled
3147         on the development builds of the GTK port.
3148
3149         * Scripts/webkitperl/FeatureList.pm:
3150
3151 2012-12-11  Jinwoo Song  <jinwoo7.song@samsung.com>
3152
3153         [EFL][WK2] MiniBrowser is not built
3154         https://bugs.webkit.org/show_bug.cgi?id=104629
3155
3156         Reviewed by Gyuyoung Kim.
3157
3158         As we are finding an elementary package in Tools/MiniBrowser/efl/CMakeList.txt since r137203,
3159         we need to check if the package is found in the same place not in the Tools/CMakeList.txt.
3160
3161         * CMakeLists.txt:
3162         * MiniBrowser/efl/CMakeLists.txt:
3163
3164 2012-12-11  Xabier Rodriguez Calvar  <calvaris@igalia.com>
3165
3166         [GTK][jhbuild] Switch to GStreamer 1.0 build
3167         https://bugs.webkit.org/show_bug.cgi?id=91727
3168
3169         Reviewed by Philippe Normand.
3170
3171         Switch build-webkit --gtk to GStreamer 1.0 support and build the
3172         necessary GStreamer git modules from JHBuild.
3173
3174         * gtk/jhbuild.modules: Added GStreamer 1.0.3 build support.
3175
3176 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
3177
3178         [GTK] Feature enabling/disabling should be possible through build-webkit
3179         https://bugs.webkit.org/show_bug.cgi?id=99271
3180
3181         Reviewed by Gustavo Noronha Silva.
3182
3183         Refactor the code in webkitdirs.pm that builds an autotools project to generate
3184         autogen.sh arguments, make arguments and installation prefix from the passed-in
3185         parameters rather than generating all of that in build-webkit. The autogen.sh
3186         arguments now contain only enable/disable flags for options that are actually
3187         configurable in configure.ac, the flag value actually reflecting whether the feature
3188         is enabled or disabled in the feature list.
3189
3190         Other features are overridable through modifying the GNUmakefile.features.am file.
3191         All these features are now stored in the build directory and upon change trigger
3192         a rerun of the autogen.sh script, pretty much like the autogen.sh arguments do.
3193
3194         The override-feature-defines script is called by autogen.sh, before GNUmakefile.in is
3195         generated when calling autoreconf. Its task is to copy the GNUmakefile.features.am.in
3196         into GNUmakefile.features.am and modify it if there's a text file in the build directory
3197         that contains all the feature defines the build-webkit script has written. If the build
3198         is not done through build-webkit (and there's no file in build directory listing all the
3199         feature defines) the script does not advance further from copying.
3200
3201         This approach is taken to overcome the rigidness of the automake system as it's impossible
3202         to effectively generate and use GNUmakefile.features.am or even GNUmakefile.features file
3203         after the autoreconf command execution in autogen.sh.
3204
3205         * Scripts/build-jsc:
3206         (buildMyProject):
3207         * Scripts/build-webkit:
3208         * Scripts/webkitdirs.pm:
3209         (runAutogenForAutotoolsProjectIfNecessary):
3210         (mustReRunAutogen):
3211         (buildAutotoolsProject):
3212         (buildGtkProject):
3213         * Scripts/webkitperl/FeatureList.pm: Update the features that are currently enabled in
3214         Source/WebCore/GNUmakefile.features.am.in but aren't in the feature list.
3215         * Scripts/webkitpy/style/checker.py: Source/WebCore/GNUmakefile.features.am.in is recognized
3216         as a text file and tabulation errors are reported. Skip the file to suppress them.
3217         * gtk/override-feature-defines: Added.
3218         (copy_feature_defines_makefile):
3219         (adjust_feature_defines_makefile):
3220         (adjust_feature_defines_makefile.override_feature_define):
3221         (override_feature_defines):
3222
3223 2012-12-10  Yury Semikhatsky  <yurys@chromium.org>
3224
3225         Memory instrumentation: make sure each edge is reported only once
3226         https://bugs.webkit.org/show_bug.cgi?id=104630
3227
3228         Reviewed by Pavel Feldman.
3229
3230         Test that outgoing edges are reported only once if we come to the same
3231         object twice: first by an address of a base class and then by a real address
3232         of the object.
