bb3533bfb72b8add1d3aed7dedc520817124601d
[WebKit-https.git] / Tools / ChangeLog
1 2012-12-19  Zeno Albisser  <zeno@webkit.org>
2
3         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
4         https://bugs.webkit.org/show_bug.cgi?id=105309
5
6         The version number is only used as a part
7         of the library name on Linux.
8
9         Reviewed by Kenneth Rohde Christiansen.
10
11         * Scripts/webkitdirs.pm:
12         (builtDylibPathForName):
13
14 2012-12-19  Alberto Garcia  <agarcia@igalia.com>
15
16         [GTK] Tests that time out are not considered as failures.
17         https://bugs.webkit.org/show_bug.cgi?id=105318
18
19         Reviewed by Martin Robinson.
20
21         Fix parsing of the GTK API tests log files.
22
23         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
24         (RunGtkAPITests.commandComplete):
25
26 2012-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
27
28         Unreviewed, rolling out r138061.
29         https://bugs.webkit.org/show_bug.cgi?id=105396
30
31         Original patch:
32         http://trac.webkit.org/changeset/138061
33         https://bugs.webkit.org/show_bug.cgi?id=97359
34
35         It is crashing on Debug bots
36
37         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
38         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
39         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
40         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
41         (AccessibilityUIElement):
42
43 2012-12-18  Alice Liu  <alice.liu@apple.com>
44
45         Test for https://bugs.webkit.org/show_bug.cgi?id=105106
46         Add SPI to WebKit1 WebFrame for hit testing
47
48         Reviewed by Dan Bernstein.
49
50         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/ElementAtPointInWebFrame.mm
51         * TestWebKitAPI/Tests/mac/ElementAtPointInWebFrame.mm: Added.
52         (TestWebKitAPI::TEST): Loads html with two divs positioned in the 2nd and 4th quadrants of the webview.
53         Then hit-tests at three points, expecting to hit the two divs and body element. 
54
55 2012-12-18  Alice Liu  <alice.liu@apple.com>
56
57         Test for https://bugs.webkit.org/show_bug.cgi?id=105262
58         Add SPI to WebKit1 WebFrame for node conversion to JSValueRef
59
60         Reviewed by Anders Carlsson.
61
62         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/JSWrapperForNodeInWebFrame.mm.mm
63         * TestWebKitAPI/Tests/mac/JSWrapperForNodeInWebFrame.mm.mm: Added.
64         (TestWebKitAPI::TEST): Tests for the correct JS wrapper for a DOMNode, provided a WebScriptWorld.
65         Loads html with a single div element. In an isolated world, creates a property on that node.
66         Also, in the standard world, creates a different property on that node.  Then tests for 4 things:
67             - Existence of the isolated property in the isolated world.
68             - Existence of the standard property in the standard world.
69             - Non-existence of the isolated property in the standard world.
70             - Non-existence of the standard property in the isolated world.
71
72 2012-12-18  Simon Fraser  <simon.fraser@apple.com>
73
74         Disable the SimplifyMarkup API test, wich asserts in debug builds.
75         Tracked by https://bugs.webkit.org/show_bug.cgi?id=105370.
76         
77         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
78         (TestWebKitAPI::TEST):
79
80 2012-12-18  Joseph Pecoraro  <pecoraro@apple.com>
81
82         [Mac] Add Build Phase to Check Headers for Inappropriate Macros (Platform.h macros)
83         https://bugs.webkit.org/show_bug.cgi?id=104279
84
85         Reviewed by David Kilzer.
86
87         Add a build phase script that checks a framework's header files for
88         Platform.h and Compiler.h macros. Also add some tests for the script.
89
90         * Scripts/check-for-inappropriate-macros-in-external-headers: Added.
91         Script to be used in build phases to check for inappropriate macros in headers.
92
93         * Scripts/test-webkitruby: Added.
94         Script that can be used to run all webkitruby tests.
95
96         * Scripts/test-webkit-scripts:
97         Also run the ruby tests.
98
99         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/fake-data-failing-expected.txt: Added.
100         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/pass-expected.txt: Added.
101         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Fail.h: Added.
102         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Pass.h: Added.
103         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/run-test.rb: Added.
104         Tests for the new script.
105
106 2012-12-18  Dominic Mazzoni  <dmazzoni@google.com>
107
108         AX: support clickPoint in DRT for chromium
109         https://bugs.webkit.org/show_bug.cgi?id=97359
110
111         Reviewed by Chris Fleizach.
112
113         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
114         several more tests can pass.
115
116         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
117         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
118         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
119         (WebTestRunner):
120         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
121         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
122         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
123         (AccessibilityUIElement):
124
125 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
126
127         Build fix after r138045.
128
129         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
130         (PerfTestsRunner._collect_tests): Use relative_path. We skip only when this path
131         is not explicitly specified. The condition was backwards but never matched so we
132         always ran the test regardless of whether it's listed in the skipped list or not.
133         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
134         (test_collect_tests_with_ignored_skipped_list): Renamed so that it actually runs.
135
136 2012-12-18  Emil A Eklund  <eae@chromium.org>
137
138         TestWebKitAPI/Tests/WebCore/LayoutUnit fails on mac
139         https://bugs.webkit.org/show_bug.cgi?id=105332
140         
141         Reviewed by Levi Weintraub.
142         
143         Create new test suite for LayoutUnit tests as the KURL tests are
144         of the incompatible TEST_F type.
145
146         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
147         (TestWebKitAPI):
148         (TestWebKitAPI::TEST_F):
149
150 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
151
152         Running a skipped test with run-perf-tests could alert the user
153         https://bugs.webkit.org/show_bug.cgi?id=105219
154
155         Reviewed by Eric Seidel.
156
157         Force run skipped performance tests explicitly specified in the command line.
158         Also warn the user when the specified paths don't exist.
159
160         * Scripts/webkitpy/layout_tests/port/base.py:
161         (Port.relative_test_filename):
162         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
163         (PerfTestsRunner._parse_args):
164         (PerfTestsRunner._collect_tests):
165         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
166         (test_collect_tests_with_skipped_list_and_files):
167
168 2012-12-18  Yaron Friedman  <yfriedman@chromium.org>
169
170         Chrome Android java output is too verbose
171         https://bugs.webkit.org/show_bug.cgi?id=105247
172
173         Reviewed by Tony Chang.
174
175         Pass -quiet to ant to suppress extraneous output.
176
177         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
178         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
179
180 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
181
182         Consider removing --pause-before-testing option
183         https://bugs.webkit.org/show_bug.cgi?id=105248
184
185         Reviewed by Eric Seidel.
186
187         Remove --pause-before-testing option.
188
189         * Tools/Scripts/webkitpy/layout_tests/port/driver.py:
190         (DriverProxy.start): Removed.
191         * Scripts/webkitpy/performance_tests/perftest.py:
192         (PerfTest.__init__):
193         (PerfTest.test_path): Renamed from path_or_url.
194         (PerfTest.run):
195         (PerfTest): Removed "driver" argument.
196         (PerfTest._run_with_driver):
197         (PerfTest.run_single):
198         (ChromiumStylePerfTest.__init__):
199         (PageLoadingPerfTest.__init__):
200         (PageLoadingPerfTest.run_single):
201         (PageLoadingPerfTest._run_with_driver): Renamed from run so that PerfTest.run can create a driver for us.
202         (ReplayPerfTest.__init__):
203         (ReplayPerfTest.prepare):
204         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
205         (TestPageLoadingPerfTest.test_run):
206         (TestPageLoadingPerfTest.test_run_with_memory_output):
207         (TestPageLoadingPerfTest.test_run_with_bad_output):
208         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
209         (PerfTestsRunner._parse_args):
210         (PerfTestsRunner._run_tests_set): Removed the code to create a driver and pause before running the test.
211         (PerfTestsRunner._run_single_test): Removed "driver" argument.
212         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
213         (run_test):
214         (test_run_test_pause_before_testing): Removed.
215
216 2012-12-18  Alberto Garcia  <agarcia@igalia.com>
217
218         [GTK] Tests that time out are not considered as failures.
219         https://bugs.webkit.org/show_bug.cgi?id=105318
220
221         Reviewed by Martin Robinson.
222
223         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
224         (RunGtkAPITests.commandComplete):
225         Consider log lines that end with TIMEOUT as errors.
226
227         * Scripts/run-gtk-tests:
228         (TestRunner.run_tests):
229         Consider timeouts as failures when returning the number of failed
230         tests.
231
232 2012-12-18  Pravin D  <pravin.d@samsung.com>
233
234         Unreviewed, adding secondary email.
235
236         * Scripts/webkitpy/common/config/committers.py:
237
238 2012-12-18  Pravin D  <pravin.d@samsung.com>
239
240         Unreviewed, promoting myself to committer's list.
241
242         * Scripts/webkitpy/common/config/committers.py:
243
244 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
245
246         [EFL] ImageDiff does not build on WebKit2 only builds
247         https://bugs.webkit.org/show_bug.cgi?id=105288
248
249         Reviewed by Kenneth Rohde Christiansen.
250
251         Moved ImageDiff to the efl/ folder inside Tools/ (so we don't pollute
252         the folder with things which are EFL-specific.
253
254         ImageDiff will be built by default now regardless of WebKit API flavor
255         you choose.
256
257         * CMakeLists.txt:
258         * DumpRenderTree/efl/CMakeLists.txt:
259         * efl/ImageDiff/CMakeLists.txt: Added.
260         * efl/ImageDiff/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/efl/ImageDiff.cpp.
261
262 2012-12-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
263
264         [EFL][WK2] When creating new windows, use the opener's context
265         https://bugs.webkit.org/show_bug.cgi?id=105283
266
267         Reviewed by Laszlo Gombos.
268
269         Use the context of the opener if exists.
270
271         * MiniBrowser/efl/main.c:
272         (on_key_down):
273         (on_window_create):
274         (window_create):
275         (elm_main):
276
277 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
278
279         [EFL][WK2] Gardening of TestWebKitAPI tests
280         https://bugs.webkit.org/show_bug.cgi?id=105268
281
282         Reviewed by Kenneth Rohde Christiansen.
283
284         Some tests were marked as failing but are passing and the newly added
285         LayoutUnit should have it's own binary because it conflicts with the
286         fixture used by KURL and is failing on the bots because of that.
287
288         * TestWebKitAPI/CMakeLists.txt:
289         * TestWebKitAPI/PlatformEfl.cmake:
290
291 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
292
293         [chromium] move webintent related methods to TestRunner library
294         https://bugs.webkit.org/show_bug.cgi?id=105164
295
296         Reviewed by Darin Fisher.
297
298         Also deleted the unused deliveredIntentFailure and
299         deliveredIntentResult methods.
300
301         * DumpRenderTree/chromium/DRTTestRunner.cpp:
302         (DRTTestRunner::DRTTestRunner):
303         * DumpRenderTree/chromium/DRTTestRunner.h:
304         (DRTTestRunner):
305         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
306         (WebKit):
307         (WebTestRunner::WebTestDelegate::setCurrentWebIntentRequest):
308         (WebTestRunner::WebTestDelegate::currentWebIntentRequest):
309         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
310         (WebKit):
311         (WebTestProxyBase):
312         (WebTestRunner::WebTestProxy::registerIntentService):
313         (WebTestRunner::WebTestProxy::dispatchIntent):
314         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
315         (WebTestRunner::TestRunner::TestRunner):
316         (WebTestRunner::TestRunner::sendWebIntentResponse):
317         (WebTestRunner):
318         (WebTestRunner::TestRunner::deliverWebIntent):
319         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
320         (TestRunner):
321         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
322         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
323         (WebTestRunner::WebTestProxyBase::registerIntentService):
324         (WebTestRunner):
325         (WebTestRunner::WebTestProxyBase::dispatchIntent):
326         * DumpRenderTree/chromium/WebViewHost.cpp:
327         (WebViewHost::setCurrentWebIntentRequest):
328         (WebViewHost::currentWebIntentRequest):
329         * DumpRenderTree/chromium/WebViewHost.h:
330         (WebKit):
331         (WebViewHost):
332
333 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
334
335         [chromium] move editing callback dumping to the WebTestProxy
336         https://bugs.webkit.org/show_bug.cgi?id=105013
337
338         Reviewed by Darin Fisher.
339
340         To decide whether or not to dump the editing callbacks, the WebTestProxy
341         needs to query the TestRunner. For this purpose, I introduce a
342         WebTestRunner class that just exposes the shouldDumpEditingCallbacks
343         method. In contrast to e.g. WebEventSender, the WebTestRunner is just
344         an interface, and it is not owned by the WebTestInterfaces, but the
345         embedder of the TestRunner library must provide it. This will change
346         once all TestRunner methods are in the library.
347
348         * DumpRenderTree/DumpRenderTree.gypi:
349         * DumpRenderTree/chromium/DRTTestRunner.cpp:
350         (DRTTestRunner::DRTTestRunner):
351         (DRTTestRunner::reset):
352         * DumpRenderTree/chromium/DRTTestRunner.h:
353         (DRTTestRunner):
354         (DRTTestRunner::taskList):
355         (DRTTestRunner::WorkQueue::taskList):
356         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
357         (WebTestRunner):
358         (WebTestInterfaces):
359         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
360         (WebKit):
361         (WebTestRunner):
362         (WebTestProxyBase):
363         (WebTestRunner::WebTestProxy::shouldBeginEditing):
364         (WebTestRunner::WebTestProxy::shouldEndEditing):
365         (WebTestRunner::WebTestProxy::shouldInsertNode):
366         (WebTestRunner::WebTestProxy::shouldInsertText):
367         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
368         (WebTestRunner::WebTestProxy::shouldDeleteRange):
369         (WebTestRunner::WebTestProxy::shouldApplyStyle):
370         (WebTestRunner::WebTestProxy::didBeginEditing):
371         (WebTestRunner::WebTestProxy::didChangeSelection):
372         (WebTestRunner::WebTestProxy::didChangeContents):
373         (WebTestRunner::WebTestProxy::didEndEditing):
374         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h: Copied from Tools/DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h.
375         (WebTestRunner):
376         (WebTestRunner::WebTestRunner::shouldDumpEditingCallbacks):
377         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
378         (WebTestRunner::TestRunner::TestRunner):
379         (WebTestRunner::TestRunner::~TestRunner):
380         (WebTestRunner):
381         (WebTestRunner::TestRunner::reset):
382         (WebTestRunner::TestRunner::shouldDumpEditingCallbacks):
383         (WebTestRunner::TestRunner::dumpEditingCallbacks):
384         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
385         (TestRunner):
386         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
387         (WebTestRunner::WebTestInterfaces::Internal::testRunner):
388         (WebTestRunner::WebTestInterfaces::Internal::setTestRunner):
389         (WebTestInterfaces::Internal):
390         (WebTestRunner::WebTestInterfaces::Internal::Internal):
391         (WebTestRunner::WebTestInterfaces::testRunner):
392         (WebTestRunner):
393         (WebTestRunner::WebTestInterfaces::setTestRunner):
394         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
395         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
396         (WebTestRunner):
397         (WebTestRunner::WebTestProxyBase::shouldBeginEditing):
398         (WebTestRunner::WebTestProxyBase::shouldEndEditing):
399         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
400         (WebTestRunner::WebTestProxyBase::shouldInsertText):
401         (WebTestRunner::WebTestProxyBase::shouldChangeSelectedRange):
402         (WebTestRunner::WebTestProxyBase::shouldDeleteRange):
403         (WebTestRunner::WebTestProxyBase::shouldApplyStyle):
404         (WebTestRunner::WebTestProxyBase::didBeginEditing):
405         (WebTestRunner::WebTestProxyBase::didChangeSelection):
406         (WebTestRunner::WebTestProxyBase::didChangeContents):
407         (WebTestRunner::WebTestProxyBase::didEndEditing):
408         * DumpRenderTree/chromium/TestShell.cpp:
409         (TestShell::initialize):
410         (TestShell::createNewWindow):
411         * DumpRenderTree/chromium/WebViewHost.cpp:
412         (WebViewHost::shouldBeginEditing):
413         (WebViewHost::shouldEndEditing):
414         (WebViewHost::shouldInsertNode):
415         (WebViewHost::shouldInsertText):
416         (WebViewHost::shouldChangeSelectedRange):
417         (WebViewHost::shouldDeleteRange):
418         (WebViewHost::shouldApplyStyle):
419         * DumpRenderTree/chromium/WebViewHost.h:
420         (WebViewHost):
421
422 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
423
424         [chromium] Fix methods in TestRunner that directly modify the settings instead of going through WebPreferences
425         https://bugs.webkit.org/show_bug.cgi?id=105154
426
427         Reviewed by Tony Chang.
428
429         If the settings are modified directly, the embedder can't keep track of
430         them and might accidentially override the changed settings.
431
432         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
433         (WebPreferences):
434         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
435         (WebTestRunner::TestRunner::TestRunner):
436         (WebTestRunner::TestRunner::setAsynchronousSpellCheckingEnabled):
437         (WebTestRunner):
438         (WebTestRunner::TestRunner::setMinimumTimerInterval):
439         (WebTestRunner::TestRunner::setTouchDragDropEnabled):
440         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
441         (TestRunner):
442         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
443         (WebTestRunner::WebPreferences::reset):
444         (WebTestRunner::WebPreferences::applyTo):
445
446 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
447
448         Unreviewed unit test skipping.
449
450         Filed https://bugs.webkit.org/show_bug.cgi?id=105253 to re-enable them.
451
452         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
453         (TestWebKitAPI):
454         (TestWebKitAPI::TEST):
455
456 2012-12-17  Dean Jackson  <dino@apple.com>
457
458         Unreviewed. Add Silvia Pfeiffer to contributor list.
459
460         * Scripts/webkitpy/common/config/committers.py:
461
462 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
463
464         build.webkit.org - add more google windows bots
465         https://bugs.webkit.org/show_bug.cgi?id=105225
466
467         Reviewed by Eric Seidel.
468
469         Adds two more windows bots (google-windows-3, a builder, and
470         google-windows-4, a tester) for the Chromium configs.
471
472         * BuildSlaveSupport/build.webkit.org-config/config.json:
473
474 2012-12-17  Eric Seidel  <eric@webkit.org>
475
476         Cleanup --profiler=perf code a little and allow up to 1000 report files (instead of 100)
477         https://bugs.webkit.org/show_bug.cgi?id=105217
478
479         Reviewed by Dirk Pranke.
480
481         Previously --profile would fail strangely if find_unused_filename returned None
482         (which would happen as soon as we had 100 report files).
483         Now we'll search through 1000 filenames before giving up (and assert if we failed to find one).
484         I also made some minor cleanups to the output of perf, and tried to make the
485         commands more self-documenting by using long-names for args.
486
487         * Scripts/webkitpy/common/system/profiler.py:
488         (SingleFileOutputProfiler.__init__):
489         (Perf.attach_to_pid):
490         (Perf.profile_after_exit):
491
492 2012-12-17  Sheriff Bot  <webkit.review.bot@gmail.com>
493
494         Unreviewed, rolling out r137198.
495         http://trac.webkit.org/changeset/137198
496         https://bugs.webkit.org/show_bug.cgi?id=105212
497
498         This patch is causing API behavior compatibility problems
499         (Requested by zdobersek on #webkit).
500
501         * DumpRenderTree/gtk/DumpRenderTree.cpp:
502         (willSendRequestCallback):
503
504 2012-12-17  Elliott Sprehn  <esprehn@chromium.org>
505
506         Unreviewed. Adding myself to committers list.
507
508         * Scripts/webkitpy/common/config/committers.py:
509
510 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
511
512         nrwt: shuffle code around for cleanup in run_webkit_tests.py
513         https://bugs.webkit.org/show_bug.cgi?id=105078
514
515         Reviewed by Ojan Vafai.
516
517         This patch reorders functions so that run_webkit_tests can be
518         understood in a top-down matter better, and to make things slightly
519         more sensible when I merge manager.py into it.
520
521         Also, this adds tests for the actual main() routine, which had several
522         bugs get through testing a week or two ago.
523
524         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
525         (main):
526         (parse_args):
527         (_set_up_derived_options):
528         (run):
529         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
530         (RunTest):
531         (RunTest.test_no_http_tests):
532         (PortTest.disabled_test_mac_lion):
533         (MainTest):
534         (MainTest.test_exception_handling):
535         (MainTest.test_exception_handling.interrupting_run):
536         (MainTest.test_exception_handling.successful_run):
537         (MainTest.test_exception_handling.successful_run.FakeRunDetails):
538         (MainTest.test_exception_handling.exception_raising_run):
539
540 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
541
542         webkitpy: move --lint-test-files code into its own module
543         https://bugs.webkit.org/show_bug.cgi?id=105077
544
545         Reviewed by Eric Seidel.
546
547         It really didn't need to be in run_webkit_tests.py anymore.
548         This patch adds a new lint-test-expectations scripts to Tools/Scripts;
549         rwt --lint-test-files is still supported for backwards compatibility
550         for now.
551
552         * Scripts/lint-test-expectations: Added.
553         * Scripts/webkitpy/layout_tests/lint_test_expectations.py: Added.
554         (lint):
555         (main):
556         * Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py: Added.
557         (FakePort):
558         (FakePort.__init__):
559         (FakePort.test_configuration):
560         (FakePort.expectations_dict):
561         (FakePort.skipped_layout_tests):
562         (FakePort.all_test_configurations):
563         (FakePort.configuration_specifier_macros):
564         (FakePort.get_option):
565         (FakeFactory):
566         (FakeFactory.__init__):
567         (FakeFactory.get):
568         (FakeFactory.all_port_names):
569         (LintTest):
570         (LintTest.test_all_configurations):
571         (LintTest.test_lint_test_files):
572         (LintTest.test_lint_test_files__errors):
573         (MainTest):
574         (MainTest.test_success):
575         (MainTest.test_success.interrupting_lint):
576         (MainTest.test_success.successful_lint):
577         (MainTest.test_success.exception_raising_lint):
578         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
579         (main):
580         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
581         (StreamTestingMixin.assertNotEmpty):
582
583 2012-12-17  Adam Klein  <adamk@chromium.org>
584
585         build-webkit: rename --template-tag to --template-element to match ENABLE #define name
586         https://bugs.webkit.org/show_bug.cgi?id=105072
587
588         Reviewed by Laszlo Gombos.
589
590         * Scripts/webkitperl/FeatureList.pm:
591
592 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
593
594         Add some unit testing for WTF::clampTo* functions
595         https://bugs.webkit.org/show_bug.cgi?id=105060
596
597         Reviewed by Emil A Eklund.
598
599         The new tests are targeted at clamping to int / unsigned and float as those are
600         the one explicitely exposed through more dedicated functions (like clampToInteger,
601         clampToPositiveInteger and clampToFloat).
602
603         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
604
605 2012-12-17  Eric Seidel  <eric@webkit.org>
606
607         rpt --profile --chromium-android throws exception
608         https://bugs.webkit.org/show_bug.cgi?id=105061
609
610         Reviewed by Daniel Bates.
611
612         Somehow popen(args) got turned into popen(*args) at some point,
613         which caused my new stringify code to fail when args was passed by name.
614         Fixed and tested.
615
616         * Scripts/webkitpy/common/system/executive.py:
617         (Executive._stringify_args):
618         (Executive):
619         (Executive.popen):
620         * Scripts/webkitpy/common/system/executive_unittest.py:
621         (ExecutiveTest.test_popen_args):
622
623 2012-12-17  Mark Pilgrim  <pilgrim@chromium.org>
624
625         Use Platform::current() instead of webKitPlatformSupport() in DumpRenderTree
626         https://bugs.webkit.org/show_bug.cgi?id=105054
627
628         Reviewed by Jochen Eisinger.
629
630         Part of a refactoring series; see tracking bug 82948.
631
632         * DumpRenderTree/chromium/NotificationPresenter.cpp:
633         (NotificationPresenter::show):
634         * DumpRenderTree/chromium/Task.cpp:
635         (postTask):
636         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
637
638 2012-12-17  Emil A Eklund  <eae@chromium.org>
639
640         Clamp values in LayoutUnit::operator/ when SATURATED_LAYOUT_ARITHMETIC is enabled
641         https://bugs.webkit.org/show_bug.cgi?id=104955
642
643         Reviewed by Julien Chaffraix.
644
645         Add tests for LayoutUnit.
646
647         * TestWebKitAPI/CMakeLists.txt:
648         * TestWebKitAPI/ForwardingHeaders: Added.
649         * TestWebKitAPI/ForwardingHeaders/WebCore: Added.
650         * TestWebKitAPI/ForwardingHeaders/WebCore/LayoutUnit.h: Added.
651         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
652         * TestWebKitAPI/TestWebKitAPI.gypi:
653         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
654         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp: Added.
655         (TestWebKitAPI):
656         (TestWebKitAPI::TEST):
657         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
658
659 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
660
661         [chromium] removed unused preferences from testRunner.overridePreference
662         https://bugs.webkit.org/show_bug.cgi?id=105169
663
664         Reviewed by Tony Chang.
665
666         Most notably this removes the code related to font family maps which is
667         now handled via window.internals.settings
668
669         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
670         (WebPreferences):
671         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
672         (WebTestRunner::TestRunner::overridePreference):
673         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
674         (TestRunner):
675         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
676         (WebTestRunner::WebPreferences::applyTo):
677
678 2012-12-17  James Robinson  <jamesr@chromium.org>
679
680         [chromium] Move creation of WebLayerTreeView to WebWidgetClient
681         https://bugs.webkit.org/show_bug.cgi?id=105071
682
683         Reviewed by Adrienne Walker.
684
685         Update WebViewHost for the new WebWidgetClient interface.
686
687         * DumpRenderTree/chromium/WebViewHost.cpp:
688         (WebViewHost::initializeLayerTreeView):
689         (WebViewHost::layerTreeView):
690         (WebViewHost::~WebViewHost):
691         * DumpRenderTree/chromium/WebViewHost.h:
692         (WebViewHost):
693
694 2012-12-17  Kenneth Rohde Christiansen  <kenneth@webkit.org>
695
696         [EFL][WK2] window_create doesn't receive the url
697         https://bugs.webkit.org/show_bug.cgi?id=105184
698
699         Reviewed by Alexis Menard.
700
701         * MiniBrowser/efl/main.c: Use the url supplied to window_create
702         (on_window_create):
703
704 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
705
706         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
707         https://bugs.webkit.org/show_bug.cgi?id=105026
708
709         Reviewed by Yury Semikhatsky.
710
711         Old SFINAE test was replaced with new one based on this article:
712         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
713
714         New test which covers this problem was added.
715
716         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
717
718 2012-12-17  Anthony Scian  <ascian@rim.com>
719
720         [Blackberry] Static code analysis warning fixes
721         https://bugs.webkit.org/show_bug.cgi?id=105129
722
723         Reviewed by Rob Buis.
724
725         Added null checks for fopen calls, no recovery attempted; access fault prevention only.
726
727         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
728         (BlackBerry::WebKit::createFile):
729         (BlackBerry::WebKit::DumpRenderTree::runTest):
730
731 2012-12-17  John Mellor  <johnme@chromium.org>
732
733         Add Anton Vayvod and Tim Volodine as contributors in committers.py
734         https://bugs.webkit.org/show_bug.cgi?id=105025
735
736         Unreviewed. Adds two contributors to committers.py.
737         They are initially contributing to the Text Autosizing feature (master bug: 84186).
738
739         * Scripts/webkitpy/common/config/committers.py:
740
741 2012-12-17  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
742
743         Unreviewed, add myself to committers.py.
744
745         * Scripts/webkitpy/common/config/committers.py:
746
747 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
748
749         [Qt] remove some unnecessary CONFIG additions
750
751         Reviewed by Simon Hausmann.
752
753         qt is already added by spec_pre.prf, warn_on and depend_includepath by
754         default_pre.prf.
755
756         * qmake/mkspecs/features/default_post.prf:
757
758 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
759
760         Unreviewed, rolling out r137892.
761         http://trac.webkit.org/changeset/137892
762         https://bugs.webkit.org/show_bug.cgi?id=105026
763
764         it broke compilation on windows
765
766         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
767
768 2012-12-14  Ilya Tikhonovsky  <loislo@chromium.org>
769
770         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
771         https://bugs.webkit.org/show_bug.cgi?id=105026
772
773         Reviewed by Yury Semikhatsky.
774
775         Old SFINAE test was replaced with new one based on this article:
776         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
777
778         New test which covers this problem was added.
779
780         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
781
782 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
783
784         [Qt] rely on automatic output directory setup
785
786         Reviewed by Simon Hausmann.
787
788         Now being a proper qt module (by virtue of having load(qt_build_config)
789         in .qmake.conf), webkit gets the path setup goodies for free.
790
791         This also fixes Makefile.api.Debug/Release trying to generate qrc_WebKit.cpp
792         at the same time, instead of the debug and release version ending up in different
793         directories as they should.
794
795         * qmake/mkspecs/features/default_pre.prf:
796
797 2012-12-17  Charles Wei  <charles.wei@torchmobile.com.cn>
798
799         [BlackBerry] Need to re-enable video track
800         https://bugs.webkit.org/show_bug.cgi?id=105155
801
802         Reviewed by George Staikos.
803
804         Just enable Video Track for BlackBerry.
805
806         * Scripts/webkitperl/FeatureList.pm:
807
808 2012-12-17  János Badics  <jbadics@inf.u-szeged.hu>
809
810         Unreviewed. Added myself to committers.py
811
812         * Scripts/webkitpy/common/config/committers.py:
813
814 2012-12-17  Jonathan Liu  <net147@gmail.com>
815
816         Fix linking of libraries with -Wl,-whole-archive
817         https://bugs.webkit.org/show_bug.cgi?id=104436
818
819         Reviewed by Simon Hausmann.
820
821         If there is more than one library in LIBS linked with
822         -Wl,-whole-archive -l... -Wl,-no-whole-archive, qmake tries to
823         merge the duplicate flags in LIBS which causes linking to fail.
824
825         Add no_smart_library_merge to CONFIG to prevent the duplicate
826         flags from being removed when using -Wl,-whole-archive.
827
828         * qmake/mkspecs/features/functions.prf:
829
830 2012-12-16  Arvid Nilsson  <anilsson@rim.com>
831
832         [BlackBerry] Adapt to new BlackBerry::Platform::TouchPoint API
833         https://bugs.webkit.org/show_bug.cgi?id=105143
834         RIM PR 171941
835
836         Reviewed by Rob Buis.
837         Internally reviewed by George Staikos.
838
839         Adapt to new method names and encapsulation of TouchPoint data members.
840
841         * DumpRenderTree/blackberry/EventSender.cpp:
842         (addTouchPointCallback):
843         (updateTouchPointCallback):
844         (touchEndCallback):
845         (releaseTouchPointCallback):
846         (sendTouchEvent):
847
848 2012-12-16  Maciej Stachowiak  <mjs@apple.com>
849
850         Add a script to compute the mean and 95% confidence interval (using two-sided t-test) of a sample set
851         https://bugs.webkit.org/show_bug.cgi?id=105148
852
853         Reviewed by Filip Pizlo.
854
855         Usage examples:
856         
857         $ sampstat --help
858         Usage: sampstat [options] [FILES]
859           Compute the mean and 95% confidence interval of a sample set.
860           Standard input or files must contain two or more decimal numbers, one per line.
861         
862         Options:
863           -h, --help            show this help message and exit
864           -u UNIT, --unit=UNIT  assume values are in units of UNIT
865           -v, --verbose         print all values (with units)
866         
867         $ sampstat -u MB memresults.txt
868         2356.90 MB +/- 101.34 MB (4.3%)
869         
870         $ sampstat -v -u MB memresults.txt
871               2318.21 MB
872               2399.56 MB
873               2352.93 MB
874         ----------------
875         Mean: 2356.90 MB +/- 101.34 MB (4.3%)
876         
877
878         * Scripts/sampstat: Added.
879         (sum): Helper function to compute the sum of a list.
880         (arithmeticMean): Compute the meam of a list.
881         (standardDeviation): Compute the sample standard deviation (unbiased estimator).
882         (standardError): Compute the sample standard error.
883         (tDist): Compute t(.025, n-1), the t-value for atwo-sided 95% confidence interval. 
884         (twoSidedConfidenceInterval): Compute the two-sided confidence interval range about the mean.
885
886 2012-12-15  Nima Ghanavatian  <nghanavatian@rim.com>
887
888         [BlackBerry] Pass key modifiers with touch and mouse events
889         https://bugs.webkit.org/show_bug.cgi?id=105108
890
891         Reviewed by Rob Buis.
892
893         Update DRT to new api.
894
895         Internally reviewed by Gen Mak.
896
897         * DumpRenderTree/blackberry/EventSender.cpp:
898         (mouseDownCallback):
899         (mouseUpCallback):
900         (mouseMoveToCallback):
901
902 2012-12-15  Zan Dobersek  <zandobersek@gmail.com>
903
904         Stop the XvfbDriver before proceeding with its starting setup
905         https://bugs.webkit.org/show_bug.cgi?id=105006
906
907         Reviewed by Philippe Normand.
908
909         Just as with the base Driver implementation, stop the XvfbDriver
910         before proceeding with setting up lock files and a new Xvfb instance.
911         This forces any driver or Xvfb instances that are currently under
912         the driver's control to be closed/killed and cleaned up after.
913
914         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
915         (XvfbDriver._start):
916
917 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
918
919         webkitpy: make port_name a required parameter to the non-test ports
920         https://bugs.webkit.org/show_bug.cgi?id=105057
921
922         Reviewed by Eric Seidel.
923
924         This is just minor code cleanup along the path to making Port
925         constructors simpler and pushing more logic into the port factory.
926
927         * Scripts/webkitpy/layout_tests/port/base.py:
928         (Port.__init__):
929         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
930         (PortTest.make_port):
931         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
932         (TestWebKitPort.__init__):
933         * Scripts/webkitpy/layout_tests/port/test.py:
934         (TestPort):
935         (TestPort.determine_full_port_name):
936         (TestPort.__init__):
937         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py:
938         (XvfbDriverTest.make_driver):
939         (XvfbDriverTest.test_stop):
940
941 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
942
943         garden-o-matic doesn't know about reftests
944         https://bugs.webkit.org/show_bug.cgi?id=101976
945
946         Reviewed by Eric Seidel.
947
948         Attempt to make garden-o-matic properly aware of reftests again.
949
950         This is re-landing r137407 with a change to
951         rebaselineWithStatusUpdates() to actually pass the filtered
952         failureInofLost to checkout.rebaseline().
953
954         Unfortunately, the testing coverage for this whole module is almost
955         non-existent, and it's not obvious that there's a good way to
956         stub out checkout.rebaseline() to test that the right thing
957         happens. I verified the change manually.
958
959         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
960         (.):
961         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
962         (.):
963         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
964
965 2012-12-14  Ryosuke Niwa  <rniwa@webkit.org>
966
967         "Running 1 DumpRenderTree over X shards" is not a helpful output
968         https://bugs.webkit.org/show_bug.cgi?id=104858
969
970         Reviewed by Dirk Pranke.
971
972         The word "shard" is not typically used in test frameworks. This word is confusing for many WebKit
973         contributors who typically don't have much experience on distributed computations and databases.
974         So only print this message out in the debug mode.
975
976         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
977         * Scripts/webkitpy/layout_tests/views/printing.py:
978         (Printer.print_workers_and_shards):
979
980 2012-12-13  Eric Seidel  <eric@webkit.org>
981
982         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
983         https://bugs.webkit.org/show_bug.cgi?id=104891
984
985         Reviewed by Ryosuke Niwa.
986
987         I also implemented a very simple "Sample" Profiler using
988         Mac OS X's /usr/bin/sample command line tool.
989
990         The real reason for this abstraction is to make it easy
991         to support both perf and pprof on linux which seem to
992         be about equally popular among those I ask in the Chrome team.
993
994         * Scripts/webkitpy/common/system/profiler.py:
995         (ProfilerFactory.create_profiler):
996         (ProfilerFactory):
997         (ProfilerFactory.available_profilers_by_name):
998         (ProfilerFactory.default_profiler_name):
999         (Sample):
1000         (Sample.__init__):
1001         (Sample.attach_to_pid):
1002         (Sample.profile_after_exit):
1003         (IProfiler.attach_to_pid):
1004         * Scripts/webkitpy/layout_tests/port/driver.py:
1005         (Driver.__init__):
1006         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1007         (PerfTestsRunner._parse_args):
1008
1009 2012-12-13  Eric Seidel  <eric@webkit.org>
1010
1011         Unreviewed, fixing typo in python unittest.
1012
1013         I changed the default profiler on linux from pprof to perf in
1014         bug 104971.  I failed to update the unittest results at that time.
1015
1016         * Scripts/webkitpy/common/system/profiler_unittest.py:
1017         (ProfilerFactoryTest.test_basic):
1018
1019 2012-12-13  Ian Vollick  <vollick@chromium.org>
1020
1021         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
1022         https://bugs.webkit.org/show_bug.cgi?id=104911
1023
1024         Reviewed by Dirk Pranke.
1025
1026         Adds two virtual tests suites to run the tests in
1027         compositing/overflow/ and scrollbars/ with the flag
1028         --enable-accelerated-overflow-scroll, which I've also plumbed through
1029         in the usual way.
1030
1031         * DumpRenderTree/chromium/DumpRenderTree.cpp:
1032         (main):
1033         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1034         (WebPreferences):
1035         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1036         (WebTestRunner::WebPreferences::reset):
1037         (WebTestRunner::WebPreferences::applyTo):
1038         * DumpRenderTree/chromium/TestShell.cpp:
1039         (TestShell::TestShell):
1040         (TestShell::resetWebSettings):
1041         * DumpRenderTree/chromium/TestShell.h:
1042         (TestShell::setAcceleratedCompositingForOverflowScrollEnabled):
1043         (TestShell):
1044         * Scripts/webkitpy/layout_tests/port/chromium.py:
1045         (ChromiumPort.virtual_test_suites):
1046
1047 2012-12-13  Eric Seidel  <eric@webkit.org>
1048
1049         Use 'perf' to profile on linux instead of google-pprof by default
1050         https://bugs.webkit.org/show_bug.cgi?id=104971
1051
1052         Reviewed by Dirk Pranke.
1053
1054         This makes Chromium Linux match Chromium Android and use the perf
1055         tool by default.  Once https://bugs.webkit.org/show_bug.cgi?id=104891
1056         lands it will be possible to use pprof again on Linux.
1057
1058         This is slightly more advanced than the AndroidPerf profiler as
1059         instead of using a timeout on "perf record" I instead watch
1060         for the termination of the target process and then control-C
1061         the 'perf record' process.  This required me to add two new
1062         methods to Executive, one to have a limited-time wait() and
1063         the second to be able to send a control-C.  I chose to add
1064         these to Executive to make them easier to mock/fix-for-win32
1065         at a later time if needed.
1066
1067         * Scripts/webkitpy/common/system/executive.py:
1068         (Executive.wait_limited):
1069         (Executive.interrupt):
1070         * Scripts/webkitpy/common/system/profiler.py:
1071         (ProfilerFactory.create_profiler):
1072         (Perf):
1073         (Perf.__init__):
1074         (Perf._perf_path):
1075         (Perf.attach_to_pid):
1076         (Perf._first_ten_lines_of_profile):
1077         (Perf.profile_after_exit):
1078
1079 2012-12-13  Julie Parent  <jparent@chromium.org>
1080
1081         Dashboard cleanup: remove usage of global g_builders.
1082         https://bugs.webkit.org/show_bug.cgi?id=104941
1083
1084         Reviewed by Dirk Pranke.
1085
1086         The dashboards use a lot of global state, which makes hacking on them
1087         complicated. This change removes the use of one such global: g_builders.
1088         In most cases, we can just use currentBuilderGroup().builders instead,
1089         which is now currentBuilders().
1090         Surprisingly, the most changes were required to the unit tests, since
1091         they were even bigger offenders of bad hygiene, relying on global state
1092         set by other tests, randomly clobbering global variables in ways the
1093         real code doesn't, etc.
1094
1095         * TestResultServer/static-dashboards/builders.js:
1096         (BuilderGroup.prototype.setup):
1097         * TestResultServer/static-dashboards/dashboard_base.js:
1098         (.switch.return):
1099         (htmlForTestTypeSwitcher):
1100         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1101         (generatePage):
1102         (getAllTestsTrie):
1103         (processTestRunsForAllBuilders):
1104         (showPopupForBuild):
1105         (generatePageForExpectationsUpdate):
1106         (loadExpectationsLayoutTests):
1107         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1108         (resetGlobals):
1109         (stubResultsByBuilder):
1110         (test):
1111         * TestResultServer/static-dashboards/loader.js:
1112         (.):
1113         * TestResultServer/static-dashboards/loader_unittests.js:
1114
1115 2012-12-13  Eric Seidel  <eric@webkit.org>
1116
1117         Callers should not have to stringify args before calling Executive run_command/popen
1118         https://bugs.webkit.org/show_bug.cgi?id=104975
1119
1120         Reviewed by Dirk Pranke.
1121
1122         One could argue that we should match the python call syntax here,
1123         but I think it's a more friendly API if we automagically handle
1124         stringification of args in run_command, etc.
1125         This removes map(unicode, args) from several callsites.
1126
1127         When I first tried to land this change, I didn't realize that
1128         Executive._command_for_printing depended on this behavior
1129         having been applied to args in run_command.  The fix is to
1130         call _stringify_args in both run_command and popen.
1131         This is slightly redundant, but given how short args have to be
1132         (due to shell limits), I don't think the double-encode check
1133         matters in practice.
1134
1135         This is slightly complicated by the fact that apache_http_server.py
1136         is the one caller in our codebase which uses shell=True.
1137         shell=True is a well-documented trail-of-tears:
1138         http://stackoverflow.com/questions/3172470/actual-meaning-of-shell-true-in-subprocess
1139         but to support this legacy (windows-only) code (which I can't easily test)
1140         I've added an if-hack to avoid stringifying the the popen(shell=True) case.
1141
1142         * Scripts/webkitpy/common/system/executive.py:
1143         (Executive.run_command):
1144         (Executive._stringify_args):
1145         (Executive.popen):
1146         * Scripts/webkitpy/common/system/executive_unittest.py:
1147         (ExecutiveTest.test_auto_stringify_args):
1148         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1149         (attach_to_pid):
1150
1151 2012-12-13  Adrienne Walker  <enne@chromium.org>
1152
1153         Unreviewed, rolling out r137645, r137646, and r137667.
1154         http://trac.webkit.org/changeset/137645
1155         http://trac.webkit.org/changeset/137646
1156         http://trac.webkit.org/changeset/137667
1157         https://bugs.webkit.org/show_bug.cgi?id=104911
1158
1159         Breaks some overflow layout tests
1160
1161         * Scripts/webkitpy/layout_tests/port/chromium.py:
1162         (ChromiumPort.virtual_test_suites):
1163
1164 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1165
1166         Unreviewed, rolling out r137661.
1167         http://trac.webkit.org/changeset/137661
1168         https://bugs.webkit.org/show_bug.cgi?id=104891
1169
1170         broke unit tests, run-webkit-tests
1171
1172         * Scripts/webkitpy/common/system/executive.py:
1173         (Executive._run_command_with_teed_output):
1174         (Executive.run_command):
1175         (Executive.popen):
1176         * Scripts/webkitpy/common/system/profiler.py:
1177         (ProfilerFactory.create_profiler):
1178         (GooglePProf.profile_after_exit):
1179         (IProfiler.attach_to_pid):
1180         * Scripts/webkitpy/layout_tests/port/driver.py:
1181         (Driver.__init__):
1182         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1183         (PerfTestsRunner._parse_args):
1184
1185 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1186
1187         nrwt: move rundetails, summarize_results into test_run_results
1188         https://bugs.webkit.org/show_bug.cgi?id=104963
1189
1190         Reviewed by Eric Seidel.
1191
1192         More cleanup; this the plain-old-data functions out of manager.py
1193         and into test_run_results next to the other aggregate data structures.
1194
1195         Also this cleans up a bunch of test code and reduces duplication.
1196
1197         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1198         (Manager.run):
1199         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1200         (ManagerTest.test_look_for_new_crash_logs):
1201         * Scripts/webkitpy/layout_tests/models/test_run_results.py:
1202         (TestRunResults.__init__):
1203         (TestRunResults.add):
1204         (RunDetails):
1205         (RunDetails.__init__):
1206         (_interpret_test_failures):
1207         (summarize_results):
1208         * Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py: Added.
1209         (get_result):
1210         (run_results):
1211         (summarized_results):
1212         (InterpretTestFailuresTest):
1213         (InterpretTestFailuresTest.setUp):
1214         (InterpretTestFailuresTest.test_interpret_test_failures):
1215         (SummarizedResultsTest):
1216         (SummarizedResultsTest.setUp):
1217         (SummarizedResultsTest.test_no_svn_revision):
1218         (SummarizedResultsTest.test_svn_revision):
1219         (SummarizedResultsTest.test_summarized_results_wontfix):
1220         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
1221         (BuildBotPrinterTests.get_printer):
1222         (BuildBotPrinterTests.test_print_unexpected_results):
1223         (BuildBotPrinterTests.test_print_results):
1224
1225 2012-12-13  Eric Seidel  <eric@webkit.org>
1226
1227         Make --profile output easier to understand for pprof on linux
1228         https://bugs.webkit.org/show_bug.cgi?id=104957
1229
1230         Reviewed by Dirk Pranke.
1231
1232         pprof is kinda mysterious, so we print some information
1233         on how to use it, and how to learn how to read the sample dump.
1234         This is similar to what we've done for perf output for Chromium Android.
1235
1236         * Scripts/webkitpy/common/system/profiler.py:
1237         (GooglePProf._pprof_path):
1238         (GooglePProf.profile_after_exit):
1239
1240 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1241
1242         Revert r137407 - it didn't actually work right.
1243         https://bugs.webkit.org/show_bug.cgi?id=101976
1244
1245         Unreviewed, build fix.
1246
1247         Turns out that rebaselining was completely broken with this change :(
1248         We're passing in a list of test names rather than the list of
1249         failureInfo objects.
1250
1251         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1252         (.):
1253         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1254         (.):
1255         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
1256
1257 2012-12-13  Antoine Labour  <piman@chromium.org>
1258
1259         [chromium] Small fixes for WebPluginContainer::setWebLayer
1260         https://bugs.webkit.org/show_bug.cgi?id=104953
1261
1262         Reviewed by James Robinson.
1263
1264         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Clean up
1265         dangling pointers, force correct invalidation of the layer.
1266
1267 2012-12-13  Eric Seidel  <eric@webkit.org>
1268
1269         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
1270         https://bugs.webkit.org/show_bug.cgi?id=104891
1271
1272         Reviewed by Dirk Pranke.
1273
1274         I also implemented a very simple "Sample" Profiler using
1275         Mac OS X's /usr/bin/sample command line tool.
1276
1277         I also moved the map(unicode, args) call into Executive.popen
1278         so that all callers don't have to do it themselves.
1279
1280         The real reason for this abstraction is to make it easy
1281         to support both perf and pprof on linux which seem to
1282         be about equally popular among those I ask in the Chrome team.
1283
1284         * Scripts/webkitpy/common/system/executive.py:
1285         (Executive.popen):
1286         * Scripts/webkitpy/common/system/profiler.py:
1287         (ProfilerFactory.create_profiler):
1288         (ProfilerFactory):
1289         (ProfilerFactory.available_profilers_by_name):
1290         (ProfilerFactory.default_profiler_name):
1291         (Sample):
1292         (Sample.__init__):
1293         (Sample.attach_to_pid):
1294         (Sample.profile_after_exit):
1295         (IProfiler.attach_to_pid):
1296         * Scripts/webkitpy/layout_tests/port/driver.py:
1297         (Driver.__init__):
1298         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1299         (PerfTestsRunner._parse_args):
1300
1301 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1302
1303         support -wk2 port names properly in webkitpy.layout_tests.port
1304         https://bugs.webkit.org/show_bug.cgi?id=104761
1305
1306         Reviewed by Eric Seidel.
1307
1308         Our support for WK2 port names was inconsistent at best; this patch
1309         modifies the code so that <port_name>-wk2 is supported for all ports that
1310         have WK2 implementations and is equivalent to specifying <port_name>
1311         and -2/--webkit-test-runner. In addition, this modifies
1312         builders.all_port_names() to include the wk2 variants.
1313
1314         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
1315         * Scripts/webkitpy/layout_tests/port/apple.py:
1316         (ApplePort.determine_full_port_name):
1317         * Scripts/webkitpy/layout_tests/port/base.py:
1318         (Port.determine_full_port_name):
1319         (Port.__init__):
1320         * Scripts/webkitpy/layout_tests/port/builders.py:
1321         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
1322         (DriverTest.make_port):
1323         * Scripts/webkitpy/layout_tests/port/mac.py:
1324         (MacPort.default_baseline_search_path):
1325         * Scripts/webkitpy/layout_tests/port/win.py:
1326         (WinPort.default_baseline_search_path):
1327         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1328         (TestRebaselineExpectations.test_rebaseline_expectations):
1329
1330 2012-12-13  Ian Vollick  <vollick@chromium.org>
1331
1332         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
1333         https://bugs.webkit.org/show_bug.cgi?id=104911
1334
1335         Reviewed by Adrienne Walker.
1336
1337         Adds two virtual tests suites to run the tests in
1338         compositing/overflow/ and scrollbars/ with the flag
1339         --enable-accelerated-compositing-for-overflow-scroll
1340
1341         * Scripts/webkitpy/layout_tests/port/chromium.py:
1342         (ChromiumPort.virtual_test_suites):
1343
1344 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1345
1346         delete extraneous self.fail in mastercfg_unittest.py
1347         https://bugs.webkit.org/show_bug.cgi?id=104940
1348
1349         Reviewed by Eric Seidel.
1350
1351         This was accidentally introduced in r137404.
1352
1353         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
1354
1355 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1356
1357         Fix a mis-named variable after r137621.
1358         https://bugs.webkit.org/show_bug.cgi?id=104871
1359
1360         Unreviewed, build fix.
1361
1362         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
1363         (BuildBotPrinter.print_results):
1364         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
1365
1366 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1367
1368         webkitpy: rename ResultSummary
1369         https://bugs.webkit.org/show_bug.cgi?id=104871
1370
1371         Reviewed by Eric Seidel.
1372
1373         This patch changes the names of ResultSummary and related modules,
1374         classes, and variables as a part of the larger clean up in bug 103824.
1375
1376         This change is basically a bunch of search and replaces:
1377         - result_summary.py -> test_run_results.py
1378         - class ResultSummary -> class TestRunResults
1379         - "result_summary" or "rs" as a generic variable name -> "run_results",
1380           when it doesn't matter (or we don't know) if this refers to the
1381           initial pass over the tests or the retry pass.
1382         - "result_summary" as a variable containing the results of the initial,
1383           complete test run -> "initial_results"
1384         - "retry_summary" as a variable containing the results of retried
1385           results -> "retry_results"
1386         - ResultSummary.results (the property) -> TestRunResults.results_by_name
1387         - ResultSummary.failures (the property) -> TestRunResults.failures_by_name
1388         - RunDetails.result_summary -> RunDetails.initial_results
1389         - RunDetails.retry_summary -> RunDetails.retry_results
1390
1391         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
1392         (LayoutTestRunner.__init__):
1393         (LayoutTestRunner.run_tests):
1394         (LayoutTestRunner._mark_interrupted_tests_as_skipped):
1395         (LayoutTestRunner._interrupt_if_at_failure_limits):
1396         (LayoutTestRunner._interrupt_if_at_failure_limits.interrupt_if_at_failure_limit):
1397         (LayoutTestRunner._update_summary_with_result):
1398         (LayoutTestRunner._handle_finished_test):
1399         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
1400         (FakePrinter.print_expected):
1401         (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
1402         (LayoutTestRunnerTests.test_update_summary_with_result):
1403         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1404         (RunDetails.__init__):
1405         (summarize_results):
1406         (Manager.run):
1407         (Manager._look_for_new_crash_logs):
1408         (Manager._tests_to_retry):
1409         (Manager._upload_json_files):
1410         (Manager._stats_trie):
1411         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1412         (ManagerTest.test_look_for_new_crash_logs):
1413         (SummarizeResultsTest):
1414         (SummarizeResultsTest.get_run_results):
1415         (SummarizeResultsTest.summarized_results):
1416         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
1417         (JSONLayoutResultsGenerator.__init__):
1418         (JSONLayoutResultsGenerator._get_modifier_char):
1419         (JSONLayoutResultsGenerator._insert_failure_summaries):
1420         (JSONLayoutResultsGenerator._get_failure_summary_entry):
1421         * Scripts/webkitpy/layout_tests/models/test_run_results.py: Renamed from Tools/Scripts/webkitpy/layout_tests/models/result_summary.py.
1422         (TestRunResults):
1423         (TestRunResults.__init__):
1424         (TestRunResults.add):
1425         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1426         (get_test_results):
1427         (MainTest.test_basic):
1428         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
1429         (BuildBotPrinter.print_results):
1430         (BuildBotPrinter.print_run_results):
1431         (BuildBotPrinter._print_run_results_entry):
1432         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
1433         (BuildBotPrinterTests.get_run_results):
1434         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
1435         (BuildBotPrinterTests):
1436         * Scripts/webkitpy/layout_tests/views/printing.py:
1437         (Printer.print_expected):
1438         (Printer._print_expected_results_of_type):
1439         (Printer.print_results):
1440         (Printer._print_timing_statistics):
1441         (Printer._print_worker_statistics):
1442         (Printer._print_aggregate_test_statistics):
1443         (Printer._print_individual_test_times):
1444         (Printer._print_directory_timings):
1445         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
1446
1447 2012-12-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1448
1449         [EFL] The WebKit2 bots are building and executing WebKit1 tests
1450         https://bugs.webkit.org/show_bug.cgi?id=104690
1451
1452         Reviewed by Csaba Osztrogonác.
1453
1454         Pass --no-webkit1 option to the EFL WK2 bots when building. This
1455         will disable WK1 API and thus, the WK1 API tests.
1456
1457         * BuildSlaveSupport/build.webkit.org-config/config.json:
1458         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1459         (CompileWebKit2Only):
1460         (BuildAndTestFactory):
1461         (BuildAndTestFactory.__init__):
1462         (BuildAndTestWebKit2OnlyFactory):
1463         Now bots using BuildAndTestWebKit2OnlyFactory (for instance only EFL) will
1464         get --no-webkit1 as a parameter on the build-webkit script.
1465         * Scripts/build-webkit:
1466         Added support for -no-webkit1 for the Qt and EFL port.
1467
1468 2012-12-13  Eric Seidel  <eric@webkit.org>
1469
1470         build-webkit --chromium-android forces Make even when update-webkit used --ninja
1471         https://bugs.webkit.org/show_bug.cgi?id=104839
1472
1473         Reviewed by Tony Chang.
1474
1475         This was recommended by Peter as part of:
1476         https://bugs.webkit.org/show_bug.cgi?id=104434#c4
1477         Fixing this makes build-webkit --chromium-android work for the common (ninja) case.
1478
1479         * Scripts/webkitdirs.pm:
1480         (buildChromium):
1481
1482 2012-12-13  Julie Parent  <jparent@chromium.org>
1483
1484         Loader_unittests should reset state between runs.
1485         https://bugs.webkit.org/show_bug.cgi?id=104862
1486
1487         Reviewed by Dirk Pranke.
1488
1489         The loader_unittests were not resetting state between runs, leaving
1490         a lot of data behind in global objects that the dashboard and tests
1491         rely on.  A handful were resetting just a few objects.  Use the common
1492         resetGlobals() method instead.
1493
1494         * TestResultServer/static-dashboards/loader_unittests.js:
1495
1496 2012-12-13  Andras Becsi  <andras.becsi@digia.com>
1497
1498         [Qt][WK2] Fix painting on Mac with retina display
1499         https://bugs.webkit.org/show_bug.cgi?id=104574
1500
1501         Reviewed by Kenneth Rohde Christiansen.
1502
1503         Remove setting the devicePixelRatio experimental property
1504         since the value is now automatically picked up from Qt.
1505
1506         * MiniBrowser/qt/qml/BrowserWindow.qml:
1507
1508 2012-12-13  Jussi Kukkonen  <jussi.kukkonen@intel.com>
1509
1510         [EFL][GTK] Don't call deprecated g_type_init when glib => 2.35
1511         https://bugs.webkit.org/show_bug.cgi?id=103209
1512
1513         Reviewed by Kenneth Rohde Christiansen.
1514
1515         g_type_init() is deprecated from 2.35.0 onwards. Don't call it
1516         in that case to avoid warnings and/or build failure.
1517
1518         * DumpRenderTree/gtk/ImageDiff.cpp:
1519         (main):
1520
1521 2012-12-13  Claudio Saavedra  <csaavedra@igalia.com>
1522
1523         [Tools] Add exception for gtk_style_context_get()
1524         https://bugs.webkit.org/show_bug.cgi?id=104896
1525
1526         Reviewed by Martin Robinson.
1527
1528         Add new exception for gtk_style_context_get().
1529
1530         * Scripts/webkitpy/style/checkers/cpp.py:
1531         (check_for_null): Add exception for gtk_style_context_get().
1532         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1533         (WebKitStyleTest.test_null_false_zero): Add a unit
1534         test for the new exception.
1535
1536 2012-12-13  Joaquim Rocha  <jrocha@igalia.com>
1537
1538         REGRESSION (r137432): The /webkit2/WebKitCookieManager/accept-policy unit test is failing
1539         https://bugs.webkit.org/show_bug.cgi?id=104790
1540
1541         Reviewed by Carlos Garcia Campos.
1542
1543         Unskip the GTK accept policy test.
1544
1545         * Scripts/run-gtk-tests:
1546         (TestRunner):
1547
1548 2012-12-13  Xingnan Wang  <xingnan.wang@intel.com>
1549
1550         Unreviewed. Adding myself to committers list.
1551
1552         * Scripts/webkitpy/common/config/committers.py:
1553
1554 2012-12-12  Roger Fong  <roger_fong@apple.com>
1555
1556         Unreviewed. Fix typo in NRWT.
1557
1558         * Scripts/webkitpy/layout_tests/port/base.py:
1559         (Port.is_reference_html_file):
1560
1561 2012-12-12  Roger Fong  <roger_fong@apple.com>
1562
1563         Fix detection of ref test files in old and new run webkit tests.
1564         https://bugs.webkit.org/show_bug.cgi?id=104850.
1565
1566         Reviewed by Timothy Horton.
1567
1568         * Scripts/old-run-webkit-tests:
1569         (isUsedInReftest): 
1570         Add notref and ref prefix and suffix cases to isUsedinReftest.
1571         * Scripts/webkitpy/layout_tests/port/base.py:
1572         (Port):
1573         (Port.is_reference_html_file): 
1574         Typo in NRWT.
1575
1576 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
1577
1578         rework ChromiumAndroidDriver.start() et. al to avoid funky layering
1579         https://bugs.webkit.org/show_bug.cgi?id=104856
1580
1581         Reviewed by Eric Seidel.
1582
1583         r137523 / bug 10357 introduced some back-and-forth hopping during
1584         driver.start() in the chromium android port. This patch cleans it
1585         up a bit.
1586
1587         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1588         (ChromiumAndroidDriver.__init__):
1589         (ChromiumAndroidDriver._start_once):
1590         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1591         * Scripts/webkitpy/layout_tests/port/driver.py:
1592         (Driver.start):
1593         (Driver._start):
1594
1595 2012-12-12  Eric Seidel  <eric@webkit.org>
1596
1597         Add support for run-perf-tests --chromium-android --profile
1598         https://bugs.webkit.org/show_bug.cgi?id=103587
1599
1600         Reviewed by Dirk Pranke.
1601
1602         This makes it possible to profile PerformanceTests on an Android device
1603         using run-perf-tests --chromium-android --profile.
1604         CPU profiling on Android is sadly non-trivial, but this goes a long
1605         ways towards making it more accessible to non-Android developers.
1606
1607         Although this script requires your device to be flashed with a
1608         userdebug build (an annoyingly non-trivial endeavor for external contributors)
1609         it endeavors to detect and explain the common errors encountered
1610         while trying to setup a device for perf-based profiling.
1611
1612         * Scripts/webkitpy/common/system/profiler.py:
1613         (GooglePProf.profile_after_exit): Discovered that pprof will just hang if passed a missing file.
1614         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1615         (AndroidPerf):
1616         (AndroidPerf.__init__):
1617         (AndroidPerf.check_configuration):
1618         (AndroidPerf.print_setup_instructions):
1619         (_file_exists_on_device):
1620         (_run_adb_command):
1621         (attach_to_pid):
1622         (_perf_version_string):
1623         (_find_perfhost_binary):
1624         (_perfhost_path):
1625         (_first_ten_lines_of_profile):
1626         (profile_after_exit):
1627         (ChromiumAndroidDriver.__init__):
1628         (ChromiumAndroidDriver.__init__.methods):
1629         (ChromiumAndroidDriver._find_or_create_symfs):
1630         (ChromiumAndroidDriver._start_once):
1631         (ChromiumAndroidDriver):
1632         (ChromiumAndroidDriver._run_post_start_tasks):
1633         (ChromiumAndroidDriver._pid_from_android_ps_output):
1634         (ChromiumAndroidDriver._pid_on_target):
1635         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1636         (ChromiumAndroidDriverTest.test_pid_from_android_ps_output):
1637         (AndroidPerfTest):
1638         (AndroidPerfTest.test_perf_output_regexp):
1639         * Scripts/webkitpy/layout_tests/port/driver.py:
1640         (Driver._setup_environ_for_driver):
1641         (Driver):
1642         (Driver._start):
1643         (Driver._run_post_start_tasks):
1644         (Driver._pid_on_target):
1645
1646 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
1647
1648         nrwt: remove deprecated ORWT compatibility command line flags
1649         https://bugs.webkit.org/show_bug.cgi?id=104301
1650
1651         Reviewed by Eric Seidel.
1652
1653         The command line flags for --noshow-results, --[no-]launch-safari,
1654         and --use-remote-links-to-tests were for compatibility with
1655         old-run-webkit-tests, but the compatibility is not needed any more
1656         since NRWT is basically default and the scripts that passed the
1657         old flags have been updated.
1658
1659         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1660         (_set_up_derived_options):
1661         (parse_args):
1662
1663 2012-12-12  John Griggs  <jgriggs@rim.com>
1664
1665         2012-12-12  John Griggs  <jgriggs@rim.com>
1666
1667         [BlackBerry] Update Media Controls
1668         https://bugs.webkit.org/show_bug.cgi?id=104443
1669         https://przilla.ott.qnx.com/bugzilla/show_bug.cgi?id=204748
1670
1671         Update Media Controls for BlackBerry platform to allow audio, embedded video and fullscreen video controls to be positioned and styled differently.
1672
1673         Reviewed by Rob Buis.
1674
1675         BlackBerry-specific update to Media Controls - No new tests required.
1676
1677         * Scripts/webkitperl/FeatureList.pm:
1678
1679 2012-12-12  Justin Novosad  <junov@google.com>
1680
1681         Adding myself to SkiaGraphics watchlist
1682
1683         Unreviewed
1684
1685         * Scripts/webkitpy/common/config/watchlist:
1686
1687 2012-12-12  Martin Robinson  <mrobinson@igalia.com>
1688
1689         [GTK] Add authentication support to DRT and fix exposed issues in the libsoup backend
1690         https://bugs.webkit.org/show_bug.cgi?id=104285
1691
1692         Reviewed by Gustavo Noronha Silva.
1693
1694         Add support to DumpRenderTree for running authentication tests. Since the DRT
1695         expects an authentication callback, we add one to DRTSupport to avoid #ifdefs
1696         in platform-independent code for GTK+.
1697
1698         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1699         (resetDefaultsToConsistentValues): Reset the authentication password and username.
1700         (authenticationCallback): Added.
1701         (createWebView): Attach the authentiation callback during startup.
1702         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
1703         (soupURIToKURL): soup_uri_to_string does not preserve passwords embedded
1704         in the URL so we add a somewhat messy method of re-adding them when they exist.
1705         It would be nice to use soupURIToKURL here, but it seems we cannot use KURL without
1706         pulling in lots of WebCore code so we use string search and replace.
1707         (TestRunner::queueLoad): Use the new helper.
1708
1709 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
1710
1711         garden-o-matic should prefer efl/ over efl-wk1/ and efl-wk2/ when rebaselining
1712         https://bugs.webkit.org/show_bug.cgi?id=102389
1713
1714         Reviewed by Tony Chang.
1715
1716         The code wasn't aware of both the efl-wk1 and efl-wk2 variants,
1717         because all_port_names() by itself doesn't tell you about the
1718         variants. Work around this for now, and filed bug 104761 to
1719         address the underlying problem.
1720
1721         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
1722         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
1723         (BaselineOptimizerTest.test_efl):
1724
1725 2012-12-12  Krzysztof Czech  <k.czech@samsung.com>
1726
1727         [EFL] Possibility to turn off accessibility feature for WebKit-EFL.
1728         https://bugs.webkit.org/show_bug.cgi?id=103036
1729
1730         Reviewed by Laszlo Gombos.
1731
1732         Guard dependencies for accessibility (ATK), so that they can be turned off.
1733
1734         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
1735         (AccessibilityUIElement):
1736         * WebKitTestRunner/PlatformEfl.cmake:
1737
1738 2012-12-12  Simon Fraser  <simon.fraser@apple.com>
1739
1740         Add some entries in the watchlist, and register me for them.
1741
1742         * Scripts/webkitpy/common/config/watchlist:
1743
1744 2012-12-12  Ilya Tikhonovsky  <loislo@chromium.org>
1745
1746         Web Inspector: Native Memory Instrumentation: remove fake root MemoryObjectInfo.
1747         https://bugs.webkit.org/show_bug.cgi?id=104796
1748
1749         Reviewed by Yury Semikhatsky.
1750
1751         It was not a good idea to introduce a fake root MemoryObjectInfo.
1752         It makes a problem when we visit an object without its own MemoryObjectType.
1753
1754         Example: RenderBox has a global pointer to a hash map.
1755         HashMap doesn't have its own object type because it is a generic container.
1756         It will inherit object type from the fake root memory object info.
1757         The same could happen for another container in another class with other MemoryObjectType.
1758
1759         This fact forces me to create custom process method for root objects
1760         because they need to have their own MemoryObjectInfo with customisable memory object type.
1761
1762         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1763
1764         Drive by fix: InstrumentedPointer* was replaced with Wrapper* because actually it is using
1765         for instrumented and not instrumented object classes.
1766
1767 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
1768
1769         [GTK] Remove the last of unnecessary configuration options in configure.ac
1770         https://bugs.webkit.org/show_bug.cgi?id=104793
1771
1772         Reviewed by Martin Robinson.
1773
1774         Removing the filters and indexed database features from the list of features
1775         being configurable through the configure script. This is not the case anymore
1776         as the corresponding configuration options are being removed.
1777
1778         * Scripts/webkitdirs.pm:
1779         (buildAutotoolsProject):
1780
1781 2012-12-12  Mario Sanchez Prada  <mario@webkit.org>
1782
1783         [GTK][WK2] New API to detect display/execution of insecure content
1784         https://bugs.webkit.org/show_bug.cgi?id=104578
1785
1786         Reviewed by Martin Robinson.
1787
1788         Add coding style exception for 'readability/enum_casing' in GTK.
1789
1790         The GTK+ API use upper case, underscore separated, words in
1791         certain types of enums (e.g. signals, properties).
1792
1793         * Scripts/webkitpy/style/checker.py: Added exception, both for
1794         WebKitGTK+ and WebKit2GTK+.
1795
1796 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
1797
1798         [chromium] Fix build on chromium-win
1799
1800         Unreviewed build-fix.
1801
1802         MSVS indeed doesn't have snprintf, adding wtf/StringExtras.h include
1803         to work around that.
1804
1805         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
1806
1807 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
1808
1809         [Qt] Fix build on Mac
1810
1811         Reviewed by Csaba Osztrogonác.
1812
1813         On Mac OS X we have sqlite as a system library available.
1814
1815         * qmake/mkspecs/features/features.prf:
1816
1817 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
1818
1819         Unreviewed GTK gardening.
1820
1821         Skipping another unit test that regressed in r137432.
1822
1823         * Scripts/run-gtk-tests:
1824         (TestRunner):
1825
1826 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
1827
1828         [Qt] Fix linkage against SQLite
1829         https://bugs.webkit.org/show_bug.cgi?id=104781
1830
1831         Reviewed by Csaba Osztrogonác.
1832
1833         Detect SQLite either through pkg-config, from the SQLITE3SRCDIR
1834         environment variable or from qtbase via a qt5.git build. Bail out early
1835         with an error message if neither option applies.
1836
1837         This removes the usage of $$QT.core.sources and the dependency on Qt's
1838         system-sqlite configure setting. We should always favour a system
1839         library over a copy in Qt.
1840
1841         * qmake/mkspecs/features/features.prf:
1842
1843 2012-12-12  Tommy Widenflycht  <tommyw@google.com>
1844
1845         MediaStream API: Change the data channel descriptor pattern to a handler pattern
1846         https://bugs.webkit.org/show_bug.cgi?id=104543
1847
1848         Reviewed by Kent Tamura.
1849
1850         Adding mocks to be able to test this refactoring.
1851
1852         * DumpRenderTree/DumpRenderTree.gypi:
1853         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.cpp: Added.
1854         (DataChannelReadyStateTask):
1855         (DataChannelReadyStateTask::DataChannelReadyStateTask):
1856         (MockWebRTCDataChannelHandler::MockWebRTCDataChannelHandler):
1857         (MockWebRTCDataChannelHandler::setClient):
1858         (MockWebRTCDataChannelHandler::bufferedAmount):
1859         (MockWebRTCDataChannelHandler::sendStringData):
1860         (MockWebRTCDataChannelHandler::sendRawData):
1861         (MockWebRTCDataChannelHandler::close):
1862         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.h: Added.
1863         (MockWebRTCDataChannelHandler):
1864         (MockWebRTCDataChannelHandler::taskList):
1865         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
1866         (RTCPeerConnectionReadyStateTask::RTCPeerConnectionReadyStateTask):
1867         (RemoteDataChannelTask::RemoteDataChannelTask):
1868         (MockWebRTCPeerConnectionHandler::createDataChannel):
1869         (MockWebRTCPeerConnectionHandler::stop):
1870         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
1871         (MockWebRTCPeerConnectionHandler):
1872
1873 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
1874
1875         [chromium] WebTestPlugin must not depend on WTF::String
1876         https://bugs.webkit.org/show_bug.cgi?id=104783
1877
1878         Reviewed by Kent Tamura.
1879
1880         In a component build, we can't access the WTF that is part of the
1881         WebKit component. Accessing WTF instead will pull in a separate copy
1882         into the TestRunner library. Instead of WTF::String::number use
1883         snprintf() to print numbers.
1884
1885         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
1886
1887 2012-12-12  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
1888
1889         [Qt] Do not automatically enable force_static_libs_as_shared when using debug_and_release
1890         https://bugs.webkit.org/show_bug.cgi?id=104562
1891
1892         Reviewed by Tor Arne Vestbø.
1893
1894         force_static_libs_as_shared creates issues on Mac if QtWebKit is built as a framework.
1895         QtWebProcess, for example, will link to the framework's release library and will rely
1896         on DYLD_IMAGE_SUFFIX to switch to the debug build. Since QtWebProcess will also links
1897         directly to the debug build of WebCore, and that QtWebKit release will link in the
1898         release build of WebCore, both the debug and release version will be loaded.
1899
1900         To fix this properly we should make sure that applications always link to the release
1901         build of dependent libraries by bypassing resolveFinalLibraryName.
1902         Disable it for now as force_static_libs_as_shared is mainly for development while
1903         debug_and_release is mainly for production builds.
1904
1905         * qmake/mkspecs/features/default_post.prf:
1906
1907 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
1908
1909         Unreviewed GTK gardening of the unit tests.
1910
1911         Skip two unit tests that regressed during the hackfest.
1912
1913         * Scripts/run-gtk-tests:
1914         (TestRunner):
1915
1916 2012-12-11  Ryosuke Niwa  <rniwa@webkit.org>
1917
1918         Add Koji Ishii to the list of contributors.
1919
1920         * Scripts/webkitpy/common/config/committers.py:
1921
1922 2012-12-11  Dominic Mazzoni  <dmazzoni@google.com>
1923
1924         AX: Make isActionSupported cross-platform.
1925         https://bugs.webkit.org/show_bug.cgi?id=104681
1926
1927         Reviewed by Chris Fleizach.
1928
1929         Replace isActionSupported, which takes a platform-specific
1930         argument, with three cross-platform methods.
1931
1932         * DumpRenderTree/AccessibilityUIElement.cpp:
1933         (isPressActionSupportedCallback):
1934         (isIncrementActionSupportedCallback):
1935         (isDecrementActionSupportedCallback):
1936         (AccessibilityUIElement::getJSClass):
1937         * DumpRenderTree/AccessibilityUIElement.h:
1938         (AccessibilityUIElement):
1939         * DumpRenderTree/blackberry/AccessibilityUIElementBlackBerry.cpp:
1940         (AccessibilityUIElement::isPressActionSupported):
1941         (AccessibilityUIElement::isIncrementActionSupported):
1942         (AccessibilityUIElement::isDecrementActionSupported):
1943         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
1944         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
1945         (WebTestRunner::AccessibilityUIElement::isPressActionSupportedCallback):
1946         (WebTestRunner):
1947         (WebTestRunner::AccessibilityUIElement::isIncrementActionSupportedCallback):
1948         (WebTestRunner::AccessibilityUIElement::isDecrementActionSupportedCallback):
1949         (WebTestRunner::AccessibilityUIElement::incrementCallback):
1950         (WebTestRunner::AccessibilityUIElement::decrementCallback):
1951         (WebTestRunner::AccessibilityUIElement::pressCallback):
1952         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
1953         (AccessibilityUIElement):
1954         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
1955         (AccessibilityUIElement::isPressActionSupported):
1956         (AccessibilityUIElement::isIncrementActionSupported):
1957         (AccessibilityUIElement::isDecrementActionSupported):
1958         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
1959         (AccessibilityUIElement::isPressActionSupported):
1960         (AccessibilityUIElement::isIncrementActionSupported):
1961         (AccessibilityUIElement::isDecrementActionSupported):
1962         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
1963         (AccessibilityUIElement::isPressActionSupported):
1964         (AccessibilityUIElement::isIncrementActionSupported):
1965         (AccessibilityUIElement::isDecrementActionSupported):
1966         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
1967         (WTR::AccessibilityUIElement::isPressActionSupported):
1968         (WTR::AccessibilityUIElement::isIncrementActionSupported):
1969         (WTR::AccessibilityUIElement::isDecrementActionSupported):
1970         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
1971         (AccessibilityUIElement):
1972         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
1973         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
1974         (WTR::AccessibilityUIElement::isPressActionSupported):
1975         (WTR):
1976         (WTR::AccessibilityUIElement::isIncrementActionSupported):
1977         (WTR::AccessibilityUIElement::isDecrementActionSupported):
1978
1979 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
1980
1981         nrwt: remove result_summary.total_tests_by_expectation
1982         https://bugs.webkit.org/show_bug.cgi?id=104756
1983
1984         Reviewed by Eric Seidel.
1985
1986         This field wasn't being used by anything.
1987
1988         * Scripts/webkitpy/layout_tests/models/result_summary.py:
1989         (ResultSummary.__init__):
1990
1991 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
1992
1993         garden-o-matic doesn't know about reftests
1994         https://bugs.webkit.org/show_bug.cgi?id=101976
1995
1996         Reviewed by Ojan Vafai.
1997
1998         This is a first attempt at fixing this, sort of a "stop-the-bleeding"
1999         fix. We will now look at the list of tests we're asked to rebaseline,
2000         and log which ones are reftests into the status area, and rebaseline
2001         the rest.
2002
2003         Longer-term, we should display feedback in the results area based on
2004         which tests we're looking at. If the test list contains no reftests,
2005         we should behave as today; if the list is all reftest, we should
2006         probably disable the button, and if the list is a mix, we should
2007         probably display an alert next to the button or something. It would
2008         also be nice to annotate which tests were reftests, maybe with a
2009         "(reftest)" next to the test name or something.
2010
2011         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2012         (.):
2013         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2014         (.):
2015         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2016
2017 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2018
2019         REGRESSION(r133380): new tests without expected file reported as failing tests on the bots
2020         https://bugs.webkit.org/show_bug.cgi?id=101469
2021
2022         Reviewed by Ryosuke Niwa.
2023
2024         Fix parsing of NRWT's output to identify missing results separately
2025         from other regressions.
2026
2027         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2028         (RunWebKitTests._parseNewRunWebKitTestsOutput):
2029         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2030         (MasterCfgTest.test_nrwt_leaks_parsing):
2031         (test_nrwt_missing_results):
2032
2033 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2034
2035         buildbot: change run-webkit-tests --no-launch-safari to --no-show-results
2036         https://bugs.webkit.org/show_bug.cgi?id=104299
2037
2038         Reviewed by Ryosuke Niwa.
2039
2040         Since --no-show-results is the more generically-named flag that NRWT uses.
2041
2042         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2043         (RunWebKitTests):
2044
2045 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2046
2047         remove --use-remote-links-to-tests from buildbot configs (except apple win)
2048         https://bugs.webkit.org/show_bug.cgi?id=104291
2049
2050         Reviewed by Csaba Osztrogonác.
2051
2052         This flag is ignored by NRWT and was just implemented for compatibility.
2053         ORWT still uses it, though.
2054
2055         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2056         (RunWebKitTests):
2057         (RunWebKitTests.start):
2058
2059 2012-12-11  Anders Carlsson  <andersca@apple.com>
2060
2061         ASSERT running platform/mac/plugins/convert-point.html on the WK2 bots
2062         https://bugs.webkit.org/show_bug.cgi?id=104735
2063
2064         Reviewed by Jon Honeycutt.
2065
2066         Do the testing from within NPP_New instead of NPN_SetWindow. Plug-ins aren't supposed to do any real work inside NPP_SetWindow,
2067         and doing so will more easily trigger the assertion failure. Ideally we should get rid of the initial sync painting pass, but this
2068         gets the test passing again!
2069
2070         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:
2071         (ConvertPoint::NPP_New):
2072
2073 2012-12-11  Julie Parent  <jparent@chromium.org>
2074
2075         Dashboard unit tests should not make xhrs.
2076         https://bugs.webkit.org/show_bug.cgi?id=104598
2077
2078         Reviewed by Dirk Pranke.
2079
2080         The unit test were requesting 25 files that they did not use.
2081         Stub out loader.request to avoid making the xhrs.
2082         Result: order of magnitude speed improvement in test run time (from ~1600ms to 160ms on my machine).
2083         
2084         * TestResultServer/static-dashboards/run-embedded-unittests.html:
2085         * TestResultServer/static-dashboards/run-unittests.html:
2086
2087 2012-12-11  Filip Pizlo  <fpizlo@apple.com>
2088
2089         Profiler should show bytecode dumps as they would have been visible to the JITs, including the profiling data that the JITs would see
2090         https://bugs.webkit.org/show_bug.cgi?id=104647
2091
2092         Reviewed by Oliver Hunt.
2093
2094         Added a "profiling" (or "p") command to show the profiling data that the JITs saw
2095         for each JIT compilation of a code block.
2096         
2097         Also added instruction counts in the "full" display and made the "full" display the
2098         default thing you see.
2099
2100         * Scripts/display-profiler-output:
2101
2102 2012-12-11  Eric Seidel  <eric@webkit.org>
2103
2104         Unreviewed, rolling out r137371.
2105         http://trac.webkit.org/changeset/137371
2106         https://bugs.webkit.org/show_bug.cgi?id=104434
2107
2108         Various scripts are not ready for out/ as the build directory,
2109         this can't work as written.
2110
2111         * Scripts/update-webkit:
2112         * Scripts/webkitdirs.pm:
2113         (buildChromium):
2114
2115 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
2116
2117         Unreviewed GTK build functionality fix.
2118
2119         Turns out the CSP feature was not enabled, that's why all the tests were failing.
2120
2121         * Scripts/webkitperl/FeatureList.pm:
2122
2123 2012-12-11  Eric Seidel  <eric@webkit.org>
2124
2125         Make Ninja the default build system for build-webkit --chromium
2126         https://bugs.webkit.org/show_bug.cgi?id=104434
2127
2128         Reviewed by Daniel Bates.
2129
2130         As far as I can tell, most Chromium developers are now using Ninja
2131         as their default build system instead of the native Xcode/Visual Studio/Make.
2132         This change makes Ninja the default for developers as well as all Chromium webkit bots.
2133
2134         Developers can pass --no-ninja to get the platform-native buildsystem instead of Ninja.
2135         It's possible we may wish to make this a per-platform default, and have
2136         Ninja only be the default for Mac/Linux.
2137
2138         There was a thread on webkit-dev discussing this change:
2139         http://lists.webkit.org/pipermail/webkit-dev/2012-December/023113.html
2140
2141         * Scripts/update-webkit:
2142
2143 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2144
2145         nrwt: record the order and run times of all of the tests in a single file
2146         https://bugs.webkit.org/show_bug.cgi?id=104184
2147
2148         Reviewed by Ryosuke Niwa.
2149
2150         Replace the N tests_run*.txt files that we output into
2151         layout-test-results with a single packed JSON trie that contains
2152         the test ordering and run statistics. This can eventually be uploaded
2153         to the TestResultsServer to replace the times.json file we currently
2154         upload as well, and will let us replay the results of a previous
2155         run fairly deterministically.
2156
2157         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2158         (Worker.__init__):
2159         (Worker.start):
2160         (Worker._run_test):
2161         (Worker.stop):
2162         (Worker._clean_up_after_test):
2163         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2164         (Manager._upload_json_files):
2165         (Manager._copy_results_html_file):
2166         (Manager):
2167         (Manager._stats_trie):
2168         (Manager._stats_trie._worker_number):
2169         * Scripts/webkitpy/layout_tests/models/test_results.py:
2170         (TestResult.__init__):
2171         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2172         (MainTest.test_missing_and_unexpected_results):
2173         (MainTest.test_retrying_and_flaky_tests):
2174         (EndToEndTest.test_end_to_end):
2175         (EndToEndTest.test_reftest_with_two_notrefs):
2176         (RebaselineTest.test_reset_results):
2177         (RebaselineTest.test_missing_results):
2178         (RebaselineTest.test_new_baseline):
2179
2180 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2181
2182         check-webkit-style complaining about TestExpectations lines a patch didn't modify
2183         https://bugs.webkit.org/show_bug.cgi?id=104687
2184
2185         Reviewed by Ojan Vafai.
2186
2187         Way back in bug 60466 / r86165, we changed the style checker
2188         for TestExpectations to ensure that the entire file linted, not
2189         just the lines modified by a patch, in order to ensure that a
2190         new line didn't cause errors elsewhere in the file (e.g., by
2191         introducing duplicate lines). At the time, errors in the
2192         expectations files were fatal and we didn't have a separate step
2193         on the bots that would lint the files to ensure things were still
2194         okay, so this was probably the right thing to do, but it could
2195         cause some patches to fail the style check even though the patches
2196         themselves were fine.
2197
2198         However, now errors aren't always fatal, and we have a separate build
2199         step to lint the files. So, we should probably not punish the
2200         patches and the style checker should filter out "unrelated" errors.
2201
2202         In theory we should be comparing the output pre- and post-patch
2203         to filter out pre-existing errors, but I don't think the style
2204         checker has that ability.
2205
2206         * Scripts/webkitpy/style/checkers/test_expectations.py:
2207         (TestExpectationsChecker.__init__):
2208         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2209         (TestExpectationsTestCase.assert_lines_lint):
2210
2211 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2212
2213         nrwt isn't rejecting unrecognized expectations
2214         https://bugs.webkit.org/show_bug.cgi?id=104701
2215
2216         Reviewed by Ojan Vafai.
2217
2218         NRWT is silently ignoring unrecognized expectations like "Text"
2219         and "Audio" (largely because I haven't fully removed support for
2220         the old syntax and keywords and restructured the code to do the
2221         proper checking). Temporarily change the tokenizer to reject
2222         unrecognized expectations until I can make those changes.
2223
2224         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2225         (TestExpectationParser._tokenize_line):
2226         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2227         (ExpectationSyntaxTests.test_unknown_expectation):
2228         (TestExpectationSerializationTests.disabled_test_string_whitespace_stripping):
2229
2230 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2231
2232         Fix a typo in my previous commit that is causing test failures.
2233
2234         Reviewed by Gustavo Noronha Silva.
2235
2236         There was an extra bit of statement appended after the end of a conditional. Somehow
2237         this compiled and caused test failures.
2238
2239         * DumpRenderTree/gtk/EventSender.cpp:
2240         (continuousMouseScrollByCallback):
2241
2242 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2243
2244         [GTK] Spurious stderr output from the event sender
2245         https://bugs.webkit.org/show_bug.cgi?id=104678
2246
2247         Reviewed by Gustavo Noronha Silva.
2248
2249         Eliminate some spurious stderr output from the EventSender by using an early return
2250         instead of an assertion that prints a warning to the console.
2251
2252         * DumpRenderTree/gtk/EventSender.cpp:
2253         (continuousMouseScrollByCallback): We don't support continuous scroll events that
2254         move by page, but we should just fail silently instead of printing an error.
2255
2256 2012-12-11  Joone Hur  <joone.hur@intel.com>
2257
2258         [GTK] Add GraphicsLayerActor
2259         https://bugs.webkit.org/show_bug.cgi?id=91940
2260
2261         Reviewed by Gustavo Noronha Silva.
2262
2263         Add GraphicsLayerActor to the list of classes that contain GObjects
2264         in the style checker script so that it does not complain about GObject
2265         coding style.
2266
2267         * Scripts/webkitpy/style/checker.py:
2268
2269 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
2270
2271         [Qt] remove hack which avoids unused .obj directories all over the place
2272
2273         Reviewed by Simon Hausmann.
2274
2275         The feature is subject to upstreaming in Qt itself in
2276         https://codereview.qt-project.org/#change,42011
2277
2278         * qmake/mkspecs/features/default_post.prf:
2279
2280 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
2281
2282         [Qt] move SQLITE3SRCDIR assignment to only use site
2283
2284         Reviewed by Simon Hausmann.
2285
2286         Target.pri includes WebCore.pri the almost first thing, so that is
2287         covered as well.
2288
2289         * qmake/mkspecs/features/default_pre.prf:
2290
2291 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
2292
2293         [Qt] delete dead include paths
2294
2295         Reviewed by Simon Hausmann.
2296
2297         followup to https://bugs.webkit.org/show_bug.cgi?id=93446
2298
2299         * DumpRenderTree/qt/DumpRenderTree.pro:
2300
2301 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2302
2303         [GTK] Implement some missing FrameLoaderClient message in DRT
2304         https://bugs.webkit.org/show_bug.cgi?id=104444
2305
2306         Reviewed by Gustavo Noronha Silva.
2307
2308         Install a FrameLoadEvent callback that prints the necessary output for some
2309         missing FrameLoader events.
2310
2311         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2312         (frameLoadEventCallback): Added.
2313         (createWebView): Install the FrameLoader event callback.
2314
2315 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
2316
2317         [Qt] change setup of QMAKE_MOC
2318
2319         Reviewed by Simon Hausmann.
2320
2321         don't load(moc) explicitly - turns out that this breaks
2322         debug_and_release builds (the feature does not like being loaded before
2323         default_post has been loaded).
2324         so instead just do a minimal call to find the moc executable.
2325
2326         as it's in fact only two files which need it, so de-centralize this call
2327         for the sake of saving some cpu cycles spent in qmake.
2328
2329         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
2330         * qmake/mkspecs/features/default_pre.prf:
2331
2332 2012-12-11  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
2333
2334         [Qt] Fix the linking of debug builds on Mac with force_static_libs_as_shared
2335         https://bugs.webkit.org/show_bug.cgi?id=104559
2336
2337         Reviewed by Tor Arne Vestbø.
2338
2339         "contains(libraries, $$TARGET): return(true)" fails for intermediate libraries since
2340         at this point TARGET may contain the "_debug" suffix.
2341
2342         Fix it by moving the TARGET adjustment later in the file.
2343
2344         * qmake/mkspecs/features/default_post.prf:
2345
2346 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2347
2348         [Coverity] [GTK] Remove some redundant null checks in DumpRenderTree
2349         https://bugs.webkit.org/show_bug.cgi?id=104570
2350
2351         Reviewed by Daniel Bates.
2352
2353         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2354         (descriptionSuitableForTestResult): Remove a redundant null check.
2355
2356 2012-12-11  Alberto Garcia  <agarcia@igalia.com>
2357
2358         Remove unused GErrors
2359         https://bugs.webkit.org/show_bug.cgi?id=104657
2360
2361         Reviewed by Martin Robinson.
2362
2363         Replace unused GErrors with null pointers.
2364
2365         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2366         (initializeFonts):
2367         * DumpRenderTree/gtk/EditingCallbacks.cpp:
2368         (dumpRange):
2369         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
2370         (WTR::inititializeFontConfigSetting):
2371
2372 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2373
2374         [GTK] Bump the jhbuild libsoup version to fix an authentication issue
2375         https://bugs.webkit.org/show_bug.cgi?id=104448
2376
2377         Reviewed by Gustavo Noronha Silva.
2378
2379         Bump the libsoup version to one from the git repository. This requires updating
2380         GLib as well which in turn requires adjusting the CFLAGS we pass when compiling
2381         the scanobj step of the gtkdoc generation.
2382
2383         * gtk/generate-gtkdoc:
2384         * gtk/jhbuild.modules:
2385
2386 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
2387
2388         Unreviewed functionality fix for the GTK port.
2389
2390         Enable the MicroData feature support for the GTK port in the FeatureList module.
2391         This feature listing now again affects which features are enabled and which disabled
2392         on the development builds of the GTK port.
2393
2394         * Scripts/webkitperl/FeatureList.pm:
2395
2396 2012-12-11  Jinwoo Song  <jinwoo7.song@samsung.com>
2397
2398         [EFL][WK2] MiniBrowser is not built
2399         https://bugs.webkit.org/show_bug.cgi?id=104629
2400
2401         Reviewed by Gyuyoung Kim.
2402
2403         As we are finding an elementary package in Tools/MiniBrowser/efl/CMakeList.txt since r137203,
2404         we need to check if the package is found in the same place not in the Tools/CMakeList.txt.
2405
2406         * CMakeLists.txt:
2407         * MiniBrowser/efl/CMakeLists.txt:
2408
2409 2012-12-11  Xabier Rodriguez Calvar  <calvaris@igalia.com>
2410
2411         [GTK][jhbuild] Switch to GStreamer 1.0 build
2412         https://bugs.webkit.org/show_bug.cgi?id=91727
2413
2414         Reviewed by Philippe Normand.
2415
2416         Switch build-webkit --gtk to GStreamer 1.0 support and build the
2417         necessary GStreamer git modules from JHBuild.
2418
2419         * gtk/jhbuild.modules: Added GStreamer 1.0.3 build support.
2420
2421 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
2422
2423         [GTK] Feature enabling/disabling should be possible through build-webkit
2424         https://bugs.webkit.org/show_bug.cgi?id=99271
2425
2426         Reviewed by Gustavo Noronha Silva.
2427
2428         Refactor the code in webkitdirs.pm that builds an autotools project to generate
2429         autogen.sh arguments, make arguments and installation prefix from the passed-in
2430         parameters rather than generating all of that in build-webkit. The autogen.sh
2431         arguments now contain only enable/disable flags for options that are actually
2432         configurable in configure.ac, the flag value actually reflecting whether the feature
2433         is enabled or disabled in the feature list.
2434
2435         Other features are overridable through modifying the GNUmakefile.features.am file.
2436         All these features are now stored in the build directory and upon change trigger
2437         a rerun of the autogen.sh script, pretty much like the autogen.sh arguments do.
2438
2439         The override-feature-defines script is called by autogen.sh, before GNUmakefile.in is
2440         generated when calling autoreconf. Its task is to copy the GNUmakefile.features.am.in
2441         into GNUmakefile.features.am and modify it if there's a text file in the build directory
2442         that contains all the feature defines the build-webkit script has written. If the build
2443         is not done through build-webkit (and there's no file in build directory listing all the
2444         feature defines) the script does not advance further from copying.
2445
2446         This approach is taken to overcome the rigidness of the automake system as it's impossible
2447         to effectively generate and use GNUmakefile.features.am or even GNUmakefile.features file
2448         after the autoreconf command execution in autogen.sh.
2449
2450         * Scripts/build-jsc:
2451         (buildMyProject):
2452         * Scripts/build-webkit:
2453         * Scripts/webkitdirs.pm:
2454         (runAutogenForAutotoolsProjectIfNecessary):
2455         (mustReRunAutogen):
2456         (buildAutotoolsProject):
2457         (buildGtkProject):
2458         * Scripts/webkitperl/FeatureList.pm: Update the features that are currently enabled in
2459         Source/WebCore/GNUmakefile.features.am.in but aren't in the feature list.
2460         * Scripts/webkitpy/style/checker.py: Source/WebCore/GNUmakefile.features.am.in is recognized
2461         as a text file and tabulation errors are reported. Skip the file to suppress them.
2462         * gtk/override-feature-defines: Added.
2463         (copy_feature_defines_makefile):
2464         (adjust_feature_defines_makefile):
2465         (adjust_feature_defines_makefile.override_feature_define):
2466         (override_feature_defines):
2467
2468 2012-12-10  Yury Semikhatsky  <yurys@chromium.org>
2469
2470         Memory instrumentation: make sure each edge is reported only once
2471         https://bugs.webkit.org/show_bug.cgi?id=104630
2472
2473         Reviewed by Pavel Feldman.
2474
2475         Test that outgoing edges are reported only once if we come to the same
2476         object twice: first by an address of a base class and then by a real address
2477         of the object.
2478
2479         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2480
2481 2012-12-10  Dana Jansens  <danakj@chromium.org>
2482
2483         [chromium] Remove WebCompositorOutputSurface implementations from DRT
2484         https://bugs.webkit.org/show_bug.cgi?id=104046
2485
2486         Reviewed by Darin Fisher.
2487
2488         The classes are now provided through WebCompositorSupport.
2489
2490         * DumpRenderTree/DumpRenderTree.gypi:
2491         * DumpRenderTree/chromium/WebViewHost.cpp:
2492         (WebViewHost::createOutputSurface):
2493         * DumpRenderTree/chromium/WebViewHostOutputSurface.cpp: Removed.
2494         * DumpRenderTree/chromium/WebViewHostOutputSurface.h: Removed.
2495         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.cpp: Removed.
2496         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.h: Removed.
2497
2498 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2499
2500         change --no-launch-safari to --no-show-results in scripts
2501         https://bugs.webkit.org/show_bug.cgi?id=104298
2502
2503         Reviewed by Eric Seidel.
2504
2505         Since Safari only exists on the Apple ports, NRWT calls the
2506         command line flag --[no-]show-results instead. This updates
2507         the scripts to use the new flag (but not the buildbot config yet).
2508
2509         * Scripts/generate-coverage-data:
2510         * Scripts/webkit-tools-completion.sh:
2511         * Scripts/webkitpy/tool/steps/runtests.py:
2512         (RunTests.run):
2513         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
2514
2515 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2516
2517         nrwt: fix mapping of AUDIO onto FAIL when printing unexpected results
2518         https://bugs.webkit.org/show_bug.cgi?id=104597
2519
2520         Reviewed by Ojan Vafai.
2521
2522         This was broken in the recent printer refactoring.
2523
2524         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
2525         (BuildBotPrinter.print_unexpected_results.add_result.is_expected):
2526         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2527         (BuildBotPrinterTests.get_result):
2528         (BuildBotPrinterTests.test_print_unexpected_results):
2529         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
2530         (BuildBotPrinterTests):
2531
2532 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2533
2534         old-run-webkit-tests: add --show-results properly
2535         https://bugs.webkit.org/show_bug.cgi?id=104608
2536
2537         Reviewed by Eric Seidel.
2538
2539         Rework r137219 / bug 104601 to use the correct perl idiom.
2540
2541         * Scripts/old-run-webkit-tests:
2542
2543 2012-12-10  Aaron Colwell  <acolwell@chromium.org>
2544
2545         Unreviewed. Adding Aaron Colwell to the contributor list.
2546
2547         * Scripts/webkitpy/common/config/committers.py:
2548
2549 2012-12-10  Anders Carlsson  <andersca@apple.com>
2550
2551         Remove WebKit2 tests on Windows
2552         https://bugs.webkit.org/show_bug.cgi?id=104607
2553
2554         Reviewed by Tim Horton.
2555
2556         Stop building WebKit2 tests and remove associated files.
2557
2558         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Removed.
2559         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Removed.
2560         * TestWebKitAPI/win/PlatformWebViewWin.cpp: Removed.
2561         * TestWebKitAPI/win/TestWebKitAPI.sln: Removed.
2562         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
2563         * TestWebKitAPI/win/TestWebKitAPIGenerated.vcproj: Removed.
2564         * TestWebKitAPI/win/TestWebKitAPIGeneratedCommon.vsprops: Removed.
2565         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Removed.
2566         * TestWebKitAPI/win/WindowMessageObserver.h: Removed.
2567         * TestWebKitAPI/win/copy-resources.cmd: Removed.
2568
2569 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2570
2571         old-run-webkit-tests: support --no-show-results for compatibility w/ NRWT
2572         https://bugs.webkit.org/show_bug.cgi?id=104601
2573
2574         Reviewed by Eric Siedel.
2575
2576         This flag is being added so we can move all of the calling scripts
2577         over to using --no-show-results instead of --no-launch-safari.
2578
2579         * Scripts/old-run-webkit-tests:
2580
2581 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2582
2583         nrwt: use the default logic for skipping platform-specific tests for chromium
2584         https://bugs.webkit.org/show_bug.cgi?id=104403
2585
2586         Reviewed by Stephen White.
2587
2588         I thought I had fixed this a while ago, but the chromium port was
2589         using custom logic to figure out which platform/* tests to run (or
2590         skip) and as a result we were getting some odd behavior (like running
2591         tests in chromium-linux on chromium-mac, but only when there was
2592         an expectation listed for the test).
2593
2594         Ports are supposed to only run the tests found under their
2595         baseline_search_path() by default.
2596
2597         Not writing a test for this; theoretically there shouldn't been a
2598         test checking that the behavior for Chromium was *different* than
2599         the other ports before.
2600
2601         * Scripts/webkitpy/layout_tests/port/chromium.py:
2602         (ChromiumPort._missing_symbol_to_skipped_tests):
2603
2604 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2605
2606         fix regression in test-webkitpy after r137189
2607         https://bugs.webkit.org/show_bug.cgi?id=104589
2608
2609         Reviewed by Eric Seidel.
2610
2611         Need to specify full port names.
2612
2613         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2614         (MainTest.test_build_check):
2615
2616 2012-12-10  Laszlo Gombos  <l.gombos@samsung.com>
2617
2618         [EFL] Change the minimum required EFL version to 1.6
2619         https://bugs.webkit.org/show_bug.cgi?id=104431
2620
2621         Reviewed by Kenneth Rohde Christiansen.
2622
2623         Change the minimum required EFL version to 1.6 from 1.7.
2624         The elementary EFL package is only required to build MiniBrowser,
2625         so I moved the required only to the Minibrowser CMake file.
2626
2627         * MiniBrowser/efl/CMakeLists.txt: In addition removed
2628         ENABLE_GLIB_SUPPORT as it is obsolete and it was always set to be on.
2629
2630 2012-12-10  Dean Jackson  <dino@apple.com>
2631
2632         Unreviewed. Adding Antoine to the contributor list so his name autocompletes in bugzilla.
2633
2634         * Scripts/webkitpy/common/config/committers.py:
2635
2636 2012-12-10  Zan Dobersek  <zandobersek@gmail.com>
2637
2638         REGRESSION (r118735): svg/dom/complex-svgView-specification.html, svg/dom/SVGViewSpec.html, svg/dom/viewspec-parser.html failing on GTK Linux 64-bit Release
2639         https://bugs.webkit.org/show_bug.cgi?id=87734
2640
2641         Reviewed by Martin Robinson.
2642
2643         When willSendRequest should return null in testing, cancel the SoupMessage
2644         instead of setting the request's URI to about:blank.
2645
2646         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2647         (willSendRequestCallback):
2648
2649 2012-12-10  Jochen Eisinger  <jochen@chromium.org>
2650
2651         [chromium] Move the test implementation of WebKit::WebPlugin to the TestRunner library
2652         https://bugs.webkit.org/show_bug.cgi?id=104537
2653
2654         Reviewed by Tony Chang.
2655
2656         In order to hide implementation details from the public TestRunner API,
2657         I moved the plugin to WebTestPluginImpl and added a static create
2658         method to WebTestPlugin. Also, I renamed the plugin from TestWebPlugin
2659         to WebTestPlugin to be more in line with other classes in the
2660         TestRunner API.
2661
2662         * DumpRenderTree/DumpRenderTree.gypi:
2663         * DumpRenderTree/chromium/TestRunner/public/WebTestPlugin.h: Added.
2664         (WebTestRunner):
2665         (WebTestPlugin):
2666         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Renamed from Tools/DumpRenderTree/chromium/TestWebPlugin.cpp.
2667         (WebTestRunner::WebTestPlugin::create):
2668         (WebTestRunner):
2669         (WebTestRunner::WebTestPlugin::WebTestPlugin):
2670         (WebTestRunner::WebTestPlugin::~WebTestPlugin):
2671         (WebTestRunner::WebTestPlugin::mimeType):
2672         * DumpRenderTree/chromium/TestWebPlugin.h: Removed.
2673         * DumpRenderTree/chromium/WebViewHost.cpp:
2674         (WebViewHost::createPlugin):
2675
2676 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2677
2678         run_webkit_tests throws exception when DRT is not present
2679         https://bugs.webkit.org/show_bug.cgi?id=104565
2680
2681         Reviewed by Jochen Eisinger.
2682
2683         Don't actually try to print the bot results if there aren't
2684         any because we exited before running any tests.
2685
2686         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2687         (main):
2688         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2689         (MainTest.test_build_check):
2690
2691 2012-12-10  Benjamin Poulain  <benjamin@webkit.org>
2692
2693         Add convenience methods to use ListHashSet for a LRU cache
2694         https://bugs.webkit.org/show_bug.cgi?id=104499
2695
2696         Reviewed by Sam Weinig.
2697
2698         Test the new methods added to ListHashSet.
2699
2700         * TestWebKitAPI/CMakeLists.txt:
2701         * TestWebKitAPI/GNUmakefile.am:
2702         * TestWebKitAPI/TestWebKitAPI.gypi:
2703         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2704         * TestWebKitAPI/Tests/WTF/ListHashSet.cpp: Added.
2705         (TestWebKitAPI):
2706         (TestWebKitAPI::TEST):
2707
2708 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2709
2710         webkit-patch print-expectations doesn't work right for platforms w/ shared expectations
2711         https://bugs.webkit.org/show_bug.cgi?id=104405
2712
2713         Reviewed by Eric Seidel.
2714
2715         If multiple ports shared a TestExpectations file, webkit-patch
2716         print-expectations would use the expectation for the first port
2717         using that file and print the result for all ports, rather than
2718         re-parse and print each port separately (e.g., we might print
2719         the chromium-linux expectation for a file for both chromium-linux
2720         and chromium-mac). Discovered while testing/debugging the fix
2721         for bug 104403.
2722
2723         * Scripts/webkitpy/layout_tests/port/test.py:
2724         (TestPort._skipped_tests_for_unsupported_features):
2725         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2726         (lint):
2727         * Scripts/webkitpy/tool/commands/queries.py:
2728         (PrintExpectations._model):
2729         * Scripts/webkitpy/tool/commands/queries_unittest.py:
2730         (PrintExpectationsTest.test_platform):
2731
2732 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
2733
2734         nrwt: fix total number of tests computed in the result summary
2735         https://bugs.webkit.org/show_bug.cgi?id=104061
2736
2737         Reviewed by Jochen Eisinger.
2738
2739         Good thing I left this open, because I actually broke it again
2740         the same way in a merge. Fix the computation of the one_line_summary
2741         again, and add tests for it (also clean up some redundant tests).
2742
2743         Also, add constants into the test port code so that the knowledge
2744         of how many tests should fail and how is next to the list of tests,
2745         rather than a totally different place.
2746
2747         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2748         (LayoutTestRunner.run_tests):
2749         * Scripts/webkitpy/layout_tests/port/test.py:
2750         (TestList.__getitem__):
2751         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2752         (get_test_results):
2753         (MainTest.setUp):
2754         (MainTest.test_basic):
2755         (MainTest.test_run_singly_actually_runs_tests):
2756         (MainTest.test_test_list_with_prefix):
2757         (MainTest.test_missing_and_unexpected_results_with_custom_exit_code.CustomExitCodePort):
2758         (MainTest.test_tolerance.ImageDiffTestPort):
2759         (EndToEndTest.parse_full_results):
2760
2761 2012-12-10  Mike West  <mkwst@chromium.org>
2762
2763         Log to console when script is blocked by sandbox attributes.
2764         https://bugs.webkit.org/show_bug.cgi?id=104365
2765
2766         Reviewed by Ojan Vafai.
2767
2768         * Scripts/webkitpy/performance_tests/perftest.py:
2769         (PerfTest):
2770             Ignoring the new console message in perftests, as it broke parsing
2771             for the Parser/html-parser.html microbenchmark.
2772
2773 2012-12-10  Filip Pizlo  <fpizlo@apple.com>
2774
2775         JSC profiling and debug dump code should use inferred names when possible
2776         https://bugs.webkit.org/show_bug.cgi?id=104519
2777
2778         Reviewed by Oliver Hunt.
2779
2780         The format I'm using for referring to a code block is now name#hash. For example,
2781         v8-crypto has something called bnpSquareTo#B5QFbU. The profiler allows you to use
2782         either the hash, the inferred name, or the combined hash and full name when referring
2783         to blocks.
2784
2785         * Scripts/display-profiler-output:
2786
2787 2012-12-09  Filip Pizlo  <fpizlo@apple.com>
2788
2789         Profiler should say things about OSR exits
2790         https://bugs.webkit.org/show_bug.cgi?id=104497
2791
2792         Reviewed by Oliver Hunt.
2793
2794         Adds support for displaying OSR exit information for full summary (just displays the
2795         counts and the number of recompilations), bytecode display (says which bytecodes
2796         exited), and DFG display (annotates disassembly with taken OSR exits and their
2797         counts).
2798
2799         * Scripts/display-profiler-output:
2800
2801 2012-12-10  Alexis Menard  <alexis@webkit.org>
2802
2803         [CSS3 Backgrounds and Borders] Remove CSS3_BACKGROUND feature flag.
2804         https://bugs.webkit.org/show_bug.cgi?id=104539
2805
2806         Reviewed by Antonio Gomes.
2807
2808         As discussed on webkit-dev it is not needed to keep this feature flag 
2809         as support for <position> type is a small feature that is already 
2810         implemented by three other UAs. It was useful while landing this 
2811         feature as partial bits were landed one after one.
2812
2813         * Scripts/webkitperl/FeatureList.pm:
2814         * qmake/mkspecs/features/features.pri:
2815
2816 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
2817
2818         Build fix.
2819
2820         Add a patchfile missing from my previous commit.
2821
2822         * gtk/patches/harfbuzz-icu-detection-fix.patch: Added.
2823
2824 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
2825
2826         REGRESSION(137127): Causes assertion failures on the debug bots
2827         https://bugs.webkit.org/show_bug.cgi?id=104549
2828
2829         Reviewed by Alejandro G. Castro.
2830
2831         After handling the new TextOutput message we should return the method
2832         handling messages received.
2833
2834         * WebKitTestRunner/TestInvocation.cpp:
2835         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Fix a missing early return.
2836
2837 2012-12-09  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
2838
2839         [Soup] utilize multipart/x-mixed-replace support recently added to libsoup
2840         https://bugs.webkit.org/show_bug.cgi?id=94515
2841
2842         Reviewed by Martin Robinson.
2843
2844         This change requires a very recent libsoup, so update our jhbuild
2845         requirements.
2846
2847         * gtk/jhbuild.modules: depend on soup version that adds the multipart/x-mixed-replace
2848         support
2849         * efl/jhbuild.modules: ditto
2850
2851 2012-12-10  Zoltan Nyul  <zoltan.nyul@intel.com>
2852
2853         Implement testRunner.setViewModeMediaFeature() in WebKitTestRunner
2854         https://bugs.webkit.org/show_bug.cgi?id=103886
2855
2856         Reviewed by Kenneth Rohde Christiansen.
2857
2858         Implement testRunner.setViewModeMediaFeature() in WebKitTestRunner.
2859         This API is required by layout tests.
2860
2861         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2862         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2863         (WTR::TestRunner::setViewModeMediaFeature):
2864         (WTR):
2865         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2866         (TestRunner):
2867
2868 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
2869
2870         [GTK] Bring Harfbuzz-ng support to Gtk
2871         https://bugs.webkit.org/show_bug.cgi?id=92098
2872
2873         Reviewed by Gustavo Noronha Silva.
2874
2875         Add HarfBuzz dependency.
2876
2877         * GNUmakefile.am: Add FreeType/HarfBuzz CFLAGS to the WebCore Internals library.
2878         * gtk/jhbuild.modules: Add HarfBuzz to the JHbuild and remove Pango, since it
2879         should no longer affect layout test results.
2880
2881 2012-12-10  Simon Hausmann  <simon.hausmann@digia.com>
2882
2883         [Qt] Use QLibraryInfo::LibraryExecutablesPath unconditionally
2884         https://bugs.webkit.org/show_bug.cgi?id=104541
2885
2886         Reviewed by Jocelyn Turcotte.
2887
2888         We now depend on a Qt 5 version that is guaranteed to have this API, so we
2889         can remove the configure checks for it.
2890
2891         * qmake/config.tests/libexecdir/libexecdir.cpp: Removed.
2892         * qmake/config.tests/libexecdir/libexecdir.pro: Removed.
2893         * qmake/mkspecs/features/features.prf:
2894
2895 2012-12-10  Víctor Manuel Jáquez Leal  <vjaquez@igalia.com>
2896
2897         REGRESSION(r135863): GtkLauncher --help no longer displays all help options
2898         https://bugs.webkit.org/show_bug.cgi?id=104382
2899
2900         Reviewed by Philippe Normand.
2901
2902         Remove gst_init() because it steals the GOptionContext.
2903         gst_init_get_option_group() already does the initialization job.
2904
2905         * GtkLauncher/main.c:
2906         (main):
2907
2908 2012-12-10  Zan Dobersek  <zandobersek@gmail.com>
2909
2910         [GTK][WTR] Implement AccessibilityUIElement::titleUIElement
2911         https://bugs.webkit.org/show_bug.cgi?id=102953
2912
2913         Reviewed by Martin Robinson.
2914
2915         Implement AccessibilityUIElement::titleUIElement in the ATK backend
2916         in the same manner it is implemented in DumpRenderTree. Both implementations
2917         are updated to unref the relation set before the final function return
2918         rather than at the end of each loop through the relations held in the said
2919         relation set.
2920
2921         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2922         (AccessibilityUIElement::titleUIElement):
2923         * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
2924         (WTR::AccessibilityUIElement::titleUIElement):
2925
2926 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
2927
2928         Fix the WebKit2 debug build
2929
2930         * WebKitTestRunner/TestInvocation.cpp:
2931         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Fix a typo of WKStringGetTypeID().
2932
2933 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
2934
2935         [WTR] Move text output accumulation to the UIProcess
2936         https://bugs.webkit.org/show_bug.cgi?id=104214
2937
2938         Reviewed by Darin Adler.
2939
2940         Instead of accumulating text output in the InjectedBundle and then sending it to the UIProcess 
2941         once a test is finished, immediately send any text output to the UIProcess. This will allow
2942         WebKitTestRunner to output text from the UIProcess as well.
2943
2944         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Remove the handling of the text output
2945         StringBuilder. Add a method to send output to the UIProcess.
2946         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
2947         (InjectedBundle):
2948         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp: Change code that appends to the StringBuilder
2949         to use the new outputText method.
2950         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2951         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Ditto.
2952         * WebKitTestRunner/InjectedBundle/atk/AccessibilityControllerAtk.cpp: Ditto.
2953         * WebKitTestRunner/TestInvocation.cpp:
2954         (WTR::TestInvocation::TestInvocation): Intialize the StringBuilder.
2955         (WTR::TestInvocation::invoke): Clear the StringBuilder when a new tests starts.
2956         (WTR::TestInvocation::dumpResults): m_textOutput is a StringBuilder now so the
2957         method of printing the text is slightly different.
2958         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Handle the new TextOutput
2959         message by appending the results to the StringBuilder.
2960         * WebKitTestRunner/TestInvocation.h:
2961         (TestInvocation): Switch the type of m_textOutput.
2962
2963 2012-12-10  Mihnea Ovidenie  <mihnea@adobe.com>
2964
2965         Unreviewed. Adding Mihai Maerean to the list of contributors, working on CSSRegions.
2966
2967         * Scripts/webkitpy/common/config/committers.py:
2968
2969 2012-12-10  Joanmarie Diggs  <jdiggs@igalia.com>
2970
2971         [GTK] accessibility/language-attribute.html is failing
2972         https://bugs.webkit.org/show_bug.cgi?id=98369
2973
2974         Reviewed by Chris Fleizach.
2975
2976         The test is failing for two reasons: Not exposing the language
2977         for accessible objects and AccessibilityUIElement::language() not
2978         being implemented.
2979
2980         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2981         (AccessibilityUIElement::language): Implemented.
2982
2983 2012-12-10  Joanmarie Diggs  <jdiggs@igalia.com>
2984
2985         [GTK] accessibility/editable-webarea-context-menu-point.html is failing
2986         https://bugs.webkit.org/show_bug.cgi?id=98364
2987
2988         Reviewed by Chris Fleizach.
2989
2990         The test is failing because the clickPoint values remain unchanged. The
2991         values remain unchanged because ATK lacks a means to keep track of this
2992         information.
2993
2994         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2995         (AccessibilityUIElement::clickPointX): Added comment that ATK doesn't have clickPoint details
2996         (AccessibilityUIElement::clickPointY): Added comment that ATK doesn't have clickPoint details
2997
2998 2012-12-09  Alejandro Piñeiro  <apinheiro@igalia.com>
2999
3000         [GTK] lack of implemention of AccessibilityUIElementGtk::isSelectable and AccessibilityUIElementGtk::isMultiSelectable
3001         https://bugs.webkit.org/show_bug.cgi?id=104481
3002
3003         Reviewed by Martin Robinson.
3004
3005         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3006         (AccessibilityUIElement::isSelectable): Implement this by checking
3007         the current state of the AtkObject.
3008         (AccessibilityUIElement::isMultiSelectable): Implement this by
3009         checking the current state of the AtkObject.
3010
3011 2012-12-09  Joanmarie Diggs  <jdiggs@igalia.com>
3012
3013         [GTK] accessibility/placeholder.html is failing
3014         https://bugs.webkit.org/show_bug.cgi?id=98373
3015
3016         Reviewed by Martin Robinson.
3017
3018         The test was failing because the placeholder text was not supported and
3019         AccessibilityUIElement::stringAttributeValue() was not implemented.
3020
3021         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3022         (coreAttributeToAtkAttribute): New convenience method to convert WebCore attribute
3023         names into AtkObject attribute names
3024         (AccessibilityUIElement::stringAttributeValue): implemented
3025
3026 2012-12-09  Dan Winship  <danw@gnome.org>
3027
3028         [GTK] Bump libxml2 requirement in jhbuild.modules
3029         https://bugs.webkit.org/show_bug.cgi?id=104477
3030
3031         Reviewed by Martin Robinson.
3032
3033         Fedora 18's Apache depends on symbols in libxml2 2.9.0, so the
3034         tests won't run if we only have libxml2 2.8.0 installed in the
3035         jhbuild root.
3036
3037         * gtk/jhbuild.modules:
3038
3039 2012-12-09  Zan Dobersek  <zandobersek@gmail.com>
3040
3041         Remove testRunner.disableImageLoading()
3042         https://bugs.webkit.org/show_bug.cgi?id=98083
3043
3044         Reviewed by Gustavo Noronha Silva.
3045
3046         Remove the disableImageLoading methods and callback.
3047         No tests use the method anymore and all ports except Wx
3048         now support this behavior through overriding the
3049         WebKitDisplayImagesKey preference.
3050
3051         * DumpRenderTree/TestRunner.cpp:
3052         (TestRunner::staticFunctions):
3053         * DumpRenderTree/TestRunner.h:
3054         (TestRunner):
3055         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
3056         (TestRunner::overridePreference): Add support for overriding the
3057         WebKitDisplayImagesKey preference which is now the preferred way
3058         of disabling image loading.
3059         * DumpRenderTree/efl/TestRunnerEfl.cpp:
3060         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
3061         * DumpRenderTree/mac/TestRunnerMac.mm:
3062         * DumpRenderTree/qt/TestRunnerQt.cpp:
3063         * DumpRenderTree/qt/TestRunnerQt.h:
3064         (TestRunner):
3065         * DumpRenderTree/win/TestRunnerWin.cpp:
3066         * DumpRenderTree/wx/TestRunnerWx.cpp:
3067
3068 2012-12-09  Joanmarie Diggs  <jdiggs@igalia.com>
3069
3070         [GTK] accessibility/label-for-control-hittest.html is failing
3071         https://bugs.webkit.org/show_bug.cgi?id=98368
3072
3073         Reviewed by Martin Robinson.
3074
3075         The test was failing because AccessibilityUIElement::elementAtPoint was
3076         not implemented.
3077
3078         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3079         (AccessibilityUIElement::elementAtPoint): implemented
3080
3081 2012-12-09  Jon Lee  <jonlee@apple.com>
3082
3083         [WK2] Move button image to injected bundle
3084         https://bugs.webkit.org/show_bug.cgi?id=104107
3085         <rdar://problem/12813842>
3086
3087         Reviewed by Simon Fraser.
3088
3089         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3090         (WTR::InjectedBundlePage::InjectedBundlePage): Update the bundle page UI client.
3091
3092 2012-12-08  Halton Huo  <halton.huo@intel.com>
3093
3094         [CMake] Add CMake style checker
3095         https://bugs.webkit.org/show_bug.cgi?id=104240
3096
3097         Reviewed by Eric Seidel.
3098
3099         The CMake style checker is followed by existing style checkers like cpp, python.
3100         Unit test is added as well.
3101
3102         * Scripts/webkitpy/style/checker.py: Use CMakeChecker CMAKE type files(.cmake or CMakeLists.txt)
3103         (FileType):
3104         (CheckerDispatcher._file_type):
3105         (CheckerDispatcher._create_checker):
3106         * Scripts/webkitpy/style/checkers/cmake.py: Added.
3107         (CMakeChecker):
3108         (CMakeChecker.__init__): 
3109         (CMakeChecker.check):
3110         (CMakeChecker._process_line):
3111         (CMakeChecker._check_trailing_whitespace):
3112         (CMakeChecker._check_no_space_cmds):
3113         (CMakeChecker._check_one_space_cmds):
3114         (CMakeChecker._check_non_lowercase_cmd):
3115         (CMakeChecker._check_indent):
3116         * Scripts/webkitpy/style/checkers/cmake_unittest.py: Added.
3117         (CMakeCheckerTest):
3118         (CMakeCheckerTest.test_init):
3119         (CMakeCheckerTest.test_init._mock_handle_style_error):
3120         (CMakeCheckerTest.test_check):
3121         (CMakeCheckerTest.test_check._mock_handle_style_error):
3122
3123 2012-12-08  Kondapally Kalyan  <kalyan.kondapally@intel.com>
3124
3125         [EFL] Enable WebGL by default.
3126         https://bugs.webkit.org/show_bug.cgi?id=102991.
3127
3128         Reviewed by Kenneth Rohde Christiansen.
3129
3130         Enable WebGL by default.
3131
3132         * Scripts/webkitperl/FeatureList.pm:
3133
3134 2012-12-07  Scott Violet  <sky@chromium.org>
3135
3136         [chromium] Remove linux theme related files and switch to default
3137         https://bugs.webkit.org/show_bug.cgi?id=103897
3138
3139         Reviewed by Dimitri Glazkov.
3140
3141         * Scripts/webkitpy/common/config/build_unittest.py:
3142         (ShouldBuildTest): linux->default
3143
3144 2012-12-07  Mike Lattanzio  <mlattanzio@rim.com>
3145
3146         [BlackBerry] Update BlackBerry Port to use new Graphics::createBuffer BufferType enum values.
3147         https://bugs.webkit.org/show_bug.cgi?id=104121
3148
3149         Reviewed by Rob Buis.
3150
3151         Update createBuffer call to reflect new BufferType enum.
3152         Internal PR 254649
3153         Internal Review: Jakob Petsovits
3154
3155         DRT is always backed for maximum compatibility.
3156
3157         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
3158         (createBitmapContextFromWebView):
3159
3160 2012-12-07  Jochen Eisinger  <jochen@chromium.org>
3161
3162         Add 'CHROMIUM_DEVEL_SANDBOX' to the environment variables to copy for layout test drivers
3163         https://bugs.webkit.org/show_bug.cgi?id=104355
3164
3165         Reviewed by Dirk Pranke.
3166
3167         This enables chromium's content_shell driver to run with the full SUID
3168         sandbox.
3169
3170         * Scripts/webkitpy/layout_tests/port/base.py:
3171         (Port.to.setup_environ_for_server):
3172
3173 2012-12-07  KyungTae Kim  <ktf.kim@samsung.com>
3174
3175         [EFL] Fix tools for WebKit2-only builds
3176         https://bugs.webkit.org/show_bug.cgi?id=103904
3177
3178         Reviewed by Laszlo Gombos.
3179
3180         The libewebkit2.so should be used for library path for EFL WebKit2.
3181
3182         * Scripts/webkitdirs.pm:
3183         (builtDylibPathForName):
3184
3185 2012-12-07  Xiaobo Wang  <xbwang@torchmobile.com.cn>
3186
3187         [BlackBerry] DRT - Update implementation of TestRunner.setCustomPolicyDelegate()
3188         https://bugs.webkit.org/show_bug.cgi?id=104243
3189
3190         RIM PR 259337
3191
3192         Reviewed by Rob Buis.
3193
3194         1. Implement calculation of navigation originating in
3195         didDecidePolicyForNavigationAction().
3196         2. Implement policy delegate for resource response.
3197         3. Enable customPolicyDelegate if waitForPolicyDelegate() is called.
3198         4. Minor optional updates in dump() and dumpFrameLoadCallbacks().
3199
3200         The patch fixed the following tests:
3201         - http/tests/download
3202         - fast/forms/mailto/formenctype-attribute-button-html.html
3203         - fast/forms/mailto/formenctype-attribute-input-html.html
3204         - fast/encoding/mailto-always-utf-8.html
3205
3206         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
3207         (BlackBerry::WebKit::DumpRenderTree::DumpRenderTree):
3208         (BlackBerry::WebKit::DumpRenderTree::resetToConsistentStateBeforeTesting):
3209         (BlackBerry::WebKit::DumpRenderTree::dump):
3210         (BlackBerry::WebKit::DumpRenderTree::didFinishLoadForFrame):
3211         (BlackBerry::WebKit::DumpRenderTree::didDecidePolicyForNavigationAction):
3212         (BlackBerry::WebKit::DumpRenderTree::didDecidePolicyForResponse):
3213         (BlackBerry::WebKit::DumpRenderTree::setCustomPolicyDelegate):
3214         * DumpRenderTree/blackberry/DumpRenderTreeBlackBerry.h:
3215         (DumpRenderTree):
3216         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
3217         (TestRunner::setCustomPolicyDelegate):
3218         (TestRunner::waitForPolicyDelegate):
3219         (TestRunner::overridePreference):
3220
3221 2012-12-07  Gwang Yoon Hwang  <ryumiel@company100.net>
3222
3223         [Qt] Adjust library name in layout_tests/port/qt.py after Qt5-r40
3224         https://bugs.webkit.org/show_bug.cgi?id=104348
3225
3226         Reviewed by Csaba Osztrogonác.
3227
3228         After Qt-5.0.0-r40, the library was renamed from QtWebKitWidgets to
3229         Qt5WebKitWidgets. run-webkit-tests were still looking for QtWebKitWidgets
3230
3231         * Scripts/webkitpy/layout_tests/port/qt.py:
3232         (QtPort._path_to_webcore_library):
3233
3234 2012-12-07  Csaba Osztrogonác  <ossy@webkit.org>
3235
3236         Unreviewed fix, unbreak run-launcher after library renaming.
3237
3238         * Scripts/webkitdirs.pm:
3239         (builtDylibPathForName):
3240
3241 2012-12-06  Csaba Osztrogonác  <ossy@webkit.org>
3242
3243         Unreviewed fix after r136896, unbreak --debug-rwt to make bots result verbose again.
3244         https://bugs.webkit.org/show_bug.cgi?id=104296
3245
3246         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3247         (main):
3248
3249 2012-12-06  Rick Byers  <rbyers@chromium.org>
3250
3251         CSS cursor property should support webkit-image-set
3252         https://bugs.webkit.org/show_bug.cgi?id=99493
3253
3254         Reviewed by Beth Dakin.
3255
3256         Add ENABLE_MOUSE_CURSOR_SCALE - disabled by default
3257
3258         * Scripts/webkitperl/FeatureList.pm:
3259         * qmake/mkspecs/features/features.pri:
3260
3261 2012-12-06  Vivek Galatage  <vivek.vg@samsung.com>
3262
3263         Unreviewed. Fixing the email address order to work correctly with bugzilla.
3264
3265         * Scripts/webkitpy/common/config/committers.py:
3266
3267 2012-12-06  Jochen Eisinger  <jochen@chromium.org>
3268
3269         [chromium] when forwarding events through the WebPluginContainer, create a UserGestureIndicator if processing a user gesture
3270         https://bugs.webkit.org/show_bug.cgi?id=104306
3271
3272         Reviewed by Dimitri Glazkov.
3273
3274         * DumpRenderTree/chromium/TestWebPlugin.cpp:
3275         (TestWebPlugin::TestWebPlugin):
3276         (TestWebPlugin::handleInputEvent): add support for printing the user gesture status
3277         * DumpRenderTree/chromium/TestWebPlugin.h:
3278         (TestWebPlugin):
3279
3280 2012-12-06  Dirk Pranke  <dpranke@chromium.org>
3281
3282         new-run-webkit-tests --lint-test-files seems to be broken
3283         https://bugs.webkit.org/show_bug.cgi?id=104296
3284
3285         Unreviewed, build fix.
3286
3287         My recent refactoring caused me to not initialize the printer
3288         before calling lint(); this patch restructures the code to make
3289         the lint routine self-contained and do all the work necessary.
3290         This will also make it easier to move this code out into a standalone
3291         file.
3292
3293         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3294         (lint):
3295         (main):
3296         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3297         (LintTest.test_all_configurations):
3298         (LintTest.test_lint_test_files):
3299         (LintTest.test_lint_test_files__errors):
3300
3301 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
3302
3303         DFG profiler should be helpful about gem installation
3304         https://bugs.webkit.org/show_bug.cgi?id=104217
3305
3306         Reviewed by Oliver Hunt.
3307         
3308         'json' and 'highline' are not installed by default on the ruby distributions with
3309         which I am familiar, and the default error messages don't make me happy. This makes
3310         display-profiler-output print a helpful message if those gems are not found.
3311
3312         * Scripts/display-profiler-output:
3313
3314 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
3315
3316         Profiler should print a helpful message if you pass the wrong arguments
3317         https://bugs.webkit.org/show_bug.cgi?id=104222
3318
3319         Reviewed by Oliver Hunt.
3320
3321         * Scripts/display-profiler-output:
3322
3323 2012-12-06  Ojan Vafai  <ojan@chromium.org>
3324
3325         Allow for CCing a secondary email address from watchlists
3326         https://bugs.webkit.org/show_bug.cgi?id=104286
3327
3328         Reviewed by Dirk Pranke.
3329
3330         Change my watchlist CC address so I can filter these separately from
3331         when people explicitly CC me and update the watchlist parser to allow that.
3332
3333         * Scripts/webkitpy/common/config/committers.py:
3334         * Scripts/webkitpy/common/config/watchlist:
3335         * Scripts/webkitpy/common/watchlist/watchlistparser.py:
3336         (WatchListParser._validate):
3337         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py:
3338         (WatchListParserTest.test_cc_rule_with_invalid_email):
3339         (WatchListParserTest.test_cc_rule_with_secondary_email):
3340
3341 2012-12-06  Max Vujovic  <mvujovic@adobe.com>
3342
3343         Unreviewed. Add myself to committers.py.
3344
3345         * Scripts/webkitpy/common/config/committers.py:
3346
3347 2012-12-06  Yael Aharon  <yael.aharon@intel.com>
3348
3349         [EFL] [WK2] fast/dom/vertical-scrollbar-in-rtl.html makes fast/regions tests flaky
3350         https://bugs.webkit.org/show_bug.cgi?id=104139
3351
3352         Reviewed by Kenneth Rohde Christiansen.
3353
3354         Resize the window back to 800x600 between tests.
3355         This should be a no-op if the test did not resize the window.
3356
3357         * WebKitTestRunner/TestController.cpp:
3358         (WTR::TestController::resetStateToConsistentValues):
3359
3360 2012-12-06  Tommy Widenflycht  <tommyw@google.com>
3361
3362         Speech Recognition API: Change the error code to a string on SpeechRecognitionError
3363         https://bugs.webkit.org/show_bug.cgi?id=104254
3364
3365         Reviewed by Adam Barth.
3366
3367         Changes the mock error function to take a string instead of an int.
3368
3369         * DumpRenderTree/chromium/DRTTestRunner.cpp:
3370         (DRTTestRunner::setMockSpeechRecognitionError):
3371         * DumpRenderTree/chromium/MockWebSpeechRecognizer.cpp:
3372         (WebTestRunner::ErrorTask::ErrorTask):
3373         (MockWebSpeechRecognizer::setError):
3374         * DumpRenderTree/chromium/MockWebSpeechRecognizer.h:
3375         (MockWebSpeechRecognizer):
3376
3377 2012-12-06  Andras Becsi  <andras.becsi@digia.com>
3378
3379         [Qt][Mac] Fix libxslt and libxml2 config tests
3380         https://bugs.webkit.org/show_bug.cgi?id=104164
3381
3382         Reviewed by Simon Hausmann.
3383
3384         We should not use pkg-config on Mac instead use direct include
3385         paths and add needed libraries to the linker options to detect
3386         libxslt and libxml2 provided by the system.
3387         Previously we would always fall back to qtxmlpatterns.
3388
3389         * qmake/config.tests/libxml2/libxml2.pro:
3390         * qmake/config.tests/libxslt/libxslt.pro:
3391
3392 2012-12-06  Csaba Osztrogonác  <ossy@webkit.org>
3393
3394         [Qt] Fix the build if the path contains "+" character
3395         https://bugs.webkit.org/show_bug.cgi?id=104120
3396
3397         Reviewed by Tor Arne Vestbø.
3398
3399         The second parameter of qmake's replace() is regular expression, so
3400         we have to escape special characters in paths passed to replace().
3401
3402         * qmake/mkspecs/features/default_post.prf:
3403         * qmake/mkspecs/features/default_pre.prf:
3404         * qmake/mkspecs/features/functions.prf:
3405
3406 2012-12-06  Sheriff Bot  <webkit.review.bot@gmail.com>
3407
3408         Unreviewed, rolling out r136825.
3409         http://trac.webkit.org/changeset/136825
3410         https://bugs.webkit.org/show_bug.cgi?id=104251
3411
3412         It made 100+ test fail on Qt-WK2 pixel bot, and 1800+ test
3413         flakey on Qt-WK2 non-pixel bot (Requested by Ossy on #webkit).
3414
3415         * WebKitTestRunner/TestController.cpp:
3416         (WTR::TestController::resetStateToConsistentValues):
3417
3418 2012-12-06  Yael Aharon  <yael.aharon@intel.com>
3419
3420         [EFL] [WK2] fast/dom/vertical-scrollbar-in-rtl.html makes fast/regions tests flaky
3421         https://bugs.webkit.org/show_bug.cgi?id=104139
3422
3423         Reviewed by Kenneth Rohde Christiansen.
3424
3425         Resize the window back to 800x600 between tests.
3426         This should be a no-op if the test did not resize the window.
3427
3428         * WebKitTestRunner/TestController.cpp:
3429         (WTR::TestController::resetStateToConsistentValues):
3430
3431 2012-12-06  Sheriff Bot  <webkit.review.bot@gmail.com>
3432
3433         Unreviewed, rolling out r136784 and r136802.
3434         http://trac.webkit.org/changeset/136784
3435         http://trac.webkit.org/changeset/136802
3436         https://bugs.webkit.org/show_bug.cgi?id=104231
3437
3438         breaks chromium canary (Requested by morrita on #webkit).
3439
3440         * Scripts/webkitpy/common/config/build_unittest.py:
3441         (ShouldBuildTest):
3442
3443 2012-12-05  Ryosuke Niwa  <rniwa@webkit.org>
3444
3445         Add myself to watch changes in PerformanceTests/*.
3446
3447         * Scripts/webkitpy/common/config/watchlist:
3448
3449 2012-12-05  Shinya Kawanaka  <shinyak@chromium.org>
3450
3451         [Chromium] DRT on Windows crashes when a testcase has a progress element in indeterminate state
3452         https://bugs.webkit.org/show_bug.cgi?id=102459
3453
3454         Reviewed by Dimitri Glazkov.
3455
3456         Since WebThemeControlDRTWin::markState does not consider indeterminate state, DRT caused a crash.
3457
3458         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
3459         (WebThemeControlDRTWin::markState):
3460
3461 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
3462
3463         DFG profiler should be able to display where a code block was inlined into, and what it inlined
3464         https://bugs.webkit.org/show_bug.cgi?id=104216
3465
3466         Reviewed by Oliver Hunt.
3467
3468         Added an "inlines" command that shows the inline stacks that a code block participated
3469         in. Also cleaned up some of the ways we print compilation units.
3470
3471         * Scripts/display-profiler-output:
3472
3473 2012-12-05  Halton Huo  <halton.huo@intel.com>
3474
3475         [CMake] Unify coding style for CMake files
3476         https://bugs.webkit.org/show_bug.cgi?id=103605
3477
3478         Reviewed by Laszlo Gombos.
3479
3480         Update cmake files(.cmake, CMakeLists.txt) with following style rules:
3481         1. Indentation
3482         1.1 Use spaces, not tabs.
3483         1.2 Four spaces as indent.
3484         2. Spacing
3485         2.1 Place one space between control statements and their parentheses.
3486             For eg, if (), else (), elseif (), endif (), foreach (),
3487             endforeach (), while (), endwhile (), break ().
3488         2.2 Do not place spaces between function and macro statements and
3489             their parentheses. For eg, macro(), endmacro(), function(),
3490             endfunction().
3491         2.3 Do not place spaces between a command or function or macro and its
3492             parentheses, or between a parenthesis and its content. For eg,
3493             message("testing") not message( "testing") or message ("testing" )
3494         2.4 No space at line ending.
3495         3. Lowercase when call commands macros and functions. For eg,
3496            add_executable() not ADD_EXECUTABLE(), set() not SET().
3497
3498         * CMakeLists.txt:
3499         * DumpRenderTree/TestNetscapePlugIn/CMakeLists.txt:
3500         * DumpRenderTree/efl/CMakeLists.txt:
3501         * EWebLauncher/CMakeLists.txt:
3502         * EWebLauncher/ControlTheme/CMakeLists.txt:
3503         * MiniBrowser/efl/CMakeLists.txt:
3504         * TestWebKitAPI/CMakeLists.txt:
3505         * TestWebKitAPI/PlatformEfl.cmake:
3506         * WebKitTestRunner/CMakeLists.txt:
3507         * WebKitTestRunner/PlatformEfl.cmake:
3508         * WinCELauncher/CMakeLists.txt:
3509         * clang/ReportMemoryUsagePlugin/CMakeLists.txt:
3510
3511 2012-12-05  Scott Violet  <sky@chromium.org>
3512
3513         [chromium] Remove linux theme related files and switch to default
3514         https://bugs.webkit.org/show_bug.cgi?id=103897
3515
3516         Reviewed by Tony Chang.
3517
3518         * Scripts/webkitpy/common/config/build_unittest.py:
3519         (ShouldBuildTest): linux->default
3520
3521 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3522
3523         nrwt: run_webkit_tests_integrationtest.py depends on being able to share a port w/ between master and worker
3524         https://bugs.webkit.org/show_bug.cgi?id=78168
3525
3526         Reviewed by Ojan Vafai.
3527
3528         This patch adds a single ordered list to ResultSummary that contains
3529         the results of every test run in order of result received, and
3530         modifies the TestResults class to contain the list of references used
3531         for each test as well as the pid of the driver; this allows us to stop
3532         having to override the test drivers to record the tests, and makes the
3533         integration tests that test rdering, retries, restarts, etc. black-box
3534         tests.
3535
3536         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
3537         (SingleTestRunner._run_rebaseline):
3538         (SingleTestRunner._compare_output):
3539         (SingleTestRunner._run_reftest):
3540         (SingleTestRunner._compare_output_with_reference):
3541         * Scripts/webkitpy/layout_tests/models/result_summary.py:
3542         (ResultSummary.__init__):
3543         (ResultSummary.add):
3544         * Scripts/webkitpy/layout_tests/models/test_results.py:
3545         (TestResult.__init__):
3546         * Scripts/webkitpy/layout_tests/port/driver.py:
3547         (DriverOutput.__init__):
3548         (Driver.run_test):
3549         * Scripts/webkitpy/layout_tests/port/test.py:
3550         (TestDriver):
3551         (TestDriver.__init__):
3552         (TestDriver.run_test):
3553         (TestDriver.stop):
3554         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3555         (get_tests_run):
3556         (get_test_batches):
3557         (get_test_results):
3558         (MainTest.test_batch_size):
3559         (MainTest.test_dryrun):
3560         (MainTest.test_natural_order):
3561         (MainTest.test_natural_order_test_specified_multiple_times):
3562         (MainTest.test_random_order):
3563         (MainTest.test_random_order_test_specified_multiple_times):
3564         (MainTest.test_no_order):
3565         (MainTest.test_no_order_test_specified_multiple_times):
3566         (MainTest.test_no_order_with_directory_entries_in_natural_order):
3567         (MainTest.test_repeat_each):
3568         (MainTest.test_ignore_flag):
3569         (MainTest.test_skipped_flag):
3570         (MainTest.test_iterations):
3571         (MainTest.test_run_chunk):
3572         (MainTest.test_run_part):
3573         (MainTest.test_run_singly):
3574         (MainTest.test_skip_failing_tests):
3575         (MainTest.test_single_file):
3576         (MainTest.test_single_file_with_prefix):
3577         (MainTest.test_single_skipped_file):
3578         (MainTest.test_test_list):
3579         (MainTest.test_test_list_with_prefix):
3580         (MainTest.test_exit_after_n_failures):
3581         (MainTest.test_exit_after_n_crashes):
3582         (MainTest.test_run_order__inline):
3583         (MainTest.test_reftest_run):
3584         (MainTest.test_reftest_run_reftests_if_pixel_tests_are_disabled):
3585         (MainTest.test_reftest_skip_reftests_if_no_ref_tests):
3586         (MainTest.test_reftest_expected_html_should_be_ignored):
3587         (MainTest.test_reftest_driver_should_run_expected_html):
3588         (MainTest.test_reftest_driver_should_run_expected_mismatch_html):
3589         (MainTest.test_no_http_tests):
3590         (MainTest.test_platform_tests_are_found):
3591
3592 2012-12-05  Dana Jansens  <danakj@chromium.org>
3593
3594         [chromium] Make WebCompositorOutputSurface an empty class when USE_CC_OUTPUT_SURFACE is defined.
3595         https://bugs.webkit.org/show_bug.cgi?id=103967
3596
3597         Reviewed by James Robinson.
3598
3599         Use the WebCompositorSupport methods to create an OutputSurface. If they
3600         fail, then fall back to the WebViewHost methods.
3601
3602         Once the chromium side lands, the WebCompositorSupport methods will stop
3603         failing, and will replace the WebViewHost ones.
3604
3605         * DumpRenderTree/chromium/WebViewHost.cpp:
3606         (WebViewHost::createOutputSurface):
3607
3608 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3609
3610         nrwt: split all of the buildbot-specific output into its own file
3611         https://bugs.webkit.org/show_bug.cgi?id=104165
3612
3613         Reviewed by Ojan Vafai.
3614
3615         Now that we have the RunDetails object we can move all of the
3616         buildbot-specific knowledge out of the core test-running code
3617         and into the outermost run_webkit_tests layer.
3618
3619         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3620         (run):
3621         (main):
3622         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3623         (passing_run):
3624         (logging_run):
3625         (run_and_capture):
3626         (MainTest.test_all):
3627         (MainTest.test_max_locked_shards):
3628         (MainTest.test_child_processes_2):
3629         (MainTest.test_child_processes_min):
3630         (MainTest.test_full_results_html):
3631         (MainTest.test_hung_thread):
3632         (MainTest.test_keyboard_interrupt):
3633         (MainTest.test_no_tests_found):
3634         (MainTest.test_no_tests_found_2):
3635         (MainTest.test_repeat_each_iterations_num_tests):
3636         (MainTest.test_run_singly_actually_runs_tests):
3637         (MainTest.test_test_list):
3638         (MainTest.test_unexpected_failures):
3639         (MainTest.test_missing_and_unexpected_results):
3640         (MainTest.test_pixel_test_directories):
3641         (MainTest.test_missing_and_unexpected_results_with_custom_exit_code):
3642         (MainTest.test_crash_with_stderr):
3643         (MainTest.test_no_image_failure_with_image_diff):
3644         (MainTest.test_crash_log):
3645         (MainTest.test_web_process_crash_log):
3646         (MainTest.test_exit_after_n_failures_upload):
3647         (MainTest.test_results_directory_absolute):
3648         (MainTest.test_results_directory_default):
3649         (MainTest.test_results_directory_relative):
3650         (MainTest.test_retrying_and_flaky_tests):
3651         (MainTest.test_reftest_should_not_use_naming_convention_if_not_listed_in_reftestlist):
3652         (MainTest.test_output_diffs):
3653         (MainTest.test_unsupported_platform):
3654         (MainTest.test_verbose_in_child_processes):
3655         (EndToEndTest.test_end_to_end):
3656         (EndToEndTest.test_reftest_with_two_notrefs):
3657         (RebaselineTest.test_reset_results):
3658         (RebaselineTest.test_missing_results):
3659         (RebaselineTest.test_new_baseline):
3660         * Scripts/webkitpy/layout_tests/views/buildbot_results.py: Added.
3661         (BuildBotPrinter):
3662         (BuildBotPrinter.__init__):
3663         (BuildBotPrinter.print_results):
3664         (BuildBotPrinter._print):
3665         (BuildBotPrinter.print_result_summary):
3666         (BuildBotPrinter._print_result_summary_entry):
3667         (BuildBotPrinter.print_unexpected_results):
3668         (BuildBotPrinter.print_unexpected_results.add_to_dict_of_lists):
3669         (BuildBotPrinter.print_unexpected_results.add_result):
3670         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py: Added.
3671         (BuildBotPrinterTests):
3672         (BuildBotPrinterTests.assertEmpty):
3673         (BuildBotPrinterTests.assertNotEmpty):
3674         (BuildBotPrinterTests.get_printer):
3675         (BuildBotPrinterTests.get_result):
3676         (BuildBotPrinterTests.get_result_summary):
3677         (BuildBotPrinterTests.test_print_unexpected_results):
3678         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
3679         * Scripts/webkitpy/layout_tests/views/printing.py:
3680         (Printer):
3681         (Printer.__init__):
3682         (Printer.print_results):
3683         (Printer._print_statistics_for_test_timings):
3684         (Printer._print_baseline):
3685         (Printer._print_debug):
3686         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
3687         (Testprinter.get_printer):
3688         (Testprinter.get_result):
3689         (Testprinter.test_configure_and_cleanup):
3690         (Testprinter.test_print_config):
3691         (Testprinter.test_print_one_line_summary):
3692         (Testprinter.test_test_status_line):
3693         (Testprinter.test_details):
3694
3695 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3696
3697         nrwt: collect all of the information about a run into a new RunDetails class
3698         https://bugs.webkit.org/show_bug.cgi?id=104158
3699
3700         Reviewed by Ojan Vafai.
3701
3702         This patch collects the information from both test passes (the
3703         main one and the retries) as well as the summarized results and
3704         the exit code into a RunDetails class, so that the integration
3705         tests can more easily tell what happened.
3706
3707         Also, change the way the --lint-test-files works slightly so that
3708         we don't need to create a printer object (and hence we need to
3709         rework the tests that are testing logging).
3710
3711         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3712         (RunDetails):
3713         (RunDetails.__init__):
3714         (Manager.run):
3715         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3716         (run):
3717         (main):
3718         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3719         (passing_run):
3720         (run_and_capture):
3721         (LintTest.test_lint_test_files):
3722         (LintTest.test_lint_test_files__errors):
3723         (MainTest.test_verbose_in_child_processes):
3724
3725 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3726
3727         nrwt: remove --no-record-results
3728         https://bugs.webkit.org/show_bug.cgi?id=104072
3729
3730         Reviewed by Ojan Vafai.
3731
3732         The command line flag was only being used in the unit tests and
3733         wasn't really needed once we started using a mock filesystem.
3734
3735         Also, reformat a bunch of long lines ...
3736
3737         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3738         (Manager.run):
3739         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3740         (parse_args):
3741         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3742         (parse_args):
3743         (passing_run):
3744         (logging_run):
3745         (MainTest.test_repeat_each_iterations_num_tests):
3746         (MainTest.test_test_list):
3747         (MainTest.test_missing_and_unexpected_results):
3748         (MainTest.test_pixel_test_directories):
3749         (MainTest.test_missing_and_unexpected_results_with_custom_exit_code):
3750         (MainTest.test_crash_with_stderr):
3751         (MainTest.test_no_image_failure_with_image_diff):
3752         (MainTest.test_crash_log):
3753         (MainTest.test_web_process_crash_log):
3754         (MainTest.test_exit_after_n_failures_upload):
3755         (MainTest.test_results_directory_absolute):
3756         (MainTest.test_results_directory_relative):
3757         (MainTest.test_reftest_should_not_use_naming_convention_if_not_listed_in_reftestlist):
3758         (MainTest.test_additional_expectations):
3759         (MainTest.test_output_diffs):
3760         (EndToEndTest.test_end_to_end):
3761         (EndToEndTest.test_reftest_with_two_notrefs):
3762         (RebaselineTest.test_reset_results):
3763         (RebaselineTest.test_missing_results):
3764         (RebaselineTest.test_new_baseline):
3765
3766 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3767
3768         nrwt: only summarize the results one way
3769         https://bugs.webkit.org/show_bug.cgi?id=104064
3770
3771         Reviewed by Ojan Vafai.
3772
3773         Change manager.summarize_results so that we always (and only)
3774         include both expected and unexpected results. Previously we
3775         would summarize the results twice, once with expected results
3776         and once without. This made the code much more confusing than
3777         it needed to be.
3778
3779         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3780         (summarize_results):
3781         (Manager.run):
3782         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3783         (ResultSummaryTest.summarized_results):
3784         (ResultSummaryTest.test_no_svn_revision):
3785         (ResultSummaryTest.test_svn_revision):
3786         (ResultSummaryTest.test_summarized_results_wontfix):
3787         * Scripts/webkitpy/layout_tests/views/printing.py:
3788         (Printer.print_results):
3789         (Printer._print_unexpected_results):
3790         (Printer._print_unexpected_results.add_result):
3791         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
3792         (Testprinter.test_print_unexpected_results.get_unexpected_results):
3793         (Testprinter):
3794
3795 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
3796
3797         DFG profiler should display more information
3798         https://bugs.webkit.org/show_bug.cgi?id=104163
3799
3800         Reviewed by Gavin Barraclough.
3801
3802         Added the following:
3803         
3804         - Distinguish between source counts and bytecode counts.
3805         
3806         - Add a "full" summary view that shows both kinds of counts, plus inline counts.
3807         
3808         - Add a "source" command to see the source.
3809         
3810         - Add a "bytecode" command to see the bytecode and counts.
3811
3812         * Scripts/display-profiler-output:
3813
3814 2012-12-05  Yong Li  <yoli@rim.com>
3815
3816         [BlackBerry] Build with libjpegturbo
3817         https://bugs.webkit.org/show_bug.cgi?id=104152
3818
3819         Reviewed by Rob Buis.
3820
3821         RIM PR# 196975.
3822         Make sure the right include path for libjpeg is included.
3823
3824         * Scripts/webkitdirs.pm:
3825         (blackberryCMakeArguments):
3826
3827 2012-12-05  Joseph Pecoraro  <pecoraro@apple.com>
3828
3829         Remove stale check in check-for-webkit-framework-include-consistency
3830         https://bugs.webkit.org/show_bug.cgi?id=103775
3831
3832         Reviewed by David Kilzer.
3833
3834         The underlying issue was fixed in r81173.
3835
3836         * Scripts/check-for-webkit-framework-include-consistency:
3837
3838 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3839
3840         nrwt: simplify the finding of tests to retry
3841         https://bugs.webkit.org/show_bug.cgi?id=104067
3842
3843         Reviewed by Ryosuke Niwa.
3844
3845         I think this makes the code a little clearer; note that
3846         we were no longer using the include_missing flag.
3847
3848         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3849         (Manager.run):
3850         (Manager._failed_test_names):
3851
3852 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3853
3854         nrwt: remove two unused parameters from a function
3855         https://bugs.webkit.org/show_bug.cgi?id=104059
3856
3857         Reviewed by Ryosuke Niwa.
3858
3859         Also remove an old, stale FIXME.
3860
3861         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3862         (interpret_test_failures):
3863         (summarize_results):
3864         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3865         (ResultSummaryTest.test_interpret_test_failures):
3866
3867 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3868
3869         nrwt: return result_summaries from LayoutTestRunner.run_tests()
3870         https://bugs.webkit.org/show_bug.cgi?id=104051
3871
3872         Reviewed by Ojan Vafai.
3873
3874         This patch completes the refactoring that makes the result summary      
3875         an object returned from the layout test runner, rather than having      
3876         it passed in and modified. This makes the split of duties between       
3877         the manager and runner clearer and makes the data flow more             
3878         functionally.                                                           
3879
3880         Also clean up some minor lint issues.
3881
3882         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
3883         (LayoutTestRunner.__init__):
3884         (LayoutTestRunner.run_tests):
3885         (LayoutTestRunner._mark_interrupted_tests_as_skipped):
3886         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
3887         (FakePrinter.print_expected):
3888         (LockCheckingRunner.__init__):
3889         (LayoutTestRunnerTests._runner):
3890         (LayoutTestRunnerTests._run_tests):
3891         (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
3892         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3893         (Manager.__init__):
3894         (Manager.run):
3895         (Manager._run_tests):
3896
3897 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
3898
3899         nrwt: move creation of the resultsummary objects into Manager._run_tests()
3900         https://bugs.webkit.org/show_bug.cgi?id=104048
3901
3902         Reviewed by Ojan Vafai.
3903
3904         This is a preparatory step to making the resultsummary just get
3905         returned from the layout_test_runner; this does most of manager-side
3906         changes.
3907
3908         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3909         (Manager._prepare_lists):
3910         (Manager.run):
3911         (Manager._run_tests):
3912
3913 2012-12-05  Leo Yang  <leoyang@rim.com>
3914
3915         [BlackBerry] Enable CSS_IMAGE_RESOLUTION
3916         https://bugs.webkit.org/show_bug.cgi?id=104132
3917
3918         Reviewed by Yong Li.
3919
3920         * Scripts/webkitperl/FeatureList.pm:
3921
3922 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
3923
3924         JSC profiler should not count executions of op_call_put_result because doing so changes DFG codegen
3925         https://bugs.webkit.org/show_bug.cgi?id=104102
3926
3927         Reviewed by Oliver Hunt.
3928
3929         Modify the profiler to not output counts for op_call_put_result, since there
3930         won't be any. Also fix a few weird bugs, like providing better error reporting
3931         when you type something incorrectly and not reporting counts for slow paths
3932         in the old JIT since those counts are actually not what you think they are
3933         (we don't actually count slow path executions separately).
3934
3935         * Scripts/display-profiler-output:
3936
3937 2012-12-05  Andras Becsi  <andras.becsi@digia.com>
3938
3939         Fix compilation for Qt5.0.0 stable branch.
3940         https://bugs.webkit.org/show_bug.cgi?id=103870
3941
3942         Reviewed by Simon Hausmann.
3943
3944         Fix the use of renamed QWindow functions.
3945         Patch by Andras Becsi <andras.becsi@digia.com>
3946
3947         * MiniBrowser/qt/BrowserWindow.cpp:
3948         (BrowserWindow::BrowserWindow):
3949         (BrowserWindow::onTitleChanged):
3950         * MiniBrowser/qt/raw/View.cpp:
3951         (View::View):
3952
3953 2012-12-05  Jussi Kukkonen  <jussi.kukkonen@intel.com>
3954
3955         [WK2] InjectedBundle should check Intent url for null
3956         https://bugs.webkit.org/show_bug.cgi?id=104117
3957
3958         Reviewed by Kenneth Rohde Christiansen.
3959
3960         WKIntentServiceInfoCopyHref() can return null, we need to check
3961         for that in InjectedBundlePage::registerIntentServiceForFrame()
3962         to avoid crashing.
3963
3964         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3965         (WTR::InjectedBundlePage::registerIntentServiceForFrame):
3966
3967 2012-12-04  Jonathan Dong  <jonathan.dong@torchmobile.com.cn>
3968
3969         Change email address of Jonathan Dong
3970         https://bugs.webkit.org/show_bug.cgi?id=103976
3971
3972         Unreviewed, change my email address from jonathan.dong@torchmobile.com.cn
3973         to jonathan.dong.webkit@gmail.com in committers list.
3974
3975         * Scripts/webkitpy/common/config/committers.py:
3976
3977 2012-12-04  Xiaobo Wang  <xbwang@torchmobile.com.cn>
3978
3979         [BlackBerry] DRT - Tests under "dumpAsText" directory should be dumped as text
3980         https://bugs.webkit.org/show_bug.cgi?id=103986
3981
3982         Reviewed by Rob Buis.
3983
3984         Also move declaration of webSettingTransaction to before any WebSettings
3985         ::setXXX() calls so that it works as expected (commit only once in the
3986         function scope).
3987
3988         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
3989         (shouldDumpAsText):
3990         (BlackBerry::WebKit::DumpRenderTree::resetToConsistentStateBeforeTesting):
3991
3992 2012-12-04  Filip Pizlo  <fpizlo@apple.com>
3993
3994         display-profiler-output should be able to show source code
3995         https://bugs.webkit.org/show_bug.cgi?id=104073
3996
3997         Reviewed by Oliver Hunt.
3998
3999         Display source code in the summary, using a one-line shortening.
4000
4001         * Scripts/display-profiler-output:
4002
4003 2012-12-04  Filip Pizlo  <fpizlo@apple.com>
4004
4005         display-profiler-output should not use reflective infocation to resolve command names
4006
4007         Rubber stamped by Mark Hahnenberg.
4008
4009         Reflection is amazing, but here it was just kind of a bit much.
4010
4011         * Scripts/display-profiler-output:
4012
4013 2012-12-02  Filip Pizlo  <fpizlo@apple.com>
4014
4015         JSC should be able to report profiling data associated with the IR dumps and disassembly
4016         https://bugs.webkit.org/show_bug.cgi?id=102999
4017
4018         Reviewed by Gavin Barraclough.
4019
4020         Added a tool that allows you to grok the output from JSC's new profiler. Currently,
4021         this still gets confused a bit about the execution counts of a method running
4022         standalone versus a method running inlined, but other than that, it's pretty cool.
4023         See the attached "sampling profiling session" attached to the bug to see it in
4024         action.
4025         
4026         Also had to feed EFL's build system.
4027
4028         * DumpRenderTree/efl/CMakeLists.txt:
4029         * Scripts/display-profiler-output: Added.
4030
4031 2012-12-04  Dirk Pranke  <dpranke@chromium.org>
4032
4033         nrwt: fix total number of tests computed in the result summary
4034         https://bugs.webkit.org/show_bug.cgi?id=104061
4035
4036         Unreviewed, build fix.
4037
4038         One of the recent changes to the resultsummary refactoring caused
4039         us to not include the number of skipped tests in the total number
4040         of tests, throwing off the totals printed in print_one_line_summary().
4041
4042         Unfortunately, given the current state of the code it is difficult
4043         to construct an accurate unit test for this. I have verified the
4044         fix by hand, and will write a test once the remaining refactoring
4045         patches land, at which point it will be trivial to test it.
4046
4047         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4048         (Manager._prepare_lists):
4049
4050 2012-12-04  Dirk Pranke  <dpranke@chromium.org>
4051
4052         nrwt is not returning the exit status properly
4053         https://bugs.webkit.org/show_bug.cgi?id=103830
4054
4055         Unreviewed, build fix.
4056
4057         The fix in r136474 wasn't quite right either, and would return
4058         the correct exceptional status, but not the correct failing status.
4059
4060         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4061
4062 2012-12-04  Dirk Pranke  <dpranke@chromium.org>
4063
4064         nrwt: make paths and test_names passed arguments in Manager._prepare_lists et al
4065         https://bugs.webkit.org/show_bug.cgi?id=104047
4066
4067         Reviewed by Eric Seidel.
4068
4069         The code becomes cleaner if we are just passing values around rather
4070         than hanging them off the manager object, helps move _prepare_lists()
4071         to a pure function, and is needed to eventually make the
4072         result_summary object something returned from runner.run_tests()
4073         (note that two more patches are needed for that to happen).
4074
4075         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4076         (Manager.__init__):
4077         (Manager._http_tests):
4078         (Manager._prepare_lists):
4079         (Manager.needs_servers):
4080         (Manager._set_up_run):
4081         (Manager.run):
4082         (Manager._run_tests):
4083         (Manager._upload_json_files):
4084         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
4085         (ManagerTest.test_needs_servers.get_manager):
4086         (ManagerTest.test_needs_servers):
4087         (ManagerTest.integration_test_needs_servers.get_manager):
4088         (ManagerTest.integration_test_needs_servers):
4089         (ManagerTest.test_look_for_new_crash_logs.get_manager):
4090         (ManagerTest):
4091         (ManagerTest.test_look_for_new_crash_logs):
4092         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
4093         (JSONLayoutResultsGenerator.__init__):
4094         (JSONLayoutResultsGenerator._get_modifier_char):
4095
4096 2012-12-04  Adam Barth  <abarth@webkit.org>
4097
4098         commit-queue can get stuck with a local commit
4099         https://bugs.webkit.org/show_bug.cgi?id=104032
4100
4101         Reviewed by Dirk Pranke.
4102
4103         Previously the commit-queue would only clean down to HEAD. If the
4104         commit-queue gets interrupted in the middle of a commit (e.g., after
4105         the local commit but before the dcommit), then it can end up with a
4106         patch locally committed. That patch will then prevent the commit-queue
4107         from functioning.
4108
4109         This patch teaches the commit-queue to clean down to the merge base,
4110         which will remove any local commits.
4111
4112         * Scripts/webkitpy/common/checkout/scm/git.py:
4113         (Git.working_directory_is_clean):
4114         (Git.clean_working_directory):
4115         * Scripts/webkitpy/tool/commands/roll_unittest.py:
4116         * Scripts/webkitpy/tool/steps/commit.py:
4117         (Commit._commit_warning):
4118         (Commit.run):
4119
4120 2012-12-04  Dirk Pranke  <dpranke@chromium.org>
4121
4122         nrwt: store only the total number of tests in the result summary
4123         https://bugs.webkit.org/show_bug.cgi?id=103963
4124
4125         Reviewed by Ryosuke Niwa.
4126
4127         The result summary only needs the total number of tests that will
4128         be run (including repetitions) in order to provide the correct stats;
4129         it doesn't actually need to know how many unique tests there are, or
4130         or how many repetitions there are.
4131
4132         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
4133         (LayoutTestRunnerTests._result_summary):
4134         (LayoutTestRunnerTests.test_update_summary_with_result):
4135         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4136         (Manager._prepare_lists):
4137         (Manager.run):
4138         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
4139         (ManagerTest.test_look_for_new_crash_logs):
4140         (ResultSummaryTest.get_result_summary):
4141         * Scripts/webkitpy/layout_tests/models/result_summary.py:
4142         (ResultSummary.__init__):
4143         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
4144         (Testprinter.get_result_summary):
4145
4146 2012-12-04  Dirk Pranke  <dpranke@chromium.org>
4147
4148         nrwt: do less work when ctrl-c'ed
4149         https://bugs.webkit.org/show_bug.cgi?id=103961
4150
4151         Reviewed by Ryosuke Niwa.
4152
4153         Do less work when we are ctrl-c'ed, because the statistics aren't
4154         that useful anyway. While we're at it, modify summarize_results
4155         to be aware that we don't always retry results (instead of pretending
4156         that the retry produced the same results the initial run did).
4157
4158         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4159         (summarize_results):
4160         (Manager.run):
4161
4162 2012-12-04  Dirk Pranke  <dpranke@chromium.org>
4163
4164         nrwt: result summaries should not require a list of tests to skip
4165         https://bugs.webkit.org/show_bug.cgi?id=103951
4166
4167         Reviewed by Ryosuke Niwa.
4168
4169         We shouldn't really be passing a list of tests to skip to the
4170         ResultSummary objects; they should be handled like any other
4171         test result as much as possible.
4172
4173         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
4174         (LayoutTestRunnerTests._result_summary):
4175         (LayoutTestRunnerTests.test_update_summary_with_result):
4176         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4177         (Manager._prepare_lists):
4178         (Manager.run):
4179         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
4180         (ManagerTest.test_look_for_new_crash_logs):
4181         (ResultSummaryTest.get_result_summary):
4182         * Scripts/webkitpy/layout_tests/models/result_summary.py:
4183         (ResultSummary.__init__):
4184         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
4185         (TestExpectations):
4186         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
4187         (Testprinter.get_result_summary):
4188
4189 2012-12-04  Dirk Pranke  <dpranke@chromium.org>
4190
4191         nrwt: derive per-thread, per-worker stats from result summaries
4192         https://bugs.webkit.org/show_bug.cgi?id=103943
4193
4194         Reviewed by Ojan Vafai.
4195
4196         This patch changes the manager and printing code to pull more           
4197         information from the ResultSummary object rather than tracking          
4198         additional statistics on the side. This allows us to consolidate        
4199         a bunch of logic and remove a bunch of free-floating lists and dicts of 
4200         of values. We add the shard_name and worker_name to each TestResult,    
4201         and then modify print_timing_statistics() and upload_json_files()       
4202         to just need the result summaries.                                      
4203
4204         Note that this patch also changes the timing statistics slightly
4205         so that the individual test times are the times to run just the
4206         tests themselves (TestResult.test_run_time) and do not include the
4207         times taken to run the ref tests, compute diffs, etc. The aggregated
4208         per-worker and per-shard statistics still use the total time to
4209         process each test (TestResult.total_run_time).
4210
4211         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
4212         (LayoutTestRunner.__init__):
4213         (LayoutTestRunner.run_tests):
4214         (LayoutTestRunner._handle_finished_test_list):
4215         (LayoutTestRunner._handle_finished_test):
4216         (Worker.handle):
4217         (Worker._run_test):
4218         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4219         (summarize_results):
4220         (Manager.run):
4221         (Manager._upload_json_files):
4222         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
4223         (ResultSummaryTest.summarized_results):
4224         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
4225         (JSONLayoutResultsGenerator.__init__):
4226         (JSONLayoutResultsGenerator._get_test_timing):
4227         * Scripts/webkitpy/layout_tests/models/test_results.py:
4228         (TestResult.__init__):
4229         * Scripts/webkitpy/layout_tests/views/printing.py:
4230         (Printer.print_results):
4231         (Printer._print_timing_statistics):
4232         (Printer):
4233         (Printer._print_worker_statistics):
4234         (Printer._print_aggregate_test_statistics):
4235         (Printer._print_individual_test_times):
4236         (Printer._print_directory_timings):
4237         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
4238         (Testprinter.test_print_unexpected_results.get_unexpected_results):
4239         (Testprinter):
4240
4241 2012-12-04  Adam Barth  <abarth@webkit.org>
4242
4243         Remove obsolete code from webkitpy
4244         https://bugs.webkit.org/show_bug.cgi?id=104029
4245
4246         Reviewed by Eric Seidel.
4247
4248         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
4249         (CommitQueueTask.run):
4250         * Scripts/webkitpy/tool/steps/commit.py:
4251         (Commit.options):
4252         * Scripts/webkitpy/tool/steps/options.py:
4253         (Options):
4254
4255 2012-12-04  Jochen Eisinger  <jochen@chromium.org>
4256
4257         [chromium] move WebPreferences and methods for modifying them to the TestRunner library
4258         https://bugs.webkit.org/show_bug.cgi?id=103841
4259
4260         Reviewed by Darin Fisher.
4261
4262         * DumpRenderTree/DumpRenderTree.gypi:
4263         * DumpRenderTree/chromium/DRTTestRunner.cpp:
4264         (DRTTestRunner::DRTTestRunner):
4265         (DRTTestRunner::reset):
4266         * DumpRenderTree/chromium/DRTTestRunner.h:
4267         (DRTTestRunner):
4268         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h: Renamed from Tools/DumpRenderTree/chromium/WebPreferences.h.
4269         (WebKit):
4270         (WebTestRunner):
4271         (WebPreferences):
4272         (WebTestRunner::WebPreferences::WebPreferences):
4273         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
4274         (WebTestRunner):
4275         (WebTestDelegate):
4276         (WebTestRunner::WebTestDelegate::localFileToDataURL):
4277         (WebTestRunner::WebTestDelegate::rewriteLayoutTestsURL):
4278         (WebTestRunner::WebTestDelegate::preferences):
4279         (WebTestRunner::WebTestDelegate::applyPreferences):
4280         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
4281         (WebTestRunner::TestRunner::TestRunner):
4282         (WebTestRunner::TestRunner::reset):
4283         (WebTestRunner::TestRunner::setUserStyleSheetEnabled):
4284         (WebTestRunner):
4285         (WebTestRunner::TestRunner::setUserStyleSheetLocation):
4286         (WebTestRunner::TestRunner::setAuthorAndUserStylesEnabled):
4287         (WebTestRunner::TestRunner::setPopupBlockingEnabled):
4288         (WebTestRunner::TestRunner::setJavaScriptCanAccessClipboard):
4289         (WebTestRunner::TestRunner::setXSSAuditorEnabled):
4290         (WebTestRunner::TestRunner::setAllowUniversalAccessFromFileURLs):
4291         (WebTestRunner::TestRunner::setAllowFileAccessFromFileURLs):
4292         (WebTestRunner::TestRunner::overridePreference):
4293         (WebTestRunner::TestRunner::setPluginsEnabled):
4294         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
4295         (TestRunner):
4296         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp: Renamed from Tools/DumpRenderTree/chromium/WebPreferences.cpp.
4297         (WebTestRunner::WebPreferences::reset):
4298         (WebTestRunner):
4299         (WebTestRunner::WebPreferences::applyTo):
4300         * DumpRenderTree/chromium/TestShell.h:
4301         (TestShell::preferences):
4302         (TestShell):
4303         * DumpRenderTree/chromium/WebViewHost.cpp:
4304         (WebViewHost::localFileToDataURL):
4305         (WebViewHost::rewriteLayoutTestsURL):
4306         (WebViewHost::preferences):
4307         (WebViewHost::applyPreferences):
4308         * DumpRenderTree/chromium/WebViewHost.h:
4309         (WebViewHost):
4310
4311 2012-12-04  Zan Dobersek  <zandobersek@gmail.com>