9400d07b5f44b97660ef7acd35a10bcbd98aca99
[WebKit-https.git] / Tools / ChangeLog
1 2013-01-10  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>
2
3         [jhbuild] Use the "configure" script to build glib
4         https://bugs.webkit.org/show_bug.cgi?id=106569
5
6         Reviewed by Martin Robinson.
7
8         So far both EFL and GTK have been using autogen.sh to build glib.
9         While that does work most of the time, it's not really needed since
10         we are downloading and building stable releases from tarballs.
11         Additionally, autogen.sh fails to work with automake 1.13 due to the
12         usage of some macros that have been removed (this has already been
13         fixed in glib master).
14
15         * efl/jhbuild.modules:
16         * gtk/jhbuild.modules:
17
18 2013-01-10  Tim 'mithro' Ansell  <mithro@mithis.com>
19
20         Changing tests to be consistent and use self.scm rather then recreating the scm object.
21         (Missed a couple)
22         https://bugs.webkit.org/show_bug.cgi?id=106536
23
24         Reviewed by Eric Seidel.
25
26         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
27
28 2013-01-10  Tony Chang  <tony@chromium.org>
29
30         Speed up supplemental dependency computation
31         https://bugs.webkit.org/show_bug.cgi?id=106503
32
33         Reviewed by Adam Barth.
34
35         Pass the IDL attributes file for generating the bindings in WebKitTestRunner.
36
37         * WebKitTestRunner/CMakeLists.txt:
38
39 2013-01-10  Zan Dobersek  <zandobersek@gmail.com>
40
41         Remove the ENABLE_ANIMATION_API feature define occurences
42         https://bugs.webkit.org/show_bug.cgi?id=106544
43
44         Reviewed by Simon Fraser.
45
46         The Animation API code was removed in r137243. The ENABLE_ANIMATION_API
47         feature define handling still lingers in various build systems and configurations
48         but is of no use, so it should be removed.
49
50         * Scripts/webkitperl/FeatureList.pm:
51         * qmake/mkspecs/features/features.pri:
52
53 2013-01-10  Carlos Garcia Campos  <cgarcia@igalia.com>
54
55         [GTK] Add support for loading web process extensions
56         https://bugs.webkit.org/show_bug.cgi?id=105631
57
58         Reviewed by Gustavo Noronha Silva.
59
60         * MiniBrowser/gtk/GNUmakefile.am:
61         * MiniBrowser/gtk/main.c:
62         (main): Set WEBKIT_INJECTED_BUNDLE env var to use the injected
63         bundle lib from build dir.
64         * Scripts/webkitpy/style/checker.py: Add exceptions for GTK+ API
65         located in WebProcess/InjectedBundle/API/gtk.
66         * gtk/generate-gtkdoc:
67         (get_webkit2_options): Scan also files in
68         WebProcess/InjectedBundle/API/gtk to generate API docs.
69         (get_webkit2_options.injected_bundle_src_path): Helper function to
70         build paths in WebProcess/InjectedBundle/API/gtk.
71
72 2013-01-10  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
73
74         [EFL][WK2] Writing immediately to created windows (open) fails
75         https://bugs.webkit.org/show_bug.cgi?id=105276
76
77         Reviewed by Kenneth Rohde Christiansen.
78
79         Do not reset 'about:blank' url for the new pages in mini browser.
80
81         * MiniBrowser/efl/main.c:
82         (window_create):
83
84 2013-01-10  Tim 'mithro' Ansell  <mithro@mithis.com>
85
86         Changing tests to be consistent and use self.scm rather then recreating the scm object.
87         https://bugs.webkit.org/show_bug.cgi?id=106536
88
89         Reviewed by Eric Seidel.
90
91         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
92
93 2013-01-10  Tim 'mithro' Ansell  <mithro@mithis.com>
94
95         Converting to unittest2 so we can use assertItemsEqual, making tests
96         not dependent on order of items.
97         https://bugs.webkit.org/show_bug.cgi?id=106532
98
99         Before:
100         Ran 1733 tests in 38.385s
101         FAILED (failures=1, errors=22)
102
103         After:
104         Ran 1733 tests in 32.298s
105         FAILED (failures=0, errors=22)
106
107         Reviewed by Eric Seidel.
108
109         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
110         * Scripts/webkitpy/thirdparty/__init__.py:
111         (AutoinstallImportHook._install_unittest2):
112
113 2013-01-10  Christophe Dumez  <christophe.dumez@intel.com>
114
115         [EFL] Add gstreamer 1.0.5 to jhbuild
116         https://bugs.webkit.org/show_bug.cgi?id=106178
117
118         Reviewed by Laszlo Gombos.
119
120         Add gstreamer 1.0.5 to EFL's jhbuild configuration.
121
122         * efl/jhbuild.modules:
123
124 2013-01-09  Alan Cutter  <alancutter@chromium.org>
125
126         Updated parameters in GCE build scripts so they now work.
127         https://bugs.webkit.org/show_bug.cgi?id=106523
128
129         Reviewed by Adam Barth.
130
131         * EWSTools/GoogleComputeEngine/build-chromium-ews.sh:
132         * EWSTools/GoogleComputeEngine/build-commit-queue.sh:
133         * EWSTools/cold-boot.sh:
134
135 2013-01-09  Mark Pilgrim  <pilgrim@chromium.org>
136
137         [Chromium] Remove some stray references to WebKitPlatformSupport.h in DumpRenderTree
138         https://bugs.webkit.org/show_bug.cgi?id=106520
139
140         Reviewed by James Robinson.
141
142         Part of a larger refactoring series. See tracking bug 82948.
143
144         * DumpRenderTree/chromium/WebThemeEngineDRTMac.h:
145         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
146
147 2013-01-09  Julie Parent  <jparent@chromium.org>
148
149         Dashboard cleanup: Remove code for currentState and window.location.hash getting out of state
150         https://bugs.webkit.org/show_bug.cgi?id=106514
151
152         This is old code, as these can no longer get out of sync.
153         permmaLinkURLHash is generated from currentState, which is set only
154         from the location.  The commment about how these can get out of sync
155         refers to a function that no longer exists.
156         
157         Reviewed by Dirk Pranke.
158
159         * TestResultServer/static-dashboards/dashboard_base.js:
160         (handleLocationChange):
161
162 2013-01-09  Julie Parent  <jparent@chromium.org>
163
164         Dashboard Cleanup: Reload the page as soon as we know we need to.
165         https://bugs.webkit.org/show_bug.cgi?id=106515
166
167         No need to continue parsing all of the data once we know that
168         we need to do a reload.
169         
170         Reviewed by Dirk Pranke.
171
172         * TestResultServer/static-dashboards/dashboard_base.js:
173         (parseParameters):
174
175 2013-01-09  Eric Seidel  <eric@webkit.org>
176
177         run-perf-tests --chromium-android --profile should show symbols for the kernel
178         https://bugs.webkit.org/show_bug.cgi?id=106498
179
180         Reviewed by Adam Barth.
181
182         Turns out this was easy, once I finally read the output from "perf report".
183         It appears there may be a bug in "perf record" on android, as it complains
184         about not being able to read from /proc/kallsyms even when
185         /proc/sys/kernel/kptr_restrict is 0.  For now I've not bothered
186         to keep /proc/sys/kernel/kptr_restrict as 0 during the actual record
187         but rather just flip it to 0 long enough to grab the /proc/kallsyms
188         and then flip it back to whatever the device had.
189
190         /proc/sys/kernel/kptr_restrict controls what /proc/kallsyms returns
191         on Linux.  /proc/kallsyms contains a mapping of kernel addresses
192         to symbol names.  Its world-readable, but will return all 0s if you
193         don't have permission to see the kernel symbols.  kptr_restrict
194         supports values 0, 1, 2.  Where 0 means "everyone can see the real symbols"
195         1 is only a specific group, and 2 is "no one, not even root".
196         By default kptr_restrict is 2 on Android it seems.
197         More kptr_restrict docs: http://lwn.net/Articles/420403/
198
199         I also took this opportunity to clean up how the perf record command
200         was built for use/display in AndroidPerf.profile_after_exit and
201         change to always using long-form names for the arguments (to hopefully help readability).
202
203         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
204         (AndroidPerf.__init__):
205         (profile_after_exit):
206         (ChromiumAndroidDriver.__init__):
207         (ChromiumAndroidDriver._update_kallsyms_cache):
208         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
209
210 2013-01-09  Julie Parent  <jparent@chromium.org>
211
212         Dashboard cleanup: remove dead code.
213         https://bugs.webkit.org/show_bug.cgi?id=106509
214
215         Reviewed by Dirk Pranke.
216
217         * TestResultServer/static-dashboards/dashboard_base.js:
218         (parseParameters):
219
220 2013-01-09  Joanmarie Diggs  <jdiggs@igalia.com>
221
222         [GTK] accessibility/aria-labelledby-overrides-label.html requires a proper baseline
223         https://bugs.webkit.org/show_bug.cgi?id=105638
224
225         Reviewed by Martin Robinson.
226
227         The test was failing for two reasons:
228         - AccessibilityRenderObject::correspondingLabelForControlElement() was
229           not ignoring the ARIA labelled-by property
230         - AccessibilityController::accessibleElementById() was not implemented
231
232         Because getting an element by ID cannot be done in the UIProcess, the
233         decision was made to expose the element's ID as an accessible attribute
234         of the object.
235
236         In addition, fixing the bug in AccessibilityRenderObject made it possible
237         to eliminate the Gtk platform-specific expectations for another test.
238
239         * DumpRenderTree/gtk/AccessibilityControllerGtk.cpp:
240         (childElementById): Added. Walks the tree looking for the AtkObject
241         Attribute which exposes the HTML element's ID.
242         (AccessibilityController::accessibleElementById): Implemented.
243         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
244         (AccessibilityUIElement::title): Modified to print out "AXTitle: " even
245         when there is no title, as is done with the Mac port.
246         * WebKitTestRunner/InjectedBundle/AccessibilityController.cpp:
247         (WTR):
248         * WebKitTestRunner/InjectedBundle/atk/AccessibilityControllerAtk.cpp:
249         (WTR::childElementById): Added. Walks the tree looking for the AtkObject
250         Attribute which exposes the HTML element's ID.
251         (WTR):
252         (WTR::AccessibilityController::accessibleElementById): Implemented.
253         * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
254         (WTR::AccessibilityUIElement::title): Modified to print out "AXTitle: "
255         even when there is no title, as is done with the Mac port.
256
257 2013-01-09  Gregg Tavares  <gman@google.com>
258
259         Add Gregg Tavares as a committer
260         https://bugs.webkit.org/show_bug.cgi?id=106492
261
262         Unreviewed.
263
264         * Scripts/webkitpy/common/config/committers.py:
265
266 2013-01-09  Roger Fong  <roger_fong@apple.com>
267
268         Enable seamless iFrames for Windows DRT.
269         https://bugs.webkit.org/show_bug.cgi?id=106468.
270
271         Rubberstamped by Timothy Horton.
272
273         * DumpRenderTree/win/DumpRenderTree.cpp:
274         (resetDefaultsToConsistentValues):
275
276 2013-01-09  Christophe Dumez  <christophe.dumez@intel.com>
277
278         [WK2] Implement eventSender.scheduleAsynchronousKeyDown
279         https://bugs.webkit.org/show_bug.cgi?id=93979
280
281         Reviewed by Benjamin Poulain.
282
283         Implement eventSender.scheduleAsynchronousKeyDown() in
284         WebKitTestRunner.
285
286         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
287         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
288         (WTR::createKeyDownMessageBody):
289         (WTR::EventSendingController::keyDown):
290         (WTR::EventSendingController::scheduleAsynchronousKeyDown):
291         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
292         (EventSendingController):
293         * WebKitTestRunner/TestController.cpp:
294         (WTR::TestController::didReceiveKeyDownMessageFromInjectedBundle):
295         (WTR::TestController::didReceiveMessageFromInjectedBundle):
296         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
297         * WebKitTestRunner/TestController.h:
298         (TestController):
299
300 2013-01-09  Jochen Eisinger  <jochen@chromium.org>
301
302         [chromium] move resource load callback dumping to TestRunner library
303         https://bugs.webkit.org/show_bug.cgi?id=106449
304
305         Reviewed by Adam Barth.
306
307         * DumpRenderTree/chromium/DRTTestRunner.cpp:
308         (DRTTestRunner::DRTTestRunner):
309         (DRTTestRunner::reset):
310         * DumpRenderTree/chromium/DRTTestRunner.h:
311         (DRTTestRunner):
312         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
313         (WebKit):
314         (WebTestRunner::WebTestDelegate::makeURLErrorDescription):
315         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
316         (WebKit):
317         (WebTestProxyBase):
318         (WebTestRunner::WebTestProxy::assignIdentifierToRequest):
319         (WebTestRunner::WebTestProxy::willRequestResource):
320         (WebTestRunner::WebTestProxy::willSendRequest):
321         (WebTestRunner::WebTestProxy::didReceiveResponse):
322         (WebTestRunner::WebTestProxy::didFinishResourceLoad):
323         (WebTestRunner::WebTestProxy::didFailResourceLoad):
324         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
325         (WebTestRunner::WebTestRunner::shouldDumpResourceLoadCallbacks):
326         (WebTestRunner::WebTestRunner::shouldDumpResourceRequestCallbacks):
327         (WebTestRunner::WebTestRunner::shouldDumpResourceResponseMIMETypes):
328         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
329         (WebTestRunner::TestRunner::TestRunner):
330         (WebTestRunner::TestRunner::reset):
331         (WebTestRunner::TestRunner::shouldDumpResourceLoadCallbacks):
332         (WebTestRunner::TestRunner::shouldDumpResourceRequestCallbacks):
333         (WebTestRunner::TestRunner::shouldDumpResourceResponseMIMETypes):
334         (WebTestRunner):
335         (WebTestRunner::TestRunner::dumpResourceLoadCallbacks):
336         (WebTestRunner::TestRunner::dumpResourceRequestCallbacks):
337         (WebTestRunner::TestRunner::dumpResourceResponseMIMETypes):
338         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
339         (TestRunner):
340         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
341         (WebTestRunner::WebTestProxyBase::reset):
342         (WebTestRunner):
343         (WebTestRunner::WebTestProxyBase::assignIdentifierToRequest):
344         (WebTestRunner::WebTestProxyBase::willRequestResource):
345         (WebTestRunner::WebTestProxyBase::willSendRequest):
346         (WebTestRunner::WebTestProxyBase::didReceiveResponse):
347         (WebTestRunner::WebTestProxyBase::didFinishResourceLoad):
348         (WebTestRunner::WebTestProxyBase::didFailResourceLoad):
349         * DumpRenderTree/chromium/WebViewHost.cpp:
350         (WebViewHost::willSendRequest):
351         (WebViewHost::makeURLErrorDescription):
352         (WebViewHost::reset):
353         * DumpRenderTree/chromium/WebViewHost.h:
354         (WebViewHost):
355
356 2013-01-09  Dan Carney  <dcarney@google.com>
357
358         [chromium] move dumpcreateview methods to testrunner library
359         https://bugs.webkit.org/show_bug.cgi?id=106434
360
361         Reviewed by Adam Barth.
362
363         * DumpRenderTree/chromium/DRTTestRunner.cpp:
364         (DRTTestRunner::DRTTestRunner):
365         (DRTTestRunner::reset):
366         * DumpRenderTree/chromium/DRTTestRunner.h:
367         (DRTTestRunner):
368         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
369         (WebTestRunner::WebTestRunner::shouldDumpCreateView):
370         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
371         (WebTestRunner::TestRunner::TestRunner):
372         (WebTestRunner::TestRunner::reset):
373         (WebTestRunner::TestRunner::shouldDumpCreateView):
374         (WebTestRunner):
375         (WebTestRunner::TestRunner::dumpCreateView):
376         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
377         (TestRunner):
378
379 2013-01-09  Csaba Osztrogonác  <ossy@webkit.org>
380
381         webkitpy cleanup: Merge additional_comment_text into comment_text
382         https://bugs.webkit.org/show_bug.cgi?id=106421
383
384         Reviewed by Adam Barth.
385
386         * Scripts/webkitpy/common/config/committervalidator.py:
387         (CommitterValidator.reject_patch_from_commit_queue):
388         (CommitterValidator.reject_patch_from_review_queue):
389         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
390         (Bugzilla.set_flag_on_attachment):
391         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
392         (MockBugzilla.set_flag_on_attachment):
393         * Scripts/webkitpy/tool/bot/feeders_unittest.py:
394         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
395         (AbstractEarlyWarningSystem._post_reject_message_on_bug):
396         * Scripts/webkitpy/tool/commands/queues_unittest.py:
397
398 2013-01-09  Mario Sanchez Prada  <mario.prada@samsung.com>
399
400         [GTK] Raise version of gdk-pixbuf in jhbuild up to 2.26.5
401         https://bugs.webkit.org/show_bug.cgi?id=106464
402
403         Reviewed by Martin Robinson.
404
405         Current version in jhbuild.modules (2.26.0) is giving trouble to
406         when trying to build WebKitGTK from scratch, due to a segmentation
407         fault error in gdk-pixbuf-query-loader during the 'make'
408         phase. Raising the version to 2.26.5 fixes this issue.
409
410         * gtk/jhbuild.modules: Version raised, checksums updated.
411
412 2013-01-09  János Badics  <jbadics@inf.u-szeged.hu>
413
414         [Qt][NRWT] Pass --timeout to DRT/WTR if a test is marked as SLOW.
415         https://bugs.webkit.org/show_bug.cgi?id=90968.
416
417         Reviewed by Csaba Osztrogonác.
418
419         Added functionality in DRT and WTR to use any timeout value while running
420         slow tests (eventually, any test). Now NRWT --time-out-ms determines the
421         timeout value for the test. Added a flag in NRWT (supports_per_test_timeout)
422         to indicate whether the current port supports setting timeout value
423         per test (it's False by default; I enabled it only on Qt).
424         Also corrected a typo in driver.py
425
426         * DumpRenderTree/DumpRenderTree.h:
427         (TestCommand::TestCommand):
428         (TestCommand):
429         * DumpRenderTree/DumpRenderTreeCommon.cpp:
430         (parseInputLine):
431         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
432         (WebCore::DumpRenderTree::processLine):
433         * Scripts/webkitpy/layout_tests/port/base.py:
434         (Port.supports_per_test_timeout):
435         * Scripts/webkitpy/layout_tests/port/driver.py:
436         (Driver.run_test):
437         (Driver._command_from_driver_input):
438         * Scripts/webkitpy/layout_tests/port/qt.py:
439         (QtPort.supports_per_test_timeout):
440         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
441         (WTR::InjectedBundle::InjectedBundle):
442         (WTR::InjectedBundle::didReceiveMessage):
443         (WTR::InjectedBundle::beginTesting):
444         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
445         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
446         (WTR::TestRunner::setCustomTimeout):
447         (WTR):
448         * WebKitTestRunner/InjectedBundle/TestRunner.h:
449         (TestRunner):
450         * WebKitTestRunner/InjectedBundle/qt/TestRunnerQt.cpp:
451         (WTR::TestRunner::initializeWaitToDumpWatchdogTimerIfNeeded):
452         * WebKitTestRunner/TestController.cpp:
453         (WTR::TestController::TestController):
454         (WTR::TestController::getCustomTimeout):
455         (WTR):
456         (WTR::TestCommand::TestCommand):
457         (TestCommand):
458         (WTR::parseInputLine):
459         (WTR::TestController::runTest):
460         (WTR::TestController::runUntil):
461         * WebKitTestRunner/TestController.h:
462         (TestController):
463         * WebKitTestRunner/TestInvocation.cpp:
464         (WTR::TestInvocation::TestInvocation):
465         (WTR::TestInvocation::setCustomTimeout):
466         (WTR):
467         (WTR::TestInvocation::invoke):
468         * WebKitTestRunner/TestInvocation.h:
469         (TestInvocation):
470
471 2013-01-08  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
472
473         Regression(r138681): Add HAVE(ACCESSIBILITY) guard to atk files
474         https://bugs.webkit.org/show_bug.cgi?id=106290
475
476         Reviewed by Martin Robinson.
477
478         Some atk files don't use HAVE(ACCESSIBILITY). It might make build errors when
479         the macro isn't enabled.
480
481         * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
482
483 2013-01-08  Martin Robinson  <mrobinson@igalia.com>
484
485         WebKitTestRunner needs support for setHandlesAuthenticationChallenges
486         https://bugs.webkit.org/show_bug.cgi?id=103653
487
488         Reviewed by Anders Carlsson.
489
490         Implement setHandlesAuthenticationChallenges in the UIProcess by sending credentials
491         and the enabled boolean to the UIProcess and then implementing the didReceiveAuthenticationChallengeInFrame
492         client callback.
493
494         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Exposed the new TestRunner
495         methods to the JavaScript API.
496         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
497         (WTR::TestRunner::setHandlesAuthenticationChallenges): Added this callback for the JavaScript API.
498         (WTR::TestRunner::setAuthenticationUsername): Ditto.
499         (WTR::TestRunner::setAuthenticationPassword): Ditto.
500         * WebKitTestRunner/InjectedBundle/TestRunner.h:
501         (TestRunner): Added new declarations for the above.
502         * WebKitTestRunner/TestController.cpp:
503         (WTR::TestController::createWebViewWithOptions): Hook up the new didReceiveSynchronousMessageFromInjectedBundle
504         callback.
505         (WTR::TestController::resetStateToConsistentValues): Reset the credentials and enabled boolean between tests.
506         (WTR::TestController::didReceiveAuthenticationChallengeInFrame): Handle the callback by using
507         the authentication information passed from the InjectedBundle.
508         * WebKitTestRunner/TestController.h:
509         (WTR::TestController::setHandlesAuthenticationChallenges): Added setter to allow TestInvocation
510         to communicate with the TestController.
511         (WTR::TestController::setAuthenticationUsername): Ditto.
512         (WTR::TestController::setAuthenticationPassword): Ditto.
513         (TestController):
514         * WebKitTestRunner/TestInvocation.cpp:
515         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Handle the messages sending authentication
516         information from the InjectedBundle.
517         (WTR::TestInvocation::outputText): Added this method so that the TestController can output
518         text to the TestInvocation text output.
519         * WebKitTestRunner/TestInvocation.h:
520         (TestInvocation): Added declaration for new method.
