[EFL] Move ImageDiff to the Tools directory root
[WebKit-https.git] / Tools / ChangeLog
1 2012-12-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2
3         [EFL] Move ImageDiff to the Tools directory root
4         https://bugs.webkit.org/show_bug.cgi?id=105305
5
6         Reviewed by Laszlo Gombos.
7
8         We previously moved this directory from the DRT directory to
9         efl/ImageDiff. The major reason for doing so was because ImageDiff does
10         not depend on DumpRenderTree but for legacy reasons was living inside the
11         DRT folder, even though it is needed when running pixel tests for WK2.
12
13         Now we are moving to Tool directory root as other ports agreed to do the same.
14
15         * CMakeLists.txt:
16         * ImageDiff/CMakeLists.txt: Added.
17         * ImageDiff/PlatformEfl.cmake: Added.
18         * ImageDiff/efl/ImageDiff.cpp: Renamed from Tools/efl/ImageDiff/ImageDiff.cpp.
19         * efl/ImageDiff/CMakeLists.txt: Removed.
20
21 2012-12-19  Zeno Albisser  <zeno@webkit.org>
22
23         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
24         https://bugs.webkit.org/show_bug.cgi?id=105309
25
26         The version number is only used as a part
27         of the library name on Linux.
28
29         Reviewed by Kenneth Rohde Christiansen.
30
31         * Scripts/webkitdirs.pm:
32         (builtDylibPathForName):
33
34 2012-12-19  Alberto Garcia  <agarcia@igalia.com>
35
36         [GTK] Tests that time out are not considered as failures.
37         https://bugs.webkit.org/show_bug.cgi?id=105318
38
39         Reviewed by Martin Robinson.
40
41         Fix parsing of the GTK API tests log files.
42
43         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
44         (RunGtkAPITests.commandComplete):
45
46 2012-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
47
48         Unreviewed, rolling out r138061.
49         https://bugs.webkit.org/show_bug.cgi?id=105396
50
51         Original patch:
52         http://trac.webkit.org/changeset/138061
53         https://bugs.webkit.org/show_bug.cgi?id=97359
54
55         It is crashing on Debug bots
56
57         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
58         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
59         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
60         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
61         (AccessibilityUIElement):
62
63 2012-12-18  Alice Liu  <alice.liu@apple.com>
64
65         Test for https://bugs.webkit.org/show_bug.cgi?id=105106
66         Add SPI to WebKit1 WebFrame for hit testing
67
68         Reviewed by Dan Bernstein.
69
70         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/ElementAtPointInWebFrame.mm
71         * TestWebKitAPI/Tests/mac/ElementAtPointInWebFrame.mm: Added.
72         (TestWebKitAPI::TEST): Loads html with two divs positioned in the 2nd and 4th quadrants of the webview.
73         Then hit-tests at three points, expecting to hit the two divs and body element. 
74
75 2012-12-18  Alice Liu  <alice.liu@apple.com>
76
77         Test for https://bugs.webkit.org/show_bug.cgi?id=105262
78         Add SPI to WebKit1 WebFrame for node conversion to JSValueRef
79
80         Reviewed by Anders Carlsson.
81
82         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/JSWrapperForNodeInWebFrame.mm.mm
83         * TestWebKitAPI/Tests/mac/JSWrapperForNodeInWebFrame.mm.mm: Added.
84         (TestWebKitAPI::TEST): Tests for the correct JS wrapper for a DOMNode, provided a WebScriptWorld.
85         Loads html with a single div element. In an isolated world, creates a property on that node.
86         Also, in the standard world, creates a different property on that node.  Then tests for 4 things:
87             - Existence of the isolated property in the isolated world.
88             - Existence of the standard property in the standard world.
89             - Non-existence of the isolated property in the standard world.
90             - Non-existence of the standard property in the isolated world.
91
92 2012-12-18  Simon Fraser  <simon.fraser@apple.com>
93
94         Disable the SimplifyMarkup API test, wich asserts in debug builds.
95         Tracked by https://bugs.webkit.org/show_bug.cgi?id=105370.
96         
97         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
98         (TestWebKitAPI::TEST):
99
100 2012-12-18  Joseph Pecoraro  <pecoraro@apple.com>
101
102         [Mac] Add Build Phase to Check Headers for Inappropriate Macros (Platform.h macros)
103         https://bugs.webkit.org/show_bug.cgi?id=104279
104
105         Reviewed by David Kilzer.
106
107         Add a build phase script that checks a framework's header files for
108         Platform.h and Compiler.h macros. Also add some tests for the script.
109
110         * Scripts/check-for-inappropriate-macros-in-external-headers: Added.
111         Script to be used in build phases to check for inappropriate macros in headers.
112
113         * Scripts/test-webkitruby: Added.
114         Script that can be used to run all webkitruby tests.
115
116         * Scripts/test-webkit-scripts:
117         Also run the ruby tests.
118
119         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/fake-data-failing-expected.txt: Added.
120         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/pass-expected.txt: Added.
121         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Fail.h: Added.
122         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Pass.h: Added.
123         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/run-test.rb: Added.
124         Tests for the new script.
125
126 2012-12-18  Dominic Mazzoni  <dmazzoni@google.com>
127
128         AX: support clickPoint in DRT for chromium
129         https://bugs.webkit.org/show_bug.cgi?id=97359
130
131         Reviewed by Chris Fleizach.
132
133         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
134         several more tests can pass.
135
136         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
137         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
138         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
139         (WebTestRunner):
140         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
141         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
142         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
143         (AccessibilityUIElement):
144
145 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
146
147         Build fix after r138045.
148
149         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
150         (PerfTestsRunner._collect_tests): Use relative_path. We skip only when this path
151         is not explicitly specified. The condition was backwards but never matched so we
152         always ran the test regardless of whether it's listed in the skipped list or not.
153         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
154         (test_collect_tests_with_ignored_skipped_list): Renamed so that it actually runs.
155
156 2012-12-18  Emil A Eklund  <eae@chromium.org>
157
158         TestWebKitAPI/Tests/WebCore/LayoutUnit fails on mac
159         https://bugs.webkit.org/show_bug.cgi?id=105332
160         
161         Reviewed by Levi Weintraub.
162         
163         Create new test suite for LayoutUnit tests as the KURL tests are
164         of the incompatible TEST_F type.
165
166         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
167         (TestWebKitAPI):
168         (TestWebKitAPI::TEST_F):
169
170 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
171
172         Running a skipped test with run-perf-tests could alert the user
173         https://bugs.webkit.org/show_bug.cgi?id=105219
174
175         Reviewed by Eric Seidel.
176
177         Force run skipped performance tests explicitly specified in the command line.
178         Also warn the user when the specified paths don't exist.
179
180         * Scripts/webkitpy/layout_tests/port/base.py:
181         (Port.relative_test_filename):
182         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
183         (PerfTestsRunner._parse_args):
184         (PerfTestsRunner._collect_tests):
185         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
186         (test_collect_tests_with_skipped_list_and_files):
187
188 2012-12-18  Yaron Friedman  <yfriedman@chromium.org>
189
190         Chrome Android java output is too verbose
191         https://bugs.webkit.org/show_bug.cgi?id=105247
192
193         Reviewed by Tony Chang.
194
195         Pass -quiet to ant to suppress extraneous output.
196
197         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
198         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
199
200 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
201
202         Consider removing --pause-before-testing option
203         https://bugs.webkit.org/show_bug.cgi?id=105248
204
205         Reviewed by Eric Seidel.
206
207         Remove --pause-before-testing option.
208
209         * Tools/Scripts/webkitpy/layout_tests/port/driver.py:
210         (DriverProxy.start): Removed.
211         * Scripts/webkitpy/performance_tests/perftest.py:
212         (PerfTest.__init__):
213         (PerfTest.test_path): Renamed from path_or_url.
214         (PerfTest.run):
215         (PerfTest): Removed "driver" argument.
216         (PerfTest._run_with_driver):
217         (PerfTest.run_single):
218         (ChromiumStylePerfTest.__init__):
219         (PageLoadingPerfTest.__init__):
220         (PageLoadingPerfTest.run_single):
221         (PageLoadingPerfTest._run_with_driver): Renamed from run so that PerfTest.run can create a driver for us.
222         (ReplayPerfTest.__init__):
223         (ReplayPerfTest.prepare):
224         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
225         (TestPageLoadingPerfTest.test_run):
226         (TestPageLoadingPerfTest.test_run_with_memory_output):
227         (TestPageLoadingPerfTest.test_run_with_bad_output):
228         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
229         (PerfTestsRunner._parse_args):
230         (PerfTestsRunner._run_tests_set): Removed the code to create a driver and pause before running the test.
231         (PerfTestsRunner._run_single_test): Removed "driver" argument.
232         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
233         (run_test):
234         (test_run_test_pause_before_testing): Removed.
235
236 2012-12-18  Alberto Garcia  <agarcia@igalia.com>
237
238         [GTK] Tests that time out are not considered as failures.
239         https://bugs.webkit.org/show_bug.cgi?id=105318
240
241         Reviewed by Martin Robinson.
242
243         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
244         (RunGtkAPITests.commandComplete):
245         Consider log lines that end with TIMEOUT as errors.
246
247         * Scripts/run-gtk-tests:
248         (TestRunner.run_tests):
249         Consider timeouts as failures when returning the number of failed
250         tests.
251
252 2012-12-18  Pravin D  <pravin.d@samsung.com>
253
254         Unreviewed, adding secondary email.
255
256         * Scripts/webkitpy/common/config/committers.py:
257
258 2012-12-18  Pravin D  <pravin.d@samsung.com>
259
260         Unreviewed, promoting myself to committer's list.
261
262         * Scripts/webkitpy/common/config/committers.py:
263
264 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
265
266         [EFL] ImageDiff does not build on WebKit2 only builds
267         https://bugs.webkit.org/show_bug.cgi?id=105288
268
269         Reviewed by Kenneth Rohde Christiansen.
270
271         Moved ImageDiff to the efl/ folder inside Tools/ (so we don't pollute
272         the folder with things which are EFL-specific.
273
274         ImageDiff will be built by default now regardless of WebKit API flavor
275         you choose.
276
277         * CMakeLists.txt:
278         * DumpRenderTree/efl/CMakeLists.txt:
279         * efl/ImageDiff/CMakeLists.txt: Added.
280         * efl/ImageDiff/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/efl/ImageDiff.cpp.
281
282 2012-12-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
283
284         [EFL][WK2] When creating new windows, use the opener's context
285         https://bugs.webkit.org/show_bug.cgi?id=105283
286
287         Reviewed by Laszlo Gombos.
288
289         Use the context of the opener if exists.
290
291         * MiniBrowser/efl/main.c:
292         (on_key_down):
293         (on_window_create):
294         (window_create):
295         (elm_main):
296
297 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
298
299         [EFL][WK2] Gardening of TestWebKitAPI tests
300         https://bugs.webkit.org/show_bug.cgi?id=105268
301
302         Reviewed by Kenneth Rohde Christiansen.
303
304         Some tests were marked as failing but are passing and the newly added
305         LayoutUnit should have it's own binary because it conflicts with the
306         fixture used by KURL and is failing on the bots because of that.
307
308         * TestWebKitAPI/CMakeLists.txt:
309         * TestWebKitAPI/PlatformEfl.cmake:
310
311 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
312
313         [chromium] move webintent related methods to TestRunner library
314         https://bugs.webkit.org/show_bug.cgi?id=105164
315
316         Reviewed by Darin Fisher.
317
318         Also deleted the unused deliveredIntentFailure and
319         deliveredIntentResult methods.
320
321         * DumpRenderTree/chromium/DRTTestRunner.cpp:
322         (DRTTestRunner::DRTTestRunner):
323         * DumpRenderTree/chromium/DRTTestRunner.h:
324         (DRTTestRunner):
325         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
326         (WebKit):
327         (WebTestRunner::WebTestDelegate::setCurrentWebIntentRequest):
328         (WebTestRunner::WebTestDelegate::currentWebIntentRequest):
329         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
330         (WebKit):
331         (WebTestProxyBase):
332         (WebTestRunner::WebTestProxy::registerIntentService):
333         (WebTestRunner::WebTestProxy::dispatchIntent):
334         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
335         (WebTestRunner::TestRunner::TestRunner):
336         (WebTestRunner::TestRunner::sendWebIntentResponse):
337         (WebTestRunner):
338         (WebTestRunner::TestRunner::deliverWebIntent):
339         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
340         (TestRunner):
341         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
342         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
343         (WebTestRunner::WebTestProxyBase::registerIntentService):
344         (WebTestRunner):
345         (WebTestRunner::WebTestProxyBase::dispatchIntent):
346         * DumpRenderTree/chromium/WebViewHost.cpp:
347         (WebViewHost::setCurrentWebIntentRequest):
348         (WebViewHost::currentWebIntentRequest):
349         * DumpRenderTree/chromium/WebViewHost.h:
350         (WebKit):
351         (WebViewHost):
352
353 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
354
355         [chromium] move editing callback dumping to the WebTestProxy
356         https://bugs.webkit.org/show_bug.cgi?id=105013
357
358         Reviewed by Darin Fisher.
359
360         To decide whether or not to dump the editing callbacks, the WebTestProxy
361         needs to query the TestRunner. For this purpose, I introduce a
362         WebTestRunner class that just exposes the shouldDumpEditingCallbacks
363         method. In contrast to e.g. WebEventSender, the WebTestRunner is just
364         an interface, and it is not owned by the WebTestInterfaces, but the
365         embedder of the TestRunner library must provide it. This will change
366         once all TestRunner methods are in the library.
367
368         * DumpRenderTree/DumpRenderTree.gypi:
369         * DumpRenderTree/chromium/DRTTestRunner.cpp:
370         (DRTTestRunner::DRTTestRunner):
371         (DRTTestRunner::reset):
372         * DumpRenderTree/chromium/DRTTestRunner.h:
373         (DRTTestRunner):
374         (DRTTestRunner::taskList):
375         (DRTTestRunner::WorkQueue::taskList):
376         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
377         (WebTestRunner):
378         (WebTestInterfaces):
379         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
380         (WebKit):
381         (WebTestRunner):
382         (WebTestProxyBase):
383         (WebTestRunner::WebTestProxy::shouldBeginEditing):
384         (WebTestRunner::WebTestProxy::shouldEndEditing):
385         (WebTestRunner::WebTestProxy::shouldInsertNode):
386         (WebTestRunner::WebTestProxy::shouldInsertText):
387         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
388         (WebTestRunner::WebTestProxy::shouldDeleteRange):
389         (WebTestRunner::WebTestProxy::shouldApplyStyle):
390         (WebTestRunner::WebTestProxy::didBeginEditing):
391         (WebTestRunner::WebTestProxy::didChangeSelection):
392         (WebTestRunner::WebTestProxy::didChangeContents):
393         (WebTestRunner::WebTestProxy::didEndEditing):
394         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h: Copied from Tools/DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h.
395         (WebTestRunner):
396         (WebTestRunner::WebTestRunner::shouldDumpEditingCallbacks):
397         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
398         (WebTestRunner::TestRunner::TestRunner):
399         (WebTestRunner::TestRunner::~TestRunner):
400         (WebTestRunner):
401         (WebTestRunner::TestRunner::reset):
402         (WebTestRunner::TestRunner::shouldDumpEditingCallbacks):
403         (WebTestRunner::TestRunner::dumpEditingCallbacks):
404         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
405         (TestRunner):
406         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
407         (WebTestRunner::WebTestInterfaces::Internal::testRunner):
408         (WebTestRunner::WebTestInterfaces::Internal::setTestRunner):
409         (WebTestInterfaces::Internal):
410         (WebTestRunner::WebTestInterfaces::Internal::Internal):
411         (WebTestRunner::WebTestInterfaces::testRunner):
412         (WebTestRunner):
413         (WebTestRunner::WebTestInterfaces::setTestRunner):
414         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
415         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
416         (WebTestRunner):
417         (WebTestRunner::WebTestProxyBase::shouldBeginEditing):
418         (WebTestRunner::WebTestProxyBase::shouldEndEditing):
419         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
420         (WebTestRunner::WebTestProxyBase::shouldInsertText):
421         (WebTestRunner::WebTestProxyBase::shouldChangeSelectedRange):
422         (WebTestRunner::WebTestProxyBase::shouldDeleteRange):
423         (WebTestRunner::WebTestProxyBase::shouldApplyStyle):
424         (WebTestRunner::WebTestProxyBase::didBeginEditing):
425         (WebTestRunner::WebTestProxyBase::didChangeSelection):
426         (WebTestRunner::WebTestProxyBase::didChangeContents):
427         (WebTestRunner::WebTestProxyBase::didEndEditing):
428         * DumpRenderTree/chromium/TestShell.cpp:
429         (TestShell::initialize):
430         (TestShell::createNewWindow):
431         * DumpRenderTree/chromium/WebViewHost.cpp:
432         (WebViewHost::shouldBeginEditing):
433         (WebViewHost::shouldEndEditing):
434         (WebViewHost::shouldInsertNode):
435         (WebViewHost::shouldInsertText):
436         (WebViewHost::shouldChangeSelectedRange):
437         (WebViewHost::shouldDeleteRange):
438         (WebViewHost::shouldApplyStyle):
439         * DumpRenderTree/chromium/WebViewHost.h:
440         (WebViewHost):
441
442 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
443
444         [chromium] Fix methods in TestRunner that directly modify the settings instead of going through WebPreferences
445         https://bugs.webkit.org/show_bug.cgi?id=105154
446
447         Reviewed by Tony Chang.
448
449         If the settings are modified directly, the embedder can't keep track of
450         them and might accidentially override the changed settings.
