29a21123a8306de72f27499a14f1e9830a386c51
[WebKit-https.git] / Tools / ChangeLog
1 2013-01-10  Ryosuke Niwa  <rniwa@webkit.org>
2
3         Add Mac WK2 EWS bots
4         https://bugs.webkit.org/show_bug.cgi?id=106590
5
6         Reviewed by Adam Barth.
7
8         Added MacWK2EWS, and made myself a watcher for MacEWS and MacWK2EWS.
9         Also add a deprecated MacWK2Port class.
10
11         * QueueStatusServer/model/queues.py:
12         (Queue):
13         * Scripts/webkitpy/common/config/ports.py:
14         (DeprecatedPort.port):
15         (MacWK2Port):
16         (MacWK2Port.run_webkit_tests_command):
17         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
18         (MacEWS):
19         (MacWK2EWS):
20         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
21         (_test_ewses):
22
23 2013-01-10  Dan Carney  <dcarney@google.com>
24
25         [chromium] move webpermissionclient related methods to TestRunner library
26         https://bugs.webkit.org/show_bug.cgi?id=106548
27
28         Reviewed by Jochen Eisinger.
29
30         * DumpRenderTree/DumpRenderTree.gypi:
31         * DumpRenderTree/chromium/DRTTestRunner.cpp:
32         (DRTTestRunner::DRTTestRunner):
33         (DRTTestRunner::reset):
34         * DumpRenderTree/chromium/DRTTestRunner.h:
35         (DRTTestRunner):
36         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
37         (WebTestRunner::WebTestDelegate::normalizeLayoutTestURL):
38         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
39         (WebKit):
40         (WebTestRunner::WebTestRunner::webPermissions):
41         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
42         (WebTestRunner::TestRunner::TestRunner):
43         (WebTestRunner::TestRunner::setDelegate):
44         (WebTestRunner):
45         (WebTestRunner::TestRunner::reset):
46         (WebTestRunner::TestRunner::webPermissions):
47         (WebTestRunner::TestRunner::dumpPermissionClientCallbacks):
48         (WebTestRunner::TestRunner::setImagesAllowed):
49         (WebTestRunner::TestRunner::setScriptsAllowed):
50         (WebTestRunner::TestRunner::setStorageAllowed):
51         (WebTestRunner::TestRunner::setPluginsAllowed):
52         (WebTestRunner::TestRunner::setAllowDisplayOfInsecureContent):
53         (WebTestRunner::TestRunner::setAllowRunningOfInsecureContent):
54         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
55         (WebTestRunner):
56         (TestRunner):
57         * DumpRenderTree/chromium/TestRunner/src/WebPermissions.cpp: Renamed from Tools/DumpRenderTree/chromium/WebPermissions.cpp.
58         (WebTestRunner):
59         (WebTestRunner::WebPermissions::WebPermissions):
60         (WebTestRunner::WebPermissions::~WebPermissions):
61         (WebTestRunner::WebPermissions::allowImage):
62         (WebTestRunner::WebPermissions::allowScriptFromSource):
63         (WebTestRunner::WebPermissions::allowStorage):
64         (WebTestRunner::WebPermissions::allowPlugins):
65         (WebTestRunner::WebPermissions::allowDisplayingInsecureContent):
66         (WebTestRunner::WebPermissions::allowRunningInsecureContent):
67         (WebTestRunner::WebPermissions::setImagesAllowed):
68         (WebTestRunner::WebPermissions::setScriptsAllowed):
69         (WebTestRunner::WebPermissions::setStorageAllowed):
70         (WebTestRunner::WebPermissions::setPluginsAllowed):
71         (WebTestRunner::WebPermissions::setDisplayingInsecureContentAllowed):
72         (WebTestRunner::WebPermissions::setRunningInsecureContentAllowed):
73         (WebTestRunner::WebPermissions::setDelegate):
74         (WebTestRunner::WebPermissions::setDumpCallbacks):
75         (WebTestRunner::WebPermissions::reset):
76         * DumpRenderTree/chromium/TestRunner/src/WebPermissions.h: Renamed from Tools/DumpRenderTree/chromium/WebPermissions.h.
77         (WebTestRunner):
78         (WebPermissions):
79         * DumpRenderTree/chromium/TestShell.cpp:
80         (TestShell::initialize):
81         (TestShell::resetTestController):
82         (TestShell::createNewWindow):
83         * DumpRenderTree/chromium/TestShell.h:
84         (TestShell):
85         * DumpRenderTree/chromium/WebViewHost.cpp:
86         (WebViewHost::normalizeLayoutTestURL):
87         * DumpRenderTree/chromium/WebViewHost.h:
88         (WebViewHost):
89
90 2013-01-10  Alan Cutter  <alancutter@chromium.org>
91
92         Split EWS tool cold-boot.sh into modular components
93         https://bugs.webkit.org/show_bug.cgi?id=106527
94
95         Reviewed by Adam Barth.
96
97         Split cold-boot.sh script up into build-vm.sh, build-repo.sh and build-boot-cmd.sh.
98         Updated GCE bot build scripts to match.
99         Unable to test scripts directly as they depend on being in the repository to work!
100
101         * EWSTools/GoogleComputeEngine/build-chromium-ews.sh:
102         * EWSTools/GoogleComputeEngine/build-commit-queue.sh:
103         * EWSTools/build-boot-cmd.sh: Copied from Tools/EWSTools/GoogleComputeEngine/build-chromium-ews.sh.
104         * EWSTools/build-repo.sh: Copied from Tools/EWSTools/cold-boot.sh.
105         * EWSTools/build-vm.sh: Renamed from Tools/EWSTools/cold-boot.sh.
106
107 2013-01-10  Julie Parent  <jparent@chromium.org>
108
109         Dashboard Cleanup: Do not generate the page if we are about to reload.
110         https://bugs.webkit.org/show_bug.cgi?id=106584
111
112         Reviewed by Dirk Pranke.
113
114         Move the logic to determine if we should generate the page into
115         parseParameters() since that is where we have the knowledge to
116         determine it, and return that value to handleLocationChange, 
117         rather than passing the changed parameters back to handleLocationChange
118         and having it make the decision.
119         
120         This is logically equivalent, save one case: when we are about to do
121         a reload, we know that there is no reason to generate the page.  Old
122         code was causing an extra page generation in this case.
123         
124         * TestResultServer/static-dashboards/dashboard_base.js:
125         (parseParameters):
126         (handleLocationChange):
127
128 2013-01-10  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>
129
130         [jhbuild] Use the "configure" script to build glib
131         https://bugs.webkit.org/show_bug.cgi?id=106569
132
133         Reviewed by Martin Robinson.
134
135         So far both EFL and GTK have been using autogen.sh to build glib.
136         While that does work most of the time, it's not really needed since
137         we are downloading and building stable releases from tarballs.
138         Additionally, autogen.sh fails to work with automake 1.13 due to the
139         usage of some macros that have been removed (this has already been
140         fixed in glib master).
141
142         * efl/jhbuild.modules:
143         * gtk/jhbuild.modules:
144
145 2013-01-10  Tim 'mithro' Ansell  <mithro@mithis.com>
146
147         Changing tests to be consistent and use self.scm rather then recreating the scm object.
148         (Missed a couple)
149         https://bugs.webkit.org/show_bug.cgi?id=106536
150
151         Reviewed by Eric Seidel.
152
153         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
154
155 2013-01-10  Tony Chang  <tony@chromium.org>
156
157         Speed up supplemental dependency computation
158         https://bugs.webkit.org/show_bug.cgi?id=106503
159
160         Reviewed by Adam Barth.
161
162         Pass the IDL attributes file for generating the bindings in WebKitTestRunner.
163
164         * WebKitTestRunner/CMakeLists.txt:
165
166 2013-01-10  Zan Dobersek  <zandobersek@gmail.com>
167
168         Remove the ENABLE_ANIMATION_API feature define occurences
169         https://bugs.webkit.org/show_bug.cgi?id=106544
170
171         Reviewed by Simon Fraser.
172
173         The Animation API code was removed in r137243. The ENABLE_ANIMATION_API
174         feature define handling still lingers in various build systems and configurations
175         but is of no use, so it should be removed.
176
177         * Scripts/webkitperl/FeatureList.pm:
178         * qmake/mkspecs/features/features.pri:
179
180 2013-01-10  Carlos Garcia Campos  <cgarcia@igalia.com>
181
182         [GTK] Add support for loading web process extensions
183         https://bugs.webkit.org/show_bug.cgi?id=105631
184
185         Reviewed by Gustavo Noronha Silva.
186
187         * MiniBrowser/gtk/GNUmakefile.am:
188         * MiniBrowser/gtk/main.c:
189         (main): Set WEBKIT_INJECTED_BUNDLE env var to use the injected
190         bundle lib from build dir.
191         * Scripts/webkitpy/style/checker.py: Add exceptions for GTK+ API
192         located in WebProcess/InjectedBundle/API/gtk.
193         * gtk/generate-gtkdoc:
194         (get_webkit2_options): Scan also files in
195         WebProcess/InjectedBundle/API/gtk to generate API docs.
196         (get_webkit2_options.injected_bundle_src_path): Helper function to
197         build paths in WebProcess/InjectedBundle/API/gtk.
198
199 2013-01-10  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
200
201         [EFL][WK2] Writing immediately to created windows (open) fails
202         https://bugs.webkit.org/show_bug.cgi?id=105276
203
204         Reviewed by Kenneth Rohde Christiansen.
205
206         Do not reset 'about:blank' url for the new pages in mini browser.
207
208         * MiniBrowser/efl/main.c:
209         (window_create):
210
211 2013-01-10  Tim 'mithro' Ansell  <mithro@mithis.com>
212
213         Changing tests to be consistent and use self.scm rather then recreating the scm object.
214         https://bugs.webkit.org/show_bug.cgi?id=106536
215
216         Reviewed by Eric Seidel.
217
218         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
219
220 2013-01-10  Tim 'mithro' Ansell  <mithro@mithis.com>
221
222         Converting to unittest2 so we can use assertItemsEqual, making tests
223         not dependent on order of items.
224         https://bugs.webkit.org/show_bug.cgi?id=106532
225
226         Before:
227         Ran 1733 tests in 38.385s
228         FAILED (failures=1, errors=22)
229
230         After:
231         Ran 1733 tests in 32.298s
232         FAILED (failures=0, errors=22)
233
234         Reviewed by Eric Seidel.
235
236         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
237         * Scripts/webkitpy/thirdparty/__init__.py:
238         (AutoinstallImportHook._install_unittest2):
239
240 2013-01-10  Christophe Dumez  <christophe.dumez@intel.com>
241
242         [EFL] Add gstreamer 1.0.5 to jhbuild
243         https://bugs.webkit.org/show_bug.cgi?id=106178
244
245         Reviewed by Laszlo Gombos.
246
247         Add gstreamer 1.0.5 to EFL's jhbuild configuration.
248
249         * efl/jhbuild.modules:
250
251 2013-01-09  Alan Cutter  <alancutter@chromium.org>
252
253         Updated parameters in GCE build scripts so they now work.
254         https://bugs.webkit.org/show_bug.cgi?id=106523
255
256         Reviewed by Adam Barth.
257
258         * EWSTools/GoogleComputeEngine/build-chromium-ews.sh:
259         * EWSTools/GoogleComputeEngine/build-commit-queue.sh:
260         * EWSTools/cold-boot.sh:
261
262 2013-01-09  Mark Pilgrim  <pilgrim@chromium.org>
263
264         [Chromium] Remove some stray references to WebKitPlatformSupport.h in DumpRenderTree
265         https://bugs.webkit.org/show_bug.cgi?id=106520
266
267         Reviewed by James Robinson.
268
269         Part of a larger refactoring series. See tracking bug 82948.
270
271         * DumpRenderTree/chromium/WebThemeEngineDRTMac.h:
272         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
273
274 2013-01-09  Julie Parent  <jparent@chromium.org>
275
276         Dashboard cleanup: Remove code for currentState and window.location.hash getting out of state
277         https://bugs.webkit.org/show_bug.cgi?id=106514
278
279         This is old code, as these can no longer get out of sync.
280         permmaLinkURLHash is generated from currentState, which is set only
281         from the location.  The commment about how these can get out of sync
282         refers to a function that no longer exists.
283         
284         Reviewed by Dirk Pranke.
285
286         * TestResultServer/static-dashboards/dashboard_base.js:
287         (handleLocationChange):
288
289 2013-01-09  Julie Parent  <jparent@chromium.org>
290
291         Dashboard Cleanup: Reload the page as soon as we know we need to.
292         https://bugs.webkit.org/show_bug.cgi?id=106515
293
294         No need to continue parsing all of the data once we know that
295         we need to do a reload.
296         
297         Reviewed by Dirk Pranke.
298
299         * TestResultServer/static-dashboards/dashboard_base.js:
300         (parseParameters):
301
302 2013-01-09  Eric Seidel  <eric@webkit.org>
303
304         run-perf-tests --chromium-android --profile should show symbols for the kernel
305         https://bugs.webkit.org/show_bug.cgi?id=106498
306
307         Reviewed by Adam Barth.
308
309         Turns out this was easy, once I finally read the output from "perf report".
310         It appears there may be a bug in "perf record" on android, as it complains
311         about not being able to read from /proc/kallsyms even when
312         /proc/sys/kernel/kptr_restrict is 0.  For now I've not bothered
313         to keep /proc/sys/kernel/kptr_restrict as 0 during the actual record
314         but rather just flip it to 0 long enough to grab the /proc/kallsyms
315         and then flip it back to whatever the device had.
316
317         /proc/sys/kernel/kptr_restrict controls what /proc/kallsyms returns
318         on Linux.  /proc/kallsyms contains a mapping of kernel addresses
319         to symbol names.  Its world-readable, but will return all 0s if you
320         don't have permission to see the kernel symbols.  kptr_restrict
321         supports values 0, 1, 2.  Where 0 means "everyone can see the real symbols"
322         1 is only a specific group, and 2 is "no one, not even root".
323         By default kptr_restrict is 2 on Android it seems.
324         More kptr_restrict docs: http://lwn.net/Articles/420403/
325
326         I also took this opportunity to clean up how the perf record command
327         was built for use/display in AndroidPerf.profile_after_exit and
328         change to always using long-form names for the arguments (to hopefully help readability).
329
330         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
331         (AndroidPerf.__init__):
332         (profile_after_exit):
333         (ChromiumAndroidDriver.__init__):
334         (ChromiumAndroidDriver._update_kallsyms_cache):
335         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
336
337 2013-01-09  Julie Parent  <jparent@chromium.org>
338
339         Dashboard cleanup: remove dead code.
340         https://bugs.webkit.org/show_bug.cgi?id=106509
341
342         Reviewed by Dirk Pranke.
343
344         * TestResultServer/static-dashboards/dashboard_base.js:
345         (parseParameters):
346
347 2013-01-09  Joanmarie Diggs  <jdiggs@igalia.com>
348
349         [GTK] accessibility/aria-labelledby-overrides-label.html requires a proper baseline
350         https://bugs.webkit.org/show_bug.cgi?id=105638
351
352         Reviewed by Martin Robinson.
353
354         The test was failing for two reasons:
355         - AccessibilityRenderObject::correspondingLabelForControlElement() was
356           not ignoring the ARIA labelled-by property
357         - AccessibilityController::accessibleElementById() was not implemented
358
359         Because getting an element by ID cannot be done in the UIProcess, the
360         decision was made to expose the element's ID as an accessible attribute
361         of the object.
362
363         In addition, fixing the bug in AccessibilityRenderObject made it possible
364         to eliminate the Gtk platform-specific expectations for another test.
365
366         * DumpRenderTree/gtk/AccessibilityControllerGtk.cpp:
367         (childElementById): Added. Walks the tree looking for the AtkObject
368         Attribute which exposes the HTML element's ID.
369         (AccessibilityController::accessibleElementById): Implemented.
370         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
371         (AccessibilityUIElement::title): Modified to print out "AXTitle: " even
372         when there is no title, as is done with the Mac port.
373         * WebKitTestRunner/InjectedBundle/AccessibilityController.cpp:
374         (WTR):
375         * WebKitTestRunner/InjectedBundle/atk/AccessibilityControllerAtk.cpp:
376         (WTR::childElementById): Added. Walks the tree looking for the AtkObject
377         Attribute which exposes the HTML element's ID.
378         (WTR):
379         (WTR::AccessibilityController::accessibleElementById): Implemented.
380         * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
381         (WTR::AccessibilityUIElement::title): Modified to print out "AXTitle: "
382         even when there is no title, as is done with the Mac port.
383
384 2013-01-09  Gregg Tavares  <gman@google.com>
385
386         Add Gregg Tavares as a committer
387         https://bugs.webkit.org/show_bug.cgi?id=106492
388
389         Unreviewed.
390
391         * Scripts/webkitpy/common/config/committers.py:
392
393 2013-01-09  Roger Fong  <roger_fong@apple.com>
394
395         Enable seamless iFrames for Windows DRT.
396         https://bugs.webkit.org/show_bug.cgi?id=106468.
397
398         Rubberstamped by Timothy Horton.
399
400         * DumpRenderTree/win/DumpRenderTree.cpp:
401         (resetDefaultsToConsistentValues):
402
403 2013-01-09  Christophe Dumez  <christophe.dumez@intel.com>
404
405         [WK2] Implement eventSender.scheduleAsynchronousKeyDown
406         https://bugs.webkit.org/show_bug.cgi?id=93979
407
408         Reviewed by Benjamin Poulain.
409
410         Implement eventSender.scheduleAsynchronousKeyDown() in
411         WebKitTestRunner.
412
413         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
414         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
415         (WTR::createKeyDownMessageBody):
416         (WTR::EventSendingController::keyDown):
417         (WTR::EventSendingController::scheduleAsynchronousKeyDown):
418         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
419         (EventSendingController):
420         * WebKitTestRunner/TestController.cpp:
421         (WTR::TestController::didReceiveKeyDownMessageFromInjectedBundle):
422         (WTR::TestController::didReceiveMessageFromInjectedBundle):
423         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
424         * WebKitTestRunner/TestController.h:
425         (TestController):
426
427 2013-01-09  Jochen Eisinger  <jochen@chromium.org>
428
429         [chromium] move resource load callback dumping to TestRunner library
430         https://bugs.webkit.org/show_bug.cgi?id=106449
431
432         Reviewed by Adam Barth.
433
434         * DumpRenderTree/chromium/DRTTestRunner.cpp:
435         (DRTTestRunner::DRTTestRunner):
436         (DRTTestRunner::reset):
437         * DumpRenderTree/chromium/DRTTestRunner.h:
438         (DRTTestRunner):
439         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
440         (WebKit):
441         (WebTestRunner::WebTestDelegate::makeURLErrorDescription):
442         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
443         (WebKit):
444         (WebTestProxyBase):
445         (WebTestRunner::WebTestProxy::assignIdentifierToRequest):
446         (WebTestRunner::WebTestProxy::willRequestResource):
447         (WebTestRunner::WebTestProxy::willSendRequest):
448         (WebTestRunner::WebTestProxy::didReceiveResponse):
449         (WebTestRunner::WebTestProxy::didFinishResourceLoad):
450         (WebTestRunner::WebTestProxy::didFailResourceLoad):
451         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
452         (WebTestRunner::WebTestRunner::shouldDumpResourceLoadCallbacks):
453         (WebTestRunner::WebTestRunner::shouldDumpResourceRequestCallbacks):
454         (WebTestRunner::WebTestRunner::shouldDumpResourceResponseMIMETypes):
455         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
456         (WebTestRunner::TestRunner::TestRunner):
457         (WebTestRunner::TestRunner::reset):
458         (WebTestRunner::TestRunner::shouldDumpResourceLoadCallbacks):
459         (WebTestRunner::TestRunner::shouldDumpResourceRequestCallbacks):
460         (WebTestRunner::TestRunner::shouldDumpResourceResponseMIMETypes):
461         (WebTestRunner):
462         (WebTestRunner::TestRunner::dumpResourceLoadCallbacks):
463         (WebTestRunner::TestRunner::dumpResourceRequestCallbacks):
464         (WebTestRunner::TestRunner::dumpResourceResponseMIMETypes):
465         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
466         (TestRunner):
467         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
468         (WebTestRunner::WebTestProxyBase::reset):
469         (WebTestRunner):
470         (WebTestRunner::WebTestProxyBase::assignIdentifierToRequest):
471         (WebTestRunner::WebTestProxyBase::willRequestResource):
472         (WebTestRunner::WebTestProxyBase::willSendRequest):
473         (WebTestRunner::WebTestProxyBase::didReceiveResponse):
474         (WebTestRunner::WebTestProxyBase::didFinishResourceLoad):
475         (WebTestRunner::WebTestProxyBase::didFailResourceLoad):
476         * DumpRenderTree/chromium/WebViewHost.cpp:
477         (WebViewHost::willSendRequest):
478         (WebViewHost::makeURLErrorDescription):
479         (WebViewHost::reset):
480         * DumpRenderTree/chromium/WebViewHost.h:
481         (WebViewHost):
482
483 2013-01-09  Dan Carney  <dcarney@google.com>
484
485         [chromium] move dumpcreateview methods to testrunner library
486         https://bugs.webkit.org/show_bug.cgi?id=106434
487
488         Reviewed by Adam Barth.
489
490         * DumpRenderTree/chromium/DRTTestRunner.cpp:
491         (DRTTestRunner::DRTTestRunner):
492         (DRTTestRunner::reset):
493         * DumpRenderTree/chromium/DRTTestRunner.h:
494         (DRTTestRunner):
495         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
496         (WebTestRunner::WebTestRunner::shouldDumpCreateView):
497         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
498         (WebTestRunner::TestRunner::TestRunner):
499         (WebTestRunner::TestRunner::reset):
500         (WebTestRunner::TestRunner::shouldDumpCreateView):
501         (WebTestRunner):
502         (WebTestRunner::TestRunner::dumpCreateView):
503         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
504         (TestRunner):
505
506 2013-01-09  Csaba Osztrogonác  <ossy@webkit.org>
507
508         webkitpy cleanup: Merge additional_comment_text into comment_text
509         https://bugs.webkit.org/show_bug.cgi?id=106421
510
511         Reviewed by Adam Barth.
512
513         * Scripts/webkitpy/common/config/committervalidator.py:
514         (CommitterValidator.reject_patch_from_commit_queue):
515         (CommitterValidator.reject_patch_from_review_queue):
516         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
517         (Bugzilla.set_flag_on_attachment):
518         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
519         (MockBugzilla.set_flag_on_attachment):
520         * Scripts/webkitpy/tool/bot/feeders_unittest.py:
521         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
522         (AbstractEarlyWarningSystem._post_reject_message_on_bug):
523         * Scripts/webkitpy/tool/commands/queues_unittest.py:
524
525 2013-01-09  Mario Sanchez Prada  <mario.prada@samsung.com>
526
527         [GTK] Raise version of gdk-pixbuf in jhbuild up to 2.26.5
528         https://bugs.webkit.org/show_bug.cgi?id=106464
529
530         Reviewed by Martin Robinson.
531
532         Current version in jhbuild.modules (2.26.0) is giving trouble to
533         when trying to build WebKitGTK from scratch, due to a segmentation
534         fault error in gdk-pixbuf-query-loader during the 'make'
535         phase. Raising the version to 2.26.5 fixes this issue.