521
522 2013-01-08  Benjamin Poulain  <benjamin@webkit.org>
523
524         Add myself as a watcher for Web Geolocation
525
526         Unreviewed.
527
528         * Scripts/webkitpy/common/config/watchlist:
529
530 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
531
532         [webkitpy] Call setup_test_run method of the parent Port interface when method is called on subclasses
533         https://bugs.webkit.org/show_bug.cgi?id=106364
534
535         Reviewed by Eric Seidel.
536
537         When the setup_test_run method is called on the subclasses of the base Port
538         interface, the parent method should also be called via super(), despite the
539         parent method not doing anything at the moment.
540
541         The according calls to super() are required in the ChromiumPort, EflPort
542         and GtkPort implementations.
543
544         * Scripts/webkitpy/layout_tests/port/chromium.py:
545         (ChromiumPort.setup_test_run):
546         * Scripts/webkitpy/layout_tests/port/efl.py:
547         (EflPort.setup_test_run):
548         * Scripts/webkitpy/layout_tests/port/gtk.py:
549         (GtkPort.setup_test_run):
550
551 2013-01-08  Julie Parent  <jparent@chromium.org>
552
553         Dashboard cleanup: Remove globals g_buildersThatFailedToLoad and g_staleBuilders
554         https://bugs.webkit.org/show_bug.cgi?id=106356
555
556         g_buildersThatFailedToLoad and g_staleBuilders were globals defined in
557         dashboard_base, assigned by Loader, and used only by dashboard_base to
558         create error messages.  Moved the variables to be privates on the Loader
559         object, moved error message creation to _getLoadingErrorMessages on the
560         Loader object, and now pass the errors back to dashboard base via the
561         resourceLoadingComplete callback.
562         
563         Also removed the now unused clearError function, it was only being used
564         by unit tests to clean up global state.
565         
566         Reviewed by Dirk Pranke.
567
568         * TestResultServer/static-dashboards/dashboard_base.js:
569         (resourceLoadingComplete):
570         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
571         * TestResultServer/static-dashboards/loader.js:
572         (.):
573         * TestResultServer/static-dashboards/loader_unittests.js:
574
575 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
576
577         [EFL][GTK] Make the PulseAudioSanitizer an object on the EflPort, GtkPort
578         https://bugs.webkit.org/show_bug.cgi?id=106354
579
580         Reviewed by Eric Seidel.
581
582         Put the PulseAudioSanitizer object on the EflPort and GtkPort interfaces
583         instead of those two inheriting from it. Also add a mock object of the
584         sanitizer that's used in unit tests.
585
586         * Scripts/webkitpy/layout_tests/port/efl.py:
587         (EflPort):
588         (EflPort.__init__):
589         (EflPort.setup_test_run):
590         (EflPort.clean_up_test_run):
591         * Scripts/webkitpy/layout_tests/port/efl_unittest.py: Also correct the
592         importing order.
593         (EflPortTest.make_port): Put a mock PulseAudioSanitizer on the instance.
594         * Scripts/webkitpy/layout_tests/port/gtk.py:
595         (GtkPort):
596         (GtkPort.__init__):
597         (GtkPort.setup_test_run):
598         (GtkPort.clean_up_test_run):
599         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py: Also correct the
600         importing order.
601         (GtkPortTest.make_port): Put a mock PulseAudioSanitizer on the instance.
602         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer.py:
603         (PulseAudioSanitizer.unload_pulseaudio_module): Stylize the method as public.
604         (PulseAudioSanitizer.restore_pulseaudio_module): Ditto.
605         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer_mock.py: Added.
606         (PulseAudioSanitizerMock): A simple mock interface for PulseAudioSanitizer.
607         (PulseAudioSanitizerMock.unload_pulseaudio_module):
608         (PulseAudioSanitizerMock.restore_pulseaudio_module):
609
610 2013-01-08  Jochen Eisinger  <jochen@chromium.org>
611
612         [chromium] move dumpFrameLoadCallbacks and friends to the TestRunner library
613         https://bugs.webkit.org/show_bug.cgi?id=106324
614
615         Reviewed by Adam Barth.
616
617         Since the frame load callbacks are part of the WebFrameClient, the
618         WebTestProxy is now intercepting both WebViewClient and WebFrameClient
619         methods.
620
621         * DumpRenderTree/chromium/DRTTestRunner.cpp:
622         (DRTTestRunner::DRTTestRunner):
623         (DRTTestRunner::reset):
624         * DumpRenderTree/chromium/DRTTestRunner.h:
625         (DRTTestRunner):
626         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
627         (WebTestInterfaces):
628         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
629         (WebKit):
630         (WebTestProxyBase):
631         (WebTestRunner):
632         (WebTestRunner::WebTestProxy::WebTestProxy):
633         (WebTestProxy):
634         (WebTestRunner::WebTestProxy::didInvalidateRect):
635         (WebTestRunner::WebTestProxy::didScrollRect):
636         (WebTestRunner::WebTestProxy::scheduleComposite):
637         (WebTestRunner::WebTestProxy::scheduleAnimation):
638         (WebTestRunner::WebTestProxy::setWindowRect):
639         (WebTestRunner::WebTestProxy::show):
640         (WebTestRunner::WebTestProxy::didAutoResize):
641         (WebTestRunner::WebTestProxy::postAccessibilityNotification):
642         (WebTestRunner::WebTestProxy::startDragging):
643         (WebTestRunner::WebTestProxy::shouldBeginEditing):
644         (WebTestRunner::WebTestProxy::shouldEndEditing):
645         (WebTestRunner::WebTestProxy::shouldInsertNode):
646         (WebTestRunner::WebTestProxy::shouldInsertText):
647         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
648         (WebTestRunner::WebTestProxy::shouldDeleteRange):
649         (WebTestRunner::WebTestProxy::shouldApplyStyle):
650         (WebTestRunner::WebTestProxy::didBeginEditing):
651         (WebTestRunner::WebTestProxy::didChangeSelection):
652         (WebTestRunner::WebTestProxy::didChangeContents):
653         (WebTestRunner::WebTestProxy::didEndEditing):
654         (WebTestRunner::WebTestProxy::registerIntentService):
655         (WebTestRunner::WebTestProxy::dispatchIntent):
656         (WebTestRunner::WebTestProxy::willPerformClientRedirect):
657         (WebTestRunner::WebTestProxy::didCancelClientRedirect):
658         (WebTestRunner::WebTestProxy::didStartProvisionalLoad):
659         (WebTestRunner::WebTestProxy::didReceiveServerRedirectForProvisionalLoad):
660         (WebTestRunner::WebTestProxy::didFailProvisionalLoad):
661         (WebTestRunner::WebTestProxy::didCommitProvisionalLoad):
662         (WebTestRunner::WebTestProxy::didReceiveTitle):
663         (WebTestRunner::WebTestProxy::didFinishDocumentLoad):
664         (WebTestRunner::WebTestProxy::didHandleOnloadEvents):
665         (WebTestRunner::WebTestProxy::didFailLoad):
666         (WebTestRunner::WebTestProxy::didFinishLoad):
667         (WebTestRunner::WebTestProxy::didChangeLocationWithinPage):
668         (WebTestRunner::WebTestProxy::didDisplayInsecureContent):
669         (WebTestRunner::WebTestProxy::didRunInsecureContent):
670         (WebTestRunner::WebTestProxy::didDetectXSS):
671         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
672         (WebTestRunner::WebTestRunner::setTestIsRunning):
673         (WebTestRunner::WebTestRunner::shouldDumpFrameLoadCallbacks):
674         (WebTestRunner::WebTestRunner::setShouldDumpFrameLoadCallbacks):
675         (WebTestRunner::WebTestRunner::shouldDumpUserGestureInFrameLoadCallbacks):
676         (WebTestRunner::WebTestRunner::stopProvisionalFrameLoads):
677         (WebTestRunner::WebTestRunner::shouldDumpTitleChanges):
678         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
679         (WebTestRunner::TestRunner::TestRunner):
680         (WebTestRunner::TestRunner::reset):
681         (WebTestRunner::TestRunner::setTestIsRunning):
682         (WebTestRunner):
683         (WebTestRunner::TestRunner::shouldDumpFrameLoadCallbacks):
684         (WebTestRunner::TestRunner::setShouldDumpFrameLoadCallbacks):
685         (WebTestRunner::TestRunner::shouldDumpUserGestureInFrameLoadCallbacks):
686         (WebTestRunner::TestRunner::stopProvisionalFrameLoads):
687         (WebTestRunner::TestRunner::shouldDumpTitleChanges):
688         (WebTestRunner::TestRunner::dumpFrameLoadCallbacks):
689         (WebTestRunner::TestRunner::dumpUserGestureInFrameLoadCallbacks):
690         (WebTestRunner::TestRunner::setStopProvisionalFrameLoads):
691         (WebTestRunner::TestRunner::dumpTitleChanges):
692         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
693         (TestRunner):
694         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
695         (WebTestInterfaces::Internal):
696         (WebTestRunner::WebTestInterfaces::Internal::webView):
697         (WebTestRunner::WebTestInterfaces::Internal::Internal):
698         (WebTestRunner::WebTestInterfaces::Internal::setWebView):
699         (WebTestRunner):
700         (WebTestRunner::WebTestInterfaces::Internal::setTestIsRunning):
701         (WebTestRunner::WebTestInterfaces::setWebView):
702         (WebTestRunner::WebTestInterfaces::setTestIsRunning):
703         (WebTestRunner::WebTestInterfaces::webView):
704         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
705         (WebTestRunner::WebTestProxyBase::willPerformClientRedirect):
706         (WebTestRunner):
707         (WebTestRunner::WebTestProxyBase::didCancelClientRedirect):
708         (WebTestRunner::WebTestProxyBase::didStartProvisionalLoad):
709         (WebTestRunner::WebTestProxyBase::didReceiveServerRedirectForProvisionalLoad):
710         (WebTestRunner::WebTestProxyBase::didFailProvisionalLoad):
711         (WebTestRunner::WebTestProxyBase::didCommitProvisionalLoad):
712         (WebTestRunner::WebTestProxyBase::didReceiveTitle):
713         (WebTestRunner::WebTestProxyBase::didFinishDocumentLoad):
714         (WebTestRunner::WebTestProxyBase::didHandleOnloadEvents):
715         (WebTestRunner::WebTestProxyBase::didFailLoad):
716         (WebTestRunner::WebTestProxyBase::didFinishLoad):
717         (WebTestRunner::WebTestProxyBase::didChangeLocationWithinPage):
718         (WebTestRunner::WebTestProxyBase::didDisplayInsecureContent):
719         (WebTestRunner::WebTestProxyBase::didRunInsecureContent):
720         (WebTestRunner::WebTestProxyBase::didDetectXSS):
721         * DumpRenderTree/chromium/TestShell.cpp:
722         (TestShell::runFileTest):
723         (TestShell::testFinished):
724         * DumpRenderTree/chromium/TestShell.h:
725         (TestShell):
726         * DumpRenderTree/chromium/WebViewHost.cpp:
727         (WebViewHost::didStartProvisionalLoad):
728         (WebViewHost::didReceiveServerRedirectForProvisionalLoad):
729         (WebViewHost::didFailProvisionalLoad):
730         (WebViewHost::didCommitProvisionalLoad):
731         (WebViewHost::didReceiveTitle):
732         (WebViewHost::didFailLoad):
733         (WebViewHost::didFinishLoad):
734         * DumpRenderTree/chromium/WebViewHost.h:
735         (WebViewHost):
736
737 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
738
739         [GTK] Lower the default timeout value for WebKitTestRunner
740         https://bugs.webkit.org/show_bug.cgi?id=106353
741
742         Reviewed by Dirk Pranke.
743
744         The default timeout value for WebKitTestRunner should be lowered.
745         The current value (80 seconds) is irrational and just leads to producing
746         text failures for layout tests that genuinely time out. This just leads
747         to more deviation from the GTK WK1 port in terms of test expectations.
748
749         The value used is the same as when running the DumpRenderTree, that's
750         6 seconds in release builds and 12 seconds in debug builds.
751
752         * Scripts/webkitpy/layout_tests/port/gtk.py:
753         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
754         (GtkPortTest.test_default_timeout_ms):
755
756 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
757
758         test-webkitpy shows failure output on my linux box
759         https://bugs.webkit.org/show_bug.cgi?id=101261
760
761         Reviewed by Eric Seidel.
762
763         Mock out the PulseAudioSanitizer methods when creating testing instances
764         of EflPort and GtkPort interfaces in unit tests. This stops throwing up
765         PulseAudio output when the sanitizer was trying to unload modules which
766         were already unloaded by another unit test that was being run in parallel.
767
768         * Scripts/webkitpy/layout_tests/port/efl_unittest.py:
769         (EflPortTest):
770         (EflPortTest.make_port):
771         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
772         (GtkPortTest):
773         (GtkPortTest.make_port):
774
775 2013-01-08  Julie Parent  <jparent@chromium.org>
776
777         Dashboard Cleanup: Remove g_hasDoneInitialPageGeneration
778         https://bugs.webkit.org/show_bug.cgi?id=106254
779
780         Remove the global variable g_hasDoneInitialPageGeneration.  It was used
781         only to figure out if we should call addBuilderLoadErrors during
782         handleLocationChange, which was called when the loader completed and on
783         any hash change, but we only ever wanted to call addBuilderLoadErrors
784         on the initial loader load.  Moved the call of addBuilderLoadErrors to
785         resourceLoadingComplete rather than handleLocationChange, so now we
786         only call it when we want to, and the check can go away completely.
787
788         Reviewed by Dirk Pranke.
789
790         * TestResultServer/static-dashboards/dashboard_base.js:
791         (addBuilderLoadErrors):
792         (resourceLoadingComplete):
793         (handleLocationChange):
794         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
795
796 2013-01-08  Dominic Mazzoni  <dmazzoni@google.com>
797
798         Add aboxhall@chromium.org to accessibility watchlist.
799
800         Unreviewed.
801
802         * Scripts/webkitpy/common/config/watchlist:
803
804 2013-01-08  Dan Winship  <danw@gnome.org>
805
806         Add Apache 2.4 support for Fedora 18
807         https://bugs.webkit.org/show_bug.cgi?id=104478
808
809         Reviewed by Martin Robinson.
810
811         Apache 2.4's httpd.conf is slightly incompatible with 2.2's, so we
812         need a separate config.
813
814         * Scripts/webkitpy/layout_tests/port/base.py:
815         (Port._apache_version):
816         (Port._apache_config_file_name_for_platform):
817         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
818         * Scripts/run-webkit-httpd:
819         * Scripts/webkitperl/httpd.pm:
820         (getApacheVersion):
821         (getDefaultConfigForTestDirectory):
822         (getHTTPDConfigPathForTestDirectory):
823
824 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
825
826         Make the override-feature-defines script compatible with Python v3
827         https://bugs.webkit.org/show_bug.cgi?id=105742
828
829         Reviewed by Martin Robinson.
830
831         Instead of printing output through a statement, use the print function and format
832         the output using the str.format method. This makes the script runnable with Python v3.
833
834         * gtk/override-feature-defines:
835         (adjust_feature_defines_makefile):
836
837 2013-01-08  Heikki Paajanen  <heikki.paajanen@palm.com>
838
839         [Qt] Fix libxml2 config test
840         https://bugs.webkit.org/show_bug.cgi?id=106317
841
842         Reviewed by Simon Hausmann.
843
844         Don't assume standard include/libxml2, libxml-2.0.pc should know
845         where it's headers are.
846
847         * qmake/config.tests/libxml2/libxml2.cpp: Fix include
848
849 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
850
851         [GTK] Make Tools/gtk/generate-gtkdoc compatible with Python 3
852         https://bugs.webkit.org/show_bug.cgi?id=106195
853
854         Reviewed by Philippe Normand.
855
856         Perform changes in the generate-gtkdoc script and common and gtkdoc modules
857         to make the Python code compatible with Python 3. This includes modifying
858         print statements, exception handling, dictionary iteration and
859         byte-sequence-to-string conversion.
860
861         * gtk/common.py:
862         (get_build_path):
863         (pkg_config_file_variable):
864         (gtk_version_of_pkg_config_file):
865         * gtk/generate-gtkdoc:
866         (print_missing_api):
867         * gtk/gtkdoc.py:
868         (GTKDoc.__init__):
869         (GTKDoc._run_command):
870
871 2013-01-02  Steve Block  <steveblock@chromium.org>
872
873         Add chromium.org email address for Steve Block.
874
875         Unreviewed.
876
877         * Scripts/webkitpy/common/config/committers.py:
878
879 2013-01-07  Filip Pizlo  <fpizlo@apple.com>
880
881         Rationalize closure call heuristics and profiling
882         https://bugs.webkit.org/show_bug.cgi?id=106270
883
884         Reviewed by Oliver Hunt.
885         
886         Add ability to use display-profiler-output via a pipe, and add the ability to dump
887         all generated code ('display *' or 'd *').
888
889         * Scripts/display-profiler-output:
890
891 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
892
893         Sorted more xcodeproj files.
894
895         * MiniBrowser/MiniBrowser.xcodeproj/project.pbxproj:
896         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
897         * WebKitLauncher/WebKitLauncher.xcodeproj/project.pbxproj:
898
899 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
900
901         Sorted xcodeproj files.
902
903         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
904         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
905
906 2013-01-07  Dan Bernstein  <mitz@apple.com>
907
908         [mac] WebKit1 clients can’t tell when a frame has been removed from the hierarchy
909         https://bugs.webkit.org/show_bug.cgi?id=106261
910
911         Reviewed by Simon Fraser.
912
913         Added a test for -[WebFrameLoadDelegate webView:didRemoveFrameFromHierarchy:].
914
915         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
916         * TestWebKitAPI/Tests/mac/WebViewDidRemoveFrameFromHierarchy.mm: Added.
917         (-[DidRemoveFrameFromHierarchyFrameLoadDelegate webView:didFinishLoadForFrame:]):
918         (-[DidRemoveFrameFromHierarchyFrameLoadDelegate webView:didRemoveFrameFromHierarchy:]):
919         (TestWebKitAPI):
920         (TestWebKitAPI::TEST):
921
922 2013-01-07  Ojan Vafai  <ojan@chromium.org>
923
924         Add the Apple MountainLion bots to garden-o-matic.
925         Eventually, we really should move away from hardcoding
926         the list of bots.
927
928         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
929
930 2013-01-07  Jochen Eisinger  <jochen@chromium.org>
931
932         [chromium] move web audio related methods to testrunner library
933         https://bugs.webkit.org/show_bug.cgi?id=106211
934
935         Reviewed by Adam Barth.
936
937         * DumpRenderTree/chromium/DRTTestRunner.cpp:
938         (DRTTestRunner::DRTTestRunner):
939         (DRTTestRunner::reset):
940         * DumpRenderTree/chromium/DRTTestRunner.h:
941         (DRTTestRunner):
942         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
943         (WebKit):
944         (WebTestRunner::WebTestRunner::shouldDumpAsAudio):
945         (WebTestRunner::WebTestRunner::audioData):
946         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
947         (WebTestRunner::TestRunner::TestRunner):
948         (WebTestRunner::TestRunner::reset):
949         (WebTestRunner::TestRunner::shouldDumpAsAudio):
950         (WebTestRunner):
951         (WebTestRunner::TestRunner::audioData):
952         (WebTestRunner::TestRunner::setAudioData):
953         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
954         (TestRunner):
955         * DumpRenderTree/chromium/TestShell.cpp:
956         (TestShell::dump):
957
958 2013-01-07  Mike West  <mkwst@chromium.org>
959
960         Make the IFRAME_SEAMLESS flag runtime-enabled.
961         https://bugs.webkit.org/show_bug.cgi?id=106213
962
963         Reviewed by Ojan Vafai.
964
965         * DumpRenderTree/chromium/TestShell.cpp:
966         (TestShell::TestShell):
967         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
968         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
969         * DumpRenderTree/gtk/DumpRenderTree.cpp:
970         (resetDefaultsToConsistentValues):
971         * DumpRenderTree/mac/DumpRenderTree.mm:
972         (resetDefaultsToConsistentValues):
973         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
974         (WebCore::WebPage::resetSettings):
975         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
976         (WTR::InjectedBundle::beginTesting):
977             Ensure that seamless IFrames are enabled for layout tests in each of
978             the ports that support seamless IFrames.
979
980 2013-01-07  Jochen Eisinger  <jochen@chromium.org>
981
982         [chromium] move dumpAsText and friends to the TestRunner library
983         https://bugs.webkit.org/show_bug.cgi?id=106222
984
985         Reviewed by Adam Barth.
986
987         This moves dumpAsText, dumpChildFramesAsText, and dumpChildFrameScrollPositions to the TestRunner library.
988
989         * DumpRenderTree/chromium/DRTTestRunner.cpp:
990         (DRTTestRunner::DRTTestRunner):
991         (DRTTestRunner::reset):
992         * DumpRenderTree/chromium/DRTTestRunner.h:
993         (DRTTestRunner):
994         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
995         (WebTestRunner::WebTestRunner::shouldDumpAsText):
996         (WebTestRunner::WebTestRunner::setShouldDumpAsText):
997         (WebTestRunner::WebTestRunner::shouldGeneratePixelResults):
998         (WebTestRunner::WebTestRunner::setShouldGeneratePixelResults):
999         (WebTestRunner::WebTestRunner::shouldDumpChildFrameScrollPositions):
1000         (WebTestRunner::WebTestRunner::shouldDumpChildFramesAsText):
1001         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1002         (WebTestRunner::TestRunner::TestRunner):
1003         (WebTestRunner::TestRunner::reset):
1004         (WebTestRunner::TestRunner::shouldDumpAsText):
1005         (WebTestRunner):
1006         (WebTestRunner::TestRunner::setShouldDumpAsText):
1007         (WebTestRunner::TestRunner::shouldGeneratePixelResults):
1008         (WebTestRunner::TestRunner::setShouldGeneratePixelResults):
1009         (WebTestRunner::TestRunner::shouldDumpChildFrameScrollPositions):
1010         (WebTestRunner::TestRunner::shouldDumpChildFramesAsText):
1011         (WebTestRunner::TestRunner::dumpAsText):
1012         (WebTestRunner::TestRunner::dumpChildFrameScrollPositions):
1013         (WebTestRunner::TestRunner::dumpChildFramesAsText):
1014         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1015         (TestRunner):
1016
1017 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
1018
1019         Build fix attempt after r138810.
1020
1021         Apparently python 2.6 is really bad at floating point computation.
1022
1023         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py:
1024         (MainTest.test_run_memory_test):
1025
1026 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
1027
1028         Extract a class to represent a test result to simplify PerfTest.parse_output and PageLoadPerfTest._run_with_driver
1029         https://bugs.webkit.org/show_bug.cgi?id=106172
1030
1031         Reviewed by Tony Chang.
1032
1033         Extracted PerfTestMetric, which represents a test result and knows how to compute statistics and convert itself
1034         to a dictionary. A metric is a type of measurement each test makes such as "Time", "Malloc", and "JSHeap".
