git-add-reviewer: Do not put the bug title and its URL in the same line in the commit...
[WebKit-https.git] / Tools / ChangeLog
1 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
2
3         git-add-reviewer: Do not put the bug title and its URL in the same line in the commit message.
4         https://bugs.webkit.org/show_bug.cgi?id=85570
5
6         Reviewed by Adam Roben.
7
8         The standard format of commit messages and ChangeLog entries has a
9         short summary in the first line, followed by the bug URL in
10         Bugzilla. Passing "%s" in the pretty format line to git-rev-list
11         merges these two lines into a single one separated by a space
12         character, which is not what we want.
13
14         Since "%s%n%n%b" is in practice "%B" (raw message), use it
15         instead, as the first two lines do not end up being merged
16         together.
17
18         * Scripts/git-add-reviewer:
19         (writeCommitMessageToFile):
20
21 2012-05-03  Ojan Vafai  <ojan@chromium.org>
22
23         Improve UI for garden-o-matic examine view when there are fewer than 5 tests.
24         https://bugs.webkit.org/show_bug.cgi?id=85566
25
26         Reviewed by Adam Barth.
27
28         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
29         (.):
30         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
31         (.):
32         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
33
34 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
35
36         webkitpy: Recognize FreeBSD as a valid platform.
37         https://bugs.webkit.org/show_bug.cgi?id=81467
38
39         Reviewed by Adam Barth.
40
41         Add PlatformInfo.is_freebsd() so that trying to use webkit-patch
42         on FreeBSD does not raise an "unrecognized platform" exception.
43
44         Most of this patch contains some unit tests for the mentioned
45         method, as currently platform-specific code is needed only on
46         Windows.
47
48         * Scripts/webkitpy/common/system/platforminfo.py:
49         (PlatformInfo.__init__):
50         (PlatformInfo.is_freebsd):
51         (PlatformInfo._determine_os_name):
52         * Scripts/webkitpy/common/system/platforminfo_mock.py:
53         (MockPlatformInfo.is_freebsd):
54         * Scripts/webkitpy/common/system/platforminfo_unittest.py:
55         (fake_platform):
56         (fake_platform.FakePlatformModule.release):
57         (TestPlatformInfo.test_real_code):
58         (TestPlatformInfo.test_os_name_and_wrappers):
59         (TestPlatformInfo.test_os_version):
60         (TestPlatformInfo.test_display_name):
61         (TestPlatformInfo.test_total_bytes_memory):
62         (TestPlatformInfo.test_free_bytes_memory):
63         * Scripts/webkitpy/layout_tests/port/factory.py:
64         (PortFactory._default_port):
65
66 2012-05-03  Ojan Vafai  <ojan@chromium.org>
67
68         Make the table headers white-space:nowrap. This decreases the vertical size,
69         which makes for more minimal embedding in garden-o-matic.
70
71         * TestResultServer/static-dashboards/flakiness_dashboard_tests.css:
72         (.table-header-content *):
73
74 2012-05-03  Eric Seidel  <eric@webkit.org>
75
76         Split build-webkit's feature option list off into its own module in preparation for autogeneration
77         https://bugs.webkit.org/show_bug.cgi?id=85548
78
79         Reviewed by Daniel Bates.
80
81         My Perl-fu is very weak.  This moves the feature option list off into a separate module
82         file so that I can easily autogenerate just that file with the fancy new generate-feature-files command.
83         It's a bit odd to have an array which we return out to build-webkit to modify, but it turns out
84         not to matter.  Also, build-webkit doesn't actually use any of the variables we're using to
85         store the results from the option parse, so those end up inaccessible from build-webkit (except by reference)
86         which is actually cleaner.  Callers can't assume that they would get a new array every time, but someone
87         with more Perl-fu than I could create something fancier than this if we need in the future.
88
89         * Scripts/build-webkit:
90         * Scripts/webkitperl/FeatureList.pm: Added.
91         (getFeatureOptionList):
92
93 2012-05-03  Dirk Pranke  <dpranke@chromium.org>
94
95         REGRESSION: run-webkit-tests failing on Chromium Mac
96         https://bugs.webkit.org/show_bug.cgi?id=85459
97
98         Unreviewed, build fix.
99
100         Turns out select doesn't throw IOErrors, it has it's own kind of
101         error :(. Trap that instead.
102
103         * Scripts/webkitpy/layout_tests/port/server_process.py:
104         (ServerProcess._wait_for_data_and_update_buffers_using_select):
105
106 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
107
108         [CMake] Rewrite FindCairo.cmake.
109         https://bugs.webkit.org/show_bug.cgi?id=84895
110
111         Reviewed by Daniel Bates.
112
113         The old approach relied on pkg-config for finding Cairo (which
114         introduced a dependency on pkg-config that could be avoided), used
115         the LibFindMacros code that we should probably remove in the
116         future and did not use the FindPackageHandleStandardArguments
117         module.
118
119         Change all that by rewriting the module.
120         - Use the pkg-config output optionally instead of requiring it
121         like LibFindMacros did.
122         - Remove the implicit dependency on FreeType which often found it
123         the wrong way via pkg-config and without considering
124         CMAKE_PREFIX_PATH.
125         - Retrieve the Cairo version by looking at cairo-version.h instead
126         of relying on pkg-config. It requires some additional code for
127         checking if the desired version has been found, but that will not
128         be needed once we start depending on CMake 2.8.3 or later.
129
130         The only downside is that FPHSA sets <UPPERCASED_NAME>_FOUND
131         instead of <Name>_FOUND, and to keep things consistent
132         Cairo_LIBRARIES and Cairo_INCLUDE_DIRS have become CAIRO_LIBRARIES
133         and CAIRO_INCLUDE_DIRS.
134
135         * DumpRenderTree/efl/CMakeLists.txt: Use CAIRO_FOO instead of
136         Cairo_FOO.
137         * EWebLauncher/CMakeLists.txt: Ditto.
138
139 2012-05-03  Ojan Vafai  <ojan@chromium.org>
140
141         Fix modifier and bug sorting on the flakiness dashboard
142         https://bugs.webkit.org/show_bug.cgi?id=85554
143
144         Reviewed by Dirk Pranke.
145
146         The sorting assumes that the column header name matches a field
147         in the testResults object and uses that field for sorting.
148         Filter the bugs out of the modifiers list and put them in a bugs field
149         and only put the non-bug modifiers in the modifiers list.
150
151         * TestResultServer/static-dashboards/flakiness_dashboard.js:
152         (populateExpectationsData):
153         (htmlForSingleTestRow):
154         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
155
156 2012-05-03  Ojan Vafai  <ojan@chromium.org>
157
158         Fix some state resetting when running unittests. HtmlForSingleTestRow
159         was flaky.
160
161         * TestResultServer/static-dashboards/flakiness_dashboard.js:
162         (htmlForSingleTestRow):
163         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
164         (resetGlobals):
165
166 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
167
168         webkitpy: Use PlatformInfo wherever possible in the User class.
169         https://bugs.webkit.org/show_bug.cgi?id=85549
170
171         Reviewed by Dirk Pranke.
172
173         Instead of relying on sys.platform to detect whether the current
174         platform is a Mac, use the PlatformInfo class instead.
175
176         * Scripts/webkitpy/common/system/user.py:
177         (User.__init__):
178         (User.edit_changelog):
179
180 2012-05-03  Ojan Vafai  <ojan@chromium.org>
181
182         Remove the percent failed column from the flakiness dashboard
183         https://bugs.webkit.org/show_bug.cgi?id=85552
184
185         Reviewed by Dirk Pranke.
186
187         This column isn't useful enough to justify the visual noise.
188
189         * TestResultServer/static-dashboards/flakiness_dashboard.js:
190         (processMissingAndExtraExpectations):
191         (tableHeaders):
192         (htmlForSingleTestRow):
193         (sortTests):
194         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
195         (test):
196
197 2012-05-03  Eric Seidel  <eric@webkit.org>
198
199         --coverage should not not use the "feature" options system in build-webkit, similarly remove special-casing for WTF_URL
200         https://bugs.webkit.org/show_bug.cgi?id=85542
201
202         Reviewed by Daniel Bates.
203
204         Remove these last two odd-men-out, so that we can generate the default feature options
205         for build-webkit using the new generate-feature-files command.
206         The Mac build system has some special casing for ENABLE_ defines which gets them
207         defined as -D on the build command, even though initially they start out as
208         environment variables ENABLE_FOO = ENABLE_FOO.  The Feature options in build-webkit
209         depend on this support, but it doesn't work for WTF_USE_ defines like WTF_USE_WTF_URL.
210         I'm told that WTF_URL is not actively being worked on, so just removing it for now.
211         It should probably come back to life as an ENABLE_, or build-webkit or the Xcode projects
212         should be made generically smarter about these USE_WTF defines.
213         (There is still WTF_USE_TILED_BACKING_STORE in the feature list, but it only seems
214         used by Qt.  If someone tried build-webkit --tiled-backing-store on Mac today it
215         would not work!)
216
217         * Scripts/build-webkit:
218
219 2012-05-03  Eric Seidel  <eric@webkit.org>
220
221         Remove the last bits of Leopard support from our perl scripts
222         https://bugs.webkit.org/show_bug.cgi?id=85545
223
224         Reviewed by Daniel Bates.
225
226         Chromium still supports Leopard (for a little while yet), but AppleMacWebKit dropped
227         Leopard support several months ago.  Remove the last bits of isLeopard() branching
228         in our Perl scripts, since no one is using this (Chromium has never used this Perl code).
229
230         * Scripts/build-webkit:
231         * Scripts/old-run-webkit-tests:
232         (countAndPrintLeaks):
233         (captureSavedCrashLog):
234         * Scripts/webkitdirs.pm:
235         (determineArchitecture):
236         (argumentsForRunAndDebugMacWebKitApp):
237
238 2012-05-03  Ojan Vafai  <ojan@chromium.org>
239
240         [Chromium] Don't show the WebKit Linux ASAN bot on the flakiness dashboard.
241         https://bugs.webkit.org/show_bug.cgi?id=85551
242
243         Reviewed by Dirk Pranke.
244
245         It's not uploading results and in the short term we don't plan to have it start.
246
247         * TestResultServer/static-dashboards/builders.js:
248         (isChromiumWebkitTipOfTreeTestRunner):
249         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
250
251 2012-05-03  Ojan Vafai  <ojan@chromium.org>
252
253         Show the test name in the dashboard if showChrome==false and we're showing multiple tests
254         https://bugs.webkit.org/show_bug.cgi?id=85243
255
256         Reviewed by Dirk Pranke.
257
258         This happens for cases where a test name matches multiple tests.
259         In practice, this only happens for virtual test suites (e.g. the canvas tests).
260
261         * TestResultServer/static-dashboards/flakiness_dashboard.js:
262         (htmlForIndividualTestOnAllBuildersWithChrome):
263         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
264         (testHtmlForIndividualTestOnAllBuildersWithChromeNonexistant):
265         (testHtmlForIndividualTestOnAllBuildersWithChrome):
266         (testHtmlForIndividualTestOnAllBuildersWithChromeWebkitMaster):
267         (testHtmlForIndividualTests):
268
269 2012-05-03  Ojan Vafai  <ojan@chromium.org>
270
271         Port the flakiness dashboard unittests to qunit
272         https://bugs.webkit.org/show_bug.cgi?id=85544
273
274         Reviewed by Adam Barth.
275
276         * TestResultServer/static-dashboards/dashboard_base.js:
277         (appendJSONScriptElements):
278         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
279         (resetGlobals):
280         (runExpectationsTest):
281         (test):
282         * TestResultServer/static-dashboards/run-unittests.html: Added.
283
284 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
285
286         webkitpy: Remove check for readline on Mac.
287         https://bugs.webkit.org/show_bug.cgi?id=85547
288
289         Reviewed by Eric Seidel.
290
291         Python 2.5 is not supported by the WebKit project anymore, and
292         Python's readline documentation does not mention the module not
293         being available in the supported versions.
294
295         * Scripts/webkitpy/common/system/user.py:
296
297 2012-05-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
298
299         [NRWT] Run performance tests with lock
300         https://bugs.webkit.org/show_bug.cgi?id=78628
301
302         Reviewed by Tony Chang.
303
304         Locking performance tests (like we do for http tests) will force them
305         to run in serial. This reduces the load of the machine when running perf
306         tests and minimizes the chances of the tests to fail, specially
307         timeouts.
308
309         * Scripts/webkitpy/layout_tests/controllers/manager.py:
310         (Manager.__init__):
311         (Manager._is_perf_test):
312         (Manager._test_requires_lock):
313         * Scripts/webkitpy/layout_tests/port/http_lock.py:
314
315 2012-05-03  Dirk Pranke  <dpranke@chromium.org>
316
317         Attempt to fix a crash on SL ...
318
319         Unreviewed, build fix.
320
321         It looks like we may be getting an IOError raised during
322         select() on chromium mac, and that error isn't being caught.
323         Pushing it into the try/catch block will hopefully catch this
324         and allow us to see why the subprocess is crashing.
325
326         * Scripts/webkitpy/layout_tests/port/server_process.py:
327         (ServerProcess._wait_for_data_and_update_buffers_using_select):
328
329 2012-05-02  Jon Honeycutt  <jhoneycutt@apple.com>
330
331         WebFrameLoaderClient::dispatchWillSendSubmitEvent() needs to be
332         implemented for WebKit2
333         https://bugs.webkit.org/show_bug.cgi?id=84304
334
335         Reviewed by Jessie Berlin.
336
337         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
338         Added new files to project.
339
340         * TestWebKitAPI/Tests/WebKit2/WillSendSubmitEvent.cpp: Added.
341         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
342         Look for the "DidReceiveWillSendSubmitEvent" message. The message body
343         will be the dictionary of form control names and values passed to the
344         injected bundle's willSendSubmitEvent handler; look for the pairs
345         "textField" -> "text field" and "passwordField" -> "password field".
346         Check that "hiddenField" is not included.
347         (TestWebKitAPI::setInjectedBundleClient):
348         (TestWebKitAPI::TEST):
349
350         * TestWebKitAPI/Tests/WebKit2/WillSendSubmitEvent_Bundle.cpp: Added.
351         (TestWebKitAPI::willSendSubmitEvent):
352         Post a message to the TestWebKitAPI process with the dictionary of form
353         control names and values that we received.
354         (TestWebKitAPI::WillSendSubmitEventTest::WillSendSubmitEventTest):
355         (TestWebKitAPI::WillSendSubmitEventTest::didCreatePage):
356         Set up the page's form client.
357
358         * TestWebKitAPI/Tests/WebKit2/auto-submitting-form.html: Added.
359
360 2012-05-03  Eric Seidel  <eric@webkit.org>
361
362         Remove dead code from build-webkit for working around old XCode dependency bug
363         https://bugs.webkit.org/show_bug.cgi?id=85535
364
365         Reviewed by Adam Barth.
366
367         Xcode used to have a bug whereby it didn't notice if you changed the #defines
368         passed on the command line to xcodebuild.  I wrote code to work around this
369         while bringing up SVG support 5-6 years ago.  I think we can finally remove this
370         code from build-webkit, since SVG is the only feature flag to ever use it
371         (modern in-progress features like MathML don't seem to be using the support)
372         an I suspect the XCode bug has long been fixed.
373
374         * Scripts/build-webkit:
375         * Scripts/webkitperl/features.pm:
376
377 2012-05-03  Andy Estes  <aestes@apple.com>
378
379         InjectedBundleControllerMac.mm needs to include config.h
380         https://bugs.webkit.org/show_bug.cgi?id=85531
381
382         Reviewed by Andreas Kling.
383
384         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm:
385
386 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
387
388         webkitpy: Use PlatformInfo whenever possible in server_process.py.
389         https://bugs.webkit.org/show_bug.cgi?id=85516
390
391         Reviewed by Eric Seidel.
392
393         Eliminate direct access to sys.platform wherever possible by using
394         PlatformInfo.is_win() instead.
395
396         * Scripts/webkitpy/layout_tests/port/server_process.py:
397         (ServerProcess.__init__): Remove the executive parameter as it was
398         not passed by any caller, obtain a SystemHost from port_obj
399         instead.
400         (ServerProcess._log): Style, remove extra empty line.
401         (ServerProcess._start):
402         (ServerProcess.stop):
403         (ServerProcess.kill):
404         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
405         (TrivialMockPort.__init__): Create a MockSystemHost used by
406         ServerProcess.
407         (TestServerProcess.test_broken_pipe): Test different platforms.
408
409 2012-05-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
410
411         webkitpy: Style checker broken when committing a PNG
412         https://bugs.webkit.org/show_bug.cgi?id=85504
413
414         Reviewed by Dirk Pranke.
415
416         * Scripts/webkitpy/style/checkers/png.py:
417         (PNGChecker.__init__):
418
419 2012-05-03  Philippe Normand  <pnormand@igalia.com>
420
421         Unreviewed, fix another call to join(), similar to the ones landed
422         in r115975.
423
424         * Scripts/webkitdirs.pm:
425         (jhbuildConfigurationChanged):
426
427 2012-05-03  Raphael Kubo da Costa  <rakuco@webkit.org>
428
429         webkitpy: Remove sys.platform accesses in CrashLogs.
430         https://bugs.webkit.org/show_bug.cgi?id=85436
431
432         Reviewed by Dirk Pranke.
433
434         Instead of directly checking for sys.platform == 'darwin', use a
435         SystemHost and call PlatformInfo.is_mac().
436
437         * Scripts/webkitpy/common/system/crashlogs.py:
438         (CrashLogs.__init__):
439         (CrashLogs.find_newest_log):
440         (CrashLogs._log_directory_darwin):
441         (CrashLogs._find_newest_log_darwin):
442         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
443         (CrashLogsTest.test_find_log_darwin):
444         * Scripts/webkitpy/layout_tests/port/mac.py:
445         (MacPort._get_crash_log):
446         * Scripts/webkitpy/layout_tests/port/test.py:
447         (TestDriver.run_test):
448         * Scripts/webkitpy/tool/commands/queries.py:
449         (execute):
450
451 2012-05-03  Philippe Normand  <pnormand@igalia.com>
452
453         [GTK] join() is wrongly used in webkitdirs.pm since r115532
454         https://bugs.webkit.org/show_bug.cgi?id=85501
455
456         Reviewed by Martin Robinson.
457
458         * Scripts/webkitdirs.pm:
459         (getJhbuildPath):
460         (buildAutotoolsProject):
461
462 2012-05-03  Zan Dobersek  <zandobersek@gmail.com>
463
464         [Gtk] ImageDiff crashes when it can't open a display
465         https://bugs.webkit.org/show_bug.cgi?id=85476
466
467         Reviewed by Martin Robinson.
468
469         Replace gdk_init() initialization call with the g_type_init()
470         call. GdkPixbufLoader doesn't actually need the current call to be
471         made and this change will also remove crashes when ImageDiff
472         is started in an environment without a DISPLAY env.
