c9bbf93eb663194b301cab05acbc72e42b59a829
[WebKit-https.git] / Tools / ChangeLog
1 2013-01-12  Jochen Eisinger  <jochen@chromium.org>
2
3         [chromium] move inspector related methods to TestRunner library
4         https://bugs.webkit.org/show_bug.cgi?id=106654
5
6         Reviewed by Adam Barth.
7
8         * DumpRenderTree/chromium/DRTTestRunner.cpp:
9         (DRTTestRunner::DRTTestRunner):
10         * DumpRenderTree/chromium/DRTTestRunner.h:
11         (DRTTestRunner):
12         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
13         (WebTestRunner::WebTestDelegate::showDevTools):
14         (WebTestRunner::WebTestDelegate::closeDevTools):
15         (WebTestRunner::WebTestDelegate::evaluateInWebInspector):
16         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
17         (WebTestRunner::TestRunner::TestRunner):
18         (WebTestRunner::TestRunner::showWebInspector):
19         (WebTestRunner):
20         (WebTestRunner::TestRunner::closeWebInspector):
21         (WebTestRunner::TestRunner::evaluateInWebInspector):
22         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
23         (TestRunner):
24         * DumpRenderTree/chromium/WebViewHost.cpp:
25         (WebViewHost::showDevTools):
26         (WebViewHost::closeDevTools):
27         (WebViewHost::evaluateInWebInspector):
28         * DumpRenderTree/chromium/WebViewHost.h:
29         (WebViewHost):
30
31 2013-01-11  Julie Parent  <jparent@chromium.org>
32
33         Builder should not be a valid parameter for stats dashboard
34         https://bugs.webkit.org/show_bug.cgi?id=106698
35
36         Reviewed by Dirk Pranke.
37         
38         Since builder is a dashboard specific parameter, dashboard_base
39         should not be handling it in handleValidHashParameterWrapper, rather,
40         each specific dashboard should in its handleValidHashParameter function.
41     
42         Note that the FIXME in dashboard_base about this was actually wrong -
43         we don't want to move it to g_crossDashboardState, because it isn't
44         actually cross dashboard state.
45         
46         In fact, flakiness and treemap have the code in 
47         handleValidHashParameter, but it is currently dead code because 
48         dashboard_base gets to it first.
49         
50         Added the proper handling to timeline_explorer, since builder is
51         valid there.
52
53         * TestResultServer/static-dashboards/dashboard_base.js:
54         * TestResultServer/static-dashboards/timeline_explorer.html:
55
56 2013-01-11  Ryosuke Niwa  <rniwa@webkit.org>
57
58         REGRESSION: buildbot mangles error messages with timestamps run-webkit-tests outputs
59         https://bugs.webkit.org/show_bug.cgi?id=106688
60
61         Reviewed by Eric Seidel.
62
63         Update the regular expression used to strip the garbage at the beginning.
64
65         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
66         (RunWebKitTests.nrwt_log_message_regexp):
67
68 2013-01-11  Pratik Solanki  <psolanki@apple.com>
69
70         TestRunner leaks when running tests
71         https://bugs.webkit.org/show_bug.cgi?id=106683
72
73         Reviewed by Simon Fraser.
74
75         Remove extra ref() that we missed out on deleting when refactoring window.layoutTestController
76         to window.testRunner in r124705. This caused the TestRunner object to be leaked when running
77         tests.
78
79         * DumpRenderTree/TestRunner.cpp:
80         (TestRunner::makeWindowObject):
81
82 2013-01-11  Julie Parent  <jparent@chromium.org>
83
84         Dashboard Cleanup: Add isLoadingComplete to the loader.Loader object.
85         https://bugs.webkit.org/show_bug.cgi?id=106247
86
87         Old code nulled out the loader instance when it completed loading and
88         then later used the fact that it was null to determine if it had loaded
89         or not.  This is not only unintuitive, but it also prevents using the
90         loader object later on.
91         
92         Added new method, used it, added unit test for it.
93         
94         Reviewed by Dirk Pranke.
95
96         * TestResultServer/static-dashboards/dashboard_base.js:
97         (resourceLoadingComplete):
98         (handleLocationChange):
99         * TestResultServer/static-dashboards/loader.js:
100         (.):
101         * TestResultServer/static-dashboards/loader_unittests.js:
102
103 2013-01-11  Ryosuke Niwa  <rniwa@webkit.org>
104
105         Try CRLF to LF change in r139407 again.
106
107         * BuildSlaveSupport/delete-stale-build-files:
108         (main):
109         (webkitBuildDirectory):
110         * BuildSlaveSupport/kill-old-processes:
111         (main):
112
113 2013-01-11  Carlos Garcia Campos  <cgarcia@igalia.com>
114
115         [GTK] Disable the unit tests timeout when running from make check
116         https://bugs.webkit.org/show_bug.cgi?id=106670
117
118         Reviewed by Philippe Normand.
119
120         * GNUmakefile.am: Pass --timeout=-1 to run-gtk-test.
121
122 2013-01-11  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>
123
124         Yet another unreviewed fix after r139431.
125
126         * Scripts/run-launcher: Pass the arguments to `jhbuild-wrapper' in
127         the correct order.
128
129 2013-01-11  Dan Carney  <dcarney@google.com>
130
131         [chromium] move some whitespace and resize related methods to TestRunner library
132         https://bugs.webkit.org/show_bug.cgi?id=106655
133
134         Reviewed by Jochen Eisinger.
135
136         * DumpRenderTree/chromium/DRTTestRunner.cpp:
137         (DRTTestRunner::DRTTestRunner):
138         * DumpRenderTree/chromium/DRTTestRunner.h:
139         (DRTTestRunner):
140         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
141         (WebKit):
142         (WebTestRunner::WebTestDelegate::setClientWindowRect):
143         (WebTestRunner::WebTestDelegate::setSelectTrailingWhitespaceEnabled):
144         (WebTestRunner::WebTestDelegate::setSmartInsertDeleteEnabled):
145         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
146         (WebTestRunner::TestRunner::TestRunner):
147         (WebTestRunner::TestRunner::setSmartInsertDeleteEnabled):
148         (WebTestRunner):
149         (WebTestRunner::TestRunner::setSelectTrailingWhitespaceEnabled):
150         (WebTestRunner::TestRunner::enableAutoResizeMode):
151         (WebTestRunner::TestRunner::disableAutoResizeMode):
152         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
153         (TestRunner):
154         * DumpRenderTree/chromium/WebViewHost.cpp:
155         (WebViewHost::setClientWindowRect):
156         * DumpRenderTree/chromium/WebViewHost.h:
157         (WebViewHost):
158
159 2013-01-11  Mario Sanchez Prada  <mario.prada@samsung.com>
160
161         Unreviewed. Update my e-mail addresses.
162
163         * Scripts/webkitpy/common/config/committers.py:
164
165 2013-01-11  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>
166
167         Another unreviewed fix after r139431.
168
169         In a way, this is also a generic bug fix, since previously isEfl()
170         would fail when called from run-efl-tests and we thus did not use
171         jhbuild to run our tests. This now resulted in us calling
172         "jhbuild-wrapper run" without specifying a platform.
173
174         * Scripts/run-efl-tests: Pass "--efl" to ARGV manually for isEfl()
175         to work.
176
177 2013-01-11  Allan Sandfeld Jensen  <allan.jensen@digia.com>
178
179         [Qt][WK1] Web Audio support
180         https://bugs.webkit.org/show_bug.cgi?id=106651
181
182         Reviewed by Jocelyn Turcotte.
183
184         Implements support for dumping audio data in DumpRenderTree.
185
186         Adds WebAudio to toggleable features in QtTestBrowser.
187
188         Enables LEGACY_WEB_AUDIO. This feature is required for 2/3 of the web audio regression tests. It is enabled by default
189         because it has no effect when WEB_AUDIO is not enabled. This combination matches the settings in FeatureList.pm.
190
191         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
192         (WebCore::DumpRenderTree::dump):
193         * DumpRenderTree/qt/TestRunnerQt.cpp:
194         (TestRunner::reset):
195         (TestRunner::overridePreference):
196         (TestRunner::setAudioData):
197         * DumpRenderTree/qt/TestRunnerQt.h:
198         (TestRunner::shouldDumpAsAudio):
199         (TestRunner::audioData):
200         (TestRunner):
201         * QtTestBrowser/launcherwindow.cpp:
202         (LauncherWindow::applyPrefs):
203         (LauncherWindow::createChrome):
204         (LauncherWindow::toggleWebAudio):
205         * QtTestBrowser/launcherwindow.h:
206         (WindowOptions::WindowOptions):
207         (WindowOptions):
208         (LauncherWindow):
209         * qmake/mkspecs/features/features.prf:
210         * qmake/mkspecs/features/features.pri:
211
212 2013-01-11  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>
213
214         Unreviewed jhbuild-related fix after r139431.
215
216         * Scripts/webkitpy/layout_tests/port/efl.py:
217         (EflPort.__init__): Pass a string to set_option_default(), since
218         shlex.split() will be called on it later.
219
220 2013-01-11  Csaba Osztrogonác  <ossy@webkit.org>
221
222         Unreviewed typo fixes after r139431.
223
224         * Scripts/run-javascriptcore-tests:
225
226 2013-01-11  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>
227
228         [jhbuild] Remove the "run-with-jhbuild" scripts.
229         https://bugs.webkit.org/show_bug.cgi?id=106572
230
231         Reviewed by Martin Robinson.
232
233         Both ports which use jhbuild were mostly duplicating the
234         `run-with-jhbuild' script in their own directories for no clear reason.
235
236         The script itself was just a thin wrapper around the `jhbuild-wrapper',
237         so just start using that directly and get rid of a useless middle man.
238
239         Adjust scripts which called run-with-jhbuild to call jhbuild-wrapper
240         instead.
241
242         * Scripts/run-javascriptcore-tests:
243         * Scripts/run-launcher:
244         * Scripts/webkitdirs.pm:
245         (runAutogenForAutotoolsProjectIfNecessary):
246         (buildAutotoolsProject):
247         (jhbuildWrapperPrefixIfNeeded):
248         (generateBuildSystemFromCMakeProject):
249         (buildCMakeGeneratedProject):
250         * Scripts/webkitpy/common/multiprocessing_bootstrap.py:
251         (run):
252         * Scripts/webkitpy/layout_tests/port/efl.py:
253         (EflPort.__init__):
254         (EflPort._image_diff_command):
255         * efl/run-with-jhbuild: Removed.
256         * gtk/run-with-jhbuild: Removed.
257
258 2013-01-11  Alan Cutter  <alancutter@chromium.org>
259
260         Allow arbitrary queue launch parameters in EWSTools/start-queue.sh
261         https://bugs.webkit.org/show_bug.cgi?id=106636
262
263         Reviewed by Adam Barth.
264
265         This parameter will allow the bot boot script to use start-queue.sh and provide an IRC password for sheriffbot.
266
267         * EWSTools/GoogleComputeEngine/build-chromium-ews.sh:
268         * EWSTools/GoogleComputeEngine/build-commit-queue.sh:
269         * EWSTools/screen-config:
270         * EWSTools/start-queue.sh:
271
272 2013-01-10  Ryosuke Niwa  <rniwa@webkit.org>
273
274         kill-old-processes and delete-stale-build-files use CRLF linebreaks
275         https://bugs.webkit.org/show_bug.cgi?id=106622
276
277         Reviewed by Simon Fraser.
278
279         Use LF instead of CRLF in kill-old-processes and delete-stale-build-files.
280         Also escape clang\+\+ correctly in kill-old-processes.
281
282         * BuildSlaveSupport/delete-stale-build-files:
283         * BuildSlaveSupport/kill-old-processes:
284
285 2013-01-10  Ryosuke Niwa  <rniwa@webkit.org>
286
287         Build fix after r139357. It was missing a comma.
288
289         * QueueStatusServer/model/queues.py:
290         (Queue):
291
292 2013-01-10  Tim 'mithro' Ansell  <mithro@mithis.com>
293
294         Fixing AuthenticationError when running test-webkitpy as a non-committer.
295         https://bugs.webkit.org/show_bug.cgi?id=106420
296
297         Reviewed by Eric Seidel.
298
299         * Scripts/webkitpy/common/checkout/scm/git.py:
300         (Git.push_local_commits_to_server):
301         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
302         * Scripts/webkitpy/common/checkout/scm/svn.py:
303         (SVNRepository):
304         (SVNRepository.has_authorization_for_realm):
305
306 2013-01-10  Christophe Dumez  <christophe.dumez@intel.com>
307
308         [EFL][jhbuild] Use tarballs for gstreamer instead of git
309         https://bugs.webkit.org/show_bug.cgi?id=106552
310
311         Reviewed by Laszlo Gombos.
312
313         Use tarballs for gstreamer instead of git in EFL's jhbuild
314         to work around intermittent network issues on our build
315         bots. Tarballs are fully cached and require no network
316         operation if their checksum matches.
317
318         * efl/jhbuild.modules:
319
320 2013-01-10  Ryosuke Niwa  <rniwa@webkit.org>
321
322         Add Mac WK2 EWS bots
323         https://bugs.webkit.org/show_bug.cgi?id=106590
324
325         Reviewed by Adam Barth.
326
327         Added MacWK2EWS, and made myself a watcher for MacEWS and MacWK2EWS.
328         Also add a deprecated MacWK2Port class.
329
330         * QueueStatusServer/model/queues.py:
331         (Queue):
332         * Scripts/webkitpy/common/config/ports.py:
333         (DeprecatedPort.port):
334         (MacWK2Port):
335         (MacWK2Port.run_webkit_tests_command):
336         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
337         (MacEWS):
338         (MacWK2EWS):
339         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
340         (_test_ewses):
341
342 2013-01-10  Dan Carney  <dcarney@google.com>
343
344         [chromium] move webpermissionclient related methods to TestRunner library
345         https://bugs.webkit.org/show_bug.cgi?id=106548
346
347         Reviewed by Jochen Eisinger.
348
349         * DumpRenderTree/DumpRenderTree.gypi:
350         * DumpRenderTree/chromium/DRTTestRunner.cpp:
351         (DRTTestRunner::DRTTestRunner):
352         (DRTTestRunner::reset):
353         * DumpRenderTree/chromium/DRTTestRunner.h:
354         (DRTTestRunner):
355         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
356         (WebTestRunner::WebTestDelegate::normalizeLayoutTestURL):
357         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
358         (WebKit):
359         (WebTestRunner::WebTestRunner::webPermissions):
360         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
361         (WebTestRunner::TestRunner::TestRunner):
362         (WebTestRunner::TestRunner::setDelegate):
363         (WebTestRunner):
364         (WebTestRunner::TestRunner::reset):
365         (WebTestRunner::TestRunner::webPermissions):
366         (WebTestRunner::TestRunner::dumpPermissionClientCallbacks):
367         (WebTestRunner::TestRunner::setImagesAllowed):
368         (WebTestRunner::TestRunner::setScriptsAllowed):
369         (WebTestRunner::TestRunner::setStorageAllowed):
370         (WebTestRunner::TestRunner::setPluginsAllowed):
371         (WebTestRunner::TestRunner::setAllowDisplayOfInsecureContent):
372         (WebTestRunner::TestRunner::setAllowRunningOfInsecureContent):
373         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
374         (WebTestRunner):
375         (TestRunner):
376         * DumpRenderTree/chromium/TestRunner/src/WebPermissions.cpp: Renamed from Tools/DumpRenderTree/chromium/WebPermissions.cpp.
377         (WebTestRunner):
378         (WebTestRunner::WebPermissions::WebPermissions):
379         (WebTestRunner::WebPermissions::~WebPermissions):
380         (WebTestRunner::WebPermissions::allowImage):
381         (WebTestRunner::WebPermissions::allowScriptFromSource):
382         (WebTestRunner::WebPermissions::allowStorage):
383         (WebTestRunner::WebPermissions::allowPlugins):
384         (WebTestRunner::WebPermissions::allowDisplayingInsecureContent):
385         (WebTestRunner::WebPermissions::allowRunningInsecureContent):
386         (WebTestRunner::WebPermissions::setImagesAllowed):
387         (WebTestRunner::WebPermissions::setScriptsAllowed):
388         (WebTestRunner::WebPermissions::setStorageAllowed):
389         (WebTestRunner::WebPermissions::setPluginsAllowed):
390         (WebTestRunner::WebPermissions::setDisplayingInsecureContentAllowed):
391         (WebTestRunner::WebPermissions::setRunningInsecureContentAllowed):
392         (WebTestRunner::WebPermissions::setDelegate):
393         (WebTestRunner::WebPermissions::setDumpCallbacks):
394         (WebTestRunner::WebPermissions::reset):
395         * DumpRenderTree/chromium/TestRunner/src/WebPermissions.h: Renamed from Tools/DumpRenderTree/chromium/WebPermissions.h.
396         (WebTestRunner):
397         (WebPermissions):
398         * DumpRenderTree/chromium/TestShell.cpp:
399         (TestShell::initialize):
400         (TestShell::resetTestController):
401         (TestShell::createNewWindow):
402         * DumpRenderTree/chromium/TestShell.h:
403         (TestShell):
404         * DumpRenderTree/chromium/WebViewHost.cpp:
405         (WebViewHost::normalizeLayoutTestURL):
406         * DumpRenderTree/chromium/WebViewHost.h:
407         (WebViewHost):
408
409 2013-01-10  Alan Cutter  <alancutter@chromium.org>
410
411         Split EWS tool cold-boot.sh into modular components
412         https://bugs.webkit.org/show_bug.cgi?id=106527
413
414         Reviewed by Adam Barth.
415
416         Split cold-boot.sh script up into build-vm.sh, build-repo.sh and build-boot-cmd.sh.
417         Updated GCE bot build scripts to match.
418         Unable to test scripts directly as they depend on being in the repository to work!
419
420         * EWSTools/GoogleComputeEngine/build-chromium-ews.sh:
421         * EWSTools/GoogleComputeEngine/build-commit-queue.sh:
422         * EWSTools/build-boot-cmd.sh: Copied from Tools/EWSTools/GoogleComputeEngine/build-chromium-ews.sh.
423         * EWSTools/build-repo.sh: Copied from Tools/EWSTools/cold-boot.sh.
424         * EWSTools/build-vm.sh: Renamed from Tools/EWSTools/cold-boot.sh.
425
426 2013-01-10  Julie Parent  <jparent@chromium.org>
427
428         Dashboard Cleanup: Do not generate the page if we are about to reload.
429         https://bugs.webkit.org/show_bug.cgi?id=106584
430
431         Reviewed by Dirk Pranke.
432
433         Move the logic to determine if we should generate the page into
434         parseParameters() since that is where we have the knowledge to
435         determine it, and return that value to handleLocationChange, 
436         rather than passing the changed parameters back to handleLocationChange
437         and having it make the decision.
438         
439         This is logically equivalent, save one case: when we are about to do
440         a reload, we know that there is no reason to generate the page.  Old
441         code was causing an extra page generation in this case.
442         
443         * TestResultServer/static-dashboards/dashboard_base.js:
444         (parseParameters):
445         (handleLocationChange):
446
447 2013-01-10  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>
448
449         [jhbuild] Use the "configure" script to build glib
450         https://bugs.webkit.org/show_bug.cgi?id=106569
451
452         Reviewed by Martin Robinson.
453
454         So far both EFL and GTK have been using autogen.sh to build glib.
455         While that does work most of the time, it's not really needed since
456         we are downloading and building stable releases from tarballs.
457         Additionally, autogen.sh fails to work with automake 1.13 due to the
458         usage of some macros that have been removed (this has already been
459         fixed in glib master).
460
461         * efl/jhbuild.modules:
462         * gtk/jhbuild.modules:
463
464 2013-01-10  Tim 'mithro' Ansell  <mithro@mithis.com>
465
466         Changing tests to be consistent and use self.scm rather then recreating the scm object.
467         (Missed a couple)
468         https://bugs.webkit.org/show_bug.cgi?id=106536
469
470         Reviewed by Eric Seidel.
471
472         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
473
474 2013-01-10  Tony Chang  <tony@chromium.org>
475
476         Speed up supplemental dependency computation
477         https://bugs.webkit.org/show_bug.cgi?id=106503
478
479         Reviewed by Adam Barth.
480
481         Pass the IDL attributes file for generating the bindings in WebKitTestRunner.
482
483         * WebKitTestRunner/CMakeLists.txt:
484
485 2013-01-10  Zan Dobersek  <zandobersek@gmail.com>
486
487         Remove the ENABLE_ANIMATION_API feature define occurences
488         https://bugs.webkit.org/show_bug.cgi?id=106544
489
490         Reviewed by Simon Fraser.
491
492         The Animation API code was removed in r137243. The ENABLE_ANIMATION_API
493         feature define handling still lingers in various build systems and configurations
494         but is of no use, so it should be removed.
495
496         * Scripts/webkitperl/FeatureList.pm:
497         * qmake/mkspecs/features/features.pri:
498
499 2013-01-10  Carlos Garcia Campos  <cgarcia@igalia.com>
500
501         [GTK] Add support for loading web process extensions
502         https://bugs.webkit.org/show_bug.cgi?id=105631
503
504         Reviewed by Gustavo Noronha Silva.
505
506         * MiniBrowser/gtk/GNUmakefile.am:
507         * MiniBrowser/gtk/main.c:
508         (main): Set WEBKIT_INJECTED_BUNDLE env var to use the injected
509         bundle lib from build dir.
510         * Scripts/webkitpy/style/checker.py: Add exceptions for GTK+ API
511         located in WebProcess/InjectedBundle/API/gtk.
512         * gtk/generate-gtkdoc:
513         (get_webkit2_options): Scan also files in
514         WebProcess/InjectedBundle/API/gtk to generate API docs.
515         (get_webkit2_options.injected_bundle_src_path): Helper function to
516         build paths in WebProcess/InjectedBundle/API/gtk.
517
518 2013-01-10  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
519
520         [EFL][WK2] Writing immediately to created windows (open) fails
521         https://bugs.webkit.org/show_bug.cgi?id=105276
522
523         Reviewed by Kenneth Rohde Christiansen.
524
525         Do not reset 'about:blank' url for the new pages in mini browser.
526
527         * MiniBrowser/efl/main.c:
528         (window_create):
529
530 2013-01-10  Tim 'mithro' Ansell  <mithro@mithis.com>
531
532         Changing tests to be consistent and use self.scm rather then recreating the scm object.
533         https://bugs.webkit.org/show_bug.cgi?id=106536
534
535         Reviewed by Eric Seidel.
536
537         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
538
539 2013-01-10  Tim 'mithro' Ansell  <mithro@mithis.com>
540
541         Converting to unittest2 so we can use assertItemsEqual, making tests
542         not dependent on order of items.