536
537         * gtk/jhbuild.modules: Version raised, checksums updated.
538
539 2013-01-09  János Badics  <jbadics@inf.u-szeged.hu>
540
541         [Qt][NRWT] Pass --timeout to DRT/WTR if a test is marked as SLOW.
542         https://bugs.webkit.org/show_bug.cgi?id=90968.
543
544         Reviewed by Csaba Osztrogonác.
545
546         Added functionality in DRT and WTR to use any timeout value while running
547         slow tests (eventually, any test). Now NRWT --time-out-ms determines the
548         timeout value for the test. Added a flag in NRWT (supports_per_test_timeout)
549         to indicate whether the current port supports setting timeout value
550         per test (it's False by default; I enabled it only on Qt).
551         Also corrected a typo in driver.py
552
553         * DumpRenderTree/DumpRenderTree.h:
554         (TestCommand::TestCommand):
555         (TestCommand):
556         * DumpRenderTree/DumpRenderTreeCommon.cpp:
557         (parseInputLine):
558         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
559         (WebCore::DumpRenderTree::processLine):
560         * Scripts/webkitpy/layout_tests/port/base.py:
561         (Port.supports_per_test_timeout):
562         * Scripts/webkitpy/layout_tests/port/driver.py:
563         (Driver.run_test):
564         (Driver._command_from_driver_input):
565         * Scripts/webkitpy/layout_tests/port/qt.py:
566         (QtPort.supports_per_test_timeout):
567         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
568         (WTR::InjectedBundle::InjectedBundle):
569         (WTR::InjectedBundle::didReceiveMessage):
570         (WTR::InjectedBundle::beginTesting):
571         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
572         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
573         (WTR::TestRunner::setCustomTimeout):
574         (WTR):
575         * WebKitTestRunner/InjectedBundle/TestRunner.h:
576         (TestRunner):
577         * WebKitTestRunner/InjectedBundle/qt/TestRunnerQt.cpp:
578         (WTR::TestRunner::initializeWaitToDumpWatchdogTimerIfNeeded):
579         * WebKitTestRunner/TestController.cpp:
580         (WTR::TestController::TestController):
581         (WTR::TestController::getCustomTimeout):
582         (WTR):
583         (WTR::TestCommand::TestCommand):
584         (TestCommand):
585         (WTR::parseInputLine):
586         (WTR::TestController::runTest):
587         (WTR::TestController::runUntil):
588         * WebKitTestRunner/TestController.h:
589         (TestController):
590         * WebKitTestRunner/TestInvocation.cpp:
591         (WTR::TestInvocation::TestInvocation):
592         (WTR::TestInvocation::setCustomTimeout):
593         (WTR):
594         (WTR::TestInvocation::invoke):
595         * WebKitTestRunner/TestInvocation.h:
596         (TestInvocation):
597
598 2013-01-08  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
599
600         Regression(r138681): Add HAVE(ACCESSIBILITY) guard to atk files
601         https://bugs.webkit.org/show_bug.cgi?id=106290
602
603         Reviewed by Martin Robinson.
604
605         Some atk files don't use HAVE(ACCESSIBILITY). It might make build errors when
606         the macro isn't enabled.
607
608         * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
609
610 2013-01-08  Martin Robinson  <mrobinson@igalia.com>
611
612         WebKitTestRunner needs support for setHandlesAuthenticationChallenges
613         https://bugs.webkit.org/show_bug.cgi?id=103653
614
615         Reviewed by Anders Carlsson.
616
617         Implement setHandlesAuthenticationChallenges in the UIProcess by sending credentials
618         and the enabled boolean to the UIProcess and then implementing the didReceiveAuthenticationChallengeInFrame
619         client callback.
620
621         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Exposed the new TestRunner
622         methods to the JavaScript API.
623         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
624         (WTR::TestRunner::setHandlesAuthenticationChallenges): Added this callback for the JavaScript API.
625         (WTR::TestRunner::setAuthenticationUsername): Ditto.
626         (WTR::TestRunner::setAuthenticationPassword): Ditto.
627         * WebKitTestRunner/InjectedBundle/TestRunner.h:
628         (TestRunner): Added new declarations for the above.
629         * WebKitTestRunner/TestController.cpp:
630         (WTR::TestController::createWebViewWithOptions): Hook up the new didReceiveSynchronousMessageFromInjectedBundle
631         callback.
632         (WTR::TestController::resetStateToConsistentValues): Reset the credentials and enabled boolean between tests.
633         (WTR::TestController::didReceiveAuthenticationChallengeInFrame): Handle the callback by using
634         the authentication information passed from the InjectedBundle.
635         * WebKitTestRunner/TestController.h:
636         (WTR::TestController::setHandlesAuthenticationChallenges): Added setter to allow TestInvocation
637         to communicate with the TestController.
638         (WTR::TestController::setAuthenticationUsername): Ditto.
639         (WTR::TestController::setAuthenticationPassword): Ditto.
640         (TestController):
641         * WebKitTestRunner/TestInvocation.cpp:
642         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Handle the messages sending authentication
643         information from the InjectedBundle.
644         (WTR::TestInvocation::outputText): Added this method so that the TestController can output
645         text to the TestInvocation text output.
646         * WebKitTestRunner/TestInvocation.h:
647         (TestInvocation): Added declaration for new method.
648
649 2013-01-08  Benjamin Poulain  <benjamin@webkit.org>
650
651         Add myself as a watcher for Web Geolocation
652
653         Unreviewed.
654
655         * Scripts/webkitpy/common/config/watchlist:
656
657 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
658
659         [webkitpy] Call setup_test_run method of the parent Port interface when method is called on subclasses
660         https://bugs.webkit.org/show_bug.cgi?id=106364
661
662         Reviewed by Eric Seidel.
663
664         When the setup_test_run method is called on the subclasses of the base Port
665         interface, the parent method should also be called via super(), despite the
666         parent method not doing anything at the moment.
667
668         The according calls to super() are required in the ChromiumPort, EflPort
669         and GtkPort implementations.
670
671         * Scripts/webkitpy/layout_tests/port/chromium.py:
672         (ChromiumPort.setup_test_run):
673         * Scripts/webkitpy/layout_tests/port/efl.py:
674         (EflPort.setup_test_run):
675         * Scripts/webkitpy/layout_tests/port/gtk.py:
676         (GtkPort.setup_test_run):
677
678 2013-01-08  Julie Parent  <jparent@chromium.org>
679
680         Dashboard cleanup: Remove globals g_buildersThatFailedToLoad and g_staleBuilders
681         https://bugs.webkit.org/show_bug.cgi?id=106356
682
683         g_buildersThatFailedToLoad and g_staleBuilders were globals defined in
684         dashboard_base, assigned by Loader, and used only by dashboard_base to
685         create error messages.  Moved the variables to be privates on the Loader
686         object, moved error message creation to _getLoadingErrorMessages on the
687         Loader object, and now pass the errors back to dashboard base via the
688         resourceLoadingComplete callback.
689         
690         Also removed the now unused clearError function, it was only being used
691         by unit tests to clean up global state.
692         
693         Reviewed by Dirk Pranke.
694
695         * TestResultServer/static-dashboards/dashboard_base.js:
696         (resourceLoadingComplete):
697         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
698         * TestResultServer/static-dashboards/loader.js:
699         (.):
700         * TestResultServer/static-dashboards/loader_unittests.js:
701
702 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
703
704         [EFL][GTK] Make the PulseAudioSanitizer an object on the EflPort, GtkPort
705         https://bugs.webkit.org/show_bug.cgi?id=106354
706
707         Reviewed by Eric Seidel.
708
709         Put the PulseAudioSanitizer object on the EflPort and GtkPort interfaces
710         instead of those two inheriting from it. Also add a mock object of the
711         sanitizer that's used in unit tests.
712
713         * Scripts/webkitpy/layout_tests/port/efl.py:
714         (EflPort):
715         (EflPort.__init__):
716         (EflPort.setup_test_run):
717         (EflPort.clean_up_test_run):
718         * Scripts/webkitpy/layout_tests/port/efl_unittest.py: Also correct the
719         importing order.
720         (EflPortTest.make_port): Put a mock PulseAudioSanitizer on the instance.
721         * Scripts/webkitpy/layout_tests/port/gtk.py:
722         (GtkPort):
723         (GtkPort.__init__):
724         (GtkPort.setup_test_run):
725         (GtkPort.clean_up_test_run):
726         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py: Also correct the
727         importing order.
728         (GtkPortTest.make_port): Put a mock PulseAudioSanitizer on the instance.
729         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer.py:
730         (PulseAudioSanitizer.unload_pulseaudio_module): Stylize the method as public.
731         (PulseAudioSanitizer.restore_pulseaudio_module): Ditto.
732         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer_mock.py: Added.
733         (PulseAudioSanitizerMock): A simple mock interface for PulseAudioSanitizer.
734         (PulseAudioSanitizerMock.unload_pulseaudio_module):
735         (PulseAudioSanitizerMock.restore_pulseaudio_module):
736
737 2013-01-08  Jochen Eisinger  <jochen@chromium.org>
738
739         [chromium] move dumpFrameLoadCallbacks and friends to the TestRunner library
740         https://bugs.webkit.org/show_bug.cgi?id=106324
741
742         Reviewed by Adam Barth.
743
744         Since the frame load callbacks are part of the WebFrameClient, the
745         WebTestProxy is now intercepting both WebViewClient and WebFrameClient
746         methods.
747
748         * DumpRenderTree/chromium/DRTTestRunner.cpp:
749         (DRTTestRunner::DRTTestRunner):
750         (DRTTestRunner::reset):
751         * DumpRenderTree/chromium/DRTTestRunner.h:
752         (DRTTestRunner):
753         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
754         (WebTestInterfaces):
755         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
756         (WebKit):
757         (WebTestProxyBase):
758         (WebTestRunner):
759         (WebTestRunner::WebTestProxy::WebTestProxy):
760         (WebTestProxy):
761         (WebTestRunner::WebTestProxy::didInvalidateRect):
762         (WebTestRunner::WebTestProxy::didScrollRect):
763         (WebTestRunner::WebTestProxy::scheduleComposite):
764         (WebTestRunner::WebTestProxy::scheduleAnimation):
765         (WebTestRunner::WebTestProxy::setWindowRect):
766         (WebTestRunner::WebTestProxy::show):
767         (WebTestRunner::WebTestProxy::didAutoResize):
768         (WebTestRunner::WebTestProxy::postAccessibilityNotification):
769         (WebTestRunner::WebTestProxy::startDragging):
770         (WebTestRunner::WebTestProxy::shouldBeginEditing):
771         (WebTestRunner::WebTestProxy::shouldEndEditing):
772         (WebTestRunner::WebTestProxy::shouldInsertNode):
773         (WebTestRunner::WebTestProxy::shouldInsertText):
774         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
775         (WebTestRunner::WebTestProxy::shouldDeleteRange):
776         (WebTestRunner::WebTestProxy::shouldApplyStyle):
777         (WebTestRunner::WebTestProxy::didBeginEditing):
778         (WebTestRunner::WebTestProxy::didChangeSelection):
779         (WebTestRunner::WebTestProxy::didChangeContents):
780         (WebTestRunner::WebTestProxy::didEndEditing):
781         (WebTestRunner::WebTestProxy::registerIntentService):
782         (WebTestRunner::WebTestProxy::dispatchIntent):
783         (WebTestRunner::WebTestProxy::willPerformClientRedirect):
784         (WebTestRunner::WebTestProxy::didCancelClientRedirect):
785         (WebTestRunner::WebTestProxy::didStartProvisionalLoad):
786         (WebTestRunner::WebTestProxy::didReceiveServerRedirectForProvisionalLoad):
787         (WebTestRunner::WebTestProxy::didFailProvisionalLoad):
788         (WebTestRunner::WebTestProxy::didCommitProvisionalLoad):
789         (WebTestRunner::WebTestProxy::didReceiveTitle):
790         (WebTestRunner::WebTestProxy::didFinishDocumentLoad):
791         (WebTestRunner::WebTestProxy::didHandleOnloadEvents):
792         (WebTestRunner::WebTestProxy::didFailLoad):
793         (WebTestRunner::WebTestProxy::didFinishLoad):
794         (WebTestRunner::WebTestProxy::didChangeLocationWithinPage):
795         (WebTestRunner::WebTestProxy::didDisplayInsecureContent):
796         (WebTestRunner::WebTestProxy::didRunInsecureContent):
797         (WebTestRunner::WebTestProxy::didDetectXSS):
798         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
799         (WebTestRunner::WebTestRunner::setTestIsRunning):
800         (WebTestRunner::WebTestRunner::shouldDumpFrameLoadCallbacks):
801         (WebTestRunner::WebTestRunner::setShouldDumpFrameLoadCallbacks):
802         (WebTestRunner::WebTestRunner::shouldDumpUserGestureInFrameLoadCallbacks):
803         (WebTestRunner::WebTestRunner::stopProvisionalFrameLoads):
804         (WebTestRunner::WebTestRunner::shouldDumpTitleChanges):
805         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
806         (WebTestRunner::TestRunner::TestRunner):
807         (WebTestRunner::TestRunner::reset):
808         (WebTestRunner::TestRunner::setTestIsRunning):
809         (WebTestRunner):
810         (WebTestRunner::TestRunner::shouldDumpFrameLoadCallbacks):
811         (WebTestRunner::TestRunner::setShouldDumpFrameLoadCallbacks):
812         (WebTestRunner::TestRunner::shouldDumpUserGestureInFrameLoadCallbacks):
813         (WebTestRunner::TestRunner::stopProvisionalFrameLoads):
814         (WebTestRunner::TestRunner::shouldDumpTitleChanges):
815         (WebTestRunner::TestRunner::dumpFrameLoadCallbacks):
816         (WebTestRunner::TestRunner::dumpUserGestureInFrameLoadCallbacks):
817         (WebTestRunner::TestRunner::setStopProvisionalFrameLoads):
818         (WebTestRunner::TestRunner::dumpTitleChanges):
819         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
820         (TestRunner):
821         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
822         (WebTestInterfaces::Internal):
823         (WebTestRunner::WebTestInterfaces::Internal::webView):
824         (WebTestRunner::WebTestInterfaces::Internal::Internal):
825         (WebTestRunner::WebTestInterfaces::Internal::setWebView):
826         (WebTestRunner):
827         (WebTestRunner::WebTestInterfaces::Internal::setTestIsRunning):
828         (WebTestRunner::WebTestInterfaces::setWebView):
829         (WebTestRunner::WebTestInterfaces::setTestIsRunning):
830         (WebTestRunner::WebTestInterfaces::webView):
831         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
832         (WebTestRunner::WebTestProxyBase::willPerformClientRedirect):
833         (WebTestRunner):
834         (WebTestRunner::WebTestProxyBase::didCancelClientRedirect):
835         (WebTestRunner::WebTestProxyBase::didStartProvisionalLoad):
836         (WebTestRunner::WebTestProxyBase::didReceiveServerRedirectForProvisionalLoad):
837         (WebTestRunner::WebTestProxyBase::didFailProvisionalLoad):
838         (WebTestRunner::WebTestProxyBase::didCommitProvisionalLoad):
839         (WebTestRunner::WebTestProxyBase::didReceiveTitle):
840         (WebTestRunner::WebTestProxyBase::didFinishDocumentLoad):
841         (WebTestRunner::WebTestProxyBase::didHandleOnloadEvents):
842         (WebTestRunner::WebTestProxyBase::didFailLoad):
843         (WebTestRunner::WebTestProxyBase::didFinishLoad):
844         (WebTestRunner::WebTestProxyBase::didChangeLocationWithinPage):
845         (WebTestRunner::WebTestProxyBase::didDisplayInsecureContent):
846         (WebTestRunner::WebTestProxyBase::didRunInsecureContent):
847         (WebTestRunner::WebTestProxyBase::didDetectXSS):
848         * DumpRenderTree/chromium/TestShell.cpp:
849         (TestShell::runFileTest):
850         (TestShell::testFinished):
851         * DumpRenderTree/chromium/TestShell.h:
852         (TestShell):
853         * DumpRenderTree/chromium/WebViewHost.cpp:
854         (WebViewHost::didStartProvisionalLoad):
855         (WebViewHost::didReceiveServerRedirectForProvisionalLoad):
856         (WebViewHost::didFailProvisionalLoad):
857         (WebViewHost::didCommitProvisionalLoad):
858         (WebViewHost::didReceiveTitle):
859         (WebViewHost::didFailLoad):
860         (WebViewHost::didFinishLoad):
861         * DumpRenderTree/chromium/WebViewHost.h:
862         (WebViewHost):
863
864 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
865
866         [GTK] Lower the default timeout value for WebKitTestRunner
867         https://bugs.webkit.org/show_bug.cgi?id=106353
868
869         Reviewed by Dirk Pranke.
870
871         The default timeout value for WebKitTestRunner should be lowered.
872         The current value (80 seconds) is irrational and just leads to producing
873         text failures for layout tests that genuinely time out. This just leads
874         to more deviation from the GTK WK1 port in terms of test expectations.
875
876         The value used is the same as when running the DumpRenderTree, that's
877         6 seconds in release builds and 12 seconds in debug builds.
878
879         * Scripts/webkitpy/layout_tests/port/gtk.py:
880         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
881         (GtkPortTest.test_default_timeout_ms):
882
883 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
884
885         test-webkitpy shows failure output on my linux box
886         https://bugs.webkit.org/show_bug.cgi?id=101261
887
888         Reviewed by Eric Seidel.
889
890         Mock out the PulseAudioSanitizer methods when creating testing instances
891         of EflPort and GtkPort interfaces in unit tests. This stops throwing up
892         PulseAudio output when the sanitizer was trying to unload modules which
893         were already unloaded by another unit test that was being run in parallel.
894
895         * Scripts/webkitpy/layout_tests/port/efl_unittest.py:
896         (EflPortTest):
897         (EflPortTest.make_port):
898         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
899         (GtkPortTest):
900         (GtkPortTest.make_port):
901
902 2013-01-08  Julie Parent  <jparent@chromium.org>
903
904         Dashboard Cleanup: Remove g_hasDoneInitialPageGeneration
905         https://bugs.webkit.org/show_bug.cgi?id=106254
906
907         Remove the global variable g_hasDoneInitialPageGeneration.  It was used
908         only to figure out if we should call addBuilderLoadErrors during
909         handleLocationChange, which was called when the loader completed and on
910         any hash change, but we only ever wanted to call addBuilderLoadErrors
911         on the initial loader load.  Moved the call of addBuilderLoadErrors to
912         resourceLoadingComplete rather than handleLocationChange, so now we
913         only call it when we want to, and the check can go away completely.
914
915         Reviewed by Dirk Pranke.
916
917         * TestResultServer/static-dashboards/dashboard_base.js:
918         (addBuilderLoadErrors):
919         (resourceLoadingComplete):
920         (handleLocationChange):
921         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
922
923 2013-01-08  Dominic Mazzoni  <dmazzoni@google.com>
924
925         Add aboxhall@chromium.org to accessibility watchlist.
926
927         Unreviewed.
928
929         * Scripts/webkitpy/common/config/watchlist:
930
931 2013-01-08  Dan Winship  <danw@gnome.org>
932
933         Add Apache 2.4 support for Fedora 18
934         https://bugs.webkit.org/show_bug.cgi?id=104478
935
936         Reviewed by Martin Robinson.
937
938         Apache 2.4's httpd.conf is slightly incompatible with 2.2's, so we
939         need a separate config.
940
941         * Scripts/webkitpy/layout_tests/port/base.py:
942         (Port._apache_version):
943         (Port._apache_config_file_name_for_platform):
944         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
945         * Scripts/run-webkit-httpd:
946         * Scripts/webkitperl/httpd.pm:
947         (getApacheVersion):
948         (getDefaultConfigForTestDirectory):
949         (getHTTPDConfigPathForTestDirectory):
950
951 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
952
953         Make the override-feature-defines script compatible with Python v3
954         https://bugs.webkit.org/show_bug.cgi?id=105742
955
956         Reviewed by Martin Robinson.
957
958         Instead of printing output through a statement, use the print function and format
959         the output using the str.format method. This makes the script runnable with Python v3.
960
961         * gtk/override-feature-defines:
962         (adjust_feature_defines_makefile):
963
964 2013-01-08  Heikki Paajanen  <heikki.paajanen@palm.com>
965
966         [Qt] Fix libxml2 config test
967         https://bugs.webkit.org/show_bug.cgi?id=106317
968
969         Reviewed by Simon Hausmann.
970
971         Don't assume standard include/libxml2, libxml-2.0.pc should know
972         where it's headers are.
973
974         * qmake/config.tests/libxml2/libxml2.cpp: Fix include
975
976 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
977
978         [GTK] Make Tools/gtk/generate-gtkdoc compatible with Python 3
979         https://bugs.webkit.org/show_bug.cgi?id=106195
980
981         Reviewed by Philippe Normand.
982
983         Perform changes in the generate-gtkdoc script and common and gtkdoc modules
984         to make the Python code compatible with Python 3. This includes modifying
985         print statements, exception handling, dictionary iteration and
986         byte-sequence-to-string conversion.
987
988         * gtk/common.py:
989         (get_build_path):
990         (pkg_config_file_variable):
991         (gtk_version_of_pkg_config_file):
992         * gtk/generate-gtkdoc:
993         (print_missing_api):
994         * gtk/gtkdoc.py:
995         (GTKDoc.__init__):
996         (GTKDoc._run_command):
997
998 2013-01-02  Steve Block  <steveblock@chromium.org>
999
1000         Add chromium.org email address for Steve Block.
1001
1002         Unreviewed.
1003
1004         * Scripts/webkitpy/common/config/committers.py:
1005
1006 2013-01-07  Filip Pizlo  <fpizlo@apple.com>
1007
1008         Rationalize closure call heuristics and profiling
1009         https://bugs.webkit.org/show_bug.cgi?id=106270
1010
1011         Reviewed by Oliver Hunt.
1012         
1013         Add ability to use display-profiler-output via a pipe, and add the ability to dump
1014         all generated code ('display *' or 'd *').