3233
3234         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3235
3236 2012-12-10  Dana Jansens  <danakj@chromium.org>
3237
3238         [chromium] Remove WebCompositorOutputSurface implementations from DRT
3239         https://bugs.webkit.org/show_bug.cgi?id=104046
3240
3241         Reviewed by Darin Fisher.
3242
3243         The classes are now provided through WebCompositorSupport.
3244
3245         * DumpRenderTree/DumpRenderTree.gypi:
3246         * DumpRenderTree/chromium/WebViewHost.cpp:
3247         (WebViewHost::createOutputSurface):
3248         * DumpRenderTree/chromium/WebViewHostOutputSurface.cpp: Removed.
3249         * DumpRenderTree/chromium/WebViewHostOutputSurface.h: Removed.
3250         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.cpp: Removed.
3251         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.h: Removed.
3252
3253 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3254
3255         change --no-launch-safari to --no-show-results in scripts
3256         https://bugs.webkit.org/show_bug.cgi?id=104298
3257
3258         Reviewed by Eric Seidel.
3259
3260         Since Safari only exists on the Apple ports, NRWT calls the
3261         command line flag --[no-]show-results instead. This updates
3262         the scripts to use the new flag (but not the buildbot config yet).
3263
3264         * Scripts/generate-coverage-data:
3265         * Scripts/webkit-tools-completion.sh:
3266         * Scripts/webkitpy/tool/steps/runtests.py:
3267         (RunTests.run):
3268         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
3269
3270 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3271
3272         nrwt: fix mapping of AUDIO onto FAIL when printing unexpected results
3273         https://bugs.webkit.org/show_bug.cgi?id=104597
3274
3275         Reviewed by Ojan Vafai.
3276
3277         This was broken in the recent printer refactoring.
3278
3279         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
3280         (BuildBotPrinter.print_unexpected_results.add_result.is_expected):
3281         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3282         (BuildBotPrinterTests.get_result):
3283         (BuildBotPrinterTests.test_print_unexpected_results):
3284         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
3285         (BuildBotPrinterTests):
3286
3287 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3288
3289         old-run-webkit-tests: add --show-results properly
3290         https://bugs.webkit.org/show_bug.cgi?id=104608
3291
3292         Reviewed by Eric Seidel.
3293
3294         Rework r137219 / bug 104601 to use the correct perl idiom.
3295
3296         * Scripts/old-run-webkit-tests:
3297
3298 2012-12-10  Aaron Colwell  <acolwell@chromium.org>
3299
3300         Unreviewed. Adding Aaron Colwell to the contributor list.
3301
3302         * Scripts/webkitpy/common/config/committers.py:
3303
3304 2012-12-10  Anders Carlsson  <andersca@apple.com>
3305
3306         Remove WebKit2 tests on Windows
3307         https://bugs.webkit.org/show_bug.cgi?id=104607
3308
3309         Reviewed by Tim Horton.
3310
3311         Stop building WebKit2 tests and remove associated files.
3312
3313         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Removed.
3314         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Removed.
3315         * TestWebKitAPI/win/PlatformWebViewWin.cpp: Removed.
3316         * TestWebKitAPI/win/TestWebKitAPI.sln: Removed.
3317         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
3318         * TestWebKitAPI/win/TestWebKitAPIGenerated.vcproj: Removed.
3319         * TestWebKitAPI/win/TestWebKitAPIGeneratedCommon.vsprops: Removed.
3320         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Removed.
3321         * TestWebKitAPI/win/WindowMessageObserver.h: Removed.
3322         * TestWebKitAPI/win/copy-resources.cmd: Removed.
3323
3324 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3325
3326         old-run-webkit-tests: support --no-show-results for compatibility w/ NRWT
3327         https://bugs.webkit.org/show_bug.cgi?id=104601
3328
3329         Reviewed by Eric Siedel.
3330
3331         This flag is being added so we can move all of the calling scripts
3332         over to using --no-show-results instead of --no-launch-safari.