543         https://bugs.webkit.org/show_bug.cgi?id=106532
544
545         Before:
546         Ran 1733 tests in 38.385s
547         FAILED (failures=1, errors=22)
548
549         After:
550         Ran 1733 tests in 32.298s
551         FAILED (failures=0, errors=22)
552
553         Reviewed by Eric Seidel.
554
555         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
556         * Scripts/webkitpy/thirdparty/__init__.py:
557         (AutoinstallImportHook._install_unittest2):
558
559 2013-01-10  Christophe Dumez  <christophe.dumez@intel.com>
560
561         [EFL] Add gstreamer 1.0.5 to jhbuild
562         https://bugs.webkit.org/show_bug.cgi?id=106178
563
564         Reviewed by Laszlo Gombos.
565
566         Add gstreamer 1.0.5 to EFL's jhbuild configuration.
567
568         * efl/jhbuild.modules:
569
570 2013-01-09  Alan Cutter  <alancutter@chromium.org>
571
572         Updated parameters in GCE build scripts so they now work.
573         https://bugs.webkit.org/show_bug.cgi?id=106523
574
575         Reviewed by Adam Barth.
576
577         * EWSTools/GoogleComputeEngine/build-chromium-ews.sh:
578         * EWSTools/GoogleComputeEngine/build-commit-queue.sh:
579         * EWSTools/cold-boot.sh:
580
581 2013-01-09  Mark Pilgrim  <pilgrim@chromium.org>
582
583         [Chromium] Remove some stray references to WebKitPlatformSupport.h in DumpRenderTree
584         https://bugs.webkit.org/show_bug.cgi?id=106520
585
586         Reviewed by James Robinson.
587
588         Part of a larger refactoring series. See tracking bug 82948.
589
590         * DumpRenderTree/chromium/WebThemeEngineDRTMac.h:
591         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
592
593 2013-01-09  Julie Parent  <jparent@chromium.org>
594
595         Dashboard cleanup: Remove code for currentState and window.location.hash getting out of state
596         https://bugs.webkit.org/show_bug.cgi?id=106514
597
598         This is old code, as these can no longer get out of sync.
599         permmaLinkURLHash is generated from currentState, which is set only
600         from the location.  The commment about how these can get out of sync
601         refers to a function that no longer exists.
602         
603         Reviewed by Dirk Pranke.
604
605         * TestResultServer/static-dashboards/dashboard_base.js:
606         (handleLocationChange):
607
608 2013-01-09  Julie Parent  <jparent@chromium.org>
609
610         Dashboard Cleanup: Reload the page as soon as we know we need to.
611         https://bugs.webkit.org/show_bug.cgi?id=106515
612
613         No need to continue parsing all of the data once we know that
614         we need to do a reload.
615         
616         Reviewed by Dirk Pranke.
617
618         * TestResultServer/static-dashboards/dashboard_base.js:
619         (parseParameters):
620
621 2013-01-09  Eric Seidel  <eric@webkit.org>
622
623         run-perf-tests --chromium-android --profile should show symbols for the kernel
624         https://bugs.webkit.org/show_bug.cgi?id=106498
625
626         Reviewed by Adam Barth.
627
628         Turns out this was easy, once I finally read the output from "perf report".
629         It appears there may be a bug in "perf record" on android, as it complains
630         about not being able to read from /proc/kallsyms even when
631         /proc/sys/kernel/kptr_restrict is 0.  For now I've not bothered
632         to keep /proc/sys/kernel/kptr_restrict as 0 during the actual record
633         but rather just flip it to 0 long enough to grab the /proc/kallsyms
634         and then flip it back to whatever the device had.
635
636         /proc/sys/kernel/kptr_restrict controls what /proc/kallsyms returns
637         on Linux.  /proc/kallsyms contains a mapping of kernel addresses
638         to symbol names.  Its world-readable, but will return all 0s if you
639         don't have permission to see the kernel symbols.  kptr_restrict
640         supports values 0, 1, 2.  Where 0 means "everyone can see the real symbols"
641         1 is only a specific group, and 2 is "no one, not even root".
642         By default kptr_restrict is 2 on Android it seems.
643         More kptr_restrict docs: http://lwn.net/Articles/420403/
644
645         I also took this opportunity to clean up how the perf record command
646         was built for use/display in AndroidPerf.profile_after_exit and
647         change to always using long-form names for the arguments (to hopefully help readability).
648
649         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
650         (AndroidPerf.__init__):
651         (profile_after_exit):
652         (ChromiumAndroidDriver.__init__):
653         (ChromiumAndroidDriver._update_kallsyms_cache):
654         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
655
656 2013-01-09  Julie Parent  <jparent@chromium.org>
657
658         Dashboard cleanup: remove dead code.
659         https://bugs.webkit.org/show_bug.cgi?id=106509
660
661         Reviewed by Dirk Pranke.
662
663         * TestResultServer/static-dashboards/dashboard_base.js:
664         (parseParameters):
665
666 2013-01-09  Joanmarie Diggs  <jdiggs@igalia.com>
667
668         [GTK] accessibility/aria-labelledby-overrides-label.html requires a proper baseline
669         https://bugs.webkit.org/show_bug.cgi?id=105638
670
671         Reviewed by Martin Robinson.
672
673         The test was failing for two reasons:
674         - AccessibilityRenderObject::correspondingLabelForControlElement() was
675           not ignoring the ARIA labelled-by property
676         - AccessibilityController::accessibleElementById() was not implemented
677
678         Because getting an element by ID cannot be done in the UIProcess, the
679         decision was made to expose the element's ID as an accessible attribute
680         of the object.
681
682         In addition, fixing the bug in AccessibilityRenderObject made it possible
683         to eliminate the Gtk platform-specific expectations for another test.
684
685         * DumpRenderTree/gtk/AccessibilityControllerGtk.cpp:
686         (childElementById): Added. Walks the tree looking for the AtkObject
687         Attribute which exposes the HTML element's ID.
688         (AccessibilityController::accessibleElementById): Implemented.
689         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
690         (AccessibilityUIElement::title): Modified to print out "AXTitle: " even
691         when there is no title, as is done with the Mac port.
692         * WebKitTestRunner/InjectedBundle/AccessibilityController.cpp:
693         (WTR):
694         * WebKitTestRunner/InjectedBundle/atk/AccessibilityControllerAtk.cpp:
695         (WTR::childElementById): Added. Walks the tree looking for the AtkObject
696         Attribute which exposes the HTML element's ID.
697         (WTR):
698         (WTR::AccessibilityController::accessibleElementById): Implemented.
699         * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
700         (WTR::AccessibilityUIElement::title): Modified to print out "AXTitle: "
701         even when there is no title, as is done with the Mac port.
702
703 2013-01-09  Gregg Tavares  <gman@google.com>
704
705         Add Gregg Tavares as a committer
706         https://bugs.webkit.org/show_bug.cgi?id=106492
707
708         Unreviewed.
709
710         * Scripts/webkitpy/common/config/committers.py:
711
712 2013-01-09  Roger Fong  <roger_fong@apple.com>
713
714         Enable seamless iFrames for Windows DRT.
715         https://bugs.webkit.org/show_bug.cgi?id=106468.
716
717         Rubberstamped by Timothy Horton.
718
719         * DumpRenderTree/win/DumpRenderTree.cpp:
720         (resetDefaultsToConsistentValues):
721
722 2013-01-09  Christophe Dumez  <christophe.dumez@intel.com>
723
724         [WK2] Implement eventSender.scheduleAsynchronousKeyDown
725         https://bugs.webkit.org/show_bug.cgi?id=93979
726
727         Reviewed by Benjamin Poulain.
728
729         Implement eventSender.scheduleAsynchronousKeyDown() in
730         WebKitTestRunner.
731
732         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
733         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
734         (WTR::createKeyDownMessageBody):
735         (WTR::EventSendingController::keyDown):
736         (WTR::EventSendingController::scheduleAsynchronousKeyDown):
737         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
738         (EventSendingController):
739         * WebKitTestRunner/TestController.cpp:
740         (WTR::TestController::didReceiveKeyDownMessageFromInjectedBundle):
741         (WTR::TestController::didReceiveMessageFromInjectedBundle):
742         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
743         * WebKitTestRunner/TestController.h:
744         (TestController):
745
746 2013-01-09  Jochen Eisinger  <jochen@chromium.org>
747
748         [chromium] move resource load callback dumping to TestRunner library
749         https://bugs.webkit.org/show_bug.cgi?id=106449
750
751         Reviewed by Adam Barth.
752
753         * DumpRenderTree/chromium/DRTTestRunner.cpp:
754         (DRTTestRunner::DRTTestRunner):
755         (DRTTestRunner::reset):
756         * DumpRenderTree/chromium/DRTTestRunner.h:
757         (DRTTestRunner):
758         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
759         (WebKit):
760         (WebTestRunner::WebTestDelegate::makeURLErrorDescription):
761         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
762         (WebKit):
763         (WebTestProxyBase):
764         (WebTestRunner::WebTestProxy::assignIdentifierToRequest):
765         (WebTestRunner::WebTestProxy::willRequestResource):
766         (WebTestRunner::WebTestProxy::willSendRequest):
767         (WebTestRunner::WebTestProxy::didReceiveResponse):
768         (WebTestRunner::WebTestProxy::didFinishResourceLoad):
769         (WebTestRunner::WebTestProxy::didFailResourceLoad):
770         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
771         (WebTestRunner::WebTestRunner::shouldDumpResourceLoadCallbacks):
772         (WebTestRunner::WebTestRunner::shouldDumpResourceRequestCallbacks):
773         (WebTestRunner::WebTestRunner::shouldDumpResourceResponseMIMETypes):
774         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
775         (WebTestRunner::TestRunner::TestRunner):
776         (WebTestRunner::TestRunner::reset):
777         (WebTestRunner::TestRunner::shouldDumpResourceLoadCallbacks):
778         (WebTestRunner::TestRunner::shouldDumpResourceRequestCallbacks):
779         (WebTestRunner::TestRunner::shouldDumpResourceResponseMIMETypes):
780         (WebTestRunner):
781         (WebTestRunner::TestRunner::dumpResourceLoadCallbacks):
782         (WebTestRunner::TestRunner::dumpResourceRequestCallbacks):
783         (WebTestRunner::TestRunner::dumpResourceResponseMIMETypes):
784         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
785         (TestRunner):
786         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
787         (WebTestRunner::WebTestProxyBase::reset):
788         (WebTestRunner):
789         (WebTestRunner::WebTestProxyBase::assignIdentifierToRequest):
790         (WebTestRunner::WebTestProxyBase::willRequestResource):
791         (WebTestRunner::WebTestProxyBase::willSendRequest):
792         (WebTestRunner::WebTestProxyBase::didReceiveResponse):
793         (WebTestRunner::WebTestProxyBase::didFinishResourceLoad):
794         (WebTestRunner::WebTestProxyBase::didFailResourceLoad):
795         * DumpRenderTree/chromium/WebViewHost.cpp:
796         (WebViewHost::willSendRequest):
797         (WebViewHost::makeURLErrorDescription):
798         (WebViewHost::reset):
799         * DumpRenderTree/chromium/WebViewHost.h:
800         (WebViewHost):
801
802 2013-01-09  Dan Carney  <dcarney@google.com>
803
804         [chromium] move dumpcreateview methods to testrunner library
805         https://bugs.webkit.org/show_bug.cgi?id=106434
806
807         Reviewed by Adam Barth.
808
809         * DumpRenderTree/chromium/DRTTestRunner.cpp:
810         (DRTTestRunner::DRTTestRunner):
811         (DRTTestRunner::reset):
812         * DumpRenderTree/chromium/DRTTestRunner.h:
813         (DRTTestRunner):
814         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
815         (WebTestRunner::WebTestRunner::shouldDumpCreateView):
816         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
817         (WebTestRunner::TestRunner::TestRunner):
818         (WebTestRunner::TestRunner::reset):
819         (WebTestRunner::TestRunner::shouldDumpCreateView):
820         (WebTestRunner):
821         (WebTestRunner::TestRunner::dumpCreateView):
822         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
823         (TestRunner):
824
825 2013-01-09  Csaba Osztrogonác  <ossy@webkit.org>
826
827         webkitpy cleanup: Merge additional_comment_text into comment_text
828         https://bugs.webkit.org/show_bug.cgi?id=106421
829
830         Reviewed by Adam Barth.
831
832         * Scripts/webkitpy/common/config/committervalidator.py:
833         (CommitterValidator.reject_patch_from_commit_queue):
834         (CommitterValidator.reject_patch_from_review_queue):
835         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
836         (Bugzilla.set_flag_on_attachment):
837         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
838         (MockBugzilla.set_flag_on_attachment):
839         * Scripts/webkitpy/tool/bot/feeders_unittest.py:
840         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
841         (AbstractEarlyWarningSystem._post_reject_message_on_bug):
842         * Scripts/webkitpy/tool/commands/queues_unittest.py:
843
844 2013-01-09  Mario Sanchez Prada  <mario.prada@samsung.com>
845
846         [GTK] Raise version of gdk-pixbuf in jhbuild up to 2.26.5
847         https://bugs.webkit.org/show_bug.cgi?id=106464
848
849         Reviewed by Martin Robinson.
850
851         Current version in jhbuild.modules (2.26.0) is giving trouble to
852         when trying to build WebKitGTK from scratch, due to a segmentation
853         fault error in gdk-pixbuf-query-loader during the 'make'
854         phase. Raising the version to 2.26.5 fixes this issue.
855
856         * gtk/jhbuild.modules: Version raised, checksums updated.
857
858 2013-01-09  János Badics  <jbadics@inf.u-szeged.hu>
859
860         [Qt][NRWT] Pass --timeout to DRT/WTR if a test is marked as SLOW.
861         https://bugs.webkit.org/show_bug.cgi?id=90968.
862
863         Reviewed by Csaba Osztrogonác.
864
865         Added functionality in DRT and WTR to use any timeout value while running
866         slow tests (eventually, any test). Now NRWT --time-out-ms determines the
867         timeout value for the test. Added a flag in NRWT (supports_per_test_timeout)
868         to indicate whether the current port supports setting timeout value
869         per test (it's False by default; I enabled it only on Qt).
870         Also corrected a typo in driver.py
871
872         * DumpRenderTree/DumpRenderTree.h:
873         (TestCommand::TestCommand):
874         (TestCommand):
875         * DumpRenderTree/DumpRenderTreeCommon.cpp:
876         (parseInputLine):
877         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
878         (WebCore::DumpRenderTree::processLine):
879         * Scripts/webkitpy/layout_tests/port/base.py:
880         (Port.supports_per_test_timeout):
881         * Scripts/webkitpy/layout_tests/port/driver.py:
882         (Driver.run_test):
883         (Driver._command_from_driver_input):
884         * Scripts/webkitpy/layout_tests/port/qt.py:
885         (QtPort.supports_per_test_timeout):
886         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
887         (WTR::InjectedBundle::InjectedBundle):
888         (WTR::InjectedBundle::didReceiveMessage):
889         (WTR::InjectedBundle::beginTesting):
890         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
891         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
892         (WTR::TestRunner::setCustomTimeout):
893         (WTR):
894         * WebKitTestRunner/InjectedBundle/TestRunner.h:
895         (TestRunner):
896         * WebKitTestRunner/InjectedBundle/qt/TestRunnerQt.cpp:
897         (WTR::TestRunner::initializeWaitToDumpWatchdogTimerIfNeeded):
898         * WebKitTestRunner/TestController.cpp:
899         (WTR::TestController::TestController):
900         (WTR::TestController::getCustomTimeout):
901         (WTR):
902         (WTR::TestCommand::TestCommand):
903         (TestCommand):
904         (WTR::parseInputLine):
905         (WTR::TestController::runTest):
906         (WTR::TestController::runUntil):
907         * WebKitTestRunner/TestController.h:
908         (TestController):
909         * WebKitTestRunner/TestInvocation.cpp:
910         (WTR::TestInvocation::TestInvocation):
911         (WTR::TestInvocation::setCustomTimeout):
912         (WTR):
913         (WTR::TestInvocation::invoke):
914         * WebKitTestRunner/TestInvocation.h:
915         (TestInvocation):
916
917 2013-01-08  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
918
919         Regression(r138681): Add HAVE(ACCESSIBILITY) guard to atk files
920         https://bugs.webkit.org/show_bug.cgi?id=106290
921
922         Reviewed by Martin Robinson.
923
924         Some atk files don't use HAVE(ACCESSIBILITY). It might make build errors when
925         the macro isn't enabled.
926
927         * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
928
929 2013-01-08  Martin Robinson  <mrobinson@igalia.com>
930
931         WebKitTestRunner needs support for setHandlesAuthenticationChallenges
932         https://bugs.webkit.org/show_bug.cgi?id=103653
933
934         Reviewed by Anders Carlsson.
935
936         Implement setHandlesAuthenticationChallenges in the UIProcess by sending credentials
937         and the enabled boolean to the UIProcess and then implementing the didReceiveAuthenticationChallengeInFrame
938         client callback.
939
940         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Exposed the new TestRunner
941         methods to the JavaScript API.
942         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
943         (WTR::TestRunner::setHandlesAuthenticationChallenges): Added this callback for the JavaScript API.
944         (WTR::TestRunner::setAuthenticationUsername): Ditto.
945         (WTR::TestRunner::setAuthenticationPassword): Ditto.
946         * WebKitTestRunner/InjectedBundle/TestRunner.h:
947         (TestRunner): Added new declarations for the above.
948         * WebKitTestRunner/TestController.cpp:
949         (WTR::TestController::createWebViewWithOptions): Hook up the new didReceiveSynchronousMessageFromInjectedBundle
950         callback.
951         (WTR::TestController::resetStateToConsistentValues): Reset the credentials and enabled boolean between tests.
952         (WTR::TestController::didReceiveAuthenticationChallengeInFrame): Handle the callback by using
953         the authentication information passed from the InjectedBundle.
954         * WebKitTestRunner/TestController.h:
955         (WTR::TestController::setHandlesAuthenticationChallenges): Added setter to allow TestInvocation
956         to communicate with the TestController.
957         (WTR::TestController::setAuthenticationUsername): Ditto.
958         (WTR::TestController::setAuthenticationPassword): Ditto.
959         (TestController):
960         * WebKitTestRunner/TestInvocation.cpp:
961         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Handle the messages sending authentication
962         information from the InjectedBundle.
963         (WTR::TestInvocation::outputText): Added this method so that the TestController can output
964         text to the TestInvocation text output.
965         * WebKitTestRunner/TestInvocation.h:
966         (TestInvocation): Added declaration for new method.
967
968 2013-01-08  Benjamin Poulain  <benjamin@webkit.org>
969
970         Add myself as a watcher for Web Geolocation
971
972         Unreviewed.
973
974         * Scripts/webkitpy/common/config/watchlist:
975
976 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
977
978         [webkitpy] Call setup_test_run method of the parent Port interface when method is called on subclasses
979         https://bugs.webkit.org/show_bug.cgi?id=106364
980
981         Reviewed by Eric Seidel.
982
983         When the setup_test_run method is called on the subclasses of the base Port
984         interface, the parent method should also be called via super(), despite the
985         parent method not doing anything at the moment.
986
987         The according calls to super() are required in the ChromiumPort, EflPort
988         and GtkPort implementations.
989
990         * Scripts/webkitpy/layout_tests/port/chromium.py:
991         (ChromiumPort.setup_test_run):
992         * Scripts/webkitpy/layout_tests/port/efl.py:
993         (EflPort.setup_test_run):
994         * Scripts/webkitpy/layout_tests/port/gtk.py:
995         (GtkPort.setup_test_run):
996
997 2013-01-08  Julie Parent  <jparent@chromium.org>
998
999         Dashboard cleanup: Remove globals g_buildersThatFailedToLoad and g_staleBuilders
1000         https://bugs.webkit.org/show_bug.cgi?id=106356
1001
1002         g_buildersThatFailedToLoad and g_staleBuilders were globals defined in
1003         dashboard_base, assigned by Loader, and used only by dashboard_base to
1004         create error messages.  Moved the variables to be privates on the Loader
1005         object, moved error message creation to _getLoadingErrorMessages on the
1006         Loader object, and now pass the errors back to dashboard base via the
1007         resourceLoadingComplete callback.
1008         
1009         Also removed the now unused clearError function, it was only being used
1010         by unit tests to clean up global state.
1011         
1012         Reviewed by Dirk Pranke.
1013
1014         * TestResultServer/static-dashboards/dashboard_base.js:
1015         (resourceLoadingComplete):
1016         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1017         * TestResultServer/static-dashboards/loader.js:
1018         (.):
1019         * TestResultServer/static-dashboards/loader_unittests.js:
1020
1021 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
1022
1023         [EFL][GTK] Make the PulseAudioSanitizer an object on the EflPort, GtkPort
1024         https://bugs.webkit.org/show_bug.cgi?id=106354
1025
1026         Reviewed by Eric Seidel.
1027
1028         Put the PulseAudioSanitizer object on the EflPort and GtkPort interfaces
1029         instead of those two inheriting from it. Also add a mock object of the
1030         sanitizer that's used in unit tests.
1031
1032         * Scripts/webkitpy/layout_tests/port/efl.py:
1033         (EflPort):
1034         (EflPort.__init__):
1035         (EflPort.setup_test_run):
1036         (EflPort.clean_up_test_run):
1037         * Scripts/webkitpy/layout_tests/port/efl_unittest.py: Also correct the
1038         importing order.
1039         (EflPortTest.make_port): Put a mock PulseAudioSanitizer on the instance.
1040         * Scripts/webkitpy/layout_tests/port/gtk.py:
1041         (GtkPort):
1042         (GtkPort.__init__):
1043         (GtkPort.setup_test_run):
1044         (GtkPort.clean_up_test_run):
1045         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py: Also correct the
1046         importing order.
1047         (GtkPortTest.make_port): Put a mock PulseAudioSanitizer on the instance.
1048         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer.py:
1049         (PulseAudioSanitizer.unload_pulseaudio_module): Stylize the method as public.
1050         (PulseAudioSanitizer.restore_pulseaudio_module): Ditto.
1051         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer_mock.py: Added.
1052         (PulseAudioSanitizerMock): A simple mock interface for PulseAudioSanitizer.