1035         A single test (e.g. html5-full-render.html) may have multiple metrics, each of which is reported as a separate
1036         test on perf-o-matic and perf. test's results page.
1037
1038         Also, use python's compute_statistics instead of reading statistics off of DriverOutput to simplify the code.
1039
1040         * Scripts/webkitpy/performance_tests/perftest.py:
1041         (PerfTestMetric): Added.
1042         (PerfTestMetric.__init__): Added.
1043         (PerfTestMetric.metric): Added.
1044         (PerfTestMetric.has_values): Added.
1045         (PerfTestMetric.legacy_chromium_bot_compatible_test_name): Added. Following the legacy convention, "Malloc" and
1046         "JSHeap" will use "test_name:Malloc" whereas "Time" uses "test_name" itself. We should eventually move these
1047         conversions to be done on the client side (perf-o-matic and results page).
1048         (PerfTestMetric.append): Added.
1049         (PerfTestMetric.to_dict): Added. Computes the statistics and generates the dictionary object that represents
1050         a result for this metric.
1051         (PerfTestMetric.metric_to_unit): Added. Each metric maps to exactly one unit (it's not one-to-one).
1052         (PerfTestMetric.compute_statistics): Moved from PerfTest.
1053
1054         (PerfTest._create_driver): Extracted from PerfTest.run.
1055         (PerfTest.run): Given an array of PerfTestMetric, generate the results dictionary. Also print out results
1056         as needed.
1057         (PerfTest._run_with_driver):
1058         (PerfTest): Removed _result_classes and _result_class_regex. "result class" is the old name for "metric".
1059         (PerfTest.parse_output): Dramatically simplified. We now just return an array of PerfTestMetric's. Also ignore
1060         all keys except "values". We rely on PerfTestMetric to compute the statistics for us.
1061
1062         (ChromiumStylePerfTest.run): Added and merged _run_with_driver. Since chromium style perf tests are so different
1063         from other perf tests, there isn't much point in sharing the code here.
1064
1065         (PageLoadingPerfTest._run_with_driver): Simplified. Just create an array of PerfTestMetric's without computing
1066         statistics for the measured values.
1067
1068         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1069         (TestPerfTestMetric): Added.
1070         (TestPerfTestMetric.test_init_set_missing_unit): Added.
1071         (TestPerfTestMetric.test_legacy_chromium_bot_compatible_test_name): Added.
1072         (TestPerfTestMetric.test_has_values): Added.
1073         (TestPerfTestMetric.test_append): Added.
1074         (TestPerfTestMetric.test_compute_statistics.compute_statistics):
1075         (TestPerfTestMetric.test_compute_statistics): Added a trivial test case.
1076
1077         (TestPerfTest._assert_results_are_correct): Updated as needed.
1078         (TestPageLoadingPerfTest.test_run): Ditto. Notice that we no longer log anything in _run_with_driver. It's done
1079         in PerfTest.run now after generating the results dictionary.
1080         (TestPageLoadingPerfTest.test_run_with_memory_output): Ditto.
1081
1082 2013-01-07  Julien Chaffraix  <jchaffraix@webkit.org>
1083
1084         Support size_t multiplication and division operators on LayoutUnit
1085         https://bugs.webkit.org/show_bug.cgi?id=83848
1086
1087         Reviewed by Emil A Eklund.
1088
1089         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
1090         Added some simple tests for the new operators using size_t.
1091
1092 2013-01-07  Csaba Osztrogonác  <ossy@webkit.org>
1093
1094         [Qt] run-webkit-tests cleanup for parallel NRWT
1095         https://bugs.webkit.org/show_bug.cgi?id=106220
1096
1097         Reviewed by Dirk Pranke.
1098
1099         * Scripts/run-webkit-tests:
1100         (useNewRunWebKitTests):
1101         * Scripts/webkitpy/layout_tests/port/qt.py:
1102         (QtPort):
1103         (QtPort.default_child_processes):
1104
1105 2013-01-07  Philippe Normand  <pnormand@igalia.com>
1106
1107         Create a GStreamerAudio watchlist and add myself to it.
1108
1109         Unreviewed.
1110
1111         * Scripts/webkitpy/common/config/watchlist:
1112
1113 2013-01-05  Ryosuke Niwa  <rniwa@webkit.org>
1114
1115         Add svn:executable to a bunch of scripts in BuildSlaveSupport.
1116         Also removed the empty mac directory.
1117
1118         * BuildSlaveSupport/built-product-archive: Added property svn:executable.
1119         * BuildSlaveSupport/download-built-product: Added property svn:executable.
1120         * BuildSlaveSupport/mac: Removed.
1121         * BuildSlaveSupport/test-result-archive: Added property svn:executable.
1122
1123 2013-01-04  Ryosuke Niwa  <rniwa@webkit.org>
1124
1125         Add a script to delete stale zero-byte build files
1126         https://bugs.webkit.org/show_bug.cgi?id=106131
1127
1128         Reviewed by Tony Chang.
1129
1130         Delete zero-byte build files on Mac port for now. It doesn't do anything on other ports.
1131
1132         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1133         (DeleteStaleBuildFiles): Added.
1134         (Factory.__init__): Add DeleteStaleBuildFiles step.
1135         * BuildSlaveSupport/delete-stale-build-files: Added.
1136         (main): Delete zero-byte files except ones intentionally generated by our build process.
1137         (webkitBuildDirectory): Copied from build-product-archive.
1138
1139 2013-01-04  Michael Pruett  <michael@68k.org>
1140
1141         [GTK] Fix build error on GTK due to r138836
1142         https://bugs.webkit.org/show_bug.cgi?id=106154
1143
1144         Reviewed by Kentaro Hara.
1145
1146         * GNUmakefile.am:
1147
1148 2013-01-04  Dan Bernstein  <mitz@apple.com>
1149
1150         [mac] WebKit1 has no equivalent of WKBundleNodeHandleCopyHTMLTableCellElementCellAbove
1151         https://bugs.webkit.org/show_bug.cgi?id=106121
1152
1153         Reviewed by Adele Peterson.
1154
1155         Added test for -[DOMTHMLTableCellElement _cellAbove].
1156
1157         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1158         * TestWebKitAPI/Tests/mac/DOMHTMLTableCellCellAbove.mm: Added.
1159         (-[HTMLTableCellElementCellAboveTest webView:didFinishLoadForFrame:]):
1160         (TestWebKitAPI):
1161         (TestWebKitAPI::expectCellAboveCell):
1162         (TestWebKitAPI::TEST):
1163         * TestWebKitAPI/Tests/mac/DOMHTMLTableCellElementCellAbove.html: Added.
1164
1165 2013-01-04  Eric Seidel  <eric@webkit.org>
1166
1167         FlakyTestReporter should be re-enabled and taught how to post patches
1168         https://bugs.webkit.org/show_bug.cgi?id=103839
1169
1170         Reviewed by Adam Barth.
1171
1172         This moves LayoutTestResultsReader off of the DeprecatedPort object.
1173         It also adds a real Port object to the EWS and CommitQueue objects
1174         using a new PatchProcessingQueue superclass.
1175         PatchProcessingQueue doesn't seem to be the cleanest insertion point
1176         but works for this purpose.
1177         I also moved _upload_results_archive_for_patch into PatchProcessingQueue
1178         to resolve a FIXME.
1179
1180         This required some hacks/shiming between the non-specific "old"
1181         port names which the EWS system is currently using, and the new
1182         port objects.  This shimming is done in PatchProcessingQueue._new_port_name_from_old.
1183
1184         Hopefully this will fix the flaky test reporter and results uploading.
1185         Certainly it's another small step towards getting rid of the old ports.py
1186         infrastructure and making the NRWT Port class common webkitpy infrastructure.
1187
1188         Now that DeprecatedPort is no longer used for the results directory
1189         I removed all the explicit results-dir code which webkit-patch was using
1190         in --non-interactive mode to tell NRWT to use a results-dir it understood. :)
1191
1192         * Scripts/webkitpy/common/config/ports.py:
1193         (DeprecatedPort.run_perl_unittests_command):
1194         * Scripts/webkitpy/common/config/ports_mock.py:
1195         (MockPort.name):
1196         * Scripts/webkitpy/common/system/outputcapture.py:
1197         (OutputCapture.assert_outputs):
1198         * Scripts/webkitpy/layout_tests/port/apple.py:
1199         (ApplePort.determine_full_port_name):
1200         * Scripts/webkitpy/tool/bot/layouttestresultsreader.py:
1201         (LayoutTestResultsReader.__init__):
1202         (LayoutTestResultsReader):
1203         (LayoutTestResultsReader._read_file_contents):
1204         (LayoutTestResultsReader._create_layout_test_results):
1205         (LayoutTestResultsReader._create_unit_test_results):
1206         (LayoutTestResultsReader.results):
1207         (LayoutTestResultsReader.archive):
1208         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
1209         (LayoutTestResultsReaderTest.test_missing_layout_test_results):
1210         (LayoutTestResultsReaderTest.test_create_unit_test_results):
1211         (test_missing_unit_test_results_path):
1212         (test_layout_test_results):
1213         (test_archive_last_layout_test_results):
1214         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1215         (AbstractEarlyWarningSystem):
1216         (AbstractEarlyWarningSystem.__init__):
1217         (AbstractEarlyWarningSystem.begin_work_queue):
1218         (AbstractEarlyWarningSystem.run_command):
1219         * Scripts/webkitpy/tool/commands/queues.py:
1220         (AbstractPatchQueue.work_item_log_path):
1221         (PatchProcessingQueue):
1222         (PatchProcessingQueue._new_port_name_from_old):
1223         (PatchProcessingQueue.begin_work_queue):
1224         (PatchProcessingQueue._upload_results_archive_for_patch):
1225         (CommitQueue):
1226         (CommitQueue.begin_work_queue):
1227         (CommitQueue.run_command):
1228         (AbstractReviewQueue):
1229         (AbstractReviewQueue.__init__):
1230         (AbstractReviewQueue.begin_work_queue):
1231         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1232         (AbstractPatchQueueTest.test_next_patch):
1233         (PatchProcessingQueueTest):
1234         (PatchProcessingQueueTest.test_upload_results_archive_for_patch):
1235         * Scripts/webkitpy/tool/steps/runtests.py:
1236         (RunTests.run):
1237         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
1238
1239 2013-01-04  Dana Jansens  <danakj@chromium.org>
1240
1241         [chromium] Remove lost-compositor-context layout tests
1242         https://bugs.webkit.org/show_bug.cgi?id=106089
1243
1244         Reviewed by James Robinson.
1245
1246         Remove the loseCompositorContext() test function.
1247
1248         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1249         (WebTestRunner::TestRunner::TestRunner):
1250         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1251         (TestRunner):
1252
1253 2013-01-04  Adam Klein  <adamk@chromium.org>
1254
1255         Remove ENABLE_MUTATION_OBSERVERS #define
1256         https://bugs.webkit.org/show_bug.cgi?id=105459
1257
1258         Reviewed by Ryosuke Niwa.
1259
1260         * Scripts/webkitperl/FeatureList.pm:
1261         * qmake/mkspecs/features/features.pri:
1262
1263 2013-01-04  Ryosuke Niwa  <rniwa@webkit.org>
1264
1265         Statistics used in perftest_unittest.py and perftest_integrationtest.py are bogus
1266         https://bugs.webkit.org/show_bug.cgi?id=106078
1267
1268         Reviewed by Tony Chang.
1269
1270         This patch updates values for individual iterations in dummy performance tests used in
1271         perftest_unittest.py and perftest_integrationtest.py so that they match up with
1272         statistics such as stdev and median reported in those tests. This change is required
1273         to resolve the bug 97510.
1274
1275         This patch also refactors integration tests by extracting *TestData classes that
1276         centralizes the definitions of each dummy performance test's text (DRT output),
1277         runner output, and parsed results (JSON) so that we may share more data between
1278         python tests.
1279
1280         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1281         (MainTest._assert_results_are_correct): Extracted from test_parse_output.
1282         Use self.assertAlmostEqual when comparing the standard deviation.
1283         (MainTest.test_parse_output):
1284         (test_parse_output_with_failing_line):
1285         (test_parse_output_with_description): Use _assert_results_are_correct.
1286         (test_parse_output_with_subtests): Ditto.
1287
1288         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py:
1289         (InspectorPassTestData): Extracted from tests below.
1290         (EventTargetWrapperTestData): Ditto.
1291         (SomeParserTestData): Ditto.
1292         (MemoryTestData): Ditto.
1293         (TestDriver.run_test): Moved test definitions out of the method.
1294         (MainTest._normalize_output): Truncate standard deviations at 5th decimal point.
1295         (MainTest._load_output_json): Extracted from create_runner.
1296         (MainTest.create_runner):
1297         (MainTest.test_run_test_set_for_parser_tests):
1298         (MainTest.test_run_memory_test):
1299         (MainTest._test_run_with_json_output):
1300         (MainTest.test_run_with_json_output):
1301         (MainTest.test_run_with_description):
1302         (MainTest.test_run_generates_json_by_default):
1303         (MainTest.test_run_merges_output_by_default):
1304         (MainTest.test_run_respects_reset_results):
1305         (MainTest.test_run_generates_and_show_results_page):
1306         (MainTest.test_run_with_slave_config_json):
1307         (MainTest.test_run_with_multiple_repositories):
1308
1309 2013-01-04  Martin Robinson  <mrobinson@igalia.com>
1310
1311         [GTK] [WebKit2] Clicking on a word that is selected in an editable area deletes it
1312         https://bugs.webkit.org/show_bug.cgi?id=106045
1313
1314         Reviewed by Carlos Garcia Campos.
1315
1316         Update test results to confirm that compositions without preedits do not trigger
1317         a composition confirmation.
1318
1319         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
1320         (TestWebKitAPI::verifyCanceledComposition):
1321
1322 2013-01-04  Sailesh Agrawal  <sail@chromium.org>
1323
1324         [Mac] [WK2] platform/mac/fast/events/numpad-keycode-mapping.html fails
1325         https://bugs.webkit.org/show_bug.cgi?id=105958
1326
1327         Reviewed by Alexey Proskuryakov.
1328
1329         Ported code from DumpRenderTree to WebKitTestRunner to map all numpad keys.
1330
1331         * WebKitTestRunner/mac/EventSenderProxy.mm:
1332         (KeyMappingEntry):
1333         (WTR):
1334         (WTR::EventSenderProxy::keyDown): Added all numpad keys.
1335
1336 2013-01-03  Chris Fleizach  <cfleizach@apple.com>
1337
1338         REGRESSION: [Mac] Intermittent crash in WTR::AccessibilityUIElement::isEqual
1339         https://bugs.webkit.org/show_bug.cgi?id=106073
1340
1341         Reviewed by Ryosuke Niwa.
1342
1343         Check that the element being compared to is not nil. 
1344         This crash is happening because one of the accessibility tests does an isEqual comparison within a callback function.
1345         It looks like occasionally, one of the elements passed in to the isEqual is nil because the notification handler is on the accessibility
1346         controller, and the notification is being delivered after the document objects are cleared. This shouldn't happen but I've seen it 
1347         sometimes before on parallel WebKitTestRunner runs.
1348
1349         * DumpRenderTree/AccessibilityUIElement.cpp:
1350         (AccessibilityUIElement::isEqual):
1351         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
1352         (WTR::AccessibilityUIElement::isEqual):
1353
1354 2013-01-03  Tim 'mithro' Ansell  <mithro@mithis.com>
1355
1356         Downloads can be flaky, so try downloading multiple times.
1357         https://bugs.webkit.org/show_bug.cgi?id=105976
1358
1359         Reviewed by Eric Seidel.
1360
1361         * Scripts/webkitpy/common/system/autoinstall.py:
1362         (AutoInstaller._download_to_stream):
1363
1364 2013-01-03  Eric Seidel  <eric@webkit.org>
1365
1366         Rename Tool.port to deprecated_port in preparation for removing more callers
1367         https://bugs.webkit.org/show_bug.cgi?id=106076
1368
1369         Reviewed by Ryosuke Niwa.
1370
1371         This resolves a long-standing FIXME to rename this method. :)
1372
1373         * Scripts/webkitpy/tool/bot/botinfo.py:
1374         (BotInfo.summary_text):
1375         * Scripts/webkitpy/tool/bot/irc_command.py:
1376         (Rollout._update_working_copy):
1377         * Scripts/webkitpy/tool/bot/layouttestresultsreader.py:
1378         (LayoutTestResultsReader._create_layout_test_results):
1379         (LayoutTestResultsReader._create_unit_test_results):
1380         (LayoutTestResultsReader._results_directory):
1381         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
1382         (test_missing_unit_test_results_path):
1383         * Scripts/webkitpy/tool/main.py:
1384         (WebKitPatch.deprecated_port):
1385         * Scripts/webkitpy/tool/mocktool.py:
1386         (MockTool.deprecated_port):
1387         * Scripts/webkitpy/tool/steps/build.py:
1388         (Build.build):
1389         * Scripts/webkitpy/tool/steps/checkstyle.py:
1390         (CheckStyle.run):
1391         * Scripts/webkitpy/tool/steps/commit.py:
1392         (Commit._check_test_expectations):
1393         * Scripts/webkitpy/tool/steps/preparechangelog.py:
1394         (PrepareChangeLog.run):
1395         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
1396         (PrepareChangeLogForDEPSRoll.run):
1397         * Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
1398         (PrepareChangeLogForRevert.run):
1399         * Scripts/webkitpy/tool/steps/runtests.py:
1400         (RunTests.run):
1401         * Scripts/webkitpy/tool/steps/steps_unittest.py:
1402         (StepsTest.test_runtests_args):
1403         * Scripts/webkitpy/tool/steps/update.py:
1404         (Update._update_command):
1405
1406 2013-01-03  Eric Seidel  <eric@webkit.org>
1407
1408         Add myself to the webkitpy watchlist.
1409
1410         Unreviewed.
1411
1412         * Scripts/webkitpy/common/config/watchlist:
1413
1414 2013-01-03  Ryosuke Niwa  <rniwa@webkit.org>
1415
1416         Cleanup perftest* tests and add a test for computing statistics
1417         https://bugs.webkit.org/show_bug.cgi?id=105685
1418
1419         Reviewed by Eric Seidel.
1420
1421         Add a test for PerfTest.compute_statistics (moved and renamed from PageLoadingPerfTest.calculate_statistics) and
1422         extracted perftestsrunner_itegrationtests.py from perftestsrunner_unittests.py.
1423
1424         Also fixed a bug in compute_statistics that the mean ('avg') value can have a large rounding errors in some cases.
1425
1426         * Scripts/webkitpy/performance_tests/perftest.py:
1427         (PerfTest.compute_statistics): Moved from PageLoadingPerfTest to prepare for the bug 97510. Also compute the mean
1428         directly from sorted_values instead of using the one from Knuth's online algorithm. This approach gives more
1429         accurate result for the mean.
1430         (PageLoadingPerfTest.run_single):
1431         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1432         (MainTest.test_compute_statistics):
1433         (MainTest.test_compute_statistics.compute_statistics): Added.
1434         (TestPageLoadingPerfTest.test_run): floatify values.
1435         (TestPageLoadingPerfTest.test_run_with_memory_output): Ditto. Also got rid of ".0" from mean values now that Python
1436         correctly recognizes them as integers.
1437
1438         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py: Copied from
1439         Tools/Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py.
1440         (TestDriver): Moved out of MainTest.
1441         (MainTest): Got rid of assertWritten and all unit tests.
1442         (MainTest._normalize_output): Renamed from normalizeFinishedTime to match the PEP8 naming convention.
1443         (MainTest.test_run_test_set_kills_drt_per_run.TestDriverWithStopCount):
1444         (MainTest.test_run_test_set_for_parser_tests):
1445         (MainTest.test_run_memory_test):
1446         (MainTest._test_run_with_json_output):
1447         (MainTest.test_run_generates_json_by_default):
1448         (MainTest.test_run_merges_output_by_default):
1449         (MainTest.test_run_respects_reset_results):
1450         (MainTest.test_run_generates_and_show_results_page): Use runner.load_output_json() instead of manually loading and
1451         parsing output JSON files. Just verify that the output path is correct instead.
1452         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1453         (MainTest): Removed all integration tests.
1454         (MainTest.create_runner): Simplified.
1455
1456 2013-01-03  Julie Parent  <jparent@chromium.org>
1457
1458         Dashboard cleanup: remove usage of global g_defaultBuilderName
1459         https://bugs.webkit.org/show_bug.cgi?id=106043
1460
1461         Use now public defaultBuilder() from the BuilderGroup, rather
1462         than global variable.
1463
1464         Also eliminates the setup() and initBuilders() functions since we
1465         no longer need to track the global.
1466
1467         Reviewed by Adam Barth.
1468
1469         * TestResultServer/static-dashboards/builders.js:
1470         (BuilderGroup.prototype.defaultBuilder):
1471         (BuilderGroup.prototype.master):
1472         * TestResultServer/static-dashboards/dashboard_base.js:
1473         (parseParameters):
1474         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1475         (updateDefaultBuilderState):
1476         * TestResultServer/static-dashboards/loader.js:
1477         (.):
1478         * TestResultServer/static-dashboards/loader_unittests.js:
1479
1480 2013-01-03  Ryosuke Niwa  <rniwa@webkit.org>
1481
1482         Disable an intermittently failing unit test for now.
1483         The failure is tracked by the bug 106051.
1484
1485         * TestWebKitAPI/Tests/mac/MemoryCachePruneWithinResourceLoadDelegate.mm:
1486         (TestWebKitAPI::TEST):
1487
1488 2013-01-03  Dan Bernstein  <mitz@apple.com>
1489
1490         No way to obtain a DOMNode given a JS wrapper for a Node
1491         https://bugs.webkit.org/show_bug.cgi?id=106033
1492
1493         Added a test for +[DOMNode _nodeFromJSWrapper:].
1494
1495         Reviewed by Sam Weinig.
1496
1497         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1498         * TestWebKitAPI/Tests/mac/DOMNodeFromJSObject.mm: Added.