3333
3334         * Scripts/old-run-webkit-tests:
3335
3336 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3337
3338         nrwt: use the default logic for skipping platform-specific tests for chromium
3339         https://bugs.webkit.org/show_bug.cgi?id=104403
3340
3341         Reviewed by Stephen White.
3342
3343         I thought I had fixed this a while ago, but the chromium port was
3344         using custom logic to figure out which platform/* tests to run (or
3345         skip) and as a result we were getting some odd behavior (like running
3346         tests in chromium-linux on chromium-mac, but only when there was
3347         an expectation listed for the test).
3348
3349         Ports are supposed to only run the tests found under their
3350         baseline_search_path() by default.
3351
3352         Not writing a test for this; theoretically there shouldn't been a
3353         test checking that the behavior for Chromium was *different* than
3354         the other ports before.
3355
3356         * Scripts/webkitpy/layout_tests/port/chromium.py:
3357         (ChromiumPort._missing_symbol_to_skipped_tests):
3358
3359 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3360
3361         fix regression in test-webkitpy after r137189
3362         https://bugs.webkit.org/show_bug.cgi?id=104589
3363
3364         Reviewed by Eric Seidel.
3365
3366         Need to specify full port names.
3367
3368         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3369         (MainTest.test_build_check):
3370
3371 2012-12-10  Laszlo Gombos  <l.gombos@samsung.com>
3372
3373         [EFL] Change the minimum required EFL version to 1.6
3374         https://bugs.webkit.org/show_bug.cgi?id=104431
3375
3376         Reviewed by Kenneth Rohde Christiansen.
3377
3378         Change the minimum required EFL version to 1.6 from 1.7.
3379         The elementary EFL package is only required to build MiniBrowser,
3380         so I moved the required only to the Minibrowser CMake file.
3381
3382         * MiniBrowser/efl/CMakeLists.txt: In addition removed
3383         ENABLE_GLIB_SUPPORT as it is obsolete and it was always set to be on.
3384
3385 2012-12-10  Dean Jackson  <dino@apple.com>
3386
3387         Unreviewed. Adding Antoine to the contributor list so his name autocompletes in bugzilla.
3388
3389         * Scripts/webkitpy/common/config/committers.py:
3390
3391 2012-12-10  Zan Dobersek  <zandobersek@gmail.com>
3392
3393         REGRESSION (r118735): svg/dom/complex-svgView-specification.html, svg/dom/SVGViewSpec.html, svg/dom/viewspec-parser.html failing on GTK Linux 64-bit Release
3394         https://bugs.webkit.org/show_bug.cgi?id=87734
3395
3396         Reviewed by Martin Robinson.
3397
3398         When willSendRequest should return null in testing, cancel the SoupMessage
3399         instead of setting the request's URI to about:blank.
3400
3401         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3402         (willSendRequestCallback):
3403
3404 2012-12-10  Jochen Eisinger  <jochen@chromium.org>
3405
3406         [chromium] Move the test implementation of WebKit::WebPlugin to the TestRunner library
3407         https://bugs.webkit.org/show_bug.cgi?id=104537
3408
3409         Reviewed by Tony Chang.
3410
3411         In order to hide implementation details from the public TestRunner API,
3412         I moved the plugin to WebTestPluginImpl and added a static create
3413         method to WebTestPlugin. Also, I renamed the plugin from TestWebPlugin
3414         to WebTestPlugin to be more in line with other classes in the
3415         TestRunner API.
3416
3417         * DumpRenderTree/DumpRenderTree.gypi:
3418         * DumpRenderTree/chromium/TestRunner/public/WebTestPlugin.h: Added.
3419         (WebTestRunner):
3420         (WebTestPlugin):
3421         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Renamed from Tools/DumpRenderTree/chromium/TestWebPlugin.cpp.
3422         (WebTestRunner::WebTestPlugin::create):
3423         (WebTestRunner):
3424         (WebTestRunner::WebTestPlugin::WebTestPlugin):
3425         (WebTestRunner::WebTestPlugin::~WebTestPlugin):
3426         (WebTestRunner::WebTestPlugin::mimeType):
3427         * DumpRenderTree/chromium/TestWebPlugin.h: Removed.