1053         (PulseAudioSanitizerMock.unload_pulseaudio_module):
1054         (PulseAudioSanitizerMock.restore_pulseaudio_module):
1055
1056 2013-01-08  Jochen Eisinger  <jochen@chromium.org>
1057
1058         [chromium] move dumpFrameLoadCallbacks and friends to the TestRunner library
1059         https://bugs.webkit.org/show_bug.cgi?id=106324
1060
1061         Reviewed by Adam Barth.
1062
1063         Since the frame load callbacks are part of the WebFrameClient, the
1064         WebTestProxy is now intercepting both WebViewClient and WebFrameClient
1065         methods.
1066
1067         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1068         (DRTTestRunner::DRTTestRunner):
1069         (DRTTestRunner::reset):
1070         * DumpRenderTree/chromium/DRTTestRunner.h:
1071         (DRTTestRunner):
1072         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
1073         (WebTestInterfaces):
1074         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
1075         (WebKit):
1076         (WebTestProxyBase):
1077         (WebTestRunner):
1078         (WebTestRunner::WebTestProxy::WebTestProxy):
1079         (WebTestProxy):
1080         (WebTestRunner::WebTestProxy::didInvalidateRect):
1081         (WebTestRunner::WebTestProxy::didScrollRect):
1082         (WebTestRunner::WebTestProxy::scheduleComposite):
1083         (WebTestRunner::WebTestProxy::scheduleAnimation):
1084         (WebTestRunner::WebTestProxy::setWindowRect):
1085         (WebTestRunner::WebTestProxy::show):
1086         (WebTestRunner::WebTestProxy::didAutoResize):
1087         (WebTestRunner::WebTestProxy::postAccessibilityNotification):
1088         (WebTestRunner::WebTestProxy::startDragging):
1089         (WebTestRunner::WebTestProxy::shouldBeginEditing):
1090         (WebTestRunner::WebTestProxy::shouldEndEditing):
1091         (WebTestRunner::WebTestProxy::shouldInsertNode):
1092         (WebTestRunner::WebTestProxy::shouldInsertText):
1093         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
1094         (WebTestRunner::WebTestProxy::shouldDeleteRange):
1095         (WebTestRunner::WebTestProxy::shouldApplyStyle):
1096         (WebTestRunner::WebTestProxy::didBeginEditing):
1097         (WebTestRunner::WebTestProxy::didChangeSelection):
1098         (WebTestRunner::WebTestProxy::didChangeContents):
1099         (WebTestRunner::WebTestProxy::didEndEditing):
1100         (WebTestRunner::WebTestProxy::registerIntentService):
1101         (WebTestRunner::WebTestProxy::dispatchIntent):
1102         (WebTestRunner::WebTestProxy::willPerformClientRedirect):
1103         (WebTestRunner::WebTestProxy::didCancelClientRedirect):
1104         (WebTestRunner::WebTestProxy::didStartProvisionalLoad):
1105         (WebTestRunner::WebTestProxy::didReceiveServerRedirectForProvisionalLoad):
1106         (WebTestRunner::WebTestProxy::didFailProvisionalLoad):
1107         (WebTestRunner::WebTestProxy::didCommitProvisionalLoad):
1108         (WebTestRunner::WebTestProxy::didReceiveTitle):
1109         (WebTestRunner::WebTestProxy::didFinishDocumentLoad):
1110         (WebTestRunner::WebTestProxy::didHandleOnloadEvents):
1111         (WebTestRunner::WebTestProxy::didFailLoad):
1112         (WebTestRunner::WebTestProxy::didFinishLoad):
1113         (WebTestRunner::WebTestProxy::didChangeLocationWithinPage):
1114         (WebTestRunner::WebTestProxy::didDisplayInsecureContent):
1115         (WebTestRunner::WebTestProxy::didRunInsecureContent):
1116         (WebTestRunner::WebTestProxy::didDetectXSS):
1117         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
1118         (WebTestRunner::WebTestRunner::setTestIsRunning):
1119         (WebTestRunner::WebTestRunner::shouldDumpFrameLoadCallbacks):
1120         (WebTestRunner::WebTestRunner::setShouldDumpFrameLoadCallbacks):
1121         (WebTestRunner::WebTestRunner::shouldDumpUserGestureInFrameLoadCallbacks):
1122         (WebTestRunner::WebTestRunner::stopProvisionalFrameLoads):
1123         (WebTestRunner::WebTestRunner::shouldDumpTitleChanges):
1124         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1125         (WebTestRunner::TestRunner::TestRunner):
1126         (WebTestRunner::TestRunner::reset):
1127         (WebTestRunner::TestRunner::setTestIsRunning):
1128         (WebTestRunner):
1129         (WebTestRunner::TestRunner::shouldDumpFrameLoadCallbacks):
1130         (WebTestRunner::TestRunner::setShouldDumpFrameLoadCallbacks):
1131         (WebTestRunner::TestRunner::shouldDumpUserGestureInFrameLoadCallbacks):
1132         (WebTestRunner::TestRunner::stopProvisionalFrameLoads):
1133         (WebTestRunner::TestRunner::shouldDumpTitleChanges):
1134         (WebTestRunner::TestRunner::dumpFrameLoadCallbacks):
1135         (WebTestRunner::TestRunner::dumpUserGestureInFrameLoadCallbacks):
1136         (WebTestRunner::TestRunner::setStopProvisionalFrameLoads):
1137         (WebTestRunner::TestRunner::dumpTitleChanges):
1138         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1139         (TestRunner):
1140         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
1141         (WebTestInterfaces::Internal):
1142         (WebTestRunner::WebTestInterfaces::Internal::webView):
1143         (WebTestRunner::WebTestInterfaces::Internal::Internal):
1144         (WebTestRunner::WebTestInterfaces::Internal::setWebView):
1145         (WebTestRunner):
1146         (WebTestRunner::WebTestInterfaces::Internal::setTestIsRunning):
1147         (WebTestRunner::WebTestInterfaces::setWebView):
1148         (WebTestRunner::WebTestInterfaces::setTestIsRunning):
1149         (WebTestRunner::WebTestInterfaces::webView):
1150         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
1151         (WebTestRunner::WebTestProxyBase::willPerformClientRedirect):
1152         (WebTestRunner):
1153         (WebTestRunner::WebTestProxyBase::didCancelClientRedirect):
1154         (WebTestRunner::WebTestProxyBase::didStartProvisionalLoad):
1155         (WebTestRunner::WebTestProxyBase::didReceiveServerRedirectForProvisionalLoad):
1156         (WebTestRunner::WebTestProxyBase::didFailProvisionalLoad):
1157         (WebTestRunner::WebTestProxyBase::didCommitProvisionalLoad):
1158         (WebTestRunner::WebTestProxyBase::didReceiveTitle):
1159         (WebTestRunner::WebTestProxyBase::didFinishDocumentLoad):
1160         (WebTestRunner::WebTestProxyBase::didHandleOnloadEvents):
1161         (WebTestRunner::WebTestProxyBase::didFailLoad):
1162         (WebTestRunner::WebTestProxyBase::didFinishLoad):
1163         (WebTestRunner::WebTestProxyBase::didChangeLocationWithinPage):
1164         (WebTestRunner::WebTestProxyBase::didDisplayInsecureContent):
1165         (WebTestRunner::WebTestProxyBase::didRunInsecureContent):
1166         (WebTestRunner::WebTestProxyBase::didDetectXSS):
1167         * DumpRenderTree/chromium/TestShell.cpp:
1168         (TestShell::runFileTest):
1169         (TestShell::testFinished):
1170         * DumpRenderTree/chromium/TestShell.h:
1171         (TestShell):
1172         * DumpRenderTree/chromium/WebViewHost.cpp:
1173         (WebViewHost::didStartProvisionalLoad):
1174         (WebViewHost::didReceiveServerRedirectForProvisionalLoad):
1175         (WebViewHost::didFailProvisionalLoad):
1176         (WebViewHost::didCommitProvisionalLoad):
1177         (WebViewHost::didReceiveTitle):
1178         (WebViewHost::didFailLoad):
1179         (WebViewHost::didFinishLoad):
1180         * DumpRenderTree/chromium/WebViewHost.h:
1181         (WebViewHost):
1182
1183 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
1184
1185         [GTK] Lower the default timeout value for WebKitTestRunner
1186         https://bugs.webkit.org/show_bug.cgi?id=106353
1187
1188         Reviewed by Dirk Pranke.
1189
1190         The default timeout value for WebKitTestRunner should be lowered.
1191         The current value (80 seconds) is irrational and just leads to producing
1192         text failures for layout tests that genuinely time out. This just leads
1193         to more deviation from the GTK WK1 port in terms of test expectations.
1194
1195         The value used is the same as when running the DumpRenderTree, that's
1196         6 seconds in release builds and 12 seconds in debug builds.
1197
1198         * Scripts/webkitpy/layout_tests/port/gtk.py:
1199         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
1200         (GtkPortTest.test_default_timeout_ms):
1201
1202 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
1203
1204         test-webkitpy shows failure output on my linux box
1205         https://bugs.webkit.org/show_bug.cgi?id=101261
1206
1207         Reviewed by Eric Seidel.
1208
1209         Mock out the PulseAudioSanitizer methods when creating testing instances
1210         of EflPort and GtkPort interfaces in unit tests. This stops throwing up
1211         PulseAudio output when the sanitizer was trying to unload modules which
1212         were already unloaded by another unit test that was being run in parallel.
1213
1214         * Scripts/webkitpy/layout_tests/port/efl_unittest.py:
1215         (EflPortTest):
1216         (EflPortTest.make_port):
1217         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
1218         (GtkPortTest):
1219         (GtkPortTest.make_port):
1220
1221 2013-01-08  Julie Parent  <jparent@chromium.org>
1222
1223         Dashboard Cleanup: Remove g_hasDoneInitialPageGeneration
1224         https://bugs.webkit.org/show_bug.cgi?id=106254
1225
1226         Remove the global variable g_hasDoneInitialPageGeneration.  It was used
1227         only to figure out if we should call addBuilderLoadErrors during
1228         handleLocationChange, which was called when the loader completed and on
1229         any hash change, but we only ever wanted to call addBuilderLoadErrors
1230         on the initial loader load.  Moved the call of addBuilderLoadErrors to
1231         resourceLoadingComplete rather than handleLocationChange, so now we
1232         only call it when we want to, and the check can go away completely.
1233
1234         Reviewed by Dirk Pranke.
1235
1236         * TestResultServer/static-dashboards/dashboard_base.js:
1237         (addBuilderLoadErrors):
1238         (resourceLoadingComplete):
1239         (handleLocationChange):
1240         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1241
1242 2013-01-08  Dominic Mazzoni  <dmazzoni@google.com>
1243
1244         Add aboxhall@chromium.org to accessibility watchlist.
1245
1246         Unreviewed.
1247
1248         * Scripts/webkitpy/common/config/watchlist:
1249
1250 2013-01-08  Dan Winship  <danw@gnome.org>
1251
1252         Add Apache 2.4 support for Fedora 18
1253         https://bugs.webkit.org/show_bug.cgi?id=104478
1254
1255         Reviewed by Martin Robinson.
1256
1257         Apache 2.4's httpd.conf is slightly incompatible with 2.2's, so we
1258         need a separate config.
1259
1260         * Scripts/webkitpy/layout_tests/port/base.py:
1261         (Port._apache_version):
1262         (Port._apache_config_file_name_for_platform):
1263         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1264         * Scripts/run-webkit-httpd:
1265         * Scripts/webkitperl/httpd.pm:
1266         (getApacheVersion):
1267         (getDefaultConfigForTestDirectory):
1268         (getHTTPDConfigPathForTestDirectory):
1269
1270 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
1271
1272         Make the override-feature-defines script compatible with Python v3
1273         https://bugs.webkit.org/show_bug.cgi?id=105742
1274
1275         Reviewed by Martin Robinson.
1276
1277         Instead of printing output through a statement, use the print function and format
1278         the output using the str.format method. This makes the script runnable with Python v3.
1279
1280         * gtk/override-feature-defines:
1281         (adjust_feature_defines_makefile):
1282
1283 2013-01-08  Heikki Paajanen  <heikki.paajanen@palm.com>
1284
1285         [Qt] Fix libxml2 config test
1286         https://bugs.webkit.org/show_bug.cgi?id=106317
1287
1288         Reviewed by Simon Hausmann.
1289
1290         Don't assume standard include/libxml2, libxml-2.0.pc should know
1291         where it's headers are.
1292
1293         * qmake/config.tests/libxml2/libxml2.cpp: Fix include
1294
1295 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
1296
1297         [GTK] Make Tools/gtk/generate-gtkdoc compatible with Python 3
1298         https://bugs.webkit.org/show_bug.cgi?id=106195
1299
1300         Reviewed by Philippe Normand.
1301
1302         Perform changes in the generate-gtkdoc script and common and gtkdoc modules
1303         to make the Python code compatible with Python 3. This includes modifying
1304         print statements, exception handling, dictionary iteration and
1305         byte-sequence-to-string conversion.
1306
1307         * gtk/common.py:
1308         (get_build_path):
1309         (pkg_config_file_variable):
1310         (gtk_version_of_pkg_config_file):
1311         * gtk/generate-gtkdoc:
1312         (print_missing_api):
1313         * gtk/gtkdoc.py:
1314         (GTKDoc.__init__):
1315         (GTKDoc._run_command):
1316
1317 2013-01-02  Steve Block  <steveblock@chromium.org>
1318
1319         Add chromium.org email address for Steve Block.
1320
1321         Unreviewed.
1322
1323         * Scripts/webkitpy/common/config/committers.py:
1324
1325 2013-01-07  Filip Pizlo  <fpizlo@apple.com>
1326
1327         Rationalize closure call heuristics and profiling
1328         https://bugs.webkit.org/show_bug.cgi?id=106270
1329
1330         Reviewed by Oliver Hunt.
1331         
1332         Add ability to use display-profiler-output via a pipe, and add the ability to dump
1333         all generated code ('display *' or 'd *').
1334
1335         * Scripts/display-profiler-output:
1336
1337 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
1338
1339         Sorted more xcodeproj files.
1340
1341         * MiniBrowser/MiniBrowser.xcodeproj/project.pbxproj:
1342         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1343         * WebKitLauncher/WebKitLauncher.xcodeproj/project.pbxproj:
1344
1345 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
1346
1347         Sorted xcodeproj files.
1348
1349         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1350         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
1351
1352 2013-01-07  Dan Bernstein  <mitz@apple.com>
1353
1354         [mac] WebKit1 clients can’t tell when a frame has been removed from the hierarchy
1355         https://bugs.webkit.org/show_bug.cgi?id=106261
1356
1357         Reviewed by Simon Fraser.
1358
1359         Added a test for -[WebFrameLoadDelegate webView:didRemoveFrameFromHierarchy:].
1360
1361         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1362         * TestWebKitAPI/Tests/mac/WebViewDidRemoveFrameFromHierarchy.mm: Added.
1363         (-[DidRemoveFrameFromHierarchyFrameLoadDelegate webView:didFinishLoadForFrame:]):
1364         (-[DidRemoveFrameFromHierarchyFrameLoadDelegate webView:didRemoveFrameFromHierarchy:]):
1365         (TestWebKitAPI):
1366         (TestWebKitAPI::TEST):
1367
1368 2013-01-07  Ojan Vafai  <ojan@chromium.org>
1369
1370         Add the Apple MountainLion bots to garden-o-matic.
1371         Eventually, we really should move away from hardcoding
1372         the list of bots.
1373
1374         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1375
1376 2013-01-07  Jochen Eisinger  <jochen@chromium.org>
1377
1378         [chromium] move web audio related methods to testrunner library
1379         https://bugs.webkit.org/show_bug.cgi?id=106211
1380
1381         Reviewed by Adam Barth.
1382
1383         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1384         (DRTTestRunner::DRTTestRunner):
1385         (DRTTestRunner::reset):
1386         * DumpRenderTree/chromium/DRTTestRunner.h:
1387         (DRTTestRunner):
1388         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
1389         (WebKit):
1390         (WebTestRunner::WebTestRunner::shouldDumpAsAudio):
1391         (WebTestRunner::WebTestRunner::audioData):
1392         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1393         (WebTestRunner::TestRunner::TestRunner):
1394         (WebTestRunner::TestRunner::reset):
1395         (WebTestRunner::TestRunner::shouldDumpAsAudio):
1396         (WebTestRunner):
1397         (WebTestRunner::TestRunner::audioData):
1398         (WebTestRunner::TestRunner::setAudioData):
1399         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1400         (TestRunner):
1401         * DumpRenderTree/chromium/TestShell.cpp:
1402         (TestShell::dump):
1403
1404 2013-01-07  Mike West  <mkwst@chromium.org>
1405
1406         Make the IFRAME_SEAMLESS flag runtime-enabled.
1407         https://bugs.webkit.org/show_bug.cgi?id=106213
1408
1409         Reviewed by Ojan Vafai.
1410
1411         * DumpRenderTree/chromium/TestShell.cpp:
1412         (TestShell::TestShell):
1413         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1414         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1415         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1416         (resetDefaultsToConsistentValues):
1417         * DumpRenderTree/mac/DumpRenderTree.mm:
1418         (resetDefaultsToConsistentValues):
1419         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1420         (WebCore::WebPage::resetSettings):
1421         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1422         (WTR::InjectedBundle::beginTesting):
1423             Ensure that seamless IFrames are enabled for layout tests in each of
1424             the ports that support seamless IFrames.
1425
1426 2013-01-07  Jochen Eisinger  <jochen@chromium.org>
1427
1428         [chromium] move dumpAsText and friends to the TestRunner library
1429         https://bugs.webkit.org/show_bug.cgi?id=106222
1430
1431         Reviewed by Adam Barth.
1432
1433         This moves dumpAsText, dumpChildFramesAsText, and dumpChildFrameScrollPositions to the TestRunner library.
1434
1435         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1436         (DRTTestRunner::DRTTestRunner):
1437         (DRTTestRunner::reset):
1438         * DumpRenderTree/chromium/DRTTestRunner.h:
1439         (DRTTestRunner):
1440         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
1441         (WebTestRunner::WebTestRunner::shouldDumpAsText):
1442         (WebTestRunner::WebTestRunner::setShouldDumpAsText):
1443         (WebTestRunner::WebTestRunner::shouldGeneratePixelResults):
1444         (WebTestRunner::WebTestRunner::setShouldGeneratePixelResults):
1445         (WebTestRunner::WebTestRunner::shouldDumpChildFrameScrollPositions):
1446         (WebTestRunner::WebTestRunner::shouldDumpChildFramesAsText):
1447         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1448         (WebTestRunner::TestRunner::TestRunner):
1449         (WebTestRunner::TestRunner::reset):
1450         (WebTestRunner::TestRunner::shouldDumpAsText):
1451         (WebTestRunner):
1452         (WebTestRunner::TestRunner::setShouldDumpAsText):
1453         (WebTestRunner::TestRunner::shouldGeneratePixelResults):
1454         (WebTestRunner::TestRunner::setShouldGeneratePixelResults):
1455         (WebTestRunner::TestRunner::shouldDumpChildFrameScrollPositions):
1456         (WebTestRunner::TestRunner::shouldDumpChildFramesAsText):
1457         (WebTestRunner::TestRunner::dumpAsText):
1458         (WebTestRunner::TestRunner::dumpChildFrameScrollPositions):
1459         (WebTestRunner::TestRunner::dumpChildFramesAsText):
1460         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1461         (TestRunner):
1462
1463 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
1464
1465         Build fix attempt after r138810.
1466
1467         Apparently python 2.6 is really bad at floating point computation.
1468
1469         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py:
1470         (MainTest.test_run_memory_test):
1471
1472 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
1473
1474         Extract a class to represent a test result to simplify PerfTest.parse_output and PageLoadPerfTest._run_with_driver
1475         https://bugs.webkit.org/show_bug.cgi?id=106172
1476
1477         Reviewed by Tony Chang.
1478
1479         Extracted PerfTestMetric, which represents a test result and knows how to compute statistics and convert itself
1480         to a dictionary. A metric is a type of measurement each test makes such as "Time", "Malloc", and "JSHeap".
1481         A single test (e.g. html5-full-render.html) may have multiple metrics, each of which is reported as a separate
1482         test on perf-o-matic and perf. test's results page.
1483
1484         Also, use python's compute_statistics instead of reading statistics off of DriverOutput to simplify the code.
1485
1486         * Scripts/webkitpy/performance_tests/perftest.py:
1487         (PerfTestMetric): Added.
1488         (PerfTestMetric.__init__): Added.
1489         (PerfTestMetric.metric): Added.
1490         (PerfTestMetric.has_values): Added.
1491         (PerfTestMetric.legacy_chromium_bot_compatible_test_name): Added. Following the legacy convention, "Malloc" and
1492         "JSHeap" will use "test_name:Malloc" whereas "Time" uses "test_name" itself. We should eventually move these
1493         conversions to be done on the client side (perf-o-matic and results page).
1494         (PerfTestMetric.append): Added.
1495         (PerfTestMetric.to_dict): Added. Computes the statistics and generates the dictionary object that represents
1496         a result for this metric.
1497         (PerfTestMetric.metric_to_unit): Added. Each metric maps to exactly one unit (it's not one-to-one).
1498         (PerfTestMetric.compute_statistics): Moved from PerfTest.
1499
1500         (PerfTest._create_driver): Extracted from PerfTest.run.
1501         (PerfTest.run): Given an array of PerfTestMetric, generate the results dictionary. Also print out results
1502         as needed.
1503         (PerfTest._run_with_driver):
1504         (PerfTest): Removed _result_classes and _result_class_regex. "result class" is the old name for "metric".
1505         (PerfTest.parse_output): Dramatically simplified. We now just return an array of PerfTestMetric's. Also ignore
1506         all keys except "values". We rely on PerfTestMetric to compute the statistics for us.
1507
1508         (ChromiumStylePerfTest.run): Added and merged _run_with_driver. Since chromium style perf tests are so different
1509         from other perf tests, there isn't much point in sharing the code here.
1510
1511         (PageLoadingPerfTest._run_with_driver): Simplified. Just create an array of PerfTestMetric's without computing
1512         statistics for the measured values.
1513
1514         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1515         (TestPerfTestMetric): Added.
1516         (TestPerfTestMetric.test_init_set_missing_unit): Added.
1517         (TestPerfTestMetric.test_legacy_chromium_bot_compatible_test_name): Added.
1518         (TestPerfTestMetric.test_has_values): Added.
1519         (TestPerfTestMetric.test_append): Added.
1520         (TestPerfTestMetric.test_compute_statistics.compute_statistics):
1521         (TestPerfTestMetric.test_compute_statistics): Added a trivial test case.