451
452         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
453         (WebPreferences):
454         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
455         (WebTestRunner::TestRunner::TestRunner):
456         (WebTestRunner::TestRunner::setAsynchronousSpellCheckingEnabled):
457         (WebTestRunner):
458         (WebTestRunner::TestRunner::setMinimumTimerInterval):
459         (WebTestRunner::TestRunner::setTouchDragDropEnabled):
460         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
461         (TestRunner):
462         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
463         (WebTestRunner::WebPreferences::reset):
464         (WebTestRunner::WebPreferences::applyTo):
465
466 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
467
468         Unreviewed unit test skipping.
469
470         Filed https://bugs.webkit.org/show_bug.cgi?id=105253 to re-enable them.
471
472         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
473         (TestWebKitAPI):
474         (TestWebKitAPI::TEST):
475
476 2012-12-17  Dean Jackson  <dino@apple.com>
477
478         Unreviewed. Add Silvia Pfeiffer to contributor list.
479
480         * Scripts/webkitpy/common/config/committers.py:
481
482 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
483
484         build.webkit.org - add more google windows bots
485         https://bugs.webkit.org/show_bug.cgi?id=105225
486
487         Reviewed by Eric Seidel.
488
489         Adds two more windows bots (google-windows-3, a builder, and
490         google-windows-4, a tester) for the Chromium configs.
491
492         * BuildSlaveSupport/build.webkit.org-config/config.json:
493
494 2012-12-17  Eric Seidel  <eric@webkit.org>
495
496         Cleanup --profiler=perf code a little and allow up to 1000 report files (instead of 100)
497         https://bugs.webkit.org/show_bug.cgi?id=105217
498
499         Reviewed by Dirk Pranke.
500
501         Previously --profile would fail strangely if find_unused_filename returned None
502         (which would happen as soon as we had 100 report files).
503         Now we'll search through 1000 filenames before giving up (and assert if we failed to find one).
504         I also made some minor cleanups to the output of perf, and tried to make the
505         commands more self-documenting by using long-names for args.
506
507         * Scripts/webkitpy/common/system/profiler.py:
508         (SingleFileOutputProfiler.__init__):
509         (Perf.attach_to_pid):
510         (Perf.profile_after_exit):
511
512 2012-12-17  Sheriff Bot  <webkit.review.bot@gmail.com>
513
514         Unreviewed, rolling out r137198.
515         http://trac.webkit.org/changeset/137198
516         https://bugs.webkit.org/show_bug.cgi?id=105212
517
518         This patch is causing API behavior compatibility problems
519         (Requested by zdobersek on #webkit).
520
521         * DumpRenderTree/gtk/DumpRenderTree.cpp:
522         (willSendRequestCallback):
523
524 2012-12-17  Elliott Sprehn  <esprehn@chromium.org>
525
526         Unreviewed. Adding myself to committers list.
527
528         * Scripts/webkitpy/common/config/committers.py:
529
530 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
531
532         nrwt: shuffle code around for cleanup in run_webkit_tests.py
533         https://bugs.webkit.org/show_bug.cgi?id=105078
534
535         Reviewed by Ojan Vafai.
536
537         This patch reorders functions so that run_webkit_tests can be
538         understood in a top-down matter better, and to make things slightly
539         more sensible when I merge manager.py into it.
540
541         Also, this adds tests for the actual main() routine, which had several
542         bugs get through testing a week or two ago.
543
544         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
545         (main):
546         (parse_args):
547         (_set_up_derived_options):
548         (run):
549         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
550         (RunTest):
551         (RunTest.test_no_http_tests):
552         (PortTest.disabled_test_mac_lion):
553         (MainTest):
554         (MainTest.test_exception_handling):
555         (MainTest.test_exception_handling.interrupting_run):
556         (MainTest.test_exception_handling.successful_run):
557         (MainTest.test_exception_handling.successful_run.FakeRunDetails):
558         (MainTest.test_exception_handling.exception_raising_run):
559
560 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
561
562         webkitpy: move --lint-test-files code into its own module
563         https://bugs.webkit.org/show_bug.cgi?id=105077
564
565         Reviewed by Eric Seidel.
566
567         It really didn't need to be in run_webkit_tests.py anymore.
568         This patch adds a new lint-test-expectations scripts to Tools/Scripts;
569         rwt --lint-test-files is still supported for backwards compatibility
570         for now.
571
572         * Scripts/lint-test-expectations: Added.
573         * Scripts/webkitpy/layout_tests/lint_test_expectations.py: Added.
574         (lint):
575         (main):
576         * Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py: Added.
577         (FakePort):
578         (FakePort.__init__):
579         (FakePort.test_configuration):
580         (FakePort.expectations_dict):
581         (FakePort.skipped_layout_tests):
582         (FakePort.all_test_configurations):
583         (FakePort.configuration_specifier_macros):
584         (FakePort.get_option):
585         (FakeFactory):
586         (FakeFactory.__init__):
587         (FakeFactory.get):
588         (FakeFactory.all_port_names):
589         (LintTest):
590         (LintTest.test_all_configurations):
591         (LintTest.test_lint_test_files):
592         (LintTest.test_lint_test_files__errors):
593         (MainTest):
594         (MainTest.test_success):
595         (MainTest.test_success.interrupting_lint):
596         (MainTest.test_success.successful_lint):
597         (MainTest.test_success.exception_raising_lint):
598         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
599         (main):
600         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
601         (StreamTestingMixin.assertNotEmpty):
602
603 2012-12-17  Adam Klein  <adamk@chromium.org>
604
605         build-webkit: rename --template-tag to --template-element to match ENABLE #define name
606         https://bugs.webkit.org/show_bug.cgi?id=105072
607
608         Reviewed by Laszlo Gombos.
609
610         * Scripts/webkitperl/FeatureList.pm:
611
612 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
613
614         Add some unit testing for WTF::clampTo* functions
615         https://bugs.webkit.org/show_bug.cgi?id=105060
616
617         Reviewed by Emil A Eklund.
618
619         The new tests are targeted at clamping to int / unsigned and float as those are
620         the one explicitely exposed through more dedicated functions (like clampToInteger,
621         clampToPositiveInteger and clampToFloat).
622
623         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
624
625 2012-12-17  Eric Seidel  <eric@webkit.org>
626
627         rpt --profile --chromium-android throws exception
628         https://bugs.webkit.org/show_bug.cgi?id=105061
629
630         Reviewed by Daniel Bates.
631
632         Somehow popen(args) got turned into popen(*args) at some point,
633         which caused my new stringify code to fail when args was passed by name.
634         Fixed and tested.
635
636         * Scripts/webkitpy/common/system/executive.py:
637         (Executive._stringify_args):
638         (Executive):
639         (Executive.popen):
640         * Scripts/webkitpy/common/system/executive_unittest.py:
641         (ExecutiveTest.test_popen_args):
642
643 2012-12-17  Mark Pilgrim  <pilgrim@chromium.org>
644
645         Use Platform::current() instead of webKitPlatformSupport() in DumpRenderTree
646         https://bugs.webkit.org/show_bug.cgi?id=105054
647
648         Reviewed by Jochen Eisinger.
649
650         Part of a refactoring series; see tracking bug 82948.
651
652         * DumpRenderTree/chromium/NotificationPresenter.cpp:
653         (NotificationPresenter::show):
654         * DumpRenderTree/chromium/Task.cpp:
655         (postTask):
656         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
657
658 2012-12-17  Emil A Eklund  <eae@chromium.org>
659
660         Clamp values in LayoutUnit::operator/ when SATURATED_LAYOUT_ARITHMETIC is enabled
661         https://bugs.webkit.org/show_bug.cgi?id=104955
662
663         Reviewed by Julien Chaffraix.
664
665         Add tests for LayoutUnit.
666
667         * TestWebKitAPI/CMakeLists.txt:
668         * TestWebKitAPI/ForwardingHeaders: Added.
669         * TestWebKitAPI/ForwardingHeaders/WebCore: Added.
670         * TestWebKitAPI/ForwardingHeaders/WebCore/LayoutUnit.h: Added.
671         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
672         * TestWebKitAPI/TestWebKitAPI.gypi:
673         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
674         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp: Added.
675         (TestWebKitAPI):
676         (TestWebKitAPI::TEST):
677         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
678
679 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
680
681         [chromium] removed unused preferences from testRunner.overridePreference
682         https://bugs.webkit.org/show_bug.cgi?id=105169
683
684         Reviewed by Tony Chang.
685
686         Most notably this removes the code related to font family maps which is
687         now handled via window.internals.settings
688
689         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
690         (WebPreferences):
691         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
692         (WebTestRunner::TestRunner::overridePreference):
693         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
694         (TestRunner):
695         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
696         (WebTestRunner::WebPreferences::applyTo):
697
698 2012-12-17  James Robinson  <jamesr@chromium.org>
699
700         [chromium] Move creation of WebLayerTreeView to WebWidgetClient
701         https://bugs.webkit.org/show_bug.cgi?id=105071
702
703         Reviewed by Adrienne Walker.
704
705         Update WebViewHost for the new WebWidgetClient interface.
706
707         * DumpRenderTree/chromium/WebViewHost.cpp:
708         (WebViewHost::initializeLayerTreeView):
709         (WebViewHost::layerTreeView):
710         (WebViewHost::~WebViewHost):
711         * DumpRenderTree/chromium/WebViewHost.h:
712         (WebViewHost):
713
714 2012-12-17  Kenneth Rohde Christiansen  <kenneth@webkit.org>
715
716         [EFL][WK2] window_create doesn't receive the url
717         https://bugs.webkit.org/show_bug.cgi?id=105184
718
719         Reviewed by Alexis Menard.
720
721         * MiniBrowser/efl/main.c: Use the url supplied to window_create
722         (on_window_create):
723
724 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
725
726         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
727         https://bugs.webkit.org/show_bug.cgi?id=105026
728
729         Reviewed by Yury Semikhatsky.
730
731         Old SFINAE test was replaced with new one based on this article:
732         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
733
734         New test which covers this problem was added.
735
736         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
737
738 2012-12-17  Anthony Scian  <ascian@rim.com>
739
740         [Blackberry] Static code analysis warning fixes
741         https://bugs.webkit.org/show_bug.cgi?id=105129
742
743         Reviewed by Rob Buis.
744
745         Added null checks for fopen calls, no recovery attempted; access fault prevention only.
746
747         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
748         (BlackBerry::WebKit::createFile):
749         (BlackBerry::WebKit::DumpRenderTree::runTest):
750
751 2012-12-17  John Mellor  <johnme@chromium.org>
752
753         Add Anton Vayvod and Tim Volodine as contributors in committers.py
754         https://bugs.webkit.org/show_bug.cgi?id=105025
755
756         Unreviewed. Adds two contributors to committers.py.
757         They are initially contributing to the Text Autosizing feature (master bug: 84186).
758
759         * Scripts/webkitpy/common/config/committers.py:
760
761 2012-12-17  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
762
763         Unreviewed, add myself to committers.py.
764
765         * Scripts/webkitpy/common/config/committers.py:
766
767 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
768
769         [Qt] remove some unnecessary CONFIG additions
770
771         Reviewed by Simon Hausmann.
772
773         qt is already added by spec_pre.prf, warn_on and depend_includepath by
774         default_pre.prf.
775
776         * qmake/mkspecs/features/default_post.prf:
777
778 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
779
780         Unreviewed, rolling out r137892.
781         http://trac.webkit.org/changeset/137892
782         https://bugs.webkit.org/show_bug.cgi?id=105026
783
784         it broke compilation on windows
785
786         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
787
788 2012-12-14  Ilya Tikhonovsky  <loislo@chromium.org>
789
790         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
791         https://bugs.webkit.org/show_bug.cgi?id=105026
792
793         Reviewed by Yury Semikhatsky.
794
795         Old SFINAE test was replaced with new one based on this article:
796         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
797
798         New test which covers this problem was added.
799
800         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
801
802 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
803
804         [Qt] rely on automatic output directory setup
805
806         Reviewed by Simon Hausmann.
807
808         Now being a proper qt module (by virtue of having load(qt_build_config)
809         in .qmake.conf), webkit gets the path setup goodies for free.
810
811         This also fixes Makefile.api.Debug/Release trying to generate qrc_WebKit.cpp
812         at the same time, instead of the debug and release version ending up in different
813         directories as they should.
814
815         * qmake/mkspecs/features/default_pre.prf:
816
817 2012-12-17  Charles Wei  <charles.wei@torchmobile.com.cn>
818
819         [BlackBerry] Need to re-enable video track
820         https://bugs.webkit.org/show_bug.cgi?id=105155
821
822         Reviewed by George Staikos.
823
824         Just enable Video Track for BlackBerry.
825
826         * Scripts/webkitperl/FeatureList.pm:
827
828 2012-12-17  János Badics  <jbadics@inf.u-szeged.hu>
829
830         Unreviewed. Added myself to committers.py
831
832         * Scripts/webkitpy/common/config/committers.py:
833
834 2012-12-17  Jonathan Liu  <net147@gmail.com>
835
836         Fix linking of libraries with -Wl,-whole-archive
837         https://bugs.webkit.org/show_bug.cgi?id=104436
838
839         Reviewed by Simon Hausmann.
840
841         If there is more than one library in LIBS linked with
842         -Wl,-whole-archive -l... -Wl,-no-whole-archive, qmake tries to
843         merge the duplicate flags in LIBS which causes linking to fail.
844
845         Add no_smart_library_merge to CONFIG to prevent the duplicate
846         flags from being removed when using -Wl,-whole-archive.
847
848         * qmake/mkspecs/features/functions.prf:
849
850 2012-12-16  Arvid Nilsson  <anilsson@rim.com>
851
852         [BlackBerry] Adapt to new BlackBerry::Platform::TouchPoint API
853         https://bugs.webkit.org/show_bug.cgi?id=105143
854         RIM PR 171941
855
856         Reviewed by Rob Buis.
857         Internally reviewed by George Staikos.
858
859         Adapt to new method names and encapsulation of TouchPoint data members.
860
861         * DumpRenderTree/blackberry/EventSender.cpp:
862         (addTouchPointCallback):
863         (updateTouchPointCallback):
864         (touchEndCallback):
865         (releaseTouchPointCallback):
866         (sendTouchEvent):
867
868 2012-12-16  Maciej Stachowiak  <mjs@apple.com>
869
870         Add a script to compute the mean and 95% confidence interval (using two-sided t-test) of a sample set
871         https://bugs.webkit.org/show_bug.cgi?id=105148
872
873         Reviewed by Filip Pizlo.
874
875         Usage examples:
876         
877         $ sampstat --help
878         Usage: sampstat [options] [FILES]
879           Compute the mean and 95% confidence interval of a sample set.
880           Standard input or files must contain two or more decimal numbers, one per line.
881         
882         Options:
883           -h, --help            show this help message and exit
884           -u UNIT, --unit=UNIT  assume values are in units of UNIT
885           -v, --verbose         print all values (with units)
886         
887         $ sampstat -u MB memresults.txt
888         2356.90 MB +/- 101.34 MB (4.3%)
889         
890         $ sampstat -v -u MB memresults.txt
891               2318.21 MB
892               2399.56 MB
893               2352.93 MB
894         ----------------
895         Mean: 2356.90 MB +/- 101.34 MB (4.3%)
896         
897
898         * Scripts/sampstat: Added.
899         (sum): Helper function to compute the sum of a list.
900         (arithmeticMean): Compute the meam of a list.
901         (standardDeviation): Compute the sample standard deviation (unbiased estimator).
902         (standardError): Compute the sample standard error.
903         (tDist): Compute t(.025, n-1), the t-value for atwo-sided 95% confidence interval. 
904         (twoSidedConfidenceInterval): Compute the two-sided confidence interval range about the mean.
905
906 2012-12-15  Nima Ghanavatian  <nghanavatian@rim.com>
907
908         [BlackBerry] Pass key modifiers with touch and mouse events
909         https://bugs.webkit.org/show_bug.cgi?id=105108
910
911         Reviewed by Rob Buis.
912
913         Update DRT to new api.
914
915         Internally reviewed by Gen Mak.
916
917         * DumpRenderTree/blackberry/EventSender.cpp:
918         (mouseDownCallback):
919         (mouseUpCallback):
920         (mouseMoveToCallback):
921
922 2012-12-15  Zan Dobersek  <zandobersek@gmail.com>
923
924         Stop the XvfbDriver before proceeding with its starting setup
925         https://bugs.webkit.org/show_bug.cgi?id=105006
926
927         Reviewed by Philippe Normand.
928
929         Just as with the base Driver implementation, stop the XvfbDriver
930         before proceeding with setting up lock files and a new Xvfb instance.
931         This forces any driver or Xvfb instances that are currently under
932         the driver's control to be closed/killed and cleaned up after.
933
934         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
935         (XvfbDriver._start):
936
937 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
938
939         webkitpy: make port_name a required parameter to the non-test ports
940         https://bugs.webkit.org/show_bug.cgi?id=105057
941
942         Reviewed by Eric Seidel.
943
944         This is just minor code cleanup along the path to making Port
945         constructors simpler and pushing more logic into the port factory.
946
947         * Scripts/webkitpy/layout_tests/port/base.py:
948         (Port.__init__):
949         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
950         (PortTest.make_port):
951         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
952         (TestWebKitPort.__init__):
953         * Scripts/webkitpy/layout_tests/port/test.py:
954         (TestPort):
955         (TestPort.determine_full_port_name):
956         (TestPort.__init__):
957         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py:
958         (XvfbDriverTest.make_driver):
959         (XvfbDriverTest.test_stop):
960
961 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
962
963         garden-o-matic doesn't know about reftests
964         https://bugs.webkit.org/show_bug.cgi?id=101976
965
966         Reviewed by Eric Seidel.
967
968         Attempt to make garden-o-matic properly aware of reftests again.
969
970         This is re-landing r137407 with a change to
971         rebaselineWithStatusUpdates() to actually pass the filtered
972         failureInofLost to checkout.rebaseline().
973
974         Unfortunately, the testing coverage for this whole module is almost
975         non-existent, and it's not obvious that there's a good way to
976         stub out checkout.rebaseline() to test that the right thing
977         happens. I verified the change manually.