3428         * DumpRenderTree/chromium/WebViewHost.cpp:
3429         (WebViewHost::createPlugin):
3430
3431 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3432
3433         run_webkit_tests throws exception when DRT is not present
3434         https://bugs.webkit.org/show_bug.cgi?id=104565
3435
3436         Reviewed by Jochen Eisinger.
3437
3438         Don't actually try to print the bot results if there aren't
3439         any because we exited before running any tests.
3440
3441         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3442         (main):
3443         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3444         (MainTest.test_build_check):
3445
3446 2012-12-10  Benjamin Poulain  <benjamin@webkit.org>
3447
3448         Add convenience methods to use ListHashSet for a LRU cache
3449         https://bugs.webkit.org/show_bug.cgi?id=104499
3450
3451         Reviewed by Sam Weinig.
3452
3453         Test the new methods added to ListHashSet.
3454
3455         * TestWebKitAPI/CMakeLists.txt:
3456         * TestWebKitAPI/GNUmakefile.am:
3457         * TestWebKitAPI/TestWebKitAPI.gypi:
3458         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3459         * TestWebKitAPI/Tests/WTF/ListHashSet.cpp: Added.
3460         (TestWebKitAPI):
3461         (TestWebKitAPI::TEST):
3462
3463 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3464
3465         webkit-patch print-expectations doesn't work right for platforms w/ shared expectations
3466         https://bugs.webkit.org/show_bug.cgi?id=104405
3467
3468         Reviewed by Eric Seidel.
3469
3470         If multiple ports shared a TestExpectations file, webkit-patch
3471         print-expectations would use the expectation for the first port
3472         using that file and print the result for all ports, rather than
3473         re-parse and print each port separately (e.g., we might print
3474         the chromium-linux expectation for a file for both chromium-linux
3475         and chromium-mac). Discovered while testing/debugging the fix
3476         for bug 104403.
3477
3478         * Scripts/webkitpy/layout_tests/port/test.py:
3479         (TestPort._skipped_tests_for_unsupported_features):
3480         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3481         (lint):
3482         * Scripts/webkitpy/tool/commands/queries.py:
3483         (PrintExpectations._model):
3484         * Scripts/webkitpy/tool/commands/queries_unittest.py:
3485         (PrintExpectationsTest.test_platform):
3486
3487 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3488
3489         nrwt: fix total number of tests computed in the result summary
3490         https://bugs.webkit.org/show_bug.cgi?id=104061
3491
3492         Reviewed by Jochen Eisinger.
3493
3494         Good thing I left this open, because I actually broke it again
3495         the same way in a merge. Fix the computation of the one_line_summary
3496         again, and add tests for it (also clean up some redundant tests).
3497
3498         Also, add constants into the test port code so that the knowledge
3499         of how many tests should fail and how is next to the list of tests,
3500         rather than a totally different place.
3501
3502         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
3503         (LayoutTestRunner.run_tests):
3504         * Scripts/webkitpy/layout_tests/port/test.py:
3505         (TestList.__getitem__):
3506         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3507         (get_test_results):
3508         (MainTest.setUp):
3509         (MainTest.test_basic):
3510         (MainTest.test_run_singly_actually_runs_tests):
3511         (MainTest.test_test_list_with_prefix):
3512         (MainTest.test_missing_and_unexpected_results_with_custom_exit_code.CustomExitCodePort):
3513         (MainTest.test_tolerance.ImageDiffTestPort):
3514         (EndToEndTest.parse_full_results):
3515
3516 2012-12-10  Mike West  <mkwst@chromium.org>
3517
3518         Log to console when script is blocked by sandbox attributes.
3519         https://bugs.webkit.org/show_bug.cgi?id=104365
3520
3521         Reviewed by Ojan Vafai.
3522
3523         * Scripts/webkitpy/performance_tests/perftest.py:
3524         (PerfTest):
3525             Ignoring the new console message in perftests, as it broke parsing
3526             for the Parser/html-parser.html microbenchmark.
3527
3528 2012-12-10  Filip Pizlo  <fpizlo@apple.com>
3529
3530         JSC profiling and debug dump code should use inferred names when possible
3531         https://bugs.webkit.org/show_bug.cgi?id=104519
3532
3533         Reviewed by Oliver Hunt.
3534
3535         The format I'm using for referring to a code block is now name#hash. For example,
3536         v8-crypto has something called bnpSquareTo#B5QFbU. The profiler allows you to use
3537         either the hash, the inferred name, or the combined hash and full name when referring
3538         to blocks.