1522
1523         (TestPerfTest._assert_results_are_correct): Updated as needed.
1524         (TestPageLoadingPerfTest.test_run): Ditto. Notice that we no longer log anything in _run_with_driver. It's done
1525         in PerfTest.run now after generating the results dictionary.
1526         (TestPageLoadingPerfTest.test_run_with_memory_output): Ditto.
1527
1528 2013-01-07  Julien Chaffraix  <jchaffraix@webkit.org>
1529
1530         Support size_t multiplication and division operators on LayoutUnit
1531         https://bugs.webkit.org/show_bug.cgi?id=83848
1532
1533         Reviewed by Emil A Eklund.
1534
1535         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
1536         Added some simple tests for the new operators using size_t.
1537
1538 2013-01-07  Csaba Osztrogonác  <ossy@webkit.org>
1539
1540         [Qt] run-webkit-tests cleanup for parallel NRWT
1541         https://bugs.webkit.org/show_bug.cgi?id=106220
1542
1543         Reviewed by Dirk Pranke.
1544
1545         * Scripts/run-webkit-tests:
1546         (useNewRunWebKitTests):
1547         * Scripts/webkitpy/layout_tests/port/qt.py:
1548         (QtPort):
1549         (QtPort.default_child_processes):
1550
1551 2013-01-07  Philippe Normand  <pnormand@igalia.com>
1552
1553         Create a GStreamerAudio watchlist and add myself to it.
1554
1555         Unreviewed.
1556
1557         * Scripts/webkitpy/common/config/watchlist:
1558
1559 2013-01-05  Ryosuke Niwa  <rniwa@webkit.org>
1560
1561         Add svn:executable to a bunch of scripts in BuildSlaveSupport.
1562         Also removed the empty mac directory.
1563
1564         * BuildSlaveSupport/built-product-archive: Added property svn:executable.
1565         * BuildSlaveSupport/download-built-product: Added property svn:executable.
1566         * BuildSlaveSupport/mac: Removed.
1567         * BuildSlaveSupport/test-result-archive: Added property svn:executable.
1568
1569 2013-01-04  Ryosuke Niwa  <rniwa@webkit.org>
1570
1571         Add a script to delete stale zero-byte build files
1572         https://bugs.webkit.org/show_bug.cgi?id=106131
1573
1574         Reviewed by Tony Chang.
1575
1576         Delete zero-byte build files on Mac port for now. It doesn't do anything on other ports.
1577
1578         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1579         (DeleteStaleBuildFiles): Added.
1580         (Factory.__init__): Add DeleteStaleBuildFiles step.
1581         * BuildSlaveSupport/delete-stale-build-files: Added.
1582         (main): Delete zero-byte files except ones intentionally generated by our build process.
1583         (webkitBuildDirectory): Copied from build-product-archive.
1584
1585 2013-01-04  Michael Pruett  <michael@68k.org>
1586
1587         [GTK] Fix build error on GTK due to r138836
1588         https://bugs.webkit.org/show_bug.cgi?id=106154
1589
1590         Reviewed by Kentaro Hara.
1591
1592         * GNUmakefile.am:
1593
1594 2013-01-04  Dan Bernstein  <mitz@apple.com>
1595
1596         [mac] WebKit1 has no equivalent of WKBundleNodeHandleCopyHTMLTableCellElementCellAbove
1597         https://bugs.webkit.org/show_bug.cgi?id=106121
1598
1599         Reviewed by Adele Peterson.
1600
1601         Added test for -[DOMTHMLTableCellElement _cellAbove].
1602
1603         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1604         * TestWebKitAPI/Tests/mac/DOMHTMLTableCellCellAbove.mm: Added.
1605         (-[HTMLTableCellElementCellAboveTest webView:didFinishLoadForFrame:]):
1606         (TestWebKitAPI):
1607         (TestWebKitAPI::expectCellAboveCell):
1608         (TestWebKitAPI::TEST):
1609         * TestWebKitAPI/Tests/mac/DOMHTMLTableCellElementCellAbove.html: Added.
1610
1611 2013-01-04  Eric Seidel  <eric@webkit.org>
1612
1613         FlakyTestReporter should be re-enabled and taught how to post patches
1614         https://bugs.webkit.org/show_bug.cgi?id=103839
1615
1616         Reviewed by Adam Barth.
1617
1618         This moves LayoutTestResultsReader off of the DeprecatedPort object.
1619         It also adds a real Port object to the EWS and CommitQueue objects
1620         using a new PatchProcessingQueue superclass.
1621         PatchProcessingQueue doesn't seem to be the cleanest insertion point
1622         but works for this purpose.
1623         I also moved _upload_results_archive_for_patch into PatchProcessingQueue
1624         to resolve a FIXME.
1625
1626         This required some hacks/shiming between the non-specific "old"
1627         port names which the EWS system is currently using, and the new
1628         port objects.  This shimming is done in PatchProcessingQueue._new_port_name_from_old.
1629
1630         Hopefully this will fix the flaky test reporter and results uploading.
1631         Certainly it's another small step towards getting rid of the old ports.py
1632         infrastructure and making the NRWT Port class common webkitpy infrastructure.
1633
1634         Now that DeprecatedPort is no longer used for the results directory
1635         I removed all the explicit results-dir code which webkit-patch was using
1636         in --non-interactive mode to tell NRWT to use a results-dir it understood. :)
1637
1638         * Scripts/webkitpy/common/config/ports.py:
1639         (DeprecatedPort.run_perl_unittests_command):
1640         * Scripts/webkitpy/common/config/ports_mock.py:
1641         (MockPort.name):
1642         * Scripts/webkitpy/common/system/outputcapture.py:
1643         (OutputCapture.assert_outputs):
1644         * Scripts/webkitpy/layout_tests/port/apple.py:
1645         (ApplePort.determine_full_port_name):
1646         * Scripts/webkitpy/tool/bot/layouttestresultsreader.py:
1647         (LayoutTestResultsReader.__init__):
1648         (LayoutTestResultsReader):
1649         (LayoutTestResultsReader._read_file_contents):
1650         (LayoutTestResultsReader._create_layout_test_results):
1651         (LayoutTestResultsReader._create_unit_test_results):
1652         (LayoutTestResultsReader.results):
1653         (LayoutTestResultsReader.archive):
1654         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
1655         (LayoutTestResultsReaderTest.test_missing_layout_test_results):
1656         (LayoutTestResultsReaderTest.test_create_unit_test_results):
1657         (test_missing_unit_test_results_path):
1658         (test_layout_test_results):
1659         (test_archive_last_layout_test_results):
1660         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1661         (AbstractEarlyWarningSystem):
1662         (AbstractEarlyWarningSystem.__init__):
1663         (AbstractEarlyWarningSystem.begin_work_queue):
1664         (AbstractEarlyWarningSystem.run_command):
1665         * Scripts/webkitpy/tool/commands/queues.py:
1666         (AbstractPatchQueue.work_item_log_path):
1667         (PatchProcessingQueue):
1668         (PatchProcessingQueue._new_port_name_from_old):
1669         (PatchProcessingQueue.begin_work_queue):
1670         (PatchProcessingQueue._upload_results_archive_for_patch):
1671         (CommitQueue):
1672         (CommitQueue.begin_work_queue):
1673         (CommitQueue.run_command):
1674         (AbstractReviewQueue):
1675         (AbstractReviewQueue.__init__):
1676         (AbstractReviewQueue.begin_work_queue):
1677         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1678         (AbstractPatchQueueTest.test_next_patch):
1679         (PatchProcessingQueueTest):
1680         (PatchProcessingQueueTest.test_upload_results_archive_for_patch):
1681         * Scripts/webkitpy/tool/steps/runtests.py:
1682         (RunTests.run):
1683         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
1684
1685 2013-01-04  Dana Jansens  <danakj@chromium.org>
1686
1687         [chromium] Remove lost-compositor-context layout tests
1688         https://bugs.webkit.org/show_bug.cgi?id=106089
1689
1690         Reviewed by James Robinson.
1691
1692         Remove the loseCompositorContext() test function.
1693
1694         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1695         (WebTestRunner::TestRunner::TestRunner):
1696         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1697         (TestRunner):
1698
1699 2013-01-04  Adam Klein  <adamk@chromium.org>
1700
1701         Remove ENABLE_MUTATION_OBSERVERS #define
1702         https://bugs.webkit.org/show_bug.cgi?id=105459
1703
1704         Reviewed by Ryosuke Niwa.
1705
1706         * Scripts/webkitperl/FeatureList.pm:
1707         * qmake/mkspecs/features/features.pri:
1708
1709 2013-01-04  Ryosuke Niwa  <rniwa@webkit.org>
1710
1711         Statistics used in perftest_unittest.py and perftest_integrationtest.py are bogus
1712         https://bugs.webkit.org/show_bug.cgi?id=106078
1713
1714         Reviewed by Tony Chang.
1715
1716         This patch updates values for individual iterations in dummy performance tests used in
1717         perftest_unittest.py and perftest_integrationtest.py so that they match up with
1718         statistics such as stdev and median reported in those tests. This change is required
1719         to resolve the bug 97510.
1720
1721         This patch also refactors integration tests by extracting *TestData classes that
1722         centralizes the definitions of each dummy performance test's text (DRT output),
1723         runner output, and parsed results (JSON) so that we may share more data between
1724         python tests.
1725
1726         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1727         (MainTest._assert_results_are_correct): Extracted from test_parse_output.
1728         Use self.assertAlmostEqual when comparing the standard deviation.
1729         (MainTest.test_parse_output):
1730         (test_parse_output_with_failing_line):
1731         (test_parse_output_with_description): Use _assert_results_are_correct.
1732         (test_parse_output_with_subtests): Ditto.
1733
1734         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py:
1735         (InspectorPassTestData): Extracted from tests below.
1736         (EventTargetWrapperTestData): Ditto.
1737         (SomeParserTestData): Ditto.
1738         (MemoryTestData): Ditto.
1739         (TestDriver.run_test): Moved test definitions out of the method.
1740         (MainTest._normalize_output): Truncate standard deviations at 5th decimal point.
1741         (MainTest._load_output_json): Extracted from create_runner.
1742         (MainTest.create_runner):
1743         (MainTest.test_run_test_set_for_parser_tests):
1744         (MainTest.test_run_memory_test):
1745         (MainTest._test_run_with_json_output):
1746         (MainTest.test_run_with_json_output):
1747         (MainTest.test_run_with_description):
1748         (MainTest.test_run_generates_json_by_default):
1749         (MainTest.test_run_merges_output_by_default):
1750         (MainTest.test_run_respects_reset_results):
1751         (MainTest.test_run_generates_and_show_results_page):
1752         (MainTest.test_run_with_slave_config_json):
1753         (MainTest.test_run_with_multiple_repositories):
1754
1755 2013-01-04  Martin Robinson  <mrobinson@igalia.com>
1756
1757         [GTK] [WebKit2] Clicking on a word that is selected in an editable area deletes it
1758         https://bugs.webkit.org/show_bug.cgi?id=106045
1759
1760         Reviewed by Carlos Garcia Campos.
1761
1762         Update test results to confirm that compositions without preedits do not trigger
1763         a composition confirmation.
1764
1765         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
1766         (TestWebKitAPI::verifyCanceledComposition):
1767
1768 2013-01-04  Sailesh Agrawal  <sail@chromium.org>
1769
1770         [Mac] [WK2] platform/mac/fast/events/numpad-keycode-mapping.html fails
1771         https://bugs.webkit.org/show_bug.cgi?id=105958
1772
1773         Reviewed by Alexey Proskuryakov.
1774
1775         Ported code from DumpRenderTree to WebKitTestRunner to map all numpad keys.
1776
1777         * WebKitTestRunner/mac/EventSenderProxy.mm:
1778         (KeyMappingEntry):
1779         (WTR):
1780         (WTR::EventSenderProxy::keyDown): Added all numpad keys.
1781
1782 2013-01-03  Chris Fleizach  <cfleizach@apple.com>
1783
1784         REGRESSION: [Mac] Intermittent crash in WTR::AccessibilityUIElement::isEqual
1785         https://bugs.webkit.org/show_bug.cgi?id=106073
1786
1787         Reviewed by Ryosuke Niwa.
1788
1789         Check that the element being compared to is not nil. 
1790         This crash is happening because one of the accessibility tests does an isEqual comparison within a callback function.
1791         It looks like occasionally, one of the elements passed in to the isEqual is nil because the notification handler is on the accessibility
1792         controller, and the notification is being delivered after the document objects are cleared. This shouldn't happen but I've seen it 
1793         sometimes before on parallel WebKitTestRunner runs.
1794
1795         * DumpRenderTree/AccessibilityUIElement.cpp:
1796         (AccessibilityUIElement::isEqual):
1797         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
1798         (WTR::AccessibilityUIElement::isEqual):
1799
1800 2013-01-03  Tim 'mithro' Ansell  <mithro@mithis.com>
1801
1802         Downloads can be flaky, so try downloading multiple times.
1803         https://bugs.webkit.org/show_bug.cgi?id=105976
1804
1805         Reviewed by Eric Seidel.
1806
1807         * Scripts/webkitpy/common/system/autoinstall.py:
1808         (AutoInstaller._download_to_stream):
1809
1810 2013-01-03  Eric Seidel  <eric@webkit.org>
1811
1812         Rename Tool.port to deprecated_port in preparation for removing more callers
1813         https://bugs.webkit.org/show_bug.cgi?id=106076
1814
1815         Reviewed by Ryosuke Niwa.
1816
1817         This resolves a long-standing FIXME to rename this method. :)
1818
1819         * Scripts/webkitpy/tool/bot/botinfo.py:
1820         (BotInfo.summary_text):
1821         * Scripts/webkitpy/tool/bot/irc_command.py:
1822         (Rollout._update_working_copy):
1823         * Scripts/webkitpy/tool/bot/layouttestresultsreader.py:
1824         (LayoutTestResultsReader._create_layout_test_results):
1825         (LayoutTestResultsReader._create_unit_test_results):
1826         (LayoutTestResultsReader._results_directory):
1827         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
1828         (test_missing_unit_test_results_path):
1829         * Scripts/webkitpy/tool/main.py:
1830         (WebKitPatch.deprecated_port):
1831         * Scripts/webkitpy/tool/mocktool.py:
1832         (MockTool.deprecated_port):
1833         * Scripts/webkitpy/tool/steps/build.py:
1834         (Build.build):
1835         * Scripts/webkitpy/tool/steps/checkstyle.py:
1836         (CheckStyle.run):
1837         * Scripts/webkitpy/tool/steps/commit.py:
1838         (Commit._check_test_expectations):
1839         * Scripts/webkitpy/tool/steps/preparechangelog.py:
1840         (PrepareChangeLog.run):
1841         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
1842         (PrepareChangeLogForDEPSRoll.run):
1843         * Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
1844         (PrepareChangeLogForRevert.run):
1845         * Scripts/webkitpy/tool/steps/runtests.py:
1846         (RunTests.run):
1847         * Scripts/webkitpy/tool/steps/steps_unittest.py:
1848         (StepsTest.test_runtests_args):
1849         * Scripts/webkitpy/tool/steps/update.py:
1850         (Update._update_command):
1851
1852 2013-01-03  Eric Seidel  <eric@webkit.org>
1853
1854         Add myself to the webkitpy watchlist.
1855
1856         Unreviewed.
1857
1858         * Scripts/webkitpy/common/config/watchlist:
1859
1860 2013-01-03  Ryosuke Niwa  <rniwa@webkit.org>
1861
1862         Cleanup perftest* tests and add a test for computing statistics
1863         https://bugs.webkit.org/show_bug.cgi?id=105685
1864
1865         Reviewed by Eric Seidel.
1866
1867         Add a test for PerfTest.compute_statistics (moved and renamed from PageLoadingPerfTest.calculate_statistics) and
1868         extracted perftestsrunner_itegrationtests.py from perftestsrunner_unittests.py.
1869
1870         Also fixed a bug in compute_statistics that the mean ('avg') value can have a large rounding errors in some cases.
1871
1872         * Scripts/webkitpy/performance_tests/perftest.py:
1873         (PerfTest.compute_statistics): Moved from PageLoadingPerfTest to prepare for the bug 97510. Also compute the mean
1874         directly from sorted_values instead of using the one from Knuth's online algorithm. This approach gives more
1875         accurate result for the mean.
1876         (PageLoadingPerfTest.run_single):
1877         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1878         (MainTest.test_compute_statistics):
1879         (MainTest.test_compute_statistics.compute_statistics): Added.
1880         (TestPageLoadingPerfTest.test_run): floatify values.
1881         (TestPageLoadingPerfTest.test_run_with_memory_output): Ditto. Also got rid of ".0" from mean values now that Python
1882         correctly recognizes them as integers.
1883
1884         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py: Copied from
1885         Tools/Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py.
1886         (TestDriver): Moved out of MainTest.
1887         (MainTest): Got rid of assertWritten and all unit tests.
1888         (MainTest._normalize_output): Renamed from normalizeFinishedTime to match the PEP8 naming convention.
1889         (MainTest.test_run_test_set_kills_drt_per_run.TestDriverWithStopCount):
1890         (MainTest.test_run_test_set_for_parser_tests):
1891         (MainTest.test_run_memory_test):
1892         (MainTest._test_run_with_json_output):
1893         (MainTest.test_run_generates_json_by_default):
1894         (MainTest.test_run_merges_output_by_default):
1895         (MainTest.test_run_respects_reset_results):
1896         (MainTest.test_run_generates_and_show_results_page): Use runner.load_output_json() instead of manually loading and
1897         parsing output JSON files. Just verify that the output path is correct instead.
1898         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1899         (MainTest): Removed all integration tests.
1900         (MainTest.create_runner): Simplified.
1901
1902 2013-01-03  Julie Parent  <jparent@chromium.org>
1903
1904         Dashboard cleanup: remove usage of global g_defaultBuilderName
1905         https://bugs.webkit.org/show_bug.cgi?id=106043
1906
1907         Use now public defaultBuilder() from the BuilderGroup, rather
1908         than global variable.
1909
1910         Also eliminates the setup() and initBuilders() functions since we
1911         no longer need to track the global.
1912
1913         Reviewed by Adam Barth.
1914
1915         * TestResultServer/static-dashboards/builders.js:
1916         (BuilderGroup.prototype.defaultBuilder):
1917         (BuilderGroup.prototype.master):
1918         * TestResultServer/static-dashboards/dashboard_base.js:
1919         (parseParameters):
1920         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1921         (updateDefaultBuilderState):
1922         * TestResultServer/static-dashboards/loader.js:
1923         (.):
1924         * TestResultServer/static-dashboards/loader_unittests.js:
1925
1926 2013-01-03  Ryosuke Niwa  <rniwa@webkit.org>
1927
1928         Disable an intermittently failing unit test for now.
1929         The failure is tracked by the bug 106051.
1930
1931         * TestWebKitAPI/Tests/mac/MemoryCachePruneWithinResourceLoadDelegate.mm:
1932         (TestWebKitAPI::TEST):
1933
1934 2013-01-03  Dan Bernstein  <mitz@apple.com>
1935
1936         No way to obtain a DOMNode given a JS wrapper for a Node
1937         https://bugs.webkit.org/show_bug.cgi?id=106033
1938
1939         Added a test for +[DOMNode _nodeFromJSWrapper:].
1940
1941         Reviewed by Sam Weinig.
1942
1943         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1944         * TestWebKitAPI/Tests/mac/DOMNodeFromJSObject.mm: Added.
1945         (TestWebKitAPI):
1946         (TestWebKitAPI::TEST):
1947
1948 2013-01-03  Emil A Eklund  <eae@chromium.org>
1949
1950         Fix overflow in LayoutUnit::ceil and floor for SATURATED_LAYOUT_ARITHMETIC
1951         https://bugs.webkit.org/show_bug.cgi?id=105961
1952
1953         Reviewed by Levi Weintraub.
1954         
1955         Add tests for LayoutUnit::ceil and floor.
1956
1957         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
1958         (TestWebKitAPI::TEST):
1959         (TestWebKitAPI):
1960
1961 2013-01-03  Julie Parent  <jparent@chromium.org>
1962
1963         Add unit test for default builder for the dashboards.
1964         https://bugs.webkit.org/show_bug.cgi?id=105972
1965
1966         Reviewed by Dirk Pranke.
1967
1968         * TestResultServer/static-dashboards/loader_unittests.js:
1969
1970 2013-01-03  Tony Chang  <tony@chromium.org>
1971
1972         Generate internal.settings from Settings.in
1973         https://bugs.webkit.org/show_bug.cgi?id=104740
1974
1975         Reviewed by Adam Barth.
1976
1977         * GNUmakefile.am: Compile new generated files on GTK+.
1978         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
1979         with both extra_sources and add_output_to_sources=false will still
1980         compile the files in extra_sources. This makes it possible to compile
1981         InternalSettingsGenerated.cpp and not be confused by
1982         InternalSettingsGenerated.idl.
1983
1984 2013-01-03  Zan Dobersek  <zandobersek@gmail.com>
1985
1986         [style] WebIDL-reflecting uppercase enums reported as style violation
1987         https://bugs.webkit.org/show_bug.cgi?id=105832
1988
1989         Reviewed by Tony Chang.
1990
1991         Recognize the enum as WebIDL-reflecting one if there's a comment present that
1992         indicates this. In enums spanning over multiple lines, the comment is expected
1993         on the first line. In enums placed in one line, the comment is expected at the
1994         end of that line.
1995
1996         WebIDL-reflecting enum values are required to be written in uppercase and as such
1997         violate the common rule of these values being written in InterCaps. Annotating
1998         enums this way makes it possible to avoid false style violation reports.
1999
2000         The comment is of form '// Web(?:Kit)?IDL enum'.
2001
2002         * Scripts/webkitpy/style/checkers/cpp.py:
2003         (_EnumState.__init__):
2004         (_EnumState.process_clean_line):
2005         (check_enum_casing):
2006         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2007         (NoNonVirtualDestructorsTest.test_enum_casing):
2008
2009 2013-01-03  Tim 'mithro' Ansell  <mithro@mithis.com>
2010
2011         Adding favicon.ico to the garden-o-matic.
2012         https://bugs.webkit.org/show_bug.cgi?id=105982
2013
2014         Reviewed by Adam Barth.
2015
2016         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2017         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon.ico:
2018
2019 2013-01-03  Jonathan Liu  <net147@gmail.com>
2020
2021         [Qt] Fix library linking order
2022         https://bugs.webkit.org/show_bug.cgi?id=105818
2023
2024         Some Qt libraries are linked before the private static libraries
2025         that depend on them which results in QtWebKit failing to link.