1499         (TestWebKitAPI):
1500         (TestWebKitAPI::TEST):
1501
1502 2013-01-03  Emil A Eklund  <eae@chromium.org>
1503
1504         Fix overflow in LayoutUnit::ceil and floor for SATURATED_LAYOUT_ARITHMETIC
1505         https://bugs.webkit.org/show_bug.cgi?id=105961
1506
1507         Reviewed by Levi Weintraub.
1508         
1509         Add tests for LayoutUnit::ceil and floor.
1510
1511         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
1512         (TestWebKitAPI::TEST):
1513         (TestWebKitAPI):
1514
1515 2013-01-03  Julie Parent  <jparent@chromium.org>
1516
1517         Add unit test for default builder for the dashboards.
1518         https://bugs.webkit.org/show_bug.cgi?id=105972
1519
1520         Reviewed by Dirk Pranke.
1521
1522         * TestResultServer/static-dashboards/loader_unittests.js:
1523
1524 2013-01-03  Tony Chang  <tony@chromium.org>
1525
1526         Generate internal.settings from Settings.in
1527         https://bugs.webkit.org/show_bug.cgi?id=104740
1528
1529         Reviewed by Adam Barth.
1530
1531         * GNUmakefile.am: Compile new generated files on GTK+.
1532         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
1533         with both extra_sources and add_output_to_sources=false will still
1534         compile the files in extra_sources. This makes it possible to compile
1535         InternalSettingsGenerated.cpp and not be confused by
1536         InternalSettingsGenerated.idl.
1537
1538 2013-01-03  Zan Dobersek  <zandobersek@gmail.com>
1539
1540         [style] WebIDL-reflecting uppercase enums reported as style violation
1541         https://bugs.webkit.org/show_bug.cgi?id=105832
1542
1543         Reviewed by Tony Chang.
1544
1545         Recognize the enum as WebIDL-reflecting one if there's a comment present that
1546         indicates this. In enums spanning over multiple lines, the comment is expected
1547         on the first line. In enums placed in one line, the comment is expected at the
1548         end of that line.
1549
1550         WebIDL-reflecting enum values are required to be written in uppercase and as such
1551         violate the common rule of these values being written in InterCaps. Annotating
1552         enums this way makes it possible to avoid false style violation reports.
1553
1554         The comment is of form '// Web(?:Kit)?IDL enum'.
1555
1556         * Scripts/webkitpy/style/checkers/cpp.py:
1557         (_EnumState.__init__):
1558         (_EnumState.process_clean_line):
1559         (check_enum_casing):
1560         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1561         (NoNonVirtualDestructorsTest.test_enum_casing):
1562
1563 2013-01-03  Tim 'mithro' Ansell  <mithro@mithis.com>
1564
1565         Adding favicon.ico to the garden-o-matic.
1566         https://bugs.webkit.org/show_bug.cgi?id=105982
1567
1568         Reviewed by Adam Barth.
1569
1570         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1571         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon.ico:
1572
1573 2013-01-03  Jonathan Liu  <net147@gmail.com>
1574
1575         [Qt] Fix library linking order
1576         https://bugs.webkit.org/show_bug.cgi?id=105818
1577
1578         Some Qt libraries are linked before the private static libraries
1579         that depend on them which results in QtWebKit failing to link.
1580
1581         Copy Qt module dependencies from QT to QT_PRIVATE when creating
1582         module to ensure the Qt libraries are linked after the private
1583         static libraries that depend on them.
1584
1585         Reviewed by Tor Arne Vestbø.
1586
1587         * qmake/mkspecs/features/default_post.prf:
1588
1589 2013-01-02  Byungwoo Lee  <bw80.lee@samsung.com>
1590
1591         [EFL][WK2] Fix tooltip bugs in MiniBrowser.
1592         https://bugs.webkit.org/show_bug.cgi?id=102209
1593
1594         Reviewed by Gyuyoung Kim.
1595
1596         Set tooltip on the elm_window instead of ewk_view, and add delay to
1597         show the tooltip window.
1598
1599         Setting tooltip on the ewk_view makes elementary error message because
1600         ewk_view is not an elementary widget. To remove the error message,
1601         tooltip should be set on the elm_window.
1602
1603         Most browsers show the tooltip after a mouse pointer stays on an
1604         element object that has tooltip text.
1605         But elm_tooltip widget itself doesn't have delay or handle mouse
1606         events to show/hide.
1607
1608         So additional codes are added to show/hide tooltip widget along with
1609         mouse events.
1610           1) When a mouse pointer stays for a second on an element which has
1611              tooltip text, then tooltip widget is displayed.
1612           2) When a mouse pointer is moved or out of ewk_view, hide the tooltip
1613              widget.
1614
1615         * MiniBrowser/efl/main.c:
1616         (_Tooltip_Information):
1617         (_Browser_Window):
1618         (on_tooltip_show):
1619         (window_tooltip_hide):
1620         (window_tooltip_update):
1621         (on_mouse_in):
1622         (on_mouse_move):
1623         (on_mouse_out):
1624         (window_free):
1625         (on_tooltip_text_set):
1626         (on_tooltip_text_unset):
1627         (window_create):
1628
1629 2013-01-02  Tony Chang  <tony@chromium.org>
1630
1631         Unreviewed, rolling out r138661.
1632         http://trac.webkit.org/changeset/138661
1633         https://bugs.webkit.org/show_bug.cgi?id=104740
1634
1635         Compile problems on EFL
1636
1637         * GNUmakefile.am:
1638         * qmake/mkspecs/features/default_post.prf:
1639
1640 2013-01-02  Tony Chang  <tony@chromium.org>
1641
1642         Generate internal.settings from Settings.in
1643         https://bugs.webkit.org/show_bug.cgi?id=104740
1644
1645         Reviewed by Adam Barth.
1646
1647         * GNUmakefile.am: Compile new generated files on GTK+.
1648         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
1649         with both extra_sources and add_output_to_sources=false will still
1650         compile the files in extra_sources. This makes it possible to compile
1651         InternalSettingsGenerated.cpp and not be confused by
1652         InternalSettingsGenerated.idl.
1653
1654 2013-01-02  Adam Barth  <abarth@webkit.org>
1655
1656         Remove old-run-webkit-tests support from webkitpy
1657         https://bugs.webkit.org/show_bug.cgi?id=105619
1658
1659         Reviewed by Eric Seidel.
1660
1661         This patch removes support for webkitpy parsing the
1662         old-run-webkit-tests results format. There doesn't seem to be any
1663         reason to keep this code around anymore now that most bots have
1664         switched to new-run-webkit-tests.
1665
1666         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1667         (Builder.fetch_layout_test_results):
1668         * Scripts/webkitpy/common/net/layouttestresults.py:
1669         (path_for_layout_test):
1670         (LayoutTestResults.results_from_string):
1671         (LayoutTestResults):
1672         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
1673         (LayoutTestResultsTest.test_results_from_string):
1674         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
1675         (test_flaky_test_failure):
1676         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
1677         (LayoutTestResultsReaderTest.test_missing_layout_test_results):
1678         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
1679         (PatchAnalysisTask._test_patch):
1680         * Scripts/webkitpy/tool/commands/queuestest.py:
1681         (QueuesTest.assert_queue_outputs):
1682
1683 2013-01-02  Heikki Paajanen  <heikki.paajanen@palm.com>
1684
1685         [Qt][WK2] Add simple UI to find text from page
1686         https://bugs.webkit.org/show_bug.cgi?id=96481
1687
1688         Reviewed by Jocelyn Turcotte.
1689
1690         Find bar that slides between navigation bar and WebView.
1691
1692         * MiniBrowser/qt/BrowserWindow.cpp:
1693         (BrowserWindow::toggleFind):
1694         * MiniBrowser/qt/BrowserWindow.h:
1695         (BrowserWindow):
1696         * MiniBrowser/qt/MiniBrowser.qrc:
1697         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1698         (MiniBrowserApplication::notify):
1699         * MiniBrowser/qt/icons/find.png: Added.
1700         * MiniBrowser/qt/qml/BrowserWindow.qml:
1701
1702 2013-01-02  Christophe Dumez  <christophe.dumez@intel.com>
1703
1704         Regression(r138603): Caused 2 webkitpy tests to fail
1705         https://bugs.webkit.org/show_bug.cgi?id=105913
1706
1707         Reviewed by Kentaro Hara.
1708
1709         Fix 2 webkitpy tests that started to fail after
1710         r138603.
1711
1712         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1713         (IRCCommandTest._sheriff_test_data_url): Make sure the
1714         returned test file URL is valid no matter what the
1715         current working directory is.
1716         * Scripts/webkitpy/tool/bot/ircbot_unittest.py:
1717         (IRCBotTest.test_help): Add "sheriffs" command to the
1718         list of available commands.
1719
1720 2013-01-01  Christophe Dumez  <christophe.dumez@intel.com>
1721
1722         [WK2] plugins/clicking-missing-plugin-fires-delegate.html is failing
1723         https://bugs.webkit.org/show_bug.cgi?id=105897
1724
1725         Reviewed by Alexey Proskuryakov.
1726
1727         Dump in WebKitTestRunner the message that is expected by layout tests
1728         when the missing plugin button is clicked.
1729
1730         * WebKitTestRunner/TestController.cpp:
1731         (WTR::TestController::createWebViewWithOptions):
1732         (WTR::TestController::unavailablePluginButtonClicked):
1733         (WTR):
1734         * WebKitTestRunner/TestController.h:
1735         (TestController):
1736
1737 2013-01-01  KwangYong Choi  <ky0.choi@samsung.com>
1738
1739         [EFL] Enable MHTML feature
1740         https://bugs.webkit.org/show_bug.cgi?id=105815
1741
1742         Reviewed by Laszlo Gombos.
1743
1744         Enable MHTML feature for EFL port by default.
1745
1746         * Scripts/webkitperl/FeatureList.pm:
1747
1748 2013-01-01  Alan Cutter  <alancutter@chromium.org>
1749
1750         sheriff-bot should know who the gardeners/sheriffs are
1751         https://bugs.webkit.org/show_bug.cgi?id=105698
1752
1753         Reviewed by Eric Seidel.
1754
1755         Added a "sheriffs" command to sheriffbot to fetch the current Chromium Webkit sheriffs.
1756
1757         * Scripts/webkitpy/tool/bot/irc_command.py:
1758         (Current):
1759         (Current._retrieve_webkit_sheriffs):
1760         (Current.execute):
1761         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1762         (IRCCommandTest.test_current):
1763         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_0.js: Added.
1764         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_1.js: Added.
1765         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_2.js: Added.
1766         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_malformed.js: Added.
1767         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_zero.js: Added.
1768
1769 2012-12-31  Carlos Garcia Campos  <cgarcia@igalia.com>
1770
1771         [GTK] Move ImageDiff to the Tools directory root
1772         https://bugs.webkit.org/show_bug.cgi?id=105421
1773
1774         Reviewed by Kenneth Rohde Christiansen.
1775
1776         * DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Removed.
1777         * GNUmakefile.am:
1778         * ImageDiff/gtk/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/gtk/ImageDiff.cpp.
1779
1780 2012-12-30  Mike West  <mkwst@chromium.org>
1781
1782         Add myself to watchlists for CSP and console logs.
1783         https://bugs.webkit.org/show_bug.cgi?id=105874
1784
1785         Reviewed by Dirk Pranke.
1786
1787         Because I really, really, really need more email.
1788
1789         * Scripts/webkitpy/common/config/committers.py:
1790             Add mkwst+watchlist@chromium.org as a contributor.
1791         * Scripts/webkitpy/common/config/watchlist:
1792             Create watchlist categories for 1) the various mechanisms used to
1793             generate console messages, and 2) CSP usage and files.
1794
1795 2012-12-30  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1796
1797         [EFL] Enable TEMPLATE_ELEMENT feature
1798         https://bugs.webkit.org/show_bug.cgi?id=105865
1799         
1800         Reviewed by Laszlo Gombos.
1801
1802         * Scripts/webkitperl/FeatureList.pm: Enable template element feature for EFL port.
1803
1804 2012-12-30  Zan Dobersek  <zandobersek@gmail.com>
1805
1806         Unreviewed GTK gardening of the unit tests.
1807
1808         Skip the TestInspectorServer unit test that's timing out completely.
1809
1810         * Scripts/run-gtk-tests:
1811         (TestRunner):
1812
1813 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
1814
1815         Web Inspector: Native Memory Instrumentation plugin: do not generate 'not instrumented' warning for instrumented mutable members.
1816         https://bugs.webkit.org/show_bug.cgi?id=105855
1817
1818         Reviewed by Vsevolod Vlasov.
1819
1820         Extract MemberExpr from ImplicitCastExpr. It happens when we instrument a mutable member because
1821         addMember expects const T& and the mutable ref to member implicitly converts into const ref to member.
1822
1823         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
1824         (ReportMemoryUsageVisitor):
1825         (clang::AddMemberCallVisitor::VisitCallExpr):
1826         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
1827         (clang::ReportMemoryUsageVisitor::checkMembersCoverage):
1828
1829 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
1830
1831         Web Inspector: Native Memory Instrumentation plugin: move function bodies out of class declarations.
1832         https://bugs.webkit.org/show_bug.cgi?id=105852
1833
1834         Reviewed by Alexander Pavlov.
1835
1836         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
1837         (clang):
1838         (AddMemberCallVisitor):
1839         (ReportMemoryUsageVisitor):
1840         (clang::ReportMemoryUsageAction::ParseArgs):
1841         (clang::AddMemberCallVisitor::VisitCallExpr):
1842         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
1843         (clang::ReportMemoryUsageVisitor::emitWarning):
1844         (clang::ReportMemoryUsageVisitor::findInstrumentationMethod):
1845         (clang::ReportMemoryUsageVisitor::needsToBeInstrumented):
1846         (clang::ReportMemoryUsageVisitor::CheckMembersCoverage):
1847
1848 2012-12-29  Zan Dobersek  <zandobersek@gmail.com>
1849
1850         [webkitpy] Omit webkitpy/thirdparty/BeautifulSoup.py from code coverage checking
1851         https://bugs.webkit.org/show_bug.cgi?id=105835
1852
1853         Reviewed by Dirk Pranke.
1854
1855         The BeautifulSoup module is a third-party import and as such we don't have much control
1856         over its testing. Because of that it should be omitted from code coverage checking.
1857
1858         * Scripts/webkitpy/test/main.py:
1859         (Tester._run_tests):
1860
1861 2012-12-28  Martin Robinson  <mrobinson@igalia.com>
1862
1863         [GTK][WK2] Add support for IME Composition
1864         https://bugs.webkit.org/show_bug.cgi?id=65093
1865
1866         Reviewed by Carlos Garcia Campos.
1867
1868         Update unit tests for GtkInputMethodFilter to note whether or not an event
1869         was a fake event synthesized for composition purposes. This is now used
1870         for WebKit2.
1871
1872         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
1873         (TestWebKitAPI::TestInputMethodFilter::sendSimpleKeyEvent):
1874         (TestWebKitAPI::TestInputMethodFilter::sendKeyEventWithCompositionResults):
1875         (TestWebKitAPI::TEST):
1876         * gtk/generate-gtkdoc:
1877         (get_webkit2_options): Skip new files in the WebKit2 source directory for documentation
1878         generation.
1879
1880 2012-12-27  Ilya Tikhonovsky  <loislo@chromium.org>
1881
1882         Web Inspector: Native Memory Instrumentation. Update clang plugin according to the current state of memory instrumentation code.
1883         https://bugs.webkit.org/show_bug.cgi?id=105800
1884
1885         Reviewed by Yury Semikhatsky.
1886
1887         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
1888         (clang::ReportMemoryUsageConsumer::ReportMemoryUsageConsumer):
1889         (clang):
1890
1891 2012-12-27  Zan Dobersek  <zandobersek@gmail.com>
1892
1893         Create a GTK build system watchlist and add myself to it
1894         https://bugs.webkit.org/show_bug.cgi?id=105769
1895
1896         Reviewed by Dirk Pranke.
1897
1898         Add a GTK build system watchlist, watching the configure.ac, GNUmakefile.am files
1899         and the input GNUmakefile.features.am.in file. The GNUmakefile.list.am files are
1900         omitted on purpose due to relatively high change rate they are exposed to.
1901
1902         The new watchlist would be of use to me as the GTK build system recently went
1903         through a series of changes that included a changed policy on adding configuration
1904         flags to configure.ac and subsequent additions to GNUmakefile.am files. A PSA on the
1905         mailing list about this change will be published later, but I'd also like to be
1906         notified of any pending changes to these files just so I can check the changes are
1907         in compliance with the new policy.
1908
1909         * Scripts/webkitpy/common/config/watchlist:
1910
1911 2012-12-26  Ilya Tikhonovsky  <loislo@chromium.org>
1912
1913         Unreviewed. Replace find("Tools") with rfind("Tools") because base dir of WebKit may have Tools word.
1914         As example ..../DevTools/src/third_party/WebKit
1915
1916         * Scripts/webkitpy/common/webkit_finder.py:
1917         (WebKitFinder.webkit_base):
1918
1919 2012-12-25  Christophe Dumez  <christophe.dumez@intel.com>
1920
1921         [EFL][WK2] Refactor snapshot taking code
1922         https://bugs.webkit.org/show_bug.cgi?id=105687
1923
1924         Reviewed by Kenneth Rohde Christiansen.
1925
1926         Call WKViewCreateSnapshot() instead of WKViewGetSnapshot()
1927         since the function was renamed to follow naming style.
1928
1929         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
1930         (WTR::PlatformWebView::windowSnapshotImage):
1931
1932 2012-12-24  Ilya Tikhonovsky  <loislo@chromium.org>
1933
1934         Web Inspector: Native Memory Instrumentation: propagate member type as edge type to the serialized heap graph.
1935         https://bugs.webkit.org/show_bug.cgi?id=105725
1936
1937         Reviewed by Yury Semikhatsky.
1938
1939         MemberType value names were adjusted according to Style Guide.
1940         Existing tests were extended with link type validation.
1941
1942         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1943
1944 2012-12-24  Xiaobo Wang  <xbwang@torchmobile.com.cn>
1945
1946         [BlackBerry] DRT - Crashed when running pixels tests
1947         https://bugs.webkit.org/show_bug.cgi?id=105711
1948
1949         RIM PR 270003
1950
1951         Reviewed by George Staikos.
1952
1953         Calling window->post() here will result in buffer swapping which is
1954         only allowed in UI thread now, and I found we don't need to call it
1955         for GL renderer.
1956
1957         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
1958         (createBitmapContextFromWebView):
1959
1960 2012-12-22  Dan Winship  <danw@gnome.org>
1961
1962         Add a patch to fix libxml2 2.9.0 regression
1963         https://bugs.webkit.org/show_bug.cgi?id=104494
1964
1965         Reviewed by Martin Robinson.
1966
1967         * gtk/jhbuild.modules:
1968         * gtk/patches/libxml2-2.9.0-dtd.patch: patch from upstream
1969
1970 2012-12-23  Allan Sandfeld Jensen  <allan.jensen@digia.com>
1971
1972         Unreviewed update of status to reviewer.
1973
1974         * Scripts/webkitpy/common/config/committers.py:
1975
1976 2012-12-23  Philippe Normand  <pnormand@igalia.com>
1977
1978         [GTK][jhbuild] Bump GStreamer to 1.0.4
1979         https://bugs.webkit.org/show_bug.cgi?id=105640
1980
1981         Reviewed by Martin Robinson.
1982
1983         * gtk/jhbuild.modules: Update GStreamer modules to 1.0.4.
1984
1985 2012-12-21  Nico Weber  <thakis@chromium.org>
1986
1987         chromium: webkit-build-directory doesn't handle ninja output dirs
1988         https://bugs.webkit.org/show_bug.cgi?id=105671
1989
1990         Reviewed by Daniel Bates.
1991
1992         Teach webkit-build-directory to detect the ninja build directory.
1993
1994         * Scripts/webkitdirs.pm:
1995         (determineBaseProductDir):
1996         (determineIsChromiumNinja):
1997
1998 2012-12-21  Ryosuke Niwa  <rniwa@webkit.org>
1999
2000         [Mountain Lion] platform/mac/editing/spelling/removing-underline-after-accepting-autocorrection-using-punctuation.html failing
2001         https://bugs.webkit.org/show_bug.cgi?id=95477
2002
2003         Reviewed by Sam Weinig.
2004
2005         Co-authored by Jia Pu.
2006
2007         Specify a fixed autocorrection dictionary in resetDefaultsToConsistentValues() so that
2008         we can have consistent autocorrection behavior when running tests.
2009
2010         * DumpRenderTree/mac/DumpRenderTree.mm:
2011         (resetDefaultsToConsistentValues):
2012         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
2013         (WTR::InjectedBundle::platformInitialize):
2014
2015 2012-12-21  Zan Dobersek  <zandobersek@gmail.com>
2016
2017         [GTK] Add WebKit1-specific TestExpectations file
2018         https://bugs.webkit.org/show_bug.cgi?id=97562
2019
2020         Reviewed by Dirk Pranke.
2021
2022         Implement the expectations_files method inside the GtkPort class that
2023         returns the list of the TestExpectations files the current test configuration
2024         requires. The method is implemented in a similar way as on the EFL port.
2025         A unit test is also added.
2026
2027         * Scripts/webkitpy/layout_tests/port/gtk.py:
2028         (GtkPort._search_paths):
2029         (GtkPort):
2030         (GtkPort.expectations_files):
2031         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
2032         (GtkPortTest.test_expectations_files):
2033
2034 2012-12-21  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
2035
2036         [EFL] Enable feature option for the CSS sticky position
2037         https://bugs.webkit.org/show_bug.cgi?id=105630
2038
2039         Reviewed by Kenneth Rohde Christiansen.
2040
2041         Enable CSS sticky position feature by default for the
2042         EFL port. This feature option was added in r138252.
2043
2044         No new tests. Existing layout tests pass.
2045
2046         * Scripts/webkitperl/FeatureList.pm:
2047
2048 2012-12-21  Jussi Kukkonen  <jussi.kukkonen@intel.com>
2049
2050         [WK2] REGRESSION(r138232): It made 4 tests timeout
2051         https://bugs.webkit.org/show_bug.cgi?id=105623
2052
2053         Reviewed by Kenneth Rohde Christiansen.
2054
2055         r138232 changed functionality so that empty text output
2056         no longer leads to a dump() call, which leads to a timeout.
2057         Instead do a text dump if there is text output or if there
2058         is no audio output.
2059
2060         * WebKitTestRunner/TestInvocation.cpp:
2061         (WTR::TestInvocation::dumpResults):
2062
2063 2012-12-21  Zeno Albisser  <zeno@webkit.org>
2064
2065         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
2066         https://bugs.webkit.org/show_bug.cgi?id=105309
2067
2068         On Windows the library name is Qt5WebKitWidgets as well.
2069         So r138136 should not have adjusted the library name for Windows.
2070
2071         Reviewed by Simon Hausmann.
2072
2073         * Scripts/webkitdirs.pm:
2074         (builtDylibPathForName):
2075
2076 2012-12-21  Zeno Albisser  <zeno@webkit.org>
2077
2078         [Qt] Add new TestExpectations for qt-5.0-mac-wk2.
2079         https://bugs.webkit.org/show_bug.cgi?id=105424
2080
2081         This change will allow us to run a very specific
2082         subset of the available LayoutTests on a mac buildbot.
2083         The new baseline search path for mac will look as follows:
2084         qt-5.0-mac-wk2 -> qt-5.0-wk2 -> qt-5.0 -> qt-mac -> qt
2085
2086         Reviewed by Dirk Pranke.
2087
2088         * Scripts/webkitpy/layout_tests/port/qt.py:
2089         (QtPort._search_paths):
2090         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2091         (QtPortTest):
2092
2093 2012-12-20  Ryuan Choi  <ryuan.choi@samsung.com>
2094
2095         [EFL] Build break with latest EFL libraries.
2096         https://bugs.webkit.org/show_bug.cgi?id=104827
2097
2098         Reviewed by Laszlo Gombos.
2099
2100         The eo EFL package is introduced and evas and ecore use it since 1.8.
2101         While introducing Eo, EFL changed several structures of Evas and Ecore
2102         from own specific class to Eo.
2103
2104         So, this patch adds FindEo and changes declaration of Evas, Evas_Object,
2105         Ecore_Timer to build with latest EFL libraries.
2106
2107         * DumpRenderTree/efl/CMakeLists.txt: Includes eo EFL library.
2108         * EWebLauncher/CMakeLists.txt: Ditto.
2109         * MiniBrowser/efl/CMakeLists.txt: Ditto.
2110         * ImageDiff/PlatformEfl.cmake: Ditto.
2111         * TestWebKitAPI/PlatformEfl.cmake: Ditto.
2112         * TestWebKitAPI/PlatformWebView.h: Modified declaration of Evas_Object.
2113         * WebKitTestRunner/InjectedBundle/TestRunner.h: Modified declaration of Ecore_Timer.
2114         * WebKitTestRunner/PlatformEfl.cmake: Includes eo EFL library.
2115         * WebKitTestRunner/PlatformWebView.h: Modified declaration of Evas_Object.
2116
2117 2012-12-20  Dirk Pranke  <dpranke@chromium.org>
2118
2119         TestExpectation fallback is broken
2120         https://bugs.webkit.org/show_bug.cgi?id=105583
2121
2122         Reviewed by Ryosuke Niwa.
2123
2124         When I added support for the -wk2 in port names, it looks
2125         like I broke the way we computed the list of TestExpectations
2126         files so that the implementation-version directory was left out :(.
2127         Fixed and cleaned up the tests.
2128
2129         * Scripts/webkitpy/layout_tests/port/base.py:
2130         (Port.expectations_files):
2131         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2132         (TestWebKitPort.__init__):
2133         (PortTestCase.test_skipped_directories_for_symbols):
2134         (test_skipped_directories_for_features):
2135         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
2136         (test_skipped_tests_for_unsupported_features_empty_test_list):
2137         (test_skipped_layout_tests):
2138         (test_expectations_files):
2139
2140 2012-12-20  Anders Carlsson  <andersca@apple.com>
2141
2142         Don't include WebKit2 headers when building TestWebKitAPI on windows
2143         https://bugs.webkit.org/show_bug.cgi?id=105582
2144
2145         Reviewed by Tim Horton.
2146
2147         * TestWebKitAPI/config.h:
2148         * TestWebKitAPI/win/main.cpp:
2149
2150 2012-12-20  Ryosuke Niwa  <rniwa@webkit.org>
2151
2152         REGRESSION(r133820?): SimplifyMarkupTest API test asserts
2153         https://bugs.webkit.org/show_bug.cgi?id=105370
2154
2155         Reviewed by Simon Fraser.
2156
2157         Re-enable the test.
2158
2159         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
2160         (TestWebKitAPI::TEST):
2161
2162 2012-12-20  Nico Weber  <thakis@chromium.org>
2163
2164         chromium nrwt: Pick the newest binary found in DEFAULT_BUILD_DIRECTORIES, not the first
2165         https://bugs.webkit.org/show_bug.cgi?id=105498
2166
2167         Reviewed by Dirk Pranke.
2168
2169         Use the newest binary available rather than an than always picking one
2170         build directory over another based on iteration order.
2171
2172         * Scripts/webkitpy/layout_tests/port/chromium.py:
2173         (ChromiumPort._static_build_path):
2174         Check for timestamps.
2175         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
2176         (ChromiumMacPortTest.test_build_path_timestamps):
2177         Test that out / xcodebuild selection happens based on timestamps
2178         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2179         (ChromiumWinPortTest.test_build_path_timestamps):
2180         Test that out / build selection happens based on timestamps
2181
2182 2012-12-19  Simon Fraser  <simon.fraser@apple.com>
2183
2184         Lots of sticky tests failing in WK2
2185         https://bugs.webkit.org/show_bug.cgi?id=105464
2186
2187         Reviewed by Beth Dakin.
2188
2189         WebKitTestRunner had a race between snapshotting in the UI process,
2190         and resettting after the test in the web process. InjectedBundle::done()
2191         was a bad place to call page()->resetAfterTest(), because of this race;
2192         it could reset the scroll position before the UI snapshot had been obtained.
2193         
2194         Fix by moving the call to page()->resetAfterTest() into didReceiveMessage(),
2195         for the "Reset" message which will come in before the next test.
2196
2197         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2198         (WTR::InjectedBundle::didReceiveMessage):
2199         (WTR::InjectedBundle::done):
2200
2201 2012-12-19  Filip Pizlo  <fpizlo@apple.com>
2202
2203         DFG speculation checks that take JumpList should consolidate OSRExits
2204         https://bugs.webkit.org/show_bug.cgi?id=105401
2205
2206         Reviewed by Oliver Hunt.
2207
2208         Update the profiler to understand that an OSR exit may have multiple
2209         candidate exit sites.
2210
2211         * Scripts/display-profiler-output:
2212
2213 2012-12-20  Eric Seidel  <eric@webkit.org>
2214
2215         EWSes do not need to dump build/test logs when things pass
2216         https://bugs.webkit.org/show_bug.cgi?id=105402
2217
2218         Reviewed by Dirk Pranke.
2219
2220         When we were first writing the commit-queue and EWS system
2221         and running them locally, we thought it was important to see the
2222         build/test output as it happened.  Now these queues have been running
2223         for years, and we don't use this feature anymore.  So this patch
2224         makes us stop tee-ing the output directly to stdout, and instead
2225         only print any output from commands when they fail.
2226
2227         I also added some logging when we execute each command
2228         since we are no longer printing the command output itself.
2229         This should make it much easier to see errors with the EWS
2230         system itself, now that all the passing-build output is removed.
2231
2232         This required updating a zillion unittests, since we have a bunch of
2233         "integration" style testing of the queue-bots.  Many of the unittests for the bots were
2234         already approximating this behavior by using a "logging" Executive.
2235         I disabled logging in those cases to avoid redundant logging in those tests.
2236
2237         * Scripts/webkitpy/common/system/executive.py:
2238         (Executive.command_for_printing):
2239         (Executive.run_command):
2240         * Scripts/webkitpy/common/system/executive_mock.py:
2241         (MockExecutive.command_for_printing):
2242         (MockExecutive.run_command):
2243         * Scripts/webkitpy/common/system/executive_unittest.py:
2244         (ExecutiveTest.test_auto_stringify_args):
2245         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2246         (EarlyWarningSytemTest._default_expected_logs):
2247         (_test_ews):
2248         (_test_ewses):
2249         * Scripts/webkitpy/tool/commands/queues.py:
2250         (AbstractQueue.run_webkit_patch):
2251         * Scripts/webkitpy/tool/commands/queues_unittest.py:
2252         (AbstractQueueTest._assert_run_webkit_patch):
2253         (CommitQueueTest.test_commit_queue):
2254         (test_rollout):
2255         (test_rollout_lands):
2256         (test_manual_reject_during_processing):
2257         (StyleQueueTest.test_style_queue_with_style_exception):
2258         (test_style_queue_with_watch_list_exception):
2259
2260 2012-12-20  Dominic Mazzoni  <dmazzoni@google.com>
2261
2262         AX: support clickPoint in DRT for chromium
2263         https://bugs.webkit.org/show_bug.cgi?id=97359
2264
2265         Reviewed by Chris Fleizach.
2266
2267         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
2268         several more tests can pass.
2269
2270         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2271         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2272         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
2273         (WebTestRunner):
2274         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
2275         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
2276         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2277         (AccessibilityUIElement):
2278
2279 2012-12-20  Dan Bernstein  <mitz@apple.com>
2280
2281         prepare-ChangeLog adds "Additional information of the change such as approach, rationale" where it doesn't make sense
2282         https://bugs.webkit.org/show_bug.cgi?id=99268
2283
2284         Reviewed by Alexey Proskuryakov.
2285
2286         * Scripts/prepare-ChangeLog:
2287         (generateNewChangeLogs): Removed the "Additional information of the change" line.
2288
2289 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
2290
2291         Unreviewed. Skip unit test /webkit/atk/getTextInParagraphAndBodyModerate
2292
2293         * Scripts/run-gtk-tests:
2294         (TestRunner):
2295
2296 2012-12-20  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2297
2298         [Qt] Move ImageDiff to the Tools directory root
2299         https://bugs.webkit.org/show_bug.cgi?id=105420
2300
2301         Reviewed by Tor Arne Vestbø.
2302
2303         Moved ImageDiff from outside DumpRenderTree because it is also a
2304         dependency when running pixel tests for WebKit2.
2305
2306         * ImageDiff/ImageDiff.pro: Renamed from Tools/DumpRenderTree/qt/ImageDiff.pro.
2307         * ImageDiff/qt/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/qt/ImageDiff.cpp.
2308         * Tools.pro:
2309
2310 2012-12-20  Zan Dobersek  <zandobersek@gmail.com>
2311
2312         [GTK] Remove the --enable-unstable-features configuration option
2313         https://bugs.webkit.org/show_bug.cgi?id=105327
2314
2315         Reviewed by Martin Robinson.
2316
2317         Add a missing feature option for the CSS sticky position feature.
2318
2319         * Scripts/webkitperl/FeatureList.pm:
2320
2321 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
2322
2323         Unreviewed. Skip unit test WebKit2.ForceRepaint that times out
2324
2325         * Scripts/run-gtk-tests:
2326         (TestRunner):
2327
2328 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
2329
2330         [EFL] MiniBrowser does not play Infinite Gangnam Style
2331         https://bugs.webkit.org/show_bug.cgi?id=103531
2332
2333         Reviewed by Kenneth Rohde Christiansen.
2334
2335         For EFL we would like this feature to be built by default.
2336
2337         * Scripts/webkitperl/FeatureList.pm:
2338
2339 2012-12-20  Csaba Osztrogonác  <ossy@webkit.org>
2340
2341         [Qt] EWS bots should call qmake always
2342         https://bugs.webkit.org/show_bug.cgi?id=90049
2343
2344         Reviewed by Simon Hausmann.
2345
2346         * Scripts/webkitdirs.pm:
2347         (buildQMakeProjects):
2348
2349 2012-12-20  Jochen Eisinger  <jochen@chromium.org>
2350
2351         [chromium] fix stray printf in TestRunner's WebTestProxy
2352         https://bugs.webkit.org/show_bug.cgi?id=105515
2353
2354         Reviewed by Ryosuke Niwa.
2355
2356         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
2357         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
2358
2359 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
2360
2361         [WK2] WebAudio WKTR support
2362         https://bugs.webkit.org/show_bug.cgi?id=95084
2363
2364         Reviewed by Kenneth Rohde Christiansen.
2365
2366         Implementing audio data dump support for WebKitTestRunner,
2367         by adding the API to the IDL, enabling IPC for transferring
2368         the audio data as bytebuffer and dumping it according
2369         to run-webkit-tests script's expectations.
2370         This enables audio testing on WebKit2 ports.
2371         Based on initial wip-patch from Philippe Normand - thanks.
2372
2373         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Adding setAudioData API.
2374         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Assign audio buffer data to a local WKDataRef after extracting it from WK2 using WKBundleDataFromUint8Array bundle API.
2375         (WTR::TestRunner::setAudioData):
2376         (WTR):
2377         * WebKitTestRunner/InjectedBundle/TestRunner.h: Add a whatToDump state for dumping Audio.
2378         (TestRunner):
2379         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Add audio to the message that's transferred to the WKTR process.
2380         (WTR::InjectedBundle::done):
2381         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: Allow setting and audio result handle to be transferred.
2382         (WTR::InjectedBundle::setAudioResult):
2383         (InjectedBundle):
2384         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2385         (WTR::InjectedBundlePage::dump): Dummy handling the Audio case in the switch, build fix for Mac.
2386         * WebKitTestRunner/TestInvocation.cpp: Audio dump support, dump audio if we haven't received a text dump from the web process.
2387         (WTR::TestInvocation::dumpResults):
2388         (WTR):
2389         (WTR::TestInvocation::dumpAudio):
2390         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Storing received audio data before dumping.
2391         * WebKitTestRunner/TestInvocation.h:
2392         (TestInvocation):
2393
2394 2012-12-20  Alberto Garcia  <agarcia@igalia.com>
2395
2396         [GTK] Count failed and timed out tests separately
2397         https://bugs.webkit.org/show_bug.cgi?id=105448
2398
2399         Reviewed by Carlos Garcia Campos.
2400
2401         Display the number of tests that fail, time out and are skipped
2402         separately in the buildbot page.
2403
2404         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2405         (RunGtkAPITests.commandComplete):
2406         (RunGtkAPITests.evaluateCommand):
2407         (RunGtkAPITests.getText2):
2408         * Scripts/run-gtk-tests:
2409         (TestRunner.run_tests):
2410
2411 2012-12-19  Ryosuke Niwa  <rniwa@webkit.org>
2412
2413         PerfTest.parse_output does too much
2414         https://bugs.webkit.org/show_bug.cgi?id=105391
2415
2416         Reviewed by Tony Chang.
2417
2418         Moved the code to filter lines into PerfTest._filter_output, which has been renamed from _filter_stderr.
2419         Also moved the code to output test results into PerfTest._run_with_driver.
2420
2421         * Scripts/webkitpy/performance_tests/perftest.py:
2422         (PerfTest.__init__): Added _description.
2423         (PerfTest.description): Added.
2424         (PerfTest._run_with_driver): Moved the code to output test description and test results.
2425         (PerfTest._filter_output): Moved and renamed from PerfTest._filter_stderr.
2426         (PerfTest.parse_output): Removed the code to output test results. Return test results and description
2427         respectively. Also removed the code that allowed some tests to omit values since all tests report each
2428         iteration now since r136492.
2429         (PerfTest.output_statistics): Removed the code to print test description, now done in _run_with_driver.
2430         (ChromiumStylePerfTest._run_with_driver): Added. Chromium style tests are sufficiently different from
2431         regular PerfTest that it doesn't make much sense to share _run_with_driver. But really, we should just
2432         get rid of this type of test altogether in favor of regular performance tests that uses runner.js.
2433         (ChromiumStylePerfTest.parse_and_log_output): Renamed from parse_output.
2434         (PageLoadingPerfTest._run_with_driver): Removed the explicit '' for the test description.
2435         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
2436         (MainTest.test_parse_output): Removed the expected logs since parse_output no longer prints out results.
2437         Also added a call to _filter_output since parse_output doesn't filter the output text anymore.
2438         (MainTest.test_parse_output_with_failing_line): Added a call to _filter_output. Also added ',' after 'Time:'
2439         so that it's not string-concatenated with the next line.
2440         (MainTest.test_parse_output_with_description): Added; a test for PerfTest.description().
2441         (MainTest.test_parse_output_with_subtests): Removed the expected logs, and added a call to _filter_output.
2442
2443 2012-12-19  Alexis Menard  <alexis@webkit.org>
2444
2445         Implement CSS parsing for CSS transitions unprefixed.
2446         https://bugs.webkit.org/show_bug.cgi?id=104804
2447
2448         Reviewed by Dean Jackson.
2449
2450         Add a new flag ENABLE_CSS_TRANSFORMS_ANIMATIONS_TRANSITIONS_UNPREFIXED
2451         to cover the work of unprefixing Transforms, Animations and 
2452         Transitions. It will let the possibility of each ports to turn it off 
2453         in their release branches until we're confident that these CSS 
2454         properties are ready to be unprefixed.
2455
2456         * Scripts/webkitperl/FeatureList.pm:
2457         * qmake/mkspecs/features/features.pri:
2458
2459 2012-12-19  Julien Chaffraix  <jchaffraix@webkit.org>
2460
2461         2 clampTo unit tests are failing
2462         https://bugs.webkit.org/show_bug.cgi?id=105253
2463
2464         Reviewed by Emil A Eklund.
2465
2466         Removed clampToUnsigned as it was actually testing exactly what
2467         clampToUnsignedLong was doing. Also made clampToIntFloat more robust
2468         to precision loss (fix suggested by Emil).
2469
2470         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
2471
2472 2012-12-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2473
2474         [EFL] Move ImageDiff to the Tools directory root
2475         https://bugs.webkit.org/show_bug.cgi?id=105305
2476
2477         Reviewed by Laszlo Gombos.
2478
2479         We previously moved this directory from the DRT directory to
2480         efl/ImageDiff. The major reason for doing so was because ImageDiff does
2481         not depend on DumpRenderTree but for legacy reasons was living inside the
2482         DRT folder, even though it is needed when running pixel tests for WK2.
2483
2484         Now we are moving to Tool directory root as other ports agreed to do the same.
2485
2486         * CMakeLists.txt:
2487         * ImageDiff/CMakeLists.txt: Added.
2488         * ImageDiff/PlatformEfl.cmake: Added.
2489         * ImageDiff/efl/ImageDiff.cpp: Renamed from Tools/efl/ImageDiff/ImageDiff.cpp.
2490         * efl/ImageDiff/CMakeLists.txt: Removed.
2491
2492 2012-12-19  Zeno Albisser  <zeno@webkit.org>
2493
2494         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
2495         https://bugs.webkit.org/show_bug.cgi?id=105309
2496
2497         The version number is only used as a part
2498         of the library name on Linux.
2499
2500         Reviewed by Kenneth Rohde Christiansen.
2501
2502         * Scripts/webkitdirs.pm:
2503         (builtDylibPathForName):
2504
2505 2012-12-19  Alberto Garcia  <agarcia@igalia.com>
2506
2507         [GTK] Tests that time out are not considered as failures.
2508         https://bugs.webkit.org/show_bug.cgi?id=105318
2509
2510         Reviewed by Martin Robinson.
2511
2512         Fix parsing of the GTK API tests log files.
2513
2514         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2515         (RunGtkAPITests.commandComplete):
2516
2517 2012-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
2518
2519         Unreviewed, rolling out r138061.
2520         https://bugs.webkit.org/show_bug.cgi?id=105396
2521
2522         Original patch:
2523         http://trac.webkit.org/changeset/138061
2524         https://bugs.webkit.org/show_bug.cgi?id=97359
2525
2526         It is crashing on Debug bots
2527
2528         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2529         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2530         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
2531         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2532         (AccessibilityUIElement):
2533
2534 2012-12-18  Alice Liu  <alice.liu@apple.com>
2535
2536         Test for https://bugs.webkit.org/show_bug.cgi?id=105106
2537         Add SPI to WebKit1 WebFrame for hit testing
2538
2539         Reviewed by Dan Bernstein.
2540
2541         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/ElementAtPointInWebFrame.mm
2542         * TestWebKitAPI/Tests/mac/ElementAtPointInWebFrame.mm: Added.
2543         (TestWebKitAPI::TEST): Loads html with two divs positioned in the 2nd and 4th quadrants of the webview.
2544         Then hit-tests at three points, expecting to hit the two divs and body element. 
2545
2546 2012-12-18  Alice Liu  <alice.liu@apple.com>
2547
2548         Test for https://bugs.webkit.org/show_bug.cgi?id=105262
2549         Add SPI to WebKit1 WebFrame for node conversion to JSValueRef
2550
2551         Reviewed by Anders Carlsson.
2552
2553         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/JSWrapperForNodeInWebFrame.mm.mm
2554         * TestWebKitAPI/Tests/mac/JSWrapperForNodeInWebFrame.mm.mm: Added.
2555         (TestWebKitAPI::TEST): Tests for the correct JS wrapper for a DOMNode, provided a WebScriptWorld.
2556         Loads html with a single div element. In an isolated world, creates a property on that node.
2557         Also, in the standard world, creates a different property on that node.  Then tests for 4 things:
2558             - Existence of the isolated property in the isolated world.
2559             - Existence of the standard property in the standard world.
2560             - Non-existence of the isolated property in the standard world.
2561             - Non-existence of the standard property in the isolated world.
2562
2563 2012-12-18  Simon Fraser  <simon.fraser@apple.com>
2564
2565         Disable the SimplifyMarkup API test, wich asserts in debug builds.
2566         Tracked by https://bugs.webkit.org/show_bug.cgi?id=105370.
2567         
2568         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
2569         (TestWebKitAPI::TEST):
2570
2571 2012-12-18  Joseph Pecoraro  <pecoraro@apple.com>
2572
2573         [Mac] Add Build Phase to Check Headers for Inappropriate Macros (Platform.h macros)
2574         https://bugs.webkit.org/show_bug.cgi?id=104279
2575
2576         Reviewed by David Kilzer.
2577
2578         Add a build phase script that checks a framework's header files for
2579         Platform.h and Compiler.h macros. Also add some tests for the script.
2580
2581         * Scripts/check-for-inappropriate-macros-in-external-headers: Added.