3539
3540         * Scripts/display-profiler-output:
3541
3542 2012-12-09  Filip Pizlo  <fpizlo@apple.com>
3543
3544         Profiler should say things about OSR exits
3545         https://bugs.webkit.org/show_bug.cgi?id=104497
3546
3547         Reviewed by Oliver Hunt.
3548
3549         Adds support for displaying OSR exit information for full summary (just displays the
3550         counts and the number of recompilations), bytecode display (says which bytecodes
3551         exited), and DFG display (annotates disassembly with taken OSR exits and their
3552         counts).
3553
3554         * Scripts/display-profiler-output:
3555
3556 2012-12-10  Alexis Menard  <alexis@webkit.org>
3557
3558         [CSS3 Backgrounds and Borders] Remove CSS3_BACKGROUND feature flag.
3559         https://bugs.webkit.org/show_bug.cgi?id=104539
3560
3561         Reviewed by Antonio Gomes.
3562
3563         As discussed on webkit-dev it is not needed to keep this feature flag
3564         as support for <position> type is a small feature that is already
3565         implemented by three other UAs. It was useful while landing this
3566         feature as partial bits were landed one after one.
3567
3568         * Scripts/webkitperl/FeatureList.pm:
3569         * qmake/mkspecs/features/features.pri:
3570
3571 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3572
3573         Build fix.
3574
3575         Add a patchfile missing from my previous commit.
3576
3577         * gtk/patches/harfbuzz-icu-detection-fix.patch: Added.
3578
3579 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3580
3581         REGRESSION(137127): Causes assertion failures on the debug bots
3582         https://bugs.webkit.org/show_bug.cgi?id=104549
3583
3584         Reviewed by Alejandro G. Castro.
3585
3586         After handling the new TextOutput message we should return the method
3587         handling messages received.
3588
3589         * WebKitTestRunner/TestInvocation.cpp:
3590         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Fix a missing early return.
3591
3592 2012-12-09  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
3593
3594         [Soup] utilize multipart/x-mixed-replace support recently added to libsoup
3595         https://bugs.webkit.org/show_bug.cgi?id=94515
3596
3597         Reviewed by Martin Robinson.
3598
3599         This change requires a very recent libsoup, so update our jhbuild
3600         requirements.
3601
3602         * gtk/jhbuild.modules: depend on soup version that adds the multipart/x-mixed-replace
3603         support
3604         * efl/jhbuild.modules: ditto
3605
3606 2012-12-10  Zoltan Nyul  <zoltan.nyul@intel.com>
3607
3608         Implement testRunner.setViewModeMediaFeature() in WebKitTestRunner
3609         https://bugs.webkit.org/show_bug.cgi?id=103886
3610
3611         Reviewed by Kenneth Rohde Christiansen.
3612
3613         Implement testRunner.setViewModeMediaFeature() in WebKitTestRunner.
3614         This API is required by layout tests.
3615
3616         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3617         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3618         (WTR::TestRunner::setViewModeMediaFeature):
3619         (WTR):
3620         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3621         (TestRunner):
3622
3623 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3624
3625         [GTK] Bring Harfbuzz-ng support to Gtk
3626         https://bugs.webkit.org/show_bug.cgi?id=92098
3627
3628         Reviewed by Gustavo Noronha Silva.
3629
3630         Add HarfBuzz dependency.
3631
3632         * GNUmakefile.am: Add FreeType/HarfBuzz CFLAGS to the WebCore Internals library.
3633         * gtk/jhbuild.modules: Add HarfBuzz to the JHbuild and remove Pango, since it
3634         should no longer affect layout test results.
3635
3636 2012-12-10  Simon Hausmann  <simon.hausmann@digia.com>
3637
3638         [Qt] Use QLibraryInfo::LibraryExecutablesPath unconditionally
3639         https://bugs.webkit.org/show_bug.cgi?id=104541
3640
3641         Reviewed by Jocelyn Turcotte.
3642
3643         We now depend on a Qt 5 version that is guaranteed to have this API, so we
3644         can remove the configure checks for it.