2026
2027         Copy Qt module dependencies from QT to QT_PRIVATE when creating
2028         module to ensure the Qt libraries are linked after the private
2029         static libraries that depend on them.
2030
2031         Reviewed by Tor Arne Vestbø.
2032
2033         * qmake/mkspecs/features/default_post.prf:
2034
2035 2013-01-02  Byungwoo Lee  <bw80.lee@samsung.com>
2036
2037         [EFL][WK2] Fix tooltip bugs in MiniBrowser.
2038         https://bugs.webkit.org/show_bug.cgi?id=102209
2039
2040         Reviewed by Gyuyoung Kim.
2041
2042         Set tooltip on the elm_window instead of ewk_view, and add delay to
2043         show the tooltip window.
2044
2045         Setting tooltip on the ewk_view makes elementary error message because
2046         ewk_view is not an elementary widget. To remove the error message,
2047         tooltip should be set on the elm_window.
2048
2049         Most browsers show the tooltip after a mouse pointer stays on an
2050         element object that has tooltip text.
2051         But elm_tooltip widget itself doesn't have delay or handle mouse
2052         events to show/hide.
2053
2054         So additional codes are added to show/hide tooltip widget along with
2055         mouse events.
2056           1) When a mouse pointer stays for a second on an element which has
2057              tooltip text, then tooltip widget is displayed.
2058           2) When a mouse pointer is moved or out of ewk_view, hide the tooltip
2059              widget.
2060
2061         * MiniBrowser/efl/main.c:
2062         (_Tooltip_Information):
2063         (_Browser_Window):
2064         (on_tooltip_show):
2065         (window_tooltip_hide):
2066         (window_tooltip_update):
2067         (on_mouse_in):
2068         (on_mouse_move):
2069         (on_mouse_out):
2070         (window_free):
2071         (on_tooltip_text_set):
2072         (on_tooltip_text_unset):
2073         (window_create):
2074
2075 2013-01-02  Tony Chang  <tony@chromium.org>
2076
2077         Unreviewed, rolling out r138661.
2078         http://trac.webkit.org/changeset/138661
2079         https://bugs.webkit.org/show_bug.cgi?id=104740
2080
2081         Compile problems on EFL
2082
2083         * GNUmakefile.am:
2084         * qmake/mkspecs/features/default_post.prf:
2085
2086 2013-01-02  Tony Chang  <tony@chromium.org>
2087
2088         Generate internal.settings from Settings.in
2089         https://bugs.webkit.org/show_bug.cgi?id=104740
2090
2091         Reviewed by Adam Barth.
2092
2093         * GNUmakefile.am: Compile new generated files on GTK+.
2094         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
2095         with both extra_sources and add_output_to_sources=false will still
2096         compile the files in extra_sources. This makes it possible to compile
2097         InternalSettingsGenerated.cpp and not be confused by
2098         InternalSettingsGenerated.idl.
2099
2100 2013-01-02  Adam Barth  <abarth@webkit.org>
2101
2102         Remove old-run-webkit-tests support from webkitpy
2103         https://bugs.webkit.org/show_bug.cgi?id=105619
2104
2105         Reviewed by Eric Seidel.
2106
2107         This patch removes support for webkitpy parsing the
2108         old-run-webkit-tests results format. There doesn't seem to be any
2109         reason to keep this code around anymore now that most bots have
2110         switched to new-run-webkit-tests.
2111
2112         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2113         (Builder.fetch_layout_test_results):
2114         * Scripts/webkitpy/common/net/layouttestresults.py:
2115         (path_for_layout_test):
2116         (LayoutTestResults.results_from_string):
2117         (LayoutTestResults):
2118         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
2119         (LayoutTestResultsTest.test_results_from_string):
2120         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
2121         (test_flaky_test_failure):
2122         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
2123         (LayoutTestResultsReaderTest.test_missing_layout_test_results):
2124         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
2125         (PatchAnalysisTask._test_patch):
2126         * Scripts/webkitpy/tool/commands/queuestest.py:
2127         (QueuesTest.assert_queue_outputs):
2128
2129 2013-01-02  Heikki Paajanen  <heikki.paajanen@palm.com>
2130
2131         [Qt][WK2] Add simple UI to find text from page
2132         https://bugs.webkit.org/show_bug.cgi?id=96481
2133
2134         Reviewed by Jocelyn Turcotte.
2135
2136         Find bar that slides between navigation bar and WebView.
2137
2138         * MiniBrowser/qt/BrowserWindow.cpp:
2139         (BrowserWindow::toggleFind):
2140         * MiniBrowser/qt/BrowserWindow.h:
2141         (BrowserWindow):
2142         * MiniBrowser/qt/MiniBrowser.qrc:
2143         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2144         (MiniBrowserApplication::notify):
2145         * MiniBrowser/qt/icons/find.png: Added.
2146         * MiniBrowser/qt/qml/BrowserWindow.qml:
2147
2148 2013-01-02  Christophe Dumez  <christophe.dumez@intel.com>
2149
2150         Regression(r138603): Caused 2 webkitpy tests to fail
2151         https://bugs.webkit.org/show_bug.cgi?id=105913
2152
2153         Reviewed by Kentaro Hara.
2154
2155         Fix 2 webkitpy tests that started to fail after
2156         r138603.
2157
2158         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
2159         (IRCCommandTest._sheriff_test_data_url): Make sure the
2160         returned test file URL is valid no matter what the
2161         current working directory is.
2162         * Scripts/webkitpy/tool/bot/ircbot_unittest.py:
2163         (IRCBotTest.test_help): Add "sheriffs" command to the
2164         list of available commands.
2165
2166 2013-01-01  Christophe Dumez  <christophe.dumez@intel.com>
2167
2168         [WK2] plugins/clicking-missing-plugin-fires-delegate.html is failing
2169         https://bugs.webkit.org/show_bug.cgi?id=105897
2170
2171         Reviewed by Alexey Proskuryakov.
2172
2173         Dump in WebKitTestRunner the message that is expected by layout tests
2174         when the missing plugin button is clicked.
2175
2176         * WebKitTestRunner/TestController.cpp:
2177         (WTR::TestController::createWebViewWithOptions):
2178         (WTR::TestController::unavailablePluginButtonClicked):
2179         (WTR):
2180         * WebKitTestRunner/TestController.h:
2181         (TestController):
2182
2183 2013-01-01  KwangYong Choi  <ky0.choi@samsung.com>
2184
2185         [EFL] Enable MHTML feature
2186         https://bugs.webkit.org/show_bug.cgi?id=105815
2187
2188         Reviewed by Laszlo Gombos.
2189
2190         Enable MHTML feature for EFL port by default.
2191
2192         * Scripts/webkitperl/FeatureList.pm:
2193
2194 2013-01-01  Alan Cutter  <alancutter@chromium.org>
2195
2196         sheriff-bot should know who the gardeners/sheriffs are
2197         https://bugs.webkit.org/show_bug.cgi?id=105698
2198
2199         Reviewed by Eric Seidel.
2200
2201         Added a "sheriffs" command to sheriffbot to fetch the current Chromium Webkit sheriffs.
2202
2203         * Scripts/webkitpy/tool/bot/irc_command.py:
2204         (Current):
2205         (Current._retrieve_webkit_sheriffs):
2206         (Current.execute):
2207         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
2208         (IRCCommandTest.test_current):
2209         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_0.js: Added.
2210         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_1.js: Added.
2211         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_2.js: Added.
2212         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_malformed.js: Added.
2213         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_zero.js: Added.
2214
2215 2012-12-31  Carlos Garcia Campos  <cgarcia@igalia.com>
2216
2217         [GTK] Move ImageDiff to the Tools directory root
2218         https://bugs.webkit.org/show_bug.cgi?id=105421
2219
2220         Reviewed by Kenneth Rohde Christiansen.
2221
2222         * DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Removed.
2223         * GNUmakefile.am:
2224         * ImageDiff/gtk/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/gtk/ImageDiff.cpp.
2225
2226 2012-12-30  Mike West  <mkwst@chromium.org>
2227
2228         Add myself to watchlists for CSP and console logs.
2229         https://bugs.webkit.org/show_bug.cgi?id=105874
2230
2231         Reviewed by Dirk Pranke.
2232
2233         Because I really, really, really need more email.
2234
2235         * Scripts/webkitpy/common/config/committers.py:
2236             Add mkwst+watchlist@chromium.org as a contributor.
2237         * Scripts/webkitpy/common/config/watchlist:
2238             Create watchlist categories for 1) the various mechanisms used to
2239             generate console messages, and 2) CSP usage and files.
2240
2241 2012-12-30  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2242
2243         [EFL] Enable TEMPLATE_ELEMENT feature
2244         https://bugs.webkit.org/show_bug.cgi?id=105865
2245         
2246         Reviewed by Laszlo Gombos.
2247
2248         * Scripts/webkitperl/FeatureList.pm: Enable template element feature for EFL port.
2249
2250 2012-12-30  Zan Dobersek  <zandobersek@gmail.com>
2251
2252         Unreviewed GTK gardening of the unit tests.
2253
2254         Skip the TestInspectorServer unit test that's timing out completely.
2255
2256         * Scripts/run-gtk-tests:
2257         (TestRunner):
2258
2259 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
2260
2261         Web Inspector: Native Memory Instrumentation plugin: do not generate 'not instrumented' warning for instrumented mutable members.
2262         https://bugs.webkit.org/show_bug.cgi?id=105855
2263
2264         Reviewed by Vsevolod Vlasov.
2265
2266         Extract MemberExpr from ImplicitCastExpr. It happens when we instrument a mutable member because
2267         addMember expects const T& and the mutable ref to member implicitly converts into const ref to member.
2268
2269         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
2270         (ReportMemoryUsageVisitor):
2271         (clang::AddMemberCallVisitor::VisitCallExpr):
2272         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
2273         (clang::ReportMemoryUsageVisitor::checkMembersCoverage):
2274
2275 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
2276
2277         Web Inspector: Native Memory Instrumentation plugin: move function bodies out of class declarations.
2278         https://bugs.webkit.org/show_bug.cgi?id=105852
2279
2280         Reviewed by Alexander Pavlov.
2281
2282         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
2283         (clang):
2284         (AddMemberCallVisitor):
2285         (ReportMemoryUsageVisitor):
2286         (clang::ReportMemoryUsageAction::ParseArgs):
2287         (clang::AddMemberCallVisitor::VisitCallExpr):
2288         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
2289         (clang::ReportMemoryUsageVisitor::emitWarning):
2290         (clang::ReportMemoryUsageVisitor::findInstrumentationMethod):
2291         (clang::ReportMemoryUsageVisitor::needsToBeInstrumented):
2292         (clang::ReportMemoryUsageVisitor::CheckMembersCoverage):
2293
2294 2012-12-29  Zan Dobersek  <zandobersek@gmail.com>
2295
2296         [webkitpy] Omit webkitpy/thirdparty/BeautifulSoup.py from code coverage checking
2297         https://bugs.webkit.org/show_bug.cgi?id=105835
2298
2299         Reviewed by Dirk Pranke.
2300
2301         The BeautifulSoup module is a third-party import and as such we don't have much control
2302         over its testing. Because of that it should be omitted from code coverage checking.
2303
2304         * Scripts/webkitpy/test/main.py:
2305         (Tester._run_tests):
2306
2307 2012-12-28  Martin Robinson  <mrobinson@igalia.com>
2308
2309         [GTK][WK2] Add support for IME Composition
2310         https://bugs.webkit.org/show_bug.cgi?id=65093
2311
2312         Reviewed by Carlos Garcia Campos.
2313
2314         Update unit tests for GtkInputMethodFilter to note whether or not an event
2315         was a fake event synthesized for composition purposes. This is now used
2316         for WebKit2.
2317
2318         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
2319         (TestWebKitAPI::TestInputMethodFilter::sendSimpleKeyEvent):
2320         (TestWebKitAPI::TestInputMethodFilter::sendKeyEventWithCompositionResults):
2321         (TestWebKitAPI::TEST):
2322         * gtk/generate-gtkdoc:
2323         (get_webkit2_options): Skip new files in the WebKit2 source directory for documentation
2324         generation.
2325
2326 2012-12-27  Ilya Tikhonovsky  <loislo@chromium.org>
2327
2328         Web Inspector: Native Memory Instrumentation. Update clang plugin according to the current state of memory instrumentation code.
2329         https://bugs.webkit.org/show_bug.cgi?id=105800
2330
2331         Reviewed by Yury Semikhatsky.
2332
2333         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
2334         (clang::ReportMemoryUsageConsumer::ReportMemoryUsageConsumer):
2335         (clang):
2336
2337 2012-12-27  Zan Dobersek  <zandobersek@gmail.com>
2338
2339         Create a GTK build system watchlist and add myself to it
2340         https://bugs.webkit.org/show_bug.cgi?id=105769
2341
2342         Reviewed by Dirk Pranke.
2343
2344         Add a GTK build system watchlist, watching the configure.ac, GNUmakefile.am files
2345         and the input GNUmakefile.features.am.in file. The GNUmakefile.list.am files are
2346         omitted on purpose due to relatively high change rate they are exposed to.
2347
2348         The new watchlist would be of use to me as the GTK build system recently went
2349         through a series of changes that included a changed policy on adding configuration
2350         flags to configure.ac and subsequent additions to GNUmakefile.am files. A PSA on the
2351         mailing list about this change will be published later, but I'd also like to be
2352         notified of any pending changes to these files just so I can check the changes are
2353         in compliance with the new policy.
2354
2355         * Scripts/webkitpy/common/config/watchlist:
2356
2357 2012-12-26  Ilya Tikhonovsky  <loislo@chromium.org>
2358
2359         Unreviewed. Replace find("Tools") with rfind("Tools") because base dir of WebKit may have Tools word.
2360         As example ..../DevTools/src/third_party/WebKit
2361
2362         * Scripts/webkitpy/common/webkit_finder.py:
2363         (WebKitFinder.webkit_base):
2364
2365 2012-12-25  Christophe Dumez  <christophe.dumez@intel.com>
2366
2367         [EFL][WK2] Refactor snapshot taking code
2368         https://bugs.webkit.org/show_bug.cgi?id=105687
2369
2370         Reviewed by Kenneth Rohde Christiansen.
2371
2372         Call WKViewCreateSnapshot() instead of WKViewGetSnapshot()
2373         since the function was renamed to follow naming style.
2374
2375         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
2376         (WTR::PlatformWebView::windowSnapshotImage):
2377
2378 2012-12-24  Ilya Tikhonovsky  <loislo@chromium.org>
2379
2380         Web Inspector: Native Memory Instrumentation: propagate member type as edge type to the serialized heap graph.
2381         https://bugs.webkit.org/show_bug.cgi?id=105725
2382
2383         Reviewed by Yury Semikhatsky.
2384
2385         MemberType value names were adjusted according to Style Guide.
2386         Existing tests were extended with link type validation.
2387
2388         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2389
2390 2012-12-24  Xiaobo Wang  <xbwang@torchmobile.com.cn>
2391
2392         [BlackBerry] DRT - Crashed when running pixels tests
2393         https://bugs.webkit.org/show_bug.cgi?id=105711
2394
2395         RIM PR 270003
2396
2397         Reviewed by George Staikos.
2398
2399         Calling window->post() here will result in buffer swapping which is
2400         only allowed in UI thread now, and I found we don't need to call it
2401         for GL renderer.
2402
2403         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
2404         (createBitmapContextFromWebView):
2405
2406 2012-12-22  Dan Winship  <danw@gnome.org>
2407
2408         Add a patch to fix libxml2 2.9.0 regression
2409         https://bugs.webkit.org/show_bug.cgi?id=104494
2410
2411         Reviewed by Martin Robinson.
2412
2413         * gtk/jhbuild.modules:
2414         * gtk/patches/libxml2-2.9.0-dtd.patch: patch from upstream
2415
2416 2012-12-23  Allan Sandfeld Jensen  <allan.jensen@digia.com>
2417
2418         Unreviewed update of status to reviewer.
2419
2420         * Scripts/webkitpy/common/config/committers.py:
2421
2422 2012-12-23  Philippe Normand  <pnormand@igalia.com>
2423
2424         [GTK][jhbuild] Bump GStreamer to 1.0.4
2425         https://bugs.webkit.org/show_bug.cgi?id=105640
2426
2427         Reviewed by Martin Robinson.
2428
2429         * gtk/jhbuild.modules: Update GStreamer modules to 1.0.4.
2430
2431 2012-12-21  Nico Weber  <thakis@chromium.org>
2432
2433         chromium: webkit-build-directory doesn't handle ninja output dirs
2434         https://bugs.webkit.org/show_bug.cgi?id=105671
2435
2436         Reviewed by Daniel Bates.
2437
2438         Teach webkit-build-directory to detect the ninja build directory.
2439
2440         * Scripts/webkitdirs.pm:
2441         (determineBaseProductDir):
2442         (determineIsChromiumNinja):
2443
2444 2012-12-21  Ryosuke Niwa  <rniwa@webkit.org>
2445
2446         [Mountain Lion] platform/mac/editing/spelling/removing-underline-after-accepting-autocorrection-using-punctuation.html failing
2447         https://bugs.webkit.org/show_bug.cgi?id=95477
2448
2449         Reviewed by Sam Weinig.
2450
2451         Co-authored by Jia Pu.
2452
2453         Specify a fixed autocorrection dictionary in resetDefaultsToConsistentValues() so that
2454         we can have consistent autocorrection behavior when running tests.
2455
2456         * DumpRenderTree/mac/DumpRenderTree.mm:
2457         (resetDefaultsToConsistentValues):
2458         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
2459         (WTR::InjectedBundle::platformInitialize):
2460
2461 2012-12-21  Zan Dobersek  <zandobersek@gmail.com>
2462
2463         [GTK] Add WebKit1-specific TestExpectations file
2464         https://bugs.webkit.org/show_bug.cgi?id=97562
2465
2466         Reviewed by Dirk Pranke.
2467
2468         Implement the expectations_files method inside the GtkPort class that
2469         returns the list of the TestExpectations files the current test configuration
2470         requires. The method is implemented in a similar way as on the EFL port.
2471         A unit test is also added.
2472
2473         * Scripts/webkitpy/layout_tests/port/gtk.py:
2474         (GtkPort._search_paths):
2475         (GtkPort):
2476         (GtkPort.expectations_files):
2477         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
2478         (GtkPortTest.test_expectations_files):
2479
2480 2012-12-21  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
2481
2482         [EFL] Enable feature option for the CSS sticky position
2483         https://bugs.webkit.org/show_bug.cgi?id=105630
2484
2485         Reviewed by Kenneth Rohde Christiansen.
2486
2487         Enable CSS sticky position feature by default for the
2488         EFL port. This feature option was added in r138252.
2489
2490         No new tests. Existing layout tests pass.
2491
2492         * Scripts/webkitperl/FeatureList.pm:
2493
2494 2012-12-21  Jussi Kukkonen  <jussi.kukkonen@intel.com>
2495
2496         [WK2] REGRESSION(r138232): It made 4 tests timeout
2497         https://bugs.webkit.org/show_bug.cgi?id=105623
2498
2499         Reviewed by Kenneth Rohde Christiansen.
2500
2501         r138232 changed functionality so that empty text output
2502         no longer leads to a dump() call, which leads to a timeout.
2503         Instead do a text dump if there is text output or if there
2504         is no audio output.
2505
2506         * WebKitTestRunner/TestInvocation.cpp:
2507         (WTR::TestInvocation::dumpResults):
2508
2509 2012-12-21  Zeno Albisser  <zeno@webkit.org>
2510
2511         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
2512         https://bugs.webkit.org/show_bug.cgi?id=105309
2513
2514         On Windows the library name is Qt5WebKitWidgets as well.
2515         So r138136 should not have adjusted the library name for Windows.
2516
2517         Reviewed by Simon Hausmann.
2518
2519         * Scripts/webkitdirs.pm:
2520         (builtDylibPathForName):
2521
2522 2012-12-21  Zeno Albisser  <zeno@webkit.org>
2523
2524         [Qt] Add new TestExpectations for qt-5.0-mac-wk2.
2525         https://bugs.webkit.org/show_bug.cgi?id=105424
2526
2527         This change will allow us to run a very specific
2528         subset of the available LayoutTests on a mac buildbot.
2529         The new baseline search path for mac will look as follows:
2530         qt-5.0-mac-wk2 -> qt-5.0-wk2 -> qt-5.0 -> qt-mac -> qt
2531
2532         Reviewed by Dirk Pranke.
2533
2534         * Scripts/webkitpy/layout_tests/port/qt.py:
2535         (QtPort._search_paths):
2536         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2537         (QtPortTest):
2538
2539 2012-12-20  Ryuan Choi  <ryuan.choi@samsung.com>
2540
2541         [EFL] Build break with latest EFL libraries.
2542         https://bugs.webkit.org/show_bug.cgi?id=104827
2543
2544         Reviewed by Laszlo Gombos.
2545
2546         The eo EFL package is introduced and evas and ecore use it since 1.8.
2547         While introducing Eo, EFL changed several structures of Evas and Ecore
2548         from own specific class to Eo.
2549
2550         So, this patch adds FindEo and changes declaration of Evas, Evas_Object,
2551         Ecore_Timer to build with latest EFL libraries.
2552
2553         * DumpRenderTree/efl/CMakeLists.txt: Includes eo EFL library.
2554         * EWebLauncher/CMakeLists.txt: Ditto.
2555         * MiniBrowser/efl/CMakeLists.txt: Ditto.
2556         * ImageDiff/PlatformEfl.cmake: Ditto.
2557         * TestWebKitAPI/PlatformEfl.cmake: Ditto.
2558         * TestWebKitAPI/PlatformWebView.h: Modified declaration of Evas_Object.
2559         * WebKitTestRunner/InjectedBundle/TestRunner.h: Modified declaration of Ecore_Timer.
2560         * WebKitTestRunner/PlatformEfl.cmake: Includes eo EFL library.
2561         * WebKitTestRunner/PlatformWebView.h: Modified declaration of Evas_Object.
2562
2563 2012-12-20  Dirk Pranke  <dpranke@chromium.org>
2564
2565         TestExpectation fallback is broken
2566         https://bugs.webkit.org/show_bug.cgi?id=105583
2567
2568         Reviewed by Ryosuke Niwa.
2569
2570         When I added support for the -wk2 in port names, it looks
2571         like I broke the way we computed the list of TestExpectations
2572         files so that the implementation-version directory was left out :(.