978
979         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
980         (.):
981         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
982         (.):
983         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
984
985 2012-12-14  Ryosuke Niwa  <rniwa@webkit.org>
986
987         "Running 1 DumpRenderTree over X shards" is not a helpful output
988         https://bugs.webkit.org/show_bug.cgi?id=104858
989
990         Reviewed by Dirk Pranke.
991
992         The word "shard" is not typically used in test frameworks. This word is confusing for many WebKit
993         contributors who typically don't have much experience on distributed computations and databases.
994         So only print this message out in the debug mode.
995
996         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
997         * Scripts/webkitpy/layout_tests/views/printing.py:
998         (Printer.print_workers_and_shards):
999
1000 2012-12-13  Eric Seidel  <eric@webkit.org>
1001
1002         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
1003         https://bugs.webkit.org/show_bug.cgi?id=104891
1004
1005         Reviewed by Ryosuke Niwa.
1006
1007         I also implemented a very simple "Sample" Profiler using
1008         Mac OS X's /usr/bin/sample command line tool.
1009
1010         The real reason for this abstraction is to make it easy
1011         to support both perf and pprof on linux which seem to
1012         be about equally popular among those I ask in the Chrome team.
1013
1014         * Scripts/webkitpy/common/system/profiler.py:
1015         (ProfilerFactory.create_profiler):
1016         (ProfilerFactory):
1017         (ProfilerFactory.available_profilers_by_name):
1018         (ProfilerFactory.default_profiler_name):
1019         (Sample):
1020         (Sample.__init__):
1021         (Sample.attach_to_pid):
1022         (Sample.profile_after_exit):
1023         (IProfiler.attach_to_pid):
1024         * Scripts/webkitpy/layout_tests/port/driver.py:
1025         (Driver.__init__):
1026         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1027         (PerfTestsRunner._parse_args):
1028
1029 2012-12-13  Eric Seidel  <eric@webkit.org>
1030
1031         Unreviewed, fixing typo in python unittest.
1032
1033         I changed the default profiler on linux from pprof to perf in
1034         bug 104971.  I failed to update the unittest results at that time.
1035
1036         * Scripts/webkitpy/common/system/profiler_unittest.py:
1037         (ProfilerFactoryTest.test_basic):
1038
1039 2012-12-13  Ian Vollick  <vollick@chromium.org>
1040
1041         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
1042         https://bugs.webkit.org/show_bug.cgi?id=104911
1043
1044         Reviewed by Dirk Pranke.
1045
1046         Adds two virtual tests suites to run the tests in
1047         compositing/overflow/ and scrollbars/ with the flag
1048         --enable-accelerated-overflow-scroll, which I've also plumbed through
1049         in the usual way.
1050
1051         * DumpRenderTree/chromium/DumpRenderTree.cpp:
1052         (main):
1053         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1054         (WebPreferences):
1055         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1056         (WebTestRunner::WebPreferences::reset):
1057         (WebTestRunner::WebPreferences::applyTo):
1058         * DumpRenderTree/chromium/TestShell.cpp:
1059         (TestShell::TestShell):
1060         (TestShell::resetWebSettings):
1061         * DumpRenderTree/chromium/TestShell.h:
1062         (TestShell::setAcceleratedCompositingForOverflowScrollEnabled):
1063         (TestShell):
1064         * Scripts/webkitpy/layout_tests/port/chromium.py:
1065         (ChromiumPort.virtual_test_suites):
1066
1067 2012-12-13  Eric Seidel  <eric@webkit.org>
1068
1069         Use 'perf' to profile on linux instead of google-pprof by default
1070         https://bugs.webkit.org/show_bug.cgi?id=104971
1071
1072         Reviewed by Dirk Pranke.
1073
1074         This makes Chromium Linux match Chromium Android and use the perf
1075         tool by default.  Once https://bugs.webkit.org/show_bug.cgi?id=104891
1076         lands it will be possible to use pprof again on Linux.
1077
1078         This is slightly more advanced than the AndroidPerf profiler as
1079         instead of using a timeout on "perf record" I instead watch
1080         for the termination of the target process and then control-C
1081         the 'perf record' process.  This required me to add two new
1082         methods to Executive, one to have a limited-time wait() and
1083         the second to be able to send a control-C.  I chose to add
1084         these to Executive to make them easier to mock/fix-for-win32
1085         at a later time if needed.
1086
1087         * Scripts/webkitpy/common/system/executive.py:
1088         (Executive.wait_limited):
1089         (Executive.interrupt):
1090         * Scripts/webkitpy/common/system/profiler.py:
1091         (ProfilerFactory.create_profiler):
1092         (Perf):
1093         (Perf.__init__):
1094         (Perf._perf_path):
1095         (Perf.attach_to_pid):
1096         (Perf._first_ten_lines_of_profile):
1097         (Perf.profile_after_exit):
1098
1099 2012-12-13  Julie Parent  <jparent@chromium.org>
1100
1101         Dashboard cleanup: remove usage of global g_builders.
1102         https://bugs.webkit.org/show_bug.cgi?id=104941
1103
1104         Reviewed by Dirk Pranke.
1105
1106         The dashboards use a lot of global state, which makes hacking on them
1107         complicated. This change removes the use of one such global: g_builders.
1108         In most cases, we can just use currentBuilderGroup().builders instead,
1109         which is now currentBuilders().
1110         Surprisingly, the most changes were required to the unit tests, since
1111         they were even bigger offenders of bad hygiene, relying on global state
1112         set by other tests, randomly clobbering global variables in ways the
1113         real code doesn't, etc.
1114
1115         * TestResultServer/static-dashboards/builders.js:
1116         (BuilderGroup.prototype.setup):
1117         * TestResultServer/static-dashboards/dashboard_base.js:
1118         (.switch.return):
1119         (htmlForTestTypeSwitcher):
1120         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1121         (generatePage):
1122         (getAllTestsTrie):
1123         (processTestRunsForAllBuilders):
1124         (showPopupForBuild):
1125         (generatePageForExpectationsUpdate):
1126         (loadExpectationsLayoutTests):
1127         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1128         (resetGlobals):
1129         (stubResultsByBuilder):
1130         (test):
1131         * TestResultServer/static-dashboards/loader.js:
1132         (.):
1133         * TestResultServer/static-dashboards/loader_unittests.js:
1134
1135 2012-12-13  Eric Seidel  <eric@webkit.org>
1136
1137         Callers should not have to stringify args before calling Executive run_command/popen
1138         https://bugs.webkit.org/show_bug.cgi?id=104975
1139
1140         Reviewed by Dirk Pranke.
1141
1142         One could argue that we should match the python call syntax here,
1143         but I think it's a more friendly API if we automagically handle
1144         stringification of args in run_command, etc.
1145         This removes map(unicode, args) from several callsites.
1146
1147         When I first tried to land this change, I didn't realize that
1148         Executive._command_for_printing depended on this behavior
1149         having been applied to args in run_command.  The fix is to
1150         call _stringify_args in both run_command and popen.
1151         This is slightly redundant, but given how short args have to be
1152         (due to shell limits), I don't think the double-encode check
1153         matters in practice.
1154
1155         This is slightly complicated by the fact that apache_http_server.py
1156         is the one caller in our codebase which uses shell=True.
1157         shell=True is a well-documented trail-of-tears:
1158         http://stackoverflow.com/questions/3172470/actual-meaning-of-shell-true-in-subprocess
1159         but to support this legacy (windows-only) code (which I can't easily test)
1160         I've added an if-hack to avoid stringifying the the popen(shell=True) case.
1161
1162         * Scripts/webkitpy/common/system/executive.py:
1163         (Executive.run_command):
1164         (Executive._stringify_args):
1165         (Executive.popen):
1166         * Scripts/webkitpy/common/system/executive_unittest.py:
1167         (ExecutiveTest.test_auto_stringify_args):
1168         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1169         (attach_to_pid):
1170
1171 2012-12-13  Adrienne Walker  <enne@chromium.org>
1172
1173         Unreviewed, rolling out r137645, r137646, and r137667.
1174         http://trac.webkit.org/changeset/137645
1175         http://trac.webkit.org/changeset/137646
1176         http://trac.webkit.org/changeset/137667
1177         https://bugs.webkit.org/show_bug.cgi?id=104911
1178
1179         Breaks some overflow layout tests
1180
1181         * Scripts/webkitpy/layout_tests/port/chromium.py:
1182         (ChromiumPort.virtual_test_suites):
1183
1184 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1185
1186         Unreviewed, rolling out r137661.
1187         http://trac.webkit.org/changeset/137661
1188         https://bugs.webkit.org/show_bug.cgi?id=104891
1189
1190         broke unit tests, run-webkit-tests
1191
1192         * Scripts/webkitpy/common/system/executive.py:
1193         (Executive._run_command_with_teed_output):
1194         (Executive.run_command):
1195         (Executive.popen):
1196         * Scripts/webkitpy/common/system/profiler.py:
1197         (ProfilerFactory.create_profiler):
1198         (GooglePProf.profile_after_exit):
1199         (IProfiler.attach_to_pid):
1200         * Scripts/webkitpy/layout_tests/port/driver.py:
1201         (Driver.__init__):
1202         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1203         (PerfTestsRunner._parse_args):
1204
1205 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1206
1207         nrwt: move rundetails, summarize_results into test_run_results
1208         https://bugs.webkit.org/show_bug.cgi?id=104963
1209
1210         Reviewed by Eric Seidel.
1211
1212         More cleanup; this the plain-old-data functions out of manager.py
1213         and into test_run_results next to the other aggregate data structures.
1214
1215         Also this cleans up a bunch of test code and reduces duplication.
1216
1217         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1218         (Manager.run):
1219         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1220         (ManagerTest.test_look_for_new_crash_logs):
1221         * Scripts/webkitpy/layout_tests/models/test_run_results.py:
1222         (TestRunResults.__init__):
1223         (TestRunResults.add):
1224         (RunDetails):
1225         (RunDetails.__init__):
1226         (_interpret_test_failures):
1227         (summarize_results):
1228         * Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py: Added.
1229         (get_result):
1230         (run_results):
1231         (summarized_results):
1232         (InterpretTestFailuresTest):
1233         (InterpretTestFailuresTest.setUp):
1234         (InterpretTestFailuresTest.test_interpret_test_failures):
1235         (SummarizedResultsTest):
1236         (SummarizedResultsTest.setUp):
1237         (SummarizedResultsTest.test_no_svn_revision):
1238         (SummarizedResultsTest.test_svn_revision):
1239         (SummarizedResultsTest.test_summarized_results_wontfix):
1240         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
1241         (BuildBotPrinterTests.get_printer):
1242         (BuildBotPrinterTests.test_print_unexpected_results):
1243         (BuildBotPrinterTests.test_print_results):
1244
1245 2012-12-13  Eric Seidel  <eric@webkit.org>
1246
1247         Make --profile output easier to understand for pprof on linux
1248         https://bugs.webkit.org/show_bug.cgi?id=104957
1249
1250         Reviewed by Dirk Pranke.
1251
1252         pprof is kinda mysterious, so we print some information
1253         on how to use it, and how to learn how to read the sample dump.
1254         This is similar to what we've done for perf output for Chromium Android.
1255
1256         * Scripts/webkitpy/common/system/profiler.py:
1257         (GooglePProf._pprof_path):
1258         (GooglePProf.profile_after_exit):
1259
1260 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1261
1262         Revert r137407 - it didn't actually work right.
1263         https://bugs.webkit.org/show_bug.cgi?id=101976
1264
1265         Unreviewed, build fix.
1266
1267         Turns out that rebaselining was completely broken with this change :(
1268         We're passing in a list of test names rather than the list of
1269         failureInfo objects.
1270
1271         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1272         (.):
1273         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1274         (.):
1275         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
1276
1277 2012-12-13  Antoine Labour  <piman@chromium.org>
1278
1279         [chromium] Small fixes for WebPluginContainer::setWebLayer
1280         https://bugs.webkit.org/show_bug.cgi?id=104953
1281
1282         Reviewed by James Robinson.
1283
1284         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Clean up
1285         dangling pointers, force correct invalidation of the layer.
1286
1287 2012-12-13  Eric Seidel  <eric@webkit.org>
1288
1289         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
1290         https://bugs.webkit.org/show_bug.cgi?id=104891
1291
1292         Reviewed by Dirk Pranke.
1293
1294         I also implemented a very simple "Sample" Profiler using
1295         Mac OS X's /usr/bin/sample command line tool.
1296
1297         I also moved the map(unicode, args) call into Executive.popen
1298         so that all callers don't have to do it themselves.
1299
1300         The real reason for this abstraction is to make it easy
1301         to support both perf and pprof on linux which seem to
1302         be about equally popular among those I ask in the Chrome team.
1303
1304         * Scripts/webkitpy/common/system/executive.py:
1305         (Executive.popen):
1306         * Scripts/webkitpy/common/system/profiler.py:
1307         (ProfilerFactory.create_profiler):
1308         (ProfilerFactory):
1309         (ProfilerFactory.available_profilers_by_name):
1310         (ProfilerFactory.default_profiler_name):
1311         (Sample):
1312         (Sample.__init__):
1313         (Sample.attach_to_pid):
1314         (Sample.profile_after_exit):
1315         (IProfiler.attach_to_pid):
1316         * Scripts/webkitpy/layout_tests/port/driver.py:
1317         (Driver.__init__):
1318         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1319         (PerfTestsRunner._parse_args):
1320
1321 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1322
1323         support -wk2 port names properly in webkitpy.layout_tests.port
1324         https://bugs.webkit.org/show_bug.cgi?id=104761
1325
1326         Reviewed by Eric Seidel.
1327
1328         Our support for WK2 port names was inconsistent at best; this patch
1329         modifies the code so that <port_name>-wk2 is supported for all ports that
1330         have WK2 implementations and is equivalent to specifying <port_name>
1331         and -2/--webkit-test-runner. In addition, this modifies
1332         builders.all_port_names() to include the wk2 variants.
1333
1334         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
1335         * Scripts/webkitpy/layout_tests/port/apple.py:
1336         (ApplePort.determine_full_port_name):
1337         * Scripts/webkitpy/layout_tests/port/base.py:
1338         (Port.determine_full_port_name):
1339         (Port.__init__):
1340         * Scripts/webkitpy/layout_tests/port/builders.py:
1341         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
1342         (DriverTest.make_port):
1343         * Scripts/webkitpy/layout_tests/port/mac.py:
1344         (MacPort.default_baseline_search_path):
1345         * Scripts/webkitpy/layout_tests/port/win.py:
1346         (WinPort.default_baseline_search_path):
1347         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1348         (TestRebaselineExpectations.test_rebaseline_expectations):
1349
1350 2012-12-13  Ian Vollick  <vollick@chromium.org>
1351
1352         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
1353         https://bugs.webkit.org/show_bug.cgi?id=104911
1354
1355         Reviewed by Adrienne Walker.
1356
1357         Adds two virtual tests suites to run the tests in
1358         compositing/overflow/ and scrollbars/ with the flag
1359         --enable-accelerated-compositing-for-overflow-scroll
1360
1361         * Scripts/webkitpy/layout_tests/port/chromium.py:
1362         (ChromiumPort.virtual_test_suites):
1363
1364 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1365
1366         delete extraneous self.fail in mastercfg_unittest.py
1367         https://bugs.webkit.org/show_bug.cgi?id=104940
1368
1369         Reviewed by Eric Seidel.
1370
1371         This was accidentally introduced in r137404.
1372
1373         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
1374
1375 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1376
1377         Fix a mis-named variable after r137621.
1378         https://bugs.webkit.org/show_bug.cgi?id=104871
1379
1380         Unreviewed, build fix.
1381
1382         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
1383         (BuildBotPrinter.print_results):
1384         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
1385
1386 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1387
1388         webkitpy: rename ResultSummary
1389         https://bugs.webkit.org/show_bug.cgi?id=104871
1390
1391         Reviewed by Eric Seidel.
1392
1393         This patch changes the names of ResultSummary and related modules,
1394         classes, and variables as a part of the larger clean up in bug 103824.
1395
1396         This change is basically a bunch of search and replaces:
1397         - result_summary.py -> test_run_results.py
1398         - class ResultSummary -> class TestRunResults
1399         - "result_summary" or "rs" as a generic variable name -> "run_results",
1400           when it doesn't matter (or we don't know) if this refers to the
1401           initial pass over the tests or the retry pass.
1402         - "result_summary" as a variable containing the results of the initial,
1403           complete test run -> "initial_results"
1404         - "retry_summary" as a variable containing the results of retried
1405           results -> "retry_results"
1406         - ResultSummary.results (the property) -> TestRunResults.results_by_name
1407         - ResultSummary.failures (the property) -> TestRunResults.failures_by_name
1408         - RunDetails.result_summary -> RunDetails.initial_results
1409         - RunDetails.retry_summary -> RunDetails.retry_results
1410
1411         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
1412         (LayoutTestRunner.__init__):
1413         (LayoutTestRunner.run_tests):
1414         (LayoutTestRunner._mark_interrupted_tests_as_skipped):
1415         (LayoutTestRunner._interrupt_if_at_failure_limits):
1416         (LayoutTestRunner._interrupt_if_at_failure_limits.interrupt_if_at_failure_limit):
1417         (LayoutTestRunner._update_summary_with_result):
1418         (LayoutTestRunner._handle_finished_test):
1419         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
1420         (FakePrinter.print_expected):
1421         (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
1422         (LayoutTestRunnerTests.test_update_summary_with_result):
1423         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1424         (RunDetails.__init__):
1425         (summarize_results):
1426         (Manager.run):
1427         (Manager._look_for_new_crash_logs):
1428         (Manager._tests_to_retry):
1429         (Manager._upload_json_files):
1430         (Manager._stats_trie):
1431         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1432         (ManagerTest.test_look_for_new_crash_logs):
1433         (SummarizeResultsTest):
1434         (SummarizeResultsTest.get_run_results):
1435         (SummarizeResultsTest.summarized_results):
1436         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
1437         (JSONLayoutResultsGenerator.__init__):
1438         (JSONLayoutResultsGenerator._get_modifier_char):
1439         (JSONLayoutResultsGenerator._insert_failure_summaries):
1440         (JSONLayoutResultsGenerator._get_failure_summary_entry):
1441         * Scripts/webkitpy/layout_tests/models/test_run_results.py: Renamed from Tools/Scripts/webkitpy/layout_tests/models/result_summary.py.
1442         (TestRunResults):
1443         (TestRunResults.__init__):
1444         (TestRunResults.add):
1445         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1446         (get_test_results):
1447         (MainTest.test_basic):
1448         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
1449         (BuildBotPrinter.print_results):
1450         (BuildBotPrinter.print_run_results):
1451         (BuildBotPrinter._print_run_results_entry):
1452         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
1453         (BuildBotPrinterTests.get_run_results):
1454         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
1455         (BuildBotPrinterTests):
1456         * Scripts/webkitpy/layout_tests/views/printing.py:
1457         (Printer.print_expected):
1458         (Printer._print_expected_results_of_type):
1459         (Printer.print_results):
1460         (Printer._print_timing_statistics):
1461         (Printer._print_worker_statistics):
1462         (Printer._print_aggregate_test_statistics):
1463         (Printer._print_individual_test_times):
1464         (Printer._print_directory_timings):
1465         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
1466
1467 2012-12-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1468
1469         [EFL] The WebKit2 bots are building and executing WebKit1 tests
1470         https://bugs.webkit.org/show_bug.cgi?id=104690
1471
1472         Reviewed by Csaba Osztrogonác.
1473
1474         Pass --no-webkit1 option to the EFL WK2 bots when building. This
1475         will disable WK1 API and thus, the WK1 API tests.
1476
1477         * BuildSlaveSupport/build.webkit.org-config/config.json:
1478         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1479         (CompileWebKit2Only):
1480         (BuildAndTestFactory):
1481         (BuildAndTestFactory.__init__):
1482         (BuildAndTestWebKit2OnlyFactory):
1483         Now bots using BuildAndTestWebKit2OnlyFactory (for instance only EFL) will
1484         get --no-webkit1 as a parameter on the build-webkit script.
1485         * Scripts/build-webkit:
1486         Added support for -no-webkit1 for the Qt and EFL port.
1487
1488 2012-12-13  Eric Seidel  <eric@webkit.org>
1489
1490         build-webkit --chromium-android forces Make even when update-webkit used --ninja
1491         https://bugs.webkit.org/show_bug.cgi?id=104839
1492
1493         Reviewed by Tony Chang.
1494
1495         This was recommended by Peter as part of:
1496         https://bugs.webkit.org/show_bug.cgi?id=104434#c4
1497         Fixing this makes build-webkit --chromium-android work for the common (ninja) case.
1498
1499         * Scripts/webkitdirs.pm:
1500         (buildChromium):
1501
1502 2012-12-13  Julie Parent  <jparent@chromium.org>
1503
1504         Loader_unittests should reset state between runs.
1505         https://bugs.webkit.org/show_bug.cgi?id=104862
1506
1507         Reviewed by Dirk Pranke.
1508
1509         The loader_unittests were not resetting state between runs, leaving
1510         a lot of data behind in global objects that the dashboard and tests
1511         rely on.  A handful were resetting just a few objects.  Use the common
1512         resetGlobals() method instead.
1513
1514         * TestResultServer/static-dashboards/loader_unittests.js:
1515
1516 2012-12-13  Andras Becsi  <andras.becsi@digia.com>
1517
1518         [Qt][WK2] Fix painting on Mac with retina display
1519         https://bugs.webkit.org/show_bug.cgi?id=104574
1520
1521         Reviewed by Kenneth Rohde Christiansen.
1522
1523         Remove setting the devicePixelRatio experimental property
1524         since the value is now automatically picked up from Qt.
1525
1526         * MiniBrowser/qt/qml/BrowserWindow.qml:
1527
1528 2012-12-13  Jussi Kukkonen  <jussi.kukkonen@intel.com>
1529
1530         [EFL][GTK] Don't call deprecated g_type_init when glib => 2.35
1531         https://bugs.webkit.org/show_bug.cgi?id=103209
1532
1533         Reviewed by Kenneth Rohde Christiansen.
1534
1535         g_type_init() is deprecated from 2.35.0 onwards. Don't call it
1536         in that case to avoid warnings and/or build failure.
1537
1538         * DumpRenderTree/gtk/ImageDiff.cpp:
1539         (main):
1540
1541 2012-12-13  Claudio Saavedra  <csaavedra@igalia.com>
1542
1543         [Tools] Add exception for gtk_style_context_get()
1544         https://bugs.webkit.org/show_bug.cgi?id=104896
1545
1546         Reviewed by Martin Robinson.
1547
1548         Add new exception for gtk_style_context_get().
1549
1550         * Scripts/webkitpy/style/checkers/cpp.py:
1551         (check_for_null): Add exception for gtk_style_context_get().
1552         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1553         (WebKitStyleTest.test_null_false_zero): Add a unit
1554         test for the new exception.
1555
1556 2012-12-13  Joaquim Rocha  <jrocha@igalia.com>
1557
1558         REGRESSION (r137432): The /webkit2/WebKitCookieManager/accept-policy unit test is failing
1559         https://bugs.webkit.org/show_bug.cgi?id=104790
1560
1561         Reviewed by Carlos Garcia Campos.
1562
1563         Unskip the GTK accept policy test.
1564
1565         * Scripts/run-gtk-tests:
1566         (TestRunner):
1567
1568 2012-12-13  Xingnan Wang  <xingnan.wang@intel.com>
1569
1570         Unreviewed. Adding myself to committers list.
1571
1572         * Scripts/webkitpy/common/config/committers.py:
1573
1574 2012-12-12  Roger Fong  <roger_fong@apple.com>
1575
1576         Unreviewed. Fix typo in NRWT.
1577
1578         * Scripts/webkitpy/layout_tests/port/base.py:
1579         (Port.is_reference_html_file):
1580
1581 2012-12-12  Roger Fong  <roger_fong@apple.com>
1582
1583         Fix detection of ref test files in old and new run webkit tests.
1584         https://bugs.webkit.org/show_bug.cgi?id=104850.
1585
1586         Reviewed by Timothy Horton.
1587
1588         * Scripts/old-run-webkit-tests:
1589         (isUsedInReftest): 
1590         Add notref and ref prefix and suffix cases to isUsedinReftest.
1591         * Scripts/webkitpy/layout_tests/port/base.py:
1592         (Port):
1593         (Port.is_reference_html_file): 
1594         Typo in NRWT.
1595
1596 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
1597
1598         rework ChromiumAndroidDriver.start() et. al to avoid funky layering
1599         https://bugs.webkit.org/show_bug.cgi?id=104856
1600
1601         Reviewed by Eric Seidel.
1602
1603         r137523 / bug 10357 introduced some back-and-forth hopping during
1604         driver.start() in the chromium android port. This patch cleans it
1605         up a bit.
1606
1607         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1608         (ChromiumAndroidDriver.__init__):
1609         (ChromiumAndroidDriver._start_once):
1610         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1611         * Scripts/webkitpy/layout_tests/port/driver.py:
1612         (Driver.start):
1613         (Driver._start):
1614
1615 2012-12-12  Eric Seidel  <eric@webkit.org>
1616
1617         Add support for run-perf-tests --chromium-android --profile
1618         https://bugs.webkit.org/show_bug.cgi?id=103587
1619
1620         Reviewed by Dirk Pranke.
1621
1622         This makes it possible to profile PerformanceTests on an Android device
1623         using run-perf-tests --chromium-android --profile.
1624         CPU profiling on Android is sadly non-trivial, but this goes a long
1625         ways towards making it more accessible to non-Android developers.
1626
1627         Although this script requires your device to be flashed with a
1628         userdebug build (an annoyingly non-trivial endeavor for external contributors)
1629         it endeavors to detect and explain the common errors encountered
1630         while trying to setup a device for perf-based profiling.
1631
1632         * Scripts/webkitpy/common/system/profiler.py:
1633         (GooglePProf.profile_after_exit): Discovered that pprof will just hang if passed a missing file.
1634         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1635         (AndroidPerf):
1636         (AndroidPerf.__init__):
1637         (AndroidPerf.check_configuration):
1638         (AndroidPerf.print_setup_instructions):
1639         (_file_exists_on_device):
1640         (_run_adb_command):
1641         (attach_to_pid):
1642         (_perf_version_string):
1643         (_find_perfhost_binary):
1644         (_perfhost_path):
1645         (_first_ten_lines_of_profile):
1646         (profile_after_exit):
1647         (ChromiumAndroidDriver.__init__):
1648         (ChromiumAndroidDriver.__init__.methods):
1649         (ChromiumAndroidDriver._find_or_create_symfs):
1650         (ChromiumAndroidDriver._start_once):
1651         (ChromiumAndroidDriver):
1652         (ChromiumAndroidDriver._run_post_start_tasks):
1653         (ChromiumAndroidDriver._pid_from_android_ps_output):
1654         (ChromiumAndroidDriver._pid_on_target):
1655         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1656         (ChromiumAndroidDriverTest.test_pid_from_android_ps_output):
1657         (AndroidPerfTest):
1658         (AndroidPerfTest.test_perf_output_regexp):
1659         * Scripts/webkitpy/layout_tests/port/driver.py:
1660         (Driver._setup_environ_for_driver):
1661         (Driver):
1662         (Driver._start):
1663         (Driver._run_post_start_tasks):
1664         (Driver._pid_on_target):
1665
1666 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
1667
1668         nrwt: remove deprecated ORWT compatibility command line flags
1669         https://bugs.webkit.org/show_bug.cgi?id=104301
1670
1671         Reviewed by Eric Seidel.
1672
1673         The command line flags for --noshow-results, --[no-]launch-safari,
1674         and --use-remote-links-to-tests were for compatibility with
1675         old-run-webkit-tests, but the compatibility is not needed any more
1676         since NRWT is basically default and the scripts that passed the
1677         old flags have been updated.
1678
1679         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1680         (_set_up_derived_options):
1681         (parse_args):
1682
1683 2012-12-12  John Griggs  <jgriggs@rim.com>
1684
1685         2012-12-12  John Griggs  <jgriggs@rim.com>
1686
1687         [BlackBerry] Update Media Controls
1688         https://bugs.webkit.org/show_bug.cgi?id=104443
1689         https://przilla.ott.qnx.com/bugzilla/show_bug.cgi?id=204748
1690
1691         Update Media Controls for BlackBerry platform to allow audio, embedded video and fullscreen video controls to be positioned and styled differently.
1692
1693         Reviewed by Rob Buis.
1694
1695         BlackBerry-specific update to Media Controls - No new tests required.
1696
1697         * Scripts/webkitperl/FeatureList.pm:
1698
1699 2012-12-12  Justin Novosad  <junov@google.com>
1700
1701         Adding myself to SkiaGraphics watchlist
1702
1703         Unreviewed
1704
1705         * Scripts/webkitpy/common/config/watchlist:
1706
1707 2012-12-12  Martin Robinson  <mrobinson@igalia.com>
1708
1709         [GTK] Add authentication support to DRT and fix exposed issues in the libsoup backend
1710         https://bugs.webkit.org/show_bug.cgi?id=104285
1711
1712         Reviewed by Gustavo Noronha Silva.
1713
1714         Add support to DumpRenderTree for running authentication tests. Since the DRT
1715         expects an authentication callback, we add one to DRTSupport to avoid #ifdefs
1716         in platform-independent code for GTK+.
1717
1718         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1719         (resetDefaultsToConsistentValues): Reset the authentication password and username.
1720         (authenticationCallback): Added.
1721         (createWebView): Attach the authentiation callback during startup.
1722         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
1723         (soupURIToKURL): soup_uri_to_string does not preserve passwords embedded
1724         in the URL so we add a somewhat messy method of re-adding them when they exist.
1725         It would be nice to use soupURIToKURL here, but it seems we cannot use KURL without
1726         pulling in lots of WebCore code so we use string search and replace.
1727         (TestRunner::queueLoad): Use the new helper.
1728
1729 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
1730
1731         garden-o-matic should prefer efl/ over efl-wk1/ and efl-wk2/ when rebaselining
1732         https://bugs.webkit.org/show_bug.cgi?id=102389
1733
1734         Reviewed by Tony Chang.
1735
1736         The code wasn't aware of both the efl-wk1 and efl-wk2 variants,
1737         because all_port_names() by itself doesn't tell you about the
1738         variants. Work around this for now, and filed bug 104761 to
1739         address the underlying problem.
1740
1741         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
1742         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
1743         (BaselineOptimizerTest.test_efl):
1744
1745 2012-12-12  Krzysztof Czech  <k.czech@samsung.com>
1746
1747         [EFL] Possibility to turn off accessibility feature for WebKit-EFL.
1748         https://bugs.webkit.org/show_bug.cgi?id=103036
1749
1750         Reviewed by Laszlo Gombos.
1751
1752         Guard dependencies for accessibility (ATK), so that they can be turned off.
1753
1754         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
1755         (AccessibilityUIElement):
1756         * WebKitTestRunner/PlatformEfl.cmake:
1757
1758 2012-12-12  Simon Fraser  <simon.fraser@apple.com>
1759
1760         Add some entries in the watchlist, and register me for them.
1761
1762         * Scripts/webkitpy/common/config/watchlist:
1763
1764 2012-12-12  Ilya Tikhonovsky  <loislo@chromium.org>
1765
1766         Web Inspector: Native Memory Instrumentation: remove fake root MemoryObjectInfo.
1767         https://bugs.webkit.org/show_bug.cgi?id=104796
1768
1769         Reviewed by Yury Semikhatsky.
1770
1771         It was not a good idea to introduce a fake root MemoryObjectInfo.
1772         It makes a problem when we visit an object without its own MemoryObjectType.
1773
1774         Example: RenderBox has a global pointer to a hash map.
1775         HashMap doesn't have its own object type because it is a generic container.
1776         It will inherit object type from the fake root memory object info.
1777         The same could happen for another container in another class with other MemoryObjectType.
1778
1779         This fact forces me to create custom process method for root objects
1780         because they need to have their own MemoryObjectInfo with customisable memory object type.
1781
1782         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1783
1784         Drive by fix: InstrumentedPointer* was replaced with Wrapper* because actually it is using
1785         for instrumented and not instrumented object classes.
1786
1787 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
1788
1789         [GTK] Remove the last of unnecessary configuration options in configure.ac
1790         https://bugs.webkit.org/show_bug.cgi?id=104793
1791
1792         Reviewed by Martin Robinson.
1793
1794         Removing the filters and indexed database features from the list of features
1795         being configurable through the configure script. This is not the case anymore
1796         as the corresponding configuration options are being removed.
1797
1798         * Scripts/webkitdirs.pm:
1799         (buildAutotoolsProject):
1800
1801 2012-12-12  Mario Sanchez Prada  <mario@webkit.org>
1802
1803         [GTK][WK2] New API to detect display/execution of insecure content
1804         https://bugs.webkit.org/show_bug.cgi?id=104578
1805
1806         Reviewed by Martin Robinson.
1807
1808         Add coding style exception for 'readability/enum_casing' in GTK.
1809
1810         The GTK+ API use upper case, underscore separated, words in
1811         certain types of enums (e.g. signals, properties).
1812
1813         * Scripts/webkitpy/style/checker.py: Added exception, both for
1814         WebKitGTK+ and WebKit2GTK+.
1815
1816 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
1817
1818         [chromium] Fix build on chromium-win
1819
1820         Unreviewed build-fix.
1821
1822         MSVS indeed doesn't have snprintf, adding wtf/StringExtras.h include
1823         to work around that.
1824
1825         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
1826
1827 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
1828
1829         [Qt] Fix build on Mac
1830
1831         Reviewed by Csaba Osztrogonác.
1832
1833         On Mac OS X we have sqlite as a system library available.
1834
1835         * qmake/mkspecs/features/features.prf:
1836
1837 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
1838
1839         Unreviewed GTK gardening.
1840
1841         Skipping another unit test that regressed in r137432.
1842
1843         * Scripts/run-gtk-tests:
1844         (TestRunner):
1845
1846 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
1847
1848         [Qt] Fix linkage against SQLite
1849         https://bugs.webkit.org/show_bug.cgi?id=104781
1850
1851         Reviewed by Csaba Osztrogonác.
1852
1853         Detect SQLite either through pkg-config, from the SQLITE3SRCDIR
1854         environment variable or from qtbase via a qt5.git build. Bail out early
1855         with an error message if neither option applies.
1856
1857         This removes the usage of $$QT.core.sources and the dependency on Qt's
1858         system-sqlite configure setting. We should always favour a system
1859         library over a copy in Qt.
1860
1861         * qmake/mkspecs/features/features.prf:
1862
1863 2012-12-12  Tommy Widenflycht  <tommyw@google.com>
1864
1865         MediaStream API: Change the data channel descriptor pattern to a handler pattern
1866         https://bugs.webkit.org/show_bug.cgi?id=104543
1867
1868         Reviewed by Kent Tamura.
1869
1870         Adding mocks to be able to test this refactoring.
1871
1872         * DumpRenderTree/DumpRenderTree.gypi:
1873         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.cpp: Added.
1874         (DataChannelReadyStateTask):
1875         (DataChannelReadyStateTask::DataChannelReadyStateTask):
1876         (MockWebRTCDataChannelHandler::MockWebRTCDataChannelHandler):
1877         (MockWebRTCDataChannelHandler::setClient):
1878         (MockWebRTCDataChannelHandler::bufferedAmount):
1879         (MockWebRTCDataChannelHandler::sendStringData):
1880         (MockWebRTCDataChannelHandler::sendRawData):
1881         (MockWebRTCDataChannelHandler::close):
1882         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.h: Added.
1883         (MockWebRTCDataChannelHandler):
1884         (MockWebRTCDataChannelHandler::taskList):
1885         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
1886         (RTCPeerConnectionReadyStateTask::RTCPeerConnectionReadyStateTask):
1887         (RemoteDataChannelTask::RemoteDataChannelTask):
1888         (MockWebRTCPeerConnectionHandler::createDataChannel):
1889         (MockWebRTCPeerConnectionHandler::stop):
1890         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
1891         (MockWebRTCPeerConnectionHandler):
1892
1893 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
1894
1895         [chromium] WebTestPlugin must not depend on WTF::String
1896         https://bugs.webkit.org/show_bug.cgi?id=104783
1897
1898         Reviewed by Kent Tamura.
1899
1900         In a component build, we can't access the WTF that is part of the
1901         WebKit component. Accessing WTF instead will pull in a separate copy
1902         into the TestRunner library. Instead of WTF::String::number use
1903         snprintf() to print numbers.
1904
1905         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
1906
1907 2012-12-12  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
1908
1909         [Qt] Do not automatically enable force_static_libs_as_shared when using debug_and_release
1910         https://bugs.webkit.org/show_bug.cgi?id=104562
1911
1912         Reviewed by Tor Arne Vestbø.
1913
1914         force_static_libs_as_shared creates issues on Mac if QtWebKit is built as a framework.
1915         QtWebProcess, for example, will link to the framework's release library and will rely
1916         on DYLD_IMAGE_SUFFIX to switch to the debug build. Since QtWebProcess will also links
1917         directly to the debug build of WebCore, and that QtWebKit release will link in the
1918         release build of WebCore, both the debug and release version will be loaded.
1919
1920         To fix this properly we should make sure that applications always link to the release
1921         build of dependent libraries by bypassing resolveFinalLibraryName.
1922         Disable it for now as force_static_libs_as_shared is mainly for development while
1923         debug_and_release is mainly for production builds.
1924
1925         * qmake/mkspecs/features/default_post.prf:
1926
1927 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
1928
1929         Unreviewed GTK gardening of the unit tests.
1930
1931         Skip two unit tests that regressed during the hackfest.
1932
1933         * Scripts/run-gtk-tests:
1934         (TestRunner):
1935
1936 2012-12-11  Ryosuke Niwa  <rniwa@webkit.org>
1937
1938         Add Koji Ishii to the list of contributors.
1939
1940         * Scripts/webkitpy/common/config/committers.py:
1941
1942 2012-12-11  Dominic Mazzoni  <dmazzoni@google.com>
1943
1944         AX: Make isActionSupported cross-platform.
1945         https://bugs.webkit.org/show_bug.cgi?id=104681
1946
1947         Reviewed by Chris Fleizach.
1948
1949         Replace isActionSupported, which takes a platform-specific
1950         argument, with three cross-platform methods.
1951
1952         * DumpRenderTree/AccessibilityUIElement.cpp:
1953         (isPressActionSupportedCallback):
1954         (isIncrementActionSupportedCallback):
1955         (isDecrementActionSupportedCallback):
1956         (AccessibilityUIElement::getJSClass):
1957         * DumpRenderTree/AccessibilityUIElement.h:
1958         (AccessibilityUIElement):
1959         * DumpRenderTree/blackberry/AccessibilityUIElementBlackBerry.cpp:
1960         (AccessibilityUIElement::isPressActionSupported):
1961         (AccessibilityUIElement::isIncrementActionSupported):
1962         (AccessibilityUIElement::isDecrementActionSupported):
1963         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
1964         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
1965         (WebTestRunner::AccessibilityUIElement::isPressActionSupportedCallback):
1966         (WebTestRunner):
1967         (WebTestRunner::AccessibilityUIElement::isIncrementActionSupportedCallback):
1968         (WebTestRunner::AccessibilityUIElement::isDecrementActionSupportedCallback):
1969         (WebTestRunner::AccessibilityUIElement::incrementCallback):
1970         (WebTestRunner::AccessibilityUIElement::decrementCallback):
1971         (WebTestRunner::AccessibilityUIElement::pressCallback):
1972         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
1973         (AccessibilityUIElement):
1974         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
1975         (AccessibilityUIElement::isPressActionSupported):
1976         (AccessibilityUIElement::isIncrementActionSupported):
1977         (AccessibilityUIElement::isDecrementActionSupported):
1978         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
1979         (AccessibilityUIElement::isPressActionSupported):
1980         (AccessibilityUIElement::isIncrementActionSupported):
1981         (AccessibilityUIElement::isDecrementActionSupported):
1982         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
1983         (AccessibilityUIElement::isPressActionSupported):
1984         (AccessibilityUIElement::isIncrementActionSupported):
1985         (AccessibilityUIElement::isDecrementActionSupported):
1986         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
1987         (WTR::AccessibilityUIElement::isPressActionSupported):
1988         (WTR::AccessibilityUIElement::isIncrementActionSupported):
1989         (WTR::AccessibilityUIElement::isDecrementActionSupported):
1990         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
1991         (AccessibilityUIElement):
1992         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
1993         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
1994         (WTR::AccessibilityUIElement::isPressActionSupported):
1995         (WTR):
1996         (WTR::AccessibilityUIElement::isIncrementActionSupported):
1997         (WTR::AccessibilityUIElement::isDecrementActionSupported):
1998
1999 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2000
2001         nrwt: remove result_summary.total_tests_by_expectation
2002         https://bugs.webkit.org/show_bug.cgi?id=104756
2003
2004         Reviewed by Eric Seidel.
2005
2006         This field wasn't being used by anything.
2007
2008         * Scripts/webkitpy/layout_tests/models/result_summary.py:
2009         (ResultSummary.__init__):
2010
2011 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2012
2013         garden-o-matic doesn't know about reftests
2014         https://bugs.webkit.org/show_bug.cgi?id=101976
2015
2016         Reviewed by Ojan Vafai.
2017
2018         This is a first attempt at fixing this, sort of a "stop-the-bleeding"
2019         fix. We will now look at the list of tests we're asked to rebaseline,
2020         and log which ones are reftests into the status area, and rebaseline
2021         the rest.
2022
2023         Longer-term, we should display feedback in the results area based on
2024         which tests we're looking at. If the test list contains no reftests,
2025         we should behave as today; if the list is all reftest, we should
2026         probably disable the button, and if the list is a mix, we should
2027         probably display an alert next to the button or something. It would
2028         also be nice to annotate which tests were reftests, maybe with a
2029         "(reftest)" next to the test name or something.
2030
2031         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2032         (.):
2033         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2034         (.):
2035         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2036
2037 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2038
2039         REGRESSION(r133380): new tests without expected file reported as failing tests on the bots
2040         https://bugs.webkit.org/show_bug.cgi?id=101469
2041
2042         Reviewed by Ryosuke Niwa.
2043
2044         Fix parsing of NRWT's output to identify missing results separately
2045         from other regressions.
2046
2047         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2048         (RunWebKitTests._parseNewRunWebKitTestsOutput):
2049         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2050         (MasterCfgTest.test_nrwt_leaks_parsing):
2051         (test_nrwt_missing_results):
2052
2053 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2054
2055         buildbot: change run-webkit-tests --no-launch-safari to --no-show-results
2056         https://bugs.webkit.org/show_bug.cgi?id=104299
2057
2058         Reviewed by Ryosuke Niwa.
2059
2060         Since --no-show-results is the more generically-named flag that NRWT uses.
2061
2062         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2063         (RunWebKitTests):
2064
2065 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2066
2067         remove --use-remote-links-to-tests from buildbot configs (except apple win)
2068         https://bugs.webkit.org/show_bug.cgi?id=104291
2069
2070         Reviewed by Csaba Osztrogonác.
2071
2072         This flag is ignored by NRWT and was just implemented for compatibility.
2073         ORWT still uses it, though.
2074
2075         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2076         (RunWebKitTests):
2077         (RunWebKitTests.start):
2078
2079 2012-12-11  Anders Carlsson  <andersca@apple.com>
2080
2081         ASSERT running platform/mac/plugins/convert-point.html on the WK2 bots
2082         https://bugs.webkit.org/show_bug.cgi?id=104735
2083
2084         Reviewed by Jon Honeycutt.
2085
2086         Do the testing from within NPP_New instead of NPN_SetWindow. Plug-ins aren't supposed to do any real work inside NPP_SetWindow,
2087         and doing so will more easily trigger the assertion failure. Ideally we should get rid of the initial sync painting pass, but this
2088         gets the test passing again!
2089
2090         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:
2091         (ConvertPoint::NPP_New):
2092
2093 2012-12-11  Julie Parent  <jparent@chromium.org>
2094
2095         Dashboard unit tests should not make xhrs.
2096         https://bugs.webkit.org/show_bug.cgi?id=104598
2097
2098         Reviewed by Dirk Pranke.
2099
2100         The unit test were requesting 25 files that they did not use.
2101         Stub out loader.request to avoid making the xhrs.
2102         Result: order of magnitude speed improvement in test run time (from ~1600ms to 160ms on my machine).
2103         
2104         * TestResultServer/static-dashboards/run-embedded-unittests.html:
2105         * TestResultServer/static-dashboards/run-unittests.html:
2106
2107 2012-12-11  Filip Pizlo  <fpizlo@apple.com>
2108
2109         Profiler should show bytecode dumps as they would have been visible to the JITs, including the profiling data that the JITs would see
2110         https://bugs.webkit.org/show_bug.cgi?id=104647
2111
2112         Reviewed by Oliver Hunt.
2113
2114         Added a "profiling" (or "p") command to show the profiling data that the JITs saw
2115         for each JIT compilation of a code block.
2116         
2117         Also added instruction counts in the "full" display and made the "full" display the
2118         default thing you see.
2119
2120         * Scripts/display-profiler-output:
2121
2122 2012-12-11  Eric Seidel  <eric@webkit.org>
2123
2124         Unreviewed, rolling out r137371.
2125         http://trac.webkit.org/changeset/137371
2126         https://bugs.webkit.org/show_bug.cgi?id=104434
2127
2128         Various scripts are not ready for out/ as the build directory,
2129         this can't work as written.
2130
2131         * Scripts/update-webkit:
2132         * Scripts/webkitdirs.pm:
2133         (buildChromium):
2134
2135 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
2136
2137         Unreviewed GTK build functionality fix.
2138
2139         Turns out the CSP feature was not enabled, that's why all the tests were failing.
2140
2141         * Scripts/webkitperl/FeatureList.pm:
2142
2143 2012-12-11  Eric Seidel  <eric@webkit.org>
2144
2145         Make Ninja the default build system for build-webkit --chromium
2146         https://bugs.webkit.org/show_bug.cgi?id=104434
2147
2148         Reviewed by Daniel Bates.
2149
2150         As far as I can tell, most Chromium developers are now using Ninja
2151         as their default build system instead of the native Xcode/Visual Studio/Make.
2152         This change makes Ninja the default for developers as well as all Chromium webkit bots.
2153
2154         Developers can pass --no-ninja to get the platform-native buildsystem instead of Ninja.
2155         It's possible we may wish to make this a per-platform default, and have
2156         Ninja only be the default for Mac/Linux.
2157
2158         There was a thread on webkit-dev discussing this change:
2159         http://lists.webkit.org/pipermail/webkit-dev/2012-December/023113.html
2160
2161         * Scripts/update-webkit:
2162
2163 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2164
2165         nrwt: record the order and run times of all of the tests in a single file
2166         https://bugs.webkit.org/show_bug.cgi?id=104184
2167
2168         Reviewed by Ryosuke Niwa.
2169
2170         Replace the N tests_run*.txt files that we output into
2171         layout-test-results with a single packed JSON trie that contains
2172         the test ordering and run statistics. This can eventually be uploaded
2173         to the TestResultsServer to replace the times.json file we currently
2174         upload as well, and will let us replay the results of a previous
2175         run fairly deterministically.
2176
2177         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2178         (Worker.__init__):
2179         (Worker.start):
2180         (Worker._run_test):
2181         (Worker.stop):
2182         (Worker._clean_up_after_test):
2183         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2184         (Manager._upload_json_files):
2185         (Manager._copy_results_html_file):
2186         (Manager):
2187         (Manager._stats_trie):
2188         (Manager._stats_trie._worker_number):
2189         * Scripts/webkitpy/layout_tests/models/test_results.py:
2190         (TestResult.__init__):
2191         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2192         (MainTest.test_missing_and_unexpected_results):
2193         (MainTest.test_retrying_and_flaky_tests):
2194         (EndToEndTest.test_end_to_end):
2195         (EndToEndTest.test_reftest_with_two_notrefs):
2196         (RebaselineTest.test_reset_results):
2197         (RebaselineTest.test_missing_results):
2198         (RebaselineTest.test_new_baseline):
2199
2200 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2201
2202         check-webkit-style complaining about TestExpectations lines a patch didn't modify
2203         https://bugs.webkit.org/show_bug.cgi?id=104687
2204
2205         Reviewed by Ojan Vafai.
2206
2207         Way back in bug 60466 / r86165, we changed the style checker
2208         for TestExpectations to ensure that the entire file linted, not
2209         just the lines modified by a patch, in order to ensure that a
2210         new line didn't cause errors elsewhere in the file (e.g., by
2211         introducing duplicate lines). At the time, errors in the
2212         expectations files were fatal and we didn't have a separate step
2213         on the bots that would lint the files to ensure things were still
2214         okay, so this was probably the right thing to do, but it could
2215         cause some patches to fail the style check even though the patches
2216         themselves were fine.
2217
2218         However, now errors aren't always fatal, and we have a separate build
2219         step to lint the files. So, we should probably not punish the
2220         patches and the style checker should filter out "unrelated" errors.
2221
2222         In theory we should be comparing the output pre- and post-patch
2223         to filter out pre-existing errors, but I don't think the style
2224         checker has that ability.
2225
2226         * Scripts/webkitpy/style/checkers/test_expectations.py:
2227         (TestExpectationsChecker.__init__):
2228         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2229         (TestExpectationsTestCase.assert_lines_lint):
2230
2231 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2232
2233         nrwt isn't rejecting unrecognized expectations
2234         https://bugs.webkit.org/show_bug.cgi?id=104701
2235
2236         Reviewed by Ojan Vafai.
2237
2238         NRWT is silently ignoring unrecognized expectations like "Text"
2239         and "Audio" (largely because I haven't fully removed support for
2240         the old syntax and keywords and restructured the code to do the
2241         proper checking). Temporarily change the tokenizer to reject
2242         unrecognized expectations until I can make those changes.
2243
2244         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2245         (TestExpectationParser._tokenize_line):
2246         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2247         (ExpectationSyntaxTests.test_unknown_expectation):
2248         (TestExpectationSerializationTests.disabled_test_string_whitespace_stripping):
2249
2250 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2251
2252         Fix a typo in my previous commit that is causing test failures.
2253
2254         Reviewed by Gustavo Noronha Silva.
2255
2256         There was an extra bit of statement appended after the end of a conditional. Somehow
2257         this compiled and caused test failures.
2258
2259         * DumpRenderTree/gtk/EventSender.cpp:
2260         (continuousMouseScrollByCallback):
2261
2262 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2263
2264         [GTK] Spurious stderr output from the event sender
2265         https://bugs.webkit.org/show_bug.cgi?id=104678
2266
2267         Reviewed by Gustavo Noronha Silva.
2268
2269         Eliminate some spurious stderr output from the EventSender by using an early return
2270         instead of an assertion that prints a warning to the console.
2271
2272         * DumpRenderTree/gtk/EventSender.cpp:
2273         (continuousMouseScrollByCallback): We don't support continuous scroll events that
2274         move by page, but we should just fail silently instead of printing an error.
2275
2276 2012-12-11  Joone Hur  <joone.hur@intel.com>
2277
2278         [GTK] Add GraphicsLayerActor
2279         https://bugs.webkit.org/show_bug.cgi?id=91940
2280
2281         Reviewed by Gustavo Noronha Silva.
2282
2283         Add GraphicsLayerActor to the list of classes that contain GObjects
2284         in the style checker script so that it does not complain about GObject
2285         coding style.
2286
2287         * Scripts/webkitpy/style/checker.py:
2288
2289 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
2290
2291         [Qt] remove hack which avoids unused .obj directories all over the place
2292
2293         Reviewed by Simon Hausmann.
2294
2295         The feature is subject to upstreaming in Qt itself in
2296         https://codereview.qt-project.org/#change,42011
2297
2298         * qmake/mkspecs/features/default_post.prf:
2299
2300 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
2301
2302         [Qt] move SQLITE3SRCDIR assignment to only use site
2303
2304         Reviewed by Simon Hausmann.
2305
2306         Target.pri includes WebCore.pri the almost first thing, so that is
2307         covered as well.
2308
2309         * qmake/mkspecs/features/default_pre.prf:
2310
2311 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
2312
2313         [Qt] delete dead include paths
2314
2315         Reviewed by Simon Hausmann.
2316
2317         followup to https://bugs.webkit.org/show_bug.cgi?id=93446
2318
2319         * DumpRenderTree/qt/DumpRenderTree.pro:
2320
2321 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2322
2323         [GTK] Implement some missing FrameLoaderClient message in DRT
2324         https://bugs.webkit.org/show_bug.cgi?id=104444
2325
2326         Reviewed by Gustavo Noronha Silva.
2327
2328         Install a FrameLoadEvent callback that prints the necessary output for some
2329         missing FrameLoader events.
2330
2331         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2332         (frameLoadEventCallback): Added.
2333         (createWebView): Install the FrameLoader event callback.
2334
2335 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
2336
2337         [Qt] change setup of QMAKE_MOC
2338
2339         Reviewed by Simon Hausmann.
2340
2341         don't load(moc) explicitly - turns out that this breaks
2342         debug_and_release builds (the feature does not like being loaded before
2343         default_post has been loaded).
2344         so instead just do a minimal call to find the moc executable.
2345
2346         as it's in fact only two files which need it, so de-centralize this call
2347         for the sake of saving some cpu cycles spent in qmake.
2348
2349         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
2350         * qmake/mkspecs/features/default_pre.prf:
2351
2352 2012-12-11  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
2353
2354         [Qt] Fix the linking of debug builds on Mac with force_static_libs_as_shared
2355         https://bugs.webkit.org/show_bug.cgi?id=104559
2356
2357         Reviewed by Tor Arne Vestbø.
2358
2359         "contains(libraries, $$TARGET): return(true)" fails for intermediate libraries since
2360         at this point TARGET may contain the "_debug" suffix.