2582         Script to be used in build phases to check for inappropriate macros in headers.
2583
2584         * Scripts/test-webkitruby: Added.
2585         Script that can be used to run all webkitruby tests.
2586
2587         * Scripts/test-webkit-scripts:
2588         Also run the ruby tests.
2589
2590         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/fake-data-failing-expected.txt: Added.
2591         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/pass-expected.txt: Added.
2592         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Fail.h: Added.
2593         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Pass.h: Added.
2594         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/run-test.rb: Added.
2595         Tests for the new script.
2596
2597 2012-12-18  Dominic Mazzoni  <dmazzoni@google.com>
2598
2599         AX: support clickPoint in DRT for chromium
2600         https://bugs.webkit.org/show_bug.cgi?id=97359
2601
2602         Reviewed by Chris Fleizach.
2603
2604         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
2605         several more tests can pass.
2606
2607         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2608         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2609         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
2610         (WebTestRunner):
2611         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
2612         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
2613         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2614         (AccessibilityUIElement):
2615
2616 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
2617
2618         Build fix after r138045.
2619
2620         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2621         (PerfTestsRunner._collect_tests): Use relative_path. We skip only when this path
2622         is not explicitly specified. The condition was backwards but never matched so we
2623         always ran the test regardless of whether it's listed in the skipped list or not.
2624         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2625         (test_collect_tests_with_ignored_skipped_list): Renamed so that it actually runs.
2626
2627 2012-12-18  Emil A Eklund  <eae@chromium.org>
2628
2629         TestWebKitAPI/Tests/WebCore/LayoutUnit fails on mac
2630         https://bugs.webkit.org/show_bug.cgi?id=105332
2631         
2632         Reviewed by Levi Weintraub.
2633         
2634         Create new test suite for LayoutUnit tests as the KURL tests are
2635         of the incompatible TEST_F type.
2636
2637         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
2638         (TestWebKitAPI):
2639         (TestWebKitAPI::TEST_F):
2640
2641 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
2642
2643         Running a skipped test with run-perf-tests could alert the user
2644         https://bugs.webkit.org/show_bug.cgi?id=105219
2645
2646         Reviewed by Eric Seidel.
2647
2648         Force run skipped performance tests explicitly specified in the command line.
2649         Also warn the user when the specified paths don't exist.
2650
2651         * Scripts/webkitpy/layout_tests/port/base.py:
2652         (Port.relative_test_filename):
2653         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2654         (PerfTestsRunner._parse_args):
2655         (PerfTestsRunner._collect_tests):
2656         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2657         (test_collect_tests_with_skipped_list_and_files):
2658
2659 2012-12-18  Yaron Friedman  <yfriedman@chromium.org>
2660
2661         Chrome Android java output is too verbose
2662         https://bugs.webkit.org/show_bug.cgi?id=105247
2663
2664         Reviewed by Tony Chang.
2665
2666         Pass -quiet to ant to suppress extraneous output.
2667
2668         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2669         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
2670
2671 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
2672
2673         Consider removing --pause-before-testing option
2674         https://bugs.webkit.org/show_bug.cgi?id=105248
2675
2676         Reviewed by Eric Seidel.
2677
2678         Remove --pause-before-testing option.
2679
2680         * Tools/Scripts/webkitpy/layout_tests/port/driver.py:
2681         (DriverProxy.start): Removed.
2682         * Scripts/webkitpy/performance_tests/perftest.py:
2683         (PerfTest.__init__):
2684         (PerfTest.test_path): Renamed from path_or_url.
2685         (PerfTest.run):
2686         (PerfTest): Removed "driver" argument.
2687         (PerfTest._run_with_driver):
2688         (PerfTest.run_single):
2689         (ChromiumStylePerfTest.__init__):
2690         (PageLoadingPerfTest.__init__):
2691         (PageLoadingPerfTest.run_single):
2692         (PageLoadingPerfTest._run_with_driver): Renamed from run so that PerfTest.run can create a driver for us.
2693         (ReplayPerfTest.__init__):
2694         (ReplayPerfTest.prepare):
2695         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
2696         (TestPageLoadingPerfTest.test_run):
2697         (TestPageLoadingPerfTest.test_run_with_memory_output):
2698         (TestPageLoadingPerfTest.test_run_with_bad_output):
2699         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2700         (PerfTestsRunner._parse_args):
2701         (PerfTestsRunner._run_tests_set): Removed the code to create a driver and pause before running the test.
2702         (PerfTestsRunner._run_single_test): Removed "driver" argument.
2703         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2704         (run_test):
2705         (test_run_test_pause_before_testing): Removed.
2706
2707 2012-12-18  Alberto Garcia  <agarcia@igalia.com>
2708
2709         [GTK] Tests that time out are not considered as failures.
2710         https://bugs.webkit.org/show_bug.cgi?id=105318
2711
2712         Reviewed by Martin Robinson.
2713
2714         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2715         (RunGtkAPITests.commandComplete):
2716         Consider log lines that end with TIMEOUT as errors.
2717
2718         * Scripts/run-gtk-tests:
2719         (TestRunner.run_tests):
2720         Consider timeouts as failures when returning the number of failed
2721         tests.
2722
2723 2012-12-18  Pravin D  <pravin.d@samsung.com>
2724
2725         Unreviewed, adding secondary email.
2726
2727         * Scripts/webkitpy/common/config/committers.py:
2728
2729 2012-12-18  Pravin D  <pravin.d@samsung.com>
2730
2731         Unreviewed, promoting myself to committer's list.
2732
2733         * Scripts/webkitpy/common/config/committers.py:
2734
2735 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2736
2737         [EFL] ImageDiff does not build on WebKit2 only builds
2738         https://bugs.webkit.org/show_bug.cgi?id=105288
2739
2740         Reviewed by Kenneth Rohde Christiansen.
2741
2742         Moved ImageDiff to the efl/ folder inside Tools/ (so we don't pollute
2743         the folder with things which are EFL-specific.
2744
2745         ImageDiff will be built by default now regardless of WebKit API flavor
2746         you choose.
2747
2748         * CMakeLists.txt:
2749         * DumpRenderTree/efl/CMakeLists.txt:
2750         * efl/ImageDiff/CMakeLists.txt: Added.
2751         * efl/ImageDiff/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/efl/ImageDiff.cpp.
2752
2753 2012-12-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2754
2755         [EFL][WK2] When creating new windows, use the opener's context
2756         https://bugs.webkit.org/show_bug.cgi?id=105283
2757
2758         Reviewed by Laszlo Gombos.
2759
2760         Use the context of the opener if exists.
2761
2762         * MiniBrowser/efl/main.c:
2763         (on_key_down):
2764         (on_window_create):
2765         (window_create):
2766         (elm_main):
2767
2768 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2769
2770         [EFL][WK2] Gardening of TestWebKitAPI tests
2771         https://bugs.webkit.org/show_bug.cgi?id=105268
2772
2773         Reviewed by Kenneth Rohde Christiansen.
2774
2775         Some tests were marked as failing but are passing and the newly added
2776         LayoutUnit should have it's own binary because it conflicts with the
2777         fixture used by KURL and is failing on the bots because of that.
2778
2779         * TestWebKitAPI/CMakeLists.txt:
2780         * TestWebKitAPI/PlatformEfl.cmake:
2781
2782 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
2783
2784         [chromium] move webintent related methods to TestRunner library
2785         https://bugs.webkit.org/show_bug.cgi?id=105164
2786
2787         Reviewed by Darin Fisher.
2788
2789         Also deleted the unused deliveredIntentFailure and
2790         deliveredIntentResult methods.
2791
2792         * DumpRenderTree/chromium/DRTTestRunner.cpp:
2793         (DRTTestRunner::DRTTestRunner):
2794         * DumpRenderTree/chromium/DRTTestRunner.h:
2795         (DRTTestRunner):
2796         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
2797         (WebKit):
2798         (WebTestRunner::WebTestDelegate::setCurrentWebIntentRequest):
2799         (WebTestRunner::WebTestDelegate::currentWebIntentRequest):
2800         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
2801         (WebKit):
2802         (WebTestProxyBase):
2803         (WebTestRunner::WebTestProxy::registerIntentService):
2804         (WebTestRunner::WebTestProxy::dispatchIntent):
2805         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
2806         (WebTestRunner::TestRunner::TestRunner):
2807         (WebTestRunner::TestRunner::sendWebIntentResponse):
2808         (WebTestRunner):
2809         (WebTestRunner::TestRunner::deliverWebIntent):
2810         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
2811         (TestRunner):
2812         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
2813         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
2814         (WebTestRunner::WebTestProxyBase::registerIntentService):
2815         (WebTestRunner):
2816         (WebTestRunner::WebTestProxyBase::dispatchIntent):
2817         * DumpRenderTree/chromium/WebViewHost.cpp:
2818         (WebViewHost::setCurrentWebIntentRequest):
2819         (WebViewHost::currentWebIntentRequest):
2820         * DumpRenderTree/chromium/WebViewHost.h:
2821         (WebKit):
2822         (WebViewHost):
2823
2824 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
2825
2826         [chromium] move editing callback dumping to the WebTestProxy
2827         https://bugs.webkit.org/show_bug.cgi?id=105013
2828
2829         Reviewed by Darin Fisher.
2830
2831         To decide whether or not to dump the editing callbacks, the WebTestProxy
2832         needs to query the TestRunner. For this purpose, I introduce a
2833         WebTestRunner class that just exposes the shouldDumpEditingCallbacks
2834         method. In contrast to e.g. WebEventSender, the WebTestRunner is just
2835         an interface, and it is not owned by the WebTestInterfaces, but the
2836         embedder of the TestRunner library must provide it. This will change
2837         once all TestRunner methods are in the library.
2838
2839         * DumpRenderTree/DumpRenderTree.gypi:
2840         * DumpRenderTree/chromium/DRTTestRunner.cpp:
2841         (DRTTestRunner::DRTTestRunner):
2842         (DRTTestRunner::reset):
2843         * DumpRenderTree/chromium/DRTTestRunner.h:
2844         (DRTTestRunner):
2845         (DRTTestRunner::taskList):
2846         (DRTTestRunner::WorkQueue::taskList):
2847         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
2848         (WebTestRunner):
2849         (WebTestInterfaces):
2850         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
2851         (WebKit):
2852         (WebTestRunner):
2853         (WebTestProxyBase):
2854         (WebTestRunner::WebTestProxy::shouldBeginEditing):
2855         (WebTestRunner::WebTestProxy::shouldEndEditing):
2856         (WebTestRunner::WebTestProxy::shouldInsertNode):
2857         (WebTestRunner::WebTestProxy::shouldInsertText):
2858         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
2859         (WebTestRunner::WebTestProxy::shouldDeleteRange):
2860         (WebTestRunner::WebTestProxy::shouldApplyStyle):
2861         (WebTestRunner::WebTestProxy::didBeginEditing):
2862         (WebTestRunner::WebTestProxy::didChangeSelection):
2863         (WebTestRunner::WebTestProxy::didChangeContents):
2864         (WebTestRunner::WebTestProxy::didEndEditing):
2865         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h: Copied from Tools/DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h.
2866         (WebTestRunner):
2867         (WebTestRunner::WebTestRunner::shouldDumpEditingCallbacks):
2868         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
2869         (WebTestRunner::TestRunner::TestRunner):
2870         (WebTestRunner::TestRunner::~TestRunner):
2871         (WebTestRunner):
2872         (WebTestRunner::TestRunner::reset):
2873         (WebTestRunner::TestRunner::shouldDumpEditingCallbacks):
2874         (WebTestRunner::TestRunner::dumpEditingCallbacks):
2875         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
2876         (TestRunner):
2877         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
2878         (WebTestRunner::WebTestInterfaces::Internal::testRunner):
2879         (WebTestRunner::WebTestInterfaces::Internal::setTestRunner):
2880         (WebTestInterfaces::Internal):
2881         (WebTestRunner::WebTestInterfaces::Internal::Internal):
2882         (WebTestRunner::WebTestInterfaces::testRunner):
2883         (WebTestRunner):
2884         (WebTestRunner::WebTestInterfaces::setTestRunner):
2885         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
2886         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
2887         (WebTestRunner):
2888         (WebTestRunner::WebTestProxyBase::shouldBeginEditing):
2889         (WebTestRunner::WebTestProxyBase::shouldEndEditing):
2890         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
2891         (WebTestRunner::WebTestProxyBase::shouldInsertText):
2892         (WebTestRunner::WebTestProxyBase::shouldChangeSelectedRange):
2893         (WebTestRunner::WebTestProxyBase::shouldDeleteRange):
2894         (WebTestRunner::WebTestProxyBase::shouldApplyStyle):
2895         (WebTestRunner::WebTestProxyBase::didBeginEditing):
2896         (WebTestRunner::WebTestProxyBase::didChangeSelection):
2897         (WebTestRunner::WebTestProxyBase::didChangeContents):
2898         (WebTestRunner::WebTestProxyBase::didEndEditing):
2899         * DumpRenderTree/chromium/TestShell.cpp:
2900         (TestShell::initialize):
2901         (TestShell::createNewWindow):
2902         * DumpRenderTree/chromium/WebViewHost.cpp:
2903         (WebViewHost::shouldBeginEditing):
2904         (WebViewHost::shouldEndEditing):
2905         (WebViewHost::shouldInsertNode):
2906         (WebViewHost::shouldInsertText):
2907         (WebViewHost::shouldChangeSelectedRange):
2908         (WebViewHost::shouldDeleteRange):
2909         (WebViewHost::shouldApplyStyle):
2910         * DumpRenderTree/chromium/WebViewHost.h:
2911         (WebViewHost):
2912
2913 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
2914
2915         [chromium] Fix methods in TestRunner that directly modify the settings instead of going through WebPreferences
2916         https://bugs.webkit.org/show_bug.cgi?id=105154
2917
2918         Reviewed by Tony Chang.
2919
2920         If the settings are modified directly, the embedder can't keep track of
2921         them and might accidentially override the changed settings.
2922
2923         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
2924         (WebPreferences):
2925         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
2926         (WebTestRunner::TestRunner::TestRunner):
2927         (WebTestRunner::TestRunner::setAsynchronousSpellCheckingEnabled):
2928         (WebTestRunner):
2929         (WebTestRunner::TestRunner::setMinimumTimerInterval):
2930         (WebTestRunner::TestRunner::setTouchDragDropEnabled):
2931         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
2932         (TestRunner):
2933         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
2934         (WebTestRunner::WebPreferences::reset):
2935         (WebTestRunner::WebPreferences::applyTo):
2936
2937 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
2938
2939         Unreviewed unit test skipping.
2940
2941         Filed https://bugs.webkit.org/show_bug.cgi?id=105253 to re-enable them.
2942
2943         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
2944         (TestWebKitAPI):
2945         (TestWebKitAPI::TEST):
2946
2947 2012-12-17  Dean Jackson  <dino@apple.com>
2948
2949         Unreviewed. Add Silvia Pfeiffer to contributor list.
2950
2951         * Scripts/webkitpy/common/config/committers.py:
2952
2953 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
2954
2955         build.webkit.org - add more google windows bots
2956         https://bugs.webkit.org/show_bug.cgi?id=105225
2957
2958         Reviewed by Eric Seidel.
2959
2960         Adds two more windows bots (google-windows-3, a builder, and
2961         google-windows-4, a tester) for the Chromium configs.
2962
2963         * BuildSlaveSupport/build.webkit.org-config/config.json:
2964
2965 2012-12-17  Eric Seidel  <eric@webkit.org>
2966
2967         Cleanup --profiler=perf code a little and allow up to 1000 report files (instead of 100)
2968         https://bugs.webkit.org/show_bug.cgi?id=105217
2969
2970         Reviewed by Dirk Pranke.
2971
2972         Previously --profile would fail strangely if find_unused_filename returned None
2973         (which would happen as soon as we had 100 report files).
2974         Now we'll search through 1000 filenames before giving up (and assert if we failed to find one).
2975         I also made some minor cleanups to the output of perf, and tried to make the
2976         commands more self-documenting by using long-names for args.
2977
2978         * Scripts/webkitpy/common/system/profiler.py:
2979         (SingleFileOutputProfiler.__init__):
2980         (Perf.attach_to_pid):
2981         (Perf.profile_after_exit):
2982
2983 2012-12-17  Sheriff Bot  <webkit.review.bot@gmail.com>
2984
2985         Unreviewed, rolling out r137198.
2986         http://trac.webkit.org/changeset/137198
2987         https://bugs.webkit.org/show_bug.cgi?id=105212
2988
2989         This patch is causing API behavior compatibility problems
2990         (Requested by zdobersek on #webkit).
2991
2992         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2993         (willSendRequestCallback):
2994
2995 2012-12-17  Elliott Sprehn  <esprehn@chromium.org>
2996
2997         Unreviewed. Adding myself to committers list.
2998
2999         * Scripts/webkitpy/common/config/committers.py:
3000
3001 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
3002
3003         nrwt: shuffle code around for cleanup in run_webkit_tests.py
3004         https://bugs.webkit.org/show_bug.cgi?id=105078
3005
3006         Reviewed by Ojan Vafai.
3007
3008         This patch reorders functions so that run_webkit_tests can be
3009         understood in a top-down matter better, and to make things slightly
3010         more sensible when I merge manager.py into it.
3011
3012         Also, this adds tests for the actual main() routine, which had several
3013         bugs get through testing a week or two ago.
3014
3015         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3016         (main):
3017         (parse_args):
3018         (_set_up_derived_options):
3019         (run):
3020         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3021         (RunTest):
3022         (RunTest.test_no_http_tests):
3023         (PortTest.disabled_test_mac_lion):
3024         (MainTest):
3025         (MainTest.test_exception_handling):
3026         (MainTest.test_exception_handling.interrupting_run):
3027         (MainTest.test_exception_handling.successful_run):
3028         (MainTest.test_exception_handling.successful_run.FakeRunDetails):
3029         (MainTest.test_exception_handling.exception_raising_run):
3030
3031 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
3032
3033         webkitpy: move --lint-test-files code into its own module
3034         https://bugs.webkit.org/show_bug.cgi?id=105077
3035
3036         Reviewed by Eric Seidel.
3037
3038         It really didn't need to be in run_webkit_tests.py anymore.
3039         This patch adds a new lint-test-expectations scripts to Tools/Scripts;
3040         rwt --lint-test-files is still supported for backwards compatibility
3041         for now.
3042
3043         * Scripts/lint-test-expectations: Added.
3044         * Scripts/webkitpy/layout_tests/lint_test_expectations.py: Added.
3045         (lint):
3046         (main):
3047         * Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py: Added.
3048         (FakePort):
3049         (FakePort.__init__):
3050         (FakePort.test_configuration):
3051         (FakePort.expectations_dict):
3052         (FakePort.skipped_layout_tests):
3053         (FakePort.all_test_configurations):
3054         (FakePort.configuration_specifier_macros):
3055         (FakePort.get_option):
3056         (FakeFactory):
3057         (FakeFactory.__init__):
3058         (FakeFactory.get):
3059         (FakeFactory.all_port_names):
3060         (LintTest):
3061         (LintTest.test_all_configurations):
3062         (LintTest.test_lint_test_files):
3063         (LintTest.test_lint_test_files__errors):
3064         (MainTest):
3065         (MainTest.test_success):
3066         (MainTest.test_success.interrupting_lint):
3067         (MainTest.test_success.successful_lint):
3068         (MainTest.test_success.exception_raising_lint):
3069         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3070         (main):
3071         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3072         (StreamTestingMixin.assertNotEmpty):
3073
3074 2012-12-17  Adam Klein  <adamk@chromium.org>
3075
3076         build-webkit: rename --template-tag to --template-element to match ENABLE #define name
3077         https://bugs.webkit.org/show_bug.cgi?id=105072
3078
3079         Reviewed by Laszlo Gombos.
3080
3081         * Scripts/webkitperl/FeatureList.pm:
3082
3083 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
3084
3085         Add some unit testing for WTF::clampTo* functions
3086         https://bugs.webkit.org/show_bug.cgi?id=105060
3087
3088         Reviewed by Emil A Eklund.
3089
3090         The new tests are targeted at clamping to int / unsigned and float as those are
3091         the one explicitely exposed through more dedicated functions (like clampToInteger,
3092         clampToPositiveInteger and clampToFloat).
3093
3094         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
3095
3096 2012-12-17  Eric Seidel  <eric@webkit.org>
3097
3098         rpt --profile --chromium-android throws exception
3099         https://bugs.webkit.org/show_bug.cgi?id=105061
3100
3101         Reviewed by Daniel Bates.
3102
3103         Somehow popen(args) got turned into popen(*args) at some point,
3104         which caused my new stringify code to fail when args was passed by name.
3105         Fixed and tested.
3106
3107         * Scripts/webkitpy/common/system/executive.py:
3108         (Executive._stringify_args):
3109         (Executive):
3110         (Executive.popen):
3111         * Scripts/webkitpy/common/system/executive_unittest.py:
3112         (ExecutiveTest.test_popen_args):
3113
3114 2012-12-17  Mark Pilgrim  <pilgrim@chromium.org>
3115
3116         Use Platform::current() instead of webKitPlatformSupport() in DumpRenderTree
3117         https://bugs.webkit.org/show_bug.cgi?id=105054
3118
3119         Reviewed by Jochen Eisinger.
3120
3121         Part of a refactoring series; see tracking bug 82948.
3122
3123         * DumpRenderTree/chromium/NotificationPresenter.cpp:
3124         (NotificationPresenter::show):
3125         * DumpRenderTree/chromium/Task.cpp:
3126         (postTask):
3127         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
3128
3129 2012-12-17  Emil A Eklund  <eae@chromium.org>
3130
3131         Clamp values in LayoutUnit::operator/ when SATURATED_LAYOUT_ARITHMETIC is enabled
3132         https://bugs.webkit.org/show_bug.cgi?id=104955
3133
3134         Reviewed by Julien Chaffraix.
3135
3136         Add tests for LayoutUnit.
3137
3138         * TestWebKitAPI/CMakeLists.txt:
3139         * TestWebKitAPI/ForwardingHeaders: Added.
3140         * TestWebKitAPI/ForwardingHeaders/WebCore: Added.
3141         * TestWebKitAPI/ForwardingHeaders/WebCore/LayoutUnit.h: Added.
3142         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
3143         * TestWebKitAPI/TestWebKitAPI.gypi:
3144         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3145         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp: Added.
3146         (TestWebKitAPI):
3147         (TestWebKitAPI::TEST):
3148         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
3149
3150 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
3151
3152         [chromium] removed unused preferences from testRunner.overridePreference
3153         https://bugs.webkit.org/show_bug.cgi?id=105169
3154
3155         Reviewed by Tony Chang.
3156
3157         Most notably this removes the code related to font family maps which is
3158         now handled via window.internals.settings
3159
3160         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
3161         (WebPreferences):
3162         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
3163         (WebTestRunner::TestRunner::overridePreference):
3164         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
3165         (TestRunner):
3166         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
3167         (WebTestRunner::WebPreferences::applyTo):
3168
3169 2012-12-17  James Robinson  <jamesr@chromium.org>
3170
3171         [chromium] Move creation of WebLayerTreeView to WebWidgetClient
3172         https://bugs.webkit.org/show_bug.cgi?id=105071
3173
3174         Reviewed by Adrienne Walker.
3175
3176         Update WebViewHost for the new WebWidgetClient interface.
3177
3178         * DumpRenderTree/chromium/WebViewHost.cpp:
3179         (WebViewHost::initializeLayerTreeView):
3180         (WebViewHost::layerTreeView):
3181         (WebViewHost::~WebViewHost):
3182         * DumpRenderTree/chromium/WebViewHost.h:
3183         (WebViewHost):
3184
3185 2012-12-17  Kenneth Rohde Christiansen  <kenneth@webkit.org>
3186
3187         [EFL][WK2] window_create doesn't receive the url
3188         https://bugs.webkit.org/show_bug.cgi?id=105184
3189
3190         Reviewed by Alexis Menard.
3191
3192         * MiniBrowser/efl/main.c: Use the url supplied to window_create
3193         (on_window_create):
3194
3195 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
3196
3197         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
3198         https://bugs.webkit.org/show_bug.cgi?id=105026
3199
3200         Reviewed by Yury Semikhatsky.
3201
3202         Old SFINAE test was replaced with new one based on this article:
3203         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
3204
3205         New test which covers this problem was added.
3206
3207         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3208
3209 2012-12-17  Anthony Scian  <ascian@rim.com>
3210
3211         [Blackberry] Static code analysis warning fixes
3212         https://bugs.webkit.org/show_bug.cgi?id=105129
3213
3214         Reviewed by Rob Buis.
3215
3216         Added null checks for fopen calls, no recovery attempted; access fault prevention only.
3217
3218         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
3219         (BlackBerry::WebKit::createFile):
3220         (BlackBerry::WebKit::DumpRenderTree::runTest):
3221
3222 2012-12-17  John Mellor  <johnme@chromium.org>
3223
3224         Add Anton Vayvod and Tim Volodine as contributors in committers.py
3225         https://bugs.webkit.org/show_bug.cgi?id=105025
3226
3227         Unreviewed. Adds two contributors to committers.py.
3228         They are initially contributing to the Text Autosizing feature (master bug: 84186).
3229
3230         * Scripts/webkitpy/common/config/committers.py:
3231
3232 2012-12-17  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3233
3234         Unreviewed, add myself to committers.py.
3235
3236         * Scripts/webkitpy/common/config/committers.py:
3237
3238 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3239
3240         [Qt] remove some unnecessary CONFIG additions
3241
3242         Reviewed by Simon Hausmann.
3243
3244         qt is already added by spec_pre.prf, warn_on and depend_includepath by
3245         default_pre.prf.
3246
3247         * qmake/mkspecs/features/default_post.prf:
3248
3249 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
3250
3251         Unreviewed, rolling out r137892.
3252         http://trac.webkit.org/changeset/137892
3253         https://bugs.webkit.org/show_bug.cgi?id=105026
3254
3255         it broke compilation on windows
3256
3257         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3258
3259 2012-12-14  Ilya Tikhonovsky  <loislo@chromium.org>
3260
3261         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
3262         https://bugs.webkit.org/show_bug.cgi?id=105026
3263
3264         Reviewed by Yury Semikhatsky.
3265
3266         Old SFINAE test was replaced with new one based on this article:
3267         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
3268
3269         New test which covers this problem was added.
3270
3271         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3272
3273 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3274
3275         [Qt] rely on automatic output directory setup
3276
3277         Reviewed by Simon Hausmann.
3278
3279         Now being a proper qt module (by virtue of having load(qt_build_config)
3280         in .qmake.conf), webkit gets the path setup goodies for free.
3281
3282         This also fixes Makefile.api.Debug/Release trying to generate qrc_WebKit.cpp
3283         at the same time, instead of the debug and release version ending up in different
3284         directories as they should.
3285
3286         * qmake/mkspecs/features/default_pre.prf:
3287
3288 2012-12-17  Charles Wei  <charles.wei@torchmobile.com.cn>
3289
3290         [BlackBerry] Need to re-enable video track
3291         https://bugs.webkit.org/show_bug.cgi?id=105155
3292
3293         Reviewed by George Staikos.
3294
3295         Just enable Video Track for BlackBerry.
3296
3297         * Scripts/webkitperl/FeatureList.pm:
3298
3299 2012-12-17  János Badics  <jbadics@inf.u-szeged.hu>
3300
3301         Unreviewed. Added myself to committers.py
3302
3303         * Scripts/webkitpy/common/config/committers.py:
3304
3305 2012-12-17  Jonathan Liu  <net147@gmail.com>
3306
3307         Fix linking of libraries with -Wl,-whole-archive
3308         https://bugs.webkit.org/show_bug.cgi?id=104436
3309
3310         Reviewed by Simon Hausmann.
3311
3312         If there is more than one library in LIBS linked with
3313         -Wl,-whole-archive -l... -Wl,-no-whole-archive, qmake tries to
3314         merge the duplicate flags in LIBS which causes linking to fail.
3315
3316         Add no_smart_library_merge to CONFIG to prevent the duplicate
3317         flags from being removed when using -Wl,-whole-archive.
3318
3319         * qmake/mkspecs/features/functions.prf:
3320
3321 2012-12-16  Arvid Nilsson  <anilsson@rim.com>
3322
3323         [BlackBerry] Adapt to new BlackBerry::Platform::TouchPoint API
3324         https://bugs.webkit.org/show_bug.cgi?id=105143
3325         RIM PR 171941
3326
3327         Reviewed by Rob Buis.
3328         Internally reviewed by George Staikos.
3329
3330         Adapt to new method names and encapsulation of TouchPoint data members.
3331
3332         * DumpRenderTree/blackberry/EventSender.cpp:
3333         (addTouchPointCallback):
3334         (updateTouchPointCallback):
3335         (touchEndCallback):
3336         (releaseTouchPointCallback):
3337         (sendTouchEvent):
3338
3339 2012-12-16  Maciej Stachowiak  <mjs@apple.com>
3340
3341         Add a script to compute the mean and 95% confidence interval (using two-sided t-test) of a sample set
3342         https://bugs.webkit.org/show_bug.cgi?id=105148
3343
3344         Reviewed by Filip Pizlo.
3345
3346         Usage examples:
3347         
3348         $ sampstat --help
3349         Usage: sampstat [options] [FILES]
3350           Compute the mean and 95% confidence interval of a sample set.
3351           Standard input or files must contain two or more decimal numbers, one per line.
3352         
3353         Options:
3354           -h, --help            show this help message and exit
3355           -u UNIT, --unit=UNIT  assume values are in units of UNIT
3356           -v, --verbose         print all values (with units)
3357         
3358         $ sampstat -u MB memresults.txt
3359         2356.90 MB +/- 101.34 MB (4.3%)
3360         
3361         $ sampstat -v -u MB memresults.txt
3362               2318.21 MB
3363               2399.56 MB
3364               2352.93 MB
3365         ----------------
3366         Mean: 2356.90 MB +/- 101.34 MB (4.3%)
3367         
3368
3369         * Scripts/sampstat: Added.
3370         (sum): Helper function to compute the sum of a list.
3371         (arithmeticMean): Compute the meam of a list.
3372         (standardDeviation): Compute the sample standard deviation (unbiased estimator).
3373         (standardError): Compute the sample standard error.
3374         (tDist): Compute t(.025, n-1), the t-value for atwo-sided 95% confidence interval. 
3375         (twoSidedConfidenceInterval): Compute the two-sided confidence interval range about the mean.
3376
3377 2012-12-15  Nima Ghanavatian  <nghanavatian@rim.com>
3378
3379         [BlackBerry] Pass key modifiers with touch and mouse events
3380         https://bugs.webkit.org/show_bug.cgi?id=105108
3381
3382         Reviewed by Rob Buis.
3383
3384         Update DRT to new api.
3385
3386         Internally reviewed by Gen Mak.
3387
3388         * DumpRenderTree/blackberry/EventSender.cpp:
3389         (mouseDownCallback):
3390         (mouseUpCallback):
3391         (mouseMoveToCallback):
3392
3393 2012-12-15  Zan Dobersek  <zandobersek@gmail.com>
3394
3395         Stop the XvfbDriver before proceeding with its starting setup
3396         https://bugs.webkit.org/show_bug.cgi?id=105006
3397
3398         Reviewed by Philippe Normand.
3399
3400         Just as with the base Driver implementation, stop the XvfbDriver
3401         before proceeding with setting up lock files and a new Xvfb instance.
3402         This forces any driver or Xvfb instances that are currently under
3403         the driver's control to be closed/killed and cleaned up after.
3404
3405         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
3406         (XvfbDriver._start):
3407
3408 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
3409
3410         webkitpy: make port_name a required parameter to the non-test ports
3411         https://bugs.webkit.org/show_bug.cgi?id=105057
3412
3413         Reviewed by Eric Seidel.
3414
3415         This is just minor code cleanup along the path to making Port
3416         constructors simpler and pushing more logic into the port factory.
3417
3418         * Scripts/webkitpy/layout_tests/port/base.py:
3419         (Port.__init__):
3420         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3421         (PortTest.make_port):
3422         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3423         (TestWebKitPort.__init__):
3424         * Scripts/webkitpy/layout_tests/port/test.py:
3425         (TestPort):
3426         (TestPort.determine_full_port_name):
3427         (TestPort.__init__):
3428         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py:
3429         (XvfbDriverTest.make_driver):
3430         (XvfbDriverTest.test_stop):
3431
3432 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
3433
3434         garden-o-matic doesn't know about reftests
3435         https://bugs.webkit.org/show_bug.cgi?id=101976
3436
3437         Reviewed by Eric Seidel.
3438
3439         Attempt to make garden-o-matic properly aware of reftests again.
3440
3441         This is re-landing r137407 with a change to
3442         rebaselineWithStatusUpdates() to actually pass the filtered
3443         failureInofLost to checkout.rebaseline().
3444
3445         Unfortunately, the testing coverage for this whole module is almost
3446         non-existent, and it's not obvious that there's a good way to
3447         stub out checkout.rebaseline() to test that the right thing
3448         happens. I verified the change manually.
3449
3450         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
3451         (.):
3452         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
3453         (.):
3454         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
3455
3456 2012-12-14  Ryosuke Niwa  <rniwa@webkit.org>
3457
3458         "Running 1 DumpRenderTree over X shards" is not a helpful output
3459         https://bugs.webkit.org/show_bug.cgi?id=104858
3460
3461         Reviewed by Dirk Pranke.
3462
3463         The word "shard" is not typically used in test frameworks. This word is confusing for many WebKit
3464         contributors who typically don't have much experience on distributed computations and databases.
3465         So only print this message out in the debug mode.
3466
3467         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3468         * Scripts/webkitpy/layout_tests/views/printing.py:
3469         (Printer.print_workers_and_shards):
3470
3471 2012-12-13  Eric Seidel  <eric@webkit.org>
3472
3473         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
3474         https://bugs.webkit.org/show_bug.cgi?id=104891
3475
3476         Reviewed by Ryosuke Niwa.
3477
3478         I also implemented a very simple "Sample" Profiler using
3479         Mac OS X's /usr/bin/sample command line tool.
3480
3481         The real reason for this abstraction is to make it easy
3482         to support both perf and pprof on linux which seem to
3483         be about equally popular among those I ask in the Chrome team.
3484
3485         * Scripts/webkitpy/common/system/profiler.py:
3486         (ProfilerFactory.create_profiler):
3487         (ProfilerFactory):
3488         (ProfilerFactory.available_profilers_by_name):
3489         (ProfilerFactory.default_profiler_name):
3490         (Sample):
3491         (Sample.__init__):
3492         (Sample.attach_to_pid):
3493         (Sample.profile_after_exit):
3494         (IProfiler.attach_to_pid):
3495         * Scripts/webkitpy/layout_tests/port/driver.py:
3496         (Driver.__init__):
3497         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3498         (PerfTestsRunner._parse_args):
3499
3500 2012-12-13  Eric Seidel  <eric@webkit.org>
3501
3502         Unreviewed, fixing typo in python unittest.
3503
3504         I changed the default profiler on linux from pprof to perf in
3505         bug 104971.  I failed to update the unittest results at that time.
3506
3507         * Scripts/webkitpy/common/system/profiler_unittest.py:
3508         (ProfilerFactoryTest.test_basic):
3509
3510 2012-12-13  Ian Vollick  <vollick@chromium.org>
3511
3512         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
3513         https://bugs.webkit.org/show_bug.cgi?id=104911
3514
3515         Reviewed by Dirk Pranke.
3516
3517         Adds two virtual tests suites to run the tests in
3518         compositing/overflow/ and scrollbars/ with the flag
3519         --enable-accelerated-overflow-scroll, which I've also plumbed through
3520         in the usual way.
3521
3522         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3523         (main):
3524         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
3525         (WebPreferences):
3526         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
3527         (WebTestRunner::WebPreferences::reset):
3528         (WebTestRunner::WebPreferences::applyTo):
3529         * DumpRenderTree/chromium/TestShell.cpp:
3530         (TestShell::TestShell):
3531         (TestShell::resetWebSettings):
3532         * DumpRenderTree/chromium/TestShell.h:
3533         (TestShell::setAcceleratedCompositingForOverflowScrollEnabled):
3534         (TestShell):
3535         * Scripts/webkitpy/layout_tests/port/chromium.py:
3536         (ChromiumPort.virtual_test_suites):
3537
3538 2012-12-13  Eric Seidel  <eric@webkit.org>
3539
3540         Use 'perf' to profile on linux instead of google-pprof by default
3541         https://bugs.webkit.org/show_bug.cgi?id=104971
3542
3543         Reviewed by Dirk Pranke.
3544
3545         This makes Chromium Linux match Chromium Android and use the perf
3546         tool by default.  Once https://bugs.webkit.org/show_bug.cgi?id=104891
3547         lands it will be possible to use pprof again on Linux.
3548
3549         This is slightly more advanced than the AndroidPerf profiler as
3550         instead of using a timeout on "perf record" I instead watch
3551         for the termination of the target process and then control-C
3552         the 'perf record' process.  This required me to add two new
3553         methods to Executive, one to have a limited-time wait() and
3554         the second to be able to send a control-C.  I chose to add
3555         these to Executive to make them easier to mock/fix-for-win32
3556         at a later time if needed.
3557
3558         * Scripts/webkitpy/common/system/executive.py:
3559         (Executive.wait_limited):
3560         (Executive.interrupt):
3561         * Scripts/webkitpy/common/system/profiler.py:
3562         (ProfilerFactory.create_profiler):
3563         (Perf):
3564         (Perf.__init__):
3565         (Perf._perf_path):
3566         (Perf.attach_to_pid):
3567         (Perf._first_ten_lines_of_profile):
3568         (Perf.profile_after_exit):
3569
3570 2012-12-13  Julie Parent  <jparent@chromium.org>
3571
3572         Dashboard cleanup: remove usage of global g_builders.
3573         https://bugs.webkit.org/show_bug.cgi?id=104941
3574
3575         Reviewed by Dirk Pranke.
3576
3577         The dashboards use a lot of global state, which makes hacking on them
3578         complicated. This change removes the use of one such global: g_builders.
3579         In most cases, we can just use currentBuilderGroup().builders instead,
3580         which is now currentBuilders().
3581         Surprisingly, the most changes were required to the unit tests, since
3582         they were even bigger offenders of bad hygiene, relying on global state
3583         set by other tests, randomly clobbering global variables in ways the
3584         real code doesn't, etc.
3585
3586         * TestResultServer/static-dashboards/builders.js:
3587         (BuilderGroup.prototype.setup):
3588         * TestResultServer/static-dashboards/dashboard_base.js:
3589         (.switch.return):
3590         (htmlForTestTypeSwitcher):
3591         * TestResultServer/static-dashboards/flakiness_dashboard.js:
3592         (generatePage):
3593         (getAllTestsTrie):
3594         (processTestRunsForAllBuilders):
3595         (showPopupForBuild):
3596         (generatePageForExpectationsUpdate):
3597         (loadExpectationsLayoutTests):
3598         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
3599         (resetGlobals):
3600         (stubResultsByBuilder):
3601         (test):
3602         * TestResultServer/static-dashboards/loader.js:
3603         (.):
3604         * TestResultServer/static-dashboards/loader_unittests.js:
3605
3606 2012-12-13  Eric Seidel  <eric@webkit.org>
3607
3608         Callers should not have to stringify args before calling Executive run_command/popen
3609         https://bugs.webkit.org/show_bug.cgi?id=104975
3610
3611         Reviewed by Dirk Pranke.
3612
3613         One could argue that we should match the python call syntax here,
3614         but I think it's a more friendly API if we automagically handle
3615         stringification of args in run_command, etc.
3616         This removes map(unicode, args) from several callsites.
3617
3618         When I first tried to land this change, I didn't realize that
3619         Executive._command_for_printing depended on this behavior
3620         having been applied to args in run_command.  The fix is to
3621         call _stringify_args in both run_command and popen.
3622         This is slightly redundant, but given how short args have to be
3623         (due to shell limits), I don't think the double-encode check
3624         matters in practice.
3625
3626         This is slightly complicated by the fact that apache_http_server.py
3627         is the one caller in our codebase which uses shell=True.
3628         shell=True is a well-documented trail-of-tears:
3629         http://stackoverflow.com/questions/3172470/actual-meaning-of-shell-true-in-subprocess
3630         but to support this legacy (windows-only) code (which I can't easily test)
3631         I've added an if-hack to avoid stringifying the the popen(shell=True) case.
3632
3633         * Scripts/webkitpy/common/system/executive.py:
3634         (Executive.run_command):
3635         (Executive._stringify_args):
3636         (Executive.popen):
3637         * Scripts/webkitpy/common/system/executive_unittest.py:
3638         (ExecutiveTest.test_auto_stringify_args):
3639         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3640         (attach_to_pid):
3641
3642 2012-12-13  Adrienne Walker  <enne@chromium.org>
3643
3644         Unreviewed, rolling out r137645, r137646, and r137667.
3645         http://trac.webkit.org/changeset/137645
3646         http://trac.webkit.org/changeset/137646
3647         http://trac.webkit.org/changeset/137667
3648         https://bugs.webkit.org/show_bug.cgi?id=104911
3649
3650         Breaks some overflow layout tests
3651
3652         * Scripts/webkitpy/layout_tests/port/chromium.py:
3653         (ChromiumPort.virtual_test_suites):
3654
3655 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3656
3657         Unreviewed, rolling out r137661.
3658         http://trac.webkit.org/changeset/137661
3659         https://bugs.webkit.org/show_bug.cgi?id=104891
3660
3661         broke unit tests, run-webkit-tests
3662
3663         * Scripts/webkitpy/common/system/executive.py:
3664         (Executive._run_command_with_teed_output):
3665         (Executive.run_command):
3666         (Executive.popen):
3667         * Scripts/webkitpy/common/system/profiler.py:
3668         (ProfilerFactory.create_profiler):
3669         (GooglePProf.profile_after_exit):
3670         (IProfiler.attach_to_pid):
3671         * Scripts/webkitpy/layout_tests/port/driver.py:
3672         (Driver.__init__):
3673         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3674         (PerfTestsRunner._parse_args):
3675
3676 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3677
3678         nrwt: move rundetails, summarize_results into test_run_results
3679         https://bugs.webkit.org/show_bug.cgi?id=104963
3680
3681         Reviewed by Eric Seidel.
3682
3683         More cleanup; this the plain-old-data functions out of manager.py
3684         and into test_run_results next to the other aggregate data structures.
3685
3686         Also this cleans up a bunch of test code and reduces duplication.
3687
3688         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3689         (Manager.run):
3690         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3691         (ManagerTest.test_look_for_new_crash_logs):
3692         * Scripts/webkitpy/layout_tests/models/test_run_results.py:
3693         (TestRunResults.__init__):
3694         (TestRunResults.add):
3695         (RunDetails):
3696         (RunDetails.__init__):
3697         (_interpret_test_failures):
3698         (summarize_results):
3699         * Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py: Added.
3700         (get_result):
3701         (run_results):
3702         (summarized_results):
3703         (InterpretTestFailuresTest):
3704         (InterpretTestFailuresTest.setUp):
3705         (InterpretTestFailuresTest.test_interpret_test_failures):
3706         (SummarizedResultsTest):
3707         (SummarizedResultsTest.setUp):
3708         (SummarizedResultsTest.test_no_svn_revision):
3709         (SummarizedResultsTest.test_svn_revision):
3710         (SummarizedResultsTest.test_summarized_results_wontfix):
3711         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3712         (BuildBotPrinterTests.get_printer):
3713         (BuildBotPrinterTests.test_print_unexpected_results):
3714         (BuildBotPrinterTests.test_print_results):
3715
3716 2012-12-13  Eric Seidel  <eric@webkit.org>
3717
3718         Make --profile output easier to understand for pprof on linux
3719         https://bugs.webkit.org/show_bug.cgi?id=104957
3720
3721         Reviewed by Dirk Pranke.
3722
3723         pprof is kinda mysterious, so we print some information
3724         on how to use it, and how to learn how to read the sample dump.