2573         Fixed and cleaned up the tests.
2574
2575         * Scripts/webkitpy/layout_tests/port/base.py:
2576         (Port.expectations_files):
2577         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2578         (TestWebKitPort.__init__):
2579         (PortTestCase.test_skipped_directories_for_symbols):
2580         (test_skipped_directories_for_features):
2581         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
2582         (test_skipped_tests_for_unsupported_features_empty_test_list):
2583         (test_skipped_layout_tests):
2584         (test_expectations_files):
2585
2586 2012-12-20  Anders Carlsson  <andersca@apple.com>
2587
2588         Don't include WebKit2 headers when building TestWebKitAPI on windows
2589         https://bugs.webkit.org/show_bug.cgi?id=105582
2590
2591         Reviewed by Tim Horton.
2592
2593         * TestWebKitAPI/config.h:
2594         * TestWebKitAPI/win/main.cpp:
2595
2596 2012-12-20  Ryosuke Niwa  <rniwa@webkit.org>
2597
2598         REGRESSION(r133820?): SimplifyMarkupTest API test asserts
2599         https://bugs.webkit.org/show_bug.cgi?id=105370
2600
2601         Reviewed by Simon Fraser.
2602
2603         Re-enable the test.
2604
2605         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
2606         (TestWebKitAPI::TEST):
2607
2608 2012-12-20  Nico Weber  <thakis@chromium.org>
2609
2610         chromium nrwt: Pick the newest binary found in DEFAULT_BUILD_DIRECTORIES, not the first
2611         https://bugs.webkit.org/show_bug.cgi?id=105498
2612
2613         Reviewed by Dirk Pranke.
2614
2615         Use the newest binary available rather than an than always picking one
2616         build directory over another based on iteration order.
2617
2618         * Scripts/webkitpy/layout_tests/port/chromium.py:
2619         (ChromiumPort._static_build_path):
2620         Check for timestamps.
2621         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
2622         (ChromiumMacPortTest.test_build_path_timestamps):
2623         Test that out / xcodebuild selection happens based on timestamps
2624         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2625         (ChromiumWinPortTest.test_build_path_timestamps):
2626         Test that out / build selection happens based on timestamps
2627
2628 2012-12-19  Simon Fraser  <simon.fraser@apple.com>
2629
2630         Lots of sticky tests failing in WK2
2631         https://bugs.webkit.org/show_bug.cgi?id=105464
2632
2633         Reviewed by Beth Dakin.
2634
2635         WebKitTestRunner had a race between snapshotting in the UI process,
2636         and resettting after the test in the web process. InjectedBundle::done()
2637         was a bad place to call page()->resetAfterTest(), because of this race;
2638         it could reset the scroll position before the UI snapshot had been obtained.
2639         
2640         Fix by moving the call to page()->resetAfterTest() into didReceiveMessage(),
2641         for the "Reset" message which will come in before the next test.
2642
2643         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2644         (WTR::InjectedBundle::didReceiveMessage):
2645         (WTR::InjectedBundle::done):
2646
2647 2012-12-19  Filip Pizlo  <fpizlo@apple.com>
2648
2649         DFG speculation checks that take JumpList should consolidate OSRExits
2650         https://bugs.webkit.org/show_bug.cgi?id=105401
2651
2652         Reviewed by Oliver Hunt.
2653
2654         Update the profiler to understand that an OSR exit may have multiple
2655         candidate exit sites.
2656
2657         * Scripts/display-profiler-output:
2658
2659 2012-12-20  Eric Seidel  <eric@webkit.org>
2660
2661         EWSes do not need to dump build/test logs when things pass
2662         https://bugs.webkit.org/show_bug.cgi?id=105402
2663
2664         Reviewed by Dirk Pranke.
2665
2666         When we were first writing the commit-queue and EWS system
2667         and running them locally, we thought it was important to see the
2668         build/test output as it happened.  Now these queues have been running
2669         for years, and we don't use this feature anymore.  So this patch
2670         makes us stop tee-ing the output directly to stdout, and instead
2671         only print any output from commands when they fail.
2672
2673         I also added some logging when we execute each command
2674         since we are no longer printing the command output itself.
2675         This should make it much easier to see errors with the EWS
2676         system itself, now that all the passing-build output is removed.
2677
2678         This required updating a zillion unittests, since we have a bunch of
2679         "integration" style testing of the queue-bots.  Many of the unittests for the bots were
2680         already approximating this behavior by using a "logging" Executive.
2681         I disabled logging in those cases to avoid redundant logging in those tests.
2682
2683         * Scripts/webkitpy/common/system/executive.py:
2684         (Executive.command_for_printing):
2685         (Executive.run_command):
2686         * Scripts/webkitpy/common/system/executive_mock.py:
2687         (MockExecutive.command_for_printing):
2688         (MockExecutive.run_command):
2689         * Scripts/webkitpy/common/system/executive_unittest.py:
2690         (ExecutiveTest.test_auto_stringify_args):
2691         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2692         (EarlyWarningSytemTest._default_expected_logs):
2693         (_test_ews):
2694         (_test_ewses):
2695         * Scripts/webkitpy/tool/commands/queues.py:
2696         (AbstractQueue.run_webkit_patch):
2697         * Scripts/webkitpy/tool/commands/queues_unittest.py:
2698         (AbstractQueueTest._assert_run_webkit_patch):
2699         (CommitQueueTest.test_commit_queue):
2700         (test_rollout):
2701         (test_rollout_lands):
2702         (test_manual_reject_during_processing):
2703         (StyleQueueTest.test_style_queue_with_style_exception):
2704         (test_style_queue_with_watch_list_exception):
2705
2706 2012-12-20  Dominic Mazzoni  <dmazzoni@google.com>
2707
2708         AX: support clickPoint in DRT for chromium
2709         https://bugs.webkit.org/show_bug.cgi?id=97359
2710
2711         Reviewed by Chris Fleizach.
2712
2713         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
2714         several more tests can pass.
2715
2716         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2717         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2718         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
2719         (WebTestRunner):
2720         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
2721         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
2722         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2723         (AccessibilityUIElement):
2724
2725 2012-12-20  Dan Bernstein  <mitz@apple.com>
2726
2727         prepare-ChangeLog adds "Additional information of the change such as approach, rationale" where it doesn't make sense
2728         https://bugs.webkit.org/show_bug.cgi?id=99268
2729
2730         Reviewed by Alexey Proskuryakov.
2731
2732         * Scripts/prepare-ChangeLog:
2733         (generateNewChangeLogs): Removed the "Additional information of the change" line.
2734
2735 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
2736
2737         Unreviewed. Skip unit test /webkit/atk/getTextInParagraphAndBodyModerate
2738
2739         * Scripts/run-gtk-tests:
2740         (TestRunner):
2741
2742 2012-12-20  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2743
2744         [Qt] Move ImageDiff to the Tools directory root
2745         https://bugs.webkit.org/show_bug.cgi?id=105420
2746
2747         Reviewed by Tor Arne Vestbø.
2748
2749         Moved ImageDiff from outside DumpRenderTree because it is also a
2750         dependency when running pixel tests for WebKit2.
2751
2752         * ImageDiff/ImageDiff.pro: Renamed from Tools/DumpRenderTree/qt/ImageDiff.pro.
2753         * ImageDiff/qt/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/qt/ImageDiff.cpp.
2754         * Tools.pro:
2755
2756 2012-12-20  Zan Dobersek  <zandobersek@gmail.com>
2757
2758         [GTK] Remove the --enable-unstable-features configuration option
2759         https://bugs.webkit.org/show_bug.cgi?id=105327
2760
2761         Reviewed by Martin Robinson.
2762
2763         Add a missing feature option for the CSS sticky position feature.
2764
2765         * Scripts/webkitperl/FeatureList.pm:
2766
2767 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
2768
2769         Unreviewed. Skip unit test WebKit2.ForceRepaint that times out
2770
2771         * Scripts/run-gtk-tests:
2772         (TestRunner):
2773
2774 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
2775
2776         [EFL] MiniBrowser does not play Infinite Gangnam Style
2777         https://bugs.webkit.org/show_bug.cgi?id=103531
2778
2779         Reviewed by Kenneth Rohde Christiansen.
2780
2781         For EFL we would like this feature to be built by default.
2782
2783         * Scripts/webkitperl/FeatureList.pm:
2784
2785 2012-12-20  Csaba Osztrogonác  <ossy@webkit.org>
2786
2787         [Qt] EWS bots should call qmake always
2788         https://bugs.webkit.org/show_bug.cgi?id=90049
2789
2790         Reviewed by Simon Hausmann.
2791
2792         * Scripts/webkitdirs.pm:
2793         (buildQMakeProjects):
2794
2795 2012-12-20  Jochen Eisinger  <jochen@chromium.org>
2796
2797         [chromium] fix stray printf in TestRunner's WebTestProxy
2798         https://bugs.webkit.org/show_bug.cgi?id=105515
2799
2800         Reviewed by Ryosuke Niwa.
2801
2802         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
2803         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
2804
2805 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
2806
2807         [WK2] WebAudio WKTR support
2808         https://bugs.webkit.org/show_bug.cgi?id=95084
2809
2810         Reviewed by Kenneth Rohde Christiansen.
2811
2812         Implementing audio data dump support for WebKitTestRunner,
2813         by adding the API to the IDL, enabling IPC for transferring
2814         the audio data as bytebuffer and dumping it according
2815         to run-webkit-tests script's expectations.
2816         This enables audio testing on WebKit2 ports.
2817         Based on initial wip-patch from Philippe Normand - thanks.
2818
2819         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Adding setAudioData API.
2820         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Assign audio buffer data to a local WKDataRef after extracting it from WK2 using WKBundleDataFromUint8Array bundle API.
2821         (WTR::TestRunner::setAudioData):
2822         (WTR):
2823         * WebKitTestRunner/InjectedBundle/TestRunner.h: Add a whatToDump state for dumping Audio.
2824         (TestRunner):
2825         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Add audio to the message that's transferred to the WKTR process.
2826         (WTR::InjectedBundle::done):
2827         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: Allow setting and audio result handle to be transferred.
2828         (WTR::InjectedBundle::setAudioResult):
2829         (InjectedBundle):
2830         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2831         (WTR::InjectedBundlePage::dump): Dummy handling the Audio case in the switch, build fix for Mac.
2832         * WebKitTestRunner/TestInvocation.cpp: Audio dump support, dump audio if we haven't received a text dump from the web process.
2833         (WTR::TestInvocation::dumpResults):
2834         (WTR):
2835         (WTR::TestInvocation::dumpAudio):
2836         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Storing received audio data before dumping.
2837         * WebKitTestRunner/TestInvocation.h:
2838         (TestInvocation):
2839
2840 2012-12-20  Alberto Garcia  <agarcia@igalia.com>
2841
2842         [GTK] Count failed and timed out tests separately
2843         https://bugs.webkit.org/show_bug.cgi?id=105448
2844
2845         Reviewed by Carlos Garcia Campos.
2846
2847         Display the number of tests that fail, time out and are skipped
2848         separately in the buildbot page.
2849
2850         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2851         (RunGtkAPITests.commandComplete):
2852         (RunGtkAPITests.evaluateCommand):
2853         (RunGtkAPITests.getText2):
2854         * Scripts/run-gtk-tests:
2855         (TestRunner.run_tests):
2856
2857 2012-12-19  Ryosuke Niwa  <rniwa@webkit.org>
2858
2859         PerfTest.parse_output does too much
2860         https://bugs.webkit.org/show_bug.cgi?id=105391
2861
2862         Reviewed by Tony Chang.
2863
2864         Moved the code to filter lines into PerfTest._filter_output, which has been renamed from _filter_stderr.
2865         Also moved the code to output test results into PerfTest._run_with_driver.
2866
2867         * Scripts/webkitpy/performance_tests/perftest.py:
2868         (PerfTest.__init__): Added _description.
2869         (PerfTest.description): Added.
2870         (PerfTest._run_with_driver): Moved the code to output test description and test results.
2871         (PerfTest._filter_output): Moved and renamed from PerfTest._filter_stderr.
2872         (PerfTest.parse_output): Removed the code to output test results. Return test results and description
2873         respectively. Also removed the code that allowed some tests to omit values since all tests report each
2874         iteration now since r136492.
2875         (PerfTest.output_statistics): Removed the code to print test description, now done in _run_with_driver.
2876         (ChromiumStylePerfTest._run_with_driver): Added. Chromium style tests are sufficiently different from
2877         regular PerfTest that it doesn't make much sense to share _run_with_driver. But really, we should just
2878         get rid of this type of test altogether in favor of regular performance tests that uses runner.js.
2879         (ChromiumStylePerfTest.parse_and_log_output): Renamed from parse_output.
2880         (PageLoadingPerfTest._run_with_driver): Removed the explicit '' for the test description.
2881         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
2882         (MainTest.test_parse_output): Removed the expected logs since parse_output no longer prints out results.
2883         Also added a call to _filter_output since parse_output doesn't filter the output text anymore.
2884         (MainTest.test_parse_output_with_failing_line): Added a call to _filter_output. Also added ',' after 'Time:'
2885         so that it's not string-concatenated with the next line.
2886         (MainTest.test_parse_output_with_description): Added; a test for PerfTest.description().
2887         (MainTest.test_parse_output_with_subtests): Removed the expected logs, and added a call to _filter_output.
2888
2889 2012-12-19  Alexis Menard  <alexis@webkit.org>
2890
2891         Implement CSS parsing for CSS transitions unprefixed.
2892         https://bugs.webkit.org/show_bug.cgi?id=104804
2893
2894         Reviewed by Dean Jackson.
2895
2896         Add a new flag ENABLE_CSS_TRANSFORMS_ANIMATIONS_TRANSITIONS_UNPREFIXED
2897         to cover the work of unprefixing Transforms, Animations and 
2898         Transitions. It will let the possibility of each ports to turn it off 
2899         in their release branches until we're confident that these CSS 
2900         properties are ready to be unprefixed.
2901
2902         * Scripts/webkitperl/FeatureList.pm:
2903         * qmake/mkspecs/features/features.pri:
2904
2905 2012-12-19  Julien Chaffraix  <jchaffraix@webkit.org>
2906
2907         2 clampTo unit tests are failing
2908         https://bugs.webkit.org/show_bug.cgi?id=105253
2909
2910         Reviewed by Emil A Eklund.
2911
2912         Removed clampToUnsigned as it was actually testing exactly what
2913         clampToUnsignedLong was doing. Also made clampToIntFloat more robust
2914         to precision loss (fix suggested by Emil).
2915
2916         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
2917
2918 2012-12-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2919
2920         [EFL] Move ImageDiff to the Tools directory root
2921         https://bugs.webkit.org/show_bug.cgi?id=105305
2922
2923         Reviewed by Laszlo Gombos.
2924
2925         We previously moved this directory from the DRT directory to
2926         efl/ImageDiff. The major reason for doing so was because ImageDiff does
2927         not depend on DumpRenderTree but for legacy reasons was living inside the
2928         DRT folder, even though it is needed when running pixel tests for WK2.
2929
2930         Now we are moving to Tool directory root as other ports agreed to do the same.
2931
2932         * CMakeLists.txt:
2933         * ImageDiff/CMakeLists.txt: Added.
2934         * ImageDiff/PlatformEfl.cmake: Added.
2935         * ImageDiff/efl/ImageDiff.cpp: Renamed from Tools/efl/ImageDiff/ImageDiff.cpp.
2936         * efl/ImageDiff/CMakeLists.txt: Removed.
2937
2938 2012-12-19  Zeno Albisser  <zeno@webkit.org>
2939
2940         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
2941         https://bugs.webkit.org/show_bug.cgi?id=105309
2942
2943         The version number is only used as a part
2944         of the library name on Linux.
2945
2946         Reviewed by Kenneth Rohde Christiansen.
2947
2948         * Scripts/webkitdirs.pm:
2949         (builtDylibPathForName):
2950
2951 2012-12-19  Alberto Garcia  <agarcia@igalia.com>
2952
2953         [GTK] Tests that time out are not considered as failures.
2954         https://bugs.webkit.org/show_bug.cgi?id=105318
2955
2956         Reviewed by Martin Robinson.
2957
2958         Fix parsing of the GTK API tests log files.
2959
2960         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2961         (RunGtkAPITests.commandComplete):
2962
2963 2012-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
2964
2965         Unreviewed, rolling out r138061.
2966         https://bugs.webkit.org/show_bug.cgi?id=105396
2967
2968         Original patch:
2969         http://trac.webkit.org/changeset/138061
2970         https://bugs.webkit.org/show_bug.cgi?id=97359
2971
2972         It is crashing on Debug bots
2973
2974         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2975         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2976         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
2977         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2978         (AccessibilityUIElement):
2979
2980 2012-12-18  Alice Liu  <alice.liu@apple.com>
2981
2982         Test for https://bugs.webkit.org/show_bug.cgi?id=105106
2983         Add SPI to WebKit1 WebFrame for hit testing
2984
2985         Reviewed by Dan Bernstein.
2986
2987         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/ElementAtPointInWebFrame.mm
2988         * TestWebKitAPI/Tests/mac/ElementAtPointInWebFrame.mm: Added.
2989         (TestWebKitAPI::TEST): Loads html with two divs positioned in the 2nd and 4th quadrants of the webview.
2990         Then hit-tests at three points, expecting to hit the two divs and body element. 
2991
2992 2012-12-18  Alice Liu  <alice.liu@apple.com>
2993
2994         Test for https://bugs.webkit.org/show_bug.cgi?id=105262
2995         Add SPI to WebKit1 WebFrame for node conversion to JSValueRef
2996
2997         Reviewed by Anders Carlsson.
2998
2999         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/JSWrapperForNodeInWebFrame.mm.mm
3000         * TestWebKitAPI/Tests/mac/JSWrapperForNodeInWebFrame.mm.mm: Added.
3001         (TestWebKitAPI::TEST): Tests for the correct JS wrapper for a DOMNode, provided a WebScriptWorld.
3002         Loads html with a single div element. In an isolated world, creates a property on that node.
3003         Also, in the standard world, creates a different property on that node.  Then tests for 4 things:
3004             - Existence of the isolated property in the isolated world.
3005             - Existence of the standard property in the standard world.
3006             - Non-existence of the isolated property in the standard world.
3007             - Non-existence of the standard property in the isolated world.
3008
3009 2012-12-18  Simon Fraser  <simon.fraser@apple.com>
3010
3011         Disable the SimplifyMarkup API test, wich asserts in debug builds.
3012         Tracked by https://bugs.webkit.org/show_bug.cgi?id=105370.
3013         
3014         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
3015         (TestWebKitAPI::TEST):
3016
3017 2012-12-18  Joseph Pecoraro  <pecoraro@apple.com>
3018
3019         [Mac] Add Build Phase to Check Headers for Inappropriate Macros (Platform.h macros)
3020         https://bugs.webkit.org/show_bug.cgi?id=104279
3021
3022         Reviewed by David Kilzer.
3023
3024         Add a build phase script that checks a framework's header files for
3025         Platform.h and Compiler.h macros. Also add some tests for the script.
3026
3027         * Scripts/check-for-inappropriate-macros-in-external-headers: Added.
3028         Script to be used in build phases to check for inappropriate macros in headers.
3029
3030         * Scripts/test-webkitruby: Added.
3031         Script that can be used to run all webkitruby tests.
3032
3033         * Scripts/test-webkit-scripts:
3034         Also run the ruby tests.
3035
3036         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/fake-data-failing-expected.txt: Added.
3037         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/pass-expected.txt: Added.
3038         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Fail.h: Added.
3039         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Pass.h: Added.
3040         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/run-test.rb: Added.
3041         Tests for the new script.
3042
3043 2012-12-18  Dominic Mazzoni  <dmazzoni@google.com>
3044
3045         AX: support clickPoint in DRT for chromium
3046         https://bugs.webkit.org/show_bug.cgi?id=97359
3047
3048         Reviewed by Chris Fleizach.
3049
3050         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
3051         several more tests can pass.
3052
3053         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
3054         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
3055         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
3056         (WebTestRunner):
3057         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
3058         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
3059         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
3060         (AccessibilityUIElement):
3061
3062 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
3063
3064         Build fix after r138045.
3065
3066         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3067         (PerfTestsRunner._collect_tests): Use relative_path. We skip only when this path
3068         is not explicitly specified. The condition was backwards but never matched so we
3069         always ran the test regardless of whether it's listed in the skipped list or not.
3070         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3071         (test_collect_tests_with_ignored_skipped_list): Renamed so that it actually runs.
3072
3073 2012-12-18  Emil A Eklund  <eae@chromium.org>
3074
3075         TestWebKitAPI/Tests/WebCore/LayoutUnit fails on mac
3076         https://bugs.webkit.org/show_bug.cgi?id=105332
3077         
3078         Reviewed by Levi Weintraub.
3079         
3080         Create new test suite for LayoutUnit tests as the KURL tests are
3081         of the incompatible TEST_F type.
3082
3083         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
3084         (TestWebKitAPI):
3085         (TestWebKitAPI::TEST_F):
3086
3087 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
3088
3089         Running a skipped test with run-perf-tests could alert the user
3090         https://bugs.webkit.org/show_bug.cgi?id=105219
3091
3092         Reviewed by Eric Seidel.
3093
3094         Force run skipped performance tests explicitly specified in the command line.
3095         Also warn the user when the specified paths don't exist.
3096
3097         * Scripts/webkitpy/layout_tests/port/base.py:
3098         (Port.relative_test_filename):
3099         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3100         (PerfTestsRunner._parse_args):
3101         (PerfTestsRunner._collect_tests):
3102         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3103         (test_collect_tests_with_skipped_list_and_files):
3104
3105 2012-12-18  Yaron Friedman  <yfriedman@chromium.org>
3106
3107         Chrome Android java output is too verbose
3108         https://bugs.webkit.org/show_bug.cgi?id=105247
3109
3110         Reviewed by Tony Chang.
3111
3112         Pass -quiet to ant to suppress extraneous output.
3113
3114         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
3115         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
3116
3117 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
3118
3119         Consider removing --pause-before-testing option
3120         https://bugs.webkit.org/show_bug.cgi?id=105248
3121
3122         Reviewed by Eric Seidel.
3123
3124         Remove --pause-before-testing option.
3125
3126         * Tools/Scripts/webkitpy/layout_tests/port/driver.py:
3127         (DriverProxy.start): Removed.
3128         * Scripts/webkitpy/performance_tests/perftest.py:
3129         (PerfTest.__init__):
3130         (PerfTest.test_path): Renamed from path_or_url.
3131         (PerfTest.run):
3132         (PerfTest): Removed "driver" argument.