2361
2362         Fix it by moving the TARGET adjustment later in the file.
2363
2364         * qmake/mkspecs/features/default_post.prf:
2365
2366 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2367
2368         [Coverity] [GTK] Remove some redundant null checks in DumpRenderTree
2369         https://bugs.webkit.org/show_bug.cgi?id=104570
2370
2371         Reviewed by Daniel Bates.
2372
2373         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2374         (descriptionSuitableForTestResult): Remove a redundant null check.
2375
2376 2012-12-11  Alberto Garcia  <agarcia@igalia.com>
2377
2378         Remove unused GErrors
2379         https://bugs.webkit.org/show_bug.cgi?id=104657
2380
2381         Reviewed by Martin Robinson.
2382
2383         Replace unused GErrors with null pointers.
2384
2385         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2386         (initializeFonts):
2387         * DumpRenderTree/gtk/EditingCallbacks.cpp:
2388         (dumpRange):
2389         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
2390         (WTR::inititializeFontConfigSetting):
2391
2392 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2393
2394         [GTK] Bump the jhbuild libsoup version to fix an authentication issue
2395         https://bugs.webkit.org/show_bug.cgi?id=104448
2396
2397         Reviewed by Gustavo Noronha Silva.
2398
2399         Bump the libsoup version to one from the git repository. This requires updating
2400         GLib as well which in turn requires adjusting the CFLAGS we pass when compiling
2401         the scanobj step of the gtkdoc generation.
2402
2403         * gtk/generate-gtkdoc:
2404         * gtk/jhbuild.modules:
2405
2406 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
2407
2408         Unreviewed functionality fix for the GTK port.
2409
2410         Enable the MicroData feature support for the GTK port in the FeatureList module.
2411         This feature listing now again affects which features are enabled and which disabled
2412         on the development builds of the GTK port.
2413
2414         * Scripts/webkitperl/FeatureList.pm:
2415
2416 2012-12-11  Jinwoo Song  <jinwoo7.song@samsung.com>
2417
2418         [EFL][WK2] MiniBrowser is not built
2419         https://bugs.webkit.org/show_bug.cgi?id=104629
2420
2421         Reviewed by Gyuyoung Kim.
2422
2423         As we are finding an elementary package in Tools/MiniBrowser/efl/CMakeList.txt since r137203,
2424         we need to check if the package is found in the same place not in the Tools/CMakeList.txt.
2425
2426         * CMakeLists.txt:
2427         * MiniBrowser/efl/CMakeLists.txt:
2428
2429 2012-12-11  Xabier Rodriguez Calvar  <calvaris@igalia.com>
2430
2431         [GTK][jhbuild] Switch to GStreamer 1.0 build
2432         https://bugs.webkit.org/show_bug.cgi?id=91727
2433
2434         Reviewed by Philippe Normand.
2435
2436         Switch build-webkit --gtk to GStreamer 1.0 support and build the
2437         necessary GStreamer git modules from JHBuild.
2438
2439         * gtk/jhbuild.modules: Added GStreamer 1.0.3 build support.
2440
2441 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
2442
2443         [GTK] Feature enabling/disabling should be possible through build-webkit
2444         https://bugs.webkit.org/show_bug.cgi?id=99271
2445
2446         Reviewed by Gustavo Noronha Silva.
2447
2448         Refactor the code in webkitdirs.pm that builds an autotools project to generate
2449         autogen.sh arguments, make arguments and installation prefix from the passed-in
2450         parameters rather than generating all of that in build-webkit. The autogen.sh
2451         arguments now contain only enable/disable flags for options that are actually
2452         configurable in configure.ac, the flag value actually reflecting whether the feature
2453         is enabled or disabled in the feature list.
2454
2455         Other features are overridable through modifying the GNUmakefile.features.am file.
2456         All these features are now stored in the build directory and upon change trigger
2457         a rerun of the autogen.sh script, pretty much like the autogen.sh arguments do.
2458
2459         The override-feature-defines script is called by autogen.sh, before GNUmakefile.in is
2460         generated when calling autoreconf. Its task is to copy the GNUmakefile.features.am.in
2461         into GNUmakefile.features.am and modify it if there's a text file in the build directory
2462         that contains all the feature defines the build-webkit script has written. If the build
2463         is not done through build-webkit (and there's no file in build directory listing all the
2464         feature defines) the script does not advance further from copying.
2465
2466         This approach is taken to overcome the rigidness of the automake system as it's impossible
2467         to effectively generate and use GNUmakefile.features.am or even GNUmakefile.features file
2468         after the autoreconf command execution in autogen.sh.
2469
2470         * Scripts/build-jsc:
2471         (buildMyProject):
2472         * Scripts/build-webkit:
2473         * Scripts/webkitdirs.pm:
2474         (runAutogenForAutotoolsProjectIfNecessary):
2475         (mustReRunAutogen):
2476         (buildAutotoolsProject):
2477         (buildGtkProject):
2478         * Scripts/webkitperl/FeatureList.pm: Update the features that are currently enabled in
2479         Source/WebCore/GNUmakefile.features.am.in but aren't in the feature list.
2480         * Scripts/webkitpy/style/checker.py: Source/WebCore/GNUmakefile.features.am.in is recognized
2481         as a text file and tabulation errors are reported. Skip the file to suppress them.
2482         * gtk/override-feature-defines: Added.
2483         (copy_feature_defines_makefile):
2484         (adjust_feature_defines_makefile):
2485         (adjust_feature_defines_makefile.override_feature_define):
2486         (override_feature_defines):
2487
2488 2012-12-10  Yury Semikhatsky  <yurys@chromium.org>
2489
2490         Memory instrumentation: make sure each edge is reported only once
2491         https://bugs.webkit.org/show_bug.cgi?id=104630
2492
2493         Reviewed by Pavel Feldman.
2494
2495         Test that outgoing edges are reported only once if we come to the same
2496         object twice: first by an address of a base class and then by a real address
2497         of the object.
2498
2499         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2500
2501 2012-12-10  Dana Jansens  <danakj@chromium.org>
2502
2503         [chromium] Remove WebCompositorOutputSurface implementations from DRT
2504         https://bugs.webkit.org/show_bug.cgi?id=104046
2505
2506         Reviewed by Darin Fisher.
2507
2508         The classes are now provided through WebCompositorSupport.
2509
2510         * DumpRenderTree/DumpRenderTree.gypi:
2511         * DumpRenderTree/chromium/WebViewHost.cpp:
2512         (WebViewHost::createOutputSurface):
2513         * DumpRenderTree/chromium/WebViewHostOutputSurface.cpp: Removed.
2514         * DumpRenderTree/chromium/WebViewHostOutputSurface.h: Removed.
2515         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.cpp: Removed.
2516         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.h: Removed.
2517
2518 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2519
2520         change --no-launch-safari to --no-show-results in scripts
2521         https://bugs.webkit.org/show_bug.cgi?id=104298
2522
2523         Reviewed by Eric Seidel.
2524
2525         Since Safari only exists on the Apple ports, NRWT calls the
2526         command line flag --[no-]show-results instead. This updates
2527         the scripts to use the new flag (but not the buildbot config yet).
2528
2529         * Scripts/generate-coverage-data:
2530         * Scripts/webkit-tools-completion.sh:
2531         * Scripts/webkitpy/tool/steps/runtests.py:
2532         (RunTests.run):
2533         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
2534
2535 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2536
2537         nrwt: fix mapping of AUDIO onto FAIL when printing unexpected results
2538         https://bugs.webkit.org/show_bug.cgi?id=104597
2539
2540         Reviewed by Ojan Vafai.
2541
2542         This was broken in the recent printer refactoring.
2543
2544         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
2545         (BuildBotPrinter.print_unexpected_results.add_result.is_expected):
2546         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2547         (BuildBotPrinterTests.get_result):
2548         (BuildBotPrinterTests.test_print_unexpected_results):
2549         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
2550         (BuildBotPrinterTests):
2551
2552 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2553
2554         old-run-webkit-tests: add --show-results properly
2555         https://bugs.webkit.org/show_bug.cgi?id=104608
2556
2557         Reviewed by Eric Seidel.
2558
2559         Rework r137219 / bug 104601 to use the correct perl idiom.
2560
2561         * Scripts/old-run-webkit-tests:
2562
2563 2012-12-10  Aaron Colwell  <acolwell@chromium.org>
2564
2565         Unreviewed. Adding Aaron Colwell to the contributor list.
2566
2567         * Scripts/webkitpy/common/config/committers.py:
2568
2569 2012-12-10  Anders Carlsson  <andersca@apple.com>
2570
2571         Remove WebKit2 tests on Windows
2572         https://bugs.webkit.org/show_bug.cgi?id=104607
2573
2574         Reviewed by Tim Horton.
2575
2576         Stop building WebKit2 tests and remove associated files.
2577
2578         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Removed.
2579         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Removed.
2580         * TestWebKitAPI/win/PlatformWebViewWin.cpp: Removed.
2581         * TestWebKitAPI/win/TestWebKitAPI.sln: Removed.
2582         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
2583         * TestWebKitAPI/win/TestWebKitAPIGenerated.vcproj: Removed.
2584         * TestWebKitAPI/win/TestWebKitAPIGeneratedCommon.vsprops: Removed.
2585         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Removed.
2586         * TestWebKitAPI/win/WindowMessageObserver.h: Removed.
2587         * TestWebKitAPI/win/copy-resources.cmd: Removed.
2588
2589 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2590
2591         old-run-webkit-tests: support --no-show-results for compatibility w/ NRWT
2592         https://bugs.webkit.org/show_bug.cgi?id=104601
2593
2594         Reviewed by Eric Siedel.
2595
2596         This flag is being added so we can move all of the calling scripts
2597         over to using --no-show-results instead of --no-launch-safari.
2598
2599         * Scripts/old-run-webkit-tests:
2600
2601 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2602
2603         nrwt: use the default logic for skipping platform-specific tests for chromium
2604         https://bugs.webkit.org/show_bug.cgi?id=104403
2605
2606         Reviewed by Stephen White.
2607
2608         I thought I had fixed this a while ago, but the chromium port was
2609         using custom logic to figure out which platform/* tests to run (or
2610         skip) and as a result we were getting some odd behavior (like running
2611         tests in chromium-linux on chromium-mac, but only when there was
2612         an expectation listed for the test).
2613
2614         Ports are supposed to only run the tests found under their
2615         baseline_search_path() by default.
2616
2617         Not writing a test for this; theoretically there shouldn't been a
2618         test checking that the behavior for Chromium was *different* than
2619         the other ports before.
2620
2621         * Scripts/webkitpy/layout_tests/port/chromium.py:
2622         (ChromiumPort._missing_symbol_to_skipped_tests):
2623
2624 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2625
2626         fix regression in test-webkitpy after r137189
2627         https://bugs.webkit.org/show_bug.cgi?id=104589
2628
2629         Reviewed by Eric Seidel.
2630
2631         Need to specify full port names.
2632
2633         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2634         (MainTest.test_build_check):
2635
2636 2012-12-10  Laszlo Gombos  <l.gombos@samsung.com>
2637
2638         [EFL] Change the minimum required EFL version to 1.6
2639         https://bugs.webkit.org/show_bug.cgi?id=104431
2640
2641         Reviewed by Kenneth Rohde Christiansen.
2642
2643         Change the minimum required EFL version to 1.6 from 1.7.
2644         The elementary EFL package is only required to build MiniBrowser,
2645         so I moved the required only to the Minibrowser CMake file.
2646
2647         * MiniBrowser/efl/CMakeLists.txt: In addition removed
2648         ENABLE_GLIB_SUPPORT as it is obsolete and it was always set to be on.
2649
2650 2012-12-10  Dean Jackson  <dino@apple.com>
2651
2652         Unreviewed. Adding Antoine to the contributor list so his name autocompletes in bugzilla.
2653
2654         * Scripts/webkitpy/common/config/committers.py:
2655
2656 2012-12-10  Zan Dobersek  <zandobersek@gmail.com>
2657
2658         REGRESSION (r118735): svg/dom/complex-svgView-specification.html, svg/dom/SVGViewSpec.html, svg/dom/viewspec-parser.html failing on GTK Linux 64-bit Release
2659         https://bugs.webkit.org/show_bug.cgi?id=87734
2660
2661         Reviewed by Martin Robinson.
2662
2663         When willSendRequest should return null in testing, cancel the SoupMessage
2664         instead of setting the request's URI to about:blank.
2665
2666         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2667         (willSendRequestCallback):
2668
2669 2012-12-10  Jochen Eisinger  <jochen@chromium.org>
2670
2671         [chromium] Move the test implementation of WebKit::WebPlugin to the TestRunner library
2672         https://bugs.webkit.org/show_bug.cgi?id=104537
2673
2674         Reviewed by Tony Chang.
2675
2676         In order to hide implementation details from the public TestRunner API,
2677         I moved the plugin to WebTestPluginImpl and added a static create
2678         method to WebTestPlugin. Also, I renamed the plugin from TestWebPlugin
2679         to WebTestPlugin to be more in line with other classes in the
2680         TestRunner API.
2681
2682         * DumpRenderTree/DumpRenderTree.gypi:
2683         * DumpRenderTree/chromium/TestRunner/public/WebTestPlugin.h: Added.
2684         (WebTestRunner):
2685         (WebTestPlugin):
2686         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Renamed from Tools/DumpRenderTree/chromium/TestWebPlugin.cpp.
2687         (WebTestRunner::WebTestPlugin::create):
2688         (WebTestRunner):
2689         (WebTestRunner::WebTestPlugin::WebTestPlugin):
2690         (WebTestRunner::WebTestPlugin::~WebTestPlugin):
2691         (WebTestRunner::WebTestPlugin::mimeType):
2692         * DumpRenderTree/chromium/TestWebPlugin.h: Removed.
2693         * DumpRenderTree/chromium/WebViewHost.cpp:
2694         (WebViewHost::createPlugin):
2695
2696 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2697
2698         run_webkit_tests throws exception when DRT is not present
2699         https://bugs.webkit.org/show_bug.cgi?id=104565
2700
2701         Reviewed by Jochen Eisinger.
2702
2703         Don't actually try to print the bot results if there aren't
2704         any because we exited before running any tests.
2705
2706         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2707         (main):
2708         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2709         (MainTest.test_build_check):
2710
2711 2012-12-10  Benjamin Poulain  <benjamin@webkit.org>
2712
2713         Add convenience methods to use ListHashSet for a LRU cache
2714         https://bugs.webkit.org/show_bug.cgi?id=104499
2715
2716         Reviewed by Sam Weinig.
2717
2718         Test the new methods added to ListHashSet.
2719
2720         * TestWebKitAPI/CMakeLists.txt:
2721         * TestWebKitAPI/GNUmakefile.am:
2722         * TestWebKitAPI/TestWebKitAPI.gypi:
2723         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2724         * TestWebKitAPI/Tests/WTF/ListHashSet.cpp: Added.
2725         (TestWebKitAPI):
2726         (TestWebKitAPI::TEST):
2727
2728 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2729
2730         webkit-patch print-expectations doesn't work right for platforms w/ shared expectations
2731         https://bugs.webkit.org/show_bug.cgi?id=104405
2732
2733         Reviewed by Eric Seidel.
2734
2735         If multiple ports shared a TestExpectations file, webkit-patch
2736         print-expectations would use the expectation for the first port
2737         using that file and print the result for all ports, rather than
2738         re-parse and print each port separately (e.g., we might print
2739         the chromium-linux expectation for a file for both chromium-linux
2740         and chromium-mac). Discovered while testing/debugging the fix
2741         for bug 104403.
2742
2743         * Scripts/webkitpy/layout_tests/port/test.py:
2744         (TestPort._skipped_tests_for_unsupported_features):
2745         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2746         (lint):
2747         * Scripts/webkitpy/tool/commands/queries.py:
2748         (PrintExpectations._model):
2749         * Scripts/webkitpy/tool/commands/queries_unittest.py:
2750         (PrintExpectationsTest.test_platform):
2751
2752 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2753
2754         nrwt: fix total number of tests computed in the result summary
2755         https://bugs.webkit.org/show_bug.cgi?id=104061
2756
2757         Reviewed by Jochen Eisinger.
2758
2759         Good thing I left this open, because I actually broke it again
2760         the same way in a merge. Fix the computation of the one_line_summary
2761         again, and add tests for it (also clean up some redundant tests).
2762
2763         Also, add constants into the test port code so that the knowledge
2764         of how many tests should fail and how is next to the list of tests,
2765         rather than a totally different place.
2766
2767         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2768         (LayoutTestRunner.run_tests):
2769         * Scripts/webkitpy/layout_tests/port/test.py:
2770         (TestList.__getitem__):
2771         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2772         (get_test_results):
2773         (MainTest.setUp):
2774         (MainTest.test_basic):
2775         (MainTest.test_run_singly_actually_runs_tests):
2776         (MainTest.test_test_list_with_prefix):
2777         (MainTest.test_missing_and_unexpected_results_with_custom_exit_code.CustomExitCodePort):
2778         (MainTest.test_tolerance.ImageDiffTestPort):
2779         (EndToEndTest.parse_full_results):
2780
2781 2012-12-10  Mike West  <mkwst@chromium.org>
2782
2783         Log to console when script is blocked by sandbox attributes.
2784         https://bugs.webkit.org/show_bug.cgi?id=104365
2785
2786         Reviewed by Ojan Vafai.
2787
2788         * Scripts/webkitpy/performance_tests/perftest.py:
2789         (PerfTest):
2790             Ignoring the new console message in perftests, as it broke parsing
2791             for the Parser/html-parser.html microbenchmark.
2792
2793 2012-12-10  Filip Pizlo  <fpizlo@apple.com>
2794
2795         JSC profiling and debug dump code should use inferred names when possible
2796         https://bugs.webkit.org/show_bug.cgi?id=104519
2797
2798         Reviewed by Oliver Hunt.
2799
2800         The format I'm using for referring to a code block is now name#hash. For example,
2801         v8-crypto has something called bnpSquareTo#B5QFbU. The profiler allows you to use
2802         either the hash, the inferred name, or the combined hash and full name when referring
2803         to blocks.
2804
2805         * Scripts/display-profiler-output:
2806
2807 2012-12-09  Filip Pizlo  <fpizlo@apple.com>
2808
2809         Profiler should say things about OSR exits
2810         https://bugs.webkit.org/show_bug.cgi?id=104497
2811
2812         Reviewed by Oliver Hunt.
2813
2814         Adds support for displaying OSR exit information for full summary (just displays the
2815         counts and the number of recompilations), bytecode display (says which bytecodes
2816         exited), and DFG display (annotates disassembly with taken OSR exits and their
2817         counts).
2818
2819         * Scripts/display-profiler-output:
2820
2821 2012-12-10  Alexis Menard  <alexis@webkit.org>
2822
2823         [CSS3 Backgrounds and Borders] Remove CSS3_BACKGROUND feature flag.
2824         https://bugs.webkit.org/show_bug.cgi?id=104539
2825
2826         Reviewed by Antonio Gomes.
2827
2828         As discussed on webkit-dev it is not needed to keep this feature flag 
2829         as support for <position> type is a small feature that is already 
2830         implemented by three other UAs. It was useful while landing this 
2831         feature as partial bits were landed one after one.
2832
2833         * Scripts/webkitperl/FeatureList.pm:
2834         * qmake/mkspecs/features/features.pri:
2835
2836 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
2837
2838         Build fix.
2839
2840         Add a patchfile missing from my previous commit.
2841
2842         * gtk/patches/harfbuzz-icu-detection-fix.patch: Added.
2843
2844 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
2845
2846         REGRESSION(137127): Causes assertion failures on the debug bots
2847         https://bugs.webkit.org/show_bug.cgi?id=104549
2848
2849         Reviewed by Alejandro G. Castro.
2850
2851         After handling the new TextOutput message we should return the method
2852         handling messages received.
2853
2854         * WebKitTestRunner/TestInvocation.cpp:
2855         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Fix a missing early return.
2856
2857 2012-12-09  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
2858
2859         [Soup] utilize multipart/x-mixed-replace support recently added to libsoup
2860         https://bugs.webkit.org/show_bug.cgi?id=94515
2861
2862         Reviewed by Martin Robinson.
2863
2864         This change requires a very recent libsoup, so update our jhbuild
2865         requirements.
2866
2867         * gtk/jhbuild.modules: depend on soup version that adds the multipart/x-mixed-replace
2868         support
2869         * efl/jhbuild.modules: ditto
2870
2871 2012-12-10  Zoltan Nyul  <zoltan.nyul@intel.com>
2872
2873         Implement testRunner.setViewModeMediaFeature() in WebKitTestRunner
2874         https://bugs.webkit.org/show_bug.cgi?id=103886
2875
2876         Reviewed by Kenneth Rohde Christiansen.
2877
2878         Implement testRunner.setViewModeMediaFeature() in WebKitTestRunner.
2879         This API is required by layout tests.
2880
2881         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2882         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2883         (WTR::TestRunner::setViewModeMediaFeature):
2884         (WTR):
2885         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2886         (TestRunner):
2887
2888 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
2889
2890         [GTK] Bring Harfbuzz-ng support to Gtk
2891         https://bugs.webkit.org/show_bug.cgi?id=92098
2892
2893         Reviewed by Gustavo Noronha Silva.
2894
2895         Add HarfBuzz dependency.
2896
2897         * GNUmakefile.am: Add FreeType/HarfBuzz CFLAGS to the WebCore Internals library.
2898         * gtk/jhbuild.modules: Add HarfBuzz to the JHbuild and remove Pango, since it
2899         should no longer affect layout test results.
2900
2901 2012-12-10  Simon Hausmann  <simon.hausmann@digia.com>
2902
2903         [Qt] Use QLibraryInfo::LibraryExecutablesPath unconditionally
2904         https://bugs.webkit.org/show_bug.cgi?id=104541
2905
2906         Reviewed by Jocelyn Turcotte.
2907
2908         We now depend on a Qt 5 version that is guaranteed to have this API, so we
2909         can remove the configure checks for it.
2910
2911         * qmake/config.tests/libexecdir/libexecdir.cpp: Removed.
2912         * qmake/config.tests/libexecdir/libexecdir.pro: Removed.
2913         * qmake/mkspecs/features/features.prf:
2914
2915 2012-12-10  Víctor Manuel Jáquez Leal  <vjaquez@igalia.com>
2916
2917         REGRESSION(r135863): GtkLauncher --help no longer displays all help options
2918         https://bugs.webkit.org/show_bug.cgi?id=104382
2919
2920         Reviewed by Philippe Normand.
2921
2922         Remove gst_init() because it steals the GOptionContext.
2923         gst_init_get_option_group() already does the initialization job.
2924
2925         * GtkLauncher/main.c:
2926         (main):
2927
2928 2012-12-10  Zan Dobersek  <zandobersek@gmail.com>
2929
2930         [GTK][WTR] Implement AccessibilityUIElement::titleUIElement
2931         https://bugs.webkit.org/show_bug.cgi?id=102953
2932
2933         Reviewed by Martin Robinson.
2934
2935         Implement AccessibilityUIElement::titleUIElement in the ATK backend
2936         in the same manner it is implemented in DumpRenderTree. Both implementations
2937         are updated to unref the relation set before the final function return
2938         rather than at the end of each loop through the relations held in the said
2939         relation set.
2940
2941         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2942         (AccessibilityUIElement::titleUIElement):
2943         * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
2944         (WTR::AccessibilityUIElement::titleUIElement):
2945
2946 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
2947
2948         Fix the WebKit2 debug build
2949
2950         * WebKitTestRunner/TestInvocation.cpp:
2951         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Fix a typo of WKStringGetTypeID().
2952
2953 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
2954
2955         [WTR] Move text output accumulation to the UIProcess
2956         https://bugs.webkit.org/show_bug.cgi?id=104214
2957
2958         Reviewed by Darin Adler.
2959
2960         Instead of accumulating text output in the InjectedBundle and then sending it to the UIProcess 
2961         once a test is finished, immediately send any text output to the UIProcess. This will allow
2962         WebKitTestRunner to output text from the UIProcess as well.
2963
2964         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Remove the handling of the text output
2965         StringBuilder. Add a method to send output to the UIProcess.
2966         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
2967         (InjectedBundle):
2968         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp: Change code that appends to the StringBuilder
2969         to use the new outputText method.
2970         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2971         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Ditto.
2972         * WebKitTestRunner/InjectedBundle/atk/AccessibilityControllerAtk.cpp: Ditto.
2973         * WebKitTestRunner/TestInvocation.cpp:
2974         (WTR::TestInvocation::TestInvocation): Intialize the StringBuilder.
2975         (WTR::TestInvocation::invoke): Clear the StringBuilder when a new tests starts.
2976         (WTR::TestInvocation::dumpResults): m_textOutput is a StringBuilder now so the
2977         method of printing the text is slightly different.
2978         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Handle the new TextOutput
2979         message by appending the results to the StringBuilder.
2980         * WebKitTestRunner/TestInvocation.h:
2981         (TestInvocation): Switch the type of m_textOutput.
2982
2983 2012-12-10  Mihnea Ovidenie  <mihnea@adobe.com>
2984
2985         Unreviewed. Adding Mihai Maerean to the list of contributors, working on CSSRegions.
2986
2987         * Scripts/webkitpy/common/config/committers.py:
2988
2989 2012-12-10  Joanmarie Diggs  <jdiggs@igalia.com>
2990
2991         [GTK] accessibility/language-attribute.html is failing
2992         https://bugs.webkit.org/show_bug.cgi?id=98369
2993
2994         Reviewed by Chris Fleizach.
2995
2996         The test is failing for two reasons: Not exposing the language
2997         for accessible objects and AccessibilityUIElement::language() not
2998         being implemented.
2999
3000         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3001         (AccessibilityUIElement::language): Implemented.
3002
3003 2012-12-10  Joanmarie Diggs  <jdiggs@igalia.com>
3004
3005         [GTK] accessibility/editable-webarea-context-menu-point.html is failing
3006         https://bugs.webkit.org/show_bug.cgi?id=98364
3007
3008         Reviewed by Chris Fleizach.
3009
3010         The test is failing because the clickPoint values remain unchanged. The
3011         values remain unchanged because ATK lacks a means to keep track of this
3012         information.
3013
3014         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3015         (AccessibilityUIElement::clickPointX): Added comment that ATK doesn't have clickPoint details
3016         (AccessibilityUIElement::clickPointY): Added comment that ATK doesn't have clickPoint details
3017
3018 2012-12-09  Alejandro Piñeiro  <apinheiro@igalia.com>
3019
3020         [GTK] lack of implemention of AccessibilityUIElementGtk::isSelectable and AccessibilityUIElementGtk::isMultiSelectable
3021         https://bugs.webkit.org/show_bug.cgi?id=104481
3022
3023         Reviewed by Martin Robinson.
3024
3025         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3026         (AccessibilityUIElement::isSelectable): Implement this by checking
3027         the current state of the AtkObject.
3028         (AccessibilityUIElement::isMultiSelectable): Implement this by
3029         checking the current state of the AtkObject.
3030
3031 2012-12-09  Joanmarie Diggs  <jdiggs@igalia.com>
3032
3033         [GTK] accessibility/placeholder.html is failing
3034         https://bugs.webkit.org/show_bug.cgi?id=98373
3035
3036         Reviewed by Martin Robinson.
3037
3038         The test was failing because the placeholder text was not supported and
3039         AccessibilityUIElement::stringAttributeValue() was not implemented.
3040
3041         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3042         (coreAttributeToAtkAttribute): New convenience method to convert WebCore attribute
3043         names into AtkObject attribute names
3044         (AccessibilityUIElement::stringAttributeValue): implemented
3045
3046 2012-12-09  Dan Winship  <danw@gnome.org>
3047
3048         [GTK] Bump libxml2 requirement in jhbuild.modules
3049         https://bugs.webkit.org/show_bug.cgi?id=104477
3050
3051         Reviewed by Martin Robinson.
3052
3053         Fedora 18's Apache depends on symbols in libxml2 2.9.0, so the
3054         tests won't run if we only have libxml2 2.8.0 installed in the
3055         jhbuild root.
3056
3057         * gtk/jhbuild.modules:
3058
3059 2012-12-09  Zan Dobersek  <zandobersek@gmail.com>
3060
3061         Remove testRunner.disableImageLoading()
3062         https://bugs.webkit.org/show_bug.cgi?id=98083
3063
3064         Reviewed by Gustavo Noronha Silva.
3065
3066         Remove the disableImageLoading methods and callback.
3067         No tests use the method anymore and all ports except Wx
3068         now support this behavior through overriding the
3069         WebKitDisplayImagesKey preference.
3070
3071         * DumpRenderTree/TestRunner.cpp:
3072         (TestRunner::staticFunctions):
3073         * DumpRenderTree/TestRunner.h:
3074         (TestRunner):
3075         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
3076         (TestRunner::overridePreference): Add support for overriding the
3077         WebKitDisplayImagesKey preference which is now the preferred way
3078         of disabling image loading.
3079         * DumpRenderTree/efl/TestRunnerEfl.cpp:
3080         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
3081         * DumpRenderTree/mac/TestRunnerMac.mm:
3082         * DumpRenderTree/qt/TestRunnerQt.cpp:
3083         * DumpRenderTree/qt/TestRunnerQt.h:
3084         (TestRunner):
3085         * DumpRenderTree/win/TestRunnerWin.cpp:
3086         * DumpRenderTree/wx/TestRunnerWx.cpp:
3087
3088 2012-12-09  Joanmarie Diggs  <jdiggs@igalia.com>
3089
3090         [GTK] accessibility/label-for-control-hittest.html is failing
3091         https://bugs.webkit.org/show_bug.cgi?id=98368
3092
3093         Reviewed by Martin Robinson.
3094
3095         The test was failing because AccessibilityUIElement::elementAtPoint was
3096         not implemented.
3097
3098         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3099         (AccessibilityUIElement::elementAtPoint): implemented
3100
3101 2012-12-09  Jon Lee  <jonlee@apple.com>
3102
3103         [WK2] Move button image to injected bundle
3104         https://bugs.webkit.org/show_bug.cgi?id=104107
3105         <rdar://problem/12813842>
3106
3107         Reviewed by Simon Fraser.
3108
3109         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3110         (WTR::InjectedBundlePage::InjectedBundlePage): Update the bundle page UI client.
3111
3112 2012-12-08  Halton Huo  <halton.huo@intel.com>
3113
3114         [CMake] Add CMake style checker
3115         https://bugs.webkit.org/show_bug.cgi?id=104240
3116
3117         Reviewed by Eric Seidel.
3118
3119         The CMake style checker is followed by existing style checkers like cpp, python.
3120         Unit test is added as well.
3121
3122         * Scripts/webkitpy/style/checker.py: Use CMakeChecker CMAKE type files(.cmake or CMakeLists.txt)
3123         (FileType):
3124         (CheckerDispatcher._file_type):
3125         (CheckerDispatcher._create_checker):
3126         * Scripts/webkitpy/style/checkers/cmake.py: Added.
3127         (CMakeChecker):
3128         (CMakeChecker.__init__): 
3129         (CMakeChecker.check):
3130         (CMakeChecker._process_line):
3131         (CMakeChecker._check_trailing_whitespace):
3132         (CMakeChecker._check_no_space_cmds):
3133         (CMakeChecker._check_one_space_cmds):
3134         (CMakeChecker._check_non_lowercase_cmd):
3135         (CMakeChecker._check_indent):
3136         * Scripts/webkitpy/style/checkers/cmake_unittest.py: Added.
3137         (CMakeCheckerTest):
3138         (CMakeCheckerTest.test_init):
3139         (CMakeCheckerTest.test_init._mock_handle_style_error):
3140         (CMakeCheckerTest.test_check):
3141         (CMakeCheckerTest.test_check._mock_handle_style_error):
3142
3143 2012-12-08  Kondapally Kalyan  <kalyan.kondapally@intel.com>
3144
3145         [EFL] Enable WebGL by default.
3146         https://bugs.webkit.org/show_bug.cgi?id=102991.
3147
3148         Reviewed by Kenneth Rohde Christiansen.
3149
3150         Enable WebGL by default.
3151
3152         * Scripts/webkitperl/FeatureList.pm:
3153
3154 2012-12-07  Scott Violet  <sky@chromium.org>
3155
3156         [chromium] Remove linux theme related files and switch to default
3157         https://bugs.webkit.org/show_bug.cgi?id=103897
3158
3159         Reviewed by Dimitri Glazkov.
3160
3161         * Scripts/webkitpy/common/config/build_unittest.py:
3162         (ShouldBuildTest): linux->default
3163
3164 2012-12-07  Mike Lattanzio  <mlattanzio@rim.com>
3165
3166         [BlackBerry] Update BlackBerry Port to use new Graphics::createBuffer BufferType enum values.
3167         https://bugs.webkit.org/show_bug.cgi?id=104121
3168
3169         Reviewed by Rob Buis.
3170
3171         Update createBuffer call to reflect new BufferType enum.
3172         Internal PR 254649
3173         Internal Review: Jakob Petsovits
3174
3175         DRT is always backed for maximum compatibility.
3176
3177         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
3178         (createBitmapContextFromWebView):
3179
3180 2012-12-07  Jochen Eisinger  <jochen@chromium.org>
3181
3182         Add 'CHROMIUM_DEVEL_SANDBOX' to the environment variables to copy for layout test drivers
3183         https://bugs.webkit.org/show_bug.cgi?id=104355
3184
3185         Reviewed by Dirk Pranke.
3186
3187         This enables chromium's content_shell driver to run with the full SUID
3188         sandbox.
3189
3190         * Scripts/webkitpy/layout_tests/port/base.py:
3191         (Port.to.setup_environ_for_server):
3192
3193 2012-12-07  KyungTae Kim  <ktf.kim@samsung.com>
3194
3195         [EFL] Fix tools for WebKit2-only builds
3196         https://bugs.webkit.org/show_bug.cgi?id=103904
3197
3198         Reviewed by Laszlo Gombos.
3199
3200         The libewebkit2.so should be used for library path for EFL WebKit2.
3201
3202         * Scripts/webkitdirs.pm:
3203         (builtDylibPathForName):
3204
3205 2012-12-07  Xiaobo Wang  <xbwang@torchmobile.com.cn>
3206
3207         [BlackBerry] DRT - Update implementation of TestRunner.setCustomPolicyDelegate()
3208         https://bugs.webkit.org/show_bug.cgi?id=104243
3209
3210         RIM PR 259337
3211
3212         Reviewed by Rob Buis.
3213
3214         1. Implement calculation of navigation originating in
3215         didDecidePolicyForNavigationAction().
3216         2. Implement policy delegate for resource response.
3217         3. Enable customPolicyDelegate if waitForPolicyDelegate() is called.
3218         4. Minor optional updates in dump() and dumpFrameLoadCallbacks().
3219
3220         The patch fixed the following tests:
3221         - http/tests/download
3222         - fast/forms/mailto/formenctype-attribute-button-html.html
3223         - fast/forms/mailto/formenctype-attribute-input-html.html
3224         - fast/encoding/mailto-always-utf-8.html
3225
3226         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
3227         (BlackBerry::WebKit::DumpRenderTree::DumpRenderTree):
3228         (BlackBerry::WebKit::DumpRenderTree::resetToConsistentStateBeforeTesting):
3229         (BlackBerry::WebKit::DumpRenderTree::dump):
3230         (BlackBerry::WebKit::DumpRenderTree::didFinishLoadForFrame):
3231         (BlackBerry::WebKit::DumpRenderTree::didDecidePolicyForNavigationAction):
3232         (BlackBerry::WebKit::DumpRenderTree::didDecidePolicyForResponse):
3233         (BlackBerry::WebKit::DumpRenderTree::setCustomPolicyDelegate):
3234         * DumpRenderTree/blackberry/DumpRenderTreeBlackBerry.h:
3235         (DumpRenderTree):
3236         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
3237         (TestRunner::setCustomPolicyDelegate):
3238         (TestRunner::waitForPolicyDelegate):
3239         (TestRunner::overridePreference):
3240
3241 2012-12-07  Gwang Yoon Hwang  <ryumiel@company100.net>
3242
3243         [Qt] Adjust library name in layout_tests/port/qt.py after Qt5-r40
3244         https://bugs.webkit.org/show_bug.cgi?id=104348
3245
3246         Reviewed by Csaba Osztrogonác.
3247
3248         After Qt-5.0.0-r40, the library was renamed from QtWebKitWidgets to
3249         Qt5WebKitWidgets. run-webkit-tests were still looking for QtWebKitWidgets
3250
3251         * Scripts/webkitpy/layout_tests/port/qt.py:
3252         (QtPort._path_to_webcore_library):
3253
3254 2012-12-07  Csaba Osztrogonác  <ossy@webkit.org>
3255
3256         Unreviewed fix, unbreak run-launcher after library renaming.
3257
3258         * Scripts/webkitdirs.pm:
3259         (builtDylibPathForName):
3260
3261 2012-12-06  Csaba Osztrogonác  <ossy@webkit.org>
3262
3263         Unreviewed fix after r136896, unbreak --debug-rwt to make bots result verbose again.
3264         https://bugs.webkit.org/show_bug.cgi?id=104296
3265
3266         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3267         (main):
3268
3269 2012-12-06  Rick Byers  <rbyers@chromium.org>
3270
3271         CSS cursor property should support webkit-image-set
3272         https://bugs.webkit.org/show_bug.cgi?id=99493
3273
3274         Reviewed by Beth Dakin.
3275
3276         Add ENABLE_MOUSE_CURSOR_SCALE - disabled by default
3277
3278         * Scripts/webkitperl/FeatureList.pm:
3279         * qmake/mkspecs/features/features.pri:
3280
3281 2012-12-06  Vivek Galatage  <vivek.vg@samsung.com>
3282
3283         Unreviewed. Fixing the email address order to work correctly with bugzilla.
3284
3285         * Scripts/webkitpy/common/config/committers.py:
3286
3287 2012-12-06  Jochen Eisinger  <jochen@chromium.org>
3288
3289         [chromium] when forwarding events through the WebPluginContainer, create a UserGestureIndicator if processing a user gesture
3290         https://bugs.webkit.org/show_bug.cgi?id=104306
3291
3292         Reviewed by Dimitri Glazkov.
3293
3294         * DumpRenderTree/chromium/TestWebPlugin.cpp:
3295         (TestWebPlugin::TestWebPlugin):
3296         (TestWebPlugin::handleInputEvent): add support for printing the user gesture status
3297         * DumpRenderTree/chromium/TestWebPlugin.h:
3298         (TestWebPlugin):
3299
3300 2012-12-06  Dirk Pranke  <dpranke@chromium.org>
3301
3302         new-run-webkit-tests --lint-test-files seems to be broken
3303         https://bugs.webkit.org/show_bug.cgi?id=104296
3304
3305         Unreviewed, build fix.
3306
3307         My recent refactoring caused me to not initialize the printer
3308         before calling lint(); this patch restructures the code to make
3309         the lint routine self-contained and do all the work necessary.
3310         This will also make it easier to move this code out into a standalone
3311         file.
3312
3313         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3314         (lint):
3315         (main):
3316         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3317         (LintTest.test_all_configurations):
3318         (LintTest.test_lint_test_files):
3319         (LintTest.test_lint_test_files__errors):
3320
3321 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
3322
3323         DFG profiler should be helpful about gem installation
3324         https://bugs.webkit.org/show_bug.cgi?id=104217
3325
3326         Reviewed by Oliver Hunt.
3327         
3328         'json' and 'highline' are not installed by default on the ruby distributions with
3329         which I am familiar, and the default error messages don't make me happy. This makes
3330         display-profiler-output print a helpful message if those gems are not found.
3331
3332         * Scripts/display-profiler-output:
3333
3334 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
3335
3336         Profiler should print a helpful message if you pass the wrong arguments
3337         https://bugs.webkit.org/show_bug.cgi?id=104222
3338
3339         Reviewed by Oliver Hunt.
3340
3341         * Scripts/display-profiler-output:
3342
3343 2012-12-06  Ojan Vafai  <ojan@chromium.org>
3344
3345         Allow for CCing a secondary email address from watchlists
3346         https://bugs.webkit.org/show_bug.cgi?id=104286
3347
3348         Reviewed by Dirk Pranke.
3349
3350         Change my watchlist CC address so I can filter these separately from
3351         when people explicitly CC me and update the watchlist parser to allow that.
3352
3353         * Scripts/webkitpy/common/config/committers.py:
3354         * Scripts/webkitpy/common/config/watchlist:
3355         * Scripts/webkitpy/common/watchlist/watchlistparser.py:
3356         (WatchListParser._validate):
3357         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py:
3358         (WatchListParserTest.test_cc_rule_with_invalid_email):
3359         (WatchListParserTest.test_cc_rule_with_secondary_email):
3360
3361 2012-12-06  Max Vujovic  <mvujovic@adobe.com>
3362
3363         Unreviewed. Add myself to committers.py.
3364
3365         * Scripts/webkitpy/common/config/committers.py:
3366
3367 2012-12-06  Yael Aharon  <yael.aharon@intel.com>
3368
3369         [EFL] [WK2] fast/dom/vertical-scrollbar-in-rtl.html makes fast/regions tests flaky
3370         https://bugs.webkit.org/show_bug.cgi?id=104139
3371
3372         Reviewed by Kenneth Rohde Christiansen.
3373
3374         Resize the window back to 800x600 between tests.
3375         This should be a no-op if the test did not resize the window.
3376
3377         * WebKitTestRunner/TestController.cpp:
3378         (WTR::TestController::resetStateToConsistentValues):
3379
3380 2012-12-06  Tommy Widenflycht  <tommyw@google.com>
3381
3382         Speech Recognition API: Change the error code to a string on SpeechRecognitionError
3383         https://bugs.webkit.org/show_bug.cgi?id=104254
3384
3385         Reviewed by Adam Barth.
3386
3387         Changes the mock error function to take a string instead of an int.
3388
3389         * DumpRenderTree/chromium/DRTTestRunner.cpp:
3390         (DRTTestRunner::setMockSpeechRecognitionError):
3391         * DumpRenderTree/chromium/MockWebSpeechRecognizer.cpp:
3392         (WebTestRunner::ErrorTask::ErrorTask):
3393         (MockWebSpeechRecognizer::setError):
3394         * DumpRenderTree/chromium/MockWebSpeechRecognizer.h:
3395         (MockWebSpeechRecognizer):
3396
3397 2012-12-06  Andras Becsi  <andras.becsi@digia.com>
3398
3399         [Qt][Mac] Fix libxslt and libxml2 config tests
3400         https://bugs.webkit.org/show_bug.cgi?id=104164
3401
3402         Reviewed by Simon Hausmann.
3403
3404         We should not use pkg-config on Mac instead use direct include
3405         paths and add needed libraries to the linker options to detect
3406         libxslt and libxml2 provided by the system.
3407         Previously we would always fall back to qtxmlpatterns.
3408
3409         * qmake/config.tests/libxml2/libxml2.pro:
3410         * qmake/config.tests/libxslt/libxslt.pro:
3411
3412 2012-12-06  Csaba Osztrogonác  <ossy@webkit.org>
3413
3414         [Qt] Fix the build if the path contains "+" character
3415         https://bugs.webkit.org/show_bug.cgi?id=104120
3416
3417         Reviewed by Tor Arne Vestbø.
3418
3419         The second parameter of qmake's replace() is regular expression, so
3420         we have to escape special characters in paths passed to replace().
3421
3422         * qmake/mkspecs/features/default_post.prf:
3423         * qmake/mkspecs/features/default_pre.prf:
3424         * qmake/mkspecs/features/functions.prf:
3425
3426 2012-12-06  Sheriff Bot  <webkit.review.bot@gmail.com>
3427
3428         Unreviewed, rolling out r136825.
3429         http://trac.webkit.org/changeset/136825
3430         https://bugs.webkit.org/show_bug.cgi?id=104251
3431
3432         It made 100+ test fail on Qt-WK2 pixel bot, and 1800+ test
3433         flakey on Qt-WK2 non-pixel bot (Requested by Ossy on #webkit).
3434
3435         * WebKitTestRunner/TestController.cpp:
3436         (WTR::TestController::resetStateToConsistentValues):
3437
3438 2012-12-06  Yael Aharon  <yael.aharon@intel.com>
3439
3440         [EFL] [WK2] fast/dom/vertical-scrollbar-in-rtl.html makes fast/regions tests flaky
3441         https://bugs.webkit.org/show_bug.cgi?id=104139
3442
3443         Reviewed by Kenneth Rohde Christiansen.
3444
3445         Resize the window back to 800x600 between tests.
3446         This should be a no-op if the test did not resize the window.
3447
3448         * WebKitTestRunner/TestController.cpp:
3449         (WTR::TestController::resetStateToConsistentValues):
3450
3451 2012-12-06  Sheriff Bot  <webkit.review.bot@gmail.com>
3452
3453         Unreviewed, rolling out r136784 and r136802.
3454         http://trac.webkit.org/changeset/136784
3455         http://trac.webkit.org/changeset/136802
3456         https://bugs.webkit.org/show_bug.cgi?id=104231
3457
3458         breaks chromium canary (Requested by morrita on #webkit).
3459
3460         * Scripts/webkitpy/common/config/build_unittest.py:
3461         (ShouldBuildTest):
3462
3463 2012-12-05  Ryosuke Niwa  <rniwa@webkit.org>
3464
3465         Add myself to watch changes in PerformanceTests/*.
3466
3467         * Scripts/webkitpy/common/config/watchlist:
3468
3469 2012-12-05  Shinya Kawanaka  <shinyak@chromium.org>
3470
3471         [Chromium] DRT on Windows crashes when a testcase has a progress element in indeterminate state
3472         https://bugs.webkit.org/show_bug.cgi?id=102459
3473
3474         Reviewed by Dimitri Glazkov.
3475
3476         Since WebThemeControlDRTWin::markState does not consider indeterminate state, DRT caused a crash.
3477
3478         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
3479         (WebThemeControlDRTWin::markState):
3480
3481 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
3482
3483         DFG profiler should be able to display where a code block was inlined into, and what it inlined
3484         https://bugs.webkit.org/show_bug.cgi?id=104216
3485
3486         Reviewed by Oliver Hunt.
3487
3488         Added an "inlines" command that shows the inline stacks that a code block participated
3489         in. Also cleaned up some of the ways we print compilation units.
3490
3491         * Scripts/display-profiler-output:
3492
3493 2012-12-05  Halton Huo  <halton.huo@intel.com>
3494
3495         [CMake] Unify coding style for CMake files
3496         https://bugs.webkit.org/show_bug.cgi?id=103605
3497
3498         Reviewed by Laszlo Gombos.
3499
3500         Update cmake files(.cmake, CMakeLists.txt) with following style rules:
3501         1. Indentation
3502         1.1 Use spaces, not tabs.
3503         1.2 Four spaces as indent.
3504         2. Spacing
3505         2.1 Place one space between control statements and their parentheses.
3506             For eg, if (), else (), elseif (), endif (), foreach (),
3507             endforeach (), while (), endwhile (), break ().
3508         2.2 Do not place spaces between function and macro statements and
3509             their parentheses. For eg, macro(), endmacro(), function(),
3510             endfunction().
3511         2.3 Do not place spaces between a command or function or macro and its
3512             parentheses, or between a parenthesis and its content. For eg,
3513             message("testing") not message( "testing") or message ("testing" )
3514         2.4 No space at line ending.
3515         3. Lowercase when call commands macros and functions. For eg,
3516            add_executable() not ADD_EXECUTABLE(), set() not SET().
3517
3518         * CMakeLists.txt:
3519         * DumpRenderTree/TestNetscapePlugIn/CMakeLists.txt:
3520         * DumpRenderTree/efl/CMakeLists.txt:
3521         * EWebLauncher/CMakeLists.txt:
3522         * EWebLauncher/ControlTheme/CMakeLists.txt:
3523         * MiniBrowser/efl/CMakeLists.txt:
3524         * TestWebKitAPI/CMakeLists.txt:
3525         * TestWebKitAPI/PlatformEfl.cmake:
3526         * WebKitTestRunner/CMakeLists.txt:
3527         * WebKitTestRunner/PlatformEfl.cmake:
3528         * WinCELauncher/CMakeLists.txt:
3529         * clang/ReportMemoryUsagePlugin/CMakeLists.txt:
3530
3531 2012-12-05  Scott Violet  <sky@chromium.org>
3532
3533         [chromium] Remove linux theme related files and switch to default
3534         https://bugs.webkit.org/show_bug.cgi?id=103897
3535
3536         Reviewed by Tony Chang.
3537
3538         * Scripts/webkitpy/common/config/build_unittest.py:
3539         (ShouldBuildTest): linux->default
3540
3541 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3542
3543         nrwt: run_webkit_tests_integrationtest.py depends on being able to share a port w/ between master and worker
3544         https://bugs.webkit.org/show_bug.cgi?id=78168
3545
3546         Reviewed by Ojan Vafai.
3547
3548         This patch adds a single ordered list to ResultSummary that contains
3549         the results of every test run in order of result received, and
3550         modifies the TestResults class to contain the list of references used
3551         for each test as well as the pid of the driver; this allows us to stop
3552         having to override the test drivers to record the tests, and makes the
3553         integration tests that test rdering, retries, restarts, etc. black-box
3554         tests.
3555
3556         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
3557         (SingleTestRunner._run_rebaseline):
3558         (SingleTestRunner._compare_output):
3559         (SingleTestRunner._run_reftest):
3560         (SingleTestRunner._compare_output_with_reference):
3561         * Scripts/webkitpy/layout_tests/models/result_summary.py:
3562         (ResultSummary.__init__):
3563         (ResultSummary.add):
3564         * Scripts/webkitpy/layout_tests/models/test_results.py:
3565         (TestResult.__init__):
3566         * Scripts/webkitpy/layout_tests/port/driver.py:
3567         (DriverOutput.__init__):
3568         (Driver.run_test):
3569         * Scripts/webkitpy/layout_tests/port/test.py:
3570         (TestDriver):
3571         (TestDriver.__init__):
3572         (TestDriver.run_test):
3573         (TestDriver.stop):
3574         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3575         (get_tests_run):
3576         (get_test_batches):
3577         (get_test_results):
3578         (MainTest.test_batch_size):
3579         (MainTest.test_dryrun):
3580         (MainTest.test_natural_order):
3581         (MainTest.test_natural_order_test_specified_multiple_times):
3582         (MainTest.test_random_order):
3583         (MainTest.test_random_order_test_specified_multiple_times):
3584         (MainTest.test_no_order):
3585         (MainTest.test_no_order_test_specified_multiple_times):
3586         (MainTest.test_no_order_with_directory_entries_in_natural_order):
3587         (MainTest.test_repeat_each):
3588         (MainTest.test_ignore_flag):
3589         (MainTest.test_skipped_flag):
3590         (MainTest.test_iterations):
3591         (MainTest.test_run_chunk):
3592         (MainTest.test_run_part):
3593         (MainTest.test_run_singly):
3594         (MainTest.test_skip_failing_tests):
3595         (MainTest.test_single_file):
3596         (MainTest.test_single_file_with_prefix):
3597         (MainTest.test_single_skipped_file):
3598         (MainTest.test_test_list):
3599         (MainTest.test_test_list_with_prefix):
3600         (MainTest.test_exit_after_n_failures):
3601         (MainTest.test_exit_after_n_crashes):
3602         (MainTest.test_run_order__inline):
3603         (MainTest.test_reftest_run):
3604         (MainTest.test_reftest_run_reftests_if_pixel_tests_are_disabled):
3605         (MainTest.test_reftest_skip_reftests_if_no_ref_tests):
3606         (MainTest.test_reftest_expected_html_should_be_ignored):
3607         (MainTest.test_reftest_driver_should_run_expected_html):
3608         (MainTest.test_reftest_driver_should_run_expected_mismatch_html):
3609         (MainTest.test_no_http_tests):
3610         (MainTest.test_platform_tests_are_found):
3611
3612 2012-12-05  Dana Jansens  <danakj@chromium.org>
3613
3614         [chromium] Make WebCompositorOutputSurface an empty class when USE_CC_OUTPUT_SURFACE is defined.
3615         https://bugs.webkit.org/show_bug.cgi?id=103967
3616
3617         Reviewed by James Robinson.
3618
3619         Use the WebCompositorSupport methods to create an OutputSurface. If they
3620         fail, then fall back to the WebViewHost methods.
3621
3622         Once the chromium side lands, the WebCompositorSupport methods will stop
3623         failing, and will replace the WebViewHost ones.
3624
3625         * DumpRenderTree/chromium/WebViewHost.cpp:
3626         (WebViewHost::createOutputSurface):
3627
3628 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3629
3630         nrwt: split all of the buildbot-specific output into its own file
3631         https://bugs.webkit.org/show_bug.cgi?id=104165
3632
3633         Reviewed by Ojan Vafai.
3634
3635         Now that we have the RunDetails object we can move all of the
3636         buildbot-specific knowledge out of the core test-running code
3637         and into the outermost run_webkit_tests layer.
3638
3639         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3640         (run):
3641         (main):
3642         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3643         (passing_run):
3644         (logging_run):
3645         (run_and_capture):
3646         (MainTest.test_all):
3647         (MainTest.test_max_locked_shards):
3648         (MainTest.test_child_processes_2):
3649         (MainTest.test_child_processes_min):
3650         (MainTest.test_full_results_html):
3651         (MainTest.test_hung_thread):