1015
1016         * Scripts/display-profiler-output:
1017
1018 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
1019
1020         Sorted more xcodeproj files.
1021
1022         * MiniBrowser/MiniBrowser.xcodeproj/project.pbxproj:
1023         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1024         * WebKitLauncher/WebKitLauncher.xcodeproj/project.pbxproj:
1025
1026 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
1027
1028         Sorted xcodeproj files.
1029
1030         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1031         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
1032
1033 2013-01-07  Dan Bernstein  <mitz@apple.com>
1034
1035         [mac] WebKit1 clients can’t tell when a frame has been removed from the hierarchy
1036         https://bugs.webkit.org/show_bug.cgi?id=106261
1037
1038         Reviewed by Simon Fraser.
1039
1040         Added a test for -[WebFrameLoadDelegate webView:didRemoveFrameFromHierarchy:].
1041
1042         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1043         * TestWebKitAPI/Tests/mac/WebViewDidRemoveFrameFromHierarchy.mm: Added.
1044         (-[DidRemoveFrameFromHierarchyFrameLoadDelegate webView:didFinishLoadForFrame:]):
1045         (-[DidRemoveFrameFromHierarchyFrameLoadDelegate webView:didRemoveFrameFromHierarchy:]):
1046         (TestWebKitAPI):
1047         (TestWebKitAPI::TEST):
1048
1049 2013-01-07  Ojan Vafai  <ojan@chromium.org>
1050
1051         Add the Apple MountainLion bots to garden-o-matic.
1052         Eventually, we really should move away from hardcoding
1053         the list of bots.
1054
1055         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1056
1057 2013-01-07  Jochen Eisinger  <jochen@chromium.org>
1058
1059         [chromium] move web audio related methods to testrunner library
1060         https://bugs.webkit.org/show_bug.cgi?id=106211
1061
1062         Reviewed by Adam Barth.
1063
1064         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1065         (DRTTestRunner::DRTTestRunner):
1066         (DRTTestRunner::reset):
1067         * DumpRenderTree/chromium/DRTTestRunner.h:
1068         (DRTTestRunner):
1069         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
1070         (WebKit):
1071         (WebTestRunner::WebTestRunner::shouldDumpAsAudio):
1072         (WebTestRunner::WebTestRunner::audioData):
1073         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1074         (WebTestRunner::TestRunner::TestRunner):
1075         (WebTestRunner::TestRunner::reset):
1076         (WebTestRunner::TestRunner::shouldDumpAsAudio):
1077         (WebTestRunner):
1078         (WebTestRunner::TestRunner::audioData):
1079         (WebTestRunner::TestRunner::setAudioData):
1080         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1081         (TestRunner):
1082         * DumpRenderTree/chromium/TestShell.cpp:
1083         (TestShell::dump):
1084
1085 2013-01-07  Mike West  <mkwst@chromium.org>
1086
1087         Make the IFRAME_SEAMLESS flag runtime-enabled.
1088         https://bugs.webkit.org/show_bug.cgi?id=106213
1089
1090         Reviewed by Ojan Vafai.
1091
1092         * DumpRenderTree/chromium/TestShell.cpp:
1093         (TestShell::TestShell):
1094         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1095         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1096         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1097         (resetDefaultsToConsistentValues):
1098         * DumpRenderTree/mac/DumpRenderTree.mm:
1099         (resetDefaultsToConsistentValues):
1100         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1101         (WebCore::WebPage::resetSettings):
1102         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1103         (WTR::InjectedBundle::beginTesting):
1104             Ensure that seamless IFrames are enabled for layout tests in each of
1105             the ports that support seamless IFrames.
1106
1107 2013-01-07  Jochen Eisinger  <jochen@chromium.org>
1108
1109         [chromium] move dumpAsText and friends to the TestRunner library
1110         https://bugs.webkit.org/show_bug.cgi?id=106222
1111
1112         Reviewed by Adam Barth.
1113
1114         This moves dumpAsText, dumpChildFramesAsText, and dumpChildFrameScrollPositions to the TestRunner library.
1115
1116         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1117         (DRTTestRunner::DRTTestRunner):
1118         (DRTTestRunner::reset):
1119         * DumpRenderTree/chromium/DRTTestRunner.h:
1120         (DRTTestRunner):
1121         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
1122         (WebTestRunner::WebTestRunner::shouldDumpAsText):
1123         (WebTestRunner::WebTestRunner::setShouldDumpAsText):
1124         (WebTestRunner::WebTestRunner::shouldGeneratePixelResults):
1125         (WebTestRunner::WebTestRunner::setShouldGeneratePixelResults):
1126         (WebTestRunner::WebTestRunner::shouldDumpChildFrameScrollPositions):
1127         (WebTestRunner::WebTestRunner::shouldDumpChildFramesAsText):
1128         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1129         (WebTestRunner::TestRunner::TestRunner):
1130         (WebTestRunner::TestRunner::reset):
1131         (WebTestRunner::TestRunner::shouldDumpAsText):
1132         (WebTestRunner):
1133         (WebTestRunner::TestRunner::setShouldDumpAsText):
1134         (WebTestRunner::TestRunner::shouldGeneratePixelResults):
1135         (WebTestRunner::TestRunner::setShouldGeneratePixelResults):
1136         (WebTestRunner::TestRunner::shouldDumpChildFrameScrollPositions):
1137         (WebTestRunner::TestRunner::shouldDumpChildFramesAsText):
1138         (WebTestRunner::TestRunner::dumpAsText):
1139         (WebTestRunner::TestRunner::dumpChildFrameScrollPositions):
1140         (WebTestRunner::TestRunner::dumpChildFramesAsText):
1141         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1142         (TestRunner):
1143
1144 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
1145
1146         Build fix attempt after r138810.
1147
1148         Apparently python 2.6 is really bad at floating point computation.
1149
1150         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py:
1151         (MainTest.test_run_memory_test):
1152
1153 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
1154
1155         Extract a class to represent a test result to simplify PerfTest.parse_output and PageLoadPerfTest._run_with_driver
1156         https://bugs.webkit.org/show_bug.cgi?id=106172
1157
1158         Reviewed by Tony Chang.
1159
1160         Extracted PerfTestMetric, which represents a test result and knows how to compute statistics and convert itself
1161         to a dictionary. A metric is a type of measurement each test makes such as "Time", "Malloc", and "JSHeap".
1162         A single test (e.g. html5-full-render.html) may have multiple metrics, each of which is reported as a separate
1163         test on perf-o-matic and perf. test's results page.
1164
1165         Also, use python's compute_statistics instead of reading statistics off of DriverOutput to simplify the code.
1166
1167         * Scripts/webkitpy/performance_tests/perftest.py:
1168         (PerfTestMetric): Added.
1169         (PerfTestMetric.__init__): Added.
1170         (PerfTestMetric.metric): Added.
1171         (PerfTestMetric.has_values): Added.
1172         (PerfTestMetric.legacy_chromium_bot_compatible_test_name): Added. Following the legacy convention, "Malloc" and
1173         "JSHeap" will use "test_name:Malloc" whereas "Time" uses "test_name" itself. We should eventually move these
1174         conversions to be done on the client side (perf-o-matic and results page).
1175         (PerfTestMetric.append): Added.
1176         (PerfTestMetric.to_dict): Added. Computes the statistics and generates the dictionary object that represents
1177         a result for this metric.
1178         (PerfTestMetric.metric_to_unit): Added. Each metric maps to exactly one unit (it's not one-to-one).
1179         (PerfTestMetric.compute_statistics): Moved from PerfTest.
1180
1181         (PerfTest._create_driver): Extracted from PerfTest.run.
1182         (PerfTest.run): Given an array of PerfTestMetric, generate the results dictionary. Also print out results
1183         as needed.
1184         (PerfTest._run_with_driver):
1185         (PerfTest): Removed _result_classes and _result_class_regex. "result class" is the old name for "metric".
1186         (PerfTest.parse_output): Dramatically simplified. We now just return an array of PerfTestMetric's. Also ignore
1187         all keys except "values". We rely on PerfTestMetric to compute the statistics for us.
1188
1189         (ChromiumStylePerfTest.run): Added and merged _run_with_driver. Since chromium style perf tests are so different
1190         from other perf tests, there isn't much point in sharing the code here.
1191
1192         (PageLoadingPerfTest._run_with_driver): Simplified. Just create an array of PerfTestMetric's without computing
1193         statistics for the measured values.
1194
1195         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1196         (TestPerfTestMetric): Added.
1197         (TestPerfTestMetric.test_init_set_missing_unit): Added.
1198         (TestPerfTestMetric.test_legacy_chromium_bot_compatible_test_name): Added.
1199         (TestPerfTestMetric.test_has_values): Added.
1200         (TestPerfTestMetric.test_append): Added.
1201         (TestPerfTestMetric.test_compute_statistics.compute_statistics):
1202         (TestPerfTestMetric.test_compute_statistics): Added a trivial test case.
1203
1204         (TestPerfTest._assert_results_are_correct): Updated as needed.
1205         (TestPageLoadingPerfTest.test_run): Ditto. Notice that we no longer log anything in _run_with_driver. It's done
1206         in PerfTest.run now after generating the results dictionary.
1207         (TestPageLoadingPerfTest.test_run_with_memory_output): Ditto.
1208
1209 2013-01-07  Julien Chaffraix  <jchaffraix@webkit.org>
1210
1211         Support size_t multiplication and division operators on LayoutUnit
1212         https://bugs.webkit.org/show_bug.cgi?id=83848
1213
1214         Reviewed by Emil A Eklund.
1215
1216         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
1217         Added some simple tests for the new operators using size_t.
1218
1219 2013-01-07  Csaba Osztrogonác  <ossy@webkit.org>
1220
1221         [Qt] run-webkit-tests cleanup for parallel NRWT
1222         https://bugs.webkit.org/show_bug.cgi?id=106220
1223
1224         Reviewed by Dirk Pranke.
1225
1226         * Scripts/run-webkit-tests:
1227         (useNewRunWebKitTests):
1228         * Scripts/webkitpy/layout_tests/port/qt.py:
1229         (QtPort):
1230         (QtPort.default_child_processes):
1231
1232 2013-01-07  Philippe Normand  <pnormand@igalia.com>
1233
1234         Create a GStreamerAudio watchlist and add myself to it.
1235
1236         Unreviewed.
1237
1238         * Scripts/webkitpy/common/config/watchlist:
1239
1240 2013-01-05  Ryosuke Niwa  <rniwa@webkit.org>
1241
1242         Add svn:executable to a bunch of scripts in BuildSlaveSupport.
1243         Also removed the empty mac directory.
1244
1245         * BuildSlaveSupport/built-product-archive: Added property svn:executable.
1246         * BuildSlaveSupport/download-built-product: Added property svn:executable.
1247         * BuildSlaveSupport/mac: Removed.
1248         * BuildSlaveSupport/test-result-archive: Added property svn:executable.
1249
1250 2013-01-04  Ryosuke Niwa  <rniwa@webkit.org>
1251
1252         Add a script to delete stale zero-byte build files
1253         https://bugs.webkit.org/show_bug.cgi?id=106131
1254
1255         Reviewed by Tony Chang.
1256
1257         Delete zero-byte build files on Mac port for now. It doesn't do anything on other ports.
1258
1259         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1260         (DeleteStaleBuildFiles): Added.
1261         (Factory.__init__): Add DeleteStaleBuildFiles step.
1262         * BuildSlaveSupport/delete-stale-build-files: Added.
1263         (main): Delete zero-byte files except ones intentionally generated by our build process.
1264         (webkitBuildDirectory): Copied from build-product-archive.
1265
1266 2013-01-04  Michael Pruett  <michael@68k.org>
1267
1268         [GTK] Fix build error on GTK due to r138836
1269         https://bugs.webkit.org/show_bug.cgi?id=106154
1270
1271         Reviewed by Kentaro Hara.
1272
1273         * GNUmakefile.am:
1274
1275 2013-01-04  Dan Bernstein  <mitz@apple.com>
1276
1277         [mac] WebKit1 has no equivalent of WKBundleNodeHandleCopyHTMLTableCellElementCellAbove
1278         https://bugs.webkit.org/show_bug.cgi?id=106121
1279
1280         Reviewed by Adele Peterson.
1281
1282         Added test for -[DOMTHMLTableCellElement _cellAbove].
1283
1284         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1285         * TestWebKitAPI/Tests/mac/DOMHTMLTableCellCellAbove.mm: Added.
1286         (-[HTMLTableCellElementCellAboveTest webView:didFinishLoadForFrame:]):
1287         (TestWebKitAPI):
1288         (TestWebKitAPI::expectCellAboveCell):
1289         (TestWebKitAPI::TEST):
1290         * TestWebKitAPI/Tests/mac/DOMHTMLTableCellElementCellAbove.html: Added.
1291
1292 2013-01-04  Eric Seidel  <eric@webkit.org>
1293
1294         FlakyTestReporter should be re-enabled and taught how to post patches
1295         https://bugs.webkit.org/show_bug.cgi?id=103839
1296
1297         Reviewed by Adam Barth.
1298
1299         This moves LayoutTestResultsReader off of the DeprecatedPort object.
1300         It also adds a real Port object to the EWS and CommitQueue objects
1301         using a new PatchProcessingQueue superclass.
1302         PatchProcessingQueue doesn't seem to be the cleanest insertion point
1303         but works for this purpose.
1304         I also moved _upload_results_archive_for_patch into PatchProcessingQueue
1305         to resolve a FIXME.
1306
1307         This required some hacks/shiming between the non-specific "old"
1308         port names which the EWS system is currently using, and the new
1309         port objects.  This shimming is done in PatchProcessingQueue._new_port_name_from_old.
1310
1311         Hopefully this will fix the flaky test reporter and results uploading.
1312         Certainly it's another small step towards getting rid of the old ports.py
1313         infrastructure and making the NRWT Port class common webkitpy infrastructure.
1314
1315         Now that DeprecatedPort is no longer used for the results directory
1316         I removed all the explicit results-dir code which webkit-patch was using
1317         in --non-interactive mode to tell NRWT to use a results-dir it understood. :)
1318
1319         * Scripts/webkitpy/common/config/ports.py:
1320         (DeprecatedPort.run_perl_unittests_command):
1321         * Scripts/webkitpy/common/config/ports_mock.py:
1322         (MockPort.name):
1323         * Scripts/webkitpy/common/system/outputcapture.py:
1324         (OutputCapture.assert_outputs):
1325         * Scripts/webkitpy/layout_tests/port/apple.py:
1326         (ApplePort.determine_full_port_name):
1327         * Scripts/webkitpy/tool/bot/layouttestresultsreader.py:
1328         (LayoutTestResultsReader.__init__):
1329         (LayoutTestResultsReader):
1330         (LayoutTestResultsReader._read_file_contents):
1331         (LayoutTestResultsReader._create_layout_test_results):
1332         (LayoutTestResultsReader._create_unit_test_results):
1333         (LayoutTestResultsReader.results):
1334         (LayoutTestResultsReader.archive):
1335         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
1336         (LayoutTestResultsReaderTest.test_missing_layout_test_results):
1337         (LayoutTestResultsReaderTest.test_create_unit_test_results):
1338         (test_missing_unit_test_results_path):
1339         (test_layout_test_results):
1340         (test_archive_last_layout_test_results):
1341         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1342         (AbstractEarlyWarningSystem):
1343         (AbstractEarlyWarningSystem.__init__):
1344         (AbstractEarlyWarningSystem.begin_work_queue):
1345         (AbstractEarlyWarningSystem.run_command):
1346         * Scripts/webkitpy/tool/commands/queues.py:
1347         (AbstractPatchQueue.work_item_log_path):
1348         (PatchProcessingQueue):
1349         (PatchProcessingQueue._new_port_name_from_old):
1350         (PatchProcessingQueue.begin_work_queue):
1351         (PatchProcessingQueue._upload_results_archive_for_patch):
1352         (CommitQueue):
1353         (CommitQueue.begin_work_queue):
1354         (CommitQueue.run_command):
1355         (AbstractReviewQueue):
1356         (AbstractReviewQueue.__init__):
1357         (AbstractReviewQueue.begin_work_queue):
1358         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1359         (AbstractPatchQueueTest.test_next_patch):
1360         (PatchProcessingQueueTest):
1361         (PatchProcessingQueueTest.test_upload_results_archive_for_patch):
1362         * Scripts/webkitpy/tool/steps/runtests.py:
1363         (RunTests.run):
1364         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
1365
1366 2013-01-04  Dana Jansens  <danakj@chromium.org>
1367
1368         [chromium] Remove lost-compositor-context layout tests
1369         https://bugs.webkit.org/show_bug.cgi?id=106089
1370
1371         Reviewed by James Robinson.
1372
1373         Remove the loseCompositorContext() test function.
1374
1375         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1376         (WebTestRunner::TestRunner::TestRunner):
1377         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1378         (TestRunner):
1379
1380 2013-01-04  Adam Klein  <adamk@chromium.org>
1381
1382         Remove ENABLE_MUTATION_OBSERVERS #define
1383         https://bugs.webkit.org/show_bug.cgi?id=105459
1384
1385         Reviewed by Ryosuke Niwa.
1386
1387         * Scripts/webkitperl/FeatureList.pm:
1388         * qmake/mkspecs/features/features.pri:
1389
1390 2013-01-04  Ryosuke Niwa  <rniwa@webkit.org>
1391
1392         Statistics used in perftest_unittest.py and perftest_integrationtest.py are bogus
1393         https://bugs.webkit.org/show_bug.cgi?id=106078
1394
1395         Reviewed by Tony Chang.
1396
1397         This patch updates values for individual iterations in dummy performance tests used in
1398         perftest_unittest.py and perftest_integrationtest.py so that they match up with
1399         statistics such as stdev and median reported in those tests. This change is required
1400         to resolve the bug 97510.
1401
1402         This patch also refactors integration tests by extracting *TestData classes that
1403         centralizes the definitions of each dummy performance test's text (DRT output),
1404         runner output, and parsed results (JSON) so that we may share more data between
1405         python tests.
1406
1407         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1408         (MainTest._assert_results_are_correct): Extracted from test_parse_output.
1409         Use self.assertAlmostEqual when comparing the standard deviation.
1410         (MainTest.test_parse_output):
1411         (test_parse_output_with_failing_line):
1412         (test_parse_output_with_description): Use _assert_results_are_correct.
1413         (test_parse_output_with_subtests): Ditto.
1414
1415         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py:
1416         (InspectorPassTestData): Extracted from tests below.
1417         (EventTargetWrapperTestData): Ditto.
1418         (SomeParserTestData): Ditto.
1419         (MemoryTestData): Ditto.
1420         (TestDriver.run_test): Moved test definitions out of the method.
1421         (MainTest._normalize_output): Truncate standard deviations at 5th decimal point.
1422         (MainTest._load_output_json): Extracted from create_runner.
1423         (MainTest.create_runner):
1424         (MainTest.test_run_test_set_for_parser_tests):
1425         (MainTest.test_run_memory_test):
1426         (MainTest._test_run_with_json_output):
1427         (MainTest.test_run_with_json_output):
1428         (MainTest.test_run_with_description):
1429         (MainTest.test_run_generates_json_by_default):
1430         (MainTest.test_run_merges_output_by_default):
1431         (MainTest.test_run_respects_reset_results):
1432         (MainTest.test_run_generates_and_show_results_page):
1433         (MainTest.test_run_with_slave_config_json):
1434         (MainTest.test_run_with_multiple_repositories):
1435
1436 2013-01-04  Martin Robinson  <mrobinson@igalia.com>
1437
1438         [GTK] [WebKit2] Clicking on a word that is selected in an editable area deletes it
1439         https://bugs.webkit.org/show_bug.cgi?id=106045
1440
1441         Reviewed by Carlos Garcia Campos.
1442
1443         Update test results to confirm that compositions without preedits do not trigger
1444         a composition confirmation.
1445
1446         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
1447         (TestWebKitAPI::verifyCanceledComposition):
1448
1449 2013-01-04  Sailesh Agrawal  <sail@chromium.org>
1450
1451         [Mac] [WK2] platform/mac/fast/events/numpad-keycode-mapping.html fails
1452         https://bugs.webkit.org/show_bug.cgi?id=105958
1453
1454         Reviewed by Alexey Proskuryakov.
1455
1456         Ported code from DumpRenderTree to WebKitTestRunner to map all numpad keys.
1457
1458         * WebKitTestRunner/mac/EventSenderProxy.mm:
1459         (KeyMappingEntry):
1460         (WTR):
1461         (WTR::EventSenderProxy::keyDown): Added all numpad keys.
1462
1463 2013-01-03  Chris Fleizach  <cfleizach@apple.com>
1464
1465         REGRESSION: [Mac] Intermittent crash in WTR::AccessibilityUIElement::isEqual
1466         https://bugs.webkit.org/show_bug.cgi?id=106073
1467
1468         Reviewed by Ryosuke Niwa.
1469
1470         Check that the element being compared to is not nil. 
1471         This crash is happening because one of the accessibility tests does an isEqual comparison within a callback function.
1472         It looks like occasionally, one of the elements passed in to the isEqual is nil because the notification handler is on the accessibility
1473         controller, and the notification is being delivered after the document objects are cleared. This shouldn't happen but I've seen it 
1474         sometimes before on parallel WebKitTestRunner runs.
1475
1476         * DumpRenderTree/AccessibilityUIElement.cpp:
1477         (AccessibilityUIElement::isEqual):
1478         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
1479         (WTR::AccessibilityUIElement::isEqual):
1480
1481 2013-01-03  Tim 'mithro' Ansell  <mithro@mithis.com>
1482
1483         Downloads can be flaky, so try downloading multiple times.
1484         https://bugs.webkit.org/show_bug.cgi?id=105976
1485
1486         Reviewed by Eric Seidel.
1487
1488         * Scripts/webkitpy/common/system/autoinstall.py:
1489         (AutoInstaller._download_to_stream):
1490
1491 2013-01-03  Eric Seidel  <eric@webkit.org>
1492
1493         Rename Tool.port to deprecated_port in preparation for removing more callers
1494         https://bugs.webkit.org/show_bug.cgi?id=106076
1495
1496         Reviewed by Ryosuke Niwa.
1497
1498         This resolves a long-standing FIXME to rename this method. :)
1499
1500         * Scripts/webkitpy/tool/bot/botinfo.py:
1501         (BotInfo.summary_text):
1502         * Scripts/webkitpy/tool/bot/irc_command.py:
1503         (Rollout._update_working_copy):
1504         * Scripts/webkitpy/tool/bot/layouttestresultsreader.py:
1505         (LayoutTestResultsReader._create_layout_test_results):
1506         (LayoutTestResultsReader._create_unit_test_results):
1507         (LayoutTestResultsReader._results_directory):
1508         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
1509         (test_missing_unit_test_results_path):
1510         * Scripts/webkitpy/tool/main.py:
1511         (WebKitPatch.deprecated_port):
1512         * Scripts/webkitpy/tool/mocktool.py:
1513         (MockTool.deprecated_port):
1514         * Scripts/webkitpy/tool/steps/build.py:
1515         (Build.build):
1516         * Scripts/webkitpy/tool/steps/checkstyle.py:
1517         (CheckStyle.run):
1518         * Scripts/webkitpy/tool/steps/commit.py:
1519         (Commit._check_test_expectations):
1520         * Scripts/webkitpy/tool/steps/preparechangelog.py:
1521         (PrepareChangeLog.run):
1522         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
1523         (PrepareChangeLogForDEPSRoll.run):
1524         * Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
1525         (PrepareChangeLogForRevert.run):
1526         * Scripts/webkitpy/tool/steps/runtests.py:
1527         (RunTests.run):
1528         * Scripts/webkitpy/tool/steps/steps_unittest.py:
1529         (StepsTest.test_runtests_args):
1530         * Scripts/webkitpy/tool/steps/update.py:
1531         (Update._update_command):
1532
1533 2013-01-03  Eric Seidel  <eric@webkit.org>
1534
1535         Add myself to the webkitpy watchlist.
1536
1537         Unreviewed.
1538
1539         * Scripts/webkitpy/common/config/watchlist:
1540
1541 2013-01-03  Ryosuke Niwa  <rniwa@webkit.org>
1542
1543         Cleanup perftest* tests and add a test for computing statistics
1544         https://bugs.webkit.org/show_bug.cgi?id=105685
1545
1546         Reviewed by Eric Seidel.
1547
1548         Add a test for PerfTest.compute_statistics (moved and renamed from PageLoadingPerfTest.calculate_statistics) and
1549         extracted perftestsrunner_itegrationtests.py from perftestsrunner_unittests.py.
1550
1551         Also fixed a bug in compute_statistics that the mean ('avg') value can have a large rounding errors in some cases.
1552
1553         * Scripts/webkitpy/performance_tests/perftest.py:
1554         (PerfTest.compute_statistics): Moved from PageLoadingPerfTest to prepare for the bug 97510. Also compute the mean
1555         directly from sorted_values instead of using the one from Knuth's online algorithm. This approach gives more
1556         accurate result for the mean.
1557         (PageLoadingPerfTest.run_single):
1558         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1559         (MainTest.test_compute_statistics):
1560         (MainTest.test_compute_statistics.compute_statistics): Added.
1561         (TestPageLoadingPerfTest.test_run): floatify values.
1562         (TestPageLoadingPerfTest.test_run_with_memory_output): Ditto. Also got rid of ".0" from mean values now that Python
1563         correctly recognizes them as integers.
1564
1565         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py: Copied from
1566         Tools/Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py.
1567         (TestDriver): Moved out of MainTest.
1568         (MainTest): Got rid of assertWritten and all unit tests.
1569         (MainTest._normalize_output): Renamed from normalizeFinishedTime to match the PEP8 naming convention.
1570         (MainTest.test_run_test_set_kills_drt_per_run.TestDriverWithStopCount):
1571         (MainTest.test_run_test_set_for_parser_tests):
1572         (MainTest.test_run_memory_test):
1573         (MainTest._test_run_with_json_output):
1574         (MainTest.test_run_generates_json_by_default):
1575         (MainTest.test_run_merges_output_by_default):
1576         (MainTest.test_run_respects_reset_results):
1577         (MainTest.test_run_generates_and_show_results_page): Use runner.load_output_json() instead of manually loading and
1578         parsing output JSON files. Just verify that the output path is correct instead.
1579         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1580         (MainTest): Removed all integration tests.
1581         (MainTest.create_runner): Simplified.
1582
1583 2013-01-03  Julie Parent  <jparent@chromium.org>
1584
1585         Dashboard cleanup: remove usage of global g_defaultBuilderName
1586         https://bugs.webkit.org/show_bug.cgi?id=106043
1587
1588         Use now public defaultBuilder() from the BuilderGroup, rather
1589         than global variable.
1590
1591         Also eliminates the setup() and initBuilders() functions since we
1592         no longer need to track the global.
1593
1594         Reviewed by Adam Barth.
1595
1596         * TestResultServer/static-dashboards/builders.js:
1597         (BuilderGroup.prototype.defaultBuilder):
1598         (BuilderGroup.prototype.master):
1599         * TestResultServer/static-dashboards/dashboard_base.js:
1600         (parseParameters):
1601         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1602         (updateDefaultBuilderState):
1603         * TestResultServer/static-dashboards/loader.js:
1604         (.):
1605         * TestResultServer/static-dashboards/loader_unittests.js:
1606
1607 2013-01-03  Ryosuke Niwa  <rniwa@webkit.org>
1608
1609         Disable an intermittently failing unit test for now.
1610         The failure is tracked by the bug 106051.
1611
1612         * TestWebKitAPI/Tests/mac/MemoryCachePruneWithinResourceLoadDelegate.mm:
1613         (TestWebKitAPI::TEST):
1614
1615 2013-01-03  Dan Bernstein  <mitz@apple.com>
1616
1617         No way to obtain a DOMNode given a JS wrapper for a Node
1618         https://bugs.webkit.org/show_bug.cgi?id=106033
1619
1620         Added a test for +[DOMNode _nodeFromJSWrapper:].
1621
1622         Reviewed by Sam Weinig.
1623
1624         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1625         * TestWebKitAPI/Tests/mac/DOMNodeFromJSObject.mm: Added.
1626         (TestWebKitAPI):
1627         (TestWebKitAPI::TEST):
1628
1629 2013-01-03  Emil A Eklund  <eae@chromium.org>
1630
1631         Fix overflow in LayoutUnit::ceil and floor for SATURATED_LAYOUT_ARITHMETIC
1632         https://bugs.webkit.org/show_bug.cgi?id=105961
1633
1634         Reviewed by Levi Weintraub.
1635         
1636         Add tests for LayoutUnit::ceil and floor.
1637
1638         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
1639         (TestWebKitAPI::TEST):
1640         (TestWebKitAPI):
1641
1642 2013-01-03  Julie Parent  <jparent@chromium.org>
1643
1644         Add unit test for default builder for the dashboards.
1645         https://bugs.webkit.org/show_bug.cgi?id=105972
1646
1647         Reviewed by Dirk Pranke.
1648
1649         * TestResultServer/static-dashboards/loader_unittests.js:
1650
1651 2013-01-03  Tony Chang  <tony@chromium.org>
1652
1653         Generate internal.settings from Settings.in
1654         https://bugs.webkit.org/show_bug.cgi?id=104740
1655
1656         Reviewed by Adam Barth.
1657
1658         * GNUmakefile.am: Compile new generated files on GTK+.
1659         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
1660         with both extra_sources and add_output_to_sources=false will still
1661         compile the files in extra_sources. This makes it possible to compile
1662         InternalSettingsGenerated.cpp and not be confused by
1663         InternalSettingsGenerated.idl.
1664
1665 2013-01-03  Zan Dobersek  <zandobersek@gmail.com>
1666
1667         [style] WebIDL-reflecting uppercase enums reported as style violation
1668         https://bugs.webkit.org/show_bug.cgi?id=105832
1669
1670         Reviewed by Tony Chang.
1671
1672         Recognize the enum as WebIDL-reflecting one if there's a comment present that
1673         indicates this. In enums spanning over multiple lines, the comment is expected
1674         on the first line. In enums placed in one line, the comment is expected at the
1675         end of that line.
1676
1677         WebIDL-reflecting enum values are required to be written in uppercase and as such
1678         violate the common rule of these values being written in InterCaps. Annotating
1679         enums this way makes it possible to avoid false style violation reports.
1680
1681         The comment is of form '// Web(?:Kit)?IDL enum'.
1682
1683         * Scripts/webkitpy/style/checkers/cpp.py:
1684         (_EnumState.__init__):
1685         (_EnumState.process_clean_line):
1686         (check_enum_casing):
1687         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1688         (NoNonVirtualDestructorsTest.test_enum_casing):
1689
1690 2013-01-03  Tim 'mithro' Ansell  <mithro@mithis.com>
1691
1692         Adding favicon.ico to the garden-o-matic.
1693         https://bugs.webkit.org/show_bug.cgi?id=105982
1694
1695         Reviewed by Adam Barth.
1696
1697         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1698         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon.ico:
1699
1700 2013-01-03  Jonathan Liu  <net147@gmail.com>
1701
1702         [Qt] Fix library linking order
1703         https://bugs.webkit.org/show_bug.cgi?id=105818
1704
1705         Some Qt libraries are linked before the private static libraries
1706         that depend on them which results in QtWebKit failing to link.
1707
1708         Copy Qt module dependencies from QT to QT_PRIVATE when creating
1709         module to ensure the Qt libraries are linked after the private
1710         static libraries that depend on them.
1711
1712         Reviewed by Tor Arne Vestbø.
1713
1714         * qmake/mkspecs/features/default_post.prf:
1715
1716 2013-01-02  Byungwoo Lee  <bw80.lee@samsung.com>
1717
1718         [EFL][WK2] Fix tooltip bugs in MiniBrowser.
1719         https://bugs.webkit.org/show_bug.cgi?id=102209
1720
1721         Reviewed by Gyuyoung Kim.
1722
1723         Set tooltip on the elm_window instead of ewk_view, and add delay to
1724         show the tooltip window.
1725
1726         Setting tooltip on the ewk_view makes elementary error message because
1727         ewk_view is not an elementary widget. To remove the error message,
1728         tooltip should be set on the elm_window.
1729
1730         Most browsers show the tooltip after a mouse pointer stays on an
1731         element object that has tooltip text.
1732         But elm_tooltip widget itself doesn't have delay or handle mouse
1733         events to show/hide.
1734
1735         So additional codes are added to show/hide tooltip widget along with
1736         mouse events.
1737           1) When a mouse pointer stays for a second on an element which has
1738              tooltip text, then tooltip widget is displayed.
1739           2) When a mouse pointer is moved or out of ewk_view, hide the tooltip
1740              widget.
1741
1742         * MiniBrowser/efl/main.c:
1743         (_Tooltip_Information):
1744         (_Browser_Window):
1745         (on_tooltip_show):
1746         (window_tooltip_hide):
1747         (window_tooltip_update):
1748         (on_mouse_in):
1749         (on_mouse_move):
1750         (on_mouse_out):
1751         (window_free):
1752         (on_tooltip_text_set):
1753         (on_tooltip_text_unset):
1754         (window_create):
1755
1756 2013-01-02  Tony Chang  <tony@chromium.org>
1757
1758         Unreviewed, rolling out r138661.
1759         http://trac.webkit.org/changeset/138661
1760         https://bugs.webkit.org/show_bug.cgi?id=104740
1761
1762         Compile problems on EFL
1763
1764         * GNUmakefile.am:
1765         * qmake/mkspecs/features/default_post.prf:
1766
1767 2013-01-02  Tony Chang  <tony@chromium.org>
1768
1769         Generate internal.settings from Settings.in
1770         https://bugs.webkit.org/show_bug.cgi?id=104740
1771
1772         Reviewed by Adam Barth.
1773
1774         * GNUmakefile.am: Compile new generated files on GTK+.
1775         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
1776         with both extra_sources and add_output_to_sources=false will still
1777         compile the files in extra_sources. This makes it possible to compile
1778         InternalSettingsGenerated.cpp and not be confused by
1779         InternalSettingsGenerated.idl.
1780
1781 2013-01-02  Adam Barth  <abarth@webkit.org>
1782
1783         Remove old-run-webkit-tests support from webkitpy
1784         https://bugs.webkit.org/show_bug.cgi?id=105619
1785
1786         Reviewed by Eric Seidel.
1787
1788         This patch removes support for webkitpy parsing the
1789         old-run-webkit-tests results format. There doesn't seem to be any
1790         reason to keep this code around anymore now that most bots have
1791         switched to new-run-webkit-tests.
1792
1793         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1794         (Builder.fetch_layout_test_results):
1795         * Scripts/webkitpy/common/net/layouttestresults.py:
1796         (path_for_layout_test):
1797         (LayoutTestResults.results_from_string):
1798         (LayoutTestResults):
1799         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
1800         (LayoutTestResultsTest.test_results_from_string):
1801         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
1802         (test_flaky_test_failure):
1803         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
1804         (LayoutTestResultsReaderTest.test_missing_layout_test_results):
1805         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
1806         (PatchAnalysisTask._test_patch):
1807         * Scripts/webkitpy/tool/commands/queuestest.py:
1808         (QueuesTest.assert_queue_outputs):
1809
1810 2013-01-02  Heikki Paajanen  <heikki.paajanen@palm.com>
1811
1812         [Qt][WK2] Add simple UI to find text from page
1813         https://bugs.webkit.org/show_bug.cgi?id=96481
1814
1815         Reviewed by Jocelyn Turcotte.
1816
1817         Find bar that slides between navigation bar and WebView.
1818
1819         * MiniBrowser/qt/BrowserWindow.cpp:
1820         (BrowserWindow::toggleFind):
1821         * MiniBrowser/qt/BrowserWindow.h:
1822         (BrowserWindow):
1823         * MiniBrowser/qt/MiniBrowser.qrc:
1824         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1825         (MiniBrowserApplication::notify):
1826         * MiniBrowser/qt/icons/find.png: Added.
1827         * MiniBrowser/qt/qml/BrowserWindow.qml:
1828
1829 2013-01-02  Christophe Dumez  <christophe.dumez@intel.com>
1830
1831         Regression(r138603): Caused 2 webkitpy tests to fail
1832         https://bugs.webkit.org/show_bug.cgi?id=105913
1833
1834         Reviewed by Kentaro Hara.
1835
1836         Fix 2 webkitpy tests that started to fail after
1837         r138603.
1838
1839         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1840         (IRCCommandTest._sheriff_test_data_url): Make sure the
1841         returned test file URL is valid no matter what the
1842         current working directory is.
1843         * Scripts/webkitpy/tool/bot/ircbot_unittest.py:
1844         (IRCBotTest.test_help): Add "sheriffs" command to the
1845         list of available commands.
1846
1847 2013-01-01  Christophe Dumez  <christophe.dumez@intel.com>
1848
1849         [WK2] plugins/clicking-missing-plugin-fires-delegate.html is failing
1850         https://bugs.webkit.org/show_bug.cgi?id=105897
1851
1852         Reviewed by Alexey Proskuryakov.
1853
1854         Dump in WebKitTestRunner the message that is expected by layout tests
1855         when the missing plugin button is clicked.
1856
1857         * WebKitTestRunner/TestController.cpp:
1858         (WTR::TestController::createWebViewWithOptions):
1859         (WTR::TestController::unavailablePluginButtonClicked):
1860         (WTR):
1861         * WebKitTestRunner/TestController.h:
1862         (TestController):
1863
1864 2013-01-01  KwangYong Choi  <ky0.choi@samsung.com>
1865
1866         [EFL] Enable MHTML feature
1867         https://bugs.webkit.org/show_bug.cgi?id=105815
1868
1869         Reviewed by Laszlo Gombos.
1870
1871         Enable MHTML feature for EFL port by default.
1872
1873         * Scripts/webkitperl/FeatureList.pm:
1874
1875 2013-01-01  Alan Cutter  <alancutter@chromium.org>
1876
1877         sheriff-bot should know who the gardeners/sheriffs are
1878         https://bugs.webkit.org/show_bug.cgi?id=105698
1879
1880         Reviewed by Eric Seidel.
1881
1882         Added a "sheriffs" command to sheriffbot to fetch the current Chromium Webkit sheriffs.
1883
1884         * Scripts/webkitpy/tool/bot/irc_command.py:
1885         (Current):
1886         (Current._retrieve_webkit_sheriffs):
1887         (Current.execute):
1888         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1889         (IRCCommandTest.test_current):
1890         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_0.js: Added.
1891         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_1.js: Added.
1892         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_2.js: Added.
1893         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_malformed.js: Added.
1894         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_zero.js: Added.
1895
1896 2012-12-31  Carlos Garcia Campos  <cgarcia@igalia.com>
1897
1898         [GTK] Move ImageDiff to the Tools directory root
1899         https://bugs.webkit.org/show_bug.cgi?id=105421
1900
1901         Reviewed by Kenneth Rohde Christiansen.
1902
1903         * DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Removed.
1904         * GNUmakefile.am:
1905         * ImageDiff/gtk/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/gtk/ImageDiff.cpp.
1906
1907 2012-12-30  Mike West  <mkwst@chromium.org>
1908
1909         Add myself to watchlists for CSP and console logs.
1910         https://bugs.webkit.org/show_bug.cgi?id=105874
1911
1912         Reviewed by Dirk Pranke.
1913
1914         Because I really, really, really need more email.
1915
1916         * Scripts/webkitpy/common/config/committers.py:
1917             Add mkwst+watchlist@chromium.org as a contributor.
1918         * Scripts/webkitpy/common/config/watchlist:
1919             Create watchlist categories for 1) the various mechanisms used to
1920             generate console messages, and 2) CSP usage and files.
1921
1922 2012-12-30  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1923
1924         [EFL] Enable TEMPLATE_ELEMENT feature
1925         https://bugs.webkit.org/show_bug.cgi?id=105865
1926         
1927         Reviewed by Laszlo Gombos.
1928
1929         * Scripts/webkitperl/FeatureList.pm: Enable template element feature for EFL port.
1930
1931 2012-12-30  Zan Dobersek  <zandobersek@gmail.com>
1932
1933         Unreviewed GTK gardening of the unit tests.
1934
1935         Skip the TestInspectorServer unit test that's timing out completely.
1936
1937         * Scripts/run-gtk-tests:
1938         (TestRunner):
1939
1940 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
1941
1942         Web Inspector: Native Memory Instrumentation plugin: do not generate 'not instrumented' warning for instrumented mutable members.
1943         https://bugs.webkit.org/show_bug.cgi?id=105855
1944
1945         Reviewed by Vsevolod Vlasov.
1946
1947         Extract MemberExpr from ImplicitCastExpr. It happens when we instrument a mutable member because
1948         addMember expects const T& and the mutable ref to member implicitly converts into const ref to member.
1949
1950         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
1951         (ReportMemoryUsageVisitor):
1952         (clang::AddMemberCallVisitor::VisitCallExpr):
1953         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
1954         (clang::ReportMemoryUsageVisitor::checkMembersCoverage):
1955
1956 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
1957
1958         Web Inspector: Native Memory Instrumentation plugin: move function bodies out of class declarations.
1959         https://bugs.webkit.org/show_bug.cgi?id=105852
1960
1961         Reviewed by Alexander Pavlov.
1962
1963         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
1964         (clang):
1965         (AddMemberCallVisitor):
1966         (ReportMemoryUsageVisitor):
1967         (clang::ReportMemoryUsageAction::ParseArgs):
1968         (clang::AddMemberCallVisitor::VisitCallExpr):
1969         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
1970         (clang::ReportMemoryUsageVisitor::emitWarning):
1971         (clang::ReportMemoryUsageVisitor::findInstrumentationMethod):
1972         (clang::ReportMemoryUsageVisitor::needsToBeInstrumented):
1973         (clang::ReportMemoryUsageVisitor::CheckMembersCoverage):
1974
1975 2012-12-29  Zan Dobersek  <zandobersek@gmail.com>
1976
1977         [webkitpy] Omit webkitpy/thirdparty/BeautifulSoup.py from code coverage checking
1978         https://bugs.webkit.org/show_bug.cgi?id=105835
1979
1980         Reviewed by Dirk Pranke.
1981
1982         The BeautifulSoup module is a third-party import and as such we don't have much control
1983         over its testing. Because of that it should be omitted from code coverage checking.
1984
1985         * Scripts/webkitpy/test/main.py:
1986         (Tester._run_tests):
1987
1988 2012-12-28  Martin Robinson  <mrobinson@igalia.com>
1989
1990         [GTK][WK2] Add support for IME Composition
1991         https://bugs.webkit.org/show_bug.cgi?id=65093
1992
1993         Reviewed by Carlos Garcia Campos.
1994
1995         Update unit tests for GtkInputMethodFilter to note whether or not an event
1996         was a fake event synthesized for composition purposes. This is now used
1997         for WebKit2.
1998
1999         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
2000         (TestWebKitAPI::TestInputMethodFilter::sendSimpleKeyEvent):
2001         (TestWebKitAPI::TestInputMethodFilter::sendKeyEventWithCompositionResults):
2002         (TestWebKitAPI::TEST):
2003         * gtk/generate-gtkdoc:
2004         (get_webkit2_options): Skip new files in the WebKit2 source directory for documentation
2005         generation.
2006
2007 2012-12-27  Ilya Tikhonovsky  <loislo@chromium.org>
2008
2009         Web Inspector: Native Memory Instrumentation. Update clang plugin according to the current state of memory instrumentation code.
2010         https://bugs.webkit.org/show_bug.cgi?id=105800
2011
2012         Reviewed by Yury Semikhatsky.
2013
2014         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
2015         (clang::ReportMemoryUsageConsumer::ReportMemoryUsageConsumer):
2016         (clang):
2017
2018 2012-12-27  Zan Dobersek  <zandobersek@gmail.com>
2019
2020         Create a GTK build system watchlist and add myself to it
2021         https://bugs.webkit.org/show_bug.cgi?id=105769
2022
2023         Reviewed by Dirk Pranke.
2024
2025         Add a GTK build system watchlist, watching the configure.ac, GNUmakefile.am files
2026         and the input GNUmakefile.features.am.in file. The GNUmakefile.list.am files are
2027         omitted on purpose due to relatively high change rate they are exposed to.
2028
2029         The new watchlist would be of use to me as the GTK build system recently went
2030         through a series of changes that included a changed policy on adding configuration
2031         flags to configure.ac and subsequent additions to GNUmakefile.am files. A PSA on the
2032         mailing list about this change will be published later, but I'd also like to be
2033         notified of any pending changes to these files just so I can check the changes are
2034         in compliance with the new policy.
2035
2036         * Scripts/webkitpy/common/config/watchlist:
2037
2038 2012-12-26  Ilya Tikhonovsky  <loislo@chromium.org>
2039
2040         Unreviewed. Replace find("Tools") with rfind("Tools") because base dir of WebKit may have Tools word.
2041         As example ..../DevTools/src/third_party/WebKit
2042
2043         * Scripts/webkitpy/common/webkit_finder.py:
2044         (WebKitFinder.webkit_base):
2045
2046 2012-12-25  Christophe Dumez  <christophe.dumez@intel.com>
2047
2048         [EFL][WK2] Refactor snapshot taking code
2049         https://bugs.webkit.org/show_bug.cgi?id=105687
2050
2051         Reviewed by Kenneth Rohde Christiansen.
2052
2053         Call WKViewCreateSnapshot() instead of WKViewGetSnapshot()
2054         since the function was renamed to follow naming style.
2055
2056         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
2057         (WTR::PlatformWebView::windowSnapshotImage):
2058
2059 2012-12-24  Ilya Tikhonovsky  <loislo@chromium.org>
2060
2061         Web Inspector: Native Memory Instrumentation: propagate member type as edge type to the serialized heap graph.
2062         https://bugs.webkit.org/show_bug.cgi?id=105725
2063
2064         Reviewed by Yury Semikhatsky.
2065
2066         MemberType value names were adjusted according to Style Guide.
2067         Existing tests were extended with link type validation.
2068
2069         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2070
2071 2012-12-24  Xiaobo Wang  <xbwang@torchmobile.com.cn>
2072
2073         [BlackBerry] DRT - Crashed when running pixels tests
2074         https://bugs.webkit.org/show_bug.cgi?id=105711
2075
2076         RIM PR 270003
2077
2078         Reviewed by George Staikos.
2079
2080         Calling window->post() here will result in buffer swapping which is
2081         only allowed in UI thread now, and I found we don't need to call it
2082         for GL renderer.
2083
2084         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
2085         (createBitmapContextFromWebView):
2086
2087 2012-12-22  Dan Winship  <danw@gnome.org>
2088
2089         Add a patch to fix libxml2 2.9.0 regression
2090         https://bugs.webkit.org/show_bug.cgi?id=104494
2091
2092         Reviewed by Martin Robinson.
2093
2094         * gtk/jhbuild.modules:
2095         * gtk/patches/libxml2-2.9.0-dtd.patch: patch from upstream
2096
2097 2012-12-23  Allan Sandfeld Jensen  <allan.jensen@digia.com>
2098
2099         Unreviewed update of status to reviewer.
2100
2101         * Scripts/webkitpy/common/config/committers.py:
2102
2103 2012-12-23  Philippe Normand  <pnormand@igalia.com>
2104
2105         [GTK][jhbuild] Bump GStreamer to 1.0.4
2106         https://bugs.webkit.org/show_bug.cgi?id=105640
2107
2108         Reviewed by Martin Robinson.
2109
2110         * gtk/jhbuild.modules: Update GStreamer modules to 1.0.4.
2111
2112 2012-12-21  Nico Weber  <thakis@chromium.org>
2113
2114         chromium: webkit-build-directory doesn't handle ninja output dirs
2115         https://bugs.webkit.org/show_bug.cgi?id=105671
2116
2117         Reviewed by Daniel Bates.
2118
2119         Teach webkit-build-directory to detect the ninja build directory.
2120
2121         * Scripts/webkitdirs.pm:
2122         (determineBaseProductDir):
2123         (determineIsChromiumNinja):
2124
2125 2012-12-21  Ryosuke Niwa  <rniwa@webkit.org>
2126
2127         [Mountain Lion] platform/mac/editing/spelling/removing-underline-after-accepting-autocorrection-using-punctuation.html failing
2128         https://bugs.webkit.org/show_bug.cgi?id=95477
2129
2130         Reviewed by Sam Weinig.
2131
2132         Co-authored by Jia Pu.
2133
2134         Specify a fixed autocorrection dictionary in resetDefaultsToConsistentValues() so that
2135         we can have consistent autocorrection behavior when running tests.
2136
2137         * DumpRenderTree/mac/DumpRenderTree.mm:
2138         (resetDefaultsToConsistentValues):
2139         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
2140         (WTR::InjectedBundle::platformInitialize):
2141
2142 2012-12-21  Zan Dobersek  <zandobersek@gmail.com>
2143
2144         [GTK] Add WebKit1-specific TestExpectations file
2145         https://bugs.webkit.org/show_bug.cgi?id=97562
2146
2147         Reviewed by Dirk Pranke.
2148
2149         Implement the expectations_files method inside the GtkPort class that
2150         returns the list of the TestExpectations files the current test configuration
2151         requires. The method is implemented in a similar way as on the EFL port.
2152         A unit test is also added.
2153
2154         * Scripts/webkitpy/layout_tests/port/gtk.py:
2155         (GtkPort._search_paths):
2156         (GtkPort):
2157         (GtkPort.expectations_files):
2158         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
2159         (GtkPortTest.test_expectations_files):
2160
2161 2012-12-21  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
2162
2163         [EFL] Enable feature option for the CSS sticky position
2164         https://bugs.webkit.org/show_bug.cgi?id=105630
2165
2166         Reviewed by Kenneth Rohde Christiansen.
2167
2168         Enable CSS sticky position feature by default for the
2169         EFL port. This feature option was added in r138252.
2170
2171         No new tests. Existing layout tests pass.
2172
2173         * Scripts/webkitperl/FeatureList.pm:
2174
2175 2012-12-21  Jussi Kukkonen  <jussi.kukkonen@intel.com>
2176
2177         [WK2] REGRESSION(r138232): It made 4 tests timeout
2178         https://bugs.webkit.org/show_bug.cgi?id=105623
2179
2180         Reviewed by Kenneth Rohde Christiansen.
2181
2182         r138232 changed functionality so that empty text output
2183         no longer leads to a dump() call, which leads to a timeout.
2184         Instead do a text dump if there is text output or if there
2185         is no audio output.
2186
2187         * WebKitTestRunner/TestInvocation.cpp:
2188         (WTR::TestInvocation::dumpResults):
2189
2190 2012-12-21  Zeno Albisser  <zeno@webkit.org>
2191
2192         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
2193         https://bugs.webkit.org/show_bug.cgi?id=105309
2194
2195         On Windows the library name is Qt5WebKitWidgets as well.
2196         So r138136 should not have adjusted the library name for Windows.
2197
2198         Reviewed by Simon Hausmann.
2199
2200         * Scripts/webkitdirs.pm:
2201         (builtDylibPathForName):
2202
2203 2012-12-21  Zeno Albisser  <zeno@webkit.org>
2204
2205         [Qt] Add new TestExpectations for qt-5.0-mac-wk2.
2206         https://bugs.webkit.org/show_bug.cgi?id=105424
2207
2208         This change will allow us to run a very specific
2209         subset of the available LayoutTests on a mac buildbot.
2210         The new baseline search path for mac will look as follows:
2211         qt-5.0-mac-wk2 -> qt-5.0-wk2 -> qt-5.0 -> qt-mac -> qt
2212
2213         Reviewed by Dirk Pranke.
2214
2215         * Scripts/webkitpy/layout_tests/port/qt.py:
2216         (QtPort._search_paths):
2217         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2218         (QtPortTest):
2219
2220 2012-12-20  Ryuan Choi  <ryuan.choi@samsung.com>
2221
2222         [EFL] Build break with latest EFL libraries.
2223         https://bugs.webkit.org/show_bug.cgi?id=104827
2224
2225         Reviewed by Laszlo Gombos.
2226
2227         The eo EFL package is introduced and evas and ecore use it since 1.8.
2228         While introducing Eo, EFL changed several structures of Evas and Ecore
2229         from own specific class to Eo.
2230
2231         So, this patch adds FindEo and changes declaration of Evas, Evas_Object,
2232         Ecore_Timer to build with latest EFL libraries.
2233
2234         * DumpRenderTree/efl/CMakeLists.txt: Includes eo EFL library.
2235         * EWebLauncher/CMakeLists.txt: Ditto.
2236         * MiniBrowser/efl/CMakeLists.txt: Ditto.
2237         * ImageDiff/PlatformEfl.cmake: Ditto.
2238         * TestWebKitAPI/PlatformEfl.cmake: Ditto.
2239         * TestWebKitAPI/PlatformWebView.h: Modified declaration of Evas_Object.
2240         * WebKitTestRunner/InjectedBundle/TestRunner.h: Modified declaration of Ecore_Timer.
2241         * WebKitTestRunner/PlatformEfl.cmake: Includes eo EFL library.
2242         * WebKitTestRunner/PlatformWebView.h: Modified declaration of Evas_Object.
2243
2244 2012-12-20  Dirk Pranke  <dpranke@chromium.org>
2245
2246         TestExpectation fallback is broken
2247         https://bugs.webkit.org/show_bug.cgi?id=105583
2248
2249         Reviewed by Ryosuke Niwa.
2250
2251         When I added support for the -wk2 in port names, it looks
2252         like I broke the way we computed the list of TestExpectations
2253         files so that the implementation-version directory was left out :(.
2254         Fixed and cleaned up the tests.
2255
2256         * Scripts/webkitpy/layout_tests/port/base.py:
2257         (Port.expectations_files):
2258         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2259         (TestWebKitPort.__init__):
2260         (PortTestCase.test_skipped_directories_for_symbols):
2261         (test_skipped_directories_for_features):
2262         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
2263         (test_skipped_tests_for_unsupported_features_empty_test_list):
2264         (test_skipped_layout_tests):
2265         (test_expectations_files):
2266
2267 2012-12-20  Anders Carlsson  <andersca@apple.com>
2268
2269         Don't include WebKit2 headers when building TestWebKitAPI on windows
2270         https://bugs.webkit.org/show_bug.cgi?id=105582
2271
2272         Reviewed by Tim Horton.
2273
2274         * TestWebKitAPI/config.h:
2275         * TestWebKitAPI/win/main.cpp:
2276
2277 2012-12-20  Ryosuke Niwa  <rniwa@webkit.org>
2278
2279         REGRESSION(r133820?): SimplifyMarkupTest API test asserts
2280         https://bugs.webkit.org/show_bug.cgi?id=105370
2281
2282         Reviewed by Simon Fraser.
2283
2284         Re-enable the test.
2285
2286         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
2287         (TestWebKitAPI::TEST):
2288
2289 2012-12-20  Nico Weber  <thakis@chromium.org>
2290
2291         chromium nrwt: Pick the newest binary found in DEFAULT_BUILD_DIRECTORIES, not the first
2292         https://bugs.webkit.org/show_bug.cgi?id=105498
2293
2294         Reviewed by Dirk Pranke.
2295
2296         Use the newest binary available rather than an than always picking one
2297         build directory over another based on iteration order.
2298
2299         * Scripts/webkitpy/layout_tests/port/chromium.py:
2300         (ChromiumPort._static_build_path):
2301         Check for timestamps.
2302         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
2303         (ChromiumMacPortTest.test_build_path_timestamps):
2304         Test that out / xcodebuild selection happens based on timestamps
2305         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2306         (ChromiumWinPortTest.test_build_path_timestamps):
2307         Test that out / build selection happens based on timestamps
2308
2309 2012-12-19  Simon Fraser  <simon.fraser@apple.com>
2310
2311         Lots of sticky tests failing in WK2
2312         https://bugs.webkit.org/show_bug.cgi?id=105464
2313
2314         Reviewed by Beth Dakin.
2315
2316         WebKitTestRunner had a race between snapshotting in the UI process,
2317         and resettting after the test in the web process. InjectedBundle::done()
2318         was a bad place to call page()->resetAfterTest(), because of this race;
2319         it could reset the scroll position before the UI snapshot had been obtained.
2320         
2321         Fix by moving the call to page()->resetAfterTest() into didReceiveMessage(),
2322         for the "Reset" message which will come in before the next test.
2323
2324         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2325         (WTR::InjectedBundle::didReceiveMessage):
2326         (WTR::InjectedBundle::done):
2327
2328 2012-12-19  Filip Pizlo  <fpizlo@apple.com>
2329
2330         DFG speculation checks that take JumpList should consolidate OSRExits
2331         https://bugs.webkit.org/show_bug.cgi?id=105401
2332
2333         Reviewed by Oliver Hunt.
2334
2335         Update the profiler to understand that an OSR exit may have multiple
2336         candidate exit sites.
2337
2338         * Scripts/display-profiler-output:
2339
2340 2012-12-20  Eric Seidel  <eric@webkit.org>
2341
2342         EWSes do not need to dump build/test logs when things pass
2343         https://bugs.webkit.org/show_bug.cgi?id=105402
2344
2345         Reviewed by Dirk Pranke.
2346
2347         When we were first writing the commit-queue and EWS system
2348         and running them locally, we thought it was important to see the
2349         build/test output as it happened.  Now these queues have been running
2350         for years, and we don't use this feature anymore.  So this patch
2351         makes us stop tee-ing the output directly to stdout, and instead
2352         only print any output from commands when they fail.
2353
2354         I also added some logging when we execute each command
2355         since we are no longer printing the command output itself.
2356         This should make it much easier to see errors with the EWS
2357         system itself, now that all the passing-build output is removed.
2358
2359         This required updating a zillion unittests, since we have a bunch of
2360         "integration" style testing of the queue-bots.  Many of the unittests for the bots were
2361         already approximating this behavior by using a "logging" Executive.
2362         I disabled logging in those cases to avoid redundant logging in those tests.
2363
2364         * Scripts/webkitpy/common/system/executive.py:
2365         (Executive.command_for_printing):
2366         (Executive.run_command):
2367         * Scripts/webkitpy/common/system/executive_mock.py:
2368         (MockExecutive.command_for_printing):
2369         (MockExecutive.run_command):
2370         * Scripts/webkitpy/common/system/executive_unittest.py:
2371         (ExecutiveTest.test_auto_stringify_args):
2372         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2373         (EarlyWarningSytemTest._default_expected_logs):
2374         (_test_ews):
2375         (_test_ewses):
2376         * Scripts/webkitpy/tool/commands/queues.py:
2377         (AbstractQueue.run_webkit_patch):
2378         * Scripts/webkitpy/tool/commands/queues_unittest.py:
2379         (AbstractQueueTest._assert_run_webkit_patch):
2380         (CommitQueueTest.test_commit_queue):
2381         (test_rollout):
2382         (test_rollout_lands):
2383         (test_manual_reject_during_processing):
2384         (StyleQueueTest.test_style_queue_with_style_exception):
2385         (test_style_queue_with_watch_list_exception):
2386
2387 2012-12-20  Dominic Mazzoni  <dmazzoni@google.com>
2388
2389         AX: support clickPoint in DRT for chromium
2390         https://bugs.webkit.org/show_bug.cgi?id=97359
2391
2392         Reviewed by Chris Fleizach.
2393
2394         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
2395         several more tests can pass.
2396
2397         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2398         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2399         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
2400         (WebTestRunner):
2401         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
2402         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
2403         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2404         (AccessibilityUIElement):
2405
2406 2012-12-20  Dan Bernstein  <mitz@apple.com>
2407
2408         prepare-ChangeLog adds "Additional information of the change such as approach, rationale" where it doesn't make sense
2409         https://bugs.webkit.org/show_bug.cgi?id=99268
2410
2411         Reviewed by Alexey Proskuryakov.
2412
2413         * Scripts/prepare-ChangeLog:
2414         (generateNewChangeLogs): Removed the "Additional information of the change" line.
2415
2416 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
2417
2418         Unreviewed. Skip unit test /webkit/atk/getTextInParagraphAndBodyModerate
2419
2420         * Scripts/run-gtk-tests:
2421         (TestRunner):
2422
2423 2012-12-20  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2424
2425         [Qt] Move ImageDiff to the Tools directory root
2426         https://bugs.webkit.org/show_bug.cgi?id=105420
2427
2428         Reviewed by Tor Arne Vestbø.
2429
2430         Moved ImageDiff from outside DumpRenderTree because it is also a
2431         dependency when running pixel tests for WebKit2.
2432
2433         * ImageDiff/ImageDiff.pro: Renamed from Tools/DumpRenderTree/qt/ImageDiff.pro.
2434         * ImageDiff/qt/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/qt/ImageDiff.cpp.
2435         * Tools.pro:
2436
2437 2012-12-20  Zan Dobersek  <zandobersek@gmail.com>
2438
2439         [GTK] Remove the --enable-unstable-features configuration option
2440         https://bugs.webkit.org/show_bug.cgi?id=105327
2441
2442         Reviewed by Martin Robinson.
2443
2444         Add a missing feature option for the CSS sticky position feature.
2445
2446         * Scripts/webkitperl/FeatureList.pm:
2447
2448 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
2449
2450         Unreviewed. Skip unit test WebKit2.ForceRepaint that times out
2451
2452         * Scripts/run-gtk-tests:
2453         (TestRunner):
2454
2455 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
2456
2457         [EFL] MiniBrowser does not play Infinite Gangnam Style
2458         https://bugs.webkit.org/show_bug.cgi?id=103531
2459
2460         Reviewed by Kenneth Rohde Christiansen.
2461
2462         For EFL we would like this feature to be built by default.
2463
2464         * Scripts/webkitperl/FeatureList.pm:
2465
2466 2012-12-20  Csaba Osztrogonác  <ossy@webkit.org>
2467
2468         [Qt] EWS bots should call qmake always
2469         https://bugs.webkit.org/show_bug.cgi?id=90049
2470
2471         Reviewed by Simon Hausmann.
2472
2473         * Scripts/webkitdirs.pm:
2474         (buildQMakeProjects):
2475
2476 2012-12-20  Jochen Eisinger  <jochen@chromium.org>
2477
2478         [chromium] fix stray printf in TestRunner's WebTestProxy
2479         https://bugs.webkit.org/show_bug.cgi?id=105515
2480
2481         Reviewed by Ryosuke Niwa.
2482
2483         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
2484         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
2485
2486 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
2487
2488         [WK2] WebAudio WKTR support
2489         https://bugs.webkit.org/show_bug.cgi?id=95084
2490
2491         Reviewed by Kenneth Rohde Christiansen.
2492
2493         Implementing audio data dump support for WebKitTestRunner,
2494         by adding the API to the IDL, enabling IPC for transferring
2495         the audio data as bytebuffer and dumping it according
2496         to run-webkit-tests script's expectations.
2497         This enables audio testing on WebKit2 ports.
2498         Based on initial wip-patch from Philippe Normand - thanks.
2499
2500         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Adding setAudioData API.
2501         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Assign audio buffer data to a local WKDataRef after extracting it from WK2 using WKBundleDataFromUint8Array bundle API.
2502         (WTR::TestRunner::setAudioData):
2503         (WTR):
2504         * WebKitTestRunner/InjectedBundle/TestRunner.h: Add a whatToDump state for dumping Audio.
2505         (TestRunner):
2506         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Add audio to the message that's transferred to the WKTR process.
2507         (WTR::InjectedBundle::done):
2508         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: Allow setting and audio result handle to be transferred.
2509         (WTR::InjectedBundle::setAudioResult):
2510         (InjectedBundle):
2511         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2512         (WTR::InjectedBundlePage::dump): Dummy handling the Audio case in the switch, build fix for Mac.
2513         * WebKitTestRunner/TestInvocation.cpp: Audio dump support, dump audio if we haven't received a text dump from the web process.
2514         (WTR::TestInvocation::dumpResults):
2515         (WTR):
2516         (WTR::TestInvocation::dumpAudio):
2517         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Storing received audio data before dumping.
2518         * WebKitTestRunner/TestInvocation.h:
2519         (TestInvocation):
2520
2521 2012-12-20  Alberto Garcia  <agarcia@igalia.com>
2522
2523         [GTK] Count failed and timed out tests separately
2524         https://bugs.webkit.org/show_bug.cgi?id=105448
2525
2526         Reviewed by Carlos Garcia Campos.
2527
2528         Display the number of tests that fail, time out and are skipped
2529         separately in the buildbot page.
2530
2531         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2532         (RunGtkAPITests.commandComplete):
2533         (RunGtkAPITests.evaluateCommand):
2534         (RunGtkAPITests.getText2):
2535         * Scripts/run-gtk-tests:
2536         (TestRunner.run_tests):
2537
2538 2012-12-19  Ryosuke Niwa  <rniwa@webkit.org>
2539
2540         PerfTest.parse_output does too much
2541         https://bugs.webkit.org/show_bug.cgi?id=105391
2542
2543         Reviewed by Tony Chang.
2544
2545         Moved the code to filter lines into PerfTest._filter_output, which has been renamed from _filter_stderr.
2546         Also moved the code to output test results into PerfTest._run_with_driver.
2547
2548         * Scripts/webkitpy/performance_tests/perftest.py:
2549         (PerfTest.__init__): Added _description.
2550         (PerfTest.description): Added.
2551         (PerfTest._run_with_driver): Moved the code to output test description and test results.
2552         (PerfTest._filter_output): Moved and renamed from PerfTest._filter_stderr.
2553         (PerfTest.parse_output): Removed the code to output test results. Return test results and description
2554         respectively. Also removed the code that allowed some tests to omit values since all tests report each
2555         iteration now since r136492.
2556         (PerfTest.output_statistics): Removed the code to print test description, now done in _run_with_driver.
2557         (ChromiumStylePerfTest._run_with_driver): Added. Chromium style tests are sufficiently different from
2558         regular PerfTest that it doesn't make much sense to share _run_with_driver. But really, we should just
2559         get rid of this type of test altogether in favor of regular performance tests that uses runner.js.
2560         (ChromiumStylePerfTest.parse_and_log_output): Renamed from parse_output.
2561         (PageLoadingPerfTest._run_with_driver): Removed the explicit '' for the test description.
2562         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
2563         (MainTest.test_parse_output): Removed the expected logs since parse_output no longer prints out results.
2564         Also added a call to _filter_output since parse_output doesn't filter the output text anymore.
2565         (MainTest.test_parse_output_with_failing_line): Added a call to _filter_output. Also added ',' after 'Time:'
2566         so that it's not string-concatenated with the next line.
2567         (MainTest.test_parse_output_with_description): Added; a test for PerfTest.description().
2568         (MainTest.test_parse_output_with_subtests): Removed the expected logs, and added a call to _filter_output.
2569
2570 2012-12-19  Alexis Menard  <alexis@webkit.org>
2571
2572         Implement CSS parsing for CSS transitions unprefixed.
2573         https://bugs.webkit.org/show_bug.cgi?id=104804
2574
2575         Reviewed by Dean Jackson.
2576
2577         Add a new flag ENABLE_CSS_TRANSFORMS_ANIMATIONS_TRANSITIONS_UNPREFIXED
2578         to cover the work of unprefixing Transforms, Animations and 
2579         Transitions. It will let the possibility of each ports to turn it off 
2580         in their release branches until we're confident that these CSS 
2581         properties are ready to be unprefixed.
2582
2583         * Scripts/webkitperl/FeatureList.pm:
2584         * qmake/mkspecs/features/features.pri:
2585
2586 2012-12-19  Julien Chaffraix  <jchaffraix@webkit.org>
2587
2588         2 clampTo unit tests are failing
2589         https://bugs.webkit.org/show_bug.cgi?id=105253
2590
2591         Reviewed by Emil A Eklund.
2592
2593         Removed clampToUnsigned as it was actually testing exactly what
2594         clampToUnsignedLong was doing. Also made clampToIntFloat more robust
2595         to precision loss (fix suggested by Emil).
2596
2597         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
2598
2599 2012-12-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2600
2601         [EFL] Move ImageDiff to the Tools directory root
2602         https://bugs.webkit.org/show_bug.cgi?id=105305
2603
2604         Reviewed by Laszlo Gombos.
2605
2606         We previously moved this directory from the DRT directory to
2607         efl/ImageDiff. The major reason for doing so was because ImageDiff does
2608         not depend on DumpRenderTree but for legacy reasons was living inside the
2609         DRT folder, even though it is needed when running pixel tests for WK2.
2610
2611         Now we are moving to Tool directory root as other ports agreed to do the same.
2612
2613         * CMakeLists.txt:
2614         * ImageDiff/CMakeLists.txt: Added.
2615         * ImageDiff/PlatformEfl.cmake: Added.
2616         * ImageDiff/efl/ImageDiff.cpp: Renamed from Tools/efl/ImageDiff/ImageDiff.cpp.
2617         * efl/ImageDiff/CMakeLists.txt: Removed.
2618
2619 2012-12-19  Zeno Albisser  <zeno@webkit.org>
2620
2621         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
2622         https://bugs.webkit.org/show_bug.cgi?id=105309
2623
2624         The version number is only used as a part
2625         of the library name on Linux.
2626
2627         Reviewed by Kenneth Rohde Christiansen.
2628
2629         * Scripts/webkitdirs.pm:
2630         (builtDylibPathForName):
2631
2632 2012-12-19  Alberto Garcia  <agarcia@igalia.com>
2633
2634         [GTK] Tests that time out are not considered as failures.
2635         https://bugs.webkit.org/show_bug.cgi?id=105318
2636
2637         Reviewed by Martin Robinson.
2638
2639         Fix parsing of the GTK API tests log files.
2640
2641         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2642         (RunGtkAPITests.commandComplete):
2643
2644 2012-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
2645
2646         Unreviewed, rolling out r138061.
2647         https://bugs.webkit.org/show_bug.cgi?id=105396
2648
2649         Original patch:
2650         http://trac.webkit.org/changeset/138061
2651         https://bugs.webkit.org/show_bug.cgi?id=97359
2652
2653         It is crashing on Debug bots
2654
2655         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2656         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2657         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
2658         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2659         (AccessibilityUIElement):
2660
2661 2012-12-18  Alice Liu  <alice.liu@apple.com>
2662
2663         Test for https://bugs.webkit.org/show_bug.cgi?id=105106
2664         Add SPI to WebKit1 WebFrame for hit testing
2665
2666         Reviewed by Dan Bernstein.
2667
2668         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/ElementAtPointInWebFrame.mm
2669         * TestWebKitAPI/Tests/mac/ElementAtPointInWebFrame.mm: Added.
2670         (TestWebKitAPI::TEST): Loads html with two divs positioned in the 2nd and 4th quadrants of the webview.
2671         Then hit-tests at three points, expecting to hit the two divs and body element. 
2672
2673 2012-12-18  Alice Liu  <alice.liu@apple.com>
2674
2675         Test for https://bugs.webkit.org/show_bug.cgi?id=105262
2676         Add SPI to WebKit1 WebFrame for node conversion to JSValueRef
2677
2678         Reviewed by Anders Carlsson.
2679
2680         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/JSWrapperForNodeInWebFrame.mm.mm
2681         * TestWebKitAPI/Tests/mac/JSWrapperForNodeInWebFrame.mm.mm: Added.
2682         (TestWebKitAPI::TEST): Tests for the correct JS wrapper for a DOMNode, provided a WebScriptWorld.
2683         Loads html with a single div element. In an isolated world, creates a property on that node.
2684         Also, in the standard world, creates a different property on that node.  Then tests for 4 things:
2685             - Existence of the isolated property in the isolated world.
2686             - Existence of the standard property in the standard world.
2687             - Non-existence of the isolated property in the standard world.
2688             - Non-existence of the standard property in the isolated world.
2689
2690 2012-12-18  Simon Fraser  <simon.fraser@apple.com>
2691
2692         Disable the SimplifyMarkup API test, wich asserts in debug builds.
2693         Tracked by https://bugs.webkit.org/show_bug.cgi?id=105370.
2694         
2695         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
2696         (TestWebKitAPI::TEST):
2697
2698 2012-12-18  Joseph Pecoraro  <pecoraro@apple.com>
2699
2700         [Mac] Add Build Phase to Check Headers for Inappropriate Macros (Platform.h macros)
2701         https://bugs.webkit.org/show_bug.cgi?id=104279
2702
2703         Reviewed by David Kilzer.
2704
2705         Add a build phase script that checks a framework's header files for
2706         Platform.h and Compiler.h macros. Also add some tests for the script.
2707
2708         * Scripts/check-for-inappropriate-macros-in-external-headers: Added.
2709         Script to be used in build phases to check for inappropriate macros in headers.
2710
2711         * Scripts/test-webkitruby: Added.
2712         Script that can be used to run all webkitruby tests.
2713
2714         * Scripts/test-webkit-scripts:
2715         Also run the ruby tests.
2716
2717         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/fake-data-failing-expected.txt: Added.
2718         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/pass-expected.txt: Added.
2719         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Fail.h: Added.
2720         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Pass.h: Added.
2721         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/run-test.rb: Added.
2722         Tests for the new script.
2723
2724 2012-12-18  Dominic Mazzoni  <dmazzoni@google.com>
2725
2726         AX: support clickPoint in DRT for chromium
2727         https://bugs.webkit.org/show_bug.cgi?id=97359
2728
2729         Reviewed by Chris Fleizach.
2730
2731         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
2732         several more tests can pass.
2733
2734         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2735         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2736         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
2737         (WebTestRunner):
2738         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
2739         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
2740         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2741         (AccessibilityUIElement):
2742
2743 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
2744
2745         Build fix after r138045.
2746
2747         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2748         (PerfTestsRunner._collect_tests): Use relative_path. We skip only when this path
2749         is not explicitly specified. The condition was backwards but never matched so we
2750         always ran the test regardless of whether it's listed in the skipped list or not.
2751         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2752         (test_collect_tests_with_ignored_skipped_list): Renamed so that it actually runs.
2753
2754 2012-12-18  Emil A Eklund  <eae@chromium.org>
2755
2756         TestWebKitAPI/Tests/WebCore/LayoutUnit fails on mac
2757         https://bugs.webkit.org/show_bug.cgi?id=105332
2758         
2759         Reviewed by Levi Weintraub.
2760         
2761         Create new test suite for LayoutUnit tests as the KURL tests are
2762         of the incompatible TEST_F type.
2763
2764         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
2765         (TestWebKitAPI):
2766         (TestWebKitAPI::TEST_F):
2767
2768 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
2769
2770         Running a skipped test with run-perf-tests could alert the user
2771         https://bugs.webkit.org/show_bug.cgi?id=105219
2772
2773         Reviewed by Eric Seidel.
2774
2775         Force run skipped performance tests explicitly specified in the command line.
2776         Also warn the user when the specified paths don't exist.
2777
2778         * Scripts/webkitpy/layout_tests/port/base.py:
2779         (Port.relative_test_filename):
2780         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2781         (PerfTestsRunner._parse_args):
2782         (PerfTestsRunner._collect_tests):
2783         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2784         (test_collect_tests_with_skipped_list_and_files):
2785
2786 2012-12-18  Yaron Friedman  <yfriedman@chromium.org>
2787
2788         Chrome Android java output is too verbose
2789         https://bugs.webkit.org/show_bug.cgi?id=105247
2790
2791         Reviewed by Tony Chang.
2792
2793         Pass -quiet to ant to suppress extraneous output.
2794
2795         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2796         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
2797
2798 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
2799
2800         Consider removing --pause-before-testing option
2801         https://bugs.webkit.org/show_bug.cgi?id=105248
2802
2803         Reviewed by Eric Seidel.
2804
2805         Remove --pause-before-testing option.
2806
2807         * Tools/Scripts/webkitpy/layout_tests/port/driver.py:
2808         (DriverProxy.start): Removed.
2809         * Scripts/webkitpy/performance_tests/perftest.py:
2810         (PerfTest.__init__):
2811         (PerfTest.test_path): Renamed from path_or_url.
2812         (PerfTest.run):
2813         (PerfTest): Removed "driver" argument.
2814         (PerfTest._run_with_driver):
2815         (PerfTest.run_single):
2816         (ChromiumStylePerfTest.__init__):
2817         (PageLoadingPerfTest.__init__):
2818         (PageLoadingPerfTest.run_single):
2819         (PageLoadingPerfTest._run_with_driver): Renamed from run so that PerfTest.run can create a driver for us.
2820         (ReplayPerfTest.__init__):
2821         (ReplayPerfTest.prepare):
2822         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
2823         (TestPageLoadingPerfTest.test_run):
2824         (TestPageLoadingPerfTest.test_run_with_memory_output):
2825         (TestPageLoadingPerfTest.test_run_with_bad_output):
2826         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2827         (PerfTestsRunner._parse_args):
2828         (PerfTestsRunner._run_tests_set): Removed the code to create a driver and pause before running the test.
2829         (PerfTestsRunner._run_single_test): Removed "driver" argument.
2830         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2831         (run_test):
2832         (test_run_test_pause_before_testing): Removed.
2833
2834 2012-12-18  Alberto Garcia  <agarcia@igalia.com>
2835
2836         [GTK] Tests that time out are not considered as failures.
2837         https://bugs.webkit.org/show_bug.cgi?id=105318
2838
2839         Reviewed by Martin Robinson.
2840
2841         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2842         (RunGtkAPITests.commandComplete):
2843         Consider log lines that end with TIMEOUT as errors.
2844
2845         * Scripts/run-gtk-tests:
2846         (TestRunner.run_tests):
2847         Consider timeouts as failures when returning the number of failed
2848         tests.
2849
2850 2012-12-18  Pravin D  <pravin.d@samsung.com>
2851
2852         Unreviewed, adding secondary email.
2853
2854         * Scripts/webkitpy/common/config/committers.py:
2855
2856 2012-12-18  Pravin D  <pravin.d@samsung.com>
2857
2858         Unreviewed, promoting myself to committer's list.
2859
2860         * Scripts/webkitpy/common/config/committers.py:
2861
2862 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2863
2864         [EFL] ImageDiff does not build on WebKit2 only builds
2865         https://bugs.webkit.org/show_bug.cgi?id=105288
2866
2867         Reviewed by Kenneth Rohde Christiansen.
2868
2869         Moved ImageDiff to the efl/ folder inside Tools/ (so we don't pollute
2870         the folder with things which are EFL-specific.
2871
2872         ImageDiff will be built by default now regardless of WebKit API flavor
2873         you choose.
2874
2875         * CMakeLists.txt:
2876         * DumpRenderTree/efl/CMakeLists.txt:
2877         * efl/ImageDiff/CMakeLists.txt: Added.
2878         * efl/ImageDiff/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/efl/ImageDiff.cpp.
2879
2880 2012-12-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2881
2882         [EFL][WK2] When creating new windows, use the opener's context
2883         https://bugs.webkit.org/show_bug.cgi?id=105283
2884
2885         Reviewed by Laszlo Gombos.
2886
2887         Use the context of the opener if exists.
2888
2889         * MiniBrowser/efl/main.c:
2890         (on_key_down):
2891         (on_window_create):
2892         (window_create):
2893         (elm_main):
2894
2895 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2896
2897         [EFL][WK2] Gardening of TestWebKitAPI tests
2898         https://bugs.webkit.org/show_bug.cgi?id=105268
2899
2900         Reviewed by Kenneth Rohde Christiansen.
2901
2902         Some tests were marked as failing but are passing and the newly added
2903         LayoutUnit should have it's own binary because it conflicts with the
2904         fixture used by KURL and is failing on the bots because of that.
2905
2906         * TestWebKitAPI/CMakeLists.txt:
2907         * TestWebKitAPI/PlatformEfl.cmake:
2908
2909 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
2910
2911         [chromium] move webintent related methods to TestRunner library
2912         https://bugs.webkit.org/show_bug.cgi?id=105164
2913
2914         Reviewed by Darin Fisher.
2915
2916         Also deleted the unused deliveredIntentFailure and
2917         deliveredIntentResult methods.
2918
2919         * DumpRenderTree/chromium/DRTTestRunner.cpp:
2920         (DRTTestRunner::DRTTestRunner):
2921         * DumpRenderTree/chromium/DRTTestRunner.h:
2922         (DRTTestRunner):
2923         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
2924         (WebKit):
2925         (WebTestRunner::WebTestDelegate::setCurrentWebIntentRequest):
2926         (WebTestRunner::WebTestDelegate::currentWebIntentRequest):
2927         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
2928         (WebKit):
2929         (WebTestProxyBase):
2930         (WebTestRunner::WebTestProxy::registerIntentService):
2931         (WebTestRunner::WebTestProxy::dispatchIntent):
2932         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
2933         (WebTestRunner::TestRunner::TestRunner):
2934         (WebTestRunner::TestRunner::sendWebIntentResponse):
2935         (WebTestRunner):
2936         (WebTestRunner::TestRunner::deliverWebIntent):
2937         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
2938         (TestRunner):
2939         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
2940         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
2941         (WebTestRunner::WebTestProxyBase::registerIntentService):
2942         (WebTestRunner):
2943         (WebTestRunner::WebTestProxyBase::dispatchIntent):
2944         * DumpRenderTree/chromium/WebViewHost.cpp:
2945         (WebViewHost::setCurrentWebIntentRequest):
2946         (WebViewHost::currentWebIntentRequest):
2947         * DumpRenderTree/chromium/WebViewHost.h:
2948         (WebKit):
2949         (WebViewHost):
2950
2951 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
2952
2953         [chromium] move editing callback dumping to the WebTestProxy
2954         https://bugs.webkit.org/show_bug.cgi?id=105013
2955
2956         Reviewed by Darin Fisher.
2957
2958         To decide whether or not to dump the editing callbacks, the WebTestProxy
2959         needs to query the TestRunner. For this purpose, I introduce a
2960         WebTestRunner class that just exposes the shouldDumpEditingCallbacks
2961         method. In contrast to e.g. WebEventSender, the WebTestRunner is just
2962         an interface, and it is not owned by the WebTestInterfaces, but the
2963         embedder of the TestRunner library must provide it. This will change
2964         once all TestRunner methods are in the library.
2965
2966         * DumpRenderTree/DumpRenderTree.gypi:
2967         * DumpRenderTree/chromium/DRTTestRunner.cpp:
2968         (DRTTestRunner::DRTTestRunner):
2969         (DRTTestRunner::reset):
2970         * DumpRenderTree/chromium/DRTTestRunner.h:
2971         (DRTTestRunner):
2972         (DRTTestRunner::taskList):
2973         (DRTTestRunner::WorkQueue::taskList):
2974         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
2975         (WebTestRunner):
2976         (WebTestInterfaces):
2977         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
2978         (WebKit):
2979         (WebTestRunner):
2980         (WebTestProxyBase):
2981         (WebTestRunner::WebTestProxy::shouldBeginEditing):
2982         (WebTestRunner::WebTestProxy::shouldEndEditing):
2983         (WebTestRunner::WebTestProxy::shouldInsertNode):
2984         (WebTestRunner::WebTestProxy::shouldInsertText):
2985         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
2986         (WebTestRunner::WebTestProxy::shouldDeleteRange):
2987         (WebTestRunner::WebTestProxy::shouldApplyStyle):
2988         (WebTestRunner::WebTestProxy::didBeginEditing):
2989         (WebTestRunner::WebTestProxy::didChangeSelection):
2990         (WebTestRunner::WebTestProxy::didChangeContents):
2991         (WebTestRunner::WebTestProxy::didEndEditing):
2992         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h: Copied from Tools/DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h.
2993         (WebTestRunner):
2994         (WebTestRunner::WebTestRunner::shouldDumpEditingCallbacks):
2995         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
2996         (WebTestRunner::TestRunner::TestRunner):
2997         (WebTestRunner::TestRunner::~TestRunner):
2998         (WebTestRunner):
2999         (WebTestRunner::TestRunner::reset):
3000         (WebTestRunner::TestRunner::shouldDumpEditingCallbacks):
3001         (WebTestRunner::TestRunner::dumpEditingCallbacks):
3002         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
3003         (TestRunner):
3004         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
3005         (WebTestRunner::WebTestInterfaces::Internal::testRunner):
3006         (WebTestRunner::WebTestInterfaces::Internal::setTestRunner):
3007         (WebTestInterfaces::Internal):
3008         (WebTestRunner::WebTestInterfaces::Internal::Internal):
3009         (WebTestRunner::WebTestInterfaces::testRunner):
3010         (WebTestRunner):
3011         (WebTestRunner::WebTestInterfaces::setTestRunner):
3012         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
3013         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
3014         (WebTestRunner):
3015         (WebTestRunner::WebTestProxyBase::shouldBeginEditing):
3016         (WebTestRunner::WebTestProxyBase::shouldEndEditing):
3017         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
3018         (WebTestRunner::WebTestProxyBase::shouldInsertText):
3019         (WebTestRunner::WebTestProxyBase::shouldChangeSelectedRange):
3020         (WebTestRunner::WebTestProxyBase::shouldDeleteRange):
3021         (WebTestRunner::WebTestProxyBase::shouldApplyStyle):
3022         (WebTestRunner::WebTestProxyBase::didBeginEditing):
3023         (WebTestRunner::WebTestProxyBase::didChangeSelection):
3024         (WebTestRunner::WebTestProxyBase::didChangeContents):
3025         (WebTestRunner::WebTestProxyBase::didEndEditing):
3026         * DumpRenderTree/chromium/TestShell.cpp:
3027         (TestShell::initialize):
3028         (TestShell::createNewWindow):
3029         * DumpRenderTree/chromium/WebViewHost.cpp:
3030         (WebViewHost::shouldBeginEditing):
3031         (WebViewHost::shouldEndEditing):
3032         (WebViewHost::shouldInsertNode):
3033         (WebViewHost::shouldInsertText):
3034         (WebViewHost::shouldChangeSelectedRange):
3035         (WebViewHost::shouldDeleteRange):
3036         (WebViewHost::shouldApplyStyle):
3037         * DumpRenderTree/chromium/WebViewHost.h:
3038         (WebViewHost):
3039
3040 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
3041
3042         [chromium] Fix methods in TestRunner that directly modify the settings instead of going through WebPreferences
3043         https://bugs.webkit.org/show_bug.cgi?id=105154
3044
3045         Reviewed by Tony Chang.
3046
3047         If the settings are modified directly, the embedder can't keep track of
3048         them and might accidentially override the changed settings.
3049
3050         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
3051         (WebPreferences):
3052         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
3053         (WebTestRunner::TestRunner::TestRunner):
3054         (WebTestRunner::TestRunner::setAsynchronousSpellCheckingEnabled):
3055         (WebTestRunner):
3056         (WebTestRunner::TestRunner::setMinimumTimerInterval):
3057         (WebTestRunner::TestRunner::setTouchDragDropEnabled):
3058         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
3059         (TestRunner):
3060         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
3061         (WebTestRunner::WebPreferences::reset):
3062         (WebTestRunner::WebPreferences::applyTo):
3063
3064 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
3065
3066         Unreviewed unit test skipping.
3067
3068         Filed https://bugs.webkit.org/show_bug.cgi?id=105253 to re-enable them.
3069
3070         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
3071         (TestWebKitAPI):
3072         (TestWebKitAPI::TEST):
3073
3074 2012-12-17  Dean Jackson  <dino@apple.com>
3075
3076         Unreviewed. Add Silvia Pfeiffer to contributor list.
3077
3078         * Scripts/webkitpy/common/config/committers.py:
3079
3080 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
3081
3082         build.webkit.org - add more google windows bots
3083         https://bugs.webkit.org/show_bug.cgi?id=105225
3084
3085         Reviewed by Eric Seidel.
3086
3087         Adds two more windows bots (google-windows-3, a builder, and
3088         google-windows-4, a tester) for the Chromium configs.
3089
3090         * BuildSlaveSupport/build.webkit.org-config/config.json:
3091
3092 2012-12-17  Eric Seidel  <eric@webkit.org>
3093
3094         Cleanup --profiler=perf code a little and allow up to 1000 report files (instead of 100)
3095         https://bugs.webkit.org/show_bug.cgi?id=105217
3096
3097         Reviewed by Dirk Pranke.
3098
3099         Previously --profile would fail strangely if find_unused_filename returned None
3100         (which would happen as soon as we had 100 report files).
3101         Now we'll search through 1000 filenames before giving up (and assert if we failed to find one).
3102         I also made some minor cleanups to the output of perf, and tried to make the
3103         commands more self-documenting by using long-names for args.
3104
3105         * Scripts/webkitpy/common/system/profiler.py:
3106         (SingleFileOutputProfiler.__init__):
3107         (Perf.attach_to_pid):
3108         (Perf.profile_after_exit):
3109
3110 2012-12-17  Sheriff Bot  <webkit.review.bot@gmail.com>
3111
3112         Unreviewed, rolling out r137198.
3113         http://trac.webkit.org/changeset/137198
3114         https://bugs.webkit.org/show_bug.cgi?id=105212
3115
3116         This patch is causing API behavior compatibility problems
3117         (Requested by zdobersek on #webkit).
3118
3119         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3120         (willSendRequestCallback):
3121
3122 2012-12-17  Elliott Sprehn  <esprehn@chromium.org>
3123
3124         Unreviewed. Adding myself to committers list.
3125
3126         * Scripts/webkitpy/common/config/committers.py:
3127
3128 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
3129
3130         nrwt: shuffle code around for cleanup in run_webkit_tests.py
3131         https://bugs.webkit.org/show_bug.cgi?id=105078
3132
3133         Reviewed by Ojan Vafai.
3134
3135         This patch reorders functions so that run_webkit_tests can be
3136         understood in a top-down matter better, and to make things slightly
3137         more sensible when I merge manager.py into it.
3138
3139         Also, this adds tests for the actual main() routine, which had several
3140         bugs get through testing a week or two ago.
3141
3142         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3143         (main):
3144         (parse_args):
3145         (_set_up_derived_options):
3146         (run):
3147         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3148         (RunTest):
3149         (RunTest.test_no_http_tests):
3150         (PortTest.disabled_test_mac_lion):
3151         (MainTest):
3152         (MainTest.test_exception_handling):
3153         (MainTest.test_exception_handling.interrupting_run):
3154         (MainTest.test_exception_handling.successful_run):
3155         (MainTest.test_exception_handling.successful_run.FakeRunDetails):
3156         (MainTest.test_exception_handling.exception_raising_run):
3157
3158 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
3159
3160         webkitpy: move --lint-test-files code into its own module
3161         https://bugs.webkit.org/show_bug.cgi?id=105077
3162
3163         Reviewed by Eric Seidel.
3164
3165         It really didn't need to be in run_webkit_tests.py anymore.
3166         This patch adds a new lint-test-expectations scripts to Tools/Scripts;
3167         rwt --lint-test-files is still supported for backwards compatibility
3168         for now.
3169
3170         * Scripts/lint-test-expectations: Added.
3171         * Scripts/webkitpy/layout_tests/lint_test_expectations.py: Added.
3172         (lint):
3173         (main):
3174         * Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py: Added.
3175         (FakePort):
3176         (FakePort.__init__):
3177         (FakePort.test_configuration):
3178         (FakePort.expectations_dict):
3179         (FakePort.skipped_layout_tests):
3180         (FakePort.all_test_configurations):
3181         (FakePort.configuration_specifier_macros):
3182         (FakePort.get_option):
3183         (FakeFactory):
3184         (FakeFactory.__init__):
3185         (FakeFactory.get):
3186         (FakeFactory.all_port_names):
3187         (LintTest):
3188         (LintTest.test_all_configurations):
3189         (LintTest.test_lint_test_files):
3190         (LintTest.test_lint_test_files__errors):
3191         (MainTest):
3192         (MainTest.test_success):
3193         (MainTest.test_success.interrupting_lint):
3194         (MainTest.test_success.successful_lint):
3195         (MainTest.test_success.exception_raising_lint):
3196         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3197         (main):
3198         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3199         (StreamTestingMixin.assertNotEmpty):
3200
3201 2012-12-17  Adam Klein  <adamk@chromium.org>
3202
3203         build-webkit: rename --template-tag to --template-element to match ENABLE #define name
3204         https://bugs.webkit.org/show_bug.cgi?id=105072
3205
3206         Reviewed by Laszlo Gombos.
3207
3208         * Scripts/webkitperl/FeatureList.pm:
3209
3210 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
3211
3212         Add some unit testing for WTF::clampTo* functions
3213         https://bugs.webkit.org/show_bug.cgi?id=105060
3214
3215         Reviewed by Emil A Eklund.
3216
3217         The new tests are targeted at clamping to int / unsigned and float as those are
3218         the one explicitely exposed through more dedicated functions (like clampToInteger,
3219         clampToPositiveInteger and clampToFloat).
3220
3221         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
3222
3223 2012-12-17  Eric Seidel  <eric@webkit.org>
3224
3225         rpt --profile --chromium-android throws exception
3226         https://bugs.webkit.org/show_bug.cgi?id=105061
3227
3228         Reviewed by Daniel Bates.
3229
3230         Somehow popen(args) got turned into popen(*args) at some point,
3231         which caused my new stringify code to fail when args was passed by name.
3232         Fixed and tested.
3233
3234         * Scripts/webkitpy/common/system/executive.py:
3235         (Executive._stringify_args):
3236         (Executive):
3237         (Executive.popen):
3238         * Scripts/webkitpy/common/system/executive_unittest.py:
3239         (ExecutiveTest.test_popen_args):
3240
3241 2012-12-17  Mark Pilgrim  <pilgrim@chromium.org>
3242
3243         Use Platform::current() instead of webKitPlatformSupport() in DumpRenderTree
3244         https://bugs.webkit.org/show_bug.cgi?id=105054
3245
3246         Reviewed by Jochen Eisinger.
3247
3248         Part of a refactoring series; see tracking bug 82948.
3249
3250         * DumpRenderTree/chromium/NotificationPresenter.cpp:
3251         (NotificationPresenter::show):
3252         * DumpRenderTree/chromium/Task.cpp:
3253         (postTask):
3254         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
3255
3256 2012-12-17  Emil A Eklund  <eae@chromium.org>
3257
3258         Clamp values in LayoutUnit::operator/ when SATURATED_LAYOUT_ARITHMETIC is enabled
3259         https://bugs.webkit.org/show_bug.cgi?id=104955
3260
3261         Reviewed by Julien Chaffraix.
3262
3263         Add tests for LayoutUnit.
3264
3265         * TestWebKitAPI/CMakeLists.txt:
3266         * TestWebKitAPI/ForwardingHeaders: Added.
3267         * TestWebKitAPI/ForwardingHeaders/WebCore: Added.
3268         * TestWebKitAPI/ForwardingHeaders/WebCore/LayoutUnit.h: Added.
3269         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
3270         * TestWebKitAPI/TestWebKitAPI.gypi:
3271         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3272         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp: Added.
3273         (TestWebKitAPI):
3274         (TestWebKitAPI::TEST):
3275         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
3276
3277 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
3278
3279         [chromium] removed unused preferences from testRunner.overridePreference
3280         https://bugs.webkit.org/show_bug.cgi?id=105169
3281
3282         Reviewed by Tony Chang.
3283
3284         Most notably this removes the code related to font family maps which is
3285         now handled via window.internals.settings
3286
3287         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
3288         (WebPreferences):
3289         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
3290         (WebTestRunner::TestRunner::overridePreference):
3291         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
3292         (TestRunner):
3293         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
3294         (WebTestRunner::WebPreferences::applyTo):
3295
3296 2012-12-17  James Robinson  <jamesr@chromium.org>
3297
3298         [chromium] Move creation of WebLayerTreeView to WebWidgetClient
3299         https://bugs.webkit.org/show_bug.cgi?id=105071
3300
3301         Reviewed by Adrienne Walker.
3302
3303         Update WebViewHost for the new WebWidgetClient interface.
3304
3305         * DumpRenderTree/chromium/WebViewHost.cpp:
3306         (WebViewHost::initializeLayerTreeView):
3307         (WebViewHost::layerTreeView):
3308         (WebViewHost::~WebViewHost):
3309         * DumpRenderTree/chromium/WebViewHost.h:
3310         (WebViewHost):
3311
3312 2012-12-17  Kenneth Rohde Christiansen  <kenneth@webkit.org>
3313
3314         [EFL][WK2] window_create doesn't receive the url
3315         https://bugs.webkit.org/show_bug.cgi?id=105184
3316
3317         Reviewed by Alexis Menard.
3318
3319         * MiniBrowser/efl/main.c: Use the url supplied to window_create
3320         (on_window_create):
3321
3322 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
3323
3324         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
3325         https://bugs.webkit.org/show_bug.cgi?id=105026
3326
3327         Reviewed by Yury Semikhatsky.
3328
3329         Old SFINAE test was replaced with new one based on this article:
3330         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
3331
3332         New test which covers this problem was added.
3333
3334         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3335
3336 2012-12-17  Anthony Scian  <ascian@rim.com>
3337
3338         [Blackberry] Static code analysis warning fixes
3339         https://bugs.webkit.org/show_bug.cgi?id=105129
3340
3341         Reviewed by Rob Buis.
3342
3343         Added null checks for fopen calls, no recovery attempted; access fault prevention only.
3344
3345         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
3346         (BlackBerry::WebKit::createFile):
3347         (BlackBerry::WebKit::DumpRenderTree::runTest):
3348
3349 2012-12-17  John Mellor  <johnme@chromium.org>
3350
3351         Add Anton Vayvod and Tim Volodine as contributors in committers.py
3352         https://bugs.webkit.org/show_bug.cgi?id=105025
3353
3354         Unreviewed. Adds two contributors to committers.py.
3355         They are initially contributing to the Text Autosizing feature (master bug: 84186).
3356
3357         * Scripts/webkitpy/common/config/committers.py:
3358
3359 2012-12-17  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3360
3361         Unreviewed, add myself to committers.py.
3362
3363         * Scripts/webkitpy/common/config/committers.py:
3364
3365 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3366
3367         [Qt] remove some unnecessary CONFIG additions
3368
3369         Reviewed by Simon Hausmann.
3370
3371         qt is already added by spec_pre.prf, warn_on and depend_includepath by
3372         default_pre.prf.
3373
3374         * qmake/mkspecs/features/default_post.prf:
3375
3376 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
3377
3378         Unreviewed, rolling out r137892.
3379         http://trac.webkit.org/changeset/137892
3380         https://bugs.webkit.org/show_bug.cgi?id=105026
3381
3382         it broke compilation on windows
3383
3384         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3385
3386 2012-12-14  Ilya Tikhonovsky  <loislo@chromium.org>
3387
3388         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
3389         https://bugs.webkit.org/show_bug.cgi?id=105026
3390
3391         Reviewed by Yury Semikhatsky.
3392
3393         Old SFINAE test was replaced with new one based on this article:
3394         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
3395
3396         New test which covers this problem was added.
3397
3398         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3399
3400 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3401
3402         [Qt] rely on automatic output directory setup
3403
3404         Reviewed by Simon Hausmann.
3405
3406         Now being a proper qt module (by virtue of having load(qt_build_config)
3407         in .qmake.conf), webkit gets the path setup goodies for free.
3408
3409         This also fixes Makefile.api.Debug/Release trying to generate qrc_WebKit.cpp
3410         at the same time, instead of the debug and release version ending up in different
3411         directories as they should.
3412
3413         * qmake/mkspecs/features/default_pre.prf:
3414
3415 2012-12-17  Charles Wei  <charles.wei@torchmobile.com.cn>
3416
3417         [BlackBerry] Need to re-enable video track
3418         https://bugs.webkit.org/show_bug.cgi?id=105155
3419
3420         Reviewed by George Staikos.
3421
3422         Just enable Video Track for BlackBerry.
3423
3424         * Scripts/webkitperl/FeatureList.pm:
3425
3426 2012-12-17  János Badics  <jbadics@inf.u-szeged.hu>
3427
3428         Unreviewed. Added myself to committers.py
3429
3430         * Scripts/webkitpy/common/config/committers.py:
3431
3432 2012-12-17  Jonathan Liu  <net147@gmail.com>
3433
3434         Fix linking of libraries with -Wl,-whole-archive
3435         https://bugs.webkit.org/show_bug.cgi?id=104436
3436
3437         Reviewed by Simon Hausmann.
3438
3439         If there is more than one library in LIBS linked with
3440         -Wl,-whole-archive -l... -Wl,-no-whole-archive, qmake tries to
3441         merge the duplicate flags in LIBS which causes linking to fail.
3442
3443         Add no_smart_library_merge to CONFIG to prevent the duplicate
3444         flags from being removed when using -Wl,-whole-archive.
3445
3446         * qmake/mkspecs/features/functions.prf:
3447
3448 2012-12-16  Arvid Nilsson  <anilsson@rim.com>
3449
3450         [BlackBerry] Adapt to new BlackBerry::Platform::TouchPoint API
3451         https://bugs.webkit.org/show_bug.cgi?id=105143
3452         RIM PR 171941
3453
3454         Reviewed by Rob Buis.
3455         Internally reviewed by George Staikos.
3456
3457         Adapt to new method names and encapsulation of TouchPoint data members.
3458
3459         * DumpRenderTree/blackberry/EventSender.cpp:
3460         (addTouchPointCallback):
3461         (updateTouchPointCallback):
3462         (touchEndCallback):
3463         (releaseTouchPointCallback):
3464         (sendTouchEvent):
3465
3466 2012-12-16  Maciej Stachowiak  <mjs@apple.com>
3467
3468         Add a script to compute the mean and 95% confidence interval (using two-sided t-test) of a sample set
3469         https://bugs.webkit.org/show_bug.cgi?id=105148
3470
3471         Reviewed by Filip Pizlo.
3472
3473         Usage examples:
3474         
3475         $ sampstat --help
3476         Usage: sampstat [options] [FILES]
3477           Compute the mean and 95% confidence interval of a sample set.
3478           Standard input or files must contain two or more decimal numbers, one per line.
3479         
3480         Options:
3481           -h, --help            show this help message and exit
3482           -u UNIT, --unit=UNIT  assume values are in units of UNIT
3483           -v, --verbose         print all values (with units)
3484         
3485         $ sampstat -u MB memresults.txt
3486         2356.90 MB +/- 101.34 MB (4.3%)
3487         
3488         $ sampstat -v -u MB memresults.txt
3489               2318.21 MB
3490               2399.56 MB
3491               2352.93 MB
3492         ----------------
3493         Mean: 2356.90 MB +/- 101.34 MB (4.3%)
3494         
3495
3496         * Scripts/sampstat: Added.
3497         (sum): Helper function to compute the sum of a list.
3498         (arithmeticMean): Compute the meam of a list.
3499         (standardDeviation): Compute the sample standard deviation (unbiased estimator).
3500         (standardError): Compute the sample standard error.
3501         (tDist): Compute t(.025, n-1), the t-value for atwo-sided 95% confidence interval. 
3502         (twoSidedConfidenceInterval): Compute the two-sided confidence interval range about the mean.
3503
3504 2012-12-15  Nima Ghanavatian  <nghanavatian@rim.com>
3505
3506         [BlackBerry] Pass key modifiers with touch and mouse events
3507         https://bugs.webkit.org/show_bug.cgi?id=105108
3508
3509         Reviewed by Rob Buis.
3510
3511         Update DRT to new api.
3512
3513         Internally reviewed by Gen Mak.
3514
3515         * DumpRenderTree/blackberry/EventSender.cpp:
3516         (mouseDownCallback):
3517         (mouseUpCallback):
3518         (mouseMoveToCallback):
3519
3520 2012-12-15  Zan Dobersek  <zandobersek@gmail.com>
3521
3522         Stop the XvfbDriver before proceeding with its starting setup
3523         https://bugs.webkit.org/show_bug.cgi?id=105006
3524
3525         Reviewed by Philippe Normand.
3526
3527         Just as with the base Driver implementation, stop the XvfbDriver
3528         before proceeding with setting up lock files and a new Xvfb instance.
3529         This forces any driver or Xvfb instances that are currently under
3530         the driver's control to be closed/killed and cleaned up after.
3531
3532         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
3533         (XvfbDriver._start):
3534
3535 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
3536
3537         webkitpy: make port_name a required parameter to the non-test ports
3538         https://bugs.webkit.org/show_bug.cgi?id=105057
3539
3540         Reviewed by Eric Seidel.
3541
3542         This is just minor code cleanup along the path to making Port
3543         constructors simpler and pushing more logic into the port factory.
3544
3545         * Scripts/webkitpy/layout_tests/port/base.py:
3546         (Port.__init__):
3547         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3548         (PortTest.make_port):
3549         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3550         (TestWebKitPort.__init__):
3551         * Scripts/webkitpy/layout_tests/port/test.py:
3552         (TestPort):
3553         (TestPort.determine_full_port_name):
3554         (TestPort.__init__):
3555         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py:
3556         (XvfbDriverTest.make_driver):
3557         (XvfbDriverTest.test_stop):
3558
3559 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
3560
3561         garden-o-matic doesn't know about reftests
3562         https://bugs.webkit.org/show_bug.cgi?id=101976
3563
3564         Reviewed by Eric Seidel.
3565
3566         Attempt to make garden-o-matic properly aware of reftests again.
3567
3568         This is re-landing r137407 with a change to
3569         rebaselineWithStatusUpdates() to actually pass the filtered
3570         failureInofLost to checkout.rebaseline().
3571
3572         Unfortunately, the testing coverage for this whole module is almost
3573         non-existent, and it's not obvious that there's a good way to
3574         stub out checkout.rebaseline() to test that the right thing
3575         happens. I verified the change manually.
3576
3577         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
3578         (.):
3579         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
3580         (.):
3581         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
3582
3583 2012-12-14  Ryosuke Niwa  <rniwa@webkit.org>
3584
3585         "Running 1 DumpRenderTree over X shards" is not a helpful output
3586         https://bugs.webkit.org/show_bug.cgi?id=104858
3587
3588         Reviewed by Dirk Pranke.
3589
3590         The word "shard" is not typically used in test frameworks. This word is confusing for many WebKit
3591         contributors who typically don't have much experience on distributed computations and databases.
3592         So only print this message out in the debug mode.
3593
3594         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3595         * Scripts/webkitpy/layout_tests/views/printing.py:
3596         (Printer.print_workers_and_shards):
3597
3598 2012-12-13  Eric Seidel  <eric@webkit.org>
3599
3600         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
3601         https://bugs.webkit.org/show_bug.cgi?id=104891
3602
3603         Reviewed by Ryosuke Niwa.
3604
3605         I also implemented a very simple "Sample" Profiler using
3606         Mac OS X's /usr/bin/sample command line tool.
3607
3608         The real reason for this abstraction is to make it easy
3609         to support both perf and pprof on linux which seem to
3610         be about equally popular among those I ask in the Chrome team.
3611
3612         * Scripts/webkitpy/common/system/profiler.py:
3613         (ProfilerFactory.create_profiler):
3614         (ProfilerFactory):
3615         (ProfilerFactory.available_profilers_by_name):
3616         (ProfilerFactory.default_profiler_name):
3617         (Sample):
3618         (Sample.__init__):
3619         (Sample.attach_to_pid):
3620         (Sample.profile_after_exit):
3621         (IProfiler.attach_to_pid):
3622         * Scripts/webkitpy/layout_tests/port/driver.py:
3623         (Driver.__init__):
3624         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3625         (PerfTestsRunner._parse_args):
3626
3627 2012-12-13  Eric Seidel  <eric@webkit.org>
3628
3629         Unreviewed, fixing typo in python unittest.
3630
3631         I changed the default profiler on linux from pprof to perf in
3632         bug 104971.  I failed to update the unittest results at that time.
3633
3634         * Scripts/webkitpy/common/system/profiler_unittest.py:
3635         (ProfilerFactoryTest.test_basic):
3636
3637 2012-12-13  Ian Vollick  <vollick@chromium.org>
3638
3639         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
3640         https://bugs.webkit.org/show_bug.cgi?id=104911
3641
3642         Reviewed by Dirk Pranke.
3643
3644         Adds two virtual tests suites to run the tests in
3645         compositing/overflow/ and scrollbars/ with the flag
3646         --enable-accelerated-overflow-scroll, which I've also plumbed through
3647         in the usual way.
3648
3649         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3650         (main):
3651         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
3652         (WebPreferences):
3653         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
3654         (WebTestRunner::WebPreferences::reset):
3655         (WebTestRunner::WebPreferences::applyTo):
3656         * DumpRenderTree/chromium/TestShell.cpp:
3657         (TestShell::TestShell):
3658         (TestShell::resetWebSettings):
3659         * DumpRenderTree/chromium/TestShell.h:
3660         (TestShell::setAcceleratedCompositingForOverflowScrollEnabled):
3661         (TestShell):
3662         * Scripts/webkitpy/layout_tests/port/chromium.py:
3663         (ChromiumPort.virtual_test_suites):
3664
3665 2012-12-13  Eric Seidel  <eric@webkit.org>
3666
3667         Use 'perf' to profile on linux instead of google-pprof by default
3668         https://bugs.webkit.org/show_bug.cgi?id=104971
3669
3670         Reviewed by Dirk Pranke.
3671
3672         This makes Chromium Linux match Chromium Android and use the perf
3673         tool by default.  Once https://bugs.webkit.org/show_bug.cgi?id=104891
3674         lands it will be possible to use pprof again on Linux.
3675
3676         This is slightly more advanced than the AndroidPerf profiler as
3677         instead of using a timeout on "perf record" I instead watch
3678         for the termination of the target process and then control-C
3679         the 'perf record' process.  This required me to add two new
3680         methods to Executive, one to have a limited-time wait() and
3681         the second to be able to send a control-C.  I chose to add
3682         these to Executive to make them easier to mock/fix-for-win32
3683         at a later time if needed.
3684
3685         * Scripts/webkitpy/common/system/executive.py:
3686         (Executive.wait_limited):
3687         (Executive.interrupt):
3688         * Scripts/webkitpy/common/system/profiler.py:
3689         (ProfilerFactory.create_profiler):
3690         (Perf):
3691         (Perf.__init__):
3692         (Perf._perf_path):
3693         (Perf.attach_to_pid):
3694         (Perf._first_ten_lines_of_profile):
3695         (Perf.profile_after_exit):
3696
3697 2012-12-13  Julie Parent  <jparent@chromium.org>
3698
3699         Dashboard cleanup: remove usage of global g_builders.
3700         https://bugs.webkit.org/show_bug.cgi?id=104941
3701
3702         Reviewed by Dirk Pranke.
3703
3704         The dashboards use a lot of global state, which makes hacking on them
3705         complicated. This change removes the use of one such global: g_builders.
3706         In most cases, we can just use currentBuilderGroup().builders instead,
3707         which is now currentBuilders().
3708         Surprisingly, the most changes were required to the unit tests, since
3709         they were even bigger offenders of bad hygiene, relying on global state
3710         set by other tests, randomly clobbering global variables in ways the
3711         real code doesn't, etc.
3712
3713         * TestResultServer/static-dashboards/builders.js:
3714         (BuilderGroup.prototype.setup):
3715         * TestResultServer/static-dashboards/dashboard_base.js:
3716         (.switch.return):
3717         (htmlForTestTypeSwitcher):
3718         * TestResultServer/static-dashboards/flakiness_dashboard.js:
3719         (generatePage):
3720         (getAllTestsTrie):
3721         (processTestRunsForAllBuilders):
3722         (showPopupForBuild):
3723         (generatePageForExpectationsUpdate):
3724         (loadExpectationsLayoutTests):
3725         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
3726         (resetGlobals):
3727         (stubResultsByBuilder):
3728         (test):
3729         * TestResultServer/static-dashboards/loader.js:
3730         (.):
3731         * TestResultServer/static-dashboards/loader_unittests.js:
3732
3733 2012-12-13  Eric Seidel  <eric@webkit.org>
3734
3735         Callers should not have to stringify args before calling Executive run_command/popen
3736         https://bugs.webkit.org/show_bug.cgi?id=104975
3737
3738         Reviewed by Dirk Pranke.
3739
3740         One could argue that we should match the python call syntax here,
3741         but I think it's a more friendly API if we automagically handle
3742         stringification of args in run_command, etc.
3743         This removes map(unicode, args) from several callsites.
3744
3745         When I first tried to land this change, I didn't realize that
3746         Executive._command_for_printing depended on this behavior
3747         having been applied to args in run_command.  The fix is to
3748         call _stringify_args in both run_command and popen.
3749         This is slightly redundant, but given how short args have to be
3750         (due to shell limits), I don't think the double-encode check
3751         matters in practice.
3752
3753         This is slightly complicated by the fact that apache_http_server.py
3754         is the one caller in our codebase which uses shell=True.
3755         shell=True is a well-documented trail-of-tears:
3756         http://stackoverflow.com/questions/3172470/actual-meaning-of-shell-true-in-subprocess
3757         but to support this legacy (windows-only) code (which I can't easily test)
3758         I've added an if-hack to avoid stringifying the the popen(shell=True) case.
3759
3760         * Scripts/webkitpy/common/system/executive.py:
3761         (Executive.run_command):
3762         (Executive._stringify_args):
3763         (Executive.popen):
3764         * Scripts/webkitpy/common/system/executive_unittest.py:
3765         (ExecutiveTest.test_auto_stringify_args):
3766         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3767         (attach_to_pid):
3768
3769 2012-12-13  Adrienne Walker  <enne@chromium.org>
3770
3771         Unreviewed, rolling out r137645, r137646, and r137667.
3772         http://trac.webkit.org/changeset/137645
3773         http://trac.webkit.org/changeset/137646
3774         http://trac.webkit.org/changeset/137667
3775         https://bugs.webkit.org/show_bug.cgi?id=104911
3776
3777         Breaks some overflow layout tests
3778
3779         * Scripts/webkitpy/layout_tests/port/chromium.py:
3780         (ChromiumPort.virtual_test_suites):
3781
3782 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3783
3784         Unreviewed, rolling out r137661.
3785         http://trac.webkit.org/changeset/137661
3786         https://bugs.webkit.org/show_bug.cgi?id=104891
3787
3788         broke unit tests, run-webkit-tests
3789
3790         * Scripts/webkitpy/common/system/executive.py:
3791         (Executive._run_command_with_teed_output):
3792         (Executive.run_command):
3793         (Executive.popen):
3794         * Scripts/webkitpy/common/system/profiler.py:
3795         (ProfilerFactory.create_profiler):
3796         (GooglePProf.profile_after_exit):
3797         (IProfiler.attach_to_pid):
3798         * Scripts/webkitpy/layout_tests/port/driver.py:
3799         (Driver.__init__):
3800         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3801         (PerfTestsRunner._parse_args):
3802
3803 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3804
3805         nrwt: move rundetails, summarize_results into test_run_results
3806         https://bugs.webkit.org/show_bug.cgi?id=104963
3807
3808         Reviewed by Eric Seidel.
3809
3810         More cleanup; this the plain-old-data functions out of manager.py
3811         and into test_run_results next to the other aggregate data structures.
3812
3813         Also this cleans up a bunch of test code and reduces duplication.
3814
3815         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3816         (Manager.run):
3817         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3818         (ManagerTest.test_look_for_new_crash_logs):
3819         * Scripts/webkitpy/layout_tests/models/test_run_results.py:
3820         (TestRunResults.__init__):
3821         (TestRunResults.add):
3822         (RunDetails):
3823         (RunDetails.__init__):
3824         (_interpret_test_failures):
3825         (summarize_results):
3826         * Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py: Added.
3827         (get_result):
3828         (run_results):
3829         (summarized_results):
3830         (InterpretTestFailuresTest):
3831         (InterpretTestFailuresTest.setUp):
3832         (InterpretTestFailuresTest.test_interpret_test_failures):
3833         (SummarizedResultsTest):
3834         (SummarizedResultsTest.setUp):
3835         (SummarizedResultsTest.test_no_svn_revision):
3836         (SummarizedResultsTest.test_svn_revision):
3837         (SummarizedResultsTest.test_summarized_results_wontfix):
3838         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3839         (BuildBotPrinterTests.get_printer):
3840         (BuildBotPrinterTests.test_print_unexpected_results):
3841         (BuildBotPrinterTests.test_print_results):
3842
3843 2012-12-13  Eric Seidel  <eric@webkit.org>
3844
3845         Make --profile output easier to understand for pprof on linux
3846         https://bugs.webkit.org/show_bug.cgi?id=104957
3847
3848         Reviewed by Dirk Pranke.
3849
3850         pprof is kinda mysterious, so we print some information
3851         on how to use it, and how to learn how to read the sample dump.
3852         This is similar to what we've done for perf output for Chromium Android.
3853
3854         * Scripts/webkitpy/common/system/profiler.py:
3855         (GooglePProf._pprof_path):
3856         (GooglePProf.profile_after_exit):
3857
3858 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3859
3860         Revert r137407 - it didn't actually work right.
3861         https://bugs.webkit.org/show_bug.cgi?id=101976
3862
3863         Unreviewed, build fix.
3864
3865         Turns out that rebaselining was completely broken with this change :(
3866         We're passing in a list of test names rather than the list of
3867         failureInfo objects.
3868
3869         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
3870         (.):
3871         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
3872         (.):
3873         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
3874
3875 2012-12-13  Antoine Labour  <piman@chromium.org>
3876
3877         [chromium] Small fixes for WebPluginContainer::setWebLayer
3878         https://bugs.webkit.org/show_bug.cgi?id=104953
3879
3880         Reviewed by James Robinson.
3881
3882         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Clean up
3883         dangling pointers, force correct invalidation of the layer.
3884
3885 2012-12-13  Eric Seidel  <eric@webkit.org>
3886
3887         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
3888         https://bugs.webkit.org/show_bug.cgi?id=104891
3889
3890         Reviewed by Dirk Pranke.
3891
3892         I also implemented a very simple "Sample" Profiler using
3893         Mac OS X's /usr/bin/sample command line tool.
3894
3895         I also moved the map(unicode, args) call into Executive.popen
3896         so that all callers don't have to do it themselves.
3897
3898         The real reason for this abstraction is to make it easy
3899         to support both perf and pprof on linux which seem to
3900         be about equally popular among those I ask in the Chrome team.
3901
3902         * Scripts/webkitpy/common/system/executive.py:
3903         (Executive.popen):
3904         * Scripts/webkitpy/common/system/profiler.py:
3905         (ProfilerFactory.create_profiler):
3906         (ProfilerFactory):
3907         (ProfilerFactory.available_profilers_by_name):
3908         (ProfilerFactory.default_profiler_name):
3909         (Sample):
3910         (Sample.__init__):
3911         (Sample.attach_to_pid):
3912         (Sample.profile_after_exit):
3913         (IProfiler.attach_to_pid):
3914         * Scripts/webkitpy/layout_tests/port/driver.py:
3915         (Driver.__init__):
3916         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3917         (PerfTestsRunner._parse_args):
3918
3919 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3920
3921         support -wk2 port names properly in webkitpy.layout_tests.port
3922         https://bugs.webkit.org/show_bug.cgi?id=104761
3923
3924         Reviewed by Eric Seidel.
3925
3926         Our support for WK2 port names was inconsistent at best; this patch
3927         modifies the code so that <port_name>-wk2 is supported for all ports that
3928         have WK2 implementations and is equivalent to specifying <port_name>
3929         and -2/--webkit-test-runner. In addition, this modifies
3930         builders.all_port_names() to include the wk2 variants.
3931
3932         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
3933         * Scripts/webkitpy/layout_tests/port/apple.py:
3934         (ApplePort.determine_full_port_name):
3935         * Scripts/webkitpy/layout_tests/port/base.py:
3936         (Port.determine_full_port_name):
3937         (Port.__init__):
3938         * Scripts/webkitpy/layout_tests/port/builders.py:
3939         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
3940         (DriverTest.make_port):
3941         * Scripts/webkitpy/layout_tests/port/mac.py:
3942         (MacPort.default_baseline_search_path):
3943         * Scripts/webkitpy/layout_tests/port/win.py:
3944         (WinPort.default_baseline_search_path):
3945         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
3946         (TestRebaselineExpectations.test_rebaseline_expectations):
3947
3948 2012-12-13  Ian Vollick  <vollick@chromium.org>
3949
3950         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
3951         https://bugs.webkit.org/show_bug.cgi?id=104911
3952
3953         Reviewed by Adrienne Walker.
3954
3955         Adds two virtual tests suites to run the tests in
3956         compositing/overflow/ and scrollbars/ with the flag
3957         --enable-accelerated-compositing-for-overflow-scroll
3958
3959         * Scripts/webkitpy/layout_tests/port/chromium.py:
3960         (ChromiumPort.virtual_test_suites):
3961
3962 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3963
3964         delete extraneous self.fail in mastercfg_unittest.py
3965         https://bugs.webkit.org/show_bug.cgi?id=104940
3966
3967         Reviewed by Eric Seidel.
3968
3969         This was accidentally introduced in r137404.
3970
3971         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
3972
3973 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3974
3975         Fix a mis-named variable after r137621.
3976         https://bugs.webkit.org/show_bug.cgi?id=104871
3977
3978         Unreviewed, build fix.
3979
3980         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
3981         (BuildBotPrinter.print_results):
3982         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3983
3984 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3985
3986         webkitpy: rename ResultSummary
3987         https://bugs.webkit.org/show_bug.cgi?id=104871
3988
3989         Reviewed by Eric Seidel.
3990
3991         This patch changes the names of ResultSummary and related modules,
3992         classes, and variables as a part of the larger clean up in bug 103824.
3993
3994         This change is basically a bunch of search and replaces:
3995         - result_summary.py -> test_run_results.py
3996         - class ResultSummary -> class TestRunResults
3997         - "result_summary" or "rs" as a generic variable name -> "run_results",
3998           when it doesn't matter (or we don't know) if this refers to the
3999           initial pass over the tests or the retry pass.
4000         - "result_summary" as a variable containing the results of the initial,
4001           complete test run -> "initial_results"
4002         - "retry_summary" as a variable containing the results of retried
4003           results -> "retry_results"
4004         - ResultSummary.results (the property) -> TestRunResults.results_by_name
4005         - ResultSummary.failures (the property) -> TestRunResults.failures_by_name
4006         - RunDetails.result_summary -> RunDetails.initial_results
4007         - RunDetails.retry_summary -> RunDetails.retry_results
4008
4009         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
4010         (LayoutTestRunner.__init__):
4011         (LayoutTestRunner.run_tests):
4012         (LayoutTestRunner._mark_interrupted_tests_as_skipped):
4013         (LayoutTestRunner._interrupt_if_at_failure_limits):
4014         (LayoutTestRunner._interrupt_if_at_failure_limits.interrupt_if_at_failure_limit):
4015         (LayoutTestRunner._update_summary_with_result):
4016         (LayoutTestRunner._handle_finished_test):
4017         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
4018         (FakePrinter.print_expected):
4019         (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
4020         (LayoutTestRunnerTests.test_update_summary_with_result):
4021         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4022         (RunDetails.__init__):
4023         (summarize_results):
4024         (Manager.run):
4025         (Manager._look_for_new_crash_logs):
4026         (Manager._tests_to_retry):
4027         (Manager._upload_json_files):
4028         (Manager._stats_trie):
4029         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
4030         (ManagerTest.test_look_for_new_crash_logs):
4031         (SummarizeResultsTest):
4032         (SummarizeResultsTest.get_run_results):
4033         (SummarizeResultsTest.summarized_results):
4034         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
4035         (JSONLayoutResultsGenerator.__init__):
4036         (JSONLayoutResultsGenerator._get_modifier_char):
4037         (JSONLayoutResultsGenerator._insert_failure_summaries):
4038         (JSONLayoutResultsGenerator._get_failure_summary_entry):
4039         * Scripts/webkitpy/layout_tests/models/test_run_results.py: Renamed from Tools/Scripts/webkitpy/layout_tests/models/result_summary.py.
4040         (TestRunResults):
4041         (TestRunResults.__init__):
4042         (TestRunResults.add):
4043         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
4044         (get_test_results):
4045         (MainTest.test_basic):
4046         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
4047         (BuildBotPrinter.print_results):
4048         (BuildBotPrinter.print_run_results):
4049         (BuildBotPrinter._print_run_results_entry):
4050         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
4051         (BuildBotPrinterTests.get_run_results):
4052         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
4053         (BuildBotPrinterTests):
4054         * Scripts/webkitpy/layout_tests/views/printing.py:
4055         (Printer.print_expected):
4056         (Printer._print_expected_results_of_type):
4057         (Printer.print_results):
4058         (Printer._print_timing_statistics):
4059         (Printer._print_worker_statistics):
4060         (Printer._print_aggregate_test_statistics):
4061         (Printer._print_individual_test_times):
4062         (Printer._print_directory_timings):
4063         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
4064
4065 2012-12-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
4066
4067         [EFL] The WebKit2 bots are building and executing WebKit1 tests
4068         https://bugs.webkit.org/show_bug.cgi?id=104690
4069
4070         Reviewed by Csaba Osztrogonác.
4071
4072         Pass --no-webkit1 option to the EFL WK2 bots when building. This
4073         will disable WK1 API and thus, the WK1 API tests.
4074
4075         * BuildSlaveSupport/build.webkit.org-config/config.json:
4076         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
4077         (CompileWebKit2Only):
4078         (BuildAndTestFactory):
4079         (BuildAndTestFactory.__init__):
4080         (BuildAndTestWebKit2OnlyFactory):
4081         Now bots using BuildAndTestWebKit2OnlyFactory (for instance only EFL) will
4082         get --no-webkit1 as a parameter on the build-webkit script.
4083         * Scripts/build-webkit:
4084         Added support for -no-webkit1 for the Qt and EFL port.
4085
4086 2012-12-13  Eric Seidel  <eric@webkit.org>
4087
4088         build-webkit --chromium-android forces Make even when update-webkit used --ninja
4089         https://bugs.webkit.org/show_bug.cgi?id=104839
4090
4091         Reviewed by Tony Chang.
4092
4093         This was recommended by Peter as part of:
4094         https://bugs.webkit.org/show_bug.cgi?id=104434#c4
4095         Fixing this makes build-webkit --chromium-android work for the common (ninja) case.
4096
4097         * Scripts/webkitdirs.pm:
4098         (buildChromium):
4099
4100 2012-12-13  Julie Parent  <jparent@chromium.org>
4101
4102         Loader_unittests should reset state between runs.
4103         https://bugs.webkit.org/show_bug.cgi?id=104862
4104
4105         Reviewed by Dirk Pranke.
4106
4107         The loader_unittests were not resetting state between runs, leaving
4108         a lot of data behind in global objects that the dashboard and tests
4109         rely on.  A handful were resetting just a few objects.  Use the common
4110         resetGlobals() method instead.
4111
4112         * TestResultServer/static-dashboards/loader_unittests.js:
4113
4114 2012-12-13  Andras Becsi  <andras.becsi@digia.com>
4115
4116         [Qt][WK2] Fix painting on Mac with retina display
4117         https://bugs.webkit.org/show_bug.cgi?id=104574
4118
4119         Reviewed by Kenneth Rohde Christiansen.
4120
4121         Remove setting the devicePixelRatio experimental property
4122         since the value is now automatically picked up from Qt.
4123
4124         * MiniBrowser/qt/qml/BrowserWindow.qml:
4125
4126 2012-12-13  Jussi Kukkonen  <jussi.kukkonen@intel.com>
4127
4128         [EFL][GTK] Don't call deprecated g_type_init when glib => 2.35
4129         https://bugs.webkit.org/show_bug.cgi?id=103209
4130
4131         Reviewed by Kenneth Rohde Christiansen.
4132
4133         g_type_init() is deprecated from 2.35.0 onwards. Don't call it
4134         in that case to avoid warnings and/or build failure.
4135
4136         * DumpRenderTree/gtk/ImageDiff.cpp:
4137         (main):
4138
4139 2012-12-13  Claudio Saavedra  <csaavedra@igalia.com>
4140
4141         [Tools] Add exception for gtk_style_context_get()
4142         https://bugs.webkit.org/show_bug.cgi?id=104896
4143
4144         Reviewed by Martin Robinson.
4145
4146         Add new exception for gtk_style_context_get().
4147
4148         * Scripts/webkitpy/style/checkers/cpp.py:
4149         (check_for_null): Add exception for gtk_style_context_get().
4150         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
4151         (WebKitStyleTest.test_null_false_zero): Add a unit
4152         test for the new exception.
4153
4154 2012-12-13  Joaquim Rocha  <jrocha@igalia.com>
4155
4156         REGRESSION (r137432): The /webkit2/WebKitCookieManager/accept-policy unit test is failing
4157         https://bugs.webkit.org/show_bug.cgi?id=104790
4158
4159         Reviewed by Carlos Garcia Campos.
4160
4161         Unskip the GTK accept policy test.
4162
4163         * Scripts/run-gtk-tests:
4164         (TestRunner):
4165
4166 2012-12-13  Xingnan Wang  <xingnan.wang@intel.com>
4167
4168         Unreviewed. Adding myself to committers list.
4169
4170         * Scripts/webkitpy/common/config/committers.py:
4171
4172 2012-12-12  Roger Fong  <roger_fong@apple.com>
4173
4174         Unreviewed. Fix typo in NRWT.
4175
4176         * Scripts/webkitpy/layout_tests/port/base.py:
4177         (Port.is_reference_html_file):
4178
4179 2012-12-12  Roger Fong  <roger_fong@apple.com>
4180
4181         Fix detection of ref test files in old and new run webkit tests.
4182         https://bugs.webkit.org/show_bug.cgi?id=104850.
4183
4184         Reviewed by Timothy Horton.
4185
4186         * Scripts/old-run-webkit-tests:
4187         (isUsedInReftest): 
4188         Add notref and ref prefix and suffix cases to isUsedinReftest.
4189         * Scripts/webkitpy/layout_tests/port/base.py:
4190         (Port):
4191         (Port.is_reference_html_file): 
4192         Typo in NRWT.
4193
4194 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
4195
4196         rework ChromiumAndroidDriver.start() et. al to avoid funky layering
4197         https://bugs.webkit.org/show_bug.cgi?id=104856
4198
4199         Reviewed by Eric Seidel.
4200
4201         r137523 / bug 10357 introduced some back-and-forth hopping during
4202         driver.start() in the chromium android port. This patch cleans it
4203         up a bit.
4204
4205         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
4206         (ChromiumAndroidDriver.__init__):
4207         (ChromiumAndroidDriver._start_once):
4208         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
4209         * Scripts/webkitpy/layout_tests/port/driver.py:
4210         (Driver.start):
4211         (Driver._start):
4212
4213 2012-12-12  Eric Seidel  <eric@webkit.org>
4214
4215         Add support for run-perf-tests --chromium-android --profile
4216         https://bugs.webkit.org/show_bug.cgi?id=103587
4217
4218         Reviewed by Dirk Pranke.
4219
4220         This makes it possible to profile PerformanceTests on an Android device
4221         using run-perf-tests --chromium-android --profile.
4222         CPU profiling on Android is sadly non-trivial, but this goes a long
4223         ways towards making it more accessible to non-Android developers.
4224
4225         Although this script requires your device to be flashed with a
4226         userdebug build (an annoyingly non-trivial endeavor for external contributors)
4227         it endeavors to detect and explain the common errors encountered
4228         while trying to setup a device for perf-based profiling.
4229
4230         * Scripts/webkitpy/common/system/profiler.py:
4231         (GooglePProf.profile_after_exit): Discovered that pprof will just hang if passed a missing file.
4232         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
4233         (AndroidPerf):
4234         (AndroidPerf.__init__):
4235         (AndroidPerf.check_configuration):
4236         (AndroidPerf.print_setup_instructions):
4237         (_file_exists_on_device):
4238         (_run_adb_command):
4239         (attach_to_pid):
4240         (_perf_version_string):
4241         (_find_perfhost_binary):
4242         (_perfhost_path):
4243         (_first_ten_lines_of_profile):
4244         (profile_after_exit):
4245         (ChromiumAndroidDriver.__init__):
4246         (ChromiumAndroidDriver.__init__.methods):
4247         (ChromiumAndroidDriver._find_or_create_symfs):
4248         (ChromiumAndroidDriver._start_once):
4249         (ChromiumAndroidDriver):
4250         (ChromiumAndroidDriver._run_post_start_tasks):
4251         (ChromiumAndroidDriver._pid_from_android_ps_output):
4252         (ChromiumAndroidDriver._pid_on_target):
4253         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
4254         (ChromiumAndroidDriverTest.test_pid_from_android_ps_output):
4255         (AndroidPerfTest):
4256         (AndroidPerfTest.test_perf_output_regexp):
4257         * Scripts/webkitpy/layout_tests/port/driver.py:
4258         (Driver._setup_environ_for_driver):
4259         (Driver):
4260         (Driver._start):
4261         (Driver._run_post_start_tasks):
4262         (Driver._pid_on_target):
4263
4264 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
4265
4266         nrwt: remove deprecated ORWT compatibility command line flags
4267         https://bugs.webkit.org/show_bug.cgi?id=104301
4268
4269         Reviewed by Eric Seidel.
4270
4271         The command line flags for --noshow-results, --[no-]launch-safari,
4272         and --use-remote-links-to-tests were for compatibility with
4273         old-run-webkit-tests, but the compatibility is not needed any more
4274         since NRWT is basically default and the scripts that passed the
4275         old flags have been updated.
4276
4277         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4278         (_set_up_derived_options):
4279         (parse_args):
4280
4281 2012-12-12  John Griggs  <jgriggs@rim.com>
4282
4283         2012-12-12  John Griggs  <jgriggs@rim.com>
4284
4285      &