3133         (PerfTest._run_with_driver):
3134         (PerfTest.run_single):
3135         (ChromiumStylePerfTest.__init__):
3136         (PageLoadingPerfTest.__init__):
3137         (PageLoadingPerfTest.run_single):
3138         (PageLoadingPerfTest._run_with_driver): Renamed from run so that PerfTest.run can create a driver for us.
3139         (ReplayPerfTest.__init__):
3140         (ReplayPerfTest.prepare):
3141         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
3142         (TestPageLoadingPerfTest.test_run):
3143         (TestPageLoadingPerfTest.test_run_with_memory_output):
3144         (TestPageLoadingPerfTest.test_run_with_bad_output):
3145         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3146         (PerfTestsRunner._parse_args):
3147         (PerfTestsRunner._run_tests_set): Removed the code to create a driver and pause before running the test.
3148         (PerfTestsRunner._run_single_test): Removed "driver" argument.
3149         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3150         (run_test):
3151         (test_run_test_pause_before_testing): Removed.
3152
3153 2012-12-18  Alberto Garcia  <agarcia@igalia.com>
3154
3155         [GTK] Tests that time out are not considered as failures.
3156         https://bugs.webkit.org/show_bug.cgi?id=105318
3157
3158         Reviewed by Martin Robinson.
3159
3160         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3161         (RunGtkAPITests.commandComplete):
3162         Consider log lines that end with TIMEOUT as errors.
3163
3164         * Scripts/run-gtk-tests:
3165         (TestRunner.run_tests):
3166         Consider timeouts as failures when returning the number of failed
3167         tests.
3168
3169 2012-12-18  Pravin D  <pravin.d@samsung.com>
3170
3171         Unreviewed, adding secondary email.
3172
3173         * Scripts/webkitpy/common/config/committers.py:
3174
3175 2012-12-18  Pravin D  <pravin.d@samsung.com>
3176
3177         Unreviewed, promoting myself to committer's list.
3178
3179         * Scripts/webkitpy/common/config/committers.py:
3180
3181 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3182
3183         [EFL] ImageDiff does not build on WebKit2 only builds
3184         https://bugs.webkit.org/show_bug.cgi?id=105288
3185
3186         Reviewed by Kenneth Rohde Christiansen.
3187
3188         Moved ImageDiff to the efl/ folder inside Tools/ (so we don't pollute
3189         the folder with things which are EFL-specific.
3190
3191         ImageDiff will be built by default now regardless of WebKit API flavor
3192         you choose.
3193
3194         * CMakeLists.txt:
3195         * DumpRenderTree/efl/CMakeLists.txt:
3196         * efl/ImageDiff/CMakeLists.txt: Added.
3197         * efl/ImageDiff/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/efl/ImageDiff.cpp.
3198
3199 2012-12-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
3200
3201         [EFL][WK2] When creating new windows, use the opener's context
3202         https://bugs.webkit.org/show_bug.cgi?id=105283
3203
3204         Reviewed by Laszlo Gombos.
3205
3206         Use the context of the opener if exists.
3207
3208         * MiniBrowser/efl/main.c:
3209         (on_key_down):
3210         (on_window_create):
3211         (window_create):
3212         (elm_main):
3213
3214 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3215
3216         [EFL][WK2] Gardening of TestWebKitAPI tests
3217         https://bugs.webkit.org/show_bug.cgi?id=105268
3218
3219         Reviewed by Kenneth Rohde Christiansen.
3220
3221         Some tests were marked as failing but are passing and the newly added
3222         LayoutUnit should have it's own binary because it conflicts with the
3223         fixture used by KURL and is failing on the bots because of that.
3224
3225         * TestWebKitAPI/CMakeLists.txt:
3226         * TestWebKitAPI/PlatformEfl.cmake:
3227
3228 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
3229
3230         [chromium] move webintent related methods to TestRunner library
3231         https://bugs.webkit.org/show_bug.cgi?id=105164
3232
3233         Reviewed by Darin Fisher.
3234
3235         Also deleted the unused deliveredIntentFailure and
3236         deliveredIntentResult methods.
3237
3238         * DumpRenderTree/chromium/DRTTestRunner.cpp:
3239         (DRTTestRunner::DRTTestRunner):
3240         * DumpRenderTree/chromium/DRTTestRunner.h:
3241         (DRTTestRunner):
3242         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
3243         (WebKit):
3244         (WebTestRunner::WebTestDelegate::setCurrentWebIntentRequest):
3245         (WebTestRunner::WebTestDelegate::currentWebIntentRequest):
3246         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
3247         (WebKit):
3248         (WebTestProxyBase):
3249         (WebTestRunner::WebTestProxy::registerIntentService):
3250         (WebTestRunner::WebTestProxy::dispatchIntent):
3251         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
3252         (WebTestRunner::TestRunner::TestRunner):
3253         (WebTestRunner::TestRunner::sendWebIntentResponse):
3254         (WebTestRunner):
3255         (WebTestRunner::TestRunner::deliverWebIntent):
3256         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
3257         (TestRunner):
3258         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
3259         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
3260         (WebTestRunner::WebTestProxyBase::registerIntentService):
3261         (WebTestRunner):
3262         (WebTestRunner::WebTestProxyBase::dispatchIntent):
3263         * DumpRenderTree/chromium/WebViewHost.cpp:
3264         (WebViewHost::setCurrentWebIntentRequest):
3265         (WebViewHost::currentWebIntentRequest):
3266         * DumpRenderTree/chromium/WebViewHost.h:
3267         (WebKit):
3268         (WebViewHost):
3269
3270 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
3271
3272         [chromium] move editing callback dumping to the WebTestProxy
3273         https://bugs.webkit.org/show_bug.cgi?id=105013
3274
3275         Reviewed by Darin Fisher.
3276
3277         To decide whether or not to dump the editing callbacks, the WebTestProxy
3278         needs to query the TestRunner. For this purpose, I introduce a
3279         WebTestRunner class that just exposes the shouldDumpEditingCallbacks
3280         method. In contrast to e.g. WebEventSender, the WebTestRunner is just
3281         an interface, and it is not owned by the WebTestInterfaces, but the
3282         embedder of the TestRunner library must provide it. This will change
3283         once all TestRunner methods are in the library.
3284
3285         * DumpRenderTree/DumpRenderTree.gypi:
3286         * DumpRenderTree/chromium/DRTTestRunner.cpp:
3287         (DRTTestRunner::DRTTestRunner):
3288         (DRTTestRunner::reset):
3289         * DumpRenderTree/chromium/DRTTestRunner.h:
3290         (DRTTestRunner):
3291         (DRTTestRunner::taskList):
3292         (DRTTestRunner::WorkQueue::taskList):
3293         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
3294         (WebTestRunner):
3295         (WebTestInterfaces):
3296         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
3297         (WebKit):
3298         (WebTestRunner):
3299         (WebTestProxyBase):
3300         (WebTestRunner::WebTestProxy::shouldBeginEditing):
3301         (WebTestRunner::WebTestProxy::shouldEndEditing):
3302         (WebTestRunner::WebTestProxy::shouldInsertNode):
3303         (WebTestRunner::WebTestProxy::shouldInsertText):
3304         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
3305         (WebTestRunner::WebTestProxy::shouldDeleteRange):
3306         (WebTestRunner::WebTestProxy::shouldApplyStyle):
3307         (WebTestRunner::WebTestProxy::didBeginEditing):
3308         (WebTestRunner::WebTestProxy::didChangeSelection):
3309         (WebTestRunner::WebTestProxy::didChangeContents):
3310         (WebTestRunner::WebTestProxy::didEndEditing):
3311         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h: Copied from Tools/DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h.
3312         (WebTestRunner):
3313         (WebTestRunner::WebTestRunner::shouldDumpEditingCallbacks):
3314         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
3315         (WebTestRunner::TestRunner::TestRunner):
3316         (WebTestRunner::TestRunner::~TestRunner):
3317         (WebTestRunner):
3318         (WebTestRunner::TestRunner::reset):
3319         (WebTestRunner::TestRunner::shouldDumpEditingCallbacks):
3320         (WebTestRunner::TestRunner::dumpEditingCallbacks):
3321         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
3322         (TestRunner):
3323         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
3324         (WebTestRunner::WebTestInterfaces::Internal::testRunner):
3325         (WebTestRunner::WebTestInterfaces::Internal::setTestRunner):
3326         (WebTestInterfaces::Internal):
3327         (WebTestRunner::WebTestInterfaces::Internal::Internal):
3328         (WebTestRunner::WebTestInterfaces::testRunner):
3329         (WebTestRunner):
3330         (WebTestRunner::WebTestInterfaces::setTestRunner):
3331         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
3332         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
3333         (WebTestRunner):
3334         (WebTestRunner::WebTestProxyBase::shouldBeginEditing):
3335         (WebTestRunner::WebTestProxyBase::shouldEndEditing):
3336         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
3337         (WebTestRunner::WebTestProxyBase::shouldInsertText):
3338         (WebTestRunner::WebTestProxyBase::shouldChangeSelectedRange):
3339         (WebTestRunner::WebTestProxyBase::shouldDeleteRange):
3340         (WebTestRunner::WebTestProxyBase::shouldApplyStyle):
3341         (WebTestRunner::WebTestProxyBase::didBeginEditing):
3342         (WebTestRunner::WebTestProxyBase::didChangeSelection):
3343         (WebTestRunner::WebTestProxyBase::didChangeContents):
3344         (WebTestRunner::WebTestProxyBase::didEndEditing):
3345         * DumpRenderTree/chromium/TestShell.cpp:
3346         (TestShell::initialize):
3347         (TestShell::createNewWindow):
3348         * DumpRenderTree/chromium/WebViewHost.cpp:
3349         (WebViewHost::shouldBeginEditing):
3350         (WebViewHost::shouldEndEditing):
3351         (WebViewHost::shouldInsertNode):
3352         (WebViewHost::shouldInsertText):
3353         (WebViewHost::shouldChangeSelectedRange):
3354         (WebViewHost::shouldDeleteRange):
3355         (WebViewHost::shouldApplyStyle):
3356         * DumpRenderTree/chromium/WebViewHost.h:
3357         (WebViewHost):
3358
3359 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
3360
3361         [chromium] Fix methods in TestRunner that directly modify the settings instead of going through WebPreferences
3362         https://bugs.webkit.org/show_bug.cgi?id=105154
3363
3364         Reviewed by Tony Chang.
3365
3366         If the settings are modified directly, the embedder can't keep track of
3367         them and might accidentially override the changed settings.
3368
3369         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
3370         (WebPreferences):
3371         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
3372         (WebTestRunner::TestRunner::TestRunner):
3373         (WebTestRunner::TestRunner::setAsynchronousSpellCheckingEnabled):
3374         (WebTestRunner):
3375         (WebTestRunner::TestRunner::setMinimumTimerInterval):
3376         (WebTestRunner::TestRunner::setTouchDragDropEnabled):
3377         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
3378         (TestRunner):
3379         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
3380         (WebTestRunner::WebPreferences::reset):
3381         (WebTestRunner::WebPreferences::applyTo):
3382
3383 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
3384
3385         Unreviewed unit test skipping.
3386
3387         Filed https://bugs.webkit.org/show_bug.cgi?id=105253 to re-enable them.
3388
3389         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
3390         (TestWebKitAPI):
3391         (TestWebKitAPI::TEST):
3392
3393 2012-12-17  Dean Jackson  <dino@apple.com>
3394
3395         Unreviewed. Add Silvia Pfeiffer to contributor list.
3396
3397         * Scripts/webkitpy/common/config/committers.py:
3398
3399 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
3400
3401         build.webkit.org - add more google windows bots
3402         https://bugs.webkit.org/show_bug.cgi?id=105225
3403
3404         Reviewed by Eric Seidel.
3405
3406         Adds two more windows bots (google-windows-3, a builder, and
3407         google-windows-4, a tester) for the Chromium configs.
3408
3409         * BuildSlaveSupport/build.webkit.org-config/config.json:
3410
3411 2012-12-17  Eric Seidel  <eric@webkit.org>
3412
3413         Cleanup --profiler=perf code a little and allow up to 1000 report files (instead of 100)
3414         https://bugs.webkit.org/show_bug.cgi?id=105217
3415
3416         Reviewed by Dirk Pranke.
3417
3418         Previously --profile would fail strangely if find_unused_filename returned None
3419         (which would happen as soon as we had 100 report files).
3420         Now we'll search through 1000 filenames before giving up (and assert if we failed to find one).
3421         I also made some minor cleanups to the output of perf, and tried to make the
3422         commands more self-documenting by using long-names for args.
3423
3424         * Scripts/webkitpy/common/system/profiler.py:
3425         (SingleFileOutputProfiler.__init__):
3426         (Perf.attach_to_pid):
3427         (Perf.profile_after_exit):
3428
3429 2012-12-17  Sheriff Bot  <webkit.review.bot@gmail.com>
3430
3431         Unreviewed, rolling out r137198.
3432         http://trac.webkit.org/changeset/137198
3433         https://bugs.webkit.org/show_bug.cgi?id=105212
3434
3435         This patch is causing API behavior compatibility problems
3436         (Requested by zdobersek on #webkit).
3437
3438         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3439         (willSendRequestCallback):
3440
3441 2012-12-17  Elliott Sprehn  <esprehn@chromium.org>
3442
3443         Unreviewed. Adding myself to committers list.
3444
3445         * Scripts/webkitpy/common/config/committers.py:
3446
3447 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
3448
3449         nrwt: shuffle code around for cleanup in run_webkit_tests.py
3450         https://bugs.webkit.org/show_bug.cgi?id=105078
3451
3452         Reviewed by Ojan Vafai.
3453
3454         This patch reorders functions so that run_webkit_tests can be
3455         understood in a top-down matter better, and to make things slightly
3456         more sensible when I merge manager.py into it.
3457
3458         Also, this adds tests for the actual main() routine, which had several
3459         bugs get through testing a week or two ago.
3460
3461         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3462         (main):
3463         (parse_args):
3464         (_set_up_derived_options):
3465         (run):
3466         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3467         (RunTest):
3468         (RunTest.test_no_http_tests):
3469         (PortTest.disabled_test_mac_lion):
3470         (MainTest):
3471         (MainTest.test_exception_handling):
3472         (MainTest.test_exception_handling.interrupting_run):
3473         (MainTest.test_exception_handling.successful_run):
3474         (MainTest.test_exception_handling.successful_run.FakeRunDetails):
3475         (MainTest.test_exception_handling.exception_raising_run):
3476
3477 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
3478
3479         webkitpy: move --lint-test-files code into its own module
3480         https://bugs.webkit.org/show_bug.cgi?id=105077
3481
3482         Reviewed by Eric Seidel.
3483
3484         It really didn't need to be in run_webkit_tests.py anymore.
3485         This patch adds a new lint-test-expectations scripts to Tools/Scripts;
3486         rwt --lint-test-files is still supported for backwards compatibility
3487         for now.
3488
3489         * Scripts/lint-test-expectations: Added.
3490         * Scripts/webkitpy/layout_tests/lint_test_expectations.py: Added.
3491         (lint):
3492         (main):
3493         * Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py: Added.
3494         (FakePort):
3495         (FakePort.__init__):
3496         (FakePort.test_configuration):
3497         (FakePort.expectations_dict):
3498         (FakePort.skipped_layout_tests):
3499         (FakePort.all_test_configurations):
3500         (FakePort.configuration_specifier_macros):
3501         (FakePort.get_option):
3502         (FakeFactory):
3503         (FakeFactory.__init__):
3504         (FakeFactory.get):
3505         (FakeFactory.all_port_names):
3506         (LintTest):
3507         (LintTest.test_all_configurations):
3508         (LintTest.test_lint_test_files):
3509         (LintTest.test_lint_test_files__errors):
3510         (MainTest):
3511         (MainTest.test_success):
3512         (MainTest.test_success.interrupting_lint):
3513         (MainTest.test_success.successful_lint):
3514         (MainTest.test_success.exception_raising_lint):
3515         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3516         (main):
3517         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3518         (StreamTestingMixin.assertNotEmpty):
3519
3520 2012-12-17  Adam Klein  <adamk@chromium.org>
3521
3522         build-webkit: rename --template-tag to --template-element to match ENABLE #define name
3523         https://bugs.webkit.org/show_bug.cgi?id=105072
3524
3525         Reviewed by Laszlo Gombos.
3526
3527         * Scripts/webkitperl/FeatureList.pm:
3528
3529 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
3530
3531         Add some unit testing for WTF::clampTo* functions
3532         https://bugs.webkit.org/show_bug.cgi?id=105060
3533
3534         Reviewed by Emil A Eklund.
3535
3536         The new tests are targeted at clamping to int / unsigned and float as those are
3537         the one explicitely exposed through more dedicated functions (like clampToInteger,
3538         clampToPositiveInteger and clampToFloat).
3539
3540         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
3541
3542 2012-12-17  Eric Seidel  <eric@webkit.org>
3543
3544         rpt --profile --chromium-android throws exception
3545         https://bugs.webkit.org/show_bug.cgi?id=105061
3546
3547         Reviewed by Daniel Bates.
3548
3549         Somehow popen(args) got turned into popen(*args) at some point,
3550         which caused my new stringify code to fail when args was passed by name.
3551         Fixed and tested.
3552
3553         * Scripts/webkitpy/common/system/executive.py:
3554         (Executive._stringify_args):
3555         (Executive):
3556         (Executive.popen):
3557         * Scripts/webkitpy/common/system/executive_unittest.py:
3558         (ExecutiveTest.test_popen_args):
3559
3560 2012-12-17  Mark Pilgrim  <pilgrim@chromium.org>
3561
3562         Use Platform::current() instead of webKitPlatformSupport() in DumpRenderTree
3563         https://bugs.webkit.org/show_bug.cgi?id=105054
3564
3565         Reviewed by Jochen Eisinger.
3566
3567         Part of a refactoring series; see tracking bug 82948.
3568
3569         * DumpRenderTree/chromium/NotificationPresenter.cpp:
3570         (NotificationPresenter::show):
3571         * DumpRenderTree/chromium/Task.cpp:
3572         (postTask):
3573         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
3574
3575 2012-12-17  Emil A Eklund  <eae@chromium.org>
3576
3577         Clamp values in LayoutUnit::operator/ when SATURATED_LAYOUT_ARITHMETIC is enabled
3578         https://bugs.webkit.org/show_bug.cgi?id=104955
3579
3580         Reviewed by Julien Chaffraix.
3581
3582         Add tests for LayoutUnit.
3583
3584         * TestWebKitAPI/CMakeLists.txt:
3585         * TestWebKitAPI/ForwardingHeaders: Added.
3586         * TestWebKitAPI/ForwardingHeaders/WebCore: Added.
3587         * TestWebKitAPI/ForwardingHeaders/WebCore/LayoutUnit.h: Added.
3588         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
3589         * TestWebKitAPI/TestWebKitAPI.gypi:
3590         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3591         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp: Added.
3592         (TestWebKitAPI):
3593         (TestWebKitAPI::TEST):
3594         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
3595
3596 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
3597
3598         [chromium] removed unused preferences from testRunner.overridePreference
3599         https://bugs.webkit.org/show_bug.cgi?id=105169
3600
3601         Reviewed by Tony Chang.
3602
3603         Most notably this removes the code related to font family maps which is
3604         now handled via window.internals.settings
3605
3606         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
3607         (WebPreferences):
3608         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
3609         (WebTestRunner::TestRunner::overridePreference):
3610         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
3611         (TestRunner):
3612         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
3613         (WebTestRunner::WebPreferences::applyTo):
3614
3615 2012-12-17  James Robinson  <jamesr@chromium.org>
3616
3617         [chromium] Move creation of WebLayerTreeView to WebWidgetClient
3618         https://bugs.webkit.org/show_bug.cgi?id=105071
3619
3620         Reviewed by Adrienne Walker.
3621
3622         Update WebViewHost for the new WebWidgetClient interface.
3623
3624         * DumpRenderTree/chromium/WebViewHost.cpp:
3625         (WebViewHost::initializeLayerTreeView):
3626         (WebViewHost::layerTreeView):
3627         (WebViewHost::~WebViewHost):
3628         * DumpRenderTree/chromium/WebViewHost.h:
3629         (WebViewHost):
3630
3631 2012-12-17  Kenneth Rohde Christiansen  <kenneth@webkit.org>
3632
3633         [EFL][WK2] window_create doesn't receive the url
3634         https://bugs.webkit.org/show_bug.cgi?id=105184
3635
3636         Reviewed by Alexis Menard.
3637
3638         * MiniBrowser/efl/main.c: Use the url supplied to window_create
3639         (on_window_create):
3640
3641 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
3642
3643         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
3644         https://bugs.webkit.org/show_bug.cgi?id=105026
3645
3646         Reviewed by Yury Semikhatsky.
3647
3648         Old SFINAE test was replaced with new one based on this article:
3649         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
3650
3651         New test which covers this problem was added.
3652
3653         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3654
3655 2012-12-17  Anthony Scian  <ascian@rim.com>
3656
3657         [Blackberry] Static code analysis warning fixes
3658         https://bugs.webkit.org/show_bug.cgi?id=105129
3659
3660         Reviewed by Rob Buis.
3661
3662         Added null checks for fopen calls, no recovery attempted; access fault prevention only.
3663
3664         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
3665         (BlackBerry::WebKit::createFile):
3666         (BlackBerry::WebKit::DumpRenderTree::runTest):
3667
3668 2012-12-17  John Mellor  <johnme@chromium.org>
3669
3670         Add Anton Vayvod and Tim Volodine as contributors in committers.py
3671         https://bugs.webkit.org/show_bug.cgi?id=105025
3672
3673         Unreviewed. Adds two contributors to committers.py.
3674         They are initially contributing to the Text Autosizing feature (master bug: 84186).
3675
3676         * Scripts/webkitpy/common/config/committers.py:
3677
3678 2012-12-17  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3679
3680         Unreviewed, add myself to committers.py.
3681
3682         * Scripts/webkitpy/common/config/committers.py:
3683
3684 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3685
3686         [Qt] remove some unnecessary CONFIG additions
3687
3688         Reviewed by Simon Hausmann.
3689
3690         qt is already added by spec_pre.prf, warn_on and depend_includepath by
3691         default_pre.prf.