473
474         * DumpRenderTree/gtk/ImageDiff.cpp:
475         (main):
476
477 2012-05-03  Zan Dobersek  <zandobersek@gmail.com>
478
479         ImageDiff should be run inside a properly established environment
480         https://bugs.webkit.org/show_bug.cgi?id=85292
481
482         Reviewed by Martin Robinson.
483
484         When creating the ImageDiff server process, pass along a freshly-set-up environment
485         in which the process should be executed in.
486
487         * Scripts/webkitpy/layout_tests/port/webkit.py:
488         (WebKitPort._start_image_diff_process):
489
490 2012-05-03  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
491
492         [Qt] Allow the web process and WTR to be paused on startup
493
494         Makes it easier to debug the web process or run-webkit-tests -2, as you
495         have ample time to attach gdb to the process.
496
497         Reviewed by Simon Hausmann.
498
499         * Scripts/webkitpy/layout_tests/port/qt.py:
500         (QtPort.setup_environ_for_server):
501         * WebKitTestRunner/qt/main.cpp:
502         (main):
503
504 2012-05-03  Csaba Osztrogonác  <ossy@webkit.org>
505
506         Password protect "Stop build" button
507         https://bugs.webkit.org/show_bug.cgi?id=81982
508
509         Reviewed by Ryosuke Niwa.
510
511         Only authenticated users should be able to use stopBuild, stopAllBuilds, cancelPendingBuild
512         features to prevent killing builds on build.webkit.org by SPAM bots.
513
514         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
515
516 2012-05-03  Dongwoo Im  <dw.im@samsung.com>
517
518         [DRT] Remove the name of parameter, visibility, from the setPageVisibility stub implementations.
519         https://bugs.webkit.org/show_bug.cgi?id=85468
520
521         Reviewed by Nikolas Zimmermann.
522
523         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
524         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
525         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
526         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
527         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
528         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
529         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
530         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
531         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
532         (LayoutTestController::setPageVisibility): Remove the name of parameter, visibility.
533
534 2012-05-03  Stephanie Lewis  <slewis@apple.com>
535
536         https://bugs.webkit.org/show_bug.cgi?id=85450 unbounded growth of JSDOMWindowShells loading pages in the same window
537         <rdar://problem/11320059> REGRESSION (r115083): PLT3 shows linear memory growth and gets slower with each run
538
539         Reviewed by Brady Eidson.
540
541         The API added for DOMWindowExtension, didCreateGlobalObjectForFrame, would create a global object
542         for every world, even those that did not need the callback.  This had the side effect of creating a
543         JSDOMWindowShell that the associated world didn't necessarily know to clean up.  Instead of creating
544         unnecessary objects change the API to globalObjectIsAvailableForFrame and do not pass the global object
545         in the API.  The object can be accessed later by those worlds which require it.
546
547         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic.cpp:
548         (TestWebKitAPI):
549         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
550         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
551         (TestWebKitAPI):
552         (DOMWindowExtensionBasic):
553         (TestWebKitAPI::DOMWindowExtensionBasic::didCreatePage):
554         (TestWebKitAPI::DOMWindowExtensionBasic::globalObjectIsAvailableForFrame):
555         (TestWebKitAPI::globalObjectIsAvailableForFrameCallback):
556
557 2012-05-03  Nikolas Zimmermann  <nzimmermann@rim.com>
558
559         Not reviewed. Fix mac build, it was missing setPageVisibility/resetPageVisibility stub implementations.
560
561         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
562         (LayoutTestController::resetPageVisibility):
563         (LayoutTestController::setPageVisibility):
564
565 2012-05-02  Alexander Færøy  <ahf@0x90.dk>
566
567         Rename deviceDPI to devicePixelRatio
568         https://bugs.webkit.org/show_bug.cgi?id=85049
569
570         Reviewed by Kenneth Rohde Christiansen.
571
572         Use 1.5 as device pixel ratio for the Qt MiniBrowser.
573
574         * MiniBrowser/qt/qml/BrowserWindow.qml:
575
576 2012-05-03  Dongwoo Im  <dw.im@samsung.com>
577
578         [EFL][DRT] Implement the LayoutTestController's methods related to the Page Visibility API.
579         https://bugs.webkit.org/show_bug.cgi?id=85347
580
581         Reviewed by Nikolas Zimmermann.
582
583         * DumpRenderTree/LayoutTestController.h:
584         (LayoutTestController):
585         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
586         (LayoutTestController::resetPageVisibility): Not implemented.
587         (LayoutTestController::setPageVisibility): Not implemented.
588         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
589         (LayoutTestController::resetPageVisibility): Reset the page visibility.
590         (LayoutTestController::setPageVisibility): Set the page visibility.
591         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
592         (LayoutTestController::resetPageVisibility): Not implemented.
593         (LayoutTestController::setPageVisibility): Not implemented.
594         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
595         (LayoutTestController::resetPageVisibility): Not implemented.
596         (LayoutTestController::setPageVisibility): Not implemented.
597         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
598         (LayoutTestController::resetPageVisibility): Not implemented.
599         (LayoutTestController::setPageVisibility): Not implemented.
600         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
601         (LayoutTestController::resetPageVisibility): Not implemented.
602         (LayoutTestController::setPageVisibility): Not implemented.
603
604 2012-05-02  Sheriff Bot  <webkit.review.bot@gmail.com>
605
606         Unreviewed, rolling out r115918.
607         http://trac.webkit.org/changeset/115918
608         https://bugs.webkit.org/show_bug.cgi?id=85452
609
610         "Broke test-webkitpy on Apple Lion bots" (Requested by estes
611         on #webkit).
612
613         * Scripts/webkitpy/common/system/crashlogs.py:
614         (CrashLogs.__init__):
615         (CrashLogs.find_newest_log):
616         (CrashLogs._log_directory_darwin):
617         (CrashLogs._find_newest_log_darwin):
618         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
619         (CrashLogsTest.test_find_log_darwin):
620         * Scripts/webkitpy/layout_tests/port/mac.py:
621         (MacPort._get_crash_log):
622         * Scripts/webkitpy/layout_tests/port/test.py:
623         (TestDriver.run_test):
624         * Scripts/webkitpy/tool/commands/queries.py:
625         (execute):
626
627 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
628
629         webkitpy: Remove sys.platform accesses in CrashLogs.
630         https://bugs.webkit.org/show_bug.cgi?id=85436
631
632         Reviewed by Dirk Pranke.
633
634         Instead of directly checking for sys.platform == 'darwin', use a
635         SystemHost and call PlatformInfo.is_mac().
636
637         * Scripts/webkitpy/common/system/crashlogs.py:
638         (CrashLogs.__init__):
639         (CrashLogs.find_newest_log):
640         (CrashLogs._log_directory_darwin):
641         (CrashLogs._find_newest_log_darwin):
642         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
643         (CrashLogsTest.test_find_log_darwin):
644         * Scripts/webkitpy/layout_tests/port/mac.py:
645         (MacPort._get_crash_log):
646         * Scripts/webkitpy/layout_tests/port/test.py:
647         (TestDriver.run_test):
648         * Scripts/webkitpy/tool/commands/queries.py:
649         (execute):
650
651 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
652
653         webkitpy: Special-case win32 calls in file_lock and executive.
654         https://bugs.webkit.org/show_bug.cgi?id=85431
655
656         Reviewed by Dirk Pranke.
657
658         Simplify accesses to sys.platform by only checking if sys.platform
659         is "win32" and removing explicit checks for the else case.
660
661         * Scripts/webkitpy/common/system/executive.py:
662         (Executive.check_running_pid):
663         * Scripts/webkitpy/common/system/file_lock.py:
664         (FileLock._create_lock):
665         (FileLock._remove_lock):
666
667 2012-05-02  Ojan Vafai  <ojan@chromium.org>
668
669         Remove pink from flakiness dashboard
670         https://bugs.webkit.org/show_bug.cgi?id=85422
671
672         Reviewed by Dirk Pranke.
673
674         The pink indicates that the test doesn't match it's listing in test_expectations.txt.
675         In practice, this just added noise and confused people.
676
677         * TestResultServer/static-dashboards/flakiness_dashboard.js:
678         (htmlForSingleTestRow):
679         (showLegend):
680         * TestResultServer/static-dashboards/flakiness_dashboard_tests.css:
681         (#legend-contents *):
682         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
683         (testHtmlForSingleTestRow):
684
685 2012-05-02  Gustavo Noronha Silva  <gns@gnome.org>
686
687         Unreviewed. Fix mistake I committed when applying review comments.
688
689         * jhbuild/jhbuild-wrapper: dependencies_path was getting 'Root'
690         appended to its path, so jhbuild itself was being cloned and
691         installed in the wrong place when using WEBKITOUTPUTDIR.
692
693 2012-05-02  Dirk Pranke  <dpranke@chromium.org>
694
695         nrwt: make ServerProcess work on chromium win
696         https://bugs.webkit.org/show_bug.cgi?id=85333
697
698         Reviewed by Ojan Vafai.
699
700         This change implements a poor man's select() that will
701         slow-spin doing non-blocking reads on the stdout and stderr
702         named pipes connecting the worker to the driver. Seems to work
703         and I have yet to see much overhead or ill effects but it
704         probably needs more testing.
705
706         * Scripts/webkitpy/layout_tests/port/chromium.py:
707         (ChromiumDriver.__init__):
708         * Scripts/webkitpy/layout_tests/port/server_process.py:
709         (ServerProcess._start):
710         (ServerProcess._wait_for_data_and_update_buffers_using_select):
711         (ServerProcess._wait_for_data_and_update_buffers_using_win32_apis):
712         (ServerProcess):
713         (ServerProcess._non_blocking_read_win32):
714         (ServerProcess._read):
715         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
716         (TestServerProcess.test_basic): Added.
717
718 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
719
720         webkitpy: Use a SystemHost instead of PlatformInfo in the PNG checker.
721         https://bugs.webkit.org/show_bug.cgi?id=85426
722
723         Reviewed by Dirk Pranke.
724
725         r115880 replaced the direct use of sys.platform with a
726         PlatformInfo object. Dirk Pranke later told me SystemHost objects
727         should be preferred, as nothing out of webkitpy.common.system
728         should create its own PlatformInfo instances.
729
730         * Scripts/webkitpy/common/system/systemhost_mock.py:
731         (MockSystemHost.__init__): Make it possible to specify a custom
732         filesystem object.
733         * Scripts/webkitpy/style/checkers/png.py:
734         (PNGChecker.__init__): Accept a host object instead of a
735         filesystem one, which is now obtained from host.
736         (PNGChecker._config_file_path): Retrieve PlatformInfo from
737         SystemHost.
738         * Scripts/webkitpy/style/checkers/png_unittest.py:
739         (PNGCheckerTest.test_init): Adjust calls to PNGChecker to its new
740         constructor signature.
741         (PNGCheckerTest.test_check): Ditto.
742
743 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
744
745         webkitpy: Remove direct checks for sys.platform == 'darwin' in the layout test classes.
746         https://bugs.webkit.org/show_bug.cgi?id=85424
747
748         Reviewed by Tony Chang.
749
750         We should reduce direct usage of sys.platform in webkitpy and use
751         PlatformInfo instead.
752
753         * Scripts/webkitpy/layout_tests/servers/http_server.py:
754         (Lighttpd._prepare_config): Call PlatformInfo.is_mac() instead of
755         checking for 'darwin' directly.
756         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
757         (HttpServerBase.__init__): Ditto.
758
759 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
760
761         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
762         https://bugs.webkit.org/show_bug.cgi?id=85419
763
764         Reviewed by Dirk Pranke.
765
766         As pointed out by Dirk Pranke, a PlatformInfo is not supposed to
767         be created directly, we should should a SystemHost instead.
768
769         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
770         (MainTest.setUp): Create a SystemHost and obtain the PlatformInfo
771         object from it instead of creating PlatformInfo directly.
772
773 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
774
775         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
776         https://bugs.webkit.org/show_bug.cgi?id=85419
777
778         Rubber-stamped by Dirk Pranke.
779
780         Follow-up patch to r115887 which fixes some capitalization problems
781         pointed out in comments 2 and 4.
782
783         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
784         (MainTest.setUp):
785         (MainTest.test_child_processes_2):
786         (MainTest.test_child_processes_min):
787         (MainTest.test_exception_raised):
788         (MainTest.test_keyboard_interrupt):
789
790 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
791
792         webkitpy: Remove direct sys.platform usage in run_webkit_tests_integrationtest.
793         https://bugs.webkit.org/show_bug.cgi?id=85419
794
795         Reviewed by Dirk Pranke.
796
797         Another part of the work towards replacing direct sys.platform
798         usage with PlatformInfo.
799
800         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
801         (MainTest.setUp): Move the setting of SHOULD_TEST_PROCESSES
802         here. Even though it is costlier, unittest.setUpClass was
803         introduced in Python 2.7.
804         (MainTest.test_child_processes_2): Use self.SHOULD_TEST_PROCESSES.
805         (MainTest.test_child_processes_min): Ditto.
806         (MainTest.test_exception_raised): Ditto.
807         (MainTest.test_keyboard_interrupt): Ditto.
808         (MainTest.test_crash_log): Check for PlatformInfo.is_mac().
809         (MainTest.test_web_process_crash_log): Ditto, and remove duplicate
810         check at the end.
811         (MainTest.test_no_http_tests): Remove check for (cygwin && Python
812         2.5.2), as we only support Python 2.6+.
813
814 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
815
816         webkitpy: Remove direct sys.platform usage in the PNG checker.
817         https://bugs.webkit.org/show_bug.cgi?id=85407
818
819         Reviewed by Tony Chang.
820
821         Work towards replacing direct sys.platform usage with
822         PlatformInfo.
823
824         * Scripts/webkitpy/common/system/platforminfo_mock.py:
825         (MockPlatformInfo.__init__): Accept os_name and os_version as
826         arguments.
827         * Scripts/webkitpy/style/checkers/png.py:
828         (PNGChecker.__init__): Create a PlatformInfo object instead of
829         using sys.platform when needed.
830         (PNGChecker._config_file_path): Check for the current platform via
831         PlatformInfo.
832         * Scripts/webkitpy/style/checkers/png_unittest.py:
833         (PNGCheckerTest.test_check): Update tests after the changes above.
834
835 2012-05-02  Raphael Kubo da Costa  <rakuco@webkit.org>
836
837         [watchlist] Update watchlist rules for EFL-related patches.
838         http://bugs.webkit.org/show_bug.cgi?id=85160
839
840         Reviewed by Eric Seidel.
841
842         * Scripts/webkitpy/common/config/watchlist: Monitor
843         Source/WTF/wtf/efl and LayoutTests/platform/efl as well.
844
845 2012-05-02  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
846
847         Revert r115191 - "Make the web view's url property follow the active url"
848
849         It uncovered/caused issues in the icon-implementation that can't be
850         easily fixed, so rolling out instead.
851
852         https://bugs.webkit.org/show_bug.cgi?id=77554
853
854         Rubber-stamped by Simon Hausmann.
855
856         * MiniBrowser/qt/qml/BrowserWindow.qml:
857         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
858         (TestWebKitAPI::TEST):
859
860 2012-05-02  Zan Dobersek  <zandobersek@gmail.com>
861
862         Reftest mismatches are providing undefined diff percent in results.html
863         https://bugs.webkit.org/show_bug.cgi?id=85283
864
865         Reviewed by Ojan Vafai.
866
867         When writing test results for a reftest mismatch, also take into account the
868         difference percent that is produced when generating image diff. This value is
869         then used when creating test results in the form of dictionaries, stored under
870         the 'image_diff_percent' key. It ultimately ends up in JSON results data,
871         providing additional information about the reftest difference.
872
873         * Scripts/webkitpy/layout_tests/controllers/manager.py:
874         (interpret_test_failures): Store the diff percent for reftest failure under the
875         'image_diff_percent' key.
876         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
877         (write_test_result): Also set the new diff_percent attribute of the reftest
878         mismatch failure to the value of the difference percent ImageDiff outputs.
879         * Scripts/webkitpy/layout_tests/models/test_failures.py:
880         (FailureReftestMismatch.__init__): Add the new diff_percent attribute with the
881         default value of None.
882         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
883         (EndToEndTest.test_reftest_with_two_notrefs): Update the test case expected results.
884
885 2012-05-02  Terry Anderson  <tdanderson@chromium.org>
886
887         Addition of a contributor to committers.py
888         https://bugs.webkit.org/show_bug.cgi?id=85324
889
890         Reviewed by Unreviewed.
891
892         * Scripts/webkitpy/common/config/committers.py:
893
894 2012-05-02  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
895
896         [EFL] EFL's LayoutTestController overridePreference implementation
897         https://bugs.webkit.org/show_bug.cgi?id=83007
898
899         Reviewed by Antonio Gomes.
900
901         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
902         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
903         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
904         (toBool): aux function converting JSStringRef to bool
905         (toInt): aux function converting JSStringRef to int
906         (LayoutTestController::overridePreference): added implementation
907
908 2012-05-01  Raphael Kubo da Costa  <rakuco@webkit.org>
909
910         webkitpy: Fix sys.platform check in find_files_unittest.
911         https://bugs.webkit.org/show_bug.cgi?id=85339
912
913         Reviewed by Ryosuke Niwa.
914
915         Follow-up to r101618 -- 'win' is not a valid sys.platform value,
916         'win32' should be used instead. The check always returned false
917         due to this, so assert_normalize() was never called.
918
919         * Scripts/webkitpy/common/find_files_unittest.py:
920         (TestWinNormalize.test_win): Add the missing import for the
921         FileSystem class and fix the sys.platform check.
922
923 2012-05-02  Zan Dobersek  <zandobersek@gmail.com>
924
925         [GTK] drag&drop related failures
926         https://bugs.webkit.org/show_bug.cgi?id=84675
927
928         Reviewed by Philippe Normand.
929
930         Rather than checking options to see whether we're running pixel tests
931         the pixel_tests argument passed to GtkDriver._start should be used as
932         it also covers cases when pixel tests should be run because of testing
933         reftests. This again creates different Xvfb processes with proper
934         display IDs for a worker that's performing both ordinary and pixel
935         tests, as was the case before changes made in r114788.
936
937         * Scripts/webkitpy/layout_tests/port/gtk.py:
938         (GtkDriver._start):
939
940 2012-05-02  Simon Hausmann  <simon.hausmann@nokia.com>
941
942         [Qt] Enable device orientation/motion with Qt 5
943
944         Reviewed by Tor Arne Vestbø.
945
946         Enable device orientation/motion features with Qt 5 if Qt Sensors is available.
947
948         * qmake/mkspecs/features/features.prf:
949
950 2012-05-01  Raphael Kubo da Costa  <rakuco@webkit.org>
951
952         webkitpy: Remove unused platform imports.
953         https://bugs.webkit.org/show_bug.cgi?id=85341
954
955         Reviewed by Eric Seidel.
956
957         Remove a few uses of `import platform' which were either rendered
958         unneeded after r105931 or were never needed at all.
959
960         * Scripts/webkitpy/common/system/executive.py:
961         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
962
963 2012-05-01  Terry Anderson  <tdanderson@chromium.org>
964
965         [chromium] Accept four parameters when dispatching a WebInputEvent::GestureTap in chromium DRT eventSender
966         https://bugs.webkit.org/show_bug.cgi?id=85289
967
968         Reviewed by Eric Seidel.
969
970         * DumpRenderTree/chromium/EventSender.cpp:
971         (EventSender::gestureEvent):
972             This change will allow for testing of WK85101. Currently
973             no tests use the extra parameters deltaX and deltaY. Added
974             a check for the size of |arguments|.
975
976 2012-05-01  Eric Seidel  <eric@webkit.org>
977
978         Add myself to the watchlist to watch all rendering changes, as well as sort the watchers in the list
979         https://bugs.webkit.org/show_bug.cgi?id=85305
980
981         Reviewed by Adam Barth.
982
983         * Scripts/webkitpy/common/config/watchlist:
984
985 2012-05-01  Sheriff Bot  <webkit.review.bot@gmail.com>
986
987         Unreviewed, rolling out r115735.
988         http://trac.webkit.org/changeset/115735
989         https://bugs.webkit.org/show_bug.cgi?id=85314
990
991         cause two tests to crash due to an ASSERTION failure
992         (Requested by zhenyao on #webkit).
993
994         * DumpRenderTree/chromium/EventSender.cpp:
995         (EventSender::gestureEvent):
996
997 2012-05-01  Terry Anderson  <tdanderson@chromium.org>
998
999         [chromium] Accept four parameters when dispatching a WebInputEvent::GestureTap in chromium DRT eventSender
1000         https://bugs.webkit.org/show_bug.cgi?id=85289
1001
1002         Reviewed by Eric Seidel.
1003
1004         * DumpRenderTree/chromium/EventSender.cpp:
1005         (EventSender::gestureEvent):
1006             This change will allow for testing of WK85101. Currently no tests use the extra
1007             parameters deltaX and deltaY.
1008
1009 2012-05-01  Zan Dobersek  <zandobersek@gmail.com>
1010
1011         Print out stderr output of ImageDiff if it is present
1012         https://bugs.webkit.org/show_bug.cgi?id=85285
1013
1014         Reviewed by Philippe Normand.
1015
1016         Log the stderr output of ImageDiff as a warning if present. This is
1017         especially helpful when ImageDiff crashes and stderr data might contain
1018         relevant information about what caused the crash.
1019
1020         * Scripts/webkitpy/layout_tests/port/webkit.py:
1021         (WebKitPort._read_image_diff):
1022
1023 2012-05-01  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1024
1025         [Qt] Add an experimental extension to set the min. contents width
1026         https://bugs.webkit.org/show_bug.cgi?id=85281
1027
1028         Reviewed by Antonio Gomes.
1029
1030         * MiniBrowser/qt/qml/BrowserWindow.qml:
1031
1032         Set the value of preferredMinimumContentsWidth to 980.
1033
1034 2012-04-30  Ojan Vafai  <ojan@chromium.org>
1035
1036         Fix tests to not depend on the actual list of builders at build.chromium.org.
1037
1038         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1039         (testHtmlForIndividualTestOnAllBuildersWithChrome):
1040         (testHtmlForIndividualTestOnAllBuildersWithChromeWebkitMaster):
1041
1042 2012-04-29  Sam Weinig  <sam@webkit.org>
1043
1044         Remove BlobBuilder
1045         https://bugs.webkit.org/show_bug.cgi?id=84036
1046
1047         Reviewed by Anders Carlsson.
1048
1049         * Scripts/build-webkit:
1050         * qmake/mkspecs/features/features.prf:
1051         Make exposing the WebKitBlobBuilder JS object conditional on a new
1052         ENABLE_LEGACY_WEBKIT_BLOB_BUILDER flag. Don't enable this for the Mac,
1053         but do for everyone else.
1054
1055 2012-04-30  Amruth Raj Padmanabhuni  <amruthraj@motorola.com>
1056
1057         [GTK] Add glib as a dependency to build glib-networking
1058         https://bugs.webkit.org/show_bug.cgi?id=85181
1059
1060         Reviewed by Martin Robinson.
1061
1062         * gtk/jhbuild.modules:
1063
1064 2012-04-30  Alexis Menard  <alexis.menard@openbossa.org>
1065
1066         Move myself to reviewers.
1067
1068         Reviewed by Andreas Kling.
1069
1070         * Scripts/webkitpy/common/config/committers.py:
1071
1072 2012-04-30  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
1073
1074         Unreviewed, move myself to reviewers
1075
1076         * Scripts/webkitpy/common/config/committers.py:
1077
1078 2012-04-30  Carlos Garcia Campos  <cgarcia@igalia.com>
1079
1080         Unreviewed. Fix make distcheck.
1081
1082         * TestWebKitAPI/GNUmakefile.am: Add missing headers to compilation
1083         and resources to EXTRA_DIST.
1084
1085 2012-04-30  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1086
1087         [GTK] DRT support for layoutTestController.removeOriginAccessWhitelistEntry
1088         https://bugs.webkit.org/show_bug.cgi?id=85156
1089
1090         Reviewed by Martin Robinson.
1091
1092         Add missing implementation removeOriginAccessWhitelistEntry to
1093         GTK's LayoutTestController.
1094
1095         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1096         (LayoutTestController::removeOriginAccessWhitelistEntry):
1097
1098 2012-04-29  Maciej Stachowiak  <mjs@apple.com>
1099
1100         REGRESSION: On Lion, run-webkit-tests changes the display color profile even when not running pixel tests
1101         https://bugs.webkit.org/show_bug.cgi?id=81729
1102
1103         Reviewed by Dan Bernstein.
1104
1105         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1106         (Manager._set_up_run): Only launch LayoutTestHelper when running pixel tests.
1107
1108 2012-04-29  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1109
1110         [GTK] DRT needs an implementation of layoutTestController.setDomainRelaxationForbiddenForURLScheme
1111         https://bugs.webkit.org/show_bug.cgi?id=85131
1112
1113         Reviewed by Martin Robinson.
1114
1115         Add missing implementation setDomainRelaxationForbiddenForURLScheme to
1116         GTK's LayoutTestController.
1117
1118         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1119         (LayoutTestController::setDomainRelaxationForbiddenForURLScheme):
1120
1121 2012-04-28  Sam Weinig  <sam@webkit.org>
1122
1123         Smooth scrolling needs a new key
1124         <rdar://problem/11331632>
1125
1126         Reviewed by Dan Bernstein.
1127
1128         * DumpRenderTree/mac/DumpRenderTree.mm:
1129         (resetDefaultsToConsistentValues):
1130         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm:
1131         (TestWebKitAPI::InjectedBundleController::platformInitialize):
1132         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
1133         (WTR::InjectedBundle::platformInitialize):
1134         Update for new key.
1135
1136 2012-04-28  Zan Dobersek  <zandobersek@gmail.com>
1137
1138         [WK2] http/tests/navigation/anchor-frames-gbk.html fails
1139         https://bugs.webkit.org/show_bug.cgi?id=76896
1140
1141         Reviewed by Darin Adler.
1142
1143         Only dump as text if currently dumping render tree. This ensures that
1144         calling dumpAsText in a test with dumpChildFramesAsText already being
1145         called doesn't override the first decision, possibly making tests fail.
1146
1147         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1148         (WTR::LayoutTestController::dumpAsText):
1149
1150 2012-04-28  Dan Winship  <danw@gnome.org>
1151
1152         Unreviewed; Add me to the SoupNetwork watchlist
1153
1154         * Scripts/webkitpy/common/config/watchlist:
1155
1156 2012-04-27  Gustavo Noronha Silva  <gns@gnome.org>
1157
1158         [GTK] jhbuild cloning is not following WEBKITOUTPUTDIR.
1159         https://bugs.webkit.org/show_bug.cgi?id=76161
1160
1161         Reviewed by Martin Robinson.
1162
1163         * Scripts/webkitdirs.pm:
1164         (getJhbuildPath): New method to obtain the jhbuild base directory,
1165         using the product base directory
1166         (jhbuildConfigurationChanged): Use the new method
1167         (buildAutotoolsProject): Ditto.
1168         * efl/jhbuildrc: Use WEBKITOUTPUTDIR when calculating the path.
1169         * gtk/jhbuildrc: Ditto.
1170         * jhbuild/jhbuild-wrapper: Ditto.
1171
1172 2012-04-27  Scott Graham  <scottmg@chromium.org>
1173
1174         [chromium] make case of npTestNetscapePlugIn.dll match case in copy rule
1175         https://bugs.webkit.org/show_bug.cgi?id=85083
1176
1177         Reviewed by Dirk Pranke.
1178
1179         Make output product_name for npTestNetscapePlugIn.dll match the case
1180         of the copy_TestNetscapePlugIn rule. This is required for ninja, which
1181         is more particular about case matching in rules than the VS build.
1182
1183         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1184
1185 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
1186
1187         [chromium] use "drt-style" output, not "test-shell-style" output, on mac and linux DRT
1188         https://bugs.webkit.org/show_bug.cgi?id=84917
1189
1190         Unreviewed, build fix.
1191
1192         Reland the change in r115453 with a fix for chromium win.
1193         The logic in ChromiumDriver was busted on windows, causing DRT
1194         to get launched without the --test-shell flag.
1195
1196         * Scripts/webkitpy/layout_tests/port/chromium.py:
1197         (ChromiumDriver):
1198         (ChromiumDriver.__init__):
1199         (ChromiumDriver._wrapper_options):
1200         (ChromiumDriver.cmd_line):
1201         (ChromiumDriver._start):
1202         (ChromiumDriver.has_crashed):
1203         (ChromiumDriver.run_test):
1204         (ChromiumDriver.stop):
1205         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1206         (ChromiumDriverTest.setUp):
1207         (ChromiumDriverTest.test_stop):
1208         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
1209         (ChromiumDriverTest.test_two_drivers):
1210
1211 2012-04-27  Peter Beverloo  <peter@chromium.org>
1212
1213         Add John Grabowski as a non-committer to committers.py
1214         https://bugs.webkit.org/show_bug.cgi?id=85080
1215
1216         Reviewed by Dirk Pranke.
1217
1218         He's being CC'ed often enough because of his involvement in the Chromium
1219         side of Chrome for Android upstreaming.
1220
1221         * Scripts/webkitpy/common/config/committers.py:
1222           add jrg@chromium.org
1223
1224 2012-04-27  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
1225
1226         [Qt] Fix minimal build.
1227         https://bugs.webkit.org/show_bug.cgi?id=85045
1228
1229         Reviewed by Tor Arne Vestbø.
1230
1231         Do not set conflicting values of ENABLE_XSLT.
1232
1233         * qmake/mkspecs/features/features.prf:
1234
1235 2012-04-27  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
1236
1237         [EFL][DRT] EFL should sanitize pulseaudio module configuration too
1238         https://bugs.webkit.org/show_bug.cgi?id=85047
1239
1240         Reviewed by Dirk Pranke.
1241
1242         PulseAudio test setup sanitization shared between EFL and GTK.
1243
1244         * Scripts/webkitpy/layout_tests/port/efl.py:
1245         (EflPort):
1246         (EflPort.setup_test_run):
1247         (EflPort.clean_up_test_run):
1248         * Scripts/webkitpy/layout_tests/port/gtk.py:
1249         (GtkPort): Moving pulseaudio unload/restore out to shared class.
1250         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer.py: Added.
1251         (PulseAudioSanitizer):
1252         (PulseAudioSanitizer._unload_pulseaudio_module):
1253         (PulseAudioSanitizer._restore_pulseaudio_module):
1254
1255 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
1256
1257         Unreviewed, revert r115453 - broke chromium win bots
1258
1259         * Scripts/webkitpy/layout_tests/port/chromium.py:
1260         (ChromiumPort._path_to_image_diff):
1261         (ChromiumDriver):
1262         (ChromiumDriver.__init__):
1263         (ChromiumDriver._wrapper_options):
1264         (ChromiumDriver.cmd_line):
1265         (ChromiumDriver.stop):
1266         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1267         (ChromiumDriverTest.setUp):
1268         (ChromiumDriverTest.test_stop):
1269         (ChromiumDriverTest.test_two_drivers):
1270         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
1271
1272 2012-04-27  Ryosuke Niwa  <rniwa@webkit.org>
1273
1274         Chromium Windows Perf bots timeout due to not output
1275         https://bugs.webkit.org/show_bug.cgi?id=84940
1276
1277         Reviewed by Dirk Pranke.
1278
1279         Dirk and I investigated the issue on the bot but we couldn't figure out what is going wrong.
1280         Since we run-perf-tests don't need any of fancy feature printer provides, just use python's
1281         built-in logging module instead. Printing out to stdout and stderr seem to work so hopefully
1282         this will fix the issue on the bot.
1283
1284         * Scripts/run-perf-tests:
1285         * Scripts/webkitpy/performance_tests/perftest.py:
1286         (PerfTest.run):
1287         (PerfTest.run_failed):
1288         (PerfTest.parse_output):
1289         (PerfTest.output_statistics):
1290         (ChromiumStylePerfTest.parse_output):
1291         (PageLoadingPerfTest.run):
1292         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1293         (MainTest.test_parse_output):
1294         (MainTest.test_parse_output_with_failing_line):
1295         (TestPageLoadingPerfTest):
1296         (TestPageLoadingPerfTest.test_run):
1297         (TestPageLoadingPerfTest.test_run_with_bad_output):
1298         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1299         (PerfTestsRunner.__init__):
1300         (PerfTestsRunner._parse_args):
1301         (PerfTestsRunner.run):
1302         (PerfTestsRunner._upload_json):
1303         (PerfTestsRunner._print_status):
1304         (PerfTestsRunner._run_tests_set):
1305         (PerfTestsRunner._run_single_test):
1306         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1307         (create_runner):
1308         (test_run_test_set):
1309         (test_run_test_set_kills_drt_per_run):
1310         (test_run_test_pause_before_testing):
1311         (test_run_test_set_for_parser_tests):
1312         (test_run_test_set_with_json_output):
1313         (test_run_test_set_with_json_source):
1314         (test_run_test_set_with_multiple_repositories):
1315         (test_upload_json):
1316         (test_parse_args):
1317
1318 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
1319
1320         [chromium] use "drt-style" output, not "test-shell-style" output, on mac and linux DRT
1321         https://bugs.webkit.org/show_bug.cgi?id=84917
1322
1323         Reviewed by Ojan Vafai.
1324
1325         This change changes the chromium new-run-webkit-tests
1326         implementation to use the code in WebKitDriver and ServerProcess
1327         by default on Mac and Linux, instead of using the old-style
1328         Chromium/test_shell style of output.
1329
1330         We will still use the test_shell style on Windows for now
1331         and you can still get the old style of behavior with
1332         new-run-webkit-tests --additional-drt-flag --test-shell
1333
1334         * Scripts/webkitpy/layout_tests/port/chromium.py:
1335         (ChromiumDriver):
1336         (ChromiumDriver.__init__):
1337         (ChromiumDriver._wrapper_options):
1338         (ChromiumDriver.cmd_line):
1339         (ChromiumDriver._start):
1340         (ChromiumDriver.has_crashed):
1341         (ChromiumDriver.run_test):
1342         (ChromiumDriver.stop):
1343         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1344         (ChromiumDriverTest.setUp):
1345         (ChromiumDriverTest.test_stop):
1346         (ChromiumDriverTest.test_two_drivers.MockDriver.__init__):
1347         (ChromiumDriverTest.test_two_drivers):
1348
1349 2012-04-27  Dirk Pranke  <dpranke@chromium.org>
1350
1351         webkitpy: fix race in unit tests for manager_worker_broker
1352         https://bugs.webkit.org/show_bug.cgi?id=84925
1353
1354         Reviewed by Ojan Vafai.
1355
1356         I have been intermittently seeing hangs when running
1357         test-webkitpy on Mac SL; it appears that there's a bug (?) in
1358         the multiprocessing module where queue references are getting
1359         leaked or not cleaned up somehow. By explicitly closing the
1360         queues, things seem to be happy, so this patch adds a cleanup()
1361         method to the message broker and modifies the manager (and the
1362         unit tests) to call it. This may get rid of the intermittent
1363         error on shutdown on Linux I've seen as well.
1364
1365         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1366         (Manager._run_tests):
1367         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
1368         (_Broker.__del__):
1369         (_Broker):
1370         (_Broker.cleanup):
1371         (_BrokerConnection.cleanup):
1372         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
1373         (_TestsMixin.test_name):
1374         (_TestsMixin.test_cancel):
1375         (_TestsMixin.test_done):
1376         (_TestsMixin.test_unknown_message):
1377
1378 2012-04-27  Gavin Peters  <gavinp@chromium.org>
1379
1380         Add new ENABLE_LINK_PRERENDER define to control the Prerendering API
1381         https://bugs.webkit.org/show_bug.cgi?id=84871
1382
1383         Reviewed by Adam Barth.
1384
1385         Prerendering is currently covered by the ENABLE_LINK_PREFETCH macro, but the new Prerendering
1386         API separates it from prefetching.  Having separate include guards lets ports enable prefetching,
1387         a relatively easy change, without needing to build the infrastructure for prerendering, which
1388         is considerably more complicated.
1389
1390 2012-04-27  Philippe Normand  <pnormand@igalia.com>
1391
1392         Unreviewed, GTK gardening.
1393
1394         * Scripts/run-gtk-tests:
1395         (TestRunner): Skip flaky WebKit2 API test.
1396
1397 2012-04-27  Philippe Normand  <pnormand@igalia.com>
1398
1399         Unreviewed, GTK gardening.
1400
1401         * Scripts/run-gtk-tests:
1402         (TestRunner): Fix another typo. So easy to get confused by a Test
1403         name and the actual name of the executable...
1404
1405 2012-04-27  Philippe Normand  <pnormand@igalia.com>
1406
1407         Unreviewed, GTK gardening.
1408
1409         * Scripts/run-gtk-tests:
1410         (TestRunner): Skip failing WebKit2 API test.
1411
1412 2012-04-27  Christophe Dumez  <christophe.dumez@intel.com>
1413
1414         [EFL] DRT should support LayoutTestController's dumpResourceLoadCallbacks()
1415         https://bugs.webkit.org/show_bug.cgi?id=81891
1416
1417         Reviewed by Philippe Normand.
1418
1419         Implement support for LayoutTestController's
1420         dumpResourceLoadCallbacks() in EFL's DumpRenderTree by catching new
1421         ewk_view signals about resource loading and print out the expected
1422         information. This allows for several test cases to be removed from the
1423         skip list.
1424
1425         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1426         (DumpRenderTreeChrome::createView):
1427         (descriptionSuitableForTestResult):
1428         (DumpRenderTreeChrome::onResponseReceived):
1429         (DumpRenderTreeChrome::onResourceLoadFinished):
1430         (DumpRenderTreeChrome::onResourceLoadFailed):
1431         (DumpRenderTreeChrome::onNewResourceRequest):
1432         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
1433         (DumpRenderTreeChrome):
1434
1435 2012-04-27  Philippe Normand  <pnormand@igalia.com>
1436
1437         Unreviewed, GTK typo fix.
1438
1439         * Scripts/run-gtk-tests:
1440         (TestRunner): Fix-up skipped WebKit2 API test name TestWKConnection.
1441
1442 2012-04-26  Carlos Garcia Campos  <cgarcia@igalia.com>
1443
1444         [SOUP] Add a way to register custom uri schemes in WebKit2
1445         https://bugs.webkit.org/show_bug.cgi?id=84130
1446
1447         Reviewed by Martin Robinson.
1448
1449         * Scripts/webkitpy/style/checkers/cpp.py:
1450         (check_identifier_name_in_declaration): Add webkit_soup prefix to
1451         the list of exceptions to the underscores in identifiers rule.
1452
1453 2012-04-26  Ryosuke Niwa  <rniwa@webkit.org>
1454
1455         Extract PerfTestFactory
1456         https://bugs.webkit.org/show_bug.cgi?id=83944
1457
1458         Reviewed by Dirk Pranke.
1459         
1460         Extracted PerfTestFactory out of PerfTest. This class creates appropriate PerfTest object given
1461         a test name and a path. Also removed dirname from the constructors of PerfTest and its subclasses.
1462
1463         Also added simple test cases to sanity check PerfTestFactory.
1464
1465         * Scripts/webkitpy/performance_tests/perftest.py:
1466         (PerfTest.__init__):
1467         (PerfTest.test_name):
1468         (ChromiumStylePerfTest.__init__):
1469         (PageLoadingPerfTest.__init__):
1470         (PageLoadingPerfTest.run):
1471         (PerfTestFactory):
1472         (PerfTestFactory.create_perf_test):
1473         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1474         (MainTest.test_parse_output):
1475         (MainTest.test_parse_output_with_failing_line):
1476         (TestPageLoadingPerfTest.test_run):
1477         (TestPageLoadingPerfTest.test_run_with_bad_output):
1478         (TestPerfTestFactory):
1479         (TestPerfTestFactory.test_regular_test):
1480         (TestPerfTestFactory.test_inspector_test):
1481         (TestPerfTestFactory.test_page_loading_test):
1482         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1483         (PerfTestsRunner):
1484         (PerfTestsRunner._collect_tests):
1485         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1486         (run_test):
1487         (_tests_for_runner):
1488
1489 2012-04-26  Dirk Pranke  <dpranke@chromium.org>
1490
1491         nrwt: clean up server process, webkit driver so chromium can use it
1492         https://bugs.webkit.org/show_bug.cgi?id=84910
1493
1494         Reviewed by Ojan Vafai.
1495
1496         This change moves the "sample a process" logic out of
1497         server_process.py and into a port-specific class (where really
1498         only the mac has an implementation), and also preemptively kills
1499         DRT when a test times out in WebKitDriver (rather than waiting through the
1500         additional delays caused by calling stop() when we would want to
1501         restart the driver generically in worker.py).
1502
1503         These changes will make it possible for the chromium port to
1504         switch over to the stock WebKitDriver implementation, at least
1505         on mac and linux.
1506
1507         * Scripts/webkitpy/layout_tests/port/base.py:
1508         (Port.sample_process):
1509         * Scripts/webkitpy/layout_tests/port/mac.py:
1510         (MacPort.sample_process):
1511         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1512         (test_helper_fails_to_stop):
1513         (test_sample_process):
1514         (test_sample_process.logging_run_command):
1515         (test_sample_process_throws_exception):
1516         (test_sample_process_throws_exception.throwing_run_command):
1517         * Scripts/webkitpy/layout_tests/port/server_process.py:
1518         (ServerProcess._log):
1519         (ServerProcess._handle_timeout):
1520         (ServerProcess.stop):
1521         (ServerProcess):
1522         (ServerProcess.kill): Here we add a method to immediately stop
1523         the process rather than trying to shut it down cleanly.
1524         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
1525         (TestServerProcess.test_broken_pipe):
1526         * Scripts/webkitpy/layout_tests/port/webkit.py:
1527         (WebKitDriver.run_test): Fix an issue where we weren't passing
1528         along any per-test args (only needed for Chromium, but still).
1529         Also, kill the driver immediately if we time out a test.
1530
1531 2012-04-26  Benjamin Poulain  <bpoulain@apple.com>
1532
1533         ObjcClass::methodsNamed() can leak if buffer is dynamically allocated
1534         https://bugs.webkit.org/show_bug.cgi?id=84668
1535
1536         Reviewed by Alexey Proskuryakov.
1537
1538         Extend ObjCPlugin to support the new layout tests of the Objective-C bridge.
1539
1540         * DumpRenderTree/mac/ObjCPlugin.m:
1541         (+[ObjCPlugin isSelectorExcludedFromWebScript:]):
1542         (+[ObjCPlugin webScriptNameForSelector:]):
1543         (-[ObjCPlugin methodMappedToLongName]):
1544         (-[ObjCPlugin testConversionColon:]):
1545         (-[ObjCPlugin _:]):
1546
1547 2012-04-26  Dimitri Glazkov  <dglazkov@chromium.org>
1548
1549         Unreviewed, rolling out r115340.
1550         http://trac.webkit.org/changeset/115340
1551         https://bugs.webkit.org/show_bug.cgi?id=83048
1552
1553         Does not work with Windows.
1554
1555         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1556         (SingleTestRunner._run_reftest):
1557         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
1558         (copy_file):
1559         * Scripts/webkitpy/layout_tests/port/base.py:
1560         (Port._parse_reftest_list):
1561         (Port.reference_files):
1562         (Port._real_tests):
1563         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1564         (PortTest.test_reference_files):
1565         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1566         (ChromiumDriverTest.test_write_command_and_read_line):
1567
1568 2012-04-26  Christophe Dumez  <christophe.dumez@intel.com>
1569
1570         [EFL] Enable VIDEO_TRACK feature
1571         https://bugs.webkit.org/show_bug.cgi?id=84830
1572
1573         Reviewed by Gustavo Noronha Silva.
1574
1575         Enable VIDEO_TRACK by default on EFL port whenever build-webkit is
1576         used.
1577
1578         * Scripts/build-webkit:
1579
1580 2012-04-08  Robert Hogan  <robert@webkit.org>
1581
1582         new-webkit-run-tests: handle ref tests from the CSS test suite
1583         https://bugs.webkit.org/show_bug.cgi?id=83048
1584
1585         Reviewed by Ryosuke Niwa.
1586
1587         This allows new-run-webkit-tests to run reference tests created for the CSS test suite. It removes a big swathe
1588         of pixel and text results that are no longer required now that the reference results are used and adds in the support
1589         files from the CSS test suite that allow the ref tests to run.
1590
1591         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1592         (SingleTestRunner._run_reftest): Assert that the reference result exists. Previously the test would
1593          run without it and report a pass!
1594         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
1595         (copy_file): ensure the output directory exists, it may not if the reference test is
1596           contained in a folder that contains nothing but reference tests.
1597         * Scripts/webkitpy/layout_tests/port/base.py:
1598         (Port.available_reference_files): Add as a helper. Handle embedded relative paths, e.g: some/../path
1599         (Port.reference_files): Only return reference results in the reftest list that actually exist on
1600           disk. Also, don't ignore potential reference results on disk that are not in the reftest list, and
1601           don't ignore them just because we have a reftest list for the test's directory.
1602         (Port._real_tests): ignore anything that looks like a test in a 'support' directory, this semantic
1603           is used by the CSS 2.1 test suite.
1604         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1605         (PortTest.test_reference_files_when_reftestlist_and_reference_file_present): Ensure that a test with a reference result
1606           on disk but no reference result listed in the directory's reftest list picks up its reference result.
1607         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1608         (ChromiumDriverTest.test_strip_uri): Test that some/test.html and some/other/../test.html are treated
1609          as the same file.
1610
1611 2012-04-26  Carlos Garcia Campos  <cgarcia@igalia.com>
1612
1613         Unreviewed. Fix GTK+ build after r115314.
1614
1615         For some reason it fails while bulding GtkVersioning for WTR, but
1616         it didn't fail neither in EWS nor in my laptop. We don't really
1617         need GtkVersioning in WTR since it already depends on GTK+3
1618         unconditionally.
1619
1620         * WebKitTestRunner/GNUmakefile.am: Remove GtkVersioning.c from
1621         compilation.
1622         * WebKitTestRunner/gtk/EventSenderProxyGtk.cpp:
1623         (WTR::EventSenderProxy::createMouseButtonEvent): Use GTK+ 3 API
1624         directly instead of getDefaultGDKPointerDevice() helper.
1625         (WTR::EventSenderProxy::keyDown): Ditto.
1626         (WTR::EventSenderProxy::mouseMoveTo): Ditto.
1627
1628 2012-04-24  Ojan Vafai  <ojan@chromium.org>
1629
1630         Show flakiness dashboard data in garden-o-matic
1631         https://bugs.webkit.org/show_bug.cgi?id=83716
1632
1633         Reviewed by Dimitri Glazkov.
1634
1635         Put an iframe below the expected/actual results in the Results view.
1636         Size the iframe to it's height.
1637
1638         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
1639         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
1640         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
1641         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1642         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
1643         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
1644         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
1645
1646 2012-04-26  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1647
1648         [EFL] [DRT] LayoutTestController needs implementation of setJavaScriptProfilingEnabled
1649         https://bugs.webkit.org/show_bug.cgi?id=84576
1650
1651         Reviewed by Martin Robinson.
1652
1653         Add missing implementation setJavaScriptProfilingEnabled to EFL's
1654         LayoutTestController in order to unskip tests rely on this setting.
1655
1656         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1657         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1658         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1659         (LayoutTestController::setJavaScriptProfilingEnabled):
1660
1661 2012-04-26  Xianzhu Wang  <wangxianzhu@chromium.org>
1662
1663         Add Jing Zhao, Min Qin, Tien-Ren Chen and Hao Zheng into contributors list.
1664
1665         Not reviewed.
1666
1667         * Scripts/webkitpy/common/config/committers.py:
1668
1669 2012-04-26  Carlos Garcia Campos  <cgarcia@igalia.com>
1670
1671         [GTK] Test TestWebKitAPI/WebKit2/TestNewFirstVisuallyNonEmptyLayout times out
1672         https://bugs.webkit.org/show_bug.cgi?id=84957
1673
1674         Reviewed by Martin Robinson.
1675
1676         * Scripts/run-gtk-tests:
1677         (TestRunner): Unskip TestNewFirstVisuallyNonEmptyLayout and
1678         TestNewFirstVisuallyNonEmptyLayoutForImages.
1679         * TestWebKitAPI/gtk/PlatformWebViewGtk.cpp:
1680         (TestWebKitAPI::PlatformWebView::PlatformWebView): Show the
1681         toplevel window containing the web view to make sure events are
1682         received.
1683
1684 2012-04-26  Carlos Garcia Campos  <cgarcia@igalia.com>
1685
1686         Unreviewed. Add bug numbers to skipped tests.
1687
1688         * Scripts/run-gtk-tests:
1689         (TestRunner):
1690
1691 2012-04-26  Carlos Garcia Campos  <cgarcia@igalia.com>
1692
1693         [GTK] Build and run TestWebKitAPI WebKit2 unit tests
1694         https://bugs.webkit.org/show_bug.cgi?id=84446
1695
1696         Reviewed by Philippe Normand.
1697
1698         * Scripts/run-gtk-tests:
1699         (TestRunner): Add TestWebKitAPI/WebKit2 to the list of directories
1700         containing tests. Unskip FunctionalTest.RefCountedStorage, since
1701         we are now using a custom main that initializes WTF threading.
1702         (TestRunner._setup_testing_environment): Set environment variables
1703         required to run WebKit2 tests.
1704         * TestWebKitAPI/GNUmakefile.am:
1705         * TestWebKitAPI/JavaScriptTest.cpp: Use
1706         JavaScriptCore/JSContextRef.h instead of
1707         JavaScriptCore/JavaScriptCore.h since it includes JSStringRefCF.h
1708         unconditionally.
1709         * TestWebKitAPI/PlatformWebView.h: Define PlatformWKView and
1710         PlatformWindow for GTK.
1711         * TestWebKitAPI/Tests/WebKit2/EvaluateJavaScript.cpp: Remove
1712         unneded header that breaks the build of GTK port.
1713         * TestWebKitAPI/Tests/WebKit2/PreventEmptyUserAgent.cpp: Use
1714         JavaScriptCore/JSContextRef.h instead of
1715         JavaScriptCore/JavaScriptCore.h since it includes JSStringRefCF.h
1716         unconditionally.
1717         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
1718         (TestWebKitAPI::TEST): Add expected values for GTK+ port.
1719         * TestWebKitAPI/gtk/InjectedBundleControllerGtk.cpp:
1720         (TestWebKitAPI):
1721         (TestWebKitAPI::InjectedBundleController::platformInitialize):
1722         * TestWebKitAPI/gtk/PlatformUtilitiesGtk.cpp: Added.
1723         (TestWebKitAPI::Util::checkTestFinished):
1724         (TestWebKitAPI::Util::run):
1725         (TestWebKitAPI::Util::sleep):
1726         (TestWebKitAPI::Util::getFilenameFromEnvironmentVariableAsUTF8):
1727         (TestWebKitAPI::Util::createInjectedBundlePath):
1728         (TestWebKitAPI::Util::createURLForResource):
1729         (TestWebKitAPI::Util::URLForNonExistentResource):
1730         (TestWebKitAPI::Util::isKeyDown):
1731         * TestWebKitAPI/gtk/PlatformWebViewGtk.cpp: Added.
1732         (TestWebKitAPI::PlatformWebView::PlatformWebView):
1733         (TestWebKitAPI::PlatformWebView::~PlatformWebView):
1734         (TestWebKitAPI::PlatformWebView::page):
1735         (TestWebKitAPI::PlatformWebView::resizeTo):
1736         (TestWebKitAPI::doKeyStroke):
1737         (TestWebKitAPI::PlatformWebView::simulateSpacebarKeyPress):
1738         (TestWebKitAPI::PlatformWebView::simulateAltKeyPress):
1739         (TestWebKitAPI::doMouseButtonEvent):
1740         (TestWebKitAPI::PlatformWebView::simulateRightClick):
1741         (TestWebKitAPI::PlatformWebView::simulateMouseMove):
1742         * TestWebKitAPI/gtk/main.cpp:
1743         (main):
1744
1745 2012-04-26  Carlos Garcia Campos  <cgarcia@igalia.com>
1746
1747         [GTK] run-gtk-tests: Use a timeout per test instead of a global timeout
1748         https://bugs.webkit.org/show_bug.cgi?id=84695
1749
1750         Reviewed by Philippe Normand.
1751
1752         It also adds a command line option to be able to pass a custom
1753         timeout value, instead of hard-coding it. The default timeout is 10
1754         seconds if no other value is passed to the script.
1755
1756         * Scripts/run-gtk-tests:
1757         (TestTimeout): Exception raised when a test times out.
1758         (TestRunner._get_child_pid_from_test_output): Helper function to
1759         get the pid of the running test from gtester output.
1760         (TestRunner._kill_process): Helper funtion ot kill a process
1761         ignoring exceptions if the process is already died.
1762         (TestRunner._run_test_command): Run the test command raising
1763         TestTimeout exception if the test doesn't finish before the given
1764         timeout in seconds.
1765         (TestRunner._run_test_command.alarm_handler): Alarm handler that
1766         raises TestTimeout exception.
1767         (TestRunner._run_test_glib): Use _run_test_command() to run the test.
1768         (TestRunner._run_test_google): Ditto.
1769         (TestRunner.run_tests): Add current test to the list of timed out
1770         test if TestTimeout exception is raised. Show the list of timed
1771         out test at the end.
1772
1773 2012-04-26  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1774
1775         [EFL] [DRT] LayoutTestController needs implementation of addOriginAccessWhitelistEntry and removeOriginAccessWhitelistEntry
1776         https://bugs.webkit.org/show_bug.cgi?id=83880
1777
1778         Reviewed by Antonio Gomes.
1779
1780         Add missing implementation addOriginAccessWhitelistEntry and
1781         removeOriginAccessWhitelistEntry to EFL's LayoutTestController
1782         in order to unskip tests which rely on this.
1783
1784         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1785         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1786         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1787         (LayoutTestController::addOriginAccessWhitelistEntry):
1788         (LayoutTestController::removeOriginAccessWhitelistEntry):
1789
1790 2012-04-26  Carlos Garcia Campos  <cgarcia@igalia.com>
1791
1792         [GTK] Add downloads support to MiniBrowser
1793         https://bugs.webkit.org/show_bug.cgi?id=73662
1794
1795         Reviewed by Philippe Normand.
1796
1797         * MiniBrowser/gtk/BrowserDownloadsBar.c: Added.
1798         (browserDownloadsBarChildRemoved): Close the downloads bar if
1799         there aren't active downloads.
1800         (browserDownloadsBarResponse): Destroy the downloads bar when
1801         closed.
1802         (browser_downloads_bar_init):
1803         (browser_downloads_bar_class_init):
1804         (browser_downloads_bar_new): Create a new downloads bar.
1805         (actionButtonClicked): Cancel the download if cancel button was
1806         clicked, or open the downloaded file if the download finished.
1807         (browserDownloadFinalize):
1808         (browser_download_init):
1809         (browser_download_class_init):
1810         (downloadReceivedResponse): Update download message using the
1811         response URI.
1812         (remainingTime): Helper function to get human readable remaining
1813         time.
1814         (downloadProgress): Update downloads progress.
1815         (downloadReceivedData):
1816         (downloadFinished): Update download message and turn download
1817         button into open.
1818         (downloadFailed): Show error message or destroy the download UI if
1819         it was cancelled.
1820         (browserDownloadNew): Create a new download widget for the given
1821         WebKitDownload.
1822         (browser_downloads_bar_add_download): Add download widget to the
1823         downloads bar.
1824         * MiniBrowser/gtk/BrowserDownloadsBar.h: Added.
1825         * MiniBrowser/gtk/BrowserWindow.c:
1826         (downloadStarted): Create a download widget and add it to the
1827         downloads bar when a new download is started.
1828         (browserWindowConstructed): Connect to
1829         WebKitWebContext::download-started signal.
1830         * MiniBrowser/gtk/GNUmakefile.am: Add new files to compilation.
1831
1832 2012-04-26  Carlos Garcia Campos  <cgarcia@igalia.com>
1833
1834         Unreviewed. Skip unit test failing in debug bot.
1835
1836         * Scripts/run-gtk-tests:
1837         (TestRunner):
1838
1839 2012-04-26  Christophe Dumez  <christophe.dumez@intel.com>
1840
1841         [EFL][DRT] "resource,request,willsend" signal needs to provide more information
1842         https://bugs.webkit.org/show_bug.cgi?id=84670
1843
1844         Reviewed by Andreas Kling.
1845
1846         Handle in EFL's DumpRenderTree the new Ewk_Frame_Resource_Messages now
1847         passed with the "resource,request,willsend" signal and dump
1848         information about the network request and the possible network
1849         response if LayoutTestController's dumpResourceLoadCallbacks() returns
1850         true.
1851
1852         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1853         (pathSuitableForTestResult):
1854         (urlSuitableForTestResult):
1855         (descriptionSuitableForTestResult):
1856         (DumpRenderTreeChrome::onWillSendRequest):
1857
1858 2012-04-25  Csaba Osztrogonác  <ossy@webkit.org>
1859
1860         [Qt][WK2] Unreviewed, rolling out r113703 because of crashing tests.
1861         http://trac.webkit.org/changeset/113703
1862         https://bugs.webkit.org/show_bug.cgi?id=83024
1863
1864         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
1865         * Tools.pro:
1866         * qmake/mkspecs/features/features.prf:
1867         * qmake/mkspecs/features/functions.prf:
1868
1869 2012-04-25  Maciej Stachowiak  <mjs@apple.com>
1870
1871         Teach LayoutTestHelper to correctly restore the color profile when it's a factory profile
1872         https://bugs.webkit.org/show_bug.cgi?id=84931
1873
1874         Reviewed by Filip Pizlo.
1875
1876         * DumpRenderTree/mac/LayoutTestHelper.m:
1877         (installLayoutTestColorProfile): Account for the possibility of the default
1878         profile being a factory profile.
1879
1880 2012-04-25  Sheriff Bot  <webkit.review.bot@gmail.com>
1881
1882         Unreviewed, rolling out r115240.
1883         http://trac.webkit.org/changeset/115240
1884         https://bugs.webkit.org/show_bug.cgi?id=84928
1885
1886         It broke everything (Requested by Ossy_HOME on #webkit).
1887
1888         * Scripts/webkitpy/layout_tests/port/gtk.py:
1889         (GtkDriver):
1890         (GtkDriver._start):
1891         (GtkDriver._start.x_filter):
1892         (GtkDriver.stop):
1893         (GtkPort._driver_class):
1894         * Scripts/webkitpy/layout_tests/port/qt.py:
1895         (QtPort._build_driver):
1896         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py: Removed.
1897
1898 2012-04-25  Dirk Pranke  <dpranke@chromium.org>
1899
1900         [chromium] change DRT's "DRT" output to match the other ports
1901         https://bugs.webkit.org/show_bug.cgi?id=84904
1902
1903         Reviewed by Kent Tamura.
1904
1905         Since we run Chromium's DRT in --test-shell mode, the
1906         non-test-shell output has never been really beaten on to make
1907         sure it's right. This fixes a few issues and will be tested
1908         with upcoming NRWT changes that'll land in a separate patch.
1909         In particular we should never output framing text like "\n"
1910         or "#EOF" outside of something outside of TestEventPrinter,
1911         and we weren't handling audio output properly at all.
1912
1913         * DumpRenderTree/chromium/TestEventPrinter.cpp:
1914         (DRTPrinter):
1915         (TestShellPrinter):
1916         (DRTPrinter::handleTextFooter):
1917         (DRTPrinter::handleAudioFooter):
1918         (DRTPrinter::handleTestFooter):
1919         (TestShellPrinter::handleAudioFooter):
1920         * DumpRenderTree/chromium/TestEventPrinter.h:
1921         (TestEventPrinter):
1922         * DumpRenderTree/chromium/TestShell.cpp:
1923         (TestShell::TestShell):
1924         (TestShell::dump):
1925
1926 2012-04-25  Joshua Bell  <jsbell@chromium.org>
1927
1928         Unreviewed: Adding Alec Flett as a contributor (non-committer).
1929
1930         * Scripts/webkitpy/common/config/committers.py:
1931
1932 2012-04-25  Abhishek Arya  <inferno@chromium.org>
1933
1934         Mark myself as a reviewer.
1935         https://bugs.webkit.org/show_bug.cgi?id=84902
1936
1937         Reviewed by Dimitri Glazkov.
1938
1939         * Scripts/webkitpy/common/config/committers.py:
1940
1941 2012-04-25  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
1942
1943         [Qt][NRWT] Run each DRT in it's own xvfb
1944         https://bugs.webkit.org/show_bug.cgi?id=77335
1945
1946         Reviewed by Dirk Pranke.
1947
1948         * Scripts/webkitpy/layout_tests/port/gtk.py:
1949         (GtkPort._driver_class):
1950         * Scripts/webkitpy/layout_tests/port/qt.py:
1951         (QtPort._driver_class):
1952         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py: Added.
1953         (XvfbDriver):
1954         (XvfbDriver._start):
1955         (XvfbDriver._start.x_filter):
1956         (XvfbDriver.stop):
1957
1958 2012-04-25  Sheriff Bot  <webkit.review.bot@gmail.com>
1959
1960         Unreviewed, rolling out r115222.
1961         http://trac.webkit.org/changeset/115222
1962         https://bugs.webkit.org/show_bug.cgi?id=84874
1963
1964         Caused test regressions on the EFL build (Requested by rakuco
1965         on #webkit).
1966
1967         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1968         (LayoutTestController::addOriginAccessWhitelistEntry):
1969         (LayoutTestController::removeOriginAccessWhitelistEntry):
1970
1971 2012-04-25  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1972
1973         [EFL] [DRT] LayoutTestController needs implementation of addOriginAccessWhitelistEntry and removeOriginAccessWhitelistEntry
1974         https://bugs.webkit.org/show_bug.cgi?id=83880
1975
1976         Reviewed by Antonio Gomes.
1977
1978         Add missing implementation addOriginAccessWhitelistEntry and
1979         removeOriginAccessWhitelistEntry to EFL's LayoutTestController
1980         in order to unskip tests which rely on this.
1981
1982         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1983         (LayoutTestController::addOriginAccessWhitelistEntry):
1984         (LayoutTestController::removeOriginAccessWhitelistEntry):
1985
1986 2012-04-25  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1987
1988         [EFL] [DRT] Reset text zoom factor before running a test
1989         https://bugs.webkit.org/show_bug.cgi?id=84861
1990
1991         Reviewed by Antonio Gomes.
1992
1993         This fix eliminates most of the current flakiness on EFL build
1994         bots, although some CSS tests will now need rebasiling.
1995
1996         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1997         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1998
1999 2012-04-25  Milian Wolff  <milian.wolff@kdab.com>
2000
2001         [Qt] add LayoutTestController::setPrinting support to Qt unit tests
2002         https://bugs.webkit.org/show_bug.cgi?id=84246
2003
2004         Reviewed by Simon Hausmann.
2005
2006         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2007         (WebCore::DumpRenderTree::dump):
2008
2009 2012-04-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2010
2011         [Qt] Make the web view's url property follow the active url
2012
2013         https://bugs.webkit.org/show_bug.cgi?id=77554
2014
2015         Update  the location bar in the minibrowser to behave
2016         a bit more like normal browsers in terms of when the url will change
2017         and how active focus is handled.
2018
2019         Reviewed by Simon Hausmann.
2020
2021         * MiniBrowser/qt/qml/BrowserWindow.qml:
2022
2023 2012-04-25  Philippe Normand  <pnormand@igalia.com>
2024
2025         Webkit build fails due to missing gstreamer include file on Kubuntu 8.04
2026         https://bugs.webkit.org/show_bug.cgi?id=81913
2027
2028         Reviewed by Tor Arne Vestbø.
2029
2030         * qmake/mkspecs/features/features.prf: Check for gstreamer-0.10
2031         and gstreamer-plugins-base >= 0.10.30 just like the GTK port
2032         does. 0.10.27 is unfortunately not recent enough to get the media
2033         player working reliably.
2034
2035
2036 2012-04-25  Xiaobo Wang  <xbwang@torchmobile.com.cn>
2037
2038         [BlackBerry] Run tests in '/http/tests/local' with local URI
2039         https://bugs.webkit.org/show_bug.cgi?id=84820
2040
2041         Reviewed by Nikolas Zimmermann.
2042
2043         Currently we run all tests under folder "/http/tests/" as HTTP
2044         tests (http://...). However in WebKit ORWT and NRWT scripts, tests in
2045         'http/tests/local' are run with local path (like '/developer/LayoutTests
2046         /http/tests/local/absolute-url-strip-whitespace.html').
2047         With this patch we can keep the same behaviour as other portings.
2048
2049         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
2050         (BlackBerry::WebKit::DumpRenderTree::isHTTPTest):
2051
2052 2012-04-25  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
2053
2054         [EFL][DRT] Reset DRT settings to their default values after a testcase was executed.
2055         https://bugs.webkit.org/show_bug.cgi?id=84709
2056
2057         Reviewed by Eric Seidel.
2058
2059         Not all the settings that can be changed by layout testcases through a JS API were 
2060         reset in DumpRenderTreeChrome::resetDefaultsToConsistentValues().
2061
2062         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2063         (defaultEditingBehavior): returns editing behavior depending on current platform.
2064         (DumpRenderTreeChrome::resetDefaultsToConsistentValues): reset some missed DRT settings
2065
2066 2012-04-24  Dirk Pranke  <dpranke@chromium.org>
2067
2068         nrwt: add ASAN handling to chromium crash logs
2069         https://bugs.webkit.org/show_bug.cgi?id=84664
2070
2071         Reviewed by Ojan Vafai.
2072
2073         * Scripts/webkitpy/layout_tests/port/chromium.py:
2074         (ChromiumPort._get_crash_log):
2075
2076 2012-04-24  Dirk Pranke  <dpranke@chromium.org>
2077
2078         webkit-patch rebaseline-test is retrieving stale expectations
2079         https://bugs.webkit.org/show_bug.cgi?id=84762
2080
2081         Reviewed by Ojan Vafai.
2082
2083         webkit-patch rebaseline-expectations is fetching files directly
2084         from the layout-test-results directory on the buildbots, rather
2085         than fetching them from the archive. This is problematic because
2086         we don't clobber the directory after each run and so you can end
2087         up fetching stale failing results.
2088
2089         This change temporarily changes the code to fetch the zip file
2090         instead -- making things much slower -- until I can confer w/
2091         abarth and ojan over the best way to fix this for the long-term.
2092
2093         * Scripts/webkitpy/tool/commands/rebaseline.py:
2094         (RebaselineTest.__init__):
2095         (RebaselineTest._zip_file_set):
2096         (RebaselineTest):
2097         (RebaselineTest._fetch_baseline):
2098         (RebaselineTest._rebaseline_test):
2099         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2100         (TestRebaseline.stub_rebaseline_test_command_and_tool):
2101         (TestRebaseline.stub_rebaseline_test_command_and_tool.FakeZipFileSet):
2102         (TestRebaseline.stub_rebaseline_test_command_and_tool.FakeZipFileSet.read):
2103         (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
2104         (test_rebaseline_updates_expectations_file):
2105         (test_rebaseline_test):
2106         (test_rebaseline_and_copy_test):
2107         (test_rebaseline_and_copy_test_no_existing_result):
2108         (test_rebaseline_and_copy_test_with_lion_result):
2109         (test_rebaseline_and_copy_no_overwrite_test):
2110
2111 2012-04-24  Ojan Vafai  <ojan@chromium.org>
2112
2113         Improve code for embedding flakiness dashboard as an iframe
2114         https://bugs.webkit.org/show_bug.cgi?id=84782
2115
2116         Reviewed by Dirk Pranke.
2117
2118         -Include the horizontal scrollbar height in the measured height of the page.
2119         -Post a message everytime the height changes instead of in response to a queryContentHeight message.
2120         This way, the embedder doesn't need to poll.
2121
2122         No new tests. This is hard to test because it involves postMessage and iframes.
2123         Once we switch the dashboard tests over to qunit, we should be able to write an
2124         asyncTest for this.
2125
2126         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2127         (generatePage):
2128         (appendHTML):
2129         (appendExpectations):
2130         (postHeightChangedMessage):
2131         (generatePageForIndividualTests.if):
2132         (generatePageForIndividualTests):
2133         * TestResultServer/static-dashboards/flakiness_dashboard_embedded.html: Added.
2134
2135 2012-04-16  Ojan Vafai  <ojan@chromium.org>
2136
2137         Replace garden-o-matic accordion UI with a panel UI
2138         https://bugs.webkit.org/show_bug.cgi?id=84081
2139
2140         Reviewed by Dimitri Glazkov.
2141
2142         This is a nicer UI to work with and makes it easier to embed the flakiness dashboard,
2143         since we can reuse the same iframe without moving it in the DOM.
2144
2145         Also, excise unnecessary jquery-isms.
2146
2147         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2148         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2149         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
2150
2151 2012-04-24  Peter Beverloo  <peter@chromium.org>
2152
2153         [Chromium] DumpRenderTree should depend on base.gyp:test_support_base
2154         https://bugs.webkit.org/show_bug.cgi?id=84729
2155
2156         Reviewed by Dirk Pranke.
2157
2158         This dependency is necessary for Android since it'll soon be
2159         calling InitAndroidOSPathStub() in it's platform support. The
2160         TestWebKitAPI and webkit_unit_tests already have this dependency.
2161
2162         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2163
2164 2012-04-24  Ojan Vafai  <ojan@chromium.org>
2165
2166         Replace vague informative statement with the list of builders where the test isn't listed
2167         https://bugs.webkit.org/show_bug.cgi?id=84779
2168
2169         Reviewed by Dirk Pranke.
2170
2171         This is more informative and more likely to be noticed. With the old UI, people never
2172         read the line and if they did, they didn't really understand what it meant. Hopefully
2173         have the list of builders will be more clear.
2174
2175         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2176         * TestResultServer/static-dashboards/flakiness_dashboard_tests.css:
2177         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2178
2179 2012-04-24  Tommy Widenflycht  <tommyw@google.com>
2180
2181         MediaStream API: Create a new flag for PeerConnection
2182         https://bugs.webkit.org/show_bug.cgi?id=84723
2183
2184         Reviewed by Dimitri Glazkov.
2185
2186         To allow Chrome to make the GetUserMedia functionality available by default we
2187         need another flag for the PeerConnection related parts. This new flag is in
2188         addition to the general media stream flag, and both need to be enabled for
2189         PeerConnection object creation.
2190
2191         * DumpRenderTree/chromium/TestShell.cpp:
2192         (TestShell::TestShell):
2193
2194 2012-04-24  Joe Mason  <jmason@rim.com>
2195
2196         [BlackBerry] Revert broken changes to authentication dialog
2197         https://bugs.webkit.org/show_bug.cgi?id=80135
2198
2199         Reviewed by Antonio Gomes.
2200
2201         The previous patches from this bug caused an infinite loop when using digest auth;
2202         apparently they were only tested with basic.
2203
2204         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
2205         * DumpRenderTree/blackberry/DumpRenderTreeBlackBerry.h:
2206         (WebCore):
2207
2208 2012-04-24  Zan Dobersek  <zandobersek@gmail.com>
2209
2210         [TestResultsServer] Add a unit test for changes in r114868
2211         https://bugs.webkit.org/show_bug.cgi?id=84713
2212
2213         Reviewed by Ojan Vafai.
2214
2215         Add a new unit test to test the isToTWebKit attribute of a BuilderGroup object
2216         that is created for either DEPS or ToT builder groups. The unit test covers changes
2217         made in r114868.
2218
2219         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2220         (testBuilderGroupIsToTWebKitAttribute.g_handleBuildersListLoaded):
2221         (testBuilderGroupIsToTWebKitAttribute.assertEquals):
2222
2223 2012-04-24  Brady Eidson  <beidson@apple.com>
2224
2225         <rdar://problem/10120155> and https://bugs.webkit.org/show_bug.cgi?id=82664
2226          Need DOMWindow mechanism to supplement UserScripts for page cache notifications
2227
2228         Reviewed by Sam Weinig.
2229
2230         Add a new bundle test that stresses the basics of the WKBundleDOMWindowExtension API.
2231
2232         It makes sure new extensions can be added at the right times, and that all of the expected
2233         callbacks are made.
2234
2235         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic.cpp: Added.
2236         (TestWebKitAPI::didReceiveMessageFromInjectedBundle): Record all messages sent from the bundle.
2237         (TestWebKitAPI::TEST): Run a page through some basic exercises, then compared messages sent from the
2238            bundle to a set of known expected messages.
2239
2240         Listen to all the new client callbacks, create some extensions, and monitor their lifetimes:
2241         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp: Added.
2242         (TestWebKitAPI::DOMWindowExtensionBasic::DOMWindowExtensionBasic):
2243         (TestWebKitAPI::DOMWindowExtensionBasic::frameLoadFinished):
2244         (TestWebKitAPI::DOMWindowExtensionBasic::sendExtensionStateMessage):
2245         (TestWebKitAPI::DOMWindowExtensionBasic::initialize):
2246         (TestWebKitAPI::DOMWindowExtensionBasic::didCreatePage):
2247         (TestWebKitAPI::DOMWindowExtensionBasic::updateExtensionStateRecord):
2248         (TestWebKitAPI::DOMWindowExtensionBasic::sendBundleMessage):
2249         (TestWebKitAPI::DOMWindowExtensionBasic::didCreateGlobalObjectForFrame):
2250         (TestWebKitAPI::DOMWindowExtensionBasic::willDisconnectDOMWindowExtensionFromGlobalObject):
2251         (TestWebKitAPI::DOMWindowExtensionBasic::didReconnectDOMWindowExtensionToGlobalObject):
2252         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyGlobalObjectForDOMWindowExtension):
2253         (TestWebKitAPI::didFinishLoadForFrameCallback):
2254         (TestWebKitAPI::didCreateGlobalObjectForFrameCallback):
2255         (TestWebKitAPI::willDisconnectDOMWindowExtensionFromGlobalObjectCallback):
2256         (TestWebKitAPI::didReconnectDOMWindowExtensionToGlobalObjectCallback):
2257         (TestWebKitAPI::willDestroyGlobalObjectForDOMWindowExtensionCallback):
2258
2259         * TestWebKitAPI/Tests/WebKit2/InjectedBundleBasic.cpp:
2260         (TestWebKitAPI::didReceiveMessageFromInjectedBundle): Make this method static, as it always should have been.
2261
2262         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2263
2264 2012-04-24  Carlos Garcia Campos  <cgarcia@igalia.com>
2265
2266         [GTK] Build and run TestWebKitAPI unit tests
2267         https://bugs.webkit.org/show_bug.cgi?id=84325
2268
2269         Reviewed by Philippe Normand.
2270
2271         We only build WTF unit tests for now. WebKit2 tests will be
2272         enabled in a follow up patch.
2273
2274         * Scripts/run-gtk-tests:
2275         (TestRunner): Add TestWebKitAPI/WTF to the list of directories
2276         containing unit tests.
2277         (TestRunner._run_test_glib): Run glib tests using gtester.
2278         (TestRunner._run_test_google): Run google tests.
2279         (TestRunner._run_test): Use _run_test_glib or _run_test_google
2280         depending on whether given test is glib or google test
2281         * TestWebKitAPI/GNUmakefile.am: Added.
2282         * TestWebKitAPI/config.h: Use #ifdef instead of # to fix a compile
2283         warning with -Wundef. Include WebKit2 headers only when building
2284         WebKit2.
2285
2286 2012-04-24  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
2287
2288         [Qt][WK2] Convert touch-point area
2289         https://bugs.webkit.org/show_bug.cgi?id=84434
2290
2291         Reviewed by Kenneth Rohde Christiansen.
2292
2293         Fix failing test after r114917 by also adding the new EventSender API
2294         to WebKitTestRunner.
2295
2296         * WebKitTestRunner/EventSenderProxy.h:
2297         (EventSenderProxy):
2298         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
2299         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
2300         (WTR::EventSendingController::setTouchPointRadius):
2301         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
2302         (EventSendingController):
2303         * WebKitTestRunner/TestController.cpp:
2304         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
2305         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
2306         (WTR::EventSenderProxy::addTouchPoint):
2307         (WTR::EventSenderProxy::updateTouchPoint):
2308         (WTR::EventSenderProxy::setTouchPointRadius):
2309         (WTR::EventSenderProxy::clearTouchPoints):
2310
2311 2012-04-24  Mikhail Naganov  <mnaganov@chromium.org>
2312
2313         [Chromium] Unreviewed test expectations update.
2314         https://bugs.webkit.org/show_bug.cgi?id=84726
2315
2316         Mark webkitpy.layout_tests.servers.http_server_unittest.TestHttpServer.test_start_cmd as failing on Windows.
2317
2318         * Scripts/webkitpy/test/test_finder.py:
2319         (TestFinder._default_names):
2320
2321 2012-04-24  Christophe Dumez  <christophe.dumez@intel.com>
2322
2323         [EFL] Enable Web Timing
2324         https://bugs.webkit.org/show_bug.cgi?id=84705
2325
2326         Reviewed by Tony Gentilcore.
2327
2328         Enable web timing by default whenever build-webkit is used.
2329
2330         * Scripts/build-webkit:
2331
2332 2012-04-23  Eriq Augustine  <eaugusti@chromium.org>
2333
2334         Return value from executed script in Chromium.
2335         https://bugs.webkit.org/show_bug.cgi?id=79851
2336
2337         Reviewed by Adam Barth.
2338
2339         Adding a function to LayoutTestController that evaluates a script in
2340         an isolated wold an returns the result of the evaluation.
2341
2342         * DumpRenderTree/LayoutTestController.cpp:
2343         (evaluateScriptInIsolatedWorldAndReturnValue):
2344         (LayoutTestController::staticFunctions):
2345         * DumpRenderTree/LayoutTestController.h:
2346         (LayoutTestController):
2347         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
2348         (LayoutTestController::evaluateScriptInIsolatedWorldAndReturnValue):
2349         * DumpRenderTree/chromium/LayoutTestController.cpp:
2350         (LayoutTestController::LayoutTestController):
2351         (LayoutTestController::setXSSAuditorEnabled):
2352         (LayoutTestController::evaluateScriptInIsolatedWorldAndReturnValue):
2353         * DumpRenderTree/chromium/LayoutTestController.h:
2354         (LayoutTestController):
2355         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2356         (LayoutTestController::evaluateScriptInIsolatedWorldAndReturnValue):
2357         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2358         (LayoutTestController::evaluateScriptInIsolatedWorldAndReturnValue):
2359         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2360         (LayoutTestController::evaluateScriptInIsolatedWorldAndReturnValue):
2361         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2362         (LayoutTestController::evaluateScriptInIsolatedWorldAndReturnValue):
2363         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2364         (LayoutTestController):
2365         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2366         (LayoutTestController::evaluateScriptInIsolatedWorldAndReturnValue):
2367         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2368         (LayoutTestController::evaluateScriptInIsolatedWorldAndReturnValue):
2369
2370 2012-04-23  Dirk Pranke  <dpranke@chromium.org>
2371
2372         run-perf-tests fail intermittently with an exception
2373         https://bugs.webkit.org/show_bug.cgi?id=83532
2374
2375         Reviewed by Ryosuke Niwa.
2376
2377         This bulletproofs port.get_option('child_processes')
2378         against the case where the option wasn't set.
2379
2380         * Scripts/webkitpy/layout_tests/port/mac.py:
2381         (MacPort._get_crash_log):
2382         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2383         (test_get_crash_log):
2384         (test_get_crash_log.fake_time_cb):
2385
2386 2012-04-23  Balazs Kelemen  <kbalazs@webkit.org>
2387
2388         [Qt] Add desktop zooming support for QQuickWebView
2389         https://bugs.webkit.org/show_bug.cgi?id=82337
2390
2391         Unreviewed typo fix after r114900.
2392
2393         * MiniBrowser/qt/BrowserWindow.cpp:
2394         (BrowserWindow::BrowserWindow): Fix the mistyped
2395         value of a zoom level.
2396
2397 2012-04-23  Hao Zheng  <zhenghao@chromium.org>
2398
2399         [chromium-android] Get rid of the trick that replaces system fallback font with DejaVuSans.
2400         https://bugs.webkit.org/show_bug.cgi?id=84580
2401
2402         Reviewed by Ojan Vafai.
2403
2404         Android's first fallback font is changed to DroidSansArabic.ttf in JB.
2405         Take the opportunity to get rid of the trick.
2406
2407         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2408         (ChromiumAndroidPort.start_helper):
2409         (ChromiumAndroidPort._push_fonts):
2410
2411 2012-04-23  Dirk Pranke  <dpranke@chromium.org>
2412
2413         test-webkitpy is failing
2414         https://bugs.webkit.org/show_bug.cgi?id=84617
2415
2416         Reviewed by Eric Seidel.
2417
2418         * Scripts/webkitpy/layout_tests/port/http_lock_unittest.py:
2419         (HttpLockTestWithRealFileSystem.clean_all_lockfile):
2420
2421 2012-04-23  Ojan Vafai  <ojan@chromium.org>
2422
2423         Split up flakiness_dashboard.html into multiple files. This is a step in the direction
2424         of making the code more maintainable (e.g. being able to use qunit for testing).
2425
2426         * TestResultServer/static-dashboards/dashboard_base.js:
2427         * TestResultServer/static-dashboards/flakiness_dashboard.css: Added.
2428         * TestResultServer/static-dashboards/flakiness_dashboard.html:
2429         * TestResultServer/static-dashboards/flakiness_dashboard.js: Copied from Tools/TestResultServer/static-dashboards/flakiness_dashboard.html.
2430         * TestResultServer/static-dashboards/flakiness_dashboard_tests.css: Added.
2431         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js: Renamed from Tools/TestResultServer/static-dashboards/flakiness_dashboard_tests.js.
2432
2433 2012-04-23  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
2434
2435         [Qt][WK2] Convert touch-point area.
2436         https://bugs.webkit.org/show_bug.cgi?id=84434
2437
2438         Reviewed by Kenneth Rohde Christiansen.
2439
2440         Add feature to set touch-point radius for emitted touch-events.
2441
2442         * DumpRenderTree/qt/EventSenderQt.cpp:
2443         (EventSender::addTouchPoint):
2444         (EventSender::updateTouchPoint):
2445         (EventSender::setTouchPointRadius):
2446         (EventSender::clearTouchPoints):
2447         * DumpRenderTree/qt/EventSenderQt.h:
2448         (EventSender):
2449
2450 2012-04-23  Christophe Dumez  <christophe.dumez@intel.com>
2451
2452         [EFL] EFL's LayoutTestController does not implement elementDoesAutoCompleteForElementWithId
2453         https://bugs.webkit.org/show_bug.cgi?id=84425
2454
2455         Reviewed by Antonio Gomes.
2456
2457         Use the new helper in DumpRenderTreeSupportEfl to properly
2458         test whether an element performs autocompletion.
2459
2460         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2461         (LayoutTestController::elementDoesAutoCompleteForElementWithId):
2462
2463 2012-04-23  Csaba Osztrogonác  <ossy@webkit.org>
2464
2465         master.cfg cleanup, remove dead codes after r92792.
2466         https://bugs.webkit.org/show_bug.cgi?id=84433
2467
2468         Relanding r114881 after typo fix in UploadBuiltProduct.
2469
2470         Reviewed by Ryosuke Niwa.
2471
2472         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2473
2474 2012-04-23  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
2475
2476         [EFL][DRT] Need to support requestAnimationFrame
2477         https://bugs.webkit.org/show_bug.cgi?id=84585
2478
2479         Reviewed by Martin Robinson.
2480
2481         Making requestAnimationFrame() support a default on the EFL build.
2482
2483         * Scripts/build-webkit:
2484
2485 2012-04-23  Balazs Kelemen  <kbalazs@webkit.org>
2486
2487         [Qt] Add desktop zooming support for QQuickWebView
2488         https://bugs.webkit.org/show_bug.cgi?id=82337
2489
2490         Reviewed by Simon Hausmann.
2491
2492         Add zooming support for MiniBrowser to
2493         test the new API.
2494
2495         * MiniBrowser/qt/BrowserWindow.cpp:
2496         (BrowserWindow::BrowserWindow):
2497         (BrowserWindow::zoomIn):
2498         (BrowserWindow::zoomOut):
2499         (BrowserWindow::keyPressEvent):
2500         (BrowserWindow::wheelEvent):
2501         * MiniBrowser/qt/BrowserWindow.h:
2502         (BrowserWindow):
2503
2504 2012-04-23  Csaba Osztrogonác  <ossy@webkit.org>
2505
2506         Unreviewed rolling out r114881, because it broke upload buildstep.
2507
2508         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2509         (ConfigureBuild.__init__):
2510         (ConfigureBuild.start):
2511         (CompileWebKit.start):
2512         (UploadBuiltProduct):
2513         (UploadBuiltProduct.determineExtraFeatures):
2514         (Factory.__init__):
2515         (BuildFactory.__init__):
2516         (TestFactory.__init__):
2517         (loadBuilderConfig):
2518
2519 2012-04-23  Csaba Osztrogonác  <ossy@webkit.org>
2520
2521         master.cfg cleanup, remove dead codes after r92792.
2522         https://bugs.webkit.org/show_bug.cgi?id=84433
2523
2524         Reviewed by Ryosuke Niwa.
2525
2526         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2527
2528 2012-04-23  Milian Wolff  <mail@milianw.de>
2529
2530         [Qt] Make DRT not dump pixel results if test is calling layoutTestController.dumpAsText()
2531         https://bugs.webkit.org/show_bug.cgi?id=84327
2532
2533         Reviewed by Csaba Osztrogonác.
2534
2535         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2536         (WebCore::DumpRenderTree::dump):
2537
2538 2012-04-22  Sriram Neelakandan  <sriram.neelakandan@gmail.com>
2539
2540         [Gtk] Added MOZ_X11 build flag for TARGET_X11
2541         https://bugs.webkit.org/show_bug.cgi?id=40785
2542
2543         Reviewed by Anders Carlsson.
2544
2545         * GNUmakefile.am:
2546
2547 2012-04-22  Zan Dobersek  <zandobersek@gmail.com>
2548
2549         [TestResultsServer] Impossible to inspect builds on non-Chromium builders
2550         https://bugs.webkit.org/show_bug.cgi?id=84544
2551
2552         Reviewed by Ryosuke Niwa.
2553
2554         Use the groupEnum parameter that is passed through to the onBuilderListLoad
2555         function rather than an undefined attribute of the BuilderGroup object. This
2556         makes it again possible to inspect builds on non-Chromium builders in the
2557         test results server by not trying to always add Chromium revision links.
2558
2559         * TestResultServer/static-dashboards/builders.js:
2560
2561 2012-04-20  Martin Robinson  <mrobinson@igalia.com>
2562
2563         [GTK] Revert changes that disabled Cairo fonts metrics hinting
2564         https://bugs.webkit.org/show_bug.cgi?id=84502
2565
2566         Reviewed by Daniel Bates.
2567
2568         Re-enable Cairo metrics hinting, which seems to ensure consistent results in
2569         reference and pixel test results. This makes our font metrics a bit different
2570         than the Mac port's, but it's a decent price to pay for consistent results.
2571
2572         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2573         (initializeGtkFontSettings):
2574         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
2575         (WTR::initializeGtkSettings):
2576
2577 2012-04-04  Jer Noble  <jer.noble@apple.com>
2578
2579         apple.com top navigation bar appears inside video during full screen exit animation
2580         https://bugs.webkit.org/show_bug.cgi?id=83095
2581
2582         Reviewed by Eric Carlson.
2583
2584         Add a new LayoutTestController function, setHasCustomFullScreenBehavior(), which allows scripts to
2585         disable the standard full screen behavior, and replace it with explicit callbacks to the relevant
2586         Document methods from within the script.  This allows subtle timing bugs to be consistently reproduced
2587         from within LayoutTests.
2588
2589         Add the new methods, hasCustomFullScreenBehavior() and setHasCustomFullScreenBehavior():
2590         * DumpRenderTree/LayoutTestController.cpp:
2591         (setHasCustomFullScreenBehaviorCallback):
2592         (LayoutTestController::staticFunctions):
2593         * DumpRenderTree/LayoutTestController.h:
2594         (LayoutTestController::setHasCustomFullScreenBehavior):
2595         (LayoutTestController::hasCustomFullScreenBehavior):
2596         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2597         (WTR::InjectedBundlePage::enterFullScreenForElement):
2598         (WTR::InjectedBundlePage::exitFullScreenForElement):
2599         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2600         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
2601         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2602         (WTR::LayoutTestController::setHasCustomFullScreenBehavior):
2603         (WTR::LayoutTestController::hasCustomFullScreenBehavior):
2604         * DumpRenderTree/chromium/LayoutTestController.cpp:
2605         (LayoutTestController::LayoutTestController):
2606         (LayoutTestController::reset):
2607         * DumpRenderTree/chromium/LayoutTestController.h:
2608         (LayoutTestController::setHasCustomFullScreenBehavior):
2609         (LayoutTestController::hasCustomFullScreenBehavior):
2610
2611         Query the new methods before calling will/did/Enter/ExitFullScreen:
2612         * DumpRenderTree/chromium/WebViewHost.cpp:
2613         (WebViewHost::enterFullScreenNow):
2614         (WebViewHost::exitFullScreenNow):
2615         * DumpRenderTree/mac/UIDelegate.mm:
2616         (-[UIDelegate webView:enterFullScreenForElement:listener:]):
2617         (-[UIDelegate webView:exitFullScreenForElement:listener:]):
2618
2619 2012-04-20  Nandor Huszka  <hnandor@inf.u-szeged.hu>
2620
2621         NRWT: option --skip-pixel-test-if-no-baseline support on DRT
2622         https://bugs.webkit.org/show_bug.cgi?id=72841
2623
2624         Reviewed by Dirk Pranke
2625
2626         This patch implements support for --skip-pixel-test-if-no-baseline generically in NRWT,
2627         instead of in every DRT and WTR implementation. We add 'should_run_pixel_test'
2628         and 'reference_files' fields to the TestInput objects, and use those flags
2629         to select the right driver to use in the DriverProxy.
2630
2631         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
2632         (SingleTestRunner.__init__):
2633         (SingleTestRunner._driver_input):
2634         (SingleTestRunner._run_reftest):
2635         * Scripts/webkitpy/layout_tests/controllers/worker.py:
2636         (Worker.handle_test_list):
2637         (Worker._update_test_input):
2638         * Scripts/webkitpy/layout_tests/models/test_input.py:
2639         (TestInput):
2640         (TestInput.__init__):
2641         (TestInput.__repr__):
2642         * Scripts/webkitpy/layout_tests/port/chromium.py:
2643         (ChromiumDriver._wrapper_options):
2644         (ChromiumDriver.run_test):
2645         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2646         (ChromiumDriverTest.test_crash_log):
2647         * Scripts/webkitpy/layout_tests/port/driver.py:
2648         (DriverInput.__init__):
2649         (Driver.__init__):
2650         (DriverProxy.__init__):
2651         (DriverProxy.run_test):
2652         (DriverProxy.start):
2653         (DriverProxy.cmd_line):
2654         * Scripts/webkitpy/layout_tests/port/gtk.py:
2655         (GtkDriver._start):
2656         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
2657         (MockDRT.run):
2658         (MockDRT.input_from_line):
2659         (MockDRT.output_for_test):
2660         (MockDRT.write_test_output):
2661         (MockChromiumDRT.input_from_line):
2662         (MockChromiumDRT.output_for_test):
2663         (MockChromiumDRT.write_test_output):
2664         * Scripts/webkitpy/layout_tests/port/webkit.py:
2665         (WebKitDriver.cmd_line):
2666         (WebKitDriver.run_test):
2667         * WebKitTestRunner/TestController.cpp:
2668         (WTR::TestController::TestController):
2669         (WTR::TestController::initialize):
2670         (WTR::TestController::runTest):
2671         * WebKitTestRunner/TestController.h:
2672         (TestController):
2673         * WebKitTestRunner/TestInvocation.cpp:
2674         (WTR::TestInvocation::TestInvocation):
2675         (WTR::TestInvocation::setIsPixelTest):
2676         * WebKitTestRunner/TestInvocation.h:
2677         (TestInvocation):
2678
2679 2012-04-20  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2680
2681         [EFL] Make it possible to define offline storage quota per origin
2682         https://bugs.webkit.org/show_bug.cgi?id=84170
2683
2684         Improved LayoutTestController::setDatabaseQuota to set quota based on
2685         origin which was causing some tests to fail. Also align the default
2686         quota for testing (5MB) with others ports.
2687
2688         When dumpDatabaseCallbacks is set to true by, also dump some database
2689         information when quota is exceed, which is also required by some tests.
2690
2691         Reviewed by Antonio Gomes.
2692
2693         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2694         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
2695         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
2696         (onExceededDatabaseQuota):
2697         (drtViewAdd):
2698         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2699         (LayoutTestController::setDatabaseQuota):
2700
2701 2012-04-20  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2702
2703         [EFL] [DRT] Send double click event from EventSender
2704         https://bugs.webkit.org/show_bug.cgi?id=82867
2705
2706         Reviewed by Daniel Bates.
2707
2708         Add support for EFL DumpRenderTree EventSender to send
2709         double click events.
2710
2711         * DumpRenderTree/efl/EventSender.cpp:
2712         (sendMouseEvent):
2713
2714 2012-04-20  Sergio Villar Senin  <svillar@igalia.com>
2715
2716         [GTK] Enable Web Timing
2717         https://bugs.webkit.org/show_bug.cgi?id=42432
2718
2719         Reviewed by Gustavo Noronha Silva.
2720
2721         Build web timing by default whenever build-webkit is used.
2722
2723         * Scripts/build-webkit:
2724
2725 2012-04-20  Kamil Blank  <k.blank@samsung.com>
2726
2727         [EFL] Add missing reset for setAuthorAndUserStylesEnabled inside DRT.
2728         https://bugs.webkit.org/show_bug.cgi?id=84430
2729
2730         Reviewed by Daniel Bates.
2731
2732         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2733         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
2734
2735 2012-04-20  Ryosuke Niwa  <rniwa@webkit.org>
2736
2737         REGRESSION(r112781): Windows build slaves are no longer including .pdb files in archives handed to testers
2738         https://bugs.webkit.org/show_bug.cgi?id=84424
2739
2740         Reviewed by Mark Rowe.
2741
2742         Let Windows and Chromium have their own list of files to copy.
2743
2744         * BuildSlaveSupport/built-product-archive:
2745         (removeDirectoryIfExists):
2746         (copyBuildFiles):
2747         (archiveBuiltProduct):
2748
2749 2012-04-20  Kamil Blank  <k.blank@samsung.com>
2750
2751         [EFL] Add setting API for author and user styles.
2752         https://bugs.webkit.org/show_bug.cgi?id=82860
2753
2754         Reviewed by Daniel Bates.
2755
2756         Implementation of setAuthorAndUserStylesEnabled.
2757
2758         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2759         (LayoutTestController::setAuthorAndUserStylesEnabled):
2760
2761 2012-04-19  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2762
2763         [EFL] Missing keycode translation for space key
2764         https://bugs.webkit.org/show_bug.cgi?id=84153
2765
2766         Reviewed by Antonio Gomes.
2767
2768         Set keyname for space key before dispatching event.
2769
2770         * DumpRenderTree/efl/EventSender.cpp:
2771         (keyNameFromJSValue):
2772
2773 2012-04-19  Carlos Garcia Campos  <cgarcia@igalia.com>
2774
2775         [GTK] Allow to use WebInspector in GtkLauncher
2776         https://bugs.webkit.org/show_bug.cgi?id=84143
2777
2778         Reviewed by Martin Robinson.
2779
2780         * GNUmakefile.am: Add new files to compilation.
2781         * GtkLauncher/LauncherInspectorWindow.c: Added.
2782         (launcherInspectorWindowFinalize):
2783         (launcher_inspector_window_init):
2784         (launcher_inspector_window_class_init):
2785         (inspectedURIChanged): Update the window title based on current
2786         inspected URI.
2787         (showInspectorWindow): Show inspector window.
2788         (closeInspectorWindow): Hide inspector window.
2789         (launcherInspectorWindowNew): Create an inspector window.
2790         (launcherInspectorWindowGetWebView): Return the inspected web view.
2791         * GtkLauncher/LauncherInspectorWindow.h: Added.
2792         * GtkLauncher/main.c:
2793         (inspectorInspectWebViewCb): Create an inspector window.
2794         (createBrowser): Connect to inspect-web-view signal of the view
2795         inspector.
2796
2797 2012-04-19  Jeremy Mao  <yujie.mao@intel.com>
2798
2799         Media Stream API: Piece of code cleanup for the chromium port
2800         https://bugs.webkit.org/show_bug.cgi?id=84328
2801
2802         Reviewed by Kent Tamura.
2803
2804         WebUserMediaClientMock use an old requestUserMedia interface and
2805         Replace it with the new interface.
2806
2807         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp:
2808         (WebKit::WebUserMediaClientMock::requestUserMedia):
2809         Remove the old requestUserMedia and replace it with the new.
2810         * DumpRenderTree/chromium/WebUserMediaClientMock.h:
2811         (WebUserMediaClientMock):
2812
2813 2012-04-19  Christophe Dumez  <christophe.dumez@intel.com>
2814
2815         [EFL] DRT console messages should contain only the file name part of local URIs
2816         https://bugs.webkit.org/show_bug.cgi?id=84236
2817
2818         Reviewed by Kentaro Hara.
2819
2820         Process console messages before printing them in EFL's DumpRenderTree
2821         them so that they contain only the file name part of local URIs. This
2822         is what is expected by test cases.
2823
2824         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
2825         (urlSuitableForTestResult):
2826         (onConsoleMessage):
2827
2828 2012-04-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2829
2830         [EFL] [DRT] Should create offline storage inside DUMPRENDERTREE_TEMP
2831         https://bugs.webkit.org/show_bug.cgi?id=84169
2832
2833         All databases were being created inside ~/.webkit which was causing
2834         flaky tests when running multiple tests in parallel, specially timeouts
2835         due to concurrency.
2836
2837         Reviewed by Eric Seidel.
2838
2839         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2840         (DumpRenderTreeChrome::initialize):
2841
2842 2012-04-19  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
2843
2844         [GTK] 3 fast/ layout tests failing after upgrading libsoup to 2.37.92
2845         https://bugs.webkit.org/show_bug.cgi?id=82484
2846
2847         Reviewed by Martin Robinson.
2848
2849         Bumping libsoup to version 2.38.1 which fixes the above WebKit bug.
2850
2851         * efl/jhbuild.modules:
2852         * gtk/jhbuild.modules:
2853
2854 2012-04-19  Lucas Forschler  <lforschler@apple.com>
2855
2856         Move apple-xserve-10 from Lion Test to SnowLeopard Debug Build
2857         
2858         * BuildSlaveSupport/build.webkit.org-config/config.json:
2859
2860 2012-04-19  Christophe Dumez  <christophe.dumez@intel.com>
2861
2862         [EFL] dispatchDidCancelClientRedirect needs implementation
2863         https://bugs.webkit.org/show_bug.cgi?id=83669
2864
2865         Reviewed by Antonio Gomes.
2866
2867         Catch the new ewk_frame "redirect,cancelled" signal in EFL's DRT and
2868         and print out information that is expected by several test cases if
2869         LayoutTestController's dumpFrameLoadCallback() returns true.
2870
2871         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2872         (DumpRenderTreeChrome::createView):
2873         (DumpRenderTreeChrome::onFrameCreated):
2874         (DumpRenderTreeChrome::onFrameRedirectCancelled):
2875         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
2876         (DumpRenderTreeChrome):
2877
2878 2012-04-18  Hironori Bono  <hbono@chromium.org>
2879
2880         Split SpellChecker::didCheck() to SpellChecker::didCheckSucceeded() and SpellChecker::didCheckCanceled()
2881         https://bugs.webkit.org/show_bug.cgi?id=83748
2882
2883         Reviewed by Ryosuke Niwa.
2884
2885         This change replaces a call for SpellChecker::didCheck() with one for
2886         SpellChecker::didCheckCenceled() because didCheck() has been split into
2887         didCheckSucceeded() and didCheckCanceled().
2888
2889         * DumpRenderTree/chromium/WebViewHost.cpp:
2890         (WebViewHost::requestCheckingOfText): Replaced didCheck with didCheckCanceled.
2891
2892 2012-04-18  Dirk Pranke  <dpranke@chromium.org>
2893
2894         Fix failure in test-webkitpy after r114583.
2895
2896         Unreviewed, build fix.
2897
2898         Updating the call site in r114583 w/o updating the test caused
2899         the test to fail. Somehow I forgot to run the unit tests :(.
2900
2901         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2902         (ChromiumDriverTest.test_crash_log):
2903
2904 2012-04-18  Filip Pizlo  <fpizlo@apple.com>
2905
2906         JSC testing should have complete coverage over typed array types
2907         https://bugs.webkit.org/show_bug.cgi?id=84302
2908
2909         Reviewed by Geoff Garen.
2910         
2911         Made run-fast-jsc print something and a newline if jsc threw an exception or
2912         otherwise returned with an error code, but did not crash.
2913
2914         * Scripts/run-fast-jsc:
2915
2916 2012-04-18  Dirk Pranke  <dpranke@chromium.org>
2917
2918         Fix crash in chromium.py after r114877.
2919
2920         Unreviewed, build fix.
2921
2922         Missed updating one call site for get_crash_log() :(.
2923
2924         * Scripts/webkitpy/layout_tests/port/chromium.py:
2925         (ChromiumDriver.run_test):
2926
2927 2012-04-18  Philippe Normand  <pnormand@igalia.com>
2928
2929         [GTK] NRWT bails out if more than one X server runs on the machine
2930         https://bugs.webkit.org/show_bug.cgi?id=84276
2931
2932         Reviewed by Dirk Pranke.
2933
2934         Set up GTK driver's Xvfb display number based on Xorg servers
2935         currently running.
2936
2937         * Scripts/webkitpy/layout_tests/port/gtk.py:
2938         (GtkDriver._start):
2939         (GtkDriver._start.x_filter):
2940
2941 2012-04-18  Dirk Pranke  <dpranke@chromium.org>
2942
2943         run-webkit-tests picked up an old crash log
2944         https://bugs.webkit.org/show_bug.cgi?id=84196
2945
2946         Reviewed by Ryosuke Niwa.
2947
2948         Modify the crash log-finding code so that we can
2949         limit our search to logs newer than X - this prevents
2950         us from finding old crash logs for a given process/pid
2951         combination even when the PIDs might've wrapped.
2952  
2953         Also fix a bug in filesystem_mock where files_under() would
2954         return paths that didn't exist.
2955
2956         * Scripts/webkitpy/common/system/crashlogs.py:
2957         (CrashLogs.find_newest_log):
2958         (CrashLogs._find_newest_log_darwin):
2959         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
2960         (CrashLogsTest.test_find_log_darwin):
2961         (CrashLogsTest.test_find_log_darwin.bad_read):
2962         (CrashLogsTest):
2963         * Scripts/webkitpy/common/system/filesystem_mock.py:
2964         (MockFileSystem.files_under):
2965         * Scripts/webkitpy/layout_tests/port/base.py:
2966         (Port._get_crash_log):
2967         * Scripts/webkitpy/layout_tests/port/gtk.py:
2968         (GtkPort._get_crash_log):
2969         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
2970         * Scripts/webkitpy/layout_tests/port/mac.py:
2971         (MacPort._get_crash_log):
2972         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2973         (PortTestCase.test_get_crash_log):
2974         * Scripts/webkitpy/layout_tests/port/webkit.py:
2975         (WebKitDriver.run_test):
2976
2977 2012-04-18  Dirk Pranke  <dpranke@chromium.org>
2978
2979         nrwt: support additional test expectations files
2980         https://bugs.webkit.org/show_bug.cgi?id=84222
2981
2982         Reviewed by Ojan Vafai.
2983
2984         This patch adds an --additional-expectations flag that can be
2985         used to point to more files that will override the default set
2986         of expectations. This will allow us to remove the
2987         'google_chrome' ports and eventually help to clean up override
2988         handling in the chromium ports and unify expectations files and
2989         Skipped files; this can also be used to specify overrides for
2990         tests that run differently on a given machine.
2991
2992         * Scripts/webkitpy/layout_tests/port/base.py:
2993         (Port.test_expectations_overrides):
2994         * Scripts/webkitpy/layout_tests/port/chromium.py:
2995         (ChromiumPort.test_expectations_overrides):
2996         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2997         (parse_args):
2998         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2999         (MainTest.test_additional_expectations):
3000
3001 2012-04-18  Jer Noble  <jer.noble@apple.com>
3002
3003         Lion Production Test failing with error: "Failed to stop httpd: pid file still exists"
3004         https://bugs.webkit.org/show_bug.cgi?id=84177
3005
3006         Reviewed by Dirk Pranke.
3007
3008         Try to delete the pid file left by Apache in the case it was killed forcefully.
3009
3010         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
3011         (LayoutTestApacheHttpd._stop_running_server):
3012
3013 2012-04-18  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
3014
3015         Clean-up WheelEvent Conversion.
3016         https://bugs.webkit.org/show_bug.cgi?id=84243
3017
3018         Reviewed by Simon Hausmann.
3019         
3020         Calculate proper wheel-delta from pixel-delta given.
3021
3022         * DumpRenderTree/qt/EventSenderQt.cpp:
3023         (EventSender::EventSender):
3024         (EventSender::mouseScrollBy):
3025         (EventSender::continuousMouseScrollBy):
3026
3027 2012-04-18  Sheriff Bot  <webkit.review.bot@gmail.com>
3028
3029         Unreviewed, rolling out r114506.
3030         http://trac.webkit.org/changeset/114506
3031         https://bugs.webkit.org/show_bug.cgi?id=84254
3032
3033         Seems to be making gtk's DumpRenderTree crash occasionally
3034         (Requested by tomz on #webkit).
3035
3036         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3037         (resetDefaultsToConsistentValues):
3038
3039 2012-04-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3040
3041         [CMake] Use jsc target instead of ONLY_BUILD_JAVASCRIPTCORE
3042         https://bugs.webkit.org/show_bug.cgi?id=84229
3043
3044         CMake caches ONLY_BUILD_JAVASCRIPTCORE=1 and will reuse the setting for
3045         subsequent builds, stopping WebCore from being built ever again.
3046
3047         Instead of trying to emulate a build target using command line options,
3048         just build the existing jsc target directly.
3049
3050         Reviewed by Rob Buis.
3051
3052         * Scripts/build-jsc:
3053
3054 2012-04-18  Simon Pena  <spena@igalia.com>
3055
3056         [GTK] r114021 triggered media flakyness
3057         https://bugs.webkit.org/show_bug.cgi?id=83874
3058
3059         Reviewed by Philippe Normand.
3060
3061         Now that there is gtk API to update the
3062         MediaPlaybackRequiresUserGesture property, set it to false from the
3063         DumpRenderTree tool when resetting defaults to consistent values.
3064
3065         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3066         (resetDefaultsToConsistentValues):
3067
3068 2012-04-18  Vivek Galatage  <vivekgalatage@gmail.com>
3069
3070         Web Inspector: Disabling Inspector causes build failure on Windows
3071         https://bugs.webkit.org/show_bug.cgi?id=83557
3072
3073         Reviewed by Pavel Feldman.
3074
3075         The calls to inspector APIs are not put under ENABLE(INSPECTOR) guard.
3076         This change brings all such API calls under the guard.
3077
3078         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3079         (WTR::LayoutTestController::showWebInspector):
3080         (WTR::LayoutTestController::closeWebInspector):
3081         (WTR::LayoutTestController::evaluateInWebInspector):
3082         (WTR::LayoutTestController::setJavaScriptProfilingEnabled):
3083         * WebKitTestRunner/TestInvocation.cpp:
3084         (WTR::TestInvocation::invoke):
3085
3086 2012-04-18  Mario Sanchez Prada  <msanchez@igalia.com>
3087
3088         [GTK] atk/textChangedNotifications API test fails
3089         https://bugs.webkit.org/show_bug.cgi?id=84010
3090
3091         Unreviewed, unskipping test after bug being fixed by r114311.
3092
3093         * Scripts/run-gtk-tests:
3094         (TestRunner): Unskip now passing test.
3095
3096 2012-04-17  Jacky Jiang  <zhajiang@rim.com>
3097
3098         [BlackBerry] Viewport metatag doesn't disable double-tap zoom
3099         https://bugs.webkit.org/show_bug.cgi?id=84199
3100
3101         Reviewed by George Staikos.
3102
3103         Revert a part of the change master_32/SHA:fa0f8ee9 and use
3104         WebPage::setUserScalable instead.
3105
3106         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
3107         (BlackBerry::WebKit::DumpRenderTree::resetToConsistentStateBeforeTesting):
3108
3109 2012-04-17  Dirk Pranke  <dpranke@chromium.org>
3110
3111         build-webkit: make ninja work with chromium linux
3112         https://bugs.webkit.org/show_bug.cgi?id=84213
3113
3114         Reviewed by Eric Seidel.
3115
3116         Fix typos so that build-webkit will use ninja to
3117         build chromium if a build.ninja file exists and is up-to-date.
3118
3119         * Scripts/webkitdirs.pm:
3120         (determineIsChromiumNinja):
3121
3122 2012-04-17  Ojan Vafai  <ojan@chromium.org>
3123
3124         Add caching to the flakiness dashboard json files
3125         https://bugs.webkit.org/show_bug.cgi?id=84198
3126
3127         Reviewed by Dirk Pranke.
3128
3129         * TestResultServer/handlers/testfilehandler.py:
3130         (GetFile._get_file_list):
3131         (GetFile._get_file_content):
3132         (GetFile._get_file_content_from_key):
3133         (GetFile._get_test_list_json):
3134         (GetFile._serve_json):
3135         Add a Last-Modified header and handle If-Modified-Since headers.
3136         (GetFile.get):
3137
3138 2012-04-17  Anders Carlsson  <andersca@apple.com>
3139
3140         Need a client callback for when the user tries to interact with an already unresponsive page
3141         https://bugs.webkit.org/show_bug.cgi?id=84201
3142         <rdar://problem/11140862>
3143
3144         Reviewed by Andreas Kling.
3145
3146         Add new client functions.
3147
3148         * MiniBrowser/mac/BrowserWindowController.m:
3149         (-[BrowserWindowController awakeFromNib]):
3150         * WebKitTestRunner/TestController.cpp:
3151         (WTR::TestController::initialize):
3152
3153 2012-04-17  Dirk Pranke  <dpranke@chromium.org>
3154
3155         nrwt: handle stopping layout test helper after a ctrl-c cleanly on the chromium port
3156         https://bugs.webkit.org/show_bug.cgi?id=84190
3157
3158         Reviewed by Ojan Vafai.
3159
3160         LayoutTestHelper would exit when the user typed ctrl-c, but we
3161         would attempt to write to it to shut it down, generating an
3162         IOError.
3163
3164         * Scripts/webkitpy/layout_tests/port/chromium.py:
3165         (ChromiumPort.stop_helper):
3166
3167 2012-04-17  Dirk Pranke  <dpranke@chromium.org>
3168
3169         rename --experimental-fully-parallel
3170         https://bugs.webkit.org/show_bug.cgi?id=84188
3171
3172         Reviewed by Ojan Vafai.
3173
3174         I thought this was done long ago; rename
3175         --experimental-fully-parallel to just --fully-parallel and
3176         remove the --no- switch that AFAIK no one uses.
3177
3178         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3179         (Manager._run_tests):
3180         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3181         (parse_args):
3182
3183 2012-04-17  Christophe Dumez  <christophe.dumez@intel.com>
3184
3185         EFL's LayoutTestController does not implement callShouldCloseOnWebView
3186         https://bugs.webkit.org/show_bug.cgi?id=82301
3187
3188         Reviewed by Martin Robinson.
3189
3190         Implement callShouldCloseOnWebView in EFL's LayoutTestController by
3191         calling shouldClose() on the FrameLoader, allowing the related test
3192         to be removed from the skip list.
3193
3194         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3195         (LayoutTestController::callShouldCloseOnWebView):
3196
3197 2012-04-17  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
3198
3199         [EFL] Add setting API to enable/disable XSSAuditor
3200         https://bugs.webkit.org/show_bug.cgi?id=83281
3201
3202         Add missing implementation setXSSAuditorEnabled to EFL's LayoutTestController
3203         in order to unskip tests in http/tests/security/xssAuditor. Also, catch the signal 
3204         'xss,detected' in DRT to enable a test, which is expecting a line containing
3205         'didDetectXSS' in the output when reflected XSS is encountered in the page.
3206
3207         Reviewed by Antonio Gomes.
3208
3209         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3210         (DumpRenderTreeChrome::createView):
3211         (DumpRenderTreeChrome::onFrameCreated):
3212         (DumpRenderTreeChrome::onDidDetectXSS):
3213         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
3214         (DumpRenderTreeChrome):
3215         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3216         (LayoutTestController::setXSSAuditorEnabled):
3217
3218 2012-04-17  Nandor Huszka  <hnandor@inf.u-szeged.hu>
3219
3220         [Qt] JSC build should handle --no-webkit2 option to avoid unwanted clean-builds
3221         https://bugs.webkit.org/show_bug.cgi?id=77533
3222
3223         Reviewed by Simon Hausmann.
3224
3225         Pick up the --no-webkit2 option from BUILD_WEBKIT_ARGS and pass on
3226         the corresponding parameter to the buildQMakeProjects function.
3227
3228         * Scripts/build-jsc:
3229
3230 2012-04-17  Christophe Dumez  <christophe.dumez@intel.com>
3231
3232         [EFL] DRT should support LayoutTestController's dumpFrameLoadCallbacks()
3233         https://bugs.webkit.org/show_bug.cgi?id=83030
3234
3235         Reviewed by Antonio Gomes.
3236
3237         Implement support for LayoutTestController's dumpFrameLoadCallbacks()
3238         in EFL's DumpRenderTree. Make use of the new frame argument passed to
3239         the "load,started" and "load,finished" signals.
3240
3241         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3242         (DumpRenderTreeChrome::createView):
3243         (DumpRenderTreeChrome::onLoadStarted):
3244         (DumpRenderTreeChrome::processWork):
3245         (DumpRenderTreeChrome::topLoadingFrameLoadFinished):
3246         (DumpRenderTreeChrome::onWebViewOnloadEvent):
3247         (DumpRenderTreeChrome::onInsecureContentRun):
3248         (DumpRenderTreeChrome::onInsecureContentDisplayed):
3249         (DumpRenderTreeChrome::onFrameCreated):
3250         (DumpRenderTreeChrome::onFrameProvisionalLoad):
3251         (DumpRenderTreeChrome::onFrameLoadCommitted):
3252         (DumpRenderTreeChrome::onFrameLoadFinished):
3253         (DumpRenderTreeChrome::onFrameLoadError):
3254         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
3255         (DumpRenderTreeChrome):
3256
3257 2012-04-17  Sheriff Bot  <webkit.review.bot@gmail.com>
3258
3259         Unreviewed, rolling out r114376.
3260         http://trac.webkit.org/changeset/114376
3261         https://bugs.webkit.org/show_bug.cgi?id=84155
3262
3263         Wrong patch version committed, asked by the patch author.
3264         (Requested by rakuco on #webkit).
3265
3266         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3267         (DumpRenderTreeChrome::createView):
3268         (DumpRenderTreeChrome::processWork):
3269         (DumpRenderTreeChrome::onLoadFinished):
3270         (DumpRenderTreeChrome::onWillSendRequest):
3271         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
3272         (DumpRenderTreeChrome):
3273
3274 2012-04-17  Christophe Dumez  <christophe.dumez@intel.com>
3275
3276         [EFL] DRT should support LayoutTestController's dumpFrameLoadCallbacks()
3277         https://bugs.webkit.org/show_bug.cgi?id=83030
3278
3279         Reviewed by Antonio Gomes.
3280
3281         Implement support for LayoutTestController's dumpFrameLoadCallbacks()
3282         in EFL's DumpRenderTree.
3283
3284         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3285         (DumpRenderTreeChrome::createView):
3286         (DumpRenderTreeChrome::processWork):
3287         (DumpRenderTreeChrome::topLoadingFrameLoadFinished):
3288         (DumpRenderTreeChrome::onWebViewOnloadEvent):
3289         (DumpRenderTreeChrome::onInsecureContentRun):
3290         (DumpRenderTreeChrome::onInsecureContentDisplayed):
3291         (DumpRenderTreeChrome::onFrameCreated):
3292         (DumpRenderTreeChrome::onFrameProvisionalLoad):
3293         (DumpRenderTreeChrome::onFrameLoadCommitted):
3294         (DumpRenderTreeChrome::onFrameLoadFinished):
3295         (DumpRenderTreeChrome::onFrameLoadError):
3296         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
3297         (DumpRenderTreeChrome):
3298
3299 2012-04-17  Michael Brüning  <michael.bruning@nokia.com>
3300
3301         [Qt] Fix builds with force_static_libs_as_shared for non-linux builds.
3302         https://bugs.webkit.org/show_bug.cgi?id=83035
3303
3304         Reviewed by Simon Hausmann.
3305
3306         The force_static_libs_as_shared feature for the Qt port will remove
3307         hide_symbols from Qt's CONFIG in order to make them visible to other
3308         libraries again.
3309
3310         Due to an evaluation order problem in qmake, the flags for the C++ compiler
3311         that are set by adding hide_symbols to the CONFIG variables get applied
3312         to qmake's C++ compiler flags directly and hence trying to remove these
3313         later will not work and cause build failures when using the force_static_libs_as_shared
3314         feature of the Qt port.
3315
3316         This is fixed by removing the flags that hide the exports from the compiler flags directly.
3317
3318         * qmake/mkspecs/features/force_static_libs_as_shared.prf:
3319
3320 2012-04-16  Ryosuke Niwa  <rniwa@webkit.org>
3321
3322         Remove CreateWebKitBuildDirectory step to fix Chromium Windows Perf bot
3323         https://bugs.webkit.org/show_bug.cgi?id=84068
3324
3325         Reviewed by Eric Seidel.
3326
3327         Remove CreateWebKitBuildDirectory from master.cfg since download-built-product
3328         automatically creates the build directory if needed.
3329
3330         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3331         (DownloadAndPerfTestFactory.__init__):
3332         (DownloadAndPerfTestWebKit2Factory.__init__):
3333
3334 2012-04-16  Gustavo Noronha Silva  <gns@gnome.org>
3335
3336         [GTK] Use configure instead of autogen.sh for gtk+
3337         https://bugs.webkit.org/show_bug.cgi?id=84112
3338
3339         Rubber-stamped by Martin Robinson.
3340
3341         * gtk/jhbuild.modules: GTK+'s autogen requires gobject-introspection;
3342         we don't want to depend on that for now, so run configure instead - we
3343         can, since we are using a tarball
3344
3345 2012-04-16  Gustavo Noronha Silva  <gns@gnome.org>
3346
3347         Unreviewed, rolling out r114322.
3348         http://trac.webkit.org/changeset/114322
3349         https://bugs.webkit.org/show_bug.cgi?id=84096
3350
3351         Didn't make the debug bot happy, so back out
3352
3353         * gtk/jhbuild.modules:
3354
3355 2012-04-16  Gustavo Noronha Silva  <gns@gnome.org>
3356
3357         [GTK] Try to fix debug bot by adding gobject-introspection to the build
3358         https://bugs.webkit.org/show_bug.cgi?id=84096
3359
3360         Unreviewed speculative build fix.
3361
3362         * gtk/jhbuild.modules: Added gobject-introspection, since modules need
3363         an up-to-date introspection m4 for running autogen.
3364
3365 2012-04-16  Gustavo Noronha Silva  <gns@gnome.org>
3366
3367         [GTK] Add pango to jhbuild
3368         https://bugs.webkit.org/show_bug.cgi?id=84086
3369
3370         Unreviewed build fix.
3371
3372         * gtk/jhbuild.modules: Add pango's latest stable release, 1.30, to our
3373         jhbuild environment. It's required for latest stable GTK+ and some
3374         systems may lack it.
3375
3376 2012-04-16  Gustavo Noronha Silva  <gns@gnome.org>
3377
3378         [GTK] fonts jhbuild module needs an empty clean target
3379         https://bugs.webkit.org/show_bug.cgi?id=84085
3380
3381         Rubber-stamped by Martin Robinson.
3382
3383         * gtk/jhbuild.modules: update to webkitgtk-test-fonts 0.0.2, which
3384         includes a phony clean target that should make warnings and errors go
3385         away.
3386
3387 2012-04-16  Gustavo Noronha Silva  <gns@gnome.org>
3388
3389         [GTK] Bump dependency on GTK+ 3.x to match reality
3390         https://bugs.webkit.org/show_bug.cgi?id=84060
3391
3392         Rubber-stamped by Martin Robinson.
3393
3394         * gtk/jhbuild.modules: Bump GTK+ to 3.4.0, plus glib and gdk-pixbuf
3395         to their GNOME 3.4 versions.
3396         * gtk/patches/gtk+-3.0.12-gtkprinter.patch: Removed. This patch
3397         is applied upstream.
3398
3399 2012-04-16  Vincent Scheib  <scheib@chromium.org>
3400
3401         Adding Contributors Brian Salomon and Tom Hudson to committers.py.
3402
3403         Unreviewed.
3404
3405         * Scripts/webkitpy/common/config/committers.py:
3406
3407 2012-04-16  Jonathan Dong  <jonathan.dong@torchmobile.com.cn>
3408
3409         [BlackBerry] Add ability for DRT to handle HTTP authentication challenges.
3410         https://bugs.webkit.org/show_bug.cgi?id=84031
3411
3412         Reviewed by Rob Buis.
3413
3414         Implemented interface function didReceiveAuthenticationChallenge in class
3415         DumpRenderTree.
3416
3417         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
3418         (drtCredentialDescription):
3419         (BlackBerry::WebKit::DumpRenderTree::didReceiveAuthenticationChallenge):
3420         (WebKit):
3421         * DumpRenderTree/blackberry/DumpRenderTreeBlackBerry.h:
3422         (WebCore):
3423         (DumpRenderTree):
3424
3425 2012-04-16  Philippe Normand  <pnormand@igalia.com>
3426
3427         build-webkit --gtk --no-webkit2 builds WK2
3428         https://bugs.webkit.org/show_bug.cgi?id=83977
3429
3430         Reviewed by Martin Robinson.
3431
3432         * Scripts/build-webkit: Fixup ./configure option to disable
3433         WebKit2 build.
3434
3435 2012-04-16  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
3436
3437         [EFL] [DRT] Feeding key events with invalid keyName
3438         https://bugs.webkit.org/show_bug.cgi?id=83797
3439
3440         Keep the keyName valid within scope of KeyEventInfo to avoid
3441         feeding key events with invalid keyName.
3442
3443         Reviewed by Philippe Normand.
3444
3445         * DumpRenderTree/efl/EventSender.cpp:
3446         (KeyEventInfo::KeyEventInfo):
3447         (KeyEventInfo):
3448         (keyPadNameFromJSValue):
3449         (keyNameFromJSValue):
3450         (createKeyEventInfo):
3451         (sendKeyDown):
3452
3453 2012-04-16  Andras Piroska  <pandras@inf.u-szeged.hu>
3454
3455         [Qt] Replace QtXmlPatterns usage with libxslt dependency
3456         <https://bugs.webkit.org/show_bug.cgi?id=76820>
3457
3458         Reviewed by Simon Hausmann.
3459
3460         The features.prf file was modified, in order to use libxml and libxslt by default.
3461         This modification only work at Qt5.
3462
3463         * qmake/config.tests/libxml2/libxml2.cpp: Added.
3464         (main):
3465         * qmake/config.tests/libxml2/libxml2.pro: Added.
3466         * qmake/config.tests/libxslt/libxslt.cpp: Added.
3467         (main):
3468         * qmake/config.tests/libxslt/libxslt.pro: Added.
3469         * qmake/mkspecs/features/features.prf:
3470         * qmake/sync.profile:
3471
3472 2012-04-16  Seokju Kwon  <seokju.kwon@samsung.com>
3473
3474         [EFL] EWebLaucher : Fix wrong parameter order in function call for multiple windows.
3475         https://bugs.webkit.org/show_bug.cgi?id=83862
3476
3477         Reviewed by Kenneth Rohde Christiansen.
3478
3479         EWebLauncher creates multiple windows, like multiple tabs when pressing F9 key.
3480         It doesn't work properly because of wrong parameter order in function call.
3481
3482         * EWebLauncher/main.c:
3483
3484 2012-04-16  Philippe Normand  <pnormand@igalia.com>
3485
3486         Unreviewed, GTK gardening.
3487
3488         * Scripts/run-gtk-tests:
3489         (TestRunner): Skip /webkit/atk/textChangedNotifications.
3490
3491 2012-04-13  Ryosuke Niwa  <rniwa@webkit.org>
3492
3493         Chromium perf bots should not build
3494         https://bugs.webkit.org/show_bug.cgi?id=83641
3495
3496         Reviewed by Eric Seidel.
3497
3498         Don't build on perf bots to improve the cycle time.
3499
3500         * BuildSlaveSupport/build.webkit.org-config/config.json:
3501
3502 2012-04-14  Zalan Bujtas  <zbujtas@gmail.com>
3503
3504         [WK2] Set WebKitTestRunner's WebProcess cache model to DocumentBrowser.
3505         https://bugs.webkit.org/show_bug.cgi?id=81945
3506
3507         Reviewed by Dan Bernstein.
3508
3509         WebProcess's default cache model is DocumentViewer. When DocumentViewer model is on,
3510         the PageCache's capacity is set to 0. Tests that are enabling 'WebKitUsesPageCachePreferenceKey' expect
3511         PageCache to be functional.
3512
3513         * WebKitTestRunner/TestController.cpp:
3514         (WTR::TestController::resetStateToConsistentValues):
3515
3516 2012-04-11  James Robinson  <jamesr@chromium.org>
3517
3518         [chromium] Remove unused compositeToTexture / compositeOffscreen setting
3519         https://bugs.webkit.org/show_bug.cgi?id=83733
3520
3521         Reviewed by Adrienne Walker.
3522
3523         Remove compositeOffscreen / compositeToTexture setting. This was an experiment that is no longer needed.
3524
3525         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3526         (main):
3527         * DumpRenderTree/chromium/TestShell.cpp:
3528         (TestShell::TestShell):
3529         (TestShell::resetWebSettings):
3530         * DumpRenderTree/chromium/TestShell.h:
3531         (TestShell):
3532         * DumpRenderTree/chromium/WebPreferences.cpp:
3533         (WebPreferences::reset):
3534         (WebPreferences::applyTo):
3535         * DumpRenderTree/chromium/WebPreferences.h:
3536         (WebPreferences):
3537
3538 2012-04-13  Ryosuke Niwa  <rniwa@webkit.org>
3539
3540         webkitpy build fix. Revert the debug change (number of runs was 2 instead of 20),
3541         and also rebaseline the test results since we ignore the first run now.
3542
3543         * Scripts/webkitpy/performance_tests/perftest.py:
3544         (PageLoadingPerfTest.run):
3545         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
3546         (TestPageLoadingPerfTest.test_run):
3547
3548 2012-04-12  Ryosuke Niwa  <rniwa@webkit.org>
3549
3550         Enable SVG page loading performance tests
3551         https://bugs.webkit.org/show_bug.cgi?id=83861
3552
3553         Reviewed by Adam Barth.
3554
3555         Add support for page loading tests. These are tests without any test running harnesses, and we use
3556         DriverOutput's test_timeout to measure the loading time.
3557
3558         Extracted PerfTest.run and PerfTest.run_failed from PerfTestRunner._run_single_test.
3559
3560         * Scripts/webkitpy/performance_tests/perftest.py:
3561         (PerfTest.run):
3562         (PerfTest):
3563         (PerfTest.run_failed):
3564         (PerfTest.parse_output):
3565         (PerfTest.output_statistics):
3566         (ChromiumStylePerfTest.parse_output):
3567         (PageLoadingPerfTest): Added.
3568         (PageLoadingPerfTest.__init__):
3569         (PageLoadingPerfTest.run): Run the driver 20 times to get the statistics for the loading time.
3570         * Scripts/webkitpy/performance_tests/perftest_unittest.py: Added.
3571         (MockPrinter):
3572         (MockPrinter.__init__):
3573         (MockPrinter.write):
3574         (MainTest):
3575         (MainTest.test_parse_output):
3576         (MainTest.test_parse_output_with_failing_line):
3577         (TestPageLoadingPerfTest):
3578         (TestPageLoadingPerfTest.assertWritten):
3579         (TestPageLoadingPerfTest.MockDriver):
3580         (TestPageLoadingPerfTest.MockDriver.__init__):
3581         (TestPageLoadingPerfTest.MockDriver.run_test):
3582         (TestPageLoadingPerfTest.test_run):
3583         (TestPageLoadingPerfTest.test_run_with_bad_output):
3584         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3585         (PerfTestsRunner):
3586         (PerfTestsRunner._collect_tests._is_test_file): Now recognizes svg tests.
3587         (PerfTestsRunner._collect_tests):
3588         (PerfTestsRunner._run_tests_set):
3589         (PerfTestsRunner._run_single_test):
3590         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3591         (test_collect_tests_with_page_load_svg):
3592         (test_collect_tests_with_page_load_svg.add_file):
3593
3594 2012-04-10  Brady Eidson  <beidson@apple.com>
3595
3596         <rdar://problem/11176921> and https://bugs.webkit.org/show_bug.cgi?id=83600
3597         Need WebKit2 API to notify whether history loads are in the page cache
3598
3599         Add a test ShouldGoToBackForwardListItem that exercises the new BundlePageLoaderClient
3600         callback as well as the new WKBundleBackForwardListItemIsInPageCache() API.
3601
3602         Reviewed by Darin Adler.
3603
3604         * TestWebKitAPI/Tests/WebKit2/ShouldGoToBackForwardListItem.cpp: Added.
3605         (TestWebKitAPI::didFinishLoadForFrame):
3606         (TestWebKitAPI::willGoToBackForwardListItem):
3607         (TestWebKitAPI::setPageLoaderClient):
3608         (TestWebKitAPI::TEST):
3609         * TestWebKitAPI/Tests/WebKit2/ShouldGoToBackForwardListItem_Bundle.cpp: Added.
3610         (ShouldGoToBackForwardListItemTest):
3611         (TestWebKitAPI::shouldGoToBackForwardListItemCallback):
3612         (TestWebKitAPI::ShouldGoToBackForwardListItemTest::ShouldGoToBackForwardListItemTest):
3613         (TestWebKitAPI::ShouldGoToBackForwardListItemTest::didCreatePage):
3614         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3615
3616         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3617         (WTR::InjectedBundlePage::InjectedBundlePage): Update for the new client.
3618
3619 2012-04-13  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
3620
3621         Group EFL bots into a category on the buildbot master
3622         https://bugs.webkit.org/show_bug.cgi?id=82712
3623
3624         Reviewed by Adam Roben.
3625
3626         Creating a new category definition on the buildbot frontend.
3627
3628         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3629         (loadBuilderConfig): Determine category for EFL bots
3630         * BuildSlaveSupport/build.webkit.org-config/templates/root.html: Add links to EFL category
3631
3632 2012-04-13  Sheriff Bot  <webkit.review.bot@gmail.com>
3633
3634         Unreviewed, rolling out r113714.
3635         http://trac.webkit.org/changeset/113714
3636         https://bugs.webkit.org/show_bug.cgi?id=83887
3637
3638         It is behaving strange on the bots and needs more
3639         investigation. (Requested by jeez_ on #webkit).
3640
3641         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
3642         (WTR::PlatformWebView::PlatformWebView):
3643
3644 2012-04-11  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
3645
3646         webkit-tools-completion: Support bugzilla email completion for webkit-patch upload --cc
3647         https://bugs.webkit.org/show_bug.cgi?id=83690
3648
3649         Reviewed by Tor Arne Vestbø.
3650
3651         Typing part of an email and pressing <TAB> will bash-complete it using the list in
3652         Tools/Scripts/webkitpy/common/config/committers.py
3653
3654         * Scripts/webkit-tools-completion.sh:
3655
3656 2012-04-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>