3725         This is similar to what we've done for perf output for Chromium Android.
3726
3727         * Scripts/webkitpy/common/system/profiler.py:
3728         (GooglePProf._pprof_path):
3729         (GooglePProf.profile_after_exit):
3730
3731 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3732
3733         Revert r137407 - it didn't actually work right.
3734         https://bugs.webkit.org/show_bug.cgi?id=101976
3735
3736         Unreviewed, build fix.
3737
3738         Turns out that rebaselining was completely broken with this change :(
3739         We're passing in a list of test names rather than the list of
3740         failureInfo objects.
3741
3742         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
3743         (.):
3744         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
3745         (.):
3746         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
3747
3748 2012-12-13  Antoine Labour  <piman@chromium.org>
3749
3750         [chromium] Small fixes for WebPluginContainer::setWebLayer
3751         https://bugs.webkit.org/show_bug.cgi?id=104953
3752
3753         Reviewed by James Robinson.
3754
3755         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Clean up
3756         dangling pointers, force correct invalidation of the layer.
3757
3758 2012-12-13  Eric Seidel  <eric@webkit.org>
3759
3760         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
3761         https://bugs.webkit.org/show_bug.cgi?id=104891
3762
3763         Reviewed by Dirk Pranke.
3764
3765         I also implemented a very simple "Sample" Profiler using
3766         Mac OS X's /usr/bin/sample command line tool.
3767
3768         I also moved the map(unicode, args) call into Executive.popen
3769         so that all callers don't have to do it themselves.
3770
3771         The real reason for this abstraction is to make it easy
3772         to support both perf and pprof on linux which seem to
3773         be about equally popular among those I ask in the Chrome team.
3774
3775         * Scripts/webkitpy/common/system/executive.py:
3776         (Executive.popen):
3777         * Scripts/webkitpy/common/system/profiler.py:
3778         (ProfilerFactory.create_profiler):
3779         (ProfilerFactory):
3780         (ProfilerFactory.available_profilers_by_name):
3781         (ProfilerFactory.default_profiler_name):
3782         (Sample):
3783         (Sample.__init__):
3784         (Sample.attach_to_pid):
3785         (Sample.profile_after_exit):
3786         (IProfiler.attach_to_pid):
3787         * Scripts/webkitpy/layout_tests/port/driver.py:
3788         (Driver.__init__):
3789         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3790         (PerfTestsRunner._parse_args):
3791
3792 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3793
3794         support -wk2 port names properly in webkitpy.layout_tests.port
3795         https://bugs.webkit.org/show_bug.cgi?id=104761
3796
3797         Reviewed by Eric Seidel.
3798
3799         Our support for WK2 port names was inconsistent at best; this patch
3800         modifies the code so that <port_name>-wk2 is supported for all ports that
3801         have WK2 implementations and is equivalent to specifying <port_name>
3802         and -2/--webkit-test-runner. In addition, this modifies
3803         builders.all_port_names() to include the wk2 variants.
3804
3805         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
3806         * Scripts/webkitpy/layout_tests/port/apple.py:
3807         (ApplePort.determine_full_port_name):
3808         * Scripts/webkitpy/layout_tests/port/base.py:
3809         (Port.determine_full_port_name):
3810         (Port.__init__):
3811         * Scripts/webkitpy/layout_tests/port/builders.py:
3812         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
3813         (DriverTest.make_port):
3814         * Scripts/webkitpy/layout_tests/port/mac.py:
3815         (MacPort.default_baseline_search_path):
3816         * Scripts/webkitpy/layout_tests/port/win.py:
3817         (WinPort.default_baseline_search_path):
3818         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
3819         (TestRebaselineExpectations.test_rebaseline_expectations):
3820
3821 2012-12-13  Ian Vollick  <vollick@chromium.org>
3822
3823         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
3824         https://bugs.webkit.org/show_bug.cgi?id=104911
3825
3826         Reviewed by Adrienne Walker.
3827
3828         Adds two virtual tests suites to run the tests in
3829         compositing/overflow/ and scrollbars/ with the flag
3830         --enable-accelerated-compositing-for-overflow-scroll
3831
3832         * Scripts/webkitpy/layout_tests/port/chromium.py:
3833         (ChromiumPort.virtual_test_suites):
3834
3835 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3836
3837         delete extraneous self.fail in mastercfg_unittest.py
3838         https://bugs.webkit.org/show_bug.cgi?id=104940
3839
3840         Reviewed by Eric Seidel.
3841
3842         This was accidentally introduced in r137404.
3843
3844         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
3845
3846 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3847
3848         Fix a mis-named variable after r137621.
3849         https://bugs.webkit.org/show_bug.cgi?id=104871
3850
3851         Unreviewed, build fix.
3852
3853         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
3854         (BuildBotPrinter.print_results):
3855         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3856
3857 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3858
3859         webkitpy: rename ResultSummary
3860         https://bugs.webkit.org/show_bug.cgi?id=104871
3861
3862         Reviewed by Eric Seidel.
3863
3864         This patch changes the names of ResultSummary and related modules,
3865         classes, and variables as a part of the larger clean up in bug 103824.
3866
3867         This change is basically a bunch of search and replaces:
3868         - result_summary.py -> test_run_results.py
3869         - class ResultSummary -> class TestRunResults
3870         - "result_summary" or "rs" as a generic variable name -> "run_results",
3871           when it doesn't matter (or we don't know) if this refers to the
3872           initial pass over the tests or the retry pass.
3873         - "result_summary" as a variable containing the results of the initial,
3874           complete test run -> "initial_results"
3875         - "retry_summary" as a variable containing the results of retried
3876           results -> "retry_results"
3877         - ResultSummary.results (the property) -> TestRunResults.results_by_name
3878         - ResultSummary.failures (the property) -> TestRunResults.failures_by_name
3879         - RunDetails.result_summary -> RunDetails.initial_results
3880         - RunDetails.retry_summary -> RunDetails.retry_results
3881
3882         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
3883         (LayoutTestRunner.__init__):
3884         (LayoutTestRunner.run_tests):
3885         (LayoutTestRunner._mark_interrupted_tests_as_skipped):
3886         (LayoutTestRunner._interrupt_if_at_failure_limits):
3887         (LayoutTestRunner._interrupt_if_at_failure_limits.interrupt_if_at_failure_limit):
3888         (LayoutTestRunner._update_summary_with_result):
3889         (LayoutTestRunner._handle_finished_test):
3890         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
3891         (FakePrinter.print_expected):
3892         (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
3893         (LayoutTestRunnerTests.test_update_summary_with_result):
3894         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3895         (RunDetails.__init__):
3896         (summarize_results):
3897         (Manager.run):
3898         (Manager._look_for_new_crash_logs):
3899         (Manager._tests_to_retry):
3900         (Manager._upload_json_files):
3901         (Manager._stats_trie):
3902         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3903         (ManagerTest.test_look_for_new_crash_logs):
3904         (SummarizeResultsTest):
3905         (SummarizeResultsTest.get_run_results):
3906         (SummarizeResultsTest.summarized_results):
3907         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
3908         (JSONLayoutResultsGenerator.__init__):
3909         (JSONLayoutResultsGenerator._get_modifier_char):
3910         (JSONLayoutResultsGenerator._insert_failure_summaries):
3911         (JSONLayoutResultsGenerator._get_failure_summary_entry):
3912         * Scripts/webkitpy/layout_tests/models/test_run_results.py: Renamed from Tools/Scripts/webkitpy/layout_tests/models/result_summary.py.
3913         (TestRunResults):
3914         (TestRunResults.__init__):
3915         (TestRunResults.add):
3916         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3917         (get_test_results):
3918         (MainTest.test_basic):
3919         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
3920         (BuildBotPrinter.print_results):
3921         (BuildBotPrinter.print_run_results):
3922         (BuildBotPrinter._print_run_results_entry):
3923         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3924         (BuildBotPrinterTests.get_run_results):
3925         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
3926         (BuildBotPrinterTests):
3927         * Scripts/webkitpy/layout_tests/views/printing.py:
3928         (Printer.print_expected):
3929         (Printer._print_expected_results_of_type):
3930         (Printer.print_results):
3931         (Printer._print_timing_statistics):
3932         (Printer._print_worker_statistics):
3933         (Printer._print_aggregate_test_statistics):
3934         (Printer._print_individual_test_times):
3935         (Printer._print_directory_timings):
3936         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
3937
3938 2012-12-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3939
3940         [EFL] The WebKit2 bots are building and executing WebKit1 tests
3941         https://bugs.webkit.org/show_bug.cgi?id=104690
3942
3943         Reviewed by Csaba Osztrogonác.
3944
3945         Pass --no-webkit1 option to the EFL WK2 bots when building. This
3946         will disable WK1 API and thus, the WK1 API tests.
3947
3948         * BuildSlaveSupport/build.webkit.org-config/config.json:
3949         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3950         (CompileWebKit2Only):
3951         (BuildAndTestFactory):
3952         (BuildAndTestFactory.__init__):
3953         (BuildAndTestWebKit2OnlyFactory):
3954         Now bots using BuildAndTestWebKit2OnlyFactory (for instance only EFL) will
3955         get --no-webkit1 as a parameter on the build-webkit script.
3956         * Scripts/build-webkit:
3957         Added support for -no-webkit1 for the Qt and EFL port.
3958
3959 2012-12-13  Eric Seidel  <eric@webkit.org>
3960
3961         build-webkit --chromium-android forces Make even when update-webkit used --ninja
3962         https://bugs.webkit.org/show_bug.cgi?id=104839
3963
3964         Reviewed by Tony Chang.
3965
3966         This was recommended by Peter as part of:
3967         https://bugs.webkit.org/show_bug.cgi?id=104434#c4
3968         Fixing this makes build-webkit --chromium-android work for the common (ninja) case.
3969
3970         * Scripts/webkitdirs.pm:
3971         (buildChromium):
3972
3973 2012-12-13  Julie Parent  <jparent@chromium.org>
3974
3975         Loader_unittests should reset state between runs.
3976         https://bugs.webkit.org/show_bug.cgi?id=104862
3977
3978         Reviewed by Dirk Pranke.
3979
3980         The loader_unittests were not resetting state between runs, leaving
3981         a lot of data behind in global objects that the dashboard and tests
3982         rely on.  A handful were resetting just a few objects.  Use the common
3983         resetGlobals() method instead.
3984
3985         * TestResultServer/static-dashboards/loader_unittests.js:
3986
3987 2012-12-13  Andras Becsi  <andras.becsi@digia.com>
3988
3989         [Qt][WK2] Fix painting on Mac with retina display
3990         https://bugs.webkit.org/show_bug.cgi?id=104574
3991
3992         Reviewed by Kenneth Rohde Christiansen.
3993
3994         Remove setting the devicePixelRatio experimental property
3995         since the value is now automatically picked up from Qt.
3996
3997         * MiniBrowser/qt/qml/BrowserWindow.qml:
3998
3999 2012-12-13  Jussi Kukkonen  <jussi.kukkonen@intel.com>
4000
4001         [EFL][GTK] Don't call deprecated g_type_init when glib => 2.35
4002         https://bugs.webkit.org/show_bug.cgi?id=103209
4003
4004         Reviewed by Kenneth Rohde Christiansen.
4005
4006         g_type_init() is deprecated from 2.35.0 onwards. Don't call it
4007         in that case to avoid warnings and/or build failure.
4008
4009         * DumpRenderTree/gtk/ImageDiff.cpp:
4010         (main):
4011
4012 2012-12-13  Claudio Saavedra  <csaavedra@igalia.com>
4013
4014         [Tools] Add exception for gtk_style_context_get()
4015         https://bugs.webkit.org/show_bug.cgi?id=104896
4016
4017         Reviewed by Martin Robinson.
4018
4019         Add new exception for gtk_style_context_get().
4020
4021         * Scripts/webkitpy/style/checkers/cpp.py:
4022         (check_for_null): Add exception for gtk_style_context_get().
4023         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
4024         (WebKitStyleTest.test_null_false_zero): Add a unit
4025         test for the new exception.
4026
4027 2012-12-13  Joaquim Rocha  <jrocha@igalia.com>
4028
4029         REGRESSION (r137432): The /webkit2/WebKitCookieManager/accept-policy unit test is failing
4030         https://bugs.webkit.org/show_bug.cgi?id=104790
4031
4032         Reviewed by Carlos Garcia Campos.
4033
4034         Unskip the GTK accept policy test.
4035
4036         * Scripts/run-gtk-tests:
4037         (TestRunner):
4038
4039 2012-12-13  Xingnan Wang  <xingnan.wang@intel.com>
4040
4041         Unreviewed. Adding myself to committers list.
4042
4043         * Scripts/webkitpy/common/config/committers.py:
4044
4045 2012-12-12  Roger Fong  <roger_fong@apple.com>
4046
4047         Unreviewed. Fix typo in NRWT.
4048
4049         * Scripts/webkitpy/layout_tests/port/base.py:
4050         (Port.is_reference_html_file):
4051
4052 2012-12-12  Roger Fong  <roger_fong@apple.com>
4053
4054         Fix detection of ref test files in old and new run webkit tests.
4055         https://bugs.webkit.org/show_bug.cgi?id=104850.
4056
4057         Reviewed by Timothy Horton.
4058
4059         * Scripts/old-run-webkit-tests:
4060         (isUsedInReftest): 
4061         Add notref and ref prefix and suffix cases to isUsedinReftest.
4062         * Scripts/webkitpy/layout_tests/port/base.py:
4063         (Port):
4064         (Port.is_reference_html_file): 
4065         Typo in NRWT.
4066
4067 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
4068
4069         rework ChromiumAndroidDriver.start() et. al to avoid funky layering
4070         https://bugs.webkit.org/show_bug.cgi?id=104856
4071
4072         Reviewed by Eric Seidel.
4073
4074         r137523 / bug 10357 introduced some back-and-forth hopping during
4075         driver.start() in the chromium android port. This patch cleans it
4076         up a bit.
4077
4078         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
4079         (ChromiumAndroidDriver.__init__):
4080         (ChromiumAndroidDriver._start_once):
4081         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
4082         * Scripts/webkitpy/layout_tests/port/driver.py:
4083         (Driver.start):
4084         (Driver._start):
4085
4086 2012-12-12  Eric Seidel  <eric@webkit.org>
4087
4088         Add support for run-perf-tests --chromium-android --profile
4089         https://bugs.webkit.org/show_bug.cgi?id=103587
4090
4091         Reviewed by Dirk Pranke.
4092
4093         This makes it possible to profile PerformanceTests on an Android device
4094         using run-perf-tests --chromium-android --profile.
4095         CPU profiling on Android is sadly non-trivial, but this goes a long
4096         ways towards making it more accessible to non-Android developers.
4097
4098         Although this script requires your device to be flashed with a
4099         userdebug build (an annoyingly non-trivial endeavor for external contributors)
4100         it endeavors to detect and explain the common errors encountered
4101         while trying to setup a device for perf-based profiling.
4102
4103         * Scripts/webkitpy/common/system/profiler.py:
4104         (GooglePProf.profile_after_exit): Discovered that pprof will just hang if passed a missing file.
4105         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
4106         (AndroidPerf):
4107         (AndroidPerf.__init__):
4108         (AndroidPerf.check_configuration):
4109         (AndroidPerf.print_setup_instructions):
4110         (_file_exists_on_device):
4111         (_run_adb_command):
4112         (attach_to_pid):
4113         (_perf_version_string):
4114         (_find_perfhost_binary):
4115         (_perfhost_path):
4116         (_first_ten_lines_of_profile):
4117         (profile_after_exit):
4118         (ChromiumAndroidDriver.__init__):
4119         (ChromiumAndroidDriver.__init__.methods):
4120         (ChromiumAndroidDriver._find_or_create_symfs):
4121         (ChromiumAndroidDriver._start_once):
4122         (ChromiumAndroidDriver):
4123         (ChromiumAndroidDriver._run_post_start_tasks):
4124         (ChromiumAndroidDriver._pid_from_android_ps_output):
4125         (ChromiumAndroidDriver._pid_on_target):
4126         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
4127         (ChromiumAndroidDriverTest.test_pid_from_android_ps_output):
4128         (AndroidPerfTest):
4129         (AndroidPerfTest.test_perf_output_regexp):
4130         * Scripts/webkitpy/layout_tests/port/driver.py:
4131         (Driver._setup_environ_for_driver):
4132         (Driver):
4133         (Driver._start):
4134         (Driver._run_post_start_tasks):
4135         (Driver._pid_on_target):
4136
4137 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
4138
4139         nrwt: remove deprecated ORWT compatibility command line flags
4140         https://bugs.webkit.org/show_bug.cgi?id=104301
4141
4142         Reviewed by Eric Seidel.
4143
4144         The command line flags for --noshow-results, --[no-]launch-safari,
4145         and --use-remote-links-to-tests were for compatibility with
4146         old-run-webkit-tests, but the compatibility is not needed any more
4147         since NRWT is basically default and the scripts that passed the
4148         old flags have been updated.
4149
4150         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4151         (_set_up_derived_options):
4152         (parse_args):
4153
4154 2012-12-12  John Griggs  <jgriggs@rim.com>
4155
4156         2012-12-12  John Griggs  <jgriggs@rim.com>
4157
4158         [BlackBerry] Update Media Controls
4159         https://bugs.webkit.org/show_bug.cgi?id=104443
4160         https://przilla.ott.qnx.com/bugzilla/show_bug.cgi?id=204748
4161
4162         Update Media Controls for BlackBerry platform to allow audio, embedded video and fullscreen video controls to be positioned and styled differently.
4163
4164         Reviewed by Rob Buis.
4165
4166         BlackBerry-specific update to Media Controls - No new tests required.
4167
4168         * Scripts/webkitperl/FeatureList.pm:
4169
4170 2012-12-12  Justin Novosad  <junov@google.com>
4171
4172         Adding myself to SkiaGraphics watchlist
4173
4174         Unreviewed
4175
4176         * Scripts/webkitpy/common/config/watchlist:
4177
4178 2012-12-12  Martin Robinson  <mrobinson@igalia.com>
4179
4180         [GTK] Add authentication support to DRT and fix exposed issues in the libsoup backend
4181         https://bugs.webkit.org/show_bug.cgi?id=104285
4182
4183         Reviewed by Gustavo Noronha Silva.
4184
4185         Add support to DumpRenderTree for running authentication tests. Since the DRT
4186         expects an authentication callback, we add one to DRTSupport to avoid #ifdefs
4187         in platform-independent code for GTK+.
4188
4189         * DumpRenderTree/gtk/DumpRenderTree.cpp:
4190         (resetDefaultsToConsistentValues): Reset the authentication password and username.
4191         (authenticationCallback): Added.
4192         (createWebView): Attach the authentiation callback during startup.
4193         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
4194         (soupURIToKURL): soup_uri_to_string does not preserve passwords embedded
4195         in the URL so we add a somewhat messy method of re-adding them when they exist.
4196         It would be nice to use soupURIToKURL here, but it seems we cannot use KURL without
4197         pulling in lots of WebCore code so we use string search and replace.
4198         (TestRunner::queueLoad): Use the new helper.
4199
4200 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
4201
4202         garden-o-matic should prefer efl/ over efl-wk1/ and efl-wk2/ when rebaselining
4203         https://bugs.webkit.org/show_bug.cgi?id=102389
4204
4205         Reviewed by Tony Chang.
4206
4207         The code wasn't aware of both the efl-wk1 and efl-wk2 variants,
4208         because all_port_names() by itself doesn't tell you about the
4209         variants. Work around this for now, and filed bug 104761 to
4210         address the underlying problem.
4211
4212         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
4213         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
4214         (BaselineOptimizerTest.test_efl):
4215
4216 2012-12-12  Krzysztof Czech  <k.czech@samsung.com>
4217
4218         [EFL] Possibility to turn off accessibility feature for WebKit-EFL.
4219         https://bugs.webkit.org/show_bug.cgi?id=103036
4220
4221         Reviewed by Laszlo Gombos.
4222
4223         Guard dependencies for accessibility (ATK), so that they can be turned off.
4224
4225         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
4226         (AccessibilityUIElement):
4227         * WebKitTestRunner/PlatformEfl.cmake:
4228
4229 2012-12-12  Simon Fraser  <simon.fraser@apple.com>
4230
4231         Add some entries in the watchlist, and register me for them.
4232
4233         * Scripts/webkitpy/common/config/watchlist:
4234
4235 2012-12-12  Ilya Tikhonovsky  <loislo@chromium.org>
4236
4237         Web Inspector: Native Memory Instrumentation: remove fake root MemoryObjectInfo.
4238         https://bugs.webkit.org/show_bug.cgi?id=104796
4239
4240         Reviewed by Yury Semikhatsky.
4241
4242         It was not a good idea to introduce a fake root MemoryObjectInfo.
4243         It makes a problem when we visit an object without its own MemoryObjectType.
4244
4245         Example: RenderBox has a global pointer to a hash map.
4246         HashMap doesn't have its own object type because it is a generic container.
4247         It will inherit object type from the fake root memory object info.
4248         The same could happen for another container in another class with other MemoryObjectType.
4249
4250         This fact forces me to create custom process method for root objects
4251         because they need to have their own MemoryObjectInfo with customisable memory object type.
4252
4253         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
4254
4255         Drive by fix: InstrumentedPointer* was replaced with Wrapper* because actually it is using
4256         for instrumented and not instrumented object classes.
4257
4258 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
4259
4260         [GTK] Remove the last of unnecessary configuration options in configure.ac
4261         https://bugs.webkit.org/show_bug.cgi?id=104793
4262
4263         Reviewed by Martin Robinson.
4264
4265         Removing the filters and indexed database features from the list of features
4266         being configurable through the configure script. This is not the case anymore
4267         as the corresponding configuration options are being removed.
4268
4269         * Scripts/webkitdirs.pm:
4270         (buildAutotoolsProject):
4271
4272 2012-12-12  Mario Sanchez Prada  <mario@webkit.org>
4273
4274         [GTK][WK2] New API to detect display/execution of insecure content
4275         https://bugs.webkit.org/show_bug.cgi?id=104578
4276
4277         Reviewed by Martin Robinson.
4278
4279         Add coding style exception for 'readability/enum_casing' in GTK.
4280
4281         The GTK+ API use upper case, underscore separated, words in
4282         certain types of enums (e.g. signals, properties).
4283
4284         * Scripts/webkitpy/style/checker.py: Added exception, both for
4285         WebKitGTK+ and WebKit2GTK+.
4286
4287 2012-12-12  Jochen Eisinger  <jochen@chromium.org>