3692
3693         * qmake/mkspecs/features/default_post.prf:
3694
3695 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
3696
3697         Unreviewed, rolling out r137892.
3698         http://trac.webkit.org/changeset/137892
3699         https://bugs.webkit.org/show_bug.cgi?id=105026
3700
3701         it broke compilation on windows
3702
3703         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3704
3705 2012-12-14  Ilya Tikhonovsky  <loislo@chromium.org>
3706
3707         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
3708         https://bugs.webkit.org/show_bug.cgi?id=105026
3709
3710         Reviewed by Yury Semikhatsky.
3711
3712         Old SFINAE test was replaced with new one based on this article:
3713         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
3714
3715         New test which covers this problem was added.
3716
3717         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3718
3719 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3720
3721         [Qt] rely on automatic output directory setup
3722
3723         Reviewed by Simon Hausmann.
3724
3725         Now being a proper qt module (by virtue of having load(qt_build_config)
3726         in .qmake.conf), webkit gets the path setup goodies for free.
3727
3728         This also fixes Makefile.api.Debug/Release trying to generate qrc_WebKit.cpp
3729         at the same time, instead of the debug and release version ending up in different
3730         directories as they should.
3731
3732         * qmake/mkspecs/features/default_pre.prf:
3733
3734 2012-12-17  Charles Wei  <charles.wei@torchmobile.com.cn>
3735
3736         [BlackBerry] Need to re-enable video track
3737         https://bugs.webkit.org/show_bug.cgi?id=105155
3738
3739         Reviewed by George Staikos.
3740
3741         Just enable Video Track for BlackBerry.
3742
3743         * Scripts/webkitperl/FeatureList.pm:
3744
3745 2012-12-17  János Badics  <jbadics@inf.u-szeged.hu>
3746
3747         Unreviewed. Added myself to committers.py
3748
3749         * Scripts/webkitpy/common/config/committers.py:
3750
3751 2012-12-17  Jonathan Liu  <net147@gmail.com>
3752
3753         Fix linking of libraries with -Wl,-whole-archive
3754         https://bugs.webkit.org/show_bug.cgi?id=104436
3755
3756         Reviewed by Simon Hausmann.
3757
3758         If there is more than one library in LIBS linked with
3759         -Wl,-whole-archive -l... -Wl,-no-whole-archive, qmake tries to
3760         merge the duplicate flags in LIBS which causes linking to fail.
3761
3762         Add no_smart_library_merge to CONFIG to prevent the duplicate
3763         flags from being removed when using -Wl,-whole-archive.
3764
3765         * qmake/mkspecs/features/functions.prf:
3766
3767 2012-12-16  Arvid Nilsson  <anilsson@rim.com>
3768
3769         [BlackBerry] Adapt to new BlackBerry::Platform::TouchPoint API
3770         https://bugs.webkit.org/show_bug.cgi?id=105143
3771         RIM PR 171941
3772
3773         Reviewed by Rob Buis.
3774         Internally reviewed by George Staikos.
3775
3776         Adapt to new method names and encapsulation of TouchPoint data members.
3777
3778         * DumpRenderTree/blackberry/EventSender.cpp:
3779         (addTouchPointCallback):
3780         (updateTouchPointCallback):
3781         (touchEndCallback):
3782         (releaseTouchPointCallback):
3783         (sendTouchEvent):
3784
3785 2012-12-16  Maciej Stachowiak  <mjs@apple.com>
3786
3787         Add a script to compute the mean and 95% confidence interval (using two-sided t-test) of a sample set
3788         https://bugs.webkit.org/show_bug.cgi?id=105148
3789
3790         Reviewed by Filip Pizlo.
3791
3792         Usage examples:
3793         
3794         $ sampstat --help
3795         Usage: sampstat [options] [FILES]
3796           Compute the mean and 95% confidence interval of a sample set.
3797           Standard input or files must contain two or more decimal numbers, one per line.
3798         
3799         Options:
3800           -h, --help            show this help message and exit
3801           -u UNIT, --unit=UNIT  assume values are in units of UNIT
3802           -v, --verbose         print all values (with units)
3803         
3804         $ sampstat -u MB memresults.txt
3805         2356.90 MB +/- 101.34 MB (4.3%)
3806         
3807         $ sampstat -v -u MB memresults.txt
3808               2318.21 MB
3809               2399.56 MB
3810               2352.93 MB
3811         ----------------
3812         Mean: 2356.90 MB +/- 101.34 MB (4.3%)
3813         
3814
3815         * Scripts/sampstat: Added.
3816         (sum): Helper function to compute the sum of a list.
3817         (arithmeticMean): Compute the meam of a list.
3818         (standardDeviation): Compute the sample standard deviation (unbiased estimator).
3819         (standardError): Compute the sample standard error.
3820         (tDist): Compute t(.025, n-1), the t-value for atwo-sided 95% confidence interval. 
3821         (twoSidedConfidenceInterval): Compute the two-sided confidence interval range about the mean.
3822
3823 2012-12-15  Nima Ghanavatian  <nghanavatian@rim.com>
3824
3825         [BlackBerry] Pass key modifiers with touch and mouse events
3826         https://bugs.webkit.org/show_bug.cgi?id=105108
3827
3828         Reviewed by Rob Buis.
3829
3830         Update DRT to new api.
3831
3832         Internally reviewed by Gen Mak.
3833
3834         * DumpRenderTree/blackberry/EventSender.cpp:
3835         (mouseDownCallback):
3836         (mouseUpCallback):
3837         (mouseMoveToCallback):
3838
3839 2012-12-15  Zan Dobersek  <zandobersek@gmail.com>
3840
3841         Stop the XvfbDriver before proceeding with its starting setup
3842         https://bugs.webkit.org/show_bug.cgi?id=105006
3843
3844         Reviewed by Philippe Normand.
3845
3846         Just as with the base Driver implementation, stop the XvfbDriver
3847         before proceeding with setting up lock files and a new Xvfb instance.
3848         This forces any driver or Xvfb instances that are currently under
3849         the driver's control to be closed/killed and cleaned up after.
3850
3851         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
3852         (XvfbDriver._start):
3853
3854 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
3855
3856         webkitpy: make port_name a required parameter to the non-test ports
3857         https://bugs.webkit.org/show_bug.cgi?id=105057
3858
3859         Reviewed by Eric Seidel.
3860
3861         This is just minor code cleanup along the path to making Port
3862         constructors simpler and pushing more logic into the port factory.
3863
3864         * Scripts/webkitpy/layout_tests/port/base.py:
3865         (Port.__init__):
3866         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3867         (PortTest.make_port):
3868         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3869         (TestWebKitPort.__init__):
3870         * Scripts/webkitpy/layout_tests/port/test.py:
3871         (TestPort):
3872         (TestPort.determine_full_port_name):
3873         (TestPort.__init__):
3874         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py:
3875         (XvfbDriverTest.make_driver):
3876         (XvfbDriverTest.test_stop):
3877
3878 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
3879
3880         garden-o-matic doesn't know about reftests
3881         https://bugs.webkit.org/show_bug.cgi?id=101976
3882
3883         Reviewed by Eric Seidel.
3884
3885         Attempt to make garden-o-matic properly aware of reftests again.
3886
3887         This is re-landing r137407 with a change to
3888         rebaselineWithStatusUpdates() to actually pass the filtered
3889         failureInofLost to checkout.rebaseline().
3890
3891         Unfortunately, the testing coverage for this whole module is almost
3892         non-existent, and it's not obvious that there's a good way to
3893         stub out checkout.rebaseline() to test that the right thing
3894         happens. I verified the change manually.
3895
3896         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
3897         (.):
3898         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
3899         (.):
3900         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
3901
3902 2012-12-14  Ryosuke Niwa  <rniwa@webkit.org>
3903
3904         "Running 1 DumpRenderTree over X shards" is not a helpful output
3905         https://bugs.webkit.org/show_bug.cgi?id=104858
3906
3907         Reviewed by Dirk Pranke.
3908
3909         The word "shard" is not typically used in test frameworks. This word is confusing for many WebKit
3910         contributors who typically don't have much experience on distributed computations and databases.
3911         So only print this message out in the debug mode.
3912
3913         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3914         * Scripts/webkitpy/layout_tests/views/printing.py:
3915         (Printer.print_workers_and_shards):
3916
3917 2012-12-13  Eric Seidel  <eric@webkit.org>
3918
3919         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
3920         https://bugs.webkit.org/show_bug.cgi?id=104891
3921
3922         Reviewed by Ryosuke Niwa.
3923
3924         I also implemented a very simple "Sample" Profiler using
3925         Mac OS X's /usr/bin/sample command line tool.
3926
3927         The real reason for this abstraction is to make it easy
3928         to support both perf and pprof on linux which seem to
3929         be about equally popular among those I ask in the Chrome team.
3930
3931         * Scripts/webkitpy/common/system/profiler.py:
3932         (ProfilerFactory.create_profiler):
3933         (ProfilerFactory):
3934         (ProfilerFactory.available_profilers_by_name):
3935         (ProfilerFactory.default_profiler_name):
3936         (Sample):
3937         (Sample.__init__):
3938         (Sample.attach_to_pid):
3939         (Sample.profile_after_exit):
3940         (IProfiler.attach_to_pid):
3941         * Scripts/webkitpy/layout_tests/port/driver.py:
3942         (Driver.__init__):
3943         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3944         (PerfTestsRunner._parse_args):
3945
3946 2012-12-13  Eric Seidel  <eric@webkit.org>
3947
3948         Unreviewed, fixing typo in python unittest.
3949
3950         I changed the default profiler on linux from pprof to perf in
3951         bug 104971.  I failed to update the unittest results at that time.
3952
3953         * Scripts/webkitpy/common/system/profiler_unittest.py:
3954         (ProfilerFactoryTest.test_basic):
3955
3956 2012-12-13  Ian Vollick  <vollick@chromium.org>
3957
3958         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
3959         https://bugs.webkit.org/show_bug.cgi?id=104911
3960
3961         Reviewed by Dirk Pranke.
3962
3963         Adds two virtual tests suites to run the tests in
3964         compositing/overflow/ and scrollbars/ with the flag
3965         --enable-accelerated-overflow-scroll, which I've also plumbed through
3966         in the usual way.
3967
3968         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3969         (main):
3970         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
3971         (WebPreferences):
3972         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
3973         (WebTestRunner::WebPreferences::reset):
3974         (WebTestRunner::WebPreferences::applyTo):
3975         * DumpRenderTree/chromium/TestShell.cpp:
3976         (TestShell::TestShell):
3977         (TestShell::resetWebSettings):
3978         * DumpRenderTree/chromium/TestShell.h:
3979         (TestShell::setAcceleratedCompositingForOverflowScrollEnabled):
3980         (TestShell):
3981         * Scripts/webkitpy/layout_tests/port/chromium.py:
3982         (ChromiumPort.virtual_test_suites):
3983
3984 2012-12-13  Eric Seidel  <eric@webkit.org>
3985
3986         Use 'perf' to profile on linux instead of google-pprof by default
3987         https://bugs.webkit.org/show_bug.cgi?id=104971
3988
3989         Reviewed by Dirk Pranke.
3990
3991         This makes Chromium Linux match Chromium Android and use the perf
3992         tool by default.  Once https://bugs.webkit.org/show_bug.cgi?id=104891
3993         lands it will be possible to use pprof again on Linux.
3994
3995         This is slightly more advanced than the AndroidPerf profiler as
3996         instead of using a timeout on "perf record" I instead watch
3997         for the termination of the target process and then control-C
3998         the 'perf record' process.  This required me to add two new
3999         methods to Executive, one to have a limited-time wait() and
4000         the second to be able to send a control-C.  I chose to add
4001         these to Executive to make them easier to mock/fix-for-win32
4002         at a later time if needed.
4003
4004         * Scripts/webkitpy/common/system/executive.py:
4005         (Executive.wait_limited):
4006         (Executive.interrupt):
4007         * Scripts/webkitpy/common/system/profiler.py:
4008         (ProfilerFactory.create_profiler):
4009         (Perf):
4010         (Perf.__init__):
4011         (Perf._perf_path):
4012         (Perf.attach_to_pid):
4013         (Perf._first_ten_lines_of_profile):
4014         (Perf.profile_after_exit):
4015
4016 2012-12-13  Julie Parent  <jparent@chromium.org>
4017
4018         Dashboard cleanup: remove usage of global g_builders.
4019         https://bugs.webkit.org/show_bug.cgi?id=104941
4020
4021         Reviewed by Dirk Pranke.
4022
4023         The dashboards use a lot of global state, which makes hacking on them
4024         complicated. This change removes the use of one such global: g_builders.
4025         In most cases, we can just use currentBuilderGroup().builders instead,
4026         which is now currentBuilders().
4027         Surprisingly, the most changes were required to the unit tests, since
4028         they were even bigger offenders of bad hygiene, relying on global state
4029         set by other tests, randomly clobbering global variables in ways the
4030         real code doesn't, etc.
4031
4032         * TestResultServer/static-dashboards/builders.js:
4033         (BuilderGroup.prototype.setup):
4034         * TestResultServer/static-dashboards/dashboard_base.js:
4035         (.switch.return):
4036         (htmlForTestTypeSwitcher):
4037         * TestResultServer/static-dashboards/flakiness_dashboard.js:
4038         (generatePage):
4039         (getAllTestsTrie):
4040         (processTestRunsForAllBuilders):
4041         (showPopupForBuild):
4042         (generatePageForExpectationsUpdate):
4043         (loadExpectationsLayoutTests):
4044         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
4045         (resetGlobals):
4046         (stubResultsByBuilder):
4047         (test):
4048         * TestResultServer/static-dashboards/loader.js:
4049         (.):
4050         * TestResultServer/static-dashboards/loader_unittests.js:
4051
4052 2012-12-13  Eric Seidel  <eric@webkit.org>
4053
4054         Callers should not have to stringify args before calling Executive run_command/popen
4055         https://bugs.webkit.org/show_bug.cgi?id=104975
4056
4057         Reviewed by Dirk Pranke.
4058
4059         One could argue that we should match the python call syntax here,
4060         but I think it's a more friendly API if we automagically handle
4061         stringification of args in run_command, etc.
4062         This removes map(unicode, args) from several callsites.
4063
4064         When I first tried to land this change, I didn't realize that
4065         Executive._command_for_printing depended on this behavior
4066         having been applied to args in run_command.  The fix is to
4067         call _stringify_args in both run_command and popen.
4068         This is slightly redundant, but given how short args have to be
4069         (due to shell limits), I don't think the double-encode check
4070         matters in practice.
4071
4072         This is slightly complicated by the fact that apache_http_server.py
4073         is the one caller in our codebase which uses shell=True.
4074         shell=True is a well-documented trail-of-tears:
4075         http://stackoverflow.com/questions/3172470/actual-meaning-of-shell-true-in-subprocess
4076         but to support this legacy (windows-only) code (which I can't easily test)
4077         I've added an if-hack to avoid stringifying the the popen(shell=True) case.
4078
4079         * Scripts/webkitpy/common/system/executive.py:
4080         (Executive.run_command):
4081         (Executive._stringify_args):
4082         (Executive.popen):
4083         * Scripts/webkitpy/common/system/executive_unittest.py:
4084         (ExecutiveTest.test_auto_stringify_args):
4085         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
4086         (attach_to_pid):
4087
4088 2012-12-13  Adrienne Walker  <enne@chromium.org>
4089
4090         Unreviewed, rolling out r137645, r137646, and r137667.
4091         http://trac.webkit.org/changeset/137645
4092         http://trac.webkit.org/changeset/137646
4093         http://trac.webkit.org/changeset/137667
4094         https://bugs.webkit.org/show_bug.cgi?id=104911
4095
4096         Breaks some overflow layout tests
4097
4098         * Scripts/webkitpy/layout_tests/port/chromium.py:
4099         (ChromiumPort.virtual_test_suites):
4100
4101 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
4102
4103         Unreviewed, rolling out r137661.
4104         http://trac.webkit.org/changeset/137661
4105         https://bugs.webkit.org/show_bug.cgi?id=104891
4106
4107         broke unit tests, run-webkit-tests
4108
4109         * Scripts/webkitpy/common/system/executive.py:
4110         (Executive._run_command_with_teed_output):
4111         (Executive.run_command):
4112         (Executive.popen):
4113         * Scripts/webkitpy/common/system/profiler.py:
4114         (ProfilerFactory.create_profiler):
4115         (GooglePProf.profile_after_exit):
4116         (IProfiler.attach_to_pid):
4117         * Scripts/webkitpy/layout_tests/port/driver.py:
4118         (Driver.__init__):
4119         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
4120         (PerfTestsRunner._parse_args):
4121
4122 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
4123
4124         nrwt: move rundetails, summarize_results into test_run_results
4125         https://bugs.webkit.org/show_bug.cgi?id=104963
4126
4127         Reviewed by Eric Seidel.
4128
4129         More cleanup; this the plain-old-data functions out of manager.py
4130         and into test_run_results next to the other aggregate data structures.
4131
4132         Also this cleans up a bunch of test code and reduces duplication.
4133
4134         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4135         (Manager.run):
4136         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
4137         (ManagerTest.test_look_for_new_crash_logs):
4138         * Scripts/webkitpy/layout_tests/models/test_run_results.py:
4139         (TestRunResults.__init__):
4140         (TestRunResults.add):
4141         (RunDetails):
4142         (RunDetails.__init__):
4143         (_interpret_test_failures):
4144         (summarize_results):
4145         * Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py: Added.
4146         (get_result):
4147         (run_results):
4148         (summarized_results):
4149         (InterpretTestFailuresTest):
4150         (InterpretTestFailuresTest.setUp):
4151         (InterpretTestFailuresTest.test_interpret_test_failures):
4152         (SummarizedResultsTest):
4153         (SummarizedResultsTest.setUp):
4154         (SummarizedResultsTest.test_no_svn_revision):
4155         (SummarizedResultsTest.test_svn_revision):
4156         (SummarizedResultsTest.test_summarized_results_wontfix):
4157         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
4158         (BuildBotPrinterTests.get_printer):
4159         (BuildBotPrinterTests.test_print_unexpected_results):
4160         (BuildBotPrinterTests.test_print_results):
4161
4162 2012-12-13  Eric Seidel  <eric@webkit.org>
4163
4164         Make --profile output easier to understand for pprof on linux
4165         https://bugs.webkit.org/show_bug.cgi?id=104957
4166
4167         Reviewed by Dirk Pranke.
4168
4169         pprof is kinda mysterious, so we print some information
4170         on how to use it, and how to learn how to read the sample dump.
4171         This is similar to what we've done for perf output for Chromium Android.
4172
4173         * Scripts/webkitpy/common/system/profiler.py:
4174         (GooglePProf._pprof_path):
4175         (GooglePProf.profile_after_exit):
4176
4177 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
4178
4179         Revert r137407 - it didn't actually work right.
4180         https://bugs.webkit.org/show_bug.cgi?id=101976
4181
4182         Unreviewed, build fix.
4183
4184         Turns out that rebaselining was completely broken with this change :(
4185         We're passing in a list of test names rather than the list of
4186         failureInfo objects.
4187
4188         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
4189         (.):
4190         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
4191         (.):
4192         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
4193
4194 2012-12-13  Antoine Labour  <piman@chromium.org>
4195
4196         [chromium] Small fixes for WebPluginContainer::setWebLayer
4197         https://bugs.webkit.org/show_bug.cgi?id=104953
4198
4199         Reviewed by James Robinson.
4200
4201         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Clean up
4202         dangling pointers, force correct invalidation of the layer.
4203
4204 2012-12-13  Eric Seidel  <eric@webkit.org>
4205
4206         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
4207         https://bugs.webkit.org/show_bug.cgi?id=104891
4208
4209         Reviewed by Dirk Pranke.
4210
4211         I also implemented a very simple "Sample" Profiler using
4212         Mac OS X's /usr/bin/sample command line tool.
4213
4214         I also moved the map(unicode, args) call into Executive.popen
4215         so that all callers don't have to do it themselves.
4216
4217         The real reason for this abstraction is to make it easy
4218         to support both perf and pprof on linux which seem to
4219         be about equally popular among those I ask in the Chrome team.
4220
4221         * Scripts/webkitpy/common/system/executive.py:
4222         (Executive.popen):
4223         * Scripts/webkitpy/common/system/profiler.py:
4224         (ProfilerFactory.create_profiler):
4225         (ProfilerFactory):
4226         (ProfilerFactory.available_profilers_by_name):
4227         (ProfilerFactory.default_profiler_name):
4228         (Sample):
4229         (Sample.__init__):
4230         (Sample.attach_to_pid):
4231         (Sample.profile_after_exit):
4232         (IProfiler.attach_to_pid):
4233         * Scripts/webkitpy/layout_tests/port/driver.py:
4234         (Driver.__init__):
4235         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
4236         (PerfTestsRunner._parse_args):
4237
4238 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
4239
4240         support -wk2 port names properly in webkitpy.layout_tests.port
4241         https://bugs.webkit.org/show_bug.cgi?id=104761
4242
4243         Reviewed by Eric Seidel.
4244
4245         Our support for WK2 port names was inconsistent at best; this patch
4246         modifies the code so that <port_name>-wk2 is supported for all ports that
4247         have WK2 implementations and is equivalent to specifying <port_name>
4248         and -2/--webkit-test-runner. In addition, this modifies
4249         builders.all_port_names() to include the wk2 variants.
4250
4251         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
4252         * Scripts/webkitpy/layout_tests/port/apple.py:
4253         (ApplePort.determine_full_port_name):
4254         * Scripts/webkitpy/layout_tests/port/base.py:
4255         (Port.determine_full_port_name):
4256         (Port.__init__):
4257         * Scripts/webkitpy/layout_tests/port/builders.py:
4258         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
4259         (DriverTest.make_port):
4260         * Scripts/webkitpy/layout_tests/port/mac.py:
4261         (MacPort.default_baseline_search_path):
4262         * Scripts/webkitpy/layout_tests/port/win.py:
4263         (WinPort.default_baseline_search_path):
4264         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
4265         (TestRebaselineExpectations.test_rebaseline_expectations):
4266
4267 2012-12-13  Ian Vollick  <vollick@chromium.org>
4268
4269         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
4270         https://bugs.webkit.org/show_bug.cgi?id=104911
4271
4272         Reviewed by Adrienne Walker.
4273
4274         Adds two virtual tests suites to run the tests in
4275         compositing/overflow/ and scrollbars/ with the flag
4276         --enable-accelerated-compositing-for-overflow-scroll
4277
4278         * Scripts/webkitpy/layout_tests/port/chromium.py:
4279         (ChromiumPort.virtual_test_suites):
4280
4281 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
4282
4283         delete extraneous self.fail in mastercfg_unittest.py
4284         https://bugs.webkit.org/show_bug.cgi?id=104940
4285
4286         Reviewed by Eric Seidel.
4287
4288         This was accidentally introduced in r137404.
4289
4